Which Of The Following Species Would You Most Expect To Be Distributed According To A Regular Pattern? A. Termites B. Stingless Bees C. Fish D. Creosote Bushes (2023)

1. Which of the following species would you most expect to be distributed ...

• Which of the following species would you most expect to be distributed according to a regular pattern? a. termites b. stingless bees c. fish d. creosote bushes ...

• The correct answer is option d. creosote bushes.The plants, especially bushes have a more regular or uniform distribution as compared to the animals. The uniform

2. CREOSOTE BUSHES most expect to be distributed according to a ...

• Which of the following species would you most expect to be distributed according to a regular pattern? a. termites b. stingless bees c. fish d. creosote bushes.

• Which of the following species would you most expect to be distributed according to a regular pattern? a. termites b. stingless bees c. fish d. creosote bushes

3. [PDF] Biodiversity and management of the Madrean Archipelago III and 7th C

• Nov 17, 2011 · Topics include climate change in the Sky Island Region, southwestern cienegas, the Northern Jaguar Reserve, amphibian conservation, biodiversity ...

4. maneatersotherod00reid_djvu.txt - Loc

• Fruits of a hundred species are grown in the greatest plenty ; the orange and the Papuan apple, the shaddock *nd lemon ; in short, almost every species of fruit ...

• ' . x>. <$? % < \ A vO o V* . b ^ A' V ' V \ ^ ,\ v *o o N % *b X . % '* - - "> 'r. V , - : v c ^ " ** \ £ ^ ,0o .**% J* * 8 I A J> ^ THE MAN-EATERS OTHER ODD PEOPLE A POPULAR DESCRIPTION OF SINGULAR RACES OF MAN. BY CAPT. MAYNE KEID, AUTHOR OF U THE DESERT HOME," U THE BUSH-BOYS, ' ETC. HGib gUostnCiraa. A NEW EDITION, WITH A MEMOIR BY R. H. STODDARD. NEW YORK: THOMAS K. KNOX & CO., Successors to James Miller, 813 Broadway. 1885. .-ft iS&^ Entered according to act of Congress, in the year 1860, by TICK NOR AND FIELDS, in the Clerk's Office of the District Court of the District of Massachusetts Entered according to act of Congress, in the year 1884, by TnOMAS It. KNOX & CO., in the office of the Librarian of Congress, at Washington. New Yobk, January 1st, 1S60. Messes. Fields, Osgood & Co.:— I accept the terms offered, and hereby concede to you the exclusive right of publ'cation. in the United States, of all my juvenile Tales of Adventure, known as Boys' Novels. MAYNE RE1D.- TROWS PRINTING AND BOOKBINDING COMPANY, MEMOIR OF MATNE REID. No one who has written books for the young during the present century ever had so large a circle of readers as Captain Mavne Reid, or ever was so well fitted by circum- stances to write the books by which he is chiefly known. His life, which was an adventurous one, was ripened with the experience of two Continents, and his temperament, which vras an ardent one, reflected the traits of two races. Irish by birth, he was American in his sympathies with the people of the New World, whose acquaintance he made at an early period, among whom he lived for years, and whose battles he helped to win. He was probably more familiar with the Southern and Western portion of the United States forty years ago than any native-born American of that time. A curious interest attaches to the life of Captain Reid, but it is not of the kind that casual biographers dwell upon. If he had written it himself it would have charmed thousands of readers, who can now merely imagine what it might have been from the glimpses of it which they obtain in his writings. It was not passed in the fierce light o; publicity, but in that simple, silent obscurity which is the lot of most men, and is their hap- piness, if they only knew it. Briefly related, the life of Captain Reid was as follows : He was born in 1818, in the north of Ireland the son of a Presbyterian clergyman, who was a type of the class which Goldsmith has described so freshly in the " Deserted Village," and was highly thought of for his labors among the poor of his neighborhood. An earnest, reverent man, to whom his calling was indeed a sacred one, he designed his son Mayne for the ministry, in the hope, no doubt, that he would be his successor. But nature had some- thing to say about that, as well as his good father He began to study for the ministry, but it was not long before he was drawn in another direction. Always a great reader, his favorite books were descriptions of travel in foreign lands, particularly those which dealt with the scenery, the people, and the resources of America. The spell which, these exercised over his imagination, joined to a love of adventure which was inherent in his temperament, and inherited, perhaps with his race, determined his career. At the age of twenty he closed his theological tomes, and girding up his loins with a stout heart he sailed from the shores of the Old World for the New. Following the spirit in his feet he landed at New Orleans, which was probably a more promising field for a young man of his talents than any Northern city, and was speedily engaged in business. The nature of this business is not stated, further than it was that of a trader ; but whatever it was it obliged this young Irishman to make long journeys into the interior of the country, which was almost a terra in- cognita. Sparsely settled, where settled at all, it was still clothed in primeval verdure — here in the endless reach of savannas, there in the depth of pathless woods, and far away to the North and the West in those monotonous ocean-like levels of land for which the speech of England has no name — the Prairies, Its population was nomadic, not to say barbaric, consisting of tribes of Indians whose hunting grounds from time immemorial the region was ; hunters and trappers, who had turned their backs upon civilization for the free, wild lL'e of nature ; men of doubtful or dangerous antecedents, who had found it con- venient to leave their country for their country's good ; and scattered about hardy pioneer communities from East- ern States, advancing waves of the great sea of emigration which is still drawing the course of empire westward. Travelling in a country like this, and among people like these, Mayne Reid passed five years of his early manhood. He was at home wherever he went, and never more so than when among the Indians of the Red River territory, with whom he spent several months, learning their lan- guage, studying their customs, and enjoying the wild and beautiful scenery of their camping grounds. Indian for the time, lie lived in their lodges, rode with them, hunted with them, and night after night sat by their blazing camp-fires listening to the warlike stories of the braves and the quaint legends of the medicine men. There was that in the blood of Mayne Reid which fitted him to lead this life at this time, and whether he knew it or not it 2 educated his genius as no other life could have done. It familiarized him with a large extent of country in the South and West ; it introduced him to men and manners which existed nowhere else; and it revealed to him the secrets or' Indian life and character. There was another side, however, to Mayne Reid than that we have touched upon, and this at the end of five years, drew him back to the average life of his kind. We find him next in Philadelphia, where he began to con- tribute stories and sketches of travel to the newspapers and magazines. Philadelphia was then the most literate city in the United States, the one in which a clever writer was at once encouraged and rewarded. Frank and warm- hearted, he made many friends there among journalists and authors. One of these friends was Edgar Allan Poe, whom he often visited at his home in Spring Garden, and concerning whom years after, when he was dead, he wrote with loving tenderness. The next episode in the career of Mayne Reid was not what one would expect from a man of letters, though it was just what might have been expected from a man of his temperament and antecedents. It grew out of the time, which was warlike, and it drove him into the army with which the United States speedily crushed the forces of the sister Republic— Mexico. He obtained a commis- sion, and served throughout the war with great bravery and distinction. This stormy episode ended with a severe wound, which he received in storming the heights of Cha- pultepec — a terrible battle which practically ended the war. A second episode of a similar character, but with a more fortunate conclusion, occurred about four years later. It grew out of another war, which, happily for us, was not on our borders, but in the heart of Europe, where the Hun- garian race had risen in insurrection against the hated power o? Austria. Their desperate valor in the face of tremen- dous odds excited the sympathy of the American people, and fired the heart of Captain Mayne Reid, who buckled on his sword once more, and sailed from New York with a body of volunteers to aid the Hungarians in their struggles for independence. They were too late, for hardly had they reached Paris before they learned that all was over : Gorgey had surrendered at Arad, and Hungary was crushed. They were at once dismissed, and Captain Reid betook himself to London. 3 The life of the Mayne rleid In whom we are most in- terested — Mayne Reid, the author — began at this time, when he was in his thirty-first year, and ended only on the day of his death, October 21, 1883. It covered one- third of a century, and was, when compared with that which had preceded it, uneventful, if not devoid of in- cident. There is not much that needs be told — not much, indeed, that can be told — in the life of a man of letters like Captain Mayne Reid. It is written in his books. Mayne Reid was one of the best known authors of his time — differing in this from many authors who are popu- lar without being known — and in the walk of fiction which he discovered for himself he is an acknowledged mas- ter. His reputation did not depend upon the admiration of the millions of young people who read his books, but upon the judgment of mature critics, to whom his delinea- tions of adventurous life were literature of no common order. His reputation as a story-teller was widely recog- nized on the Continent,, where he was accepted as an authority in regard to the customs of the pioneers and the guerilla warfare of the Indian tribes, and was warmly praised for his freshness, his novelty, and his hardy origi- nality. The people of France and Germany delighted in this soldier -writer. " There was not a word in his books which a school-boy could not safely read aloud to his mother and sisters." So says a late English critic, to which another adds, that if he has somewhat gone out of fashion of late years, the more's the pity lor the school-boy of the period. What Deioe is in Robinson Crusoe — realistic idyl of island solitude — that, in his romantic stories of wilder- ness life, is his great scholar, Captain Mayne Reid. R. H. Stoddard. 4 CONTENTS. Page MAN-EATERS OF THE FEEGEE ISLANDS 5 MUNDRUCUS, OR BEHEADERS 30 THE CENTAURS OF THE " GRAN CHACO " . . . . .57 BOSJESMEN, OR BUSHMEN 81 THE AMAZONIAN INDIANS ....... Ill THE WATER-DWELLERS OF MARACAIBO .... 137 THE ESQUIMAUX 161 THE TONGANS. OR FRIENDLY ISLAKD5S3 194 THE TURCOMANS 218 THE OTTOMACS, OR DIRT-EATERS ...... 244 THE COMANCHES, OR PRAIRIE INDIANS 268 THE PEHUENCHES, OR PAMPAS INDIANS ..... 290 THE YAMPARICOS, OR ROOT-DIGGERS .... 309 THE GUARAONS, OR PALM-DWELLERS . « . • . 341 THE LAPLANDERS 359 THE ANDAMANERS, OR MUD-BEDAUBEBS . . . . . 388 THB PATAGONIAN GIANTS • . 411 THE FUEGIAN DWARFS * . • 439 THE MAN-EATERS OF THE FEEGEE ISLANDS. Have I a reader who has not heard of the " King of the Cannibal Islands?" I think I may take it for granted that there is not one in my large circle of boy- readers who has not heard of that royal anthropopha- gist, that " mighty king " who, — "in one hut, Had fifty wives as black as sut y And fifty of a double smut — That King of the Cannibal Islands.*' And yet, strange as it may appear, the old song was no exaggeration — neither as regards the number of his wives, nor any other particular relating to King " Musty- fusty-shang." On the contrary, it presents a picture of the life and habits of his polygamous majesty that is, alas ! too ludicrously like the truth. Though the king of the Cannibal Islands has been long known by reputation, people never had any very definite idea in what quarter of the world his majesty's dominions lay. Being,' as the name implies, an island- kingdom, it was to be looked for of course, in some part of the ocean ; and the Pacific Ocean or Great South 6 THE FEEJJEES, OR Sea was generally regarded as that in which it wa* situated ; but whether it was the Tonga Islands, or the Marquesas, or the Loo-Choos, or the Soo-loos — or some other group, that was entitled to the distinction of being the man-eating community, with the man-eating king at their head — was not very distinctly ascertained up to a recent period. On this head there u uncertainty no longer. Though in several groups of South-Sea Island? the horrible propensity is known to exist, yet the man* eaters, par excellence, the real bona-jide followers of the habit, are the Feegees. Beyond doubt these are th< greatest cannibals in all creation, their islands the true " Cannibal Islands," and their king no other than " Mus- ty-fusty-shang " himself. Alas ! the subject is too serious to jest upon, and it is not without pain that we employ our pen upon it. The truth must needs be told ; and there is no reason why the world should not know how desperately wicked men may become under the influence of a despotism that leaves the masses in the power of the irresponsible few, with no law, either moral or physical, to restrain their unbridled passions. You will find the Feegee Islands, in the Pacific Ocean, in the latitude of 18° south. This parallel passes nearly through the centre of the group. Theil longitude is remarkable : it is the complement of the meridian of Greenwich — the line 180°. Therefore, when it is noon in London, it is midnight among the Feegees. Take the intersection of these two lines, 18° latitude and 180° longitude as a centre ; describe an imaginary circle, with a diameter of 300 miles ; its cir- cumference, with the slight exception of a small outly* MAN-EATERS. 7 ing group, w ill enclose, in a " ring fence," as if were, the whole Feegee archipelago. The group numbers, in all, no fewer than 225 islands and islets, of which between 80 and 90 are at present inhabited — the whole population being not much under 200,000. The estimates of writers diner widely on this point; some state 150,000 — others, more than double this amount. There is reason to believe that 150,000 is too low. Say, then, 200,000 ; since the old adage : " In medias res," is generally true. Only two of the islands are large, — " Viti," and "Vanua." Viti is 90 miles long, by 50 in breadth, and Vanua 100 by 25. Some are what are known as " coral islands ; " others are " volcanic," presenting all varieties of mountain aspect, rugged and sublime. A few of the mountain-peaks attain the elevation of 5,000 feet above sea-level, and every form is known — table- topped, dome-shaped, needle, and conical. In fact, no group in the Pacific affords so many varieties of form and aspect, as are to be observed in the Feegee archi- pelago. In sailing through these islands, the most love- ly landscapes open out before the eye, the most pictu- resque groupings of rocks, ridges, and mountain-peaks, ravines filled with luxuriant vegetation, valleys covered with soft verdure, so divinely fair as to appear the abode of angelic beings. " So beautiful was their as- pect," writes one who visited them, " that I could scarce- ly bring my mind to the realizing sense of the well- known fa?t, that they were the abode of a savage, ferocious, and treacherous race of cannibals." Such, Bias! is the fact, well known, as the writer observes. Perhaps to no part of the world has Nature beeD 8 TEE FEEGEES, OR more bountiful than to the Feegee Islands. She has here poured out her favors in very profusion ; and the cornucopia might be regarded as an emblem of the land. The richest products of a tropic vegetation flour- ish in an abundance elsewhere unknown, and the growth of valuable articles of food is almost spon- taneous. Many kinds are really of spontaneous pro- duction ; and those under cultivation are almost end- less in numbers and variety. Yams grow to the length of six feet, weighing one hundred pounds each ! and several varieties are cultivated. The sweet potato reaches the weight of five or six pounds, and the " taro " {Arum esculentum) also produces a root of enormous size, which forms the staple article of the Feegeean's food. Still another great tuber, weighing twenty or thirty pounds, and used as a liquorice, is the produce of the " massawe," or ti-tree (draccena termi- nalis) ; and the root of the piper methisticum often at- tains the weight of one hundred and forty pounds! This last is possessed of highly narcotic properties ; and is the njaterial universally used in the distillation, or rather brewing, of the native drink called " yaqona " — the " kava " of the South-Sea voyagers. Bread-fruit grows in abundance : there being no less than nine va- rieties of this celebrated tree upon the different islands of the group, each producing a distinct kind of fruit ; and what is equally remarkable, of the musacece — the plantain and banana — there are in the Feegee isles thirty different kinds, either of spontaneous growth, or cultivated ! All these are well distinguished from one another, and bear distinct appellations. Three kinda c r cocoa-palm add to the extraordinary variety of vege MAN-EATERS. 9 table food, as well as to the picturesquenes3 of the scenery ; but there is no lack of lovely forms in the vegetation, where the beautiful ti-tree grows, — where the fern and the screw-pines flourish, — where plan- tains and bananas unfold their broad bright leaves to the sun ; where arums spread their huge fronds min- gling with the thick succulent blades of the bromelia, and where pawpaws, shaddocks, orange and lime-trees exhibit every hue of foliage, from deep green to the most brilliant golden. Fruits of a hundred species are grown in the greatest plenty ; the orange and the Papuan apple, the shaddock *nd lemon ; in short, almost every species of fruit that will flourish in a tropical clime. In addition, many in- digenous and valuable kinds, both of roots and fruits, are peculiar to the Feegee group, yet unknown and unculti- vated in any other part of the world. Even the very cloth of the country- — and a beautiful fabric it makes — is the product of an indigenous tree, the " malo " or paper-mulberry (Bromonetia papyri/era), the "tapa" of voyagers. Not only the material for dresses, but the tapestry for the adornment of their temples, the curtains and hangings of their houses, are all obtained from this valuable tree. We have not space for a more detailed account of the productions of these isles. It would fill a volume to describe with any degree of minuteness the various genera and species of its plants alone. Enough has been said to show how bountiful, or rather how prodi- gal, nature has been to the islands of the Feegeean Archipelago. Of the animal kingdom the^e is not much to be said 10 THE FEEGEES, OR Of quadrupeds there is the usual paucity of species thai is noticed everywhere throughout the Polynesian islands. Dogs and pigs are kept ; the latter in considerable num- bers, as the tiesh forms an important article of food ; but they are not indigenous to the Feegee group, though the period of their introduction is unknown. Two or three small rodents are the only quadrupeds yet known to be true natives of the soil. Reptiles are alike scarce in species, — though the turtle is common upon the coasts, and its fishery forms the regular occupation of a par- ticular class of the inhabitants. The species of birds are more numerous, and there are parrots, peculiar to the islands, of rich and beautiful plumage. But we are not allowed to dwell upon these subjects. Interesting as may be the zoology and botany of the Feegeean Archipelago, both sink into insignificance when brought into comparison with its ethnology, — the natural history of its human inhabitants ; — a subject of deep, but alas ! of a terribly painful interest. By inquiry into the condition and character of these people, we shall see how little they have deserved the fa vers which nature has so bounteously bestowed upon them. In the portrait of the Feegeean you will expect some- thing frightfully hideous, — knowing, as you already do, that he is an eater of human flesh, — a man of' gigantic stature, swarthy skin, bloodshot eyes, gaunt, bony jaws, and terrific aspect. You will expect this man to be described as being nakea\ — or only with the skin of a wild beast upon his shoulders, — building no house, manufacturing no household or other utensils, and armed vri*h a huge knotted club, which he is ever ready to use: — a m<\n who dwells in a cavern, sleeps indifferently w MAN-EATERS. 1 1 the open air or under the shelter of a bush ; in short, a true savage. That is the sort of creature you expect me to describe, and I confess that just such a physical aspect —just such a condition of personal hideousness — would be exactly in keeping with the moral deformity of the Feegeean. You would furthermore expect this savage to be almost devoid of intellectual power, — altogether wanting in moral sen=;e, — without knowledge of right and wrong, — without knowledge of any kind, — with- out ideas. It seems but natural you should look for such Characteristics in a cannibal. The portrait I am about to paint will disappoint you. I do hot regret it, since it enables me to bring forward another testimony that man in his original nature is not a being of such desperate wickedness. That simple and primitive state, which men glibly call savage, is not the condition favorable to cannibalism. I know that it is to such people that the habit is usually ascribed, but quite erroneously. The Andaman islander has been blamed with it simply becauses he chances to go naked, and looks, as he is, hungry and emaciated. The charge is proved false. The Bushman of South Africa has en- joyed a similar reputation. It also turns out to be a libel. The Carib long lived under the imputation, sim- ply because he presented a fierce front to the Spanish tyrant, who would have enslaved him ; and we have heard the same stigma cast upon a dozen other tribes, the lowest savages being usually selected ; in other words, those whose condition appeared the most wretch- ed. In such cases the accusation has ever been found, upon investigation, to be erroneous. In the c^ost primitive state in which man appear* 12 THE FEEGEES, OR npon the earth, he is either without social organization altogether, or if any do exist, it is either patriarchal of republican. Neither of these conditions is favorable to the development of vice, — much less the most horrible of all vices. It will not do to quote the character of the Bushman, or certain other of the low tribes, to refute this state- ment. These are not men in their primitive state ascending upward, but a condition altogether the reverse. They are the decaying remnants of some corrupt civili- zation, sinking back into the dust out of which they were created. No — and I am happy to say it — man, as he origi- nally came from the hands of the Creator, has no such horrid propensity as cannibalism. In his primitive state he has never been known to practise it, — except when the motives have been such as have equally tempted men professing the highest civilization, — but this cannot be considered cannibalism. Where that exists in its true unmitigated form, — and unhappily it does so, — the early stages of social organization must have been passed ; the republican and patriarchal forms must both have given place to the absolute and monarchical. This condition of things is absolutely necessary, before man can obtain sufficient power to prey upon his fellow-man to the extent of eating him. There can be no " canni- bal " without a " king." So far from the Feegeean cannibals being savages, at cording to the ordinary acceptation of the term, they are in reality the very reverse. If we adhere to the usual meaning of the word civilization, understanding by it a people possessing an intelligent knowledge of arts, living MAN-EATERS. ] 3 In well-built houses, fabricating fine goods, tilling tneh lands in a scientific and successful manner, practising the little politenesses and accomplishments of social life, — if these be the criteria of civilization, then it is no more than the truth to say that the standard possessed by the Feegee islanders is incomparably above that of the lower orders of most European nations. It is startling to reflect — startling as sad — that a people possessed of such intellectual power, and who have ever exercised it to a wonderful extent, in arts, manufactures, and even in the accomplishing of their own persons, should at the same time exhibit moral traits of such an opposite character. An atrocious cruelty, — an instinct for oppression, brutal and ferocious, — a heart pitiless as that of the fiend himself, — a hand ever ready to strike the murderous blow, even though the victim be a brother, — lips that lie in every word they speak, — a tongue ever bent on barbaric boasting, — a bosom that beats only with sentiments of treachery and abject cow- ardice, — these are the revolting characteristics of the Feegeean. Dark as is his skin, his soul is many shades darker. It is time, however, to descend to a more particular delineation of this man-eating monster ; and first, we shall give a description of his personal appearance. The Feegeeans are above the average height of Eu- ropeans or white men : men of six feet are common among them, though few reach the height of six feet six. Corpulent persons are not common, though large and muscular men abound. Their figure corresponds more nearly to that of the white man than any other race known. The p *opertions of their limbs resemble thosf U THE FEEGEES, OR f northern Europeans, though some are narrower across the loins. Their chests are broad and sinewy, and theif Btout limbs and short, well-set necks are conspicuous characters. The outline of the face is a good oval ; the mouth large, with white teeth regularly arranged — ah ! those horrid teeth ! — the nose is well-shaped, with full nostrils ; yet quite distinct, as are the lips also, from the type of the African negro. Indeed, with the exception of their color, they bear very little resemblance to the negro, — that is, the thick-lipped, flat-nosed negro of our fancy ; for there are negro tribes in Africa whose fea- tures are as fine as those ot the Feegeeans, or even as our own. In color of skin the Feegeean is nearly, if not quite, as dark as the negro ; but it may be remarked that there are different shades, as there are also among pure Ethiopians. In the Feegee group there are many men of mulatto color, but these are not of the original Fee- gee stock. They are either a mixed offspring with the Tonga islander, or pure-bred Tonga islanders themselves who for the past two hundred years have been insinuat- ing themselves into the social compact of the Feegee- ans. These light-colored people are mostly found on the eastern or windward side of the Feegee group, — that is, the side towards Tonga itself, — and the trade-winds will account for their immigration, which was at first purely accidental. They at present play a conspicuous part in ♦he affairs of the Feegeeans, being in favor with the lings and great chiefs, partly on account of their being Letter sailors than the native Feegeeans, and partly on account of other services which these tyrants require them to perform. In some arts the Tongans are superior to the Feegeeans, but not in all. In pottery, wood-carving MAN-EATERS. 13 making of mats or baskets, and the manufacture of the tapa cloth, the Feegeeans stand unrivalled over all the Pacific Ocean. We need say no more of the Tongans here ; they are elsewhere described. Those dwelling in Feegee are not all fixed there for life. Some are so, and these are called Tonga-Feegenans ; the others are only visitors, giving their services temporarily to th.3 Feegeean chiefs, or occupied in ship-building, — in constructing those great war canoes that have been the astonishment of South-Sea voyagers, and which Feegee sends forth from her dock-yards in the greatest perfection. These, when finished by the Tongan strangers, are used to carry them back to their own islands, that lie about three hundred miles to the windward (southeast). But to continue the portrait of the Feegeean. We have touched almost every part of it except the hair ; but this requires a most elaborate limning, such as the owner himself gives it. In its natural state the head of the Feegeean is covered by a mass of black hair long, frizzled, and bushy, sometimes encroaching on the forehead, and joined by whiskers to a thick, round, or pointed beard, to which mustaches are often added. Black is, of course, the natural colcr of the hair, but tt is not always worn of this hue. Other colors are thought more becoming ; and the hair, both of the men and women, is dyed in a variety of ways, lime burning it to a redish or whity-brown shade. A turmeric-yel- low, or even a vermilion-red are not uncommon colors but all these keep varying, according to the change of fashions at court ! Commodore Wilkes, who has given a good deal of 16 THE FEEGEES, Oft his time to an exploration of the Feegee Islands, states that the Feegee hair, in its natural condition, is straight, and not "frizzled," as described above — he says that the frizzling is the work of the barber; but the Com- modore is altogether mistaken in this idea. Thousands of Feegeans, whose hair was never touched by a bar- ber, nor dressed even by themselves, exhibit this pecu- liarity. We regret to add that this is only one of a thousand erroneous statements which the Commodore has made during his gigantic exploration. He may have been excellent at his own speciality of making soundings and laying down charts ; but on all matters pertaining to natural history or ethnology, the worthy Commodore appears to have been purblind, and, indeed, his extensive staff of naturalists of every kind have produced far less than might have been expected from such excellent opportunities as they enjoyed. The ob- servations of the Commodore will not stand the test of time, and cannot be depended upon as safe guides, ex- cepting in those cases where he was an actual eye- witness. About his truthful intentions there can be no doubt whatever. Of one very peculiar performance among the Fee- gees he appears to have had actual demonstration, and as he has described this with sufficient minute- ness, we shall copy his account ; though, after what we have said, we should apologize largely for the lib- erty. The performance referred to is that of " barber- izing" a barbarian monarch, and may be taken as a proof of high civilization among the Feegees. It will be seen that, with the exception of the tabooed fingers, there is not much difference between a barber of Bond MAN-EATERS. 17 Street and an artist of like calling in the Cannibal Islands. "The chiefs in particular," writes Commodore Wilkes^ u P av g r eat attention to the dressing of their heads, and for this purpose all of them have barbers, whose sole occupation is the care of their masters' heads. These barbers are called a-vu-ni-ulu. They are attached to the household of the chiefs in numbers of from two to a dozen. The duty is held to be of so sacred a nature, that their hands are tabooed from all other employment, and they are not even permitted to feed themselves. To dress the head of a chief requires several hours. The hair is made to spread out from the head, on every side, to a distance that is often eight inches. The beard, which is also carefully nursed, often reachea the breast, and when a Feegeean has these important parts of his person well dressed, he exhibits a degree of conceit that is not a little amusing. "In the process of dressing the hair it is well anointed with oil, mixed with a carbonaceous black, until it is completely saturated. The barber then takes the hair- pin, which is a long and slender rod, made of tortoise- shell or bone, and proceeds to twitch almost every sepa- rate hair. This causes it to frizzle and stand erect The bush of hair is then trimmed smooth by singeing it, until it has the appearance of an immense wig. When this has been finished, a piece of tapa, so fine as to resemble tissue-paper, is wound in light folds around it, to protect the hair from the dew or dust. This covering, which has the look of a turban, is called sola, and none but the chiefs are allowed to wear it guy attempt to assume this head-dress by a kaisi t o 18 THE FEEGEES, OR common person, would be immediately punished with death. Tie sala, when taken proper care of, will last three wee'is or a month, and the hair is not dressed except when it is removed; but the high chiefs and dandies seldom allow a day to pass without changing the sala and having the hair put in order." With this account, we conclude our description of the Feegeean's person. His costume is of the simplest kind, and easily described. With the men it is merely a strip of " tapa " or " malo " cloth passed several times round the waist, and the ends left to hang down in front. The length of the hanging ends determines the rank of the wearer, and only in the case of kings or great chiefs are they allowed to touch the ground. A turban of the finest tapa cloth among the great mop of hair is another badge of rank, worn only by kings and chiefs; and this head-dress, which adds greatly to the dignified appearance of the wearer, is not always coiffed in the same fashion, but each chief adapts it to his owr, or the prevailing taste of the court. The dress of the women is a mere waist-belt, with a fringe from six to ten inches in length. It is worn longer after they have become wives, sometimes reaching near the knee, and forming a very picturesque garment. It is called the "liku," and many of them are manufactured with sur- prising skill and neatness, the material being obtained from various climbing plants of the forest. Under the tt liku r the women are tattooed, and there only. Their men, on the contrary, do not undergo the tattoo ; but on grand occasions paint their faces and bodies in the most fanciful colors and patterns. The kings and some chiefs suspend from their neck* MAN-EATERS. 19 •hell ornaments — often as large as a dining-plate — that hang down upon the breast. Some, instead of thia> wear a necklace of whales' teeth, carved to resemble claws, and bearing a very close resemblance to the necklaces of the Prairie Indians, made of the claws of the grizzly bear. Another kind of necklace — per- haps more appropriate to the Feegee — is a string of human teeth ; and this kind is not unfrequently worn by these ferocious dandies. It must not be supposed that the scantiness of the Feegeean costume arises from poverty or stinginess on the part of the wearer. Nothing of the kind. It is simply because such is the fashion of the time. Were it otherwise, he could easily supply the materials, but he does not wish it otherwise. His climate is an eter- nal summer, and he has no need to encumber his body with extraneous clothing. With the exception of the turban upon his head, his king is as naked as himself. You may suppose that the Feegeans have but little notions of modesty ; but, strange as it may appear, this is in reality not one of their failings. They regard the " malo " and " liku " as the most modest of garments ; and a man or woman seen in the streets without these scanty coverings would be in danger of being clubbed to death ! It must be acknowledged that they are not altogether depraved — for in this respect they present the most astounding anomaly. Certain virtues are ascribed to them, and as I have painted only the dark side of their character, it is but fair to give tr > other. Indeed, it is a pleasure to do this — though tliere is not enough of the favorable to make any great alteration in the pio 20 THE FEEGEES, OR hire. The whole character is so well described by on6 of the most acute observers who ha° } r et visited the South Seas — the Wesleyan missionary Williams — that we borrow the description. "The aspect of the Feegeean," says Mr. Williams, "with reference to his mental character, so far from supporting the decision which would thrust him almost out of mankind, presents many points of great interest, showing that, if an ordinary amount of attention were bestowed on him, he would take no mean rank in the human family, to which, hitherto, he has been a dis- grace. Dull, barren stupidity forms no part of his char- acter. His feelings are acute, but not lasting ; his emotions easily roused, but transient ; he can love tru- ly, and hate deeply ; he can sympathize with thorough sincerity, and feign with consummate skill ; his fidelity and loyalty are strong and enduring, while his revenge never dies, but waits to avail itself of circumstances, or of the blackest treachery, to accomplish its purj ose. His senses are keen, and so well employed, that he often excels the white man in ordinary things. Tact has been called i ready cash/ and of this the native of Feegee has a full share, enabling him to surmount at once many difficulties, and accomplish many tasks, that would have 'fixed' an Englishman. Tools, cord, or packing materials, he finds directly, where the wKte man would be at a loss for either ; and nature seen * to him but a general store for his use, where the article he wants is always within reach. " In social diplomacy the Feegeean is very cautious and clever. That he ever paid a visit merely en passant, is hard to be believed. If no request leaves his lips, he MAN-EATERS. 2 1 has brought the desire, and only waits for a good chance to present it now, or prepare the way for its favorable ^reception at some other time. His face and voice are all pleasantness ; and he has the rare skill of rinding out just the subject on which you most like to talk, or sees at once whether you desire silence. Rarely will he fail to read your countenance ; and the case must be urgent indeed which obliges him to ask a favor when he sees a frowu The more important he feels his business the more earnestly he protests that he has none at all ; and the subject uppermost in his thoughts comes last to his lips, or is not even named; for he will make a second, or even a third visit, rather than risk a failure through precipitancy. He seems to read other men by intuition, especially where selfishness or lust are prominent traits. If it serves his purpose, he will study difficult and pe- culiar characters, reserving the results for future use; if afterwards he wish to please them, he will know how, and if to annoy them, it will be done most exactly. "His sense of hearing is acute, and by a stroke of his nail he judges the ripeness of fruits, or soundness of various substances." From what source the Feegeean has sprung is purely a matter of conjecture. He has no history, — not even a tradition of when his ancestors first peopled the Archi- pelago in which we now find him. Of his race we have not a much clearer knowledge. Speculation places Irim in the same family as the " Papuan Negro," and he has some points of resemblance to this race, in the color and frizzled hair; but there is as much difference between the wretched native of West Australia and the finely-devel- oped Feegeean as there is between the stunted Laplander 22 THE FEEGEES, OR and the stalwart Norwegian ; nor is the coarse rough skin ef the true Papuan to be recognized in the smooth, glo?*y epidermis of the Feegee Islander. This, however^ ma) be the result of better living ; and certainly among the mountain-tribes of the Feegees, who lead lives of greater privation and hardship, the approach to the Pa- puan appearance is observable. It is hardly necessary to add that the Feegeean is of a race quite distinct from that known as the Polynesian or South-Sea Islander. This last is different not only in form, complexion, and language, but also in many important mental character- istics. It is to this race the Tongans belong, and its pecularities will be sketched in treating of that people. Were we to enter upon a minute description of the manners and customs of the Fegees, — of their mode of house and canoe building, — of their arts and manu- factures, for they possess both, — of their implements of agriculture and domestic use, — of their weapons of war, — their ceremonies of religion and court etiquette, — our task would require more space than is here allotted to us : it would in fact be as much as to describe the complete social economy of a civilized nation ; and a whole volume would scarce suffice to contain such a de- scription. In a sketch like the present, the account of these people requires to be given in the most condensed and synoptical form, and only those points can be touched upor that may appear of the greatest interest. It must be remembered that the civilization of the FeegeeH — of course, I allude to their proficiency in the industrial arts — is entirely an indigenous growth. They have borrowed ideas from the Tongans, — as the Tongans have also from them, — but both are native MAN-EATERS. 23 productions of the South Sea, and not derived from any of the so-called great centres of civilization. Such as hav r e sprung from these' sources are of modern date, and make but a small feature in the panorama of Feegeean life. The houses they build are substantial, and suitable to their necessities. We cannot stay to note the archi- tecture minutely. The private dwellings are usually about twenty-five feet long by fifteen in breadth, the interior forming one room, but with a sort of elevated divan at the end, sometimes screened with beautiful " tapa " curtains, and serving as the dormitory. The ground-plan of the house is that of an oblong square, — or, to speak more properly, a parallelogram. The walls are constructed of timber, — being straight posts of cocoa-palm, tree-fern, bamboo, or bread-fruit, — the spaces between closely warped or otherwise filled in with reeds of cane or calamus. The thatch is of the leaves of the wild or cultivated sugar-cane, — sometimes of a pandanus, — thickly laid on, especially near the eaves, where it is carefully cropped, exposing an edge of from one to two feet in thickness. The roof has four faces, — that is, it is a u hip roof." It is made with a very steep pitch, and comes down low, projecting far over the heads of the upright timbers. This gives a sort of shaded veranda all around the house, and throws the' rain quite clear of the walls. The ridge-pole is a pecu- liar feature ; it is fastened to the ridge of the thatch by strong twisted ropes, that give it an ornamental appear- and ; and its carved ends project at both gables, or rather, over the " hip roofs," to the length of a foot, or more ; it is further ornamented by white shells, these of the cyprea ovula being most used for the purpose. The 24 THE FEE GEES, OR Feegee house presents altogether a picturesque and noi inelegant appearance. The worst feature is the low door. There are usually two of them, neither in each house being over three feet in height. The Feegee assigns no reason why his door is made so low ; but as he is fre- quently in expectation of a visitor, with a murderous bludgeon in his grasp, it is possible this may have some- thing to do with his making the entrance so difficult. The houses of the chiefs, and the great council-house, or temple, — called the " Bure," — are built precisely in the same style ; only that both are larger, and the doors, walls, and ridge-poles more elaborately ornamented. The fashionable style of decoration is a plaiting of cocoa fibre, or " sinnet," which is worked and woven around the posts in regular figures of " relievo." The house described is not universal throughout all the group. There are many " orders " of architecture, and that prevailing in the Windward Islands is different from the style of the Leeward, and altogether of a better kind. Different districts have different forms. In one you may see a village looking like an assemblage of wicker baskets, while in another you might fancy it a collection of rustic arbors. A third seems a collection oblong hayricks, with holes in their sides ; while, in a fourth these ricks are conical. It will be seen that, with this variety in house- building, it would be a tedious task to illustrate the complete architecture of Feegeeans. Even Master Ruskin himself would surrender it up in despair. Equally tedious would it be to describe the various implements or utensils which a Feegee house contains, The furniture is simple enough. There are neither MAN-EATERS. 25 chairs, tables, nor bedsteads. The bed is a beautiful mat spread on the dais, or divan ; and in the houses of the rich the floors are covered with a similar carpet. These mats are of the finest texture, far superior to those made elsewhere. The materials used are the Hibiscus tiliaceus, Pandanus odoi'atissimus, and a spe- cies of rush. They are in great abundance in every house, — even the poorest person having his mat to sit or lie upon ; and it is they that serve for the broad- Bpreading sails of the gigantic canoes. In addition to the mats, plenty of tapa-cloth may be seen, and baskets of every shape and size, — the wicker being obtained from the rattan (Jlagellaria), and other sources. One piece of furniture deserves especial mention, — this is the pillow upon which the Feegee lord^ lays his head when he goes to sleep. It presents but little claim tc the appellation of a downy pillow ; since it is a mere cylinder of hard polished wood, with short arched pedes- tals to it, to keep it firmly in its place. Its object is to keep the great frizzled mop from being tossed or dis- arranged, during the hours of repose ; and Feegeean vanity enables the owner of the mop to endure this flinty bolster with the most uncomplaining equanimity. If he were possessed of the slightest spark of conscience; even this would be soft, compared with any pillow upon which he might rest his guilty head. In addition to the baskets, other vessels meet the eye. These are of pottery, as varied in shape and size as they are in kind. There are pots and pans, bowls, dishes, cups and saucers, jars and bottles, — many of them of rare and curious designs, — some red, some ornamented with a glaze obtained from the gum of the 26 THE FEEGEES, OR kauri pine, — for this tree is also an indigenous produo* tion of the Feegee Islands. Though no potter's wheel is known to the Feegees, the proportions of their vessels are as just and true, and their polish as complete, as if Stafford had produced them. There are cooking pots to be seen of immense size. These are jars formed with mouths wide enough to admit the largest joint. I dare not mention the kind of joint that is frequently cooked in those great caldrons. Ugh ! the horrid pots ! Their implements are equally varied and numerous, — some for manufacturing purposes, and others for agriculture. The latter are of the simplest kind. The Feegee plough is merely a pointed stick inserted deeply into the ground, and kept moving about till a lump of the soil is broken upward. This is crushed into mould, first by a light club, and afterwards pulverized with the fingers. The process is slow, but fast enough for the Feegeean, whose farm is only a garden. He requires no plough, neither bullocks nor horses. With taro-roots and sweet potatoes that weigh ten pounds each, yams and yaqonas over one hundred, and plantains producing bunches of a hundred and fifty fruits to the single head, why need he trouble himself by breaking up more sur- face ? His single acre yields him as much vegetable wealth as fifty would to an English firmer ! It is not to be supposed that he has it all to himself; no, nor half of it either ; nor yet the fifth part of it At least four fifths of his sweat has to be expended in tax or tithe ; and this brings us to the form of his gov- ernment. We shall not dwell long upon this subject Suffice it to say that the great body of the people are in a condition of abject serfdom, — worse than slavery MAN-EATERS. 27 itself. They own nothing that they can call their own, •—not their wives, — not their daughters, — not even their lives ! All these may be taken from them at any hour. There is no law against despoiling them, — no check upon the will and pleasure of their chiefs or superiors ; and, as these constitute a numerous body, the poor canaille have no end of ruffian despoilers. It is an every-day act for a chief to rob, or club to death, one of the common people ! and no unfrequent occurrence to be himself clubbed to death by his superior, the king ! Of these kings there are eight in Feegee, — not one, as the old song has it ; but the words of the ballad will apply to each of them with sufficient appropriateness. Any one of them will answer to the character of " Musty- fusty-shang ! " These kings have their residences on various island?, and the different parts of the group are distributed some- what irregularly under their rule. Some islands, or parts of islands, are only tributary to them ; others con- nected by a sort of deferential alliance ; and there are communities quite independent, and living under the arbitrary sway of their own chieftains. The kings are not all of equal power or importance ; but in this respect there have been many changes, even during the Fee- geean historical period, — which extends back only to the beginning of the present century. Sometimes one is th 3 most influential, sometimes another; and in most cases the pre-eminence is obtained by him who possesses the greatest amount of truculence and treachery. He who is most successful in murdering his rivals, and rid- ding himself of opposition, by the simple application of the club, usually succeeds in becoming for the time head 28 THE FEEGEES, OK u king of the Cannibal Islands." I do not mean that he reigns over the whole Archipelago. No king has yet succeeded in uniting all the islands under one govern- ment He only gets so far as to be feared everywhere, and to have tributary presents, and all manner of debas- ing compliments offered to him. These kings have all their courts and court etiquette, just as their " royal brothers " elsewhere ; and the ceremonials observed are quite as complicated and degrading to the dignity of man. The punishment for neglecting their observance is rather more severe in Feegee than elsewhere. For a decided or wilful non-compliance, the skull of the de- linquent is frequently crushed in by the club of his majesty himself, — even in presence of a full " drawing- room." Lesser or accidental mistakes, or even the ex- hibition of an ungraceful gaucherie, are punished by the loss of a finger : the consequence of which is, that in Feegee there are many fingers missing ! Indeed, a com- plete set is rather the exception than the rule. If a king or great chief should chance to miss his foot and slip down, it is the true ton for all those who are near or around him to fall likewise, — the crowd coming down, literally like a " thousand of bricks ! " I might detail a thousand customs to show how far the dignity of the human form is debased and disgraced upon Feegee soil ; but the subject could be well illus- trated nearer home. Flunkey ism is a fashion unfortu- nately not confined to the Feegeean archipelago ; and though the forms in which it exhibits itself there may be different, (he sentiment is still the same. It must evei appear where men are politically unequal, — wherever there is a ^lass possessed of hereditary privileges. MAN-EATERS. 29 I come to the last, — the darkest feature in the Fee- geean character, — the horrid crime and custom of can- nibalism. I could paint a picture, and fill up the details with the testimony of scores of eyewitnesses, — a pic- ture that would cause your heart to weep. It is too horrid to be given here. My pen declines the office; and. therefore, I must leave the painful story untold MDNDRUCUS, OR BEHEADIiRS. In our general sketch of the Amazonian Indians it was stated that there were some few tribes who differed in certain customs from all the rest, and who might even be regarded as odd among the odd. One of these tribes is the Mundrucu, which, from its numbers and warlike strength, almost deserves to be styled a nation. It is, at all events, a powerful confederacy, of different tribes, linked together in one common nationality, and including in their league other Indians which the Mundrucus themselves first conquered, and afterwards associated with themselves on terms of equality ; in other words, " annexed " them. The same sort of annexation or alli- ance is common among the tribes of North America ; as in the case of the powerful Coma tche nation, who extend their protecting alliance over the Wacoes, Washites, and Cayguaas or Kioways. The Mahue is the principal tribe that is patronized in this fashion by the Mundrucus, and the two together number at least 20,000 souls. Before the days of the Portuguese slave-hunting, the Mundrucus occupied the south bank of the Amazon, from the mouth of the Tapajos to that of the Madeira. Thia MUNDRUCUS. 31 mfanious traffic had the effect of clearing the banks of the great river of its native inhabitants, — except sucb of them as chose to submit to slavery, or become neo phytes, by adopting the monkish faith. Neither of these courses appeared pleasing in the eyes of the Mundrucus, and they adopted the only alternative that was likely to insure their independence, — by withdrawing from the dangerous proximity of the sanguinary slave-trade. Tins retreat of the Mundrucus, however, was by no means an ignominious flight. The withdrawal was vol- untary on their part, and not compulsory, as was the case with weaker tribes. From the earliest times they had presented a firm front to the Portuguese encroach- ments, and the latter were even forced into a sort of nefarious alliance with them. The leaving the Amazon on the part of the Mundrucus was rather the result of a negotiation, by which they conceded their territory — between the mouths of the Tapajojs and Madeira — to the Brazilian government ; and to this hour they are not exactly unfriendly to Brazilian whites, though to the mulattoes and negroes, who constitute a large proportion of the Brazilian population, the Mundrucu knows no other feeling than that of a deadly hostility. The origin of their hatred of the Brazilian blacks is to be found in sl revolt which occurred in the provinces of the Lower Amazon (at Para) in 1835. It was a caste revolution against whites, but more especially against European Portuguese. In this affair tht Mundrucus were em- ployed against the darker-skinned rebels — the Cabanas, as they were called — and did great service in putting down the rebellion. Hence they retain a lingering spark of friendship for their ci-devant white allies ; oi 32 MUNDRUCUS, perhaps if would be more correct to say they do rot actually hate them, but carry on a little commerce with their traders. For all that, they occasionally cut the throats of a few of the latter, — especially those who do not come to deal directly with them, but who pass through their country in going from the Amazon to the diamond mines of Brazil. These last are called Mon- $aos, and their business is to carry supplies from the towns on the Amazon (Santarem and Para) to the mi- ners of gold and washers of diamonds in the district of Matto Grosso, of which Cuiaba is the capital. Their route is by water and " portage " up the Tapajos river, and through the territory of the dreaded Mundrucus, — requiring a journey of six months, as perilous and toil- some as it is tedious. The present residence of the Mundrucus is between the Tapajos and Madeira, as formerly, but far up on both rivers. On the Tapajos, above what are known as the " Caxoeiras," or Cataracts, their villages are found. There they dwell, free from all molestation en the part of the whites ; their borders extending widely around them, and limited only by contact with those of other warlike tribes like themselves, who are theur deadly enemies. Among these last are the Muras, who dwell at the mouths of the Madeira and Rio Negro. The Mundrucus build the malocca, elsewhere de- scribed ; only in their case it is not used as a dwelling, but rather as a grand arsenal, a council-chamber, a ball- room, and, if need be, a fortress. When fearing an attack, all sleep in it " under arms." It is a structure of large size and great strength, usually rendered more unassailable by being*" chinked " and plastered with OR BEHEADERS. 33 tiay. It is in this building that are deposited those hor- rid trophies which have given to the Mundi ticus their terrible title of decapitadores, or " beheaders." The title and its origin shall be presently explained. Around the great malocca the huts are placed, form- ing a village, and in these the people ordinarily dwell. The Mundrucus are not without ample means of sub- sistence. Like most other Amazonian tribes, they cul- tivate a little manioc, plantains, and even maize ; and they know how to prepare the farinha meal, and, unfortunately, also the detestable chicha, the universal beverage of the South American aborigines. They have their vessels of calabash — both of the vegetable and arborescent kinds — and a full set of implements and utensils for the field and kitchen. Their war weapons are those common to other Amazonian tribes, and they sometimes also carry the spear. They have canoes of hollow trees ; and, of course, fishing and hunting are the employments of the men, — the women, as almost every- where e^lse among Indians, doing the drudgery, — the tilling and reaping, the " hewing of wood and the draw- ing of water," the making the household utensils and using them, — all such offices being beneath the dignity of the " lordly," or rather lazy savage. I have said that they carry on a commercial inter- course with the white traders. It is not of much magni- tude, and their exports consist altogether of the native and spontaneous productions of the soil, sarsaparilla being one of the chief articles*. They gather this (the women and children do) during six months of the year. The other six months no industry is followed, — as this period is spent in hostile^ excursions against the neigh 34 MUNDRUCU5, boring tribes. Their imports consist of iron tools and pieces for weapons ; but they more especially barter the product of their labor for ornamental gewgaws, — such as savages universally admire and desire. Their Bars* parilla is good, and much sought for in the medical market. Every one is acquainted with the nature and charac- ter of this valuable medicinal root, the appearance of which must also be known to almost everybody, — since it is so very common for our druggists to display the bundles of it in their shop windows. Perhaps every one is not acquainted with the fact, that the sarsaparilla root is the product of a great many different species of plants most of them of the genus Simlax, but not a few belong ino- to plants of other genera, as those of Carex and Her- reria the roots of which are also sold as sarsaparilla. The species of simlax are widely distributed throughout the whole torrid zone, in Asia, Africa, and America, and some kinds are found growing many degrees outside the tropics, — as is the case in Virginia and the vaiiev of the Mississippi, and also on the other side of the Pacific on the great continent-island of Australia. The best sarsaparilla, however, is that which is pr* duced in tropical countries, and. especially in moist sue* tions, where the atmosphere is at once hot and humid Tt requires these conditions to concentrate the virtue of its sap, and render it more active. It would be idle to give a list of the different specie? of simlax that furnish the sarsaparilla root of the phar ma:opeia. There is an almost endless number of thein, and they are equally varied in respect to excellence of quality ; some kinds are in reality almost worthless, and OR BEHEADERS. 35 for this reason, in using it as a medicine, great, care should bf taken in the selection of the species. Like all other articles, either of food or medicine, the valu- able kinds are the scarcest ; the reason in this case being that the best sarsaparilla is found in situations * not only difficult of access, but where the gathering of ite root is attended with considerable danger, from the unhealthy nature of the climate and the hostility of the savages in whose territory it grows. As to the quan- tity that may be obtained, there is no limit, on the score of any scarcity of the plant itself, since it is found throughout all the countries of tropical America plenteously distributed both in species and individual plants. Such quantities of it grow along the banks of some South-American rivers, that the Indians have a belief that those streams known as black waters — such as the Rio Negro and others — derive their peculiar color from the roots of this plant. This, however, is an erroneous supposition, as there are many of the white-water rivers that run through regions abundant- ly supplied with the sarsaparilla root. The black water, therefore, must arise from some other cause, as yet un« known. As observed, the sarsaparilla of the Mundrucu eoun try is of the very best quality. It i? the Simlax pa pyracea of Soiret, and is known in commerce as th« " Lisbon," or " Brazilian." It is a climbing plant, or under-shrub, the stem of which is flattened and angu- lar, with rows of prickles standing along the prominent edges. Its leaves are of an oval acuminated shape, and marked with longitudinal nerves. It shoots up without any support, to a height of fifteen or twentjr 36 MUNDRUCUS, feet, after which it embraces the surrounding branches of trees and spreads to a great distance in every direc- tion. The main root sends out many long tendrils, all of like thickness, covered with a brownish bark, of sometimes of a dark-gray color. These tendrils are ^ fibrous, and about as thick as a quill. They present a constant tendency to become crooked, and they are also wrinkled longitudinally, with here and there some smaller lateral fibres branching off from the sides. It is in the bark or epidermis of the rhizomes that the medicinal virtue lies; but the tendrils — both rhi- zome and bark — are collected together, and no at- tempt is made to separate them, until they have reached their commercial destination. Indeed, even these are sold together, the mode t)f preparing the root being left to the choice of the consumer, or the apothecary who procures it. The Mundrucus collect it during the six months of the rainy season, partly because during the remaining six they are otherwise employed, and partly for the reason that, in the time of rain, the roots are more easily extracted from the damp soil. The process sim- ply consists in digging them up or dragging them out of the earth — the latter mode especially where the tendrils He near the surface, and they will pull up without breaking. If the main root be not dug out, it will send forth new tendrils, which in a short time would yield a new crop ; but the improvident savages make no prudential calculations of this kind — present convenience forming their sole consideration; and on this account both the root and plant are generally d© stroyed by them during the operation of 'collecting. OR BEHEAIERS. 37 As already stated, thi3 labor devolves upon thfl ^onien, who are also assisted in it by their children They proceed into the depths of the forest — where che siinlax grows in greatest abundance — and aftei collecting as much root as they can carry home witi them, they return with their bundles to the malocca When fresh gathered the sarsaparilla is heavy enough - — partly on account of the sap which it then contains, and partly from the quantity of the mud or earth that adheres to the corrugated surface of the roots. It is extremely probable that in this fresh state the virtue of the sarsaparilla, as a blood-purifier, is much greater than after it has passed through the channels of commerce; and the writer of this sketch has some reason, derived from personal experience, to believe that such is the case. Certain it is, that the reputa- tion of this invaluable drug is far less in countries where the plant does not grow, than in those where it is common and can be obtained in its fresh state. In all parts of Spanish America its virtues are un- questioned, and experience has led to a more extensive use of it there than elsewhere. It is probable, there- fore, that the virtue exists in the juice rather than the cortical integument of the rhizome ; and this of course would be materially altered and deteriorated, if not altogether destroyed, in the process of exsiccation, which must necessarily take place in the time required for transporting it to distant parts of the world. In the European pharmacopeia it is the epidermis of the root which is supposed to contain the ^anitarv principle ; and this, which is of a mucilaginous nature and slightly bit- ter taste, is employed, both in decoctions and infusions, 38 MUNDRUCUS, as a tonic and alterative. In America, however, it is generally taken for what is termed purifying the blood — for .he same purpose as the rhizomes o^ the Lauriu sassafras and other plants are used ; but the sarsaparilia is generally considered the best, and it certainly is the best of all known medicines for this purpose. Why it Las fallen in the estimation of the Old World practition- ers, or why it never obtained so great a reputation as it has in America, may arise from two circumstances. First, that the root offered for sale is generally the pro duct of the less valuable species ; and second, that the sap, and not the rhizome, may be the part that contains the virtuous principle. When the collected roots have been kept for awhile they become dry and light, and for the convenience of stowage and carriage — an important consideration to the trader in his eight-ton garratea — it is necessary to have the roots done up in packages of a uniform length and thickness. These packages are formed by laying the roots side by side, and doubling in the ends of the longer ones. A bundle of the proper size for stowage contains an arroba of twenty-five pounds, though the weight varies according to the condition of the root. Uniformity in size is the chief object aimed at, and the bundles are made of a round or cylindrical shape, about five inches in diameter, and something more than a yard in length. They are trimmed off small at the ends — so as to admit of stowage without leaving any empty Bpace between two tiers of them — and each bundle ij tightly corded round from one end to the other with a * sipo," or creeping plan t. Tt has been stated that this 'sipo" is a root of the OR BEHEADERS. 39 sarsaparilla itself, with the baik scrap id off; and, in- deed, its ovrn root would serve well enough — wpt$ it not that putting it to such a use would destroy its medi- cinal value, and thus cause a considerable waste of the costly material. The sarsaparilla is not to be had for nothing even upon the banks of the Tapajos. A bundle of the best quality does not -leave the hands of the Mun- drucn until about four dollars' worth of exchange com- modities have been put into them, which would bring the price of it to something over sixpence a pound. He is, therefore, a little particular about wasting a material that has cost him — or rather his wife and children — so much trouble in collecting. His cordage is obtained more cheaply, and consists of the long, flexi- ble roots of a species of pothos, which roots — being what are termed aerial and not buried in the ground — require no labor or digging to get at them. It is only necessary to stretch up the hand, and pull them down from the tops of lofty trees, from which they hang like streamers, often to the length of a hundred feet. These are toughened by the bark being scraped off; and when that is done they are ready for use, and serve not only to tie up the bundles of sarsaparilla, but for many other purposes in the domestic economy of the Mundrucus. . In addition to the sarsaparilla, the Mundrucu fur- nishes the trader with several other items of commercial value — for his climate, although one of the most un- healthy in all the Amazon region, on account of its great heat and humidity, is for that very reason one of the most fertile. Nearly all those trojical vegetable pro- ducts which are characteristics of Brazilian export com merce can here be produced of the most luxuriant kind 40 MUNDRUCUS, but it is only those that grow spontaneously at his very doors that tempt the Mundrucu to take .the trouble of collecting them. There is one article, however, which he not only takes some trouble to collect, but also to manufacture into an item of commercial exchange — a very rare item indeed. This is the guarana, which is manufactured from the fruit of a tree almost peculiar to the Mundrucu territory — since nowhere is it found so abundantly as on the Tapajos. It is so prized in the Brazilian settlements as to command almost its weight in silver when trans- ported thither. It is the constituent element of a drink, which has a stimulating effect on the system, somewhat more powerful than tea or coffee. It will prevent sleep; but its most valuable property is, that it is a good feb- rifuge, equal to the best quinine. Guar ana is prepared from the seeds of an inga — one of the Mimosacce. It is a low, wide-spreading tree like most of the mimosa family. The legumes are gathered, and the seeds roasted in them. The latter are then taken out, and T *fter being ground to powder, are mixed with water so as to make a tough paste, which is moulded into little bricks, and when dried is ready for use. The beverage is then prepared by scraping a table-spoonful of dust from the brick, and mixing it with about a pint of water ; and the diy paste, keeping for any length of time, is ready whenever wanted. The guarana bush grows elswhere in the Amazon valley, and on some headwaters of the Orinoco, where certain tribes also know how to prepare the drink. But U is sparingly distributed, and is nowhere so common as on the upper Tapajos ; hence its high price in tht /)R BEHEADERS. 41 markets of llrazil Tlie Mundrucu manufactures it, not only fcr "fcsaie use," but for "exportation." He prepares another singular article of luxury, and this he makes exclusively for his own use, — not for the gratification of his lips or palate, but foi his nose, — in other words, a snuff. Do not fancy, ho\ sver, that it is snuff of the ordinary kind — the pulverL ed produce of innocent tobacco. No such thing ; but a composition of such a powerful and stimulating character, that he who inhales it feels as if struck by an electric shock ; his body trembles ; his eyes start forward as if they would forsake their sockets ; his limbs fail to support him ; and he drops to the earth like one in a state of intoxi- cation ! For a short time he is literally mad ; but the fit is soon over, — lasting usually only a few minutes, — and then a feeling of renewed strength, courage, and joyousness succeeds. Such are the consequences of taking snuff with a Mundrucu. And now to describe the nature of the substance which produces these powerful effects. Like the guarana this snuff is a preparation, having for its basis the seeds of a leguminous tree. This time, however, it is an acacia, not an inga. It is the acacia niopo ; so called because u niopo " is the name given to the ~nuff itself by certain tribes (the Ottomacs and others), who, like the Mundrucus, are snuff-takers. It is also called curupa, and the apparatus for preparing and taking it — for there is an apparatus of an exten- tensive kind — is termed parica, in the general language {lingoa geral) of the Amazonian regions. We shall describe the preparation, the apparatus, ani the ceremonial. 4-2 MUNDRUCUS, The pods of the Acacia niopo — a small tree, witi very delicate pinnate leaves — are plucked when ripe. They are then cut into small pieces and flung into a •vessel of water. In this they remain until macerated, and until the seeds have turned black. These are then picked out, pounded in a mortar, which is usually the pericarp of the sapugaia, or "monkey-pot" tree (Lecy- this ollaria). The pounding reduces them to a paste, which is taken up, clapped between the hands and formed into little cakes — but not until it has been mixed with some manioc flour, some lime from a burnt shell (a helix), and a little juice from the fresh leaves of the "abuta" — a menispermous plant of the genus Cocculus. The cakes are then dried or " barbecued " upon a primitive gridiron — the bars of which are sap- lings of hard wood — and when well-hardened the snuff is ready for the " box." In a box it is actually earned — usually one made out of some rare and beautiful shell. The ceremonial of taking the snuff is the most singular part of the performance. When a Mundrucu feels in- clined for a " pinch " — though it is something more than a pinch that he inhales when he does feel inclined — 1:3 takes the cake out of the box, scrapes off about a spoon- ful of it into a shallow, saucer-shaped vessel of the cala- bash kind, and then spreads the powder all over the bottom of the vessel in a regular u stratification." The spreading is not performed by the fingers, but with a tiny, pencil-like brush made out of the bristles of the great ant-eater (Myrmecophaga jubata). He is in no hurry, but takes his time, — for as you may guess from its effects, the performance is not one 98 OR BEHEADERS. ! > often repeated as that of ordinary snuff-taking. When the niopo dust is laid to his liking, another implement is brought into play, the construction of which it is also necessary to describe. It is a " machine " of six to eight inches in length, and is made of two quills from the wing of the gaviao real, or "harpy eagle" (Harpyia destructor). These quills are placed side by side for the greater part of their length, forming two parallel tubes, and they are thus neatly whipped together by a thread. At one end they are pressed apart so as to di- verge to a width corresponding to the breadth between the Mundrucu's nostrils, — where it is intended they shall be placed during the ceremony of snuff-taking. And thus are they placed, — one end of each quill being slightly intruded within the line of the septum, while the other end rests upon the snuff, or wanders over the surface of the saucer, till all the powder placed there is drawn up and inhaled, producing the convulsive effects already detailed. The shank-bone of a species of bird — thought to be be a plover — is sometimes used instead of the quills. It is hollow, and has a forking-tube at the end. This kind is not common or easily obtained, for the niopo- taksr who has one, esteems it as the most valuable item of his apparatus. Snufifing the niopo is not exclusively confined to the Mund r ucu. We have seen elsewhere that it is also a habit of the dirt-eating Ottomacs ; and "other tribes on the upper Amazon practise it. But the Mahiies, already mentioned as the allies of the Mundrucus, are the most confirmed snuff-takers of all. Anther odd custom of the Mundrucus is their haDil 44 MUNDRUCUS, of u tatooing." I speak of real tatooing, — that is, mark ing the skin with dots and lines that cannot be effaced, in contradistinction to mere painting, or staining, whicb can easily be washed off. The ' Mundrucus paint also, with the anotto, huitoc, caru'xi, and other pigments, but in this they only follow the practice of hundreds of other tribes. The true tatoo is a far different affair, and scarce- ly known among the aborigines of America, though com- mon enough in the islands of the South Sea. A few other Indian tribes practise it to a limited extent, — as is elsewhere stated, — but among the Mundrucus it is an " institution ;" and painful though the process be, it has to be endured by every one in the nation, " every moth- er's son," and daughter as well, that are cursed with a Mundrucu for their father. It is upon the young people the infliction is performed, — when they are about eight or ten years of age. The tatoo has been so often described, that I should not repeat it here ; but there are a few " points " peculiar to Mundrucu tatooing, and a few others, not elsewhere understood. The performance is usually the work of certain old crones, who, from long practice, have acquired great skill in the art. The chief instrument used is a comb of thorns, — not a single thorn, as is gererally stated, — but a tier or row of them set comb-fashion. These thorns are the spines of the "murumuru," or "pupunha" palm (Gullielmia sptcicsa). Humboldt states that this palm is smooth anj spineless, but in this the great, good man was m error. Its trunk is so covered with thorns or spines, that when the Indians require to clinib it — for the purpose of OR BEHEAPEKS. 45 procuring the valuable fruits, which they cat \ariouslj prepared — they have to erect a staging, or rude sort of ladder, to be able to get at them. The comb, then, is pressed down upon the skin of the " tatooee," till all the points have penetrated the flesh, an I a row of holes is laid open, from which the blood flows profusely. As soon as this can be wiped off, ashes of a burnt gum or pitch are rubbed into the wounds, which, when healed, appear like so many dots of a deep bluish or black color. In this way the young Mundru- cus, both boys and girls, get those regular rows of dotted lines, which traverse their forehead and cheeks, their arms and limbs, breasts, and bodies in such eccentric fashion. It has often been asked how these lines of dots were carried over the skin in such straight and symmet- rical rows, forming regular parallel lines, or other geo- metrical patterns. The " comb " will explain the mys- tery. The tatoo, with a few strings of shell-beads or neck- laces, and bracelets of monkey and jaguar teeth, is all the dress which is permitted to the Mundrucu belle. In Mundrucu-land it is the reverse of what is practised among civilized people : the men are the exponents of the fashions, and keep exclusively to themselves the cos- metics and bijouterie. Not contented with being tatooed, these also paint their bodies, "by way of " overcoat," and also adorn themselves with the bright feathers of birds. They wear on their heads the beautiful circlet of macaw plumes, and on grand occasions appear in the magnificent " feather dress," so leng celebrated as the peculiar cos- tume of the tropical- forest Indian. These dresses their women weave and border, at a sacrifice of much tedioui 46 MUNDRUCUS, labor. They also ornament their arms and kgs witk rows of feathers around them, the tips turned upward and backward. The tatooing is confined to the Mundrucus proper,— their allies, the Mahues not following the practice, but contenting themselves with a simple " coat " of paint. It is difficult to say what motive first inducted human beings into this singular and barbarous custom. It is easier to tell why it is still followed, and the " why " is answered by saying that the Mundrucus " scarify " them- selves, because their fathers did so before them. Many a custom among civilized nations, but^ little less ridicu- lous, if we could only think so, rests upon a similar basis. Perhaps our modern abominable hat — though it has a different origin — is not less ludicrous than the tatooed patterns of the savage. Certainly it is quite equal to it in ugliness, and is likely to rival it in permanence, — to our sorrow be it said. But even ice deal slightly in the tatoo. Our jolly Jack would be nobody in the foreca^lo without " Polly," in blue, upon his weather-beaten breast, and the foul anchor upon his arm. But the Mundrucu baptizes his unfortunate offspring in a still more savage fashion. The tattoo may be termed the baptism in blood, performed at the tender age of ten. When the youth — fortunately it does not extend to the weaker sex — has attained to the age of eighteen, he has *hen to undergo the tocandeira, which deserves to be called the baptism of fire ! This too merits description. When the Mundrucu youth would become a candidate for manhood, a pair of u gloves " is prepared for him. These consist of two pieces of a palm-tree bark, with the pith hollowed out, but OR BEHEADE&5. 4 7 left in at one end. The hollow part is of sufficient diam- eter to draw over the hands loosely, a.nd so long as to reach up to mid-arm, after the fashion of gauntlets. The " gloves " being got ready, are nearly tilled with ants, not only the venomous red ants, but ail other spe- cies, large or small, that can either bite or sting, of which tropical South America possesses an endless variety. With this " lining " the " mittens " are ready for use, and the " novice " is compelled to draw them on. Should he refuse, or even exhibit a ilisposition to shrink from the fiery trial, he is a lost man From that hour he need never hold up his head, much less offer his hand and heart, for there is not a maiden in all Mundrucu-land that would listen to his softest speech. He is forever de- barred from the pleasure of becoming a benedict. Of course he does not refuse, but plunging his hands into the u mittens," into the very midst of the crawling host, he sets about the ceremony. He must keep on the gloves till he has danced before every door in the village. He must sing as if from very joy ; and there is plenty of music to accompany liim, drum.i and fifes, and human voices, — for his parents and relatives are by his side encouraging him with their pongs and gestures. He is in pain, — in positive agony, — for these venomous ants both sting and bite, and have been bury at both from the very first moment. Each moment his agony grows more intense, his suf- ferings more acute, for the poison is thrilling through his veins, — he turns pale, . — his eyes become blood-cast, — his breast quivers with emotion and his limbs tremble beneath him ; but despite all this, woe to him if he utter ft cry of weakness ! It would brand him with an eternp' 48 MUNDRUCUS, stigma, — he would never be suffered to cany tht Mun drucu lance to battle, — to poise upon its point the ghast« ly trophy of the Beheaders. On, on, through the howling throng, amidst friends and relatives with faces anxious as his own ; on to the sound of the shrill-piping reed and the hoarse booming of the Indian drum ; on till he Stands in front of the cabin of the chief! There again fche song is sung, the "jig" is danced, both proudly pro- longed till the strength of the performer becomes com- pletely exhausted. Then, and not till then, the gloves are thrown aside, and the wearer falls back, into the arms of his friends, " sufficiently punished ! " This is the hour of congratulation. Girls gather round him, and fling their tatooed arms about his neck. They cluster and cling upon him, singing Ins song of triumph ; but just at that crisis he is not in the mood for soft ca- resses ; and, escaping from their blandishments, he makes a rush towards the river. On reaching its bank he plunges bodily in, and there remains up to his neck in the water, till the cooling fluid has to some extent eased his aching arms, and tranquillized the current of his boiling blood. When he emerges from the water, he is a man, fit stuff for a Mundrucu warrior, and eligible to the hand of a Mundrucu maiden. It may bo remarked that this terrible ordeal of the Mundrucus, though, perhaps, peculiar among South- American Indians, has its parallel among certain tribes of the north, — the Mandans and others, as detailed by Catlin, one of the most acute of ethnological observers. The scalp trophy, too, of the Northern Indian has its analogy in a Mundrucu custom — that which distinguish- es him most of all, and which has won far him the terri ble title of " Beheader." OR BEHEADERS. 49 Thi3 singular appellation is now to be explained. When a Mundrucu has succeeded in killing an enemy, he is not, like Ins northern compeer, satisfied with only Jhe skin of the head. He must hove the whole head^ scalp and skull, bones, brains, and all ! And he takes all, severing the head with his knife by a clean cut across the small of the neck, and leaving the trunk to the 7ul« ture king. With the ghastly trophy poised upon the point of his lance, he returns triumphant to the malocca to receive the greetings of his tribe and the praises of Lis chief. But the warlike exploit requires a memento — some token by which he may perpetuate its fame. The art of printing does not exist among the Mundrucus, and there is no friendly pen to record the deed. It has been done, — behold the evidence ! much clearer than often accom- panies the exploits of civilized heroes. There is the evi- dence of an enemy slain ; there is the grim, gory voucher, palpable both to sight and touch — proof positive that there is a dead body somewhere. Of course, such evidence is sufficient for the present ; but how about the future ? As time passes, the feat may i 3 forgotten, as great deeds are elsewhere. Somebody may e/en deny it. Some slanderous tongue may whisper, or insinuate, or openly declare that it was no exploit after all — that there was no dead man ; for the vultures by this time would have removed the body, and the white ants (termites) would have equally extinguished all traces of the bones. How, then, are the proofs to be preserved ? By preserving the head ! And this is the very idea that is in the mind of the Mundrucu warrior. He is resolved not to permit his exploit to be burhd in fion ; j burying the head ot" his enemy. That tongue, I tell the tale to ] : that pallid [h, perhaps, it may become a little >hri veiled a." will still be smooth em -now that 'oo. and to be identified as the skin of an en- eim s Kfundrucu, yet unborn, will road in the eon; of that grinning and gory witness, the /> prowess. The head, therefore, must be and it i< preserved with as much care as the 'herisheil portrait of a famous ancestor. The relic i< >\ /. as it^ out of affection tor him The brains and eye-balls are removed, to I we?< of desiccation ; but false eyes are inserted, .100 ::.- nngue, teetli, and ears, Scalp, skull, and ha only retained, but .1 " out in tlu most approved style of fashion. The long rnbed out, parted, and ar- : feathers >( roek-.-oek and macaw are planted behind the ears and twisted in tlie hanging tresses. An ornamental string pass* gh the tongue, and the trophy is suspended trotv the beams of the great maloeea. It is not permitted to remain there. Li. -ome dark niche of this a — this |Uiri Westminster — it might be overlooked and forgotten To prevent this it is of: en br :h. and receives mam an air jng. On all ? is does it appear, poised upon the point of the v. md even in peaeei it may be seen — along with hundreds of its like — placed in the circular row are;, manioc cl ling its demure countenance te the tabors o( the held. OK BEHEADEKS. 51 It is not a little singular that this custom of embalm- ing the heads of their enemies is found among the Dyak? af Borneo, and the process in both places is ludicrously similar. Another rare coincidence occurs between the Amazonian tribes and the Bornean savages, viz. in both being provided with the blow -gun. The gravitana of the American tribes is almost identical with the sum- pitan of Borneo. It furnishes a further proof of our theory regarding an original connection between the American Indians and the savages of the great South Sea. The Mundrucu is rarely ill off in the way of food. When he is so, it is altogether his own fault, and charge- able to his indolent disposition. The soil of his territory is of the most fertile kind, and produces many kinds of edible fruits spontaneously, as the nuts of the pupunha palm and the splendid fruits of the Bertholetia excelsa, or juvia-tree, known in Europe as " Brazil nuts." Of these then are two kinds, as mentioned elsewhere, the sec- ond being a tree of the genus Lecythys, — the Lecyihys ollaria, or " monkey-pot " tree. It obtains this trivial name from the circumstance, first, of its great pericarp, almost as large as a child's head, having a movable top or lid, which falls off when the fruit Vipens ; and second- ly from the monkeys being often seen drawing the seeds or nuts out of that part of the shell which remains attached to the tree, and which, bearing a considerable resemblance to a pot in its shape, is thus very appro- priately designated the pot of the monkeys. The com- mon Indian name of the monkey-pot tree is sapucaya, and the nuts of this species are so called in commerce, though tQey are also termed Brazil-nuts. They are of u 52 MUNDRUCUS, "more agreeable flavor than the true Brazil-nuts, and nol so easily obtained, as the Lecythys is less generally dis- tributed over the Amazonian valley. It requires a pecu- liar soil, and grows only in those tracts that are subjeei to the annual inundations of the rivers. The true Brazil-nuts are the "juvia" trees of ths Indians ; and the season for collecting them is one of the harvests of the Mundrucu people. The great pericarps — resembling large cocoa-nuts when stripped of the fibres — do not open and shed their seeds, as is the case with the monkey-pot tree. The whole fruit falls at once ; and as it is very he*avy, and the branches on which it grows are often nearly a hundred feet from the ground, it may easily be imagined that it comes down like a ten-pound shot ; in fact, one of them falling upon the head of a Mundrucu would be very likely to crush his cranium, as a bullet would an egg-shell ; and such accidents not unfrequently occur to persons passing im- prudently under the branches of the Bertholetia when its nuts are ripe. Sometimes the monkeys, when on the ground looking after those that have fallen, become vic- tims to the like accident ; but these creatures are cun- ning reasoners, and being by experience aware of the danger, will scarce ever go under a juvia-tree, but when passing one always make a wide circuit around it. The monkeys cannot of themselves open the great pericarp, as they do that of the "sapucuya," but are crafty enough to get at the precious contents, notwithstanding. Li doing this they avail themselves of the help of other creatures, that have also a motive in opening the javia shells — cavies and other small rodent animals, whoee teeth, formed for this very purpose, enable them OR BEHEADERS. 53 lo gna\s a note in the ligneous pericarps, hard and thick as the} are. Meanwhile the monkeys, squatted around, wa tch the operation in a careless, nonchalant sort of way, as if they had no concern whatever in the result ; but aa sooii as they perceive that an entrance has been effected, big enough to admit their hand, they rush forward, drive off the weaker crea ure, who has been so long and la boriously at work, and take possession of the prize. Neither does the Mundrucu nut-gatherer get posses- sion of the juvia fruit without a certain degree of danger and toil. He has to climb the tallest trees, to secure the whole crop at one time ; and while engaged in collecting those upon the ground, he is in danger of a blow from odd ones that are constantly falling. To secure his skull against accidents, he wears upon his head a thick wooden cap or helmet, — after the fashion of the hats worn by our firemen, — and he is always careful to keep his body in an upright attitude, stooping as seldom as he can avoid doing so, lest he might get a thump between the shoul- ders, or upon the spine of his back, which would be very likely to flatten him out upon the earth. These Brazil- nuts furnish the Mundrucu with a portion of his food, — as they also do many other tribes of Amazonian Indians, — and they are also an item of Indian commerce, being collected from among the different tribes by the Portu- guese and Spanish traders. But the Mundrucu does not depend altogether on the spontaneous productions of the forest, which at best furnish only a precarious supply. He does something in the agricultural line, — cultivating a little manioc root, with plantains, yams, and other tropical plants that pro- duce an enormoi :s yield with the very slightest trouble 54 MUNDRUCUS, or attention ; and this is exactly what suite Lim. A few days spent by the little community in the yam patch — « or rather, by the women and children, for these are the agricultural laborers in Mundrucu land — is sufficient to ensure an abundant supply of breadstuff for the whole year. With regard to flesh-meat he is net so well off, for the domesiic animals, and oxen more especially, do not thrive in the Amazon country. In Mundrucu land, the carnivorous jaguar, aided by flies and vampire bats, would soon destroy them, even if the Indian had the inclination to raise them, which he has not. Instead of beef, therefore, he contents himself with fish, and occasionally a steak from the great tapir, or a griskin of manati. Birds, too, furnish him with an occasional meal ; but the staple article of his flesh diet is obtained from the quadrumana, — the numerous spe- cies of monkeys with which his forests abound. These he obtains by shooting them down from the trees with his bow and arrows, and also by various other hunting devices. His mode of cooking them is sufficiently peculiar to be described. A large log fire is first kindled and per- mitted to burn until a sufficient quantity of red cinders are produced. Over these cinders a grating is erected with green saplings of wood, laid parallel to each other like the bars of a gridiron, and upon this the u joint" is laid. Nothing is done to the monkey before its being placed on the gridiron. Its skin is not removed, and even the intestines are not always taken out. The fire will singe off the hair sufficiently to content a Mundrucu stomach, and the hide is broiled and eaten with the flesh. It is thus literally " came con cuero." OR BEHEADERS. 55 It may bo observed that this forest gridiron, 01 a bar- becue," as it is properly termed, is not an idea exclu- sively confined to South America. It is in use among the Indians of the north, and various uncivilized tribes in other parts of the world. Sometimes the Mundrucu does not take the trouble to construct the gridiron. When on the march in some warlike expedition that will not allow time for being particular about the mode of cooking, the joint is broiled upon a spit over the common fire. The spit is simply a stick, sharpened at both ends, one of which impales the monkey, and the other is stuck into the ground. The stick is then set with a lean towards the lire, so as to bring the carcass over the blaze. While on the spit the monkey appears in a sitting position, with its head up- ward, and its long tail hanging along the sapling, — just as if it were still living, and in one of its most natural attitudes, clinging to the branch of a tree ! The sight it sufficiently comical ; but sometimes a painful spectacle has been witnessed, — painful to any one but a savage : when the young of the monkey has been captured along with its dam, and still recognizing the form of its parent, — even when all the hair has been singed off, and the skin has become calcined by the fire, — is seen rushing forward into the very flames, and with plaintive cry in- viting the maternal embrace ! Such an affecting incident has been often witnessed amid the forests of Amazonia. We conclude our sketch of the Mundrucus, by stating that their form of government is despotic, though not to an extreme degree. The " tushao," or chief, has con- siderable power, though it is not absolute, and does not extend to the taking of life, — unless the object of hw 56 MUNDRUCUS. displeasure be a slave, and many of these are held in abject bondage among the Mundrucus. The Mundrucu religion resembles that of many other tribes both in North and South America. It consists in absurd ceremonies, and appeals to the good and evil spirits of the other world, and is mixed up with a vast deal of quackery in relation to the ills that afflict the Mundruou in this life. In other words, it is a combina- tion of the priest and doctor united in one, that arch- charlatan known to the North-American Indians a* the " Medicine-man/' and among the Mundrucus a& tfet -Page." THE CENTAURS OF THE GRAN CHACO/' I have elsewhere stated that a broad band of in- dependent Indian territory — that is, territory never really subdued or possessed by the Spaniards — trav- erses the interior of South America, extending longitu- dinally throughout the whole continent. Beginning at Cape Horn, it ends in the peninsula of the free Goajiros, which projects into the Caribbean Sea, — in other words, it is nearly 5,000 miles in length. In breadth it varies much. In Patagonia and a portion of the Pampas country it extends from the Atlantic to the Pacific, and it is of still wider extent on the latitude of the Amazon river, where the whole country, from the Atlantic to the Peruvian Andes, — with the exception of some thinly- placed Brazilian settlements, — is occupied by tribes of independent Indians. At either point this territory will appear — upon maps — to be interrupted by tracts of country possessing civilized settlements. The name? of towns and villages are set as thickly as if the country were well peopled ; and numerous roads are traced, forming a labyrinthine network upon the paper. A broad belt of this kind extends from the Lower Parana (La Plata) to the Andes of Chili, constituting the uppei 58 THE CENTAURS OF provinces of the " Argentine Confederation ; " ano&ei apparently joins the settlements of Bolivia and Brazil* and again in the north, the provinces of Venezuela ap- near to be united to those of New Granada. All this, however, is more apparent than real. The towns upon the maps are in general mere rancherio£, or collections o. ? huts ; some of them are the names of furti fied posts, and a large proportion are but ruins, — the ruins of monkish mission settlements long since gone to destruction, and with little else than the name on the map to testify that they ever had an existence. The roads are no roads at all, nothing more than tracings on the chart showing the general route of travel. Even across the Argentine provinces — where this nomenclature appears thickest upon the map — the hoi ^-Indian of the Pampas extends his forays at will; his "range " meeting, and, in some cases, " dove-tailing" into that of the tribes dwelling upon the northern side of these settlements. The latter, in their turn, carry their plundering expeditions across to the Campos Parexis, on the head- waters of the Amazon, whence stretches the in- dependent territory, far and wide to the Amazon itself; thence to the Orinoco, and across the Llanos to the shores of the Maracaibo Gulf — the free range of the independ- ent Goajiros. This immense belt of territory, then, is in actual pos* session of the aborigines. Although occupied at a few points by the white race, - - Spanish and Portuguese, — ■ the occupation scarce deserves the name. The settle- ments are sparse and rather retrograde than progressive. The Indian ranges through and around them, wherever and whenever his inclination leads k^n ; and only when THE ' GRAN CHACO." 59 somt humiliating treaty has secured him a temporary respite from hostilities does the colonist enjoy tranquillity. At other times he lives in continual dread, scarce daring to trust hinnelf beyond the immediate vicinity of his fcouse or village, both of winch he has been under the ecessity of fortifying. It is tru<3 that at one period of South American his- tory things were not quite so bad. When the Spanish nation was at the zenith of its power a different condi- tion existed ; but even then, in the territory indicated, there were large tracts circumstanced just as at the pres- ent hour, — tracts which the Spaniards, with all their boasted warlike strength, were unable even to explore, much less to subdue. One of these was that which f orms the subject of our sketch, " El Gran Chaco." Of all the tracts of wild territory existing in South America, and known by the different appellations of Pampas, Paramos, Campos Parexis, the Puna, the Pa* jonal, Llanos, and Montanas, there is none possessed of a greater interest than that of El Gran Chaco, — perhaps not one that equals it in this respect. It is interesting, not only from having a peculiar soil, climate, and pro- ductions, but quite as much from the character and his- tory of its inhabitants, both of which present us' with traits and episodes truly romantic. The " Gran Chaco " is 200,000 square miles in extent, or twice the size of the British Isles. Its eastern boun- dary is well defined, being the Paraguay river, and itd continuation the Parana, down to the point where the lat- ter receives one of its great western tributaries, the Sala- do ; and this last is usually regarded as the southern and vestera boundary of the Chaco. Northward its limits 60 THE CENTAUKS OF are ecarcely so definite ; though the highlands of Bolivia and the old missionary province of Chiquitos, forming the water-shed between the rivers of the La Plata and the Amazonian basins — may be geographically regarded as the termination of the Chaco in that direction. North and south it extends through eleven degrees of latitude ; east and west it is of unequal breadth, — sometimes ex- panding, sometimes contracting, according to the ability of the white settlers along it borders to maintain their frontier. On its eastern side, as already stated, the fron- tier is definite, and terminates on the banks of the Para- guay and Parana. East of this line — coinciding almost with a meridian of longitude — the Indian of the Gran Chaco does not roam, the well-settled province of Corri- entes and the dictatorial government of Paraguay pre- senting a firmer front of resistance ; but neither does the colonist of these countries think of crossing to the west- ern bank of the boundary river to form any establishment there. He dares not even set his foot upon the territory of the Chaco. For a thousand miles, up and down, the $wo races, European and American, hold the opposite banks of this great stream. They gaze across at each other : the one from the portico of his well-built man- sion, or perhaps from the street of his town ; the other, standing by his humble " toldo," or mat-covered tent, — more probably, upon the back of his half-wild horse, reined up for a moment on some projecting promontory that commands the view of the river. And thus have these two races gazed at each other for three centuries, with little other intercourse passing between them than that of a deadly hostility. The surface of the Gran Chaco is throughout of 8 THE 'GRAN CHACO." 61 champaign character. It may be described as a vast plain. It is not, however, a continuation of the Pampas, since the two are separated by a more broken tract of country, in which lie the sierras of Cordova and San Luis, with the Argentine settlements already mentioned. Besides, the two great plains differ essentially in their character, even to a greater extent than do the Pampas themselves from the desert steppes of Patagonia. Only a few of the animal and vegetable productions of the Gran Chaco are identical with those of the Pampas, and its Indian inhabitants are altogether unlike the sanguinary savages of the more southern plain. The Chaco, ap- proaching man) degrees nearer to the equator, is more tropical in its character ; in fact, the northern portion of it is truly so, lying as it does within the torrid zone, and presenting the aspect of a tropical vegetation. Every inch of the Chaco is within the palm region ; but in its northern half these beautiful trees abound in numberless species, yet unknown to the botanist, and forming the characteristic features of the landscape. Some grow in forests of many miles in extent, others only in " clumps," with open, grass-covered plains between, while still other species mingle their graceful fronds with the leaves and branches of dicotyledonous trees, or elapsed in the em- brace of luxuriant llianas and parasitical climbers form groves of the most variegated verdure and fantastic out- lines. With such groves the whole surface of the Chaco country is enamelled ; the intervals between being occu- pied by plains of rich waving grass, now and then tractf of morass covered with tail and elegant reeds, a few arid spots bristling with singular forms of algarobia and caC' tin, and, in some peaces, isolated rocky moimds, of dome G2 THE CENTAURS OF or conical shape, rising above the general level of tl c plains, as if intended to be used as watch-towers for theil guardianship and safety. Such are the landscapes which the Grand Chaco pre sents to the eye — far different from the bald and uni- form monotony exhibited in the aspect of either Prairie or Pampa ; far grander and lovelier than either — in point of scenic loveliness, perhaps, unequalled on earth. No wonder, then, that the Indian of South America esteems it as an earthly Elysium ; no wonder that the Spaniard dreams of it as such, — though to the Spanish priest and the Spanish soldier it has ever proved more of a Purgatory than a Paradise. Both have entered upon its borders, but neither has been able to dwell within its domain ; and the attempts at its conquest, by sword and cross, have been alike unsuccessful, — equally and fatally repulsed, throughout a period of more than three hundred years. At this hour, as at the time of the Peruvian conquest, — as on the day when the ships of Mendoza sailed up the waters of the Parana, — the Gran Chaco is an unconquered country, owned by its aboriginal inhabitants, and by them alone. It is true that it is claimed, both by Spaniard and Portuguese ; and by no less than four separate claimants belonging to these two nationalities. Brazil and Bolivia, Paraguay and the Argentine Confederation, all assert their title to a slice of this earthly paradise ; and even quarrel as to how their boundary lines should intersect it ! There is something extremely ludicrous in these claims, — since neither one nor other of the four powers can show the slightest basis for them. Not one of the ai ran pretend to the claim of conquest ; and far less eaD THE " GRAN CHACO." 63 they rest their rights upon the basis of occupation or possession. So far from possessing the land> not. one of them dare set foot over its borders ; and they are only too well pleased if its present occupants are contented to remain within them. The claim, therefore, of both Spaniard and Portuguese, has no higher title, than that some three hundred and fifty years ago it was given them by the Pope, — a title not less ludicrous than their kissing the Pope's toe to obtain it ! In the midst of these four conflicting claimants, there appears a fifth, and that is the real owner, — the " red Indian " himself. His claim has " three points of the law " in his favor, — possession, — and perhaps the fourth, too, — the power to keep possession. At all events, he has held it for three hundred years against all odds and all comers ; and who knows that he may not hold it for three hundred years more ? — only, it is to be hoped, for a different use, and under the influence of a more progressive civilization. The Indian, "then, is the undoubted lord of the " Gran Chaco." Let us drop in upon him, and see what sort of an Indian he is, and how he manages this majestic domain. After baring feasted our eyes upon the rich scenery of the land, — upon the verdant plains, mottled with copses of " quebracho " and clumps of the Caranday palm, — upon landscapes that resemble the most lordly parks, we look around for the mansions ai id the owners. The mansion is not there, but the owner stands before ws. We are at once struck by his appearance : his per* sou tall and straight as a reed, his frame muscular G4 THE CENTAURS OF his limbs round and well-proportioned, piereing coal black eyes, well-formed features, and slightly aquiline nose, — and perhaps we are a little surprised at the light color of his skin. In this we note a decided peculiarity which distinguishes him from most other tribes of his race. It is not a red Indian we behold, nor yet a cop* ver-colored savage ; but a man whose complexion is scarce darker than that of the mulatto, and not at all deeper in hue than many a Spaniard of Andalusian descent, who boasts possession of the purest " sangre azul ; " not one shade darker than thousands of Portu- guese dwelling upon the other side of the Brazilian fron- tier. And remember, that it is the true skin of the Chaco Indian we have before our view, — and not a painted one, — for here, almost for the first time, do we encoun- ter the native complexion of the aboriginal, undisfigured by those horrid pigments which in these pages have so often glared before the eyes of our readers. Of paint, the Chaco Indian scarce knows the use ; or, at all events, employs it sparingly, and only at interval*. on very particular and ceremonial occasions. We are spared, therefore, the describing his escutcheon, and a positive relief it is. It would be an interesting inquiry to trace out the cause of his thus abstaining from a custom almost uni- versal among his race. Why does he abjure the paint? Is it because he cannot afford it, or that it is not pro- curable in his country ? No ; neither of these can be offered as a reason. The "annotto" bush (Bixa 07 el land), and the wild-indigo, abound in his territory ; and Ve knows how to extract the colors of both, — for hi* THE " GRAN CHACON 05 women do extract them, and use them in dying the yam 3f their webs. Other dyewoods — a multitude of others — he could easily obtain ; and even the cochineal cactus, with its gaudy vermilion parasite, is indigenous to his land. It cannot be the scarcity of the material that pre- vents him from employing it, — what then ? The cause is unexplained ; but may it not be that this romantic savage, otherwise more highly gifted than the rest of his race, is endowed also with a truer sense of the beautiful and becoming ? Quien sake ? Let it not be understood, however, that he is altogether free from the " taint," — for he does paint sometimes, as already admitted ; and it must be remembered, more- over, that the Chaco Indians are not all of one tribe, nor of one community. There are many associations of them scattered over the face of this vast plain, who are not all alike, either in their habits or customs, but, on the con- trary, very unlike ; who are not even at all times friendly with each other, but occupied with feuds and vendetta* of the most deadly description. Some of these tribes paint most frightfully, while others of them go &till far< ther, and scarify their faces with the indelible tattoo, — a custom that in America is almost confined to the In- dians of the Chaco and a few tribes on the southern tributaries of the Amazon. Happily this custom is on the decline : the men practise it no longer ; but, by a singular' perversity of taste, it is still universal among the women, and no Chaco belle would be esteemed beau- tiful without a cross of bluish-black dots upon her fore- head, a hue of like points extending from the angle of each eye to the ears, with a variety of similar markings upon Jder cheeks, arms, and bosom. All this is done 66 THE CENT AUKS OF with the point of a thorn, — the spine of a mimosa, ci of the caraguatay aloe ; and the dark purple color is obtained by infusing charcoal into the fresh and bleeding punctures. It is an operation that requires days to com- plete, and the pain from it is of the most acute and pro- longed character, enduring until the poisoned wounds become cicatrized. And yet it is borne without a mur- mur, — just as people in civilized life bear the painfu] application of hair-dyes and tweezers. I need not say that the hair of the Chaco Indian does not need to be dyed, — that is, unless he were to fancy having it of a white, or a red, or yellow color, — not an uncommon fancy among savages. His taste, however, does not run that way any more than among civilized dandies, and he is contented with its natural hue, which is that of the raven's wing. But he is not contented to leave it to its natural growth. Only a portion of it, — that which covers the upper part of his head, — is permitted to retain its full length and flowing glories. For the remainder, he has a peculiar tonsure of his own ; and the hair immediately over the forehead — and sometimes a stripe running all around above the ears, to the back of the head — is either close shaven with a sharp shell, or plucked entirely out by a pair of horn tweezers of native manufacture. Were it not that the long and luxuriant tresses that still remain, — covering his crown, as with a crest, — the shorn circle would assimilate him to some orders of friars ; but, not- withstanding the similarity of tonsure, there is not much resemblance between a Chaco Indian and a brother of ths crucifix and cowl. This mode of " dressing the hair " is not altogether pei THE "GRAN CHACO." 67 J culiar to Die Indian of the Gran Chaco. It is also prac- tised by certain prairie tribes, — the Osage, Pawnee, and two or three others ; but all these carry the " razor " a little higher up, leaving a mere patch, or " scalp-look," upon the crown. The Cliaco tribes are beardless by nature ; and if a few hairs chance to show themselves upon cheek or chin, they are carefully " wed " out. In a like fashion both men and women serve their eyebrows and lashes, — sac- rificing these undoubted ornaments, as they say, to a principle of utility, since they allege that they can see better without them! They laugh at white men, who preserve these appendages, calling them " ostrich-eyed/* — from a resemblance which they perceive between hairy brows and the stiff, hair-like feathers that bristle round the eyes of the rhea, or American ostrich, — a well-known denizen of the Gran Chaco. The costume of the Chaco Indian is one of exceeding simplicity ; and in this again we observe a peculiar trait of his mind> Instead of the tawdry and tinsel orna- ments, in which most savages delight to array them- selves, he is contented with a single strip of cloth, folded tightly around his loins. It is usually either a piece of white cotton, or of wool woven in a tri-color of red, white, and blue, and of hues so brilliant, as to produce altogether a pretty effect. The wear of the women scarce differs from that of the men, and the covering of both, scant as it is, is neither inelegant nor immodest. It is well adapted to their mode of life, and to their climate, which is that of an eternal spring. When cold winds sweep over their grassy plains, they seek protection under the folds of a more ample covering, with which 08 THE CENTAURY OP they are provided, — a cloak usually made of the eofl fur of the " 0111™," or South American otter, or a robe of the beautiful spotted skin of the jaguar. They wear neither head-dress nor chanssure, — neither pendants from the nose, nor the hideous lip ornaments seen among other tribes of South America ; but many of them pierce the ears ; and more especially the women, who split the deli- cate lobes, and insert into them spiral appendages of rolled palm-leaf, that hang dangling to their very shoul- ders. It will be observed, thei^fore, that among the Chaco tribes the women disfigure themselves more than the men, and all, no doubt, in the interest of fashion. It will be seen that the simple dress we have described leaves the limbs and most part of the body bare. To the superficial observer it might be deemed an inelegant costume, and perhaps so it would be among Europeans, or so-called " whites." The deformed figures of Euro- pean people — deformed by ages of toil and monarchical serfdom — would ill bear exposure to the light, neither would the tripe-colored skin, of which they are so com- monly conceited. A very different impression is pro- duced by the rich brunette hue, — bronze, if you will, — especially when, as in the case of the Chaco Indian, it covers a body of proper shape, with arms and limbs in symmetrical proportion. Then, and then only, does costly clothing appear superfluous, and the eye at once admits that there is no fashion on earth equal to that of the human form itself. Above all does it appear graceful on horseback, and almost universally in this attitude doi i s the Chaco Indian exhibit it. Scarce ever may we meet him afoot, but always 0*1 the back of his beautiful horse, — the two THE "GKAN CHACO.' o9 together presenting the aspect of the Centaur. And probably in the resemblance he approaches nearer to the true ideal of the Grecian myth, than any other horse- man in the world ; for the Chaco Indians differ not only from oilier "horse Indians" in their mode of equitation, but aLo from every other equestrian people. The ab- surd high-peaked saddles of Tartar and Arab, with their gaudy trappings, are unknown to him, — unknown, too, the ridiculous paraphernalia, half-hiding the horse, in use among Mexicans, SoutJ}- American Spaniards, and even the Indians of other bribes, — despised by him the plated bits, the embroidered bridles, and the tinkling spurs, so tickling to the vanity of other New-World equestrians. The Chaco horseman needs no such accessories to his elegance. Saddle he has none, or only the slightest patch of jaguar-skin, — spurs and stirrups are alike ab- sent. Naked he sits upon his naked horse, the beautiful curvature of whose form is interrupted by no extraneous trappings, — even the thong that guides him scarce ob- servable from its slightness. Who then can deny his resemblance to the centaur? Thus mounted, with no other saddle than that de- scribed, no bridle but a thin strip of raw hide looped around the lower jaw of his horse, he will gallop wildly over the plain, wheel in graceful curves to avoid the bur- rows of the viscacha, pass at full speed through the close- standing and often thorny trunks of the palms, or, if need be, stand erect upon the withers of his horse, like a " star rider " of the Hippodrome. In this attitude he looks abroad for his enemies, or the game of which he may be in search ; and, thus elevated above surrounding object^ hs discovers the ostrich far oif upov the plain, the large 70 THE CEKTAUKS OF deer (cervus ca?npcstris), and the beautiful spotted roe* bucks that browse in countless herds upon the grass-cov a*red savannas. The dwelling of the Chaco Indian is a tent, not cov ered wiih skins, but usually with mats woven from the epidermis of young leaves of a palm-tree. It is set up by two long uprights and a ridge-pole, over which the mat is suspended — very much after the fashion of the tente d'abri used by Zouave soldiers. His bed is a ham- mock, swung between the upright poles, or oftener, be- tween two palm-trees growing near. He only seeks shelter in his tent when it rains, and he prevents its floor getting wet by digging a trench around the outside. He cares little for exposure to the sun ; but his wife is more delicate, and usually carries over her head a large bunch of rhea feathers, a la parasol, which protects her faca from the hot scorching beams. The tent does not stand long in one situation. Ample as is the supply which Nature affords in the wilds of the Chaco, it is not all poured out in any one place. This would be too much convenience, and would result in an evil consequence. The receiver of such a benefit would soon become indolent, from the absence of all necessity for exertion ; and not only his health, but his moral nature, would suffer from such abundance. Fortunately no such fate is likely to befall the Indian of the Chaco. The food upon which he subsists is de- rived from many varied sources, a few of which only are to be found in any one particular place, and each only at its own season of the year. For instance, upon the dry plainr he pursues the rhea and viscacha, the jaguar, puma, and partridges ; in woods and marshy { laces the dhTerenf THE " GRAN CHACO." 7 1 gpecies of wild hogs (peccaries). On the banks of rivers he encounters the tapir and capivara, and in their wa- ters, fish, utrias, geese, and ducks. In the denser forest- covered tracts he must look for the various kinds cf monkeys, which also constitute a portion of his food. When he would gather the legumes, of the algarobias — of several species — or collects the sugary sap of th<* caraguatay, he must visit the tracts where the mimosa and bromelias alone flourish ; and then he employs much of his time in searching for the nests of wild bees, from the honey of which and the seeds of the algarobia he distils a pleasant but highly intoxicating drink. To his credit, however, he uses this but sparingly, and only upon grand occasions of ceremony ; how different from the bestial chicha-drinking revellers of the Pampas ! These numerous journeys, and the avocations connect* with them, hinder the Chaco Indian from falling into hab- its of idleness, and preserve his health to a longevity thai is remarkable : so much so, that " to live as long as a Chaco Indian/' has become a proverbial expression in the settlements of South America. The old Styrian monk Dobrezhoffer has chronicled the astounding facts, that among these people a man of eighty is reckoned to be in the prime of manhood ; that a hundred years is accounted a common age ; and that many of them are still hale and hearty at the age of one hundred and twenty! Allowing for a little ex- aggeration in the statements of the monk, it is neverthe- less certain that the Indians of the Gran Chaco, partly owing to their fine climate, and partly to their mode of life and subsistence, enjoy health and strength to a very old age, and to a degree unknown in less-favored regions 72 THE CENTAURS OF of the world. Of this there is ample and trustworthy testimony. The food of the Chaco Indian is of a simple character, and he makes no use either of salt or spices. He is usu ally the owner of a small herd of cattle and a few sheep, which he has obtained by plundering the neighboring set- tlements of the Spaniards. It is towards those of the south and west that he generally directs his hostile fo- rays ; for he is at peace with the riverine provinces, — Brazilian, Paraguayan, and Correntine. In these excursions he travels long distances, crossing many a fordless stream and river, and taking along with him wife, children, tents, and utensils, in short, everything which he possesses. He fords the streams by swimming using one hand to guide his horse. With this hand he can also propel himself, while in the other he carries Ins long lance, on the top of which he poises any object he does not wish should be wetted. A " balza," called " pe- lota," made of bull's hide, and more like a square box than a boat, carries over the house utensils and the pup- pies, of which there are always a large number. The " precious baby " is also a passenger by the balza. The pelota is propelled, or rather, pulled over, by means of a tiller-rope, held in the teeth of a strong swimmer, or tied to the tail of a horse ; and thus the crossing is effected. Returning with his plunder — with herds of horned cattle or flocks of sheep — not unfrequently with human captives, women and children, the crossing becomes more difficult ; but he is certain to effect it without loss, and almost without danger of being overtaken in the pur- suit. Hi? freebooting habits should not be censured too THE " GRAN CHACO." 73 gravely. Many extenuating circumstances must be taken into consideration, — bis wrongs and sanguinary persecutions. It must be remembered that the hostili- ties commenced on the opposite side ; and with the In- dian the habit is not altogether indigenous, but rather the result of the principle of retaliation. He is near kindred to the Incas, — in fact, some of the Chaco tribes are remnants of the scattered Peruvian race, and he still remembers the sanguinary slaughter of his an- cestors by the Pizarros and Almagros. Therefore, using the phraseology of the French tribunals, we may say there are " extenuating circumstances in his favor." One circumstance undoubtedly speaks trumpet-tongued for the Chaco Indian ; and that is, he does not torture his cap- tives, even when white men have fallen into Ins hands! As to the captive women and children, their treatment is rather gentle than otherwise ; in fact, they are adopted into the tribe, and share, alike with the rest, the pleas- ures as well as the hardships of a savage life. When the Chaco Indian possesses homed cattle and sheep, he eats mutton and beef; but if these are want- ing, he must resort to the chase. He captures deer and ostriches by running them down with his swift steed, and piercing them with his long spear ; and occasionally he uses the bolas. For smaller game he employs the bow and arrow, and fish are also caught by shooting them with arrows. The Chaco Indian is the owner of a breed of dcgs, and large packs of these animals may be seen around his camping-ground, or following the cavalcade in its removal from plrxe to place. They are small creatures, — supposed to ae derived from a European stock, but 74 THE CENTAURS OF they are wonderfully prolific, the femah often bringing forth twelve puppies at a birth. They burrow in the ground, and subsist on the offal of the camp. They are used in running down the spotted roebuck, in hunting tin :apivara, the great ant-bear, viscachas, and other email animals. The tapir is taken in traps, and also speared, when the opportunity offers. His flesh is rel- ished by the Chaco Indian, but his hide is of more consequence, as from it bags, whips, and various other articles can be manufactured. The peccary of two spe- cies (dicotyles torquatus and collaris) is also pursued by the dogs, and speared by the hunter while pausing to bay the yelping pack ; and the great American tiger (jaguar) is killed in a like manner. The slaying of this fierce and powerful quadruped is one of the feats of the Chaco hunter, and both its skin and flesh are articles of eager demand. The latter is particularly sought for ; as by eating the flesh of so strong and courageous a creature the Indian fancies his own strength and courage will be increased. When a jaguar is killed, its carcass becomes the common property of all ; and each indi- vidual of the tribe must have his slice, or " griskin," — however small the piece may be after such multiplied subdivision ! For the same reason, the flesh of the wild boar is relished ; also that of the ant-bear — one of the most courageous of animals, — and of the tapir, on ac- count of its great strength. • The bread of the Chaco Indian is derived, as before mentioned, from several species of mimosas, called in- definitely algarobiasj and by the missionary monks known as " St. John's bread." Palms of various kind* furnish edible nuts; and there are many trees in the THE "GRAN CHACD" 75 Chaco forests that produce luscious fruits. With these the Indian varies his diet, and also with wild honey, — ■ (i most important article, for reasons alreadj assigned Ir -he Chaco there are stingless bees, of numerous dis tinct species, --a proof of the many blossoms which bloom as it were " unseen " in that flowery El\ sium. The honey of these bees — of some of the species in particular — is 'known to be of the finest and purest quality. In the Spanish settlements it commands the highest price, and is very difficult to be obtained, — for the Chaco Indian is but little given to commerce, and only occasionally brings it to market. He has but few wants to satisfy, and cares not for the tinsel of the tra- der : hence it is that most of the honey he gathers is reserved for his own use. He searches for the bees 1 nest by observing the flight of the insect, as it passes back and forward over the wild parterre ; and his keen- ness of sight — far surpassing that of a European — enables him to trace its movements in the air, and follow it to its hoard. He alleges that he could not accomplish this so well, were he encumbered with eyebrows and lashes, and offers this as one of his reasons for extract- ing these hirsute appendages. There may be something in what he says, — strange as it sounds to the ear of one who is not a bee-hunter. He finds the nest at length, — sometimes in a hollow tree, sometimes upon a branch, — the latter kind of nest being a large mass, of a substance like blotting-paper, and hanging suspended from the twigs. Sometimes he traces the insect to a subterranean dwelling ; but it must be remarked that all these art different species of bees, that build their nests and ^on Itruct the cells of their honeycombs each in its own 76 THE CENTAURS OF favorite place, and according to its own fashion. The bee-hunter cares not how — so long as he can find the nest ; though he would prefer being guided to one built upon a species of thick octagonal cactus, known as the habitat of the bee " tosimi." This preference is caused by the simple fact — that of all the honey in the Chaco. that of the bee " tosimi " is the sweetest. It is to be regretted that, with his many virtues, and his fine opportunity of exercising them, the Chaco In- dian will not consent to remain in peace and good-will with all men. It seems a necessity of his nature to have an occasional shy at some enemy, whether white or of liis own complexion. But, indeed, it would be ridicu- lous to censure him for this, since it appears also to be a vice universal among mankind ; for where is the tribe or nation, savage or civilized, who does not practise it, whenever it feels bold enough or strong enough to do so ? The Chaco Indian is not alone in his disregard of of the sixth commandment, — not the only being on earth who too frequently goes forth to battle. He has two distinct kinds of enemies, — one of Euro- pean, the other of his own race, — almost of his own kindred, you would say. But it must be remembered that there are several distinct tribes dwelling in the Chaco ; who, although presenting a certain similitude, are in many respects widely dissimilar ; and, so far from forming one nation, or living in harmonious alliance with each other, are more frequently engaged in the mpst deadly hostilities. Their wars are all conducted on horseback, — all cavalry skirmishes, — the Chaco Indian disdaining to touch the ground with his foot. Dis- mounted he would feel himself vanquished, — as much out of his element as a fish out of water ! THE "GRAN CHACO." 77 His wai weapons are of a primitive kind ; they are the bow and lance, and a species of club, known in Spanish phraselogy as the "macana." This last weapon is also found in the hands of several of the Amazonian tribes, though differing slightly in its construction. The "macana" of the Chaco Indian is a short, stout piece of heavy iron-wood, — usually a species known as the qucbracha, or " axe-breaker," which grows plentifully throughout the Paraguayan countries. Numerous spe- cies are termed " quebracha " in Spanish-American coun- tries, as there are numerous " iron-woods." That of Par- aguay, like most others that have obtained this name, is a species of ebony -wood, or lignum vitse, — in short, a true guaiacum. The wood is hard, solid, and heavy almost as metal ; and therefore juM the very stuff foi a war-club. The macana of the Chaco Indian is short, — not much over two feet in length, and is used both for striking in the hand and throwing to a distance. It is thicker, and of course heavier, at both extremities ; and the mode of grasping it is round the narrow part in the middle. The Indian youths, while training for war, practise throwing the macana, as other people play at skittles or quoits. The lazo and bolas are both in the hands of the Chaco tribes, but these contrivances are used sparingly, and more for hunting than war. They rarely trouble them- selves with them on a real war expedition. Their chief weapons against an enemy are their long lances, — for these are far the most effective arms for a man mounted on horseback. Those of the Chaco In- dian are of enormous length, their shafts being often fifteen foet from butt to barb. They use thsm also when 78 THE CENTAURS OF mounting on horseback, in a fashion peculiar to them* selves. They mount by the right side, contrary to our European mode ; nor is there the slightest resemblance in any other respect between the two fashions of getting into the saddle. With the Chaco Indian there is no put- ting toes into stirrups, — no tugging at the poor steed's withers, — no clinging or climbing into the seat. He places the butt of his lance upon the ground, grasps it a little above his head with the right hand, and then rais- ing his lithe body with an elastic spring, he drops like a cat upon the spine of his well-trained steed. A word, — a touch of his knee, or other well-understood signal, — and the animal is off like an arrow. When the Chaco Indian goes to war against the whites, his arms are those already described. He is not yet initiated into the use of guns and gunpowder, though he often experiences their deadly effects. Indeed, the won der is that he could have maintained his independence so long, with such weapons opposed to him. Gunpowder has often given cowards the victory over brave men ; but the Chaco Indian, even without gunpowder, has managed somehow or other to preserve his freedom. When he makes an expedition against the white set- tlements, he carries no shield or other defensive armor. He did so at one period of his history ; but experience has taught him that these contrivances are of little use against leaden bullets ; and he has thrown them away, taking them up again, however, when he goes to war with enemies of his own kind. In attacking a settlement or village of the whites, one of his favorite strategic plans is to set the houses on lire ; and in this he very often succeeds, — almost cer- TIIE "GRAN CHACO." 79 tainly when the thatch chances to be dry. His plan is to project an arrow with a piece of blazing cotton fas- tened near the head. For this purpose he uses the strongest kind of bow, and lying upon his back, bends it with his feet. By this means a much longer range is obtained, and the aim is of little consequence, so long as the arrow falls upon the roof a house. On going to war with a hostile tribe of his own kind and color, he equips himself in a manner altogether dif ferent. His face is then painted most frightfully, and in the most hideous designs that his imagination can suggest, while his body is almost entirely covered by a complete suit of mail. The thick hide of the tapir furnishes him with the materials for helmet, cuirass, cuisses, greaves, everything, — and underneath is a lining of jaguar-skin- Thus accoutred he is in little danger from the arrows oi the enemy, though he is also sadly encumbered in the management of his horse ; and were he upon a plunder- ing expedition against the whites, such an encumbrance would certainly bring him to grief. He knows that very well, and therefore he never goes in such guise upon any foray that is directed towards the settlements. The Chaco Indian has now been at peace with his eastern neighbors — both Spaniards and Portuguese -— for a considerable length of time ; but he still keeps up hostility with the settlements on the south, — those of Cordova and San Luis, — and often returns from these wretched provinces laden with booty. If he should chance to bring away anything that is of no use tc him, or that may appear superfluous in his savage home, *— a harp or guitar, a piece of costly furniture, or even a handsome horse, — he is not required to throw it away 80 THE CENTAURS. he knows that he can find purchasers on the other side of the river, — among the Spanish merchants of Cor* rientes or Paraguay, who are ready at any time to become the receivers of the property stolen from their kindred of the south ! Such queer three-cornered dealings are also earned on in the northern countries of Spanish America, — in the provinces of Chihuahua, New Leon, and New Mexico. They are there called " cosas de Mexico." It appears tfiey are equally " cosas de Paraguay." BOSJESMEN, OR BDSHMEN. Perhaps no race of people has more piqu'xl tho curiosity of the civilized world than those little yellow savages of South Africa, known as the Bushmen. From the first hour in which European nations became ac- quainted with their existence, a keen interest was ex- cited by the stories told of their peculiar character and habits ; and although they have been visited by many travellers, and many descriptions have been given of them, it is but truth to say, that the interest in them has not yet abated, and the Bushmen of Africa are al- most as great a curiosity at this hour as they were when Di Gama first doubled the Cape. Indeed, there is no reason why this should not be, for the habits and personal appearance of these savages are just now as they were then, and our familiarity with them is not much greater. Whatever has been added to our knowledge of their character, has tended rather to increase than diminish our curiosity. At first the tales related of them were supposed to be filled with wilful exaggerations, and the early travellers were accused of dealing too much in the marvellous. This is a V3ry common accusation brought against the 82 BOS.TESMEN, OR early travellers ; and in some instances it is a just one But in regard to the accounts given of the Bushmen and their habits there has been far less exaggeration than might be supposed ; and the more insight we ob- tain into their peculiar customs and modes of subsistence, the more do we become satisfied that almost everything alleged of them is true. In fact, il would be difficult for the most inventive genius to contrive a fanciful ac- count, that would be much more curious or interesting than the real and bond fide truth that can be told about this most peculiar people. Where do the Bushmen dwell ? what is their coun- try? These are questions not so easily answered, as in reality they are not supposed to possess any country at all, any more than the wild animals amidst which they roam, and upon whom they prey. There is no Bushman's country upon the map, though several spots in Southern Africa have at times received this desig- nation. It is not possible, therefore, to delineate the boundaries of their country, since it has no boundaries, any more than that of the wandering Gypsies of Europe. If the Bushmen, however, have no country in the proper sense of the word, they have a " range," and one of the most extensive character — since it covers the whole southern portion of the African cont merit, from the Cape of Good Hope to the twentieth degree of south latitude, extending east and west from the country of the Caffres to the Atlantic Ocean. Until lately it was be- lieved that the Bushman-range did not extend far to the north of the Orange river; but this has proved an er- roneous idea. They have recently u turned up M in the land o^ the Dammaras, and also in the great Kalahari BUSHMEN. 83 desert, hundreds of miles north from the Orange river and it is not certain that they do not range still nearer to the equatorial line — though it maybe remarked that the country in that direction does noc favor the suppo- sition, not being of the peculiar nature of a Bushman's country. The Bushman requires a desert for his dwell- ing-place. It is an absolute necessity of his nature, as it is to the os'rich and many species of animals ; and north of the twentieth degree of latitude, South Africa does not appear to be of this character. The heroic Livingstone has dispelled the long-cherished illusion of the Geography about the " Great-sanded level" of these interior regions ; and, instead, disclosed to the world a fertile land, well watered, and covered with a profuse and luxuriant vegetation. In such a land there will be no Bushmen. The limits we have allowed them, however, are suffi- ciently large, — fifteen degrees of latitude, and an equally extensive range from east to west. It must not be sup- posed, however, that they 'populate this vast territory. On the contrary, they are only distributed over it in spots , in little communities, that have no relationship or ' connection with one another, but are separated by wide intervals, sometimes of hundreds of miles in extent. It is only in the desert tracts of South Africa that the Bushmen exist, — » in the karoos, and treeless, waterless plains — among the barren ridges and rocky defiles — in the ravines formed by the beds of dried-up rivers — in situations so sterile, so remote, so wild and inhospitable as to offer a home to no other human being save the Bushman himself. If we state more particularly *he localities where the 84 BOSJESMEN, OR haunts of the Bushman are to be found, we m,ty specify the barren lands on both sides of the Orange Liver, — including most of its head-waters, and down to its mouth, — and also the Great Kalahari desert. Through all tliis extensive region the kraals of the Bushmen may be encountered. At one time they were common enough within the limits of the Cape colony itself, and some half-caste remnants still exist in the more remote dis- tricts ; but the cruel persecution of the boers has had the effect of extirpating these unfortunate savages ; and, like the elephant, the ostrich, and the eland, the true wild Bushman is now only to be met with beyond the fron- tiers of the colony. About the origin of the Bushmen we can offer no opinion. They are generally considered as a branch of the great Hottentot family ; but this theory is far from being an established fact. When South Africa was first discovered and colonized, both Hottentots and Bushmen were found there, differing from each other just as they differ at this day ; and though there are some striking points of resemblance between them, there are also points of dissimilarity that are equally as strik- ing, if we regard the two people as one. In personal appearance there is a certain general likeness : that is, both are woolly-haired, and both have a Chinese cast of features, especially in the form and expression of the eye. Their color too is nearly the same; but, on the other hand, the Hottentots are larger than the Bushmen. It is not in their persons, however, that the most essential points of dissimilarity are to be looked for, but rather in their mental characters ; and here we observe distinc- tions so marked and antithetical, that it is difficult to BUSHMEN. 85 reconcile them with the fact that these two, people are of one race. Whether a different habit of life has pro- duced this distinctive character, or whether it has in- fluenced the habits of life, are questions not easily an- swered. We only know that a strange anomaly exists — the anomaly of two people being personally alike — that is, possessing physical characteristics that seem to prove them of the same race, while intellectually, as we shall presently see, they have scarce one character in common. The slight resemblance that exists between the languages of the two is not to be regarded as a proof of their common origin. It only shows that they have long lived in juxtaposition, or contiguous to each other ; a fact which cannot be denied. In giving a more particular description of the Bush- man, it will be seen in what respect he resembles the true Hottentot, and in what he differs from him, both physically and mentally, and this description may now be given. The Bushman is the smallest man with whom we are acquainted ; and if the terms " dwarf " and " pigmy " may be applied to any race of human beings, the South- Afri- can Bushmen presents the fairest claim to these titles. He stands only 4 feet 6 inches upon his naked soles — never more than 4 feet 9, and not unfrequently is he encountered of still less height — even so diminutive as 4 ieet 2. His wife is of still shorter stature, and this Lilliputian lady is often the mother of children when the crown of her head is just 3 feet 9 inches above the sole* of her feet. It has been a very common thing to con- tradict the assertion that these people are such pigmies in stature, and even Dr. Livingstone has done so in hi# 8G BOSJESMEN, OR late magnificent work. The doctor states, very jocosely, that they are "not dwarfish — that the specimens brought to Europe have been selected, like costermongcrs' dogs, for their extreme ugliness." But the doctor forgets that it is not from " the speci- mens brought to Europe " that the above standard of the Bushman's height has been derived, but from the testi- mony of numerous travellers — many of them as trust- worthy as the doctor himself — from actual measurements made by them upon the spot. It is hardly to be believed that such men as Sparmann and Burchell, Barrow and Lichtenstein, Harris, Campbell, Patterson, arid a dozen others that might be mentioned, should all give an erro- neous testimony on this subject. These travellers have differed notoriously on other points, but in this they all agree, that a Bushman of five feet in height is a tall man in his tribe. Dr. Livingstone speaks of Bushmen " six feet high," and these are the tribes lately discovered liv- ing so far north as the Lake Nagami. It is doubtful whether these are Bushmen at all. Indeed, the descrip- tion given by the doctor, not only of their height and the color of their skin, but also some hints about their intel- lectual character, would lead to the belief that he has mistaken some other people for Bushmen. It must be remembered that the experience of this great traveller has been chiefty among the Bechuana tribes, and his knowledge of the Bushman proper does not appear to be either accurate or extensive. No man is expected to know everybody ; and amid the profusion of new facts, which the doctor has so liberally laid before the world, it would be strange if a few inaccuracies should not occur Perhaps we should have more confidence if IhLi BUSHMEN. 87 was the only one we are enabled to detect ; but the doc- tor also denies that there is anything either terrific or majestic in the " roaring of the lion." Thus speaks he : u The same feeling which has induced the modern painter to caricature the lion has led the sentimentalist to con* sidcr the lion's roar as the most terrific of all earthly sounds. We hear of the ' majestic roar of the king of beasts.' To talk of the majestic roar of the lion is mere majestic twaddle." The doctor is certainly in error -here. Does he sup- pose that any one is ignorant of the character of the lion's roar? Does he fancy that no one has ever heard it but himself? If it be necessary to go to South Africa to take the true measure of a Bushman, it is not neces- sary to make that long journey in order to obtain a cor- rect idea of the compass of the lion's voice. We can hear it at home in all its modulations ; and any on«r wno has ever visited the Zoological Gardens in Regent's Park — nay, any one who chances to live within half a mile of that magnificent menagerie — will be very much disposed to doubt the correctness of the doctor's asser- tion. If there be a sound upon the earth above all others " majestic," a noise above all others " terrific," it is certainly the roar of the lion. Ask Albert Terrace and St. John's Wood ! But let us not be too severe upon the doctor. The Irorld i3 indebted to him much more than to any other modern traveller, and all great men indulge occasion- ally in the .luxury of an eccentric opinion. We have brought the point forward here for a special purpose, — to illustrate a too much neglected truth. Error is not always on the side of exaggeration ; but is sometimes 88 BOSJESMEN, OR also found in the opposite extreme of a too-s j*ueaiuish moderation. We find the learned Professor Lichtenstein ridiculing poor old Hernandez, the natural historian, of Mexico, for having given a description of certain fabu- lous animals — fabulous, he terms them, because to him they were odd and unknown. But it turns out that the old author was right, and the animals exist! How many similar misconceptions might be recorded of the BufFons, and other closet philosophers — urged," too, with the most bitter zeal ! Incredulity carried too far is but another form of credulity. But to return to our proper theme, and complete the portrait of the Bushman. We have given his height. It is in tolerable proportion to his other dimensions. When young, he appears stout enough ; but this is only when a mere boy. At the age of sixteen he has reached all the manhood he is ever destined to attain ; and then his flesh disappears; his body assumes a meagre outline; his arms and limbs grow thin ; the calf disappears from his legs ; the plumpness from his cheeks ; and altogether he becomes as wretched-looking an object as it is possi- ble to conceive in human shape. Older, his skin grows dry, corrugated, and scaly ; his bones protrude ; and his knee, elbow, and ankle-joints appear like horny knobs placed at the ends of what more resemble long straight sticks than the arms and limbs of a human being. The color of this creature may be designated a yellow* brown, though it is not easy to determine it to a shade. The Bushman appears darker than he really is ; since his skin serves him for a towel, and every species of dirt that discommodes his fingers he gets rid of by wip- ing it off on his arms, sides, or breast. The remit i^ BUSHMEN. 89 that his whole body is usually coated over with a stratum of grease and filth, which has led to the belief that he regularly anoints himself — a custom common among many savage tribes. This, however, the Bushman does not do : the smearing toilet is merely occasional or ac- cidental, and consists simply in the fat of whatever flesh he lias been eating being transferred from his fingers to the cuticle of his body. This is never washed off again — for water never touches the Bushman's hide. Such a use of water is entirely unknown to him, not even for washing his face. Should he have occasion to cleanse his hands — which the handling of gum or some like substance sometimes compels him to do — he performs the operation, not with soap and water, but with the dry dung of cattle or some wild animal. A little rubbing of this upon his skin is all the purification the Bushman believes to be needed. Of course, the dirt darkens his complexion ; but he has the vanity at times to brighten it up — not by making it whiter — but rather a brick-red. A little ochreous earth produces the color he requires ; and with this he smears his body all over — not excepting even the crown of his head, and the scant stock of wool that covers it. Bushmen have been washed. It requires some scruV bing, and a plentiful application either of soda or soap, to reach the true skin and bring out the natural color; but the experiment has been made, and the result proves that the Bushman is not so black as, under ordinary cir- cumstances, he appears. A yellow hue shines through the epidermis, somewhat like the color of the Chinese, or a European in the worst stage of jaundice — th 3 (ye 90 BOSJESMLN, OR only not having that complexion. Indeed, the features of the Bushman, as well as the Hottentot, bear a strong similarity to those of the Chinese, and the Bushman's eye is essentially of the Mongolian type. His hair, however, is entirely of another character. Instead of being long, straight, and lank, it is short, crisp, and curly, — in reality, wool. Its scantiness is a character- istic ; and in this respect the Bushman differs from the woolly-haired tribes both of Africa and Australasia. These generally have " fleeces " in profusion, whereas both Hottentot and Bushman have not enough to half cover their scalps ; and between the little knot-like 'kinks" there are wide spaces without a single hair apon them. The Bushman's " wool " is naturally black, but red ochre and the sun soon convert the color into a burnt reddish hue. The Bushman has no beard, or other hairy encum- brances, Were they to grow, he would root them out as useless inconveniences. He has a low-bridged nose, with wide flattened nostrils ; an eye that appears a mere flit between the eyelids ; a pair of high cheek-bones, and a receding forehead. His lips are not thick, as in the negro, and he is furnished with a set of fine white teeth, which, as he grows older, do not decay, but pre- sent the singular phenomenon of being regularly worn down to the stumps — as occurs to the teeth of sheep and other ruminant animals. Notwithstanding the small stature of the B islunan, Lis frame is wiry and capable of great endurance. He is also as agile as an antelope. From the description above given, it will be inferred that the Bushman is no beauty. Neither is the Bush- BUSHMEN. 01 ivoraan; but, on the contrary, both having passed the period of youth, become absolutely ugly, — the woman, if possible, more so than the man. And yet, strange to say, many of the Bush-girls, when young, have a cast of prettiness almost amounting to beauty. It is difficult to tell in what this beauty con- sists. Something, perhaps, in the expression of the oblique almond-shaped eye, ana the small well-formed mouth and lips, with the shining white teeth. Their limbs, too, at this early age, are often well rounded ; and many of them exhibit forms that might serve as models for a .sculptor. Their feet are especially well- shaped, and, in point of size, they are by far the small- est in the world. Had the Chinese ladies been gifted by nature with such little feet, they might have been spared the torture of compressing them. The foot, of a Bushwoman rarely measures so much as six inches in length ; and full-grown girls have been seen, whose feet, submitted to the test of an actual measurement, proved but a very lirtie over four inches ! Intellectually, the Bushman does not rank so low as is generally believed. He has a quick, cheerful mind, that appears ever on the alert, — as may be judged by the constant play of his little piercing black eye, — and though he does not always display much skill in the manufacture of his weapons, he can do so if he pleases. Some tribes construct their bows, arrows, fish- baskets, and other implements and utensils with admi- rable ingenuity ; but in general the Bushman takes no pride in fancy weapons. He prefers having them effec- tive, and to this end he gives proof of his skill in the manufacture of most deadly poisons with which to an tint 92 BOSJESMEN, OK ois arrow,;. Furthermore, he is ever active and ready for action ; and in this his mind is in complete contrast with that of the Hottentot, with whom indolence is a predominant and well-marked characteristic. The Bush- man, on the contrary, is always on the qui vive ; alwa) r s ready to he doing where there is anything to do ; and there is not much opportunity for him to he idle> as he rarely ever knows where the next meal is to come from. The ingenuity which he displays in the capture of vari- ous kinds of game, — far exceeding that of other hunting tribes of Africa, — as also the cunning exhibited by him while engaged in cattle-stealing and other plundering forays, prove an intellectual capacity more than pro- portioned to his diminutive body ; and, in short, in nearly every mental characteristic does he differ from the supposed cognate race — the Hottentot. It would be hardly just to give the Bushman a char- acter for high courage ; but, on the other hand, it would be as unjust to charge him with cowardice. Small as he is, he shows plenty of " pluck," and when brought to bay, his motto is, " No surrender." He will light to the death, discharging his poisoned arrows as long as he is able to bend a bow. Indeed, he has generally been treated to shooting, or clubbing to death, wher- ever and whenever caught, and he knows nothing of quarter. Just as a badger he ends his life, — his last struggle being an attempt to do injury to his assailant This trait in his character has, no doubt, been strength- ened by the inhuman treatment that, for a century, he has been receiving from the brutal boers of the colonial 5 on tier. The costume of the Bushman is of the most primitive BUSHMEN. D3 character, — differing only from that worn by our firai parents, in that the fig-leaf used by the men is a patch of jackal-skin, and that of the women a sort of fringe or bunch of leather thongs, suspended around the waist by a strap, and hanging down to the knees. It is in reality a little apron of dressed skin ; or, to speak more accu- rately, two of them, one above the other, both cut into narrow strips or thongs, from below the waist downward. Other clothing than this they have none, if we except a little skin kaross, or cloak, which is worn over their shoulders ; — that of the women being provided with a bag or hood at the top, that answers the naked " piccaninny " for a nest or cradle. Sandals protect their feet from the sharp stones, and these are of the rudest description, — merely a piece of the thick hide cut a little longer and broader than the soles of the feat, and fastened at the toes and round the ankles by thongs of sinews. An attempt at ornament is displayed in a leathern skullcap,, or more commonly a circlet around the head, upon which are sewed a number of " cowries," or small shells of the Cyprea moneta. It is diificult to say where these shells are procured, — as they are not the product of the Bushman's country, but are only found on the far shores of the Indian Ocean. Most probably he obtains them by barter, and after they have passed through many hands ; but they must cost the Bushman dear, as he sets the highest value upon them. Other ornaments consist of old brass or copper buttons, attached to the little curls of his woolly hair ; and, among the women, strings of little pieces of ostrich egg-shells, fashioned to resemble beads ; besides a per- fect load of leathern bracelets on the arms, and a like 04 BOSJESMEN, OR profusion of similar circlets on the limbs, often reaching from the knee to the ankle-joint. Red ochre over the face and hair is the fashionable toilette, and a perfumery is obtained by rubbing the skin with the powdered leaves of the " buku " plant, a species of diosma. According to a quaint old writer, this causes them to " stink like a poppy," and would be highly objectionable, were it not preferable to the odor which they have without it. They do not tattoo, nor yet perforate the ears, lips, or nose, — practices so common aniDng savage tribes. Some instances of nose-piercing have been observed, with the usual appendage of a piece of wood or porcupine's quill inserted in the septum, but this is a custom rather of the Caffres than Bushmen. Among the latter it is rare. A grand ornament is obtained by smearing the face and head with a shining micaceous paste, which is procured from a cave in one particular part of the Bushman's range ; but this, being a " far-fetched " article, is pro- portionably scarce and dear. It is only a fine belle who can afford to give herself a coat of blink-slip, — as this sparkling pigment is called by the colonists. Many of the women, and men as well, carry in their hands the bushy tail of a jackal. The purpose is to fan off the flies, and serve also as a <; wipe," to disembarrass their bodies of perspiration when the weather chances to be over hot. The domicile of the Bushman next merits description. It is quite as simple and primitive as his dress, and gives him about equal trouble in its construction. If a cave or cleft can be found in the rocks, of sufficient capacity to admit his own body and those of his family BUSHMEN. 05 »— never a very large one — he builds no house. The cave contents him, be it ever so tight a squeeze. If there be no cave handy, an overhanging rock will an- swer equally as well. He regards not the open sides, nor the draughts. It is only the rain which he does net relish ; and any sort of a shed, that will shelter him from that, will serve him for a dwelling. If neither cave, crevice, nor impending cliff can be found in the neigh- borhood, he then resorts to the alternative of house- building; and his style of architecture does not differ greatly from that of the orang-outang. A bush is chosen that grows near to two or three others, — the brauches of all meeting in a common centre. Of these branches the builder takes advantage, fastening them together at the ends, and wattling some into the others. Over this framework a quantity of grass is scattered in such a fashion as to cast off a good shower of rain, and then the " carcass " of the building is considered complete. The inside work remains yet to be done, and that is next set about. A large roundish or oblong hole is scraped out in the middle of the floor. It is made wide enough and deep enough to hold the bodies of three or four Bush- people, though a single large Caffre or Dutchman would scarcely find room in it. Into this hole is flung a quantity of dry grass, and arranged so as to present the appearance of a gigantic nest. This nest, or lair, be- comes the bed of the Bushman, his wife, or wives, — for he frequently keeps two, — and the other members of his family. Coiled together like monkeys, and covered with their skin karosses, they all sleep in it, — whether " sweetly ,; or " soundly," I shall not take upon me to determine. 96 BOSJESMEN, OP. It is supposed to be this fashion of literally " sleeping in the bush," as also the mode by which he skulks and hides among bushes, — invariably ta&ing to them when oursued, — that has given origin to the name Bushman, ^r Bosjesman, as it is in the language of the colonial Dutch. This derivation is probable enough, and no Setter has been offered. The Bushman sometimes constructs himself a more elaborate dwelling ; that is, some Bushmen ; — for it should be remarked that there are a great many tribes or communities of these people, ajid they are not all so very low in the scale of civilization. None, how- ever, ever arrive at the building of a house, — not even a hut. A tent is their highest effort in the building line, and that is of the rudest description, scarce deserv- ing the name. Its covering is a mat, which they weave out of a species of rush that grows along some of the desert streams ; and in the fabrication of the covering they display far more ingenuity than in the planning or construction of the tent itself. The mat, in fact, is simply laid over two poles, that are bent into the form of an arch, by having both ends stuck into the ground. A second piece of matting closes up one end : and the other, left open, serves for the entrance. As a door is not deemed necessary, no further construction is re- quired, and the tent is "pitched" complete. It only remains to scoop out the sand, and make the nest as already described. It is said that the Goths drew their ideas of archi- tecture from the aisles of the cak forest ; the Chinese from their Mongolian tents ; and the Egyptians from their caves in the rocks. Beyond a doubt, the Bush- man has borrowed his from the nest of the ostrich 1 BUSHMEN. 07 It now becomes necessary to inquire how the Bush* man spends his time ? how he obtains subsistence ? and what is the nature of his food ? All these questions can be answered, though at first it may appear difficult to answer them. Dwelling, as he always does, in the very heart of the desert, remote from forests that might fur- nish him with some sort of food — trees that might j ield fruit, — far away from a fertile soil, with no knowledge of agriculture, even if it were near, — with no flocks or herds ; neither sheep, cattle, horses, nor swine, — no domestic animals but his lean, diminutive dogs, — how does this Bushman procure enough to eat ? What are his sources of supply ? We shall see. Being neither a grazier nor a farmer, he has other means of subsistence, — though it must be confessed that they are of a precarious character, and often during his life does the Bushman find himself on the very threshold of starvation. This, however, results less from the parsimony of Nature than the Bushman's own improvident habits, — a trait in his character which is, perhaps, more strongly developed in him than any other. We shall have occasion to refer to it presently. His first and chief mode of procuring his food is by the chase : for, although he is surrounded by the sterile wilderness, he is not the only animated being who has chosen the desert for his home. Several species of birds — one the largest of all — and quadrupeds, share frith the Bushman the solitude and safety of this deso- late region. The rhinoceros can dwell there ; and in numerous streams are found the huge hippopotami ; whilst quaggas, zebras, and several species of antelope frequent the desert plains as their favorite a stamping " i)S BOSJESMEN, OR ground. Some of these animals can live almof.t vyithout water ; but when they do require it, what to th mi is a gallop of fifty mil* s to some well-known u vley " or pool ? It will be seen, therefore, that the desert has its numer- ous denizens. All these are objects of the Bushman's pursuit, who follows them with incessant pertinacity — s if he were a beast of prey, furnished by Nature with le most carnivorous propensities. In the capture of these animals he displays an almost credible dexterity and cunning. His mode of ap- proaching the sly ostrich, by disguising himself in the skin of one of these birds, is so well known that I need not describe it here ; but the ruses he adopts for captur- ing or killing other sorts of game are many of them equally ingenious. The pit-trap is one of his favorite contrivances ; and this, too, has been often described, — but often very erroneously. The pit is not a large hollow, — as is usually asserted, — but rather of dimen- sions proportioned to the size of the animal that is ex- pected to fall into it. For game like the rhinoceros or eland antelope, it is dug of six feet in length and three in width at the top ; gradually narrowing to the bottom, where it ends in a trench of only twelve inches broad. Six or seven feet is considered deep enough ; and the animal, once into it, gets so wedged at the narrow bofc- torn part as to be unable to make use of its legs for the purpose of springing out again. Sometimes a sharp stake or two are used, with the view of impaling the victim ; but this plan is not always adopted. There is not much danger of a quadruped that drops in ever getting out again, till he is dragged out by the Bushman in the shape of a carcass. BUSHMEN. 99 The Bushman's ingenuity does not end here. Be« sides the construction of the trap, it is necessary the game should be guided into it. Were this not done, the pit might remain a long time empty, and, as a necessary consequence, so too might the belly of the Bushman. In the wide plain few of the gregarious animals have a path which they follow habitually ; only where there iff a pool may such beaten trails be found, and of these the Bushman also avails himself; but they are not enough. Some artificial means must be used to make the trapa pay — for they are not constructed without much labor and patience. The plan adopted by the Bushman to accomplish this exhibits some points of originality. He first chooses a part of the plain which lies between two mountains. No matter if these be distant from each other : a mile, or even two, will not deter the Bushman from his design. By the help of his whole tribe — men, women, and children — he constructs a fence from one mountain to the other. The material used is whatever may be most ready to the hand : stones, sods, brush, or dead timber, if this be convenient. No matter how rude the fence : it need not either be very high. He leaves several gaps in it ; and the wild animals, however easily they might leap over such a puny barrier, will, in their ordinary way, prefer to walk leisurely througk the gaps. In each of these, however, there is a danger ous hole — dangerous from its depth as well as from t) i cunning way in which it is concealed from the view - in short, in each gap there is a pit-fall. No one- at least no animal except the elephant — would ever dus* pect its presence ; the grass seems to grow over it, and the sand lies unturned, just as elsewhere upon the plain 100 BOSJESMEN, OR What quadriped could detect the cheat ? Not any on« except the sagacious elephant. The stupid eland tum- bles through ; the gemsbok goes under ; and the rhi- noceros rushes into it as if destined to destruction. V he Bushman sees this from his elevated perch, glides for- ward over the ground, and spears the struggling victim with his poisoned assagai. Besides the above method of capturing game the Bushman also uses the bow and arrows. This is a weapon in which he is greatly skilled ; and although both bow and arrows are as tiny as if intended for children's toys, they are among the deadliest of weapons, their fatal effect lies not in the size of the wound they are capable of inflicting, but in the peculiar mode in which the barbs of the arrows are prepared. I need hardly add that they are dipped in poison ; — for who has not heard of the poisoned arrows of the African Bushmen ? Both bow and arrows are usually rude enough in their construction, and would appear but a trumpery affair, were it not for a knowledge of their effects. The bow is a mere round stick, about three feet long, and slightly bent by means of its string of twisted sinews. The arrows are mere reeds, tipped with pieces of bone, with a split ostrich-quill lapped behind the head, and answering for a barb. This arrow the Bushman can shoot with tolerable certainty to a distance of a hundred yards, and he can even project it farther by giving a slight elevation to his aim. It signifies not whether the force with which it strikes the object be ever so slight, if it onlj makes an entrance. Even a scratch from itg point will sometimes prove fatal. EOSSMEN. 101 Of course .he danger dwells altogether in the poison* Were it not for that, the Bushman, from his dwarfish stature and pigmy strength, would be a harmless creature indeed. The poison he well knows how to prepare, and he can make it of the most " potent spell," when the " ma- terials " are within his reach. For this purpose he makes use of both vegetable and animal substances, and a mineral is also employed ; but the last is not a poison, . and is only used to give consistency to the liquid, so that it may the better adhere to the arrow. The vegetable substances are of various kinds. Some are botanically known : the bulb of Amaryllis disticha, — the gum of a Euphorbia, — the sap of a species of sumac (Bhus),—* and the nuts of a shrubby plant, by the colonists called Woolf-gift (Wolf-poison). , The animal substance is the fluid found in the fangs of venomous serpents, several species of which serve the purpose of the Bushman : as the little " Horned Snake," — so called from the scales rising prominently over its eyes ; the " Yellow Snake/' or South African Cobra (Naga haje) ; the " Puff Adder," and others. From all these he obtains the ingredients of his deadly ointment, and mixes them, not all together ; for he cannot always procure them all in any one region of the country in which he dwells. He makes his poison, also, of different degrees of potency, according to the purpose for which he intends it ; whether for hunting or war. With sixty or seventy little arrows, well imbued with this fatal mixture, and carefully placed in his quiver of tree-bark or skin, — or, what is not uncommon, stuck like a coro- aet around his head, — he sallies forth, ready to rtejd 102 BOSJESMEN, OR destruction either to game, animals, or o human fin©* mies. Of these last he has no lack. Eveiy man, not a Bushman, he deems his enemy ; and he has some reason for thinking so. Truly may it be said of him, as of Ishmael, that his " hand is against every man, and every man's hand against him ; " and such has been his un- happy history for ages. Not alone have the boers been his pursuers and oppressors, but all others upon his borders who are strong enough to attack him, — colo- nists, Caffres, and Bechuanas, all alike, — not even ex- cepting his supposed kindred, the Hottentots. Not only does no fellow-feeling exist between Bushman and Hot- tentot, but, strange to say, they hate each other with the most rancorous hatred. The Bushman will plunder a Namaqua Hottentot, a Griqua, or a Gonaqua, — plunder and murder him with as much ruthlessness, or even more, than he would the hated CafFre or boer. All are alike his enemies, — all to be plundered and massacred, whenever met, and the thing appears possible. We are speaking of plunder. This is another source of supply to the Bushman, though one that is not always to be depended upon. It is his most dangerous method of obtaining a livelihood, and often costs him his life. He only resorts to it when all other resources fail him, and food is no longer to be obtained by the chase. He makes an expedition into the settlements, — either of the frontier boers, Caffres, or Hottentots, — whichever chance to live most convenient to his haunts. The ex- pedition, of course, is by night, and conducted, not as an open foray, but in secret, and by stealth. The cattle are stolen, not reeved, and driven off while the owner and his people are asleep. BUSHMEN. 103 In the morning, or as soon as the loss is discovered, a pursuit is at once set on foot. A dozen men, mounted and armed with long muskets (roers), take the spoor of the spoilers, and follow it as fast as their horses will carry them. A dozen boers, or even half that number, is considered a match for a whole tribe of Bushmen, in any fight which may occur in the open plain, as the boers make use of their long-range guns at such a dis- tance that the Bushmen are shot down without being able to use their poisoned arrows ; and if the thieves have the fortune to be overtaken before they have got far into the desert, they stand a good chance of being terribly chastised. There is no quarter shown them. Such a thing as mercy is never dreamt of, — no sparing of lives any more than if they were a pack of hyenas. The Bush- men may escape to the rocks, such of them as are not hit by the bullets ; and there the boers know it would be idle to follow them. Like the klipspringer antelope, the little savages can bound from rock to rock, and cliff to cliff, or hide like partridges among crevices, where neither man nor horse can pursue them. Even upon the level plain — if it chance to be stony or intersected with breaks and ravines — a horseman would endeavor to overtake them in vain, for these yellow imps are as swift as ostriches. When the spoilers scatter thus, the boer may recover his cattle, but in what condition? That he has sur- mised already, without going among the herd. He does not expect to drive home one half of them ; perhaps not one head. On reaching the flock he finds there is not one without a wound of some kind or other : a gash in 104 BOSJESMEN, OR the flank, the cut of a knife, the stab of an assagai, or a poisoned arrow — intended for the boer himself — stick ing between the ribs. This is the sad spectacle thai meets his eyes ; but he never reflects that it is the result of his own cruelty, — he never regards it in the light of retribution. Had he not first hunted the Bushman to make him a slave, to make bondsmen and bondsmaids of his sons and daughters, to submit them to the ca- price and tyranny of his great, strapping frau, perhaps his cattle would have been browsing quietly in his fields. The poor Bushman, in attempting to take them, followed but his instincts of hunger : in yielding them up he obeyed but the promptings of revenge. It is not always that the Bushman is thus overtaken. He frequently succeeds in carrying the whole herd to his desert fastness ; -and the skill which he exhibits in getting them there is perfectly surprising. The cattle themselves are more afraid of him than of a wild beast, and run at his approach ;, but the Bushman, swifter than they, can glide all around them, and keep them moving at a rapid rate. He uses stratagem also to obstruct or baffle the pur- suit. The route he takes is through the driest part of the desert, — if possible, where water does not exist at all. The cattle suffer from thirst, and bellow from the pain ; but the Bushman cares not .or that, so long as he is himself served. But how is he served ? There is no water, and a Bushman can no more go without drink- ing than a boer : how then does he provide for himself on these long expeditions ? All has been pre-arranged. Whi.e off to the settle- ments-, the Bushman's wife has leen busy. The whola BUSHMEN. 105 kraal of women — young and old — have made an ex- cursion half-way across the desert, each carrying ostrich egg-shells, as much as her kaross will hold, each shell full of water. These have been deposited at intervals along the route in secret spots known by marks to the Bushmen, and this accomplished the women return homo again. In this way the plunderer obtains his supply of water, and thus is he enabled to continue his journey over I he arid Karroo. The pursuers become appalled. They are suffering from thirst — their horses sinking under them. Perhaps they have lost their way ? It would be madness to pro- ceed further. " Let the cattle go this time ! " and with this disheartening reflection they give up the pursuit, turn the heads of their horses, and ride homeward. There is a feast at the Bushman's kraal — and such a feast ! not one ox is slaughtered, but a score of them all at once. They kill them, as if from very wantonness ; and they no longer eat, but raven on the flesh. For days the feasting is kept up almost continuou ,iy, — even at night they must wake up to have a midnight meal ! and thus runs the tale, till every ox has been eaten. They have not the slightest idea of a provision for the fu- ture ; even the lower animals seem wiser in this respect. They do not think of keeping a few of the plundered cattle at pasture to serve them for a subsequent occasion. They give the poor brutes neither food nor drink ; but having penned them up in some defile of the rocks, leave them to moan and bellow, to drop down and die. On goes the feasting, till all are finished ; and even if the flesh has turned putrid, this forms not the slightest tbjection : it is eaten all the same. 106 BOSJESMKN, OR The kraal now exhibits an altered spectacle. The starved meagre wretches, who were seen flitting among its tents but a week ago, have all disappeared. ^Plump bodies and distended abdomens are the order of the day ; and the profile of the Bushwoman, taken from the neck to the knees, now exhibits the outline of the letter S. The little imps leap about, tearing raw flesh, — their yellow cheeks besmeared with blood, — and the lean curs seem to have been exchanged for a pack of fat, petted poodles. But this scene must some time come to an end, and at length it does end. All the flesh is exhausted, and the bones picked clean. A complete reaction comes over the spirit of the Bushman. He falls into a state of languor, — the only time when he knows such a feeling, — and he keeps his kraal, and remains idle for clays. Often he sleeps for twenty-four hours at a time, and wakes only to go to sleep again. He need not rouse himself with the idea of getting something to eat : there is not a morsel in the whole kraal, and he knows it. He lies still, there- fore, — weakened with hunger, and overcome with the drowsiness of a terrible lassitude. Fortunate for him, while in this state, if those bold vultures — attracted by the debris of his feast, and now high wheeling in the air — be not perceived from afar ; fortunate if they do not discover the whereabouts of his kraal to the vengeful pursuer. If they should do so, he has made his last foray and his last feast. When the absolute danger of starvation at length compels our Bushman to bestir himself, he seems to recover a little of his energy, and once more takes to hunting, or, if near a stream, endeavors to catch a few BUSHME1S. 107 fish. Should both these resources fail, he has another, — without which he would most certainly starve, — and perhaps this may be considered his most important source of supply, since it is the most constant, and can be depended on at nearly all seasons of the year. Weak- ened with hunger, then, and scarce equal to any severei labor, he goes out hunting — this time insects, not quad- rupeds. With a stout stick inserted into a stone at cne end and pointed at the other, he proceeds to the nests of the white ants (termites), and using the point of the stick, — the stone serving by its weight to aid the force of the blow, — he breaks open the hard, gummy clay of which the hillock is formed. Unless the aard-vark anc the 'pangolin — two very different kinds of ant-eaters — have been there before him, he finds the chambers filled with the eggs of the ants, the insects themselves, and perhaps large quantities of their larvce. All are equally secured by the Bushman, and either devoured on the spot, or collected into a skin bag, and carried back to his kraal. He hunts also another species of ants that do not build nests or " hillocks," but bring forth their young ui hol- lows under the ground. These make long galleries or covered ways just under the surface, and at certain pe- riods — which the Bushman knows by unmistakable signs — they become very active, and traverse these underground galleries in thousands. If the passages were to be opened above, the ants would soon make off to their caves, and but a very few could be captured. The Bushman, knowing this, adopts a stratagem. With the stick already mentioned he pierces holes of a good depth down; and works the stick about, until the sides of th€ 108 BOSJESMEN, OR holes are smooth and even. These he in tends shall servo him as pitfalls ; and they are therefore made in the cov- ered ways along which the insects are passing. The result is, that the little creatures, not suspecting the ex^ istence of these deep wells, tumble head foremost into them, and are unable to mount up the steep smooth sides again, so that in a few minutes the hole will be filled with aits, which the Bushman scoops out at his leisure. Another source of supply which he has, and also a pretty constant one, consists of various roots of th6 tuberous kind, but more especially bulbous roots, which grow in the desert. They are several species of Ixias and MesembryanthemuniSj — some of them producing bulbs of a large size, and deeply buried underground. Half the Bushman's and Bushwoman's time is occupied in digging for these roots ; and the spade employed is the stone-headed staff already described. Ostrich eggs also furnish the Bushman with many a meal ; and the hu^e shells of these eggs serve him for water-vessels, cups, and dishes. He is exceedingly ex- pert in tracking up the ostrich, and discovering its nest. Sometimes he finds a nest in the absence of the birds ; and in a case of this kind he pursues a course of con- duct that is peculiarly Bushman. Having removed all the eggs to a distance, and concealed them under some bush, he returns to the nest and ensconces himself in it. His diminutive body, when close squatted, cannot be perceived from a distance, especially when there are a few bushes around the nest, as there usually are. Thus concealed he awaits the return of the birds, holding his bow and poisoned arrows ready to salute them as t^ion as they come within range. By this ruse he is BUSHMEN. 109 almost certain of killing either the cock or hen, and not unfrequently both — when they do not return together. Lizards and land-tortoises often furnish the Bushman with a meal ; arid the shell of the latter serves him also for a dish ; but his period of greatest plenty is when the locusts appear. Then, indeed, the Bushman is no longer in want of a meal ; and while these creatures re- main with him, he knows no hunger. He grows fat in a trice, and his curs keep pace with him — for they too greedily devour the locusts. Were the locusts a con- stant, or even an annual visitor, the Bushman would be a rich man — at all events his wants would be amply supplied. Unfortunately for him, but fortunately for everybody else, these terrible destroyers of vegetation only come now and then — several years often inter- vening between their visits. The Bushmen have no religion whatever ; no form of marriage — any more than mating together like wild beasts; but they appear to have some respect for the memory of their dead, since they bury them — usually erecting a large pile of stones, or "cairn," over the body. They are far from being of a melancholy mood. Though crouching in their dens and caves during the day, in dread of the boers and other enemies, they come forth at night to chatter and make merry. During fine moonlights they dance all night, keeping up the ball till morning ; and in their kraals may be seen a circular spot — beaten hard and smooth with their feet — where these dances are performed. They have no form of government' — not so much as A head man or chief. Even the father of the family 110 BOSJESMEN, OR BUSHMEN. possesses no authority, except such as superior strength may give him ; and when his sons are grown up and become as strong as he is, this of course also ceases. They have no tribal organization ; the small com- munities in which they live being merely so many in- dividuals aceidently brought together, often quarrelling and separating from one another. These communities rarely number over a hundred individuals, since, from the nature of their country, a large number could not find subsistence in any one place. It follows, therefore, that the Bushman race must ever remain widely scat- tered — so long as they pursue their present mode of life — and no influence has ever been able to win them from it. Missionary efforts made among them have all proved fruitless. The desert seems to have been cre- ated for them, as they for the desert ; and when trans- ferred elsewhere, to dwell amidst scenes of civilized life 5 tbey always yearn to return to their wilderness hoise. Truly are these pigmy savages an odd people ! THK AMAZONIAN INDIANS. In glancing at the map of the American continent, we are struck by a remarkable analogy between the geographical features of its two great divisions — the North and the South, — an analogy amounting almost to a symmetrical parallelism. Each has its "mighty" mountains — the Cordilleras of the Andes in the south, and the Cordilleras of the Sierra Madre (Rocky Mountains) in the north — with all the varieties of volcano and eternal snow. Each has its secondary chain : in the north, the Nevadas of Cali- fornia and Oregon ; in the south, the Sierras of Carac- cas and the group of Guiana ; and, if you wish to render the parallelism complete, descend to a lower ele- vation, and set the Alleghanies of the United States against the mountains of Brazil — both alike detached from all the others. In the comparison we have exhausted the mountain- chains of both divisions of the continent. If we pro- ceed further, and carry it into minute detail, we shall find the same correspondence — ridge for ridge, chain for cnain, peak for peak ; — in short, a most singular equilibrium, as if there had been a design that one half of this great continent should balance the other ! 112 THE AMAZONIAN INDIANS. From the mountains let us proceed to the riv^rSj and see how they will correspond. Here, again, we dis- cover a like parallelism, amounting almost to a rivalry. Each continent (for it is proper to style them so) con- tains the largest river in the world. If we make length the standard, the north claims precedence for the Mia sissippi ,* if volume of water is to be the criterion, the south is entitled to it upon the merits of the Amazon. Each, too, has its numerous branches, spreading into a mighty " tree " ; and these, either singly or combined, form a curious equipoise both in length andgnagnitude. We have only time to set list against list, tributaries of the great northern river against tributaries of its great southern compeer, — the Ohio and Illinois, the Yellow- stone and Platte, the Kansas and Osage, the Arkansas and Red, against the Madeira and Purus, the Ucayali and Huallaga, the Japura and Negro, the Xingu and Tapajos. Of other river systems, the St. Lawrence may be placed against the La Plata, the Oregon against the Orinoco, the Mackenzie against the Magdalena, and the Rio Bravo del Norte against the Tocantins ; while the two Colorado^ — the Brazos and Alabama — mid their respective rivals in the Essequibo, the Paranahybo, the Pedro, and the Patagonian Negro ; and the San Francisco of California, flowing over sands cf gold, is balanced by its homonyme of Brazil, that has its origin in the land of diamonds. To an endless list might the comparison be carried. We pass to the plains. Prairies in the north, Hanoi and pampas in the south, almost identical in character. Of the plateaux or table-lands, those of Mexico, La THE AMAZONIAN INDiANS. 113 Puebla, Perote, and silver Potosi in the north ; those of Quito, Bogota, Cusco, and gold Potosi in the south ; of the desert plains, Utah and the Llano Estacado against Ataeama and the deserts of Patagonia. Even the Great Salt Lake has its parallel in Titicaca; while the "Sali- nas " of New Mexico and the upland prairies, are rep- resented by similar deposits in the Gran Chaco and the Pampas. We arrive finally at the forests. Though unlike in other respects, we have here also a rivalry in magni- tude, — between the vast timbered expanse stretching from Arkansas to the Atlantic shores, and that which covers the valley of the Amazon. These were the two greatest forests on the face of the earth. I say were, for one of them no longer exists ; at least, it is no longer a continuous tract, but a collection of forests, opened by the axe, and intersected by the clearings of the colonist. The other still stands in all its virgin beauty and pri-i, meval vigor, untouched by the axe, undefined by fire, its path scarce trodden by human feet, its silent depths to this hour unexplored. It is with this forest and its denizens we have to do. Here then let us terminate the catalogue of similitudes, and concentrate our attention "upon the particular subject of our sketch. The whole valley of the Amazon — in other words, the tract watered by this great river and its tributaries — may be described as one unbroken forest. We now know the borders of this forest with considerable exact- ness, but to trace them here would require a too length- ened detail. Suffice it to say, that lengthwise it extends from the mouth of the Amazon to the foothills of the 114 THE AMAZONIAN INDIAN*. Peruvian Andes, a distance of 2,500 miles. Tn breadth it varies, beginning on the Atlantic coast with a breadth of 400 miles, which widens towards the central part of the continent till it attains to 1,500, and again narrowing to about 1,000, where it touches the eastern slope of the Andes. That form of leaf known to botanists as " obovate " will give a good idea of the figure of the great Amazon forest, supposing the small end or shank to rest on the Atlantic, and the broad end to extend along the semi- circular concavity of the Andes, from Bolivia on the south to New Granada on the north. In all this vast expanse of territory there is scarce an acre of open ground, if we except the water-surface of the rivers and their bordering " lagoons," which, were they to bear their due proportions on a map, could scarce be repre- sented by the narrowest lines, or the most inconspicuous dots. The grass plains which embay the forest on its southern ed^e along the banks of some of its Brazilian tributaries, or those which proceed like spurs from the Llanos of Venezuela, do not in any place approach the Amazon itself*, and there are many points on the great river which may be taken as centres, and around which circles may be drawn, having diameters 1,000 miles in length, the circumferences of which will enclose nothing hut timbered land. The main stream of the Amazon, though it intersects this grand forest, does not bisect it, speaking witH mathematical precision. There is rather more timbered surface to the southward than that which extends northward, though the inequality of the two division* •: .ot great. It would not be much of an error to say tha*. : A.na^on .i : cuts the foiest in halved TflE AMAZONIAN INDIANS. 1 ! 5 At its mouth, however, this would not apply ; since for the first 300 miles above the embouchure of the river the country on the northern side is destitute of timber. This is occasioned by the projecting spurs of (he Guiana mountains, which on that side approach the Amazon in the shape of naked ridges and grass-covered hills and plains. It is not necessary to say that the great forest of tha Amazon is a tropical one — since the river itself, through- out its whole cburse, almost traces the line of the equator. Its vegetation, therefore, is emphatically of a tropical character ; and in this respect it differs essentially from that of North America, or rather, we should say, of Can- ada and the United States. It is necessary to make thi? limitation, because the forests of the tropical parts of North America, including the West-Indian islands, pre sent a great similitude to that of the Amazon. It is not only in the genera and species of trees that the sylva of the temperate zone differs from that of the torrid ; but there is a very remarkable difference in the distribution of these genera and species. In a great forest of the north, it is not uncommon to find a large tract covered with a single species of trees, — as with pines, oaks, poplars, or the red cedar (Juniperus Virginiana) This arrangement is rather the rule than the exception ; whereas, in the tropical forest, the rule is reversed, ex- cept in the case of two or three species of palms (Mau- ritia and JZutco-pe), which sometimes exclusively cover large tracts of surface. Of other trees, it is rare to find even a clump or grove standing together — often only two or three trees, and still more frequently, a single individual is observed, separated from tnose of its gwd 116 THE AMAZONIAN INDIANS. kind by hundreds of others, all differing in order, genua, and species. I note this peculiarity of the tropic forest, because it exercises, as may easily be imagine**, a direct influence upon the economy of its human occupants — whether these be savage or civilized. Even the habits of the lower animals — beasts and birds — are subject to a similar influence. It would be out of place here to enumerate the differ- ent kinds of trees that compose this mighty wood, — a «bare catalogue of then' names would alone fill many pages, — and it would be safe to say that if the list were given as now known to botanists, it would comprise scarce half the species that actually exist in the valley of the Amazon. In real truth, this vast Garden of God io yet unexplored by man. Its border walks and edges- have alone been examined ; and the enthusiastic botanist need not fear that he is too late in the field. A hundred years will elapse before this grand parterre can be ex- hausted. At present, a thorough examination of the botany of the Amazon valley would be difficult, if not altogether impossible, even though conducted on a grand and ex- pensive scale. There are several reasons for tins. Its woods are in many places absolutely impenetrable — on account either of the thick tangled undergrowth, or from the damp, spongy nature of the soil. There are no roads that could be traversed by horse or man ; and the few paths are known only to the wild savage, — not always passable even by him. Travelling can only be done by water, either upon the great rivers, or by the narrow creeks (igaripes) or lagoons ; and a journey per- formed in this fashion must needs be both tedious and THE AMAZONIAN INDIANS. 117 indirect, allowing but a limited opportunity for observa- tion. Horses can scarce be said to exist in the country, suid cattle are equally rare — a few only are found in one or two of the large Portuguese settlements on the main river — and the jaguars and blood-sucking bats offer a direct impediment to their increase. Contrary to the general belief, the tropical forest is not the home of the larger mammalia : it is not their proper habitat, nor are they found in it. In the Amazon forest but few species exist, and these not numerous in individuals. There are no vast herds — as of buffaloes on the prai- ries of North America, or of antelopes in Africa. The tapir alone attains to any considerable size, — exceeding that of the ass, — but its numbers are few Three or four species of small deer represent the ruminants, and the hog of the Amazon is the peccary. Of these there are at least three species. Where the forest impinges on the mountain regions of Peru, bears are found of at least two kinds, but not on the lower plains of the great " Montana," — for by this general designation is the vast expanse of the Amazon country known among the Peruvian people. "Montes" and " montanas," lit- erally signifying " mountains," are not so understood among Spanish Americans. With them the "montes" and a montanas " are tracts of forest-covered country, and that of the Amazon valley is the " Montana " pa? excellence. Sloths of several species, and opossums of sfll greater variety, are found all over the Montana, but both thinly distributed as regards the number of individuals. A similar remark applies to the ant-eaters or " ant-bears," of which there are four kinds, — to the armadillos, tb« US THE AMAZONIAN INDIANS. u agoutis/' and the " cavies," one of which last, the eapi* bara, it the largest rodent upon earth. This, with its kindred genus, the " paca," is not so rare in individual numbers, but, on the contrary, appears in large herds upon the borders of the rivers and lagoons. A porcu- pine, several species of spinous rats, an otter, two or three kinds of badger-like animals (the potto and coat is) , a " honey -bear " (Galera barbara), and a fox, or wild dog, are widely distributed throughout the Montana. Everywhere exists the jaguar, both the black and spotted varieties, and the puma has there his lurking- place. Smaller cats, both spotted and striped, are nu- merous in species, and squirrels of several kinds, with bats, complete the list of the terrestrial mammalia. Of all the lower animals, monkeys are the most common, for to them the Montana is a congenial home. They abound not only in species, but in the number of individuals, and their ubiquitous presence contributes to enliven the woods. At least thirty different kinds of them exist in the Amazon valley, from the u coatas," and other howlers as large as baboons, to the tiny little " ouistitis " and " saimiris," not bigger than squirrels or rats. While we must admit a paucity in the species of the quadrupeds of the Amazon, the same remark does not apply to the birds. In the ornithological department of natural history, a fulness and richness here exist, per- haps not equalled elsewhere. The most singular and graceful forms, combined with the most brilliant plumage, are everywhere presented to the eye, in the parrots and great macaws, the toucans, trogons, and tanagers, the shrikes, humming-1 irds, and orioles ; and even in the THE AMAZONIAN INDIANS. Hi) rultures and eagles : for here are found th 3 most beau- tiful of predatory birds, — the king vulture and the harpy eagle. Of the feathered creatures existing in the valley of the Amazon there are not less than one thousand different species, of which only one half have yet been caught or described. Reptiles are equally abundant — the serpent family being represented by numerous species, from the great water boa (anaconda), often yards in length, to the tiny and beautiful but venomous lachesis, or coral snake, not thicker than the shank of a tobacco-pipe. The lizards range through a like gradation, beginning with the huge "jacare," or crocodile, of several species, and ending with the turquoise-blue anolius, not bigger than a newt. The waters too are rich in species of their peculiar inhabitants — of which the most remarkable and valu- able are the manatees (two or three species), the great and smaller turtles, the porpoises of various kinds, and an endless catalogue of the finny tribes that frequent the rivers of the tropics. It is mainly from this source, and not from four-footed creatures of the forest, that the human denizen of the great Montana draws his supply of food, — at least that portion of it which may be termed the u meaty." Were it not for the manatee, the great porpoise, and other large fish, he would often have to " eat his bread dry." And now it is his turn to be " talked about." I need not inform you that the aborigines who inhabit the valley of the Amazon, are all of the so-called Indian race — though there are so many distinct tribes of them that almost every river of any considerable magnitude has a tribe of its own. In some cases a number of these 120 THE AMAZONIAN INDIANS. tribes belong to one nationality ; that is, several of them may be found speaking nearly the same language, though living apart from each other ; and of these larger di- visions or nationalities there are several occupying the different districts of the Montana. The tribes even of the same nationality do not always present a uniform appearance. There are darker and fairer tribes ; some in which the average standard of height is less than among Europeans ; and others where it equals or ex- ceeds this. There are tribes again where both men and women are ill-shaped and ill-favored — though these are few — and other tribes where both sexes exhibit a con- siderable degree of personal beauty. Some tribes are even distinguished for their good looks, the men pre- senting models of manly form, while the women are equally attractive by the regularity of their features, and the graceful modesty of expression that adorns them, A minute detail of the many peculiarities in which the numerous tribes of the Amazon differ from one another would fill a large volume ; and in a sketch like the present, which is meant to include them all, it would not be possible to give such a detail. Nor indeed would it serve any good purpose ; for although there are many points of difference between the different tribes, yet these are generally of slight importance, and are far more than counterbalanced by the multitude of resemblances. So numerous are these last, as to create a strong idio- syncrasy in the tribes of the Amazon, which not only entitles them to be classed together in an ethnological point of view, but which separates them from all the other Indians of America. Of course, the non -posses- sion of the horse — they do not even know the animal THE \MAZONTAN INDIANS. 121 — at once broach? distinguishes them from the Horse Indians, both of "ho Northern and Southern divisions of the continent. It would H Mile here to discuss the question as to whether ♦Jic Amazonian Indians have all a common origin. Tt is evident they have not. We know that many of them are from Peru and Bogota — runaways from Spanish oppression. We know that others mi- grated from the south — equally fugitives from the still more brutal and barbarous domination of the Portu- guese. And still others were true aboriginals of the 6oil, or if emigrants, when and whence came they? An idle question, never to be satisfactorily answered. There they now are, and as they are only shall we here consider them. Notwithstanding the different sources whence they sprang, we find them, as I have already said, stamped with a certain idiosyncrasy, the result, no doubt, of the like circumstances which surround them. One or two tribes alone, whose habits are somewhat "odder" than the rest, have been treated to a separate chapter ; but for the others, whatever is said of one, will, with very slight alteration, stand good for the whole of the Ama- zonian tribes. Let it be understood that we are dis- coursing only of those known as the "Indios bravos," the fierce, brave, savage, or wild Indians — as you may choose to translate the phrase, — a phrase used through- out all Spanish America to distinguish those tribes, o? sections of tribes, who refused obedience to Spanish tyranny, and who preserve to this hour their native in- dependence and freedom. In contradistinction to the u Indios braves " are the " Indios mansos," or " turn* 122 TRF AMAZONIAN INDIANS. Indians," who submitted tamely both to the cross ?r:d sword, and now enjoy a rude demi-semi-civilization, un- der the joint protectorate of priests and soldiers. Be- tween these two kinds of American aborigines, there is as much difference as between a lord and liis serf — the true savage representing the former and the demi- semi-civilized savage approximating more nearly to the latter. The meddling monk has made a complete fail- ure of it. His ends were purely political, and the result has proved ruinous to all concerned ; — instead of civil- izing the savage, he has positively demoralized him. It is not of his neophytes, the " Indios mansos," we are now writing, but of the " infidels," who would not hearken to his voice or listen to his teachings — those who could never be brought within " sound of the bell." Both " kinds " dwell within the valley of the Amazon, but in different places. The " Indios mansos " may be found along the banks of the main stream, from its source to its mouth — but more especially on its upper waters, where it runs through Spanish (Peruvian) ter- ritory. There they dwell in little villages or collections of huts, ruled by the missionary monk with iron rod, and performing for him all the offices of the menial slave. Their resources are few, not even equalling those of their wild but independent brethren ; and their cus- toms and religion exhibit a ludicrous melange of sav- agery and civilization. Farther down the river, the " Indio manso " is a " tapuio," a hireling of the Portu- guese, or to speak more correctly, a slave ; for the lattei treats him as such, considers him as such, and though there is a law against it, often drags him from his forest- home and keeps him in life-long bondage. Any human THE AMAZONIAN INDIAN?- 123 law would be a dead letter among such while-skins as are to be encountered upon the banks of the Amazoa Fortunately they are but few ; a town or two on the lower Amazon and Rio Negro, — some w retched vil- lages between, — scattered estancias along the banks — with here and there a paltry post of " militarios," dig- nified by the name of a " fort : " these alone speak the progress of the Portuguese civilization throughout a pe- riod of three centuries ! From all these settlements the wild Indian keeps away. He is never found near them — he is never seen by travellers, not even by the settlers. You may descend the mighty Amazon from its source to its mouth, and rot once set your eyes upon the true son of the forest — the " Indio bravo." Coming in contact only with the neophyte of the Spanish missionary, and the skulking tapuio of the Portuguese trader, you might, bring away a very erroneous impression of the charac- ter of an Amazonian Indian. Where is he to be seen ? where dwells he ? what-like is his home ? what sort of a house does he build ? His costume ? his arms ? his occupation ? his habits ? These are the questions you would put. They shall all be answered, but briefly as possible — since our limited 6pace requires brevity. - The wild Indian, then, is not to be found upon the Amazon itself, though there are long reaches of the river where he is free to roam — hundreds of miles without either town or estancia. He hunts, and occa- eionall} fishes by the great water, but does not there make his dwelling — though in days gone b\, its shores were his favorite place of residence. These happy day* 124 THE AMAZONIAN INDIANS. were before the time when Orellana floated down paftl the door of his " maloeca ' — before that dark hour when the Brazilian slave-hunter found his way into the waters of the mighty Solimoes. This last event was the cause of his disappearance. It drove him from the shores of his beloved river-sea ; forced him to withdraw his dwell- ing from observation, and rebuild it far up, on those tributaries where he might live a more peaceful life, secure from the trafficker in human flesh. Hence it is that the home of the Amazonian Indian is now to be nought for — not on the Amazon itself, but on its tribu- tary streams — on the " canos " and " igaripes," the ca- nals and lagoons that, with a labyrinthine ramification, intersect the mighty forest of the Montana. Here dwells he, and here is he to be seen by any one bold enough to visit him in his fastness home. How is he domiciled ? Is there anything peculiar about the style of his house or his village ? Eminently peculiar; for in this respect he differs from all the other savage people of whom we have yet written, or of whom we may have occasion to write. Let us proceed at once to describe his dwelling. It is not a tent, nor is it a hut, nor a cabin, nor a cottage, nor yet a cave ! His dwelling can hardly be termed a house, nor his village a collection of houses — since both house and village are one and the same, and both are so pe- culiar, that we have no name for such a structure in civilized lands, unless we should call it a " barrack." But even this appellation would give but an erroneous idea of the Amazonian dwelling ; and therefore we shall use that by which it is kno vn in tr e " Lingoa geral/ and ca 1 ! it a maloeca THE AMAZONIAN INDIANS. 125 1>/ such name is his house (or vilhge rather) known among the tapuios and traders of the Amazon. Since it is both house and village at the same time, it must needs be a large structure ; and so is it, large enough to contain the whole tribe — or at least the section of it that has chosen one particular spot for their residence. It is the property of the whole community, built by the labor of all, and used as their common dwelling—™ though each family has its own section specially set apart for itself. It will thus be seen that the Amazo- nian savage is, to some extent, a disciple of the Social- ist school. I have not space to enter into a minute account of the architecture of the malocca. Suffice it to say, that it is an immense temple-like building, raised upon timber uprights, so smooth and straight as to resemble columns. Ths beams and rafters are also straight and smooth, and are held in their places by " sipos " (tough creeping plants), which are whipped around the joints with a neatness and compactness equal to that used in the rig- ging of a ship. The roof is a thatch of palm-leaves, laid on with great regularity, and brought very low down at the eaves, so as to give to the whole structure the ap- pearance of a gigantic beehive. The walls are built of split palms or bamboos, placed so closely together as. to be impervious to either bullet or arrows. The plan is a parallelogram, with a semicircle at one end ; and the building is large enough to accommodate the whole communicy, often numbering more than a hundred individuals. On grand festive occasions several neighboring communities can find room enough in it — even for dancing — and three or four hundred individuals 126 THE AMAZONIAN INDIANS. not unfrequently assemble under the roof if a singi€ malocca. Inside the arrangements are curious. There is a wide hall or avenue in the middle — that extends from end to end throughout the whole length of the parallelogram — and on both sides of the hall is a row of partitions, separated from each other by split palms or canes, closely placed. Each of these sections is the abode of a family, and the place of deposit for the hammocks, clay pots, calabash-cups, dishes, baskets, weapons, and ornaments, which are the private property of each. The hall is used for the larger cooking utensils — such as the great clay ovens and pans for baking the cassava, and boiling the caxire or chicha. This is also a neutral ground, where the children play, and where the dancing is clone on the occasion of grand " balls " and other ceremonial festivals. The common doorway L m the gable end, and is six feet wide by ten in height. It remains open during the day, but is closed at night by a mat of palm fibre sus- pended from the top. There is another and smaller doorway at the semicircular end; but this is for the private use of the chief, who appropriates the whole section of the semicircle to himself and his family. Of course the above is only the general outline of a malocca. A more particular description would not an- swer for that of all the tribes of the Amazon. Among different communities, and in different parts of the Mon- tana, the malocco varies in size, shape, and the materials of which it is built ; and there are some tribes who live in separate huts. These exceptions, however, are few, and as a general thing, that above described is the style THE AMAZONIAN INDIANS. 127 of habitation throughout the whole Montana, from the confines of Peru to the shores of the Atlantic. North and south we encounter this singular house-village, from the head-waters of the Rio Negro to the highlands of Brazil. Most of the Amazonian tribes follow agriculture, and anderstood the art of tillage before the coming of the Spaniards. They practise it, however, to a very lim- ited extent. They cultivate a little manioc, and know how to manufacture it into farinha or cassava bread. They plant the musacece and yam, and understand the distillation of various drinks, both from the plantain and several kinds of palms. They can make pottery from clay, — shaping it into various forms, neither rude nor inelegant, — and from the trees and parasitical twiners that surround their dwellings, they manufacture an end- less variety of neat implements and utensils. Their canoes are hollow trunks of trees sufficiently well shaped, and admirably adapted to their mode of travelling — which is almost exclusively by water, by the numerous canos and igaripes, which are the roads and paths of their country — often as narrow and intri cate as paths by land. The Indians of the tropic forest dress in the very light- est costume. Of course each tribe has its own fashion ; but a mere belt of cotton cloth, or the inner bark of a tree, passed round the waist and between the limbs, is all the covering they care for. It is the guayuco. Some wear a =kirt of tree-bark, and, on grand occasions, fea.her tunics are seen, and also plume head-dresses, made of the brilliant wing and tail feathers of parrots and macaws. Circlets of these also adorn the arms and lirabs. All th« 128 THE AMAZOriAN LILIANS. tribes paint, using the anotto, carnto, and sereral other dyes which they obtain from various kinds of trees, else- where more particularly described. There are one or two tribes who. tattoo their skins but this strange practice is far less common among the American Indians than with the natives of the Pacific isles. In the manufacture of their various household utensils and implements, as well as their weapons for war and the chase, many tribes of Amazonian Indians display an in- genuity that would do credit to the most accomplished artisans. The hammocks made by them have been ad- mired everywhere ; and it is from the valley of the Ama- zon that most of these are obtained, so much prized in the cities of Spanish and Portuguese America. They are the special manufacture of the women, the men only employing their mechanical skill on their weapons. The hammock, " rede," or " maqueira," is manufac- tured out of strings obtained from the young leaves of several species of palms. The astrocaryum, or " tucum " palm furnishes this cordage, but a still better quality is obtained from the "miriti" (Mauritia jiexuosa). The unopened leaf, which forms a thick pointed column grow- ing up out of the crown of the tree, is cut off at the base, and this being pulled apart, is shaken dexterously until the tender leaflets fall out. These being stripped of their outer covering, leave behind a thin tissue of a pale- yellowish color which is the fibre for making the cordage. After being tied in bundles this fibre is left awhile to dry, and is then twisted by being rolled between the hand and the hip or thigh. The women perform this process with great dexterity. Taking two strands of fibre between THE AMAZONIAN INDIANS. 129 the forefinger and thumb of the left hand, they lay them separated a little along the thigh ; a roll downward givea them a twist, and then being adroitly brought together, a roll upwards completes the making of the cord. Fifty fathoms in a day is considered a good day's spinning. The cords are afterwards dyed of various colors, to ren- der them more ornamental when woven into the ma- queira. The making of this is a simple process. Two horizon- tal rods are placed at about seven feet apart, over which the cord is passed some fifty or sixty times, thus forming the " woof." The warp is then worked in by knotting ;he cross strings at equal distances apart, until there are enough. Two strong cords are then inserted where the rods pass through, and these being firmly looped, so as to draw all the parallel strings together, the rod is pulled out, and the hammock is ready to be used. Of course, with very fine " redes," and those intended to be disposed of to the traders, much pains are taken in the selection of the materials, the dyeing the cord, and the weaving it into the hammock. Sometimes very ex- pensive articles are made ornamented with the brilliant feathers of birds cunningly woven among the meshes and along the borders. Besides making the hammock, which is the universal couch of the Amazonian Indian, the women also manu- facture a variety of beautiful baskets. Many speck s of palms and calamus supply them with materials for this purpose, one of the best being the " Iu " palm (Astroca ryum acaide). They also make many implements and utensils, some for cultivating the plantains, melons, and manioc root, and others for manufacturing the last-named ICO THE AMAZONIAN INDIANS. vegetable into their favorite "farinha" (cassava). The Indians understood how to separate the poisonous juice of this valuable root from its wholesome farina before the arrival of white men among them ; and the process by which they accomplish this purpose has remained without change up to the present hour, in fact, it is almost the same as that practised by the Spaniards and Portu- guese, who simply adopted the Indian method. The ~work is performed by the women, and thus: the roots are brought home from the manioc " patch " in baskets, and then washed and peeled. The peeling is usually performed by the teeth ; after that the roots are grated, the grater being a large wooden slab about three feet long, a foot wide, a little hollowed out, and the hollow part covered all over with sharp pieces of quartz set in regular diamond-shaped patterns. Sometime a cheaper grater is obtained by using the aerial root of the pashiuba palm (Iriartea exhorhiza), which, being thickly covered over with hard spinous protuberances, serves admirably for the purpose. The grated pulp is next placed to dry upon a sieve, made of the rind of a water-plant, and is afterwards put into a long elastic cylinder-shaped basket or net, of the bark of the " jacitara " palm (Desmoncus macroacan* thus). This is the tipiti ; and at its lower end there is a strong loop, through which a stout pole is passed; while the tipiti itself, when filled with pulp, is hung up to the branch of a tree, or to a firm peg in the wall. One end of the pole is then rested against some project- ing point, that serves as a fulcrum, while the Indian woman, having seated herself upon the oilier end, with her infant in her arms, or perhaps some work in he! THE AMAZONIAN INDIANS. 131 kttnas act 3 as the lever power. Her weight draws the Bides of the tipiti together, until it assumes the form of an inverted cone ; and thus the juice is gradually pressed out of the pulp, and drops into a vessel placed underneath to receive it. The mother must be careful that the little imp does not escape from under her eye, and perchance quench its thirst out of the vessel below. If such an accident were to take place, in a very few minutes she would have to grieve for a lost child ; since the sap of the manioc root, the variety most cultivated by the Indians, is a deadly poison. This is the " yuc- ca amarga," or bitter manioc ; the " yucca dulce," or sweet kind, being quite innoxious, even if eaten in its raw state. The remaind^/ of the process consists in placing the grated pulp — now sufficiently dry — on a large pan or oven, and submitting it to the action of the fire. It is then thought sufficiently good for Indian use ; but much of it is afterwards prepared for commerce, under different names, and sold as semonilla (erroneously called semolina), sago, and even as arrowroot. At the bottom of that poisonous tub, a sediment has all the while been forming. . That is the starch of the manioc root — the tapioca of commerce : of course that is not thrown away. The men of the tropic forest spend their lives in doing very little. They are idle and not much disposed to work — only when war or the chase calls them forth do they throw aside for awhile their indolent habit, and exhibit a little activity. They hunt with the bow and arrow, and fish with a barpoon spear, ne:s, and sometimes by poisoning the 132 THE AMAZONIAN INDIANS. water with the juice of a vine called barbasco. Tin u peixe boy," " vaca marina," or " manatee," — all three names being synonymes — is one of the chief animals of their pursuit. All the waters of the Amazon valley abound with manatees, probably of several species, and these large creatures are captured by the harpoon, just as seals or walrus are taken. Porpoises also frequent the South-American rivers and large fresh-water fish of numerous species. The game hunted by the Ama- zonian Indians can scarcely be termed noble. We have seen that the large mammalia are few, and thinly dis- tributed in the tropical forest. With the exception of the jaguar and peccary, the chase is limited to small quadrupeds — as the capibara, the paca, agouti — to many kinds of monkeys, and an immense variety of birds. The monkey is the most common game, and is not only eaten by all the Amazonian Indians, but by most of them considered as the choicest of food. In procuring their game the hunters sometimes use the common bow and arrow, but most of the tribes are in possession of a weapon which they prefer to all others for this particular purpose. It is an implement of death bo original in its character and so singular in its con- struction as to deserve a special and minute description. The weapon I allude to is the "blow-gun," called "pucuna" by the Indians themselves, "gravitana" by the Spaniards, and " cerbatana " by the Portuguese of Brazil. When the Amazonian Indian wishes to manufacture for himself a pucuna he goes out into the forest and searches for two tall, straight stems of the "pashiuba miri" pa\m (Iriartea setigera). These he requires of THE AMAZOKLAN INDIANS. 133 Buck thickness that one can be contained within the other. Having found what he wants, he cuts both down and carries them home to his molocca. Neither of them is of such dimensions as to render this either impossible or difficult. He now takes a long slender rod — already prepared for the purpose — and with this pushes out the pith from both stems, just as boys do when preparing their pop-guns from the stems of the elder- tree. The rod thus used is obtained from another species of Iriartea palm, of which the wood is very hard and tough. A little tuft of fern-root, fixed upon the end of the rod, is then drawn backward and forward through the tubes, OEtil both are cleared of any pith which may have ad- hered to the interior; and both are polished by this process to the smoothness of ivory. The palm of smaller diameter, being scraped to a proper size, is now inserted into the tube of the larger, the object being to correct any crookedness in either, should there be such ; and if this does not succeed, both are whipped to some straight beam or post, and thus left till they become straight. One end of the bore, from the nature of the tree, is always smaller than the other ; and to this end is fitted a mouth-piece of two peccary tusks to concen Irate the breath of the hunter when blowing into the tube. The other end is the muzzle ; and near this, on the top, a sight is placed, usually a tooth of the " paca " or some other rodent animal. This sight is glued on with a gum which another tropic tree furnishes. Over the outside, when desirous of giving the weapon an ornamental finish, the maker winds spirally a shining creeper, and then the pucuna is ready for action. 134 THE AYAZOXIAN INDIANS Sometimes only a single shank of palm is used, and instead of the pith being pushed out, the item is split into two equal parts tliroughout its whole extent. The heart substance being then removed, the two pieces are brought together, like the two divisions of a cedarwood pencil, and tightly bound with a sipo. The pucuna is usually about an inch and a half in diameter at the thickest end, and the bore about equal to that of a pistol of ordinary calibre. In length, how- ever, the weapon varies from eight to twelve feet. This singular instrument is designed, not for propel- ling a bullet, but an arrow ; but as this arrow differs altogether from the common kind it also needs to be described. The blow- of the Esquimaux, what ingots of gold are are to Europeans, and worth while inquir- ing if a few bars of the last-mentioned metal were laid loosely and carelessly upon the pavements of London, how long they would be in changing their owners ? Theft should be regarded along with the amount of temptation ; and it appears even in these recorded cases that only a few of the Esquimaux took part in it. I apprehend that something more thar a few Londoners would be found picking up the golden ingots. How many thieves have we among us, with no greater temptation than THE ESQUIMAUX. 1G7 a cheap cotton kerchief ? — more than a few it is to to feared. In truth, the Esquimaux are by no means the savagea they have been represented. The only important point in which they at all assimilate to the purely savage state is in the Pithiness of their persons, and perhaps also in the fact of their eating much of their food (iish and flesh- meat) in a raw state. For the latter habit, however, they are partially indebted to the circumstances in which they are placed — fires or cookery being at times alto- gether impossible. They are not the only people who have been forced to eat raw flesh ; and Europeans who have travelled in that inhospitable country soon get used to the practice, at the same time getting quite cured of their dig out for it. It is certainly not correct to characterize the Esqui- maux as mere savages. On the contrary, they may be regarded as a civilized people, that is, so far as civiliza- tion is permitted by the rigorous climate in which they live ; and it would be safe to affirm that a colony of the most polished people in Europe, established as the Esqui- maux are, and left solely to their own resources, would in a single generation exhibit a civilization not one degree higher than that now met with among the Esquimaux, [ndeed, the fact is already established : the Danish and Norwegian colonists of West Greenland, though backed by constant intercourse with their mother-land, are but little more civilized than the " Skellings," who are their neighbors. In reality, the Esquimaux have made the most of the circumstances in which they are placed, aud continue to do so. Among them agriculture is impossible, else they 168 TEE ESQUIMAUX. would long since have taken to it. So too is commerce \ and as to m anufactures, it is doubtful whether Europeans could excel them under like circumstances. Whatever raw material their country produces, is by them both strongly and neatly fabricated, as indicated by the sur- prising skill with which they make their dresses, their boats, their implements for hunting and fishing ; and in these accomplishments — the only ones practicable under their hyperborean heaven — they are perfect adepts. In such arts civilized Europeans are perfect simpletons to them, and the theories of fireside speculators, so lately promulgated in our newspapers, that Sir John Franklin and his crew could not fail to procure a living where the simple Esquimaux were able to make a home, betrayed only ignorance of the condition of these people. In truth, white men would starve, where the Esquimaux could live in luxurious abundance, so far superior to ours is their knowledge both of fishing and the chase. It is a well-recorded fact, that while our Arctic voyagers, at their winter stations, provided with good guns, nets, and every appliance, could but rarely kill a reindeer or capture a seal, the Esquimaux obtained both in abun- dance, and apparently without an effort ; and we shall presently note the causes of their superiority in this respect The very dress of the Esquimaux is a proof of their superiority over other savages. At no season of the year do they go either naked, or even " ragged." They have their changes to suit the seasons, — their summer dress, and one of a warmer kind for winter. Both arf made in a most complicated manner ; and the prepara* tion of the material, as well as the manner by which if THE ESQUIMAUX. 169 is put together, prove the Esquimaux women — for they are alike the tailors anil dressmakers — to be among the best seamstresses in the world. Captain Lyon r one of the most observant of Arctic voyagers, has given a description of the costume of tne Esquimaux of Savage Island, and those of Repulse Bay, where be wintered, and his account is so graphic and minute in details, that it would be idle to alter a word of his language. His description, with slight differences in make and material, will answer pretty accurately for the costume of the whole race. " The clothes of both sexes are principally composed of fine and well-prepared reindeer pelts ; the skins of bears, seals, wolves, foxes, and marmottes, are also used. The seal-skins are seldom employed for any part of the dress except boots and shoes, as being more capable of resisting water, and of far greater durability than other leather. " The general winter dress of the men is an ample outer coat of deer-skin, having no opening in front, and a large hood, which is drawn over the head at pleasure This hood is invariably bordered with white fur from the thighs of the deer, and thus presents a lively con- trast to the dark face which it encircles. The front or belly part of the coat is cut off square with the upper part of the thighs, but behind it is formed into a broad skirt, rounded at the lower end, which reaches to within a few inches of the ground. The lower edges and tail9 of these dresses are in some cases bordered witJi bands of fur of an opposite col )r to the body ; and it is a favor- ite ornament to hang a fringe of little strips of skin be- neath the border. The embellishments give a verj 170 THE ESQUIMAUX. pleasing appearance to the dress. It is customary in blowing weather to tie a piece of skin or cord tight round the waist of the coat ; but in other cases the dress hanpr? loose. " Within the covering I have just described is another, of precisely the same form ; but though destitute of orna- ments of leather, it has frequently little strings of beads hanging to it from the shoulders or small of the back. This dress is of thinner skin, and acts as a shirt, the hairy part being placed near the body : it is the in-doors habit. When walking, the tail is tied up by two strings to the back, so that it may not incommode the legs. Be- sides these two coats, they have also a large cloak, or, in fact, an open deer-skin, with sleeves : this, from its size, is more frequently used as a blanket ; and I but once saw it worn by a man at the ship, although the women throw it over their shoulders to shelter them- selves and children while sitting on the sledge. " The trowsers, which are tightly tied round the loins, have no waistbands, but depend entirely by the drawing- string ; they are generally of deer-skin, and ornamented in the same manner as the coats. One of the most favorite patterns is an arrangement of the skins of deer's legs, so as to form very pretty stripes. As with the< jack:ts, there, are two pair of these indispensables, reaching no lower than the knee-cap, which is a cause of great distress in cold weather, as that part is fre- quently severely frost-bitten ; yet, with all their expe- rience of this bad contrivance, they will not add an inch to the established length. " The boots reach to the bottom of the breechus, whiot hang loosely over them. In these, as in other parts of rHE ESQUIMAUX. 171 the drer>s, are many varieties of color, material, aid pat- tern, yet in shape they never vary. The general winter boots are of deer-skin ; one having the hair next the leg, and I he other with the fur outside. A pair of soft slip- pers of the same kind are worn between the two pair of boots, and outside of all a strong seal-skin shoe is pulled to the* height of the ankle, where it is tightly secured by a drawing-string. For hunting excursions, or in sum? mer when the country is thawed, one pair of boots only >=; worn. They are of sealskin, and so well sewed and prepared without the hair, that although completely sat- urated, they allow no water to pass through them. The soles are generally of the tough hide of the walrus, or of the large seal called Oo-ghioo, so that the feet*are well protected in walking over rough ground. Slippers are sometimes worn outside. In both cases the boots are tightly fastened round the instep with a thong of leather. The mittens in common use are of deer-skin, with the hair inside ; but, in fact, every kind of skin is lsed for them. They are extremely comfortable when Iry ; but if once wetted and frozen again, in the winter afford as little protection to the hands as a case of ice would do. In summer, and in fishing, excellent seal- skin mittens are used, and have the same power of resist- ing water as the boots of which I have just spoken. The dresses I have just described are chiefly used in winter. During the summer it is customary to wear coats, boots, and even breeches, composed of the prepared skins of ducks, with the feathers next the bo(Jy. These are com- fortable, light, and easily prepared. The few ornaments in their possession are worn by the men. These are •ome baudeaus which encircle the head, And are cou> 172 THE ESQUIMAJX. posed of various-colored leather, plaited in a mosaic pattern, and in some cases having l?"man hair woven ir. them, as a contrast to the white skins. From the lower edge foxes' teeth hang suspended, arranged as a fringe across the forehead. Some wear a musk-ox tooth, a bit of ivory, or a small piece of bone. " The clothing of the women is of the same materials as that of the men, but in shape almost every part is different from the male dress. An inner jacket is worn next the skin, and the fur of the other is outside. The hind-flap, or tail, is of the same form before described, but there is also a small flap in front, extending about half-way down the thigh. The coats have each an im- mense hood, which, as well as covering the head, answers the purpose of a child's cradle for two or three years after the birth of an infant. In order to keep the bur- den of the child from drawing the dress tight across the throat, a contrivance, in a great measure resembling the slings of a soldier's knapsack, is affixed to the collar or neck part, whence it passes beneath the hood, crosses, and, being brought under the arms, is secured on each side the breast by a wooden button. The shoulders of the women's coat have a bag-like space, for the purpose of facilitating the removal of the child from the hood round to the breast without taking it out of the jacket. " A girdle is sometimes worn round the waist : it an- swers the double purposes of comfort and ornament , being composed of what \hej consider valuable trinkets, such as foxes' bones (those of the rableeaghioo), or sometimes of the ears of deer, which hang in pairs to the number of twenty or thirty, and are trophies of the nkill of the hunter, to vhom the wearer is allied The THE ESQUIMAUX. 173 inexpressibles of the women are in the same form aa those of the men, but they are not o.rnamented by the same curioub arrangement of colors ; the front part is generally of white, and the back of dark fur. The manner of securing them at the waist is also the same ; but the drawing-strings are of much greater length, being suffered to hang down by one side, and their ends are frequently ornamented with some pendent jewel, such as a grinder or two of the musk-ox, a piece of ivory, a small ball of wood, or a perforated stone. "The boots of the fair sex are, without dispute, the most extraordinary part of their equipment, and are of such an immense size as to resemble leather sacks, and to give a most deformed, and, at the same time, ludicrous appearance to the whole figure, the bulky part being at the knee ; the upper end is formed into a pointed flap, which, covering the front of the thigh, is secured by a button or knot within the waistband of the breeches. " Some of these ample articles of apparel are com- posed with considerable taste, of various colored skins ; they also have them of parchment, — seals' leather. Two pairs are worn ; and the feet have also a pair of seal- skin slippers, which fit close, and are tightly tied round the ankle. " Children have no kind of clothing, but he naked in their mothers' hoods until two or three years of age, when they are stuffed into a little dress, generally of fawn-skin, which has jacket and breeches in one, the back part being open ; into these they are pushed, when a string or two closes all up again. A cap forms an indispensable part of the equipment, and is generally of lome fantastical shape ; the skin of a fawn's head is a 174 THE ESQUIMAU* favorite material in the composition, and is FometuHJH Been with the ears perfect ; the nose and holes for the eyes lying along the crown of the wearer's head, which in consequence, looks like that of an animal." The same author also gives a most graphic description of the curious winter dwellings of the Esquimaux, which on many parts of the coast are built out of the (.nly materials to be had, — ice and s?iow I Snow for the walls and ice for the windows ! you might fancy the house of the Esquimaux to be a very cold dwelling ; such, however, is by no means its character. "The entrance to the dwellings," says Captain Lyon, "was by a hole, about a yard in diameter, which led through a low-arched passage of sufficient breadth for two to pass in a stooping posture, and about sixteen feet in length; another hole then presented itself, and led through a similarly-shaped, but shorter passage, having at its termination a round opening, about two feet across. Up this hole we crept one step, and found ourselves in a dome about seven feet in height, and as many in diam- eter, from whence the three dwelling-places, with arched roofs, were entered. It must be observed that this is the description of a large hut, the smaller ones, containing one or two families, have the domes somewhat differently arranged. " Each dwelling might be averaged at fourteen or six- teen feet in diameter by six or seven in height, but as snow alone was used in their construction, and was always at hand, it might be supposed that there was no particular size, that being of course at the option of the builder. The laying of the arch was performed in sucb 11 manner as would have satisfied the niort regular aitist, THE ESQUIMAUX. 175 ih»; key- piece on the top, being a large square slab The blocks of snow used in the buildings were from four to six inches in thickness, and about a couple of feet in length, carefully pared with a large knife. Where tvfO famihes occupied a dome, a seat was raised on either side, two feet in height. These raised places were used as beds, and covered in the first place with whalebone, 6prigs of andromeda, or pieces of seals'-skin, over these were spread deer-pelts and deer-skin clothes, which had a very warm appearance. The pelts were used as blan- kets, and many of them had ornamental fringes of leather sewed round their edges. " Each dwelling-place was illumined by a broad piece of transparent fresh-water ice, of about two feet hi diam- eter, which formed part of the roof, and was placed over the door. These windows gave a most pleasing light, free from glare, and something like that which is thrown thixmgh ground glass. We soon learned that the build- ing of a house was but the work of an hour or two, and that a couple of men — one to cut the slabs and the other to lay them — were laborers sufficient. " For the support of the lamps and cooking apparatus, a mound of snow is erected for each family ; and when the master has two wives or a mother, both have an independent place, one at each end of the bench. " I find it impossible to attempt describing everything at a second visit, and shall therefore only give an account of those articles of furniture which must be always the same, and with which, in five minutes, any one might be acquainted. A frame, composed of two or three broken fishing-spears, supported in the first place a large hoop of wood 01 bone, across which an open- meshed and ill- 176 THE ESQUIMAUX. made net was ?pread or worked for the reception of wet cr damp clothes, skins, etc., which could be dried by the heat of the lamp. On this contrivance the master of each hut placed his glomes on entering, first carefully iearing them of snow. " From the frame above -mentioned, one or more colHn- shapsd stone pots were suspendea over lamps of the same material, crescent-shaped, and having a ridge extending along their back ; the bowl part was filled with blubber, and the oil and wicks were ranged close together along the edge. The wicks were made of moss and trimmed by a piece of asbestos, stone, or wood ; near at hand a large bundle of moss was hanging for a future supply. The lamps were supported by sticks, bones, or pieces of horn, at a sufficient height to admit an oval pot of wood or whalebone beneath, in order to catch any oil that might drop from them. The lamps varied consid- erably in size, from two feet to six inches in length, and the pots were equally irregular, holding from two or three gallons to half a pint. Although I have mentioned a kind of scaffolding, these people did not all possess so grand an establishment, many being contented to suspend l heir pot to a piece of bone stuck in the wall of the hut. One young woman was quite a caricature in this way : she was the inferior wife of a young man, whose senior lady was of a large size, and had a corresponding lamp, etc., at one corner ; while she herself, being short and fat, had a lamp the size of half a dessert -plate, and a pot which held a pint only. u Almost every family was possessed of a large wood- en tray, resembling those used by butchers in Eng- land ; its offices, however, as w 3 soon perceived, were THE ESQUIMAUX. 177 aaofV various, some containing raw flesh of seals and blubber, and others, skins, which were steeping in urine, A quantity of variously-sized bowls of whalebone, wood, or skin, completed the list of vessels, and it was evident that they were made to contain anything? The Esquimaux use two kinds of boats, — the " oo- miak " and " kayak.'' The oomiak is merely a large specieo of punt, used exclusively by the women ; but the kayak is a triumph in the art of naval architecture, and is as elegant as it is ingenious. It is about twenty-five feet in length, and less than two in breadth of beam. In shape it has been compared to a weaver's shuttle, though it tapers much more elegantly than this piece of ma- chinery. It is decked from stem to stern, excepting a circular hole very nearly amidships, and this round hatch- way is just large enough to admit the body of an Esqui- maux in a sitting posture. Around the rim of the circle is a little ridge, sometimes higher in front than at the back, and this ridge is often ornamented with a hoop of ivory. A flat piece of wood runs along each side of the frame, and is, in fact, the only piece of any strength in a kayak. Its depth in the centre is four or five incles, and its thickness about three fourths of an inch ; it tapers to a point at the commencement of the stem and stern pro- jections. Sixty-four ribs are fastened to this gunwale piece ; seven slight rods run the whole length of the bot- tom and outside the ribs. The bottom is rounded, and has no keel ; twenty-two little beams or cross-pieces keep the frame on a stretch above, and one strong batten runs along the centre, from stem to stern, being, of course, dis continued at thn seat part. The ribs are made of ground willow, -Uso of whalebone, or, if it can be procured, of 178 T1TE ESQUIMAUX. good-grained wood. The whole contrivance does not weigb over fifty or sixty pounds; so that a man easily carries his kayak on his head, which, by the form of the ran, be can do without the assistance of his hands. An Esquimaux prides himself in the neat appearance of his boat, and has a warm skin placed in its bottom to sit on. His posture ; s with the legs pointed forward, and he cannot change his position without the assistance of another person ; in all cases where a weight is to be fifted, an alteration of stowage, or any movement to be made, it is customary for two kayaks to lie together ; and the paddle of each being placed across the other, they form a steady double boat. An inflated seal's bladder forms, invariably, part of the equipage of a canoe, and the weapons are confined in their places by small lines of whalebone, stretched tigjitly across the upper cover- ing, so as to receive the points or handles of the spears beneath them. Flesh is frequently stowed within the stem or stern, as are also birds and eggs ; but a seal, al- though round, and easily made to roll, is so neatly bal- anced on the upper part of the boat as seldom to require a lashing. When Esquimaux are not paddling, their bal- ance must be nicely preserved, and a trembling motion is always observable in the boat. The most difficult posi* tion for managing a kayak is when going before the wind, and with a little swell running. Any inattention would instantly, by exposing the broadside, overturn this frail vessel. The dexterity with winch they are turned, the velocity of their way, and the extreme elegance of form of the kayaks, render an Esquimaux of the highest inter- est when sitting independently, aud urging Ids course to^ wards his prey. THE ESQUIMAUX. 179 a The paddle is double-bladed, nhu feet \iree inches ifi length, small at the grasp, and widening to four inehea at the blades, which are thin, and edged with ivory for strength as well as ornament. "The next object of importance to the boat is the sledge, which finds occupation during at least three Fourths of the year. A man who possesses both this uid a canoe is considered a person of property. To #ve a particular description of the sledge would be im- possible, as there are no two actually alike ; and the ma- trials of which they are composed are as various as f heir form. The best are made of the jaw-bones of the whale, sawed to about two inches in thickness, and in depth from six inches to a foot. These are the runners, and are shod with a thin plant of the same material ; the side-pieces are connected by means of bones, pieces of wood, or deers' horns, lashed across, with a few inches space between each, and they yield to any great strain which the sledge may receive. The general breadth of the upper part of the sledge is about twenty inches ; but the runners lean inwards, and therefore at bottom it is rather greater. The length of bone sledges is from four feet to fourteen. Their weight is necessarily great ; and one of moderate size, that is to say, about ten or twelve feet, was found to be two hundred and seventeen pounds. The skin of the walrus is very commonly used during the coldest part of the winter, as being hard-frozen, and resembling an inch board, with ten time the strength, for runners. Another ingenious contrivance is, by casing mos? and earth hi seal's skin, so that by pouring a little water a round hard bolster is easily formed. Across all these kinds of runners there is the same arrangement of tones, 180 THE ESQUIMAUX. sticks, &c, on the upper part ; and the surface which passes over the snow is coated with ice, by mix; tig snow with fresL water, which assists greatly in lightening the load for the dogs, as it slides forwards with ease. Boyi frequently amuse themselves by yoking several dogs to a small piece of seal's skin, and sitting on i ,, holding by the traces. Their plan is then to set off at full speed, and he who bears the greatest number of bumps before he relinquishes his hold is considered a very fine fellow. " The Esquimaux possess various kinds of spears, but their difference is chiefly in consequence of the sub- stances of which they are composed, and not in their general form. " One called ka-te-teek, is a large and strong-handled spear, with an ivory point made for despatching any wounded animal in the water. It is never thrown, but has a place appropriated for it on the kayak. " The oonak is a lighter kind than the former ; also ivory-headed. It has a bladder fastened to it, and has a loose head with a line attached ; this being darted into an animal, is instantly liberated from the handle winch gives the impetus. Some few of these weapons are con- structed of the solid ivory of the unicorn's horn, about four feet in length, and remarkably well rounded and polished. " Ip-poo-too-yoo, is another kind of hand-spear, vary- ing but little from the one last described. It has, how- ever, no appendages. " The Noogh-wit is of two kinds ; but both are used for striking birds, young animals, or fish. The first hsa a double fork at the extremity, and there are three other barbed ones at abci * half its length, diveiging \n differ* THE ESQUIMAUX. 181 nt direction.;, so that if the end pair should miss, soma of the centre ones might strike. The second kind ha? mly thrue barbed forks at the head. All the points arassions. For this species of hunting, the bow far excels any other weapon ; even the rifle is inferior to it. Sometimes the Esquimaux take the deer in large numbers, by hunting them with dogs, driving the herd into some defile or cut de sac among the rocks, and then killing them at will with their arrows and jave- lins. This, however, is an exceptional case, as such natural " pounds " are not always at hand. The In- dians farther south construct artificial enclosures ; but in the Esquimaux -country there is neither time nor material for such elaborate contrivances. The Esquimaux who dwell in those parts frequented by the musk-oxen, hunt these animals very much as they do the reindeer; but killing a musk bull, or cow either, is a feat of far grander magnitude, and requires more address than shooting a tiny deer. I have said that the Esquimaux do not, even in these hunting excursions, stray very far into the inte- rior. There is a good reason for their keeping close to the seashore. Were they to penetrate far into the tand they would be in danger of meeting with their bitter ^bemen, the Thine Indians, who in this region fclao hunt reindeer and musk-oxen. War to the knife T1IE ESQUIMAUX. 187 18 the practice between these two races )f people, and has ever been since the first knowledge of either. They often meet in conflict upon the rivers inland, and th^se conflicts are of so cruel and sanguinary a nature as to imbue each with a wholesome fear of the other. The Indians, however, dread ihe Esquimaux more than the latter fear them ; and up to a late period took good care never to approach their coasts ; but the musket and rifle have now got iato the hands of some of the northern tribes, who avail themselves of these superior weapons, not only to keep the Esqui- maux at bay, but also to render them more cautious about extending their range towards the interior. When the dreary winter begins to make its appear- ance, and the reindeer grow scarce upon the snow- covered plains, the Esquimaux return to their winter villages upon the coast. Quadrupeds and birds no longer occupy their whole attention, for the drift of their thoughts is now turned towards the inhabitants of the great deep. The seal and the walrus are hence* forth the main objects of pursuit. Perhaps during the summer, when the water was open, they may have visited the shore for the purpose of capturing that great giant of the icy seas — a whale. If so, and they have been successful in only one or two captures, they may look forward to a winter of plenty — since the flesh of a full-grown whale, or, better still, -a brace of such ample creatures, would be sufficient *o *eed a whole tribe for months. They have no curing process for tins immense carcass , they stand in need of none. Neither salt nor smoking is required in their climate. Jack Frost is their provision 188 THE ESQUIMAUX. curer, and performs the task without putting them eithei to trouble or expense. It is only necessary for them to hoist tin; great flitches upon scaffolds, already erected for the purpose, so as to keep the meat from the wolves, wolverines, foxes, and their own half-starved dogs. From their aerial larder they can cut a piece of blubber when- ever they feel hungry, or they have a mind to eat, and this mind they are in so long as a morsel is left. Their mode of capturing a whale is quite different from that practised by the whale-fishers. When the huge creature is discovered near, the whole tribe salty forth, and suiTound it in their kayaks ; they then hurl darts into its body, but histead of these having long lines at- tached to them, they are provided with seal skins sewed up air-tight and inflated, like bladders. When a number of these become attached to the body of the whale, the animal, powerful though he be, finds great difficulty in sinking far down, or even progressing rapidly through the water. He soon rises to the surface, and the seal- skin buoys indicate his whereabouts to the occupants of the kayaks, who in their swift little crafts, soon dart up to him again, and shoot a fresh volley into his body. In this way the whale is soon " wearied out," and then falls a victim to their larger spears, just as in the case wb^.ra a capture is made by regular whalers. I need scarcely add that a success of this kind is hailed as a jubilee of the tribe, since it not only brings a benefit to the whole community, but is also a piece of fortune of somewhat rare occurrence. YiThen no whales have been taken, the long, dark win- ter may just'y be looked forward to with some solici- tude ; and it is then that the Esquimaux requires to put THE ESQUIMAUX. I $9 forth all his skill and energies for the capture of the wal- rus or th i seal — the latter of which may be regarded as ;he staff of his life, furnishing him not only with food, but with ligJ-t, fuel, and clothing for his body and limbs. Of the seals that inhabit the Polar Seas there are sev- eral species ; but the common seal ( Calocephalus vituli- na) and the harp-seal ( O. Grcenlandlcus) are those most numerous, and consequently the principal object of pursuit. The Esquimaux uses various stratagems for taking these creatures, according to the circumstances in which they may be encountered ; and simpletons as the seals may appear, they are by no means easy of capture. They are usually very shy and suspicious, even in places where man has never been seen by them. They have other enemies, especially in the great polar bear ; and the dread of this tyrant of the icy seas keeps them ever on the alert. Notwithstanding their watchfulness, how- ever, both the bear and the biped make great havoc among them, and each year hundreds of thousands of them are destroyed. The bear, in capturing seals, exhibits a skill and cun- ling scarce excelled by that of the rational being him- self. When this great quadruped perceives a seal bask- ing on the edge of an ice-field, he makes his approaches, not by rushing directly towards it, which he well knows would defeat his purpose. If once seen by the seal, the latter has only to betake himself to the water, where it can soon sink or swim beyond the reach of the bear. To prevent this, the bear gets well to leeward, and then div- ing below the surface, makes his approaches under water, now and then cautiously raising his head to get the tru« 190 THE ESQUIMAUX. bearings of his intended victim. After a number of these subaqueous " reaches," he gets .ilose in to the edge o^ the floe in such a position as to cut off the seal's re- treat to the water. A single spring brings him on the ice, and then, before the poor seal has time to make a brace of flounders, it finds itself locked in the deadly em- brace of the bear. When seals are thus detected asleep, the Esquimaux approaches them in his kayak, taking care to paddle cautiously and silently If he succeed in getting between them and the open water, he kills them in the ordinary way — by simply knocking them on the snou' with a club, or piercing them with a spear. Sometimes, however, the seal goes to sleep on the surface of the open water. Then the approach is made in a similar manner by means of the kayak, and the animal is struck with a harpoon. But a single blow does not always kill a seal, especially if it be a large one, an I the blow has been ill- directed. In such cases the animal would undoubtedly make his escape, and carry the harpoon along with it, which would be a serious loss to the owner, who does not obtain such weapons without great difficulty. To pre- vent this, the Esquimaux uses a < ontrivance similar to that employed in the capture of the whale, — that is, he attaches a float or buoy to his harpoon by means of a cord, and this so impedes the passage of the seal through the 7/ater, that it can neither dive nor swim to any very great distance. The float is usually a walrus bladder inflated in the ordinary way, and wherever the seal may go, the float betrays its track, enabling the Esquimaux to follow it in his shuttle-shaped kayak, and pierce it again wil surer aim. In winter, when the sea is quite covered with ice, yon THE ESQUIMAUX. 191 might fancy that the seal-fishery would be at an end, fol the seal is essentially a marine animal ; and although it can exist upon the ice or on dry land, it could not svliist there. Access to the water it must have, in ord?r to procure its food, winch consists of small lish and nob lusks. Of course, when the ice forms on the surface, the seal is in its true element — the water underneath — but when this ice becomes, as it often does, a full yard in thickness, extending over hundreds of miles of the sea, how then is the seal to be got at ? It could not be reached at all ; and at such a season the Esquimau? people would undoubtedly starve, were it not for a habit peculiar to this animal, which, happily for them, brings it within their reach. Though the seal can live under water like a fish, and probably could pass a whole white, under the ice without much inconvenience, it likes now and then to take a little fresh air, and have a quiet nap upon the upper surface ir. the open air. With this design it breaks a hole through the ice, while the latter is yet thin, and this hole it keeps carefully open during the whole winter, clearing out each new crust as it forms. No matter to what thickness the *ce may attain, this hole always forms a breathing-place for the seal, and a passage by which he may reach the upper surface, and indulge himself in his favorite siesta in the open air. Knowing this habit, the Esquimaux takes advantage of it to make the seal his captive. When the animal is discovered on the ice, the hunter approaches with the greatest stealth and caution. This is absolutely neccessary : for if the enemy is perceived, or makes the slightest noise, the wary seal flounders rapidly into his kolr, and fa lost beyond redemption. Tf badly frightened, 192 THE ESQUIMAUX. he will not appear for a long time, denying himself open air exercise until the patience of his persecutor id quite worn out, and the coast is again clear. In making his approaches, the hunter uses all his art, not only taking advantage of every inequality — such as snow-drifts and ice-hillocks — to conceal himself ; but he also practises an ingenious deception by dressing him- self in the skin of a seal of like species, giving his body the figure of the animal, and counterfeiting its motions, by floundering clumsily over the ice, and oscillating his head from side to side, just as seals are seen to do. This deception often proves successful, when the hun- ter under any other shape would in vain endeavor to get within striking distance of his prey. When seals are scarce, and the supply greatly needed, the Esquimaux often lies patiently for hours together on the edge of a seal-hole waiting for the animal to come up. In ordei to give it time to get well out upon the ice, the hunter conceals himself behind a heap of snow, which he has collected and piled up for the purpose. A float-stick, ingeniously placed in the water of the breathing-hole, serves as a signal to tell when the seal is mounting through his trap-like passage, the motion of the stick betraying its ascent. The hunter then gets himself into the right attitude to strike, and summons all his energies for the encounter. Even during the long, dark night of winter this mode of capturing the seal is practised. The hunter, having discovered a breathing-hole — which its dark color en- aDies mm u/ *md — proceeds in the following manner he scrapes away the snow from around it, and lifting up some water pours it on the ice, so as to make a circle THE ESQUIMAUX. 193 »f a darker hue around the orifice. He then makes a 6ort of cake of pure white snow, and with this cover*. the hole as with a lid. In the centre of this lid he punches a small opening with the shaft-end of his spear, and then sits down and patiently awaits the issut. The seal ascends unsuspiciously as before. The dark water, bubbling up through the small central orifice, betrays its approach, which can be perceived even in the darkest night. The hunter does not wait for its climbing out upon the ice. Perhaps if he did so, the suspicious creature might detect the device, and dive down again. But it is not allowed time for reflection. Before it can turn its unwieldly body, the heavy spear of the hunter — struck through the yielding snow — descends upon its skull, and kills it on the instant. The great " walrus " or " morse " ( Trichecus rosma* rus) is another important product of the Polar Seas, and is hunted by the Esquimaux with great assiduity. This splendid amphibious animal is taken by contrivances very similar to those used for the seal ; but the capture of a walrus is an event of importance, second only to the striking of a whale. Its great carcass not only supplies food to a whole village, but an oil superior to that of the whale, besides various other useful articles. Its skin, bones, and intestines are employed by the Esquimaux for many domestic purposes, — and, in addition, there are the huge molar tusks, that furnish one of the most valuable ivories of commerce, from which are manufac- tured those beautiful sets of teeth, of dazzling whiteness, that, gleaming between vermilion lips, you may often see at a ball or an evening party ! THE TONGANS, OK FRIENDLY ISLANDERS It is a pleasure to pass out of the company of the ferocious Feegees into that of another people, which, though near neighbors of the former, are different from them in almost every respect, — I mean the Tongans, or Friendly Islanders. This appellation scarce requires to be explained. Every one knows tha it was bestowed upon them hy the celebrated navigator Uook, — who al- though not the actual discoverer of the Tonga group was the first who thoroughly explored these islands, and gave any reliable account of them to the civilized world. Tasman, who might be termed the " Dutch Captain Cook," is allowed to be their discoverer, so long ago as 1643 ; though there is reason to believe that some of the Spanish explorers from Peru may have touched at these islands before his time. Tasman, however, has fixed the record of his visit, and is therefore entitled to the credit of the discovery, — as he is also to that ui' Australia, New Zealand, Van Diemen's Land, and other now well-known islands of the Southwestern Pacific Tasman bestowed upon three of the Tonga group the names — Amsterdam, Rotterdam, and Middleburgh ; but fortunately, geographers have acted in this matter with THE TONGANS. 195 better taste than is their wont ; and Ta>man's Dutch national titles Lave fallen into disuse, — while the true native names of the islands have been restored to the map. This is what should be done with other Pacific islands as well ; for it is difficult to conceive anything in worse taste than such titles as the Caroline and Loy- alty Isles, Prince William's Land, King George's Island, and the ten thousand Albert and Victoria Lands which the genius of flattery, or rather flunkeyism, has so lib- erally distributed over the face of the earth. The title of Friendly Isles, bestowed by Cook upon the Tonga archipelago, deserves to live ; since it is not only appro- priate, but forms the record of a pleasant fact, — the pacific character of our earliest intercourse with these interesting people. It may be here remarked, that Mr. Wylde and other superficial map-makers have taken a most unwarrantable liberty with this title. Instead of leaving it as bestowed by the great navigator, — applicable to the Tonga archi- pelago alone, — they have stretched it to include that of the Samoans, and — would it be believed — that of the Feegees ? It is hardly necessary to point out the extreme absurdity of such a classification : since it would be diffi- cult to find two nationalities much more unlike than those of Tonga and Feegee. That they have many cus- toms in common, is due (unfortunately for the Tongans) to the intercourse which proximity has produced ; but in an ethnological sense, white is not a greater contract to black, nor good to evil, than that which exists between a Tongan and a Feegeean. Cook never visited tne Feegee archipelago, — he only saw some of these peo- ple while at Tonga-taboo, and heard of their country as 196 THE TONGANS, OR being a large island. Had he visited that island, — oj rather that group of ovei two hundred islands, — it is not at all likely he would have seen reason to extend to them the title which the map-makers have thought fit to bestow. Instead of " Friendly Islands," he might by way of contrast have called them the " Hostile Isles," or given tL mi that — above all others most appropriate, and which they truly deserve to bear — that old title celebrated in song ! the " Cannibal Islands." An ob- server so acute as Cook could scarce have overlooked the appropriateness of the appellation. The situation of the Tonga, or Friendly Isles, is easily registered in the memory. The parallel of 20° south, and the meridian of 175° west, very nearly inter- sect each other in Tofoa, which may be regarded as the central island of the group. It will thus be seen that their central point is 5° east and 2° south of the centre of the Feegeean archipelago, and the nearest islands of the two groups are about three hundred miles apart. It is worthy of observation, however, that the Tonga Isles have the advantage, as regards the wind. The trades are in their favor ; and from Tonga to Feegee, if we employ a landsman's phraseology, it is " down hill," while it is all " up hill " in the contrary direction. The consequence is, that many Tongans are constantly making voyages to the Feegee group, — a large number of them having settled there (as stated elsewhere), — while but a limited number of Feegeeans find their way to the Friendly Islands. Thu'e is another reason for this unequally-balanced migration : and that is, that the Tongans are much bolder and better sailors than their western neighbors ; for although the Feegees far excel FRIENDLY ISLANDERS. 197 any other South-Sea islanders in the art of building tLeir canoes (or ships as they might reasonably be called), yet they are as far behind many others in the art of sailing them. Their superiority in ship-building may be attributed, partly, to the excellent materials which these islands abundantly afford ; though \his is not the sole cause. However much we may deny to the Feegeeans the possession of moral qualities, we are at the same time forced to admit their great intellectual capacity, — as exhibited in the advanced state of their arts and manu- factures. In intellectual capacity, however, the Friendly Islanders are their equals ; and the superiority of the Feegeeans even in " canoe architecture " is no longer acknowledged. It is true the Tongans go to the Feegee group for most of their large double vessels ; but that is for the reasons already stated, — the greater abundance and superior quality of the timber and other materials produced there. In the Feegee " dockyards," the Ton- gans build for themselves; and have even improved upon the borrowed pattern. This intercourse, — partaking somewhat of the chap acter of an alliance, — although in some respects advan- tageous to the Friendly Islanders, may be regarded, upon the whole, as unfortunate for them. If it has im- proved their knowledge in arts and manufactures, it has far more than counterbalanced this advantage by the damage done to their moral character. It is always much easier to make proselytes to vice than to virtue, — as is proved in this instance : for his intercourse with the ferocious Feegee has done much to deteriorate the character of the Tongan. From that source he tas in* 198 THE TONGA ITS, OR bibed a fondness for war and other wicked customs ; and in all probability, had this influence been permitted to continue uninterrupted for a few years longer, the horrid habit of cannibalism — though entirely repugnant to the natural disposition of the Tongans — would have become common among them. Indeed, there can be little doubt that this would have been the ultimate consequence of the alliance ; for already its precursors — human sacri- fices and the vengeful immolation of enemies — had made their appearance upon the Friendly Islands. Happily for the Tongan, another influence — that of the missionaries — came just in time to avert this dire catastrophe ; and, although this missionary interference has not been the best of its kind, it is still preferable to the paganism which it has partially succeeded in subduing. The Tongan archipelago is much less extensive than that of the Feegees, — the islands being of a limited number, and only five or six of them of any consid- erable size. Tongataboo, the largest, is about ninety miles in circumference. From the most southern of the group Eoo, to Vavau at the other extremity, it stretches, northerly or northeasterly, about two hundred miles, in ft nearly direct line. The islands are all, with one or two exceptions, low-lying, their surface being diversified by a few hillocks or mounds, of fifty or sixty feet in height, most of which have the appearance of being artificial. Some of the smaller islets, as Kao, are mountains of some six hundred feet elevation, rising directly out of the sea ; while Tofoa, near the eastern edge of the archipelago, presents the appearance oi an elevated table-laud. The larger nunilxr of them arc FRIENDLY ISLANDERS 199 clothed with a rich tropical vegetation, both n&turai and cultivated, and their botany includes most of the species common to the other islands of the South Sea. We find the cocoa, and three other species of palm, the pandanus, the bread-fruit in varieties, as also the use- ful musacaa?, — the plantain, and banana. The ti-tree (Draccena terminalis), the paper-mulberry (Broussonetia papyri/era), the sugar-cane, yams of many kinds, the tree yielding the well-known turmeric, the beautiful casuarina, and a hundred other sorts of plants, shrubs, or trees, valuable for the product of their roots or fruits, their sap and pith, of their trunks and branches, their leaves and the fibrous material of their bark. As a scenic decoration to the soil, there is no part of the world where more lovely landscapes are produced by the aid of a luxuriant vegetation. They are perhaps not equal in picturesque effect to those of the Feegee group, — where mountains form an adjunct to the scenery, — but in point of soft, quiet beauty, the land- scapes of the Tonga Islands are not surpassed by any others in the tropical world ; and with the climate they enjoy — that of an endless summer — they might well answer to the description of the " abode of the Blessed." And, indeed, when Tasman first looked upon these islands, they perhaps merited the title more than any other spot on the habitable globe ; for, if any people on this earth might be esteemed happy and blessed, Enrely it was the inhabitants of these fair isles of the far Southern Sea. Tasman even records the remarkable feet, that he saw no arms among them, — no weapons of war ! and perhaps, at that time, neither the detestable trade nor its implements ^ere known to them. Alas 200 THE TONGANS, OR in little more than a century afterwards, this peajefu aspect was no longer presented. When the great Eng- lish navigator visited these islands, he found the war- club and spear in the hands of the people, both ot Feegee pattern, and undoubtedly of the same ill-omened origin. The personal appearance of the Friendly Islanders differs not a great deal from that of the other South-Sea tribes or nations. Of course we speak only of the true Polynesians of the brown complexion, without reference to the black-skinned islanders — as the Feegees and others of the Papuan stock. The two have neither re- semblance nor relationship to one another ; and it would not be difficult to show that they are of a totally distinct origin. As for the blacks, it is not even certain that they are themselves of one original stock ; for the splen- didly-developed cannibal of Feegee presents very few features in common with the wretched kangaroo-eater of West Australia. Whether the black islanders (or Melanesians as they have been designated) originally came from one source, is still a question for ethnolo- gists ; but there can be no doubt as to the direction whence they entered upon the colonization of the Pa- cific. That was certainly upon its western border, be- yond which they have not made much progress : since the Feegeean archipelago is at the present time their most advanced station to the eastward. The brown or Polynesian races, on the contrary, began their migra- tions from the eastern border of the great ocean — in other words, they came from America ; and the so- called Indians of America are, in my opinion, the pro* genxtors, not the descendants, of thes<3 people of thfl FftlKNbLY ISLANDERS. 201 O'-ean world. If learned ethnologists will give their attention to this view of the subject, and disembarrass their minds of that fabulous old fancy, about an original stock situated somewhere (they know not exactly where) upon the steppes of Asia, they will perhaps arrive at a more rational hypothesis about the peopling of the so- called new worlds, both the American and Oceanic They will be able to prove — what might be here done if space would permit — that the Polynesians are emi- grants from tropical America, and that the Sandwich Islanders came originally from California, and not the Californians from the island homes of Hawaii. It is of slight importance here how this question may be viewed. Enough to know that ,the natives of the Tonga group bear a strong resemblance to those of the other Polynesian archipelagos — to the Otaheitans and New Zealanders, but most of all to the inhabitants of the Samoan or Navigators' Islands, of whom, indeed, they may be regarded as a branch, with a separate political and geographical existence. Their language also confirms the affinity, as it is merely a dialect of the common tongue spoken by all the Polynesians. Whatever difference exists between the Tongans and other Polynesians in point of personal appearance, is in favor of the former. The men are generally regarded as the best-looking of all South-Sea Islanders, and the women among the fairest of their sex. Many of them would be accounted beautiful in any part of the world j and as a general rule, they possess personal beauty in a far higher degree than the much-talked-of Otaheitans. The Tongans are of tall stature — rather above than under that of European nations. Men of six feet are 202 THE TONGANS, OR common enough ; though few are seen of what might be termed gigantic proportions. In fact, the true medium size is almost universal, and the excess in either direc- tion forms the exception. The bulk of their bodies is m perfect proportion to their height. Unlike the bkek Feegeeans — who are often bony and gaunt — the Tou- gans possess well-rounded arms and limbs ; and the hands and feet, especially those of the women, are small and elegantly shaped. To give a delineation of their features would be a difficult task — since these are so varied in different individuals, that it would be almost impossible to select a good typical face. Indeed the same might be said of nearly every nation on the face of the earth ; and the difficulty will be understood by your making an attempt to describe some face that will answer for every set of features in a large town, or even a small village ; or still, with greater limitation, for the different individuals of a single family. Just such a variety there will be found among the faces of the Friendly Islanders, as you might note in the inhabitants of an English town or county; and hence the difficulty of making a correct likeness. A few characteristic points, however, may be given, both as to their features and complexion. Their lips are scarcely ever of a thick or negro form ; and although the noses are in general rounded at the end, this rule is not universal ; — many have genuine Roman coses, and what may be termed a full set of the best Italian features. There is also less difference between the sexes in regard to their features than is usually seen elsewhere — those of the women being only dis- tinguished by their less size. FRIENDLY ISLANDERS. 20S The forms of the women constitute a more marked distinction ; and among the beauties of Tonga are many that might be termed models in respect to shape and proportions. In color, the Tongans are lighter than most other South-Sea Islanders. Some of the better classes of women — those least exposed to the open air — show skins of a light olive tint; and the children of all are nearly white after birth. They become browner less from age than exposure to the sun ; for, as soon as they are able to be abroad, they scarce ever afterwards enter under the shadow of a roof, except during the hours of night. The Tongans have good eyes and teeth ; but in this respect they are not superior to many other Oceanic tribes — even the black Feegeeans possessing both eyes and " ivories " scarce surpassed anywhere. The Ton- gans, however, have the advantage of their dusky neigh- bors in the matter of hair — their heads being clothed with a luxuriant growth of true hair. Sometimes it is quite straight, as among the American Indians, but oftener with a slight wave or undulation, or a curl ap- proaching, but never quite arriving at the condition of " crisp." His hair in its natural color is jet black ; and it is to be regretted that the Tongans have not the good taste to leave it to its natural hue. On the contrary, their fash- ion is to stain it of a reddish-brown, a purple or an orange. The brown is obtained by the application of burnt coral, the purple from a vegetable dye applied poultice-fashion to the hair, and the orange is produced by a copious lathering of common turmeric, — with which the women also sometimes anoint their bodies, and those 204 THE TONGANS, OR of their children. This fashion of hair-dyeing ia alsc common to the Feegees, and whether they obtained it from the Tongans, or the Tongans ft om them, is an un- settled point. The more probable hypothesis would be, that among many other ugly customs, it had its origin in Feegee-land, — where, however, the people as3ign a reason for practising it very different from the mere motive of ornament. They allege that it also serves a useful purpose, in preventing the too great fructification of a breed of parasitic insects, — that would other- wise find the immense mop of the frizzly Feegeean a most convenient dwelling-place, and a secure asylum from danger. This may have had something to do with the origin ot the custom ; but once established for pur- poses of utility, it is now confirmed, and kept up by the Tongans as a useless ornament. Their taste in the color runs exactly counter to that of European fashionables. What a pity.it is that the two could not make an ex- change of hair ! Then both parties, like a pair of adver« tisements in the " Times," would exactly fit each other. Besides the varied fashion in colors, there is also great variety in the styles in which the Tongans wear their hair. Some cut it short on one side of their head, leav- ing it at full length on the other ; some shave a smaD patch, or cut off only a single lock ; wliile others — and these certainly display the best taste — leave it to grew out in all its full luxuriance. In this, again, we find the European fashion reversed, for the women are those who wear it shortest. The men, although they are not with- out beard, usually crop this appendage very close, or shave it off altogether, — a piece of shell, or rather a pair of shells, serving them for a razor. FRIENDLY ISLANDERS. 205 The mode is to place the thin edge of one .-jhell un derneath the hair, — just as a hair-cutter does his comb, — and with the edge of the other applied above, the hairs are rasped through and divided. There are regu- lar barbers for this purpose, who by practice have been rendered exceedingly dexterous in its performance ; and the victim of the operation alleges that there i? little or no pain produced, — at all events, it does no', biing the tears to his eyes, as a dull razor often does with us poor thin-skinned Europeans ! The di-ess of the Tongans is very similar to that of the Otaheitans, so often described and well known ; but we cannot pass it here without remarking a notable peculiarity on the part of the Polynesian people, as exhibited in the character of their costume. The native tribes of almost all other warm climates content them- selves with the most scant covering, — generally with no covering at all, but rarely with anything that may be termed a skirt. In South America most tribes wear the " guayuco," — a mere strip around the loins, and among the Feegees the " malo " or " masi " of the men, and the scant " liku " of the women are the only excuse for a modest garment. In Africa we find tribes equally des- titute of clothing, and the same remark will apply to the tropical countries all around the glol)e. Here, however, amongst a people dwelling in the middle of a vast ocean, — isolated from the whole civilized world, we find a nat ural instinct of modesty that does credit to their character, and is even in keeping with that character, as first ob- served by voyagers to the South Seas. Whatever acts of indelicacy may be alleged against the Otaheitans, this has been much exaggerated by their interc* urse with immoral 206 THE TONGANS, OR white men; but none of such criminal conduct can be charged against the natives of the Friendly Isles. Ob the contrary, the behavior of these, both among them« selves and in presence of European visitors, has been ever characterized by a modesty that would shame either Regent Street or Ratcliffe Highway. A description of the national costume of the Tongans, though often given, is not unworthy of a place here; and we shall give it as briefly as a proper understanding of it will allow. There is but one " garment " to be de- scribed, and that is the " pareu," which will be better understood, perhaps, by calling it a " petticoat." The material is usually of " tapa " cloth, — a fabric of native manufacture, to be described hereafter, — and the cut- ting out is one of the simplest of performances, requiring neither a tailor for the men, nor a dressmaker for the other sex, for every one can make their own pareu. It needs only to clip a piece of " tapa " cloth in the form of an "oblong square" — an ample one, being about two yards either way. This is wrapped round the body, — the middle part against the small of the back, — and then both ends brought round to the front are lapped over each other as far as they will go, producing, of course, a double fold of the cloth. A girdle is next tied around the waist, — usually a cord of ornamental plait ; and this divides the piece of tapa into body and skirt. The latter is of such a length as to stretch below the calf of theVleg, — sometimes down to the rnkle, — and the upper part or body would reach to the shoulders, if the weather required it, and often does when the ?nis- itonaries require it. But not at any other time : such au ungraceful mode of wearing the pareu was nevei FRIENDLY ISLANDERS. 207 intended by the simple Tongans, who never dreamt of there being an\ immodesty in their fashion until told of it by their puritanical preceptors ! Tongan-fashion, the pareu is a sort of tunic, and a most graceful garment to boot ; Methodist fashion, it becomes a gown or rather a sleeveless wrapper that re- sembles a sack. But if the body part is not to be used in this way r how, you will ask, is it to be disposed of? Is it allowed to hang down outside, like the gown of a slattern woman, who has only half got into it ? No such thing. The natural arrangement is both simple and peculiar ; and produces, moreover, a costume that is not only characteristic but graceful to the eye that once be- comes used to it. The upper half of the tapa cloth is neatly folded or turned, until it becomes a thick roll; and this roll, brought round the body, just above the girdle, is secured in that position. The swell thus pro- duced causes the waist to appear smaller by contrast ; and the effect of a well-formed bust, rising above the roll of tapa cloth, is undoubtedly striking and elegant In cold weather, but more especially at night, the roll is taken out, and the shoulders are then covered ; for it is to be observed that the pareu, worn by day as a dress, is also, kept on at night as a sleeping-gown, more especially by those who possess only a limited wardrobe. It is not always the cold that requires it to be kept on at night. It is more used, at this time, as a protection against the mosquitoes, that abound amidst the luxuriant vegetation of the Tongan Islands. The " pareu " is not always made of the " tapa ' 1 cloth. Fine mats, woven from the fibres of the screw- pine \j>andanus), are equally in vogue ; and, upon fed 208 THE TONGANS, OR tivc occasions, a full-dress pareu is embellished with red feather-work, adding greatly to the elegance and pic- turesqueness of its appearance. A coarser and scantier pareu is to be seen among the poorer people, the mate* rial of which is a rough tapa, fabricated from the bark of the bread-fruit, and not unfrequently this is only a mere strip wrapped around the loins ; in other words, a u malo," " maro," or " maso," — as it is indifferently written in the varied orthography of the voyagers. Having described this only and unique garment, we have finished with the costume of the Tongan Islanders, both men and women, — for both wear the pareu alike. The head is almost universally uncovered ; and no head-dress is ever worn unless a cap of feathers by the great chiefs, and this only upon rare and grand occa- sions. It is a sort of chaplet encircling the head, and deeper in front than behind. Over the forehead the plumes stand up to a height of twelve or fifteen inches, gradually lowering on each side as the ray extends backward beyond the ears. The main row is made with the beautiful tail-plumes of the tropic bird Phaeton cethereusj while the front or fillet part of the cap is ornamented with the scarlet feathers of a species of parrot. The head-dress of the women consists simply of fresh flowers : a profusion of which — among others the beau- tiful blossoms of the orange — - is always easily obtained. An ear-pendant is also worn, — a piece of ivory of about two inches in length, passed through two holes, pierced in the lobe of the ear for this purpose. The pendant hangs horizontally, the two holes balancing it, and keeping it in position. A necklace also of pearl* FRIENDLY ISLANDERS. 2 r K jhells, shaped into beads, is worn. Sometimes a string of the seeds of the pandanus is added, and an additional ornament is an armlet of mother-o' -pearl, fashioned into the form of a ring. Only the men tattoo themselves ; and the process is confined to that portion of the body from the waist to the thighs, which is always covered with the pareu. The practice of tattooing perhaps first originated in the desire to equalize age with youth, and to hide an ugly physiognomy. But the Tongan Islander has no ugliness to conceal, and both men and women have had the good taste to refrain from disfiguring the fair features which nature has so bountifully bestowed upon tKem. The only marks of tattoo to be seen upon the women are a few fine lines upon the palms of their hands ; nor do they disfigure their fair skins with the hideous pigments so much in use among other tribes, of vhat we are in the habit of terming savages. They anoint the body with a fine oil procured from the cocoanut, and which is also perfumed by various kinds of flowers that are allowed to macerate in the oil ; but this toilet is somewhat expensive, and is only practised by the better classes of the community. All, however, both rich and poor, are addicted to habits of extreme cleanliness, and bathing in fresh water is a frequent performance. They object to bathing in the sea ; and when they do so, always finish the bath by pouring fresh water over their bodies, — a practice which they allege prevents the skin from becoming rough, which the sea-water would otherwise make it. House architecture in the Tongan Islands is in rather a backward state. They have produced no Wrens noi Inigo Joneses ; but this arises from a natural cause 210 fHE TONGANS, OR They have no need fcr great architects, — • scarce anj deed for houses either, — and only the richer Tongans erect any dwelling more pretentious than a mere shed A few posts of palm-trunks are set up, and upon these are placed the cross-beams, rafters, and roof. Pandanus leaves, or those of the sugar-cane, form the thatch ; and the sides are left open underneath. In the houses of the chiefs and more wealthy people there are walls of pandanus mats, fastened to the uprights ; and some of these houses are of considerable size and neatly built. The interiors are kept scrupulously clean, — the floors being covered with beautiful mats woven in colored patterns, and presenting all the gay appearance of costly carpeting. There are neither chairs nor tables. The men sit tailor-fashion, and the women in a reclining posture, with both limbs turned a little to one side and backwards. A curious enclosure or partition is formed by setting a stiff mat, of about two feet width, upon its edge, — the roll at each end steadying it and keeping it hi an upright position. The utensils to be observed are dishes, bowls, and cups, — usually of calabash or cocoa-shells, — and an endless variety of baskets of the most ingenious plait and construction. The " stool-pillow " is also used ; but differing from that of the Fee^ees in the horizontal piece having a hollow to receive the head. Many kinds of musical instruments may be seen, — the Pan- dean pipes, the nose-flute, and various kinds of bamboo drums, all of which have been minutely described by travellers. I am sorry to add that war-clubs and spears for a similar purpose are also to be observed conspicu ous among the more useful implements of peace. Bowi FKIENDLY ISLANDERS. 211 and arrows, too, are common ; but these are only em- ploy 3d for shooting birds and small lodents, especially rats, that are very numerous and destructive to the crops. For food, tho Tongans have the pig, — the same variety as is so generally distributed throughout the Oceanic Islands. It is stated that the Feegeeans ob- tained tins animal from the Friendly Isles ; but I am of opinion that in this case the benefit came the othei way, as the Sus Papua is more likely to have entered the South Sea from its leeward rather than its wind- ward side. In all likelihood the dog may have been derived from the eastern edge ; but the pigs and poultry would seem to be of western origin, — western as re- gards the position of the Pacific. The principal food of the Friendly Islanders, how- ever, is of a vegetable nature, and consists of yams, breadfruit, taro, plantains, sweet potatoes, and, in fact, most of those roots and fruits common to the other islands of the Pacific. Fish also forms an important article of their food. They drink the " kava," or juice of the Piper methisticum — or rather of its roots chewed to a pulp ; but they rarely indulge to that excess ob- served among the Feegees, and they are not over fond of the drink, except as a means of producing ft species of intoxication which gives them a momentary pleasure. Many of them, especially the women, make wry faces while partaking of it ; and no wonder they do, for it ia at best a disgusting beverage. The time of the Tongan Islanders is passed pleasantly enough, when there is no wicked war upon hand. The men employ themselves in cultivating the groin d o* 212 THE TOtfGANS, OK fishing ; and here the woman is no longer the mera slave and drudge — as almost universally elsewhere among savage or even semi-civilized nations. This is a great fact, which tells a wondrous tale — which speaks trumpet-tongued to the credit of the Tongan Islander. Not only do the men share the labor with their more delicate companions, but everything else — their food, conversation, and every enjoyment of life. Both par- take alike — eat together, drink together, and join at once in the festive ceremony. In their grand dances — or balls as they might more properly be termed — the women play an important part ; and these exhibi- tions, though in the open air, are got up with an ele- gance and eclat that w T ould not disgrace the most fash- ionable ball-room in Christendom. Their dances, in- deed, are far more graceful than anything ever seen either at " Almacks " or the " Jarclin MabiUe." The principal employment of the men is in the cul- tivation of their yam and plantain grounds, many of which extend to the size of fields, with fences that would almost appear to have been erected as orna- ments. These are of canes, closely set, raised to the height of six feet — wide spaces being left betw r een the fences of different owmers to serve as roads for the whole community. In the midst of thes^ fields stand the sheds, or houses, surrounded by splendid forms of tropic vegetation, and forming pictures of a softly beau tiful character. The men also occupy themselves in the construction of their canoes, — to procure the large ones, making a voyage as already stated, to the Feegee Islands, and sometimes remaining absent for several years. FRIENDLY ISLANDERS. 213 Thsse, however, are usually professional boat-builders, and form but a very small proportion of the forty thou- sand people who inhabit the different islands of the Ton gan archipelago. The men also occasionally occupy themselves in weav- ing mats and wicker baskets, and carving fancy toys out of wood and shells ; but the chief part of the manufac- turing business is in the hands of the women — more especially the making of the tapa-cloth, already so often mentioned. An account of the manufacture may be here introduced, with the proviso, that it is carried on not only by the women of the Feegee group, but by those of nearly all the other Polynesian Islands. There are slight differences in the mode of -manufacture, as well as in the quality of the fabric ; but the account here given, both of the making and dyeing, will answer pretty nearly for all. The bark of the malo-tree, or " paper-mulberry," is taken off in strips, as long as possible, and then steeped in water, to facilitate the separation of the epidermis, which is effected by a large volute shell. In this state it is kept for some time, although fit for immediate use. A log, flattened on the upper side, is so fixed as to spring a little, and on this the strips of bark — or masi, as it is called — are beaten with an iki, or mallet, about two inches square, and grooved longitudinally on three of its sides. Two lengths of the wet masi are generally beaten together, in order to secure greater strength — the gluten which they contain being sufficient to keep their fibres united. A two-inch strip can thus be beaten out to the w r idth of a foot and a half; but the length is at the same time reduced. The pieces are neatly 214 THE 10XGANS, OR lapped together with the starch of the taro, or arrow-root, boiled whole ; and thus reach a length of many yards, The " widths " are- also joined by the same means later- ally, so as to form pieces of fifteen or thirty feet square s and upon these, the ladies exhaust their ornamenting skill. The middle of the square is printed with a red- brown, by the following process : — Upon a convex board, several feet long, are arranged parallel, at about a finger-width apart, thin straight slips of bamboo, a quarter of an inch wide. By the side of these, curved pieces, formed of the mid-rib of cocoanut leaflets, are arranged. On the board thus prepared the cloth is laid, and rubbed over with a dye obtained from the laud (Aleurites triloba). The cloth of course, takes the dye ^\pon those parts which receive pressure, being sup- ported by the slips beneath ; and thus shows the same pattern in the color employed. A stronger preparation of the same dye, laid on with a sort of brush, is used to divide the square into oblong compartments, with large lound or radiated dots in the centre. The kesa, or dye, when good, dries bright. Blank borders, two or three feet wide, are still left on two sides of the square ; and to elaborate the ornamentation of these, so as to excite applause, is the pride of every lady. There is now an entire change of apparatus. The operator works on a plain board ; the red dye gives place to a jet black ; the pattern is now formed of a strip of banana-loaf placed on the upper surface of the cloth. Out of the leaf is cut the pattern — not more than an inch long — which the lady wishes to print upon the border, and holds by tier first and middle finger, pressing it down with the thumb. Then taking a soft pad of cloth, steeped in die FRIENDLY ISLANDERS. 215 dye, in her right hand, she rubs it firmly over the stencil, and a (air, sharp figure is made. The practised fingers of the operator move quickly, but it is, after all, a te- dious process. I regret to add, that the men employ themselves in an art of less utility: the manufacture of war weapons — clubs and spears — which the people of the different islands, and even those of the same, too often brandish against one another. This war spirit is entirely owing to their intercourse" with the ferocious Feegees, whose boasting and ambitious spirit they, are too prone to emu- late. In fact, their admiration of the Feegee habits is something surprising ; and can only be accounted for by the fact, that while visiting these savages and professed warriors, the Tongans have become imbued with a cer- tain fear of them. They acknowledge the more reck- less spirit of their allies, and are also aware that in intellectual capacity the black men are not inferior to themselves. They certainly are inferior in courage, as in every good moral quality ; but the Tongans can hard- ly believe this, since their cruel and ferocious conduct seems to give color to the contrary idea. In fact, it is this that inspires them with a kind of respect, which has no other foundation than a vague sense of fear. Hence they endeavor to emulate the actions that produce this fear, and this leads them to go to war with one another. It is to be regretted that the missionaries have sap- plied them with a motive. Their late wars are solely due to missionary influence, — for Methodism upon the Toagan Islands has adopted one of the doctrine? of Mahomet, and believes in the faith being propagated by the fsword A usurper, who wishes to be king over the 216 THE TONGANS, OR whole group, his embraced the Methodist form of Chris tianity, and linked himself with its teachers, — who offer to aid him with all their influence ; and these formerly peaceful islands now present the painful spectacle of a divided nationality. — the " Christian party," and the " DeviFs party." The object of conquest on the part of the former is to place the DeviFs party under the absolute sovereignty of a despot, whose laws will be dictated by his missionary ministers. Of the mildness of these laws we have already some specimens, which of course extend only to the " Christianized." One of them, which refers to the mode of wearing the pareu, has been already hinted at, — and another is a still more off-hand piece of legislation : being an edict that no one hereafter shall be permitted to smoke tobacco, under pain of a most severe punishment. When it is considered that the Tongan Islander enjoys the " weed " (and grows it too) more than almost any other smoker in creation, the severity of the " taboo " may be understood. But it is very certain, if his Metho- dist majesty were once firmly seated on his throne, bluet laws than this would speedily be proclaimed. The American Commodore Wilkes found things in this war- like attitude when he visited the Tongan Islands ; but perceiving that the right was clearly on the side of the " DeviFs party," declined to interfere ; or rather, his interference, which would Lave speedily brought peace, was rejected by the Christian party, instigated by the sanguinary spirit of their " Christian " teachers. Not so, Captain Croker, of Her Britannic Majesty's service, who came shortly after. This unreflecting officer — loath to believe that royalty could be in the wrong — WWII a^ 'II Vila! FRIENDLY ISLANDERS ?2- at once took side with the king and Christians, an dashed headlong into the affair. The melancholy resulf is well known v It ended by Captain Croker leaving his body upon the field, alongside those of many of his bra?e tars ; and a disgraceful retreat of the Christian party beyond the reach of their enemies. This interference of a British war-vessel in the affairs of the Tongan Islanders, offers a strong contrast to our conduct when in presence of the Feegees. There we have the fact recorded of British officers being eye- witnesses of the most horrid scenes, — wholesale mur- der and cannibalism, — with full power to stay the crime and full authority to punish it, — that authority which would have been freely given them by the accord and acclamation of the whole civilized world, — and yet they stood by, in the character of idle spectators, fearful of breaking through the delicate icy line of non-inter" vention ! A strange theory it seems, that murder is no longer murder, when the murderer and his victim chance to be of a different nationality from our own ! It is a distinc- tion too delicate to bear the investigation of the philo- Bophic mind ; and perhaps will yet yield to a truer appre- ciation of the principles of justice. There was no such squeamishness displayed when royalty required support upon the Tongan Islands ; nor ever is there when self- interest demands it otherwise. Mercy and justice may both fail to disarrange the hypocritical fallacy of non- intervention ; but the principle always breaks down at the call of political convenience. THE TURCOMANS. Asia has been remarkable, from the earliest fcme^ for having a large population without ,iny fixed place of residence, but who lead a nomade or wandering life. It is not the only quarter of the globe where this kind of people are found : as there are many nomade nations in Africa, especially in the northern division of it ; and if we take the Indian race into consideration, we find that both the North and South American continents have their tribes of wandering people. It is in Asia, nevertheless, that we find this unsettled mode of life carried out to its greatest extent, — it is there that we find those great pastoral tribes, — or " hordes," as they have been termed, — who at different historical periods have not only increased to the numerical strength of large nationalities, but have also been powerful enough to overrun adjacent empires, pushing their conquer even into Europe itself. Such were the invasions of the Mongols under Zenghis Khan, the Tartars under Ti- mour, and the Turks, whose degenerate descendants now so feebly hold the vast territory won by iheii wandering ancestors. The pastoral life, indeed, has its charms, that render THE TURCOMANS. 219 it atiinctive to the natural disposition of man, and wher- ever the opportunity offers of following it, this life will I e preferred to any other. It affords to man an abun- dant supply of all his most prominent wants, without requiring from him any very severe exertion, either of mind or body ; and, considering the natural indolence of Asiatic people, it is not to be wondered at that so many of them betake themselves to this mode of existence. Their country, moreover, is peculiarly favorable to the development of a pastoral race. Perhaps not one third of the surface of the Asiatic continent is adapted to agriculture. At least one half 'of it is occupied by tree- less, waterless plains, many of which have ail the char- acters of a desert, where an agricultural people could not exist, or, at all events, where their labor would be rewarded by only the most scant and precarious returns. Even a pastoral people in these regions would find but a sorry subsistence, were they confined to one spot ; for the luxurious herbage which, for the most part, char- acterizes the great savanna plains of America, is either altogether wanting upon the steppes of Asia, or at best very meagre and inconstant. A fixed abode is therefore impossible, except in the most fertile tracts or oases: elsewhere, the nomad life is a necessity arising from the circumstances of the soil. It would be difficult to define exactly the limits of the territory occupied by the wandering races in Asia ; but in a general way it may be said that the whole central portion of the continent is thus peopled : indeed, much more than the central portion, — for, if we except the rich agricultural countries of Hindostan and China. wilJh 220 THE TURCOMANS. a small portion of Persia, Arabia, and Turkey, the whol« of Asia is of this character. The countries known as Balk and Bokara, Yarkand and Khiva, with several others of equal note, are merely the central points of oases, — large towns, supported rather by commerce than by the produce of agriculture, and having nomad tribes dwelling within sight of their walls. Even the present boundaries of Asiatic Turkey, Arabia and Persia, contain within them a large proportion of nomadic popu- lation ; and the same is true of Eastern Poland and Russia in Europe. A portion of the AfFghan and Belo- ehee country is also inhabited by nomad people. These wandering .people are of many different types and races of men ; but there is a certain similarity in the habits and customs of all : as might be expected from the similar circumstances in which they are placed. It is always the more sterile steppes that are thus oc- cupied ; and tins is easily accounted for : where fertile districts occur the nomad life is no longer necessary. Even a wandering tribe, entering upon such a tract, would no longer have a motive for leaving it, and would soon become attached to the soil, — in other words, would cease to be wanderers ; and whether they turned their attention to the pursuit of agriculture, or not, they would be certain to give up their tent-life, and fix themselves in a permanent abode. This has been the history of many Asiatic tribes ; but there are many others, again, who from time immemorial, have shown a repugnance to the idea of fixing themselves to the soil. They prefer the free roving life which the desert enables them to indulge in ; and wandering from place to place as the thoice of posture guides them, occupy themselves en- 1HE TURCOMANS. 221 &i jly in feeding their flocks and herds, — the sole means of their subsistence. These never have been, and nevef co lid be, induced to reside in towns or villages. Nor is it that they have been driven into these desert tracts to seek shelter from political oppression, — as is the case with some of the native tribes of Africa and America. On the contrary, these Asiatic nomads are more often the aggressors than the objects of aggression. It is rather a matter of choice and propensity with them : as with those tribes of the Arabian race, — known as " Bedouins." The proportion of the Asiatic wandering population to those who dwell in towns, or fixed habitations, varies according to the nature of the country. In many ex- tensive tracts, the former greatly exceed the latter ; and the more sterile steppes are almost exclusively oc- cupied by them. In general, they acknowledge the sov- ereignty of some of the great powers, — such as the empires of China, Russia, and Turkey, the kingdom of Persia, or that of several powerful khans, as those of Khiva and Bokara ; but this sovereignty is, for the most part, little more than nominal, and their allegiance is readily thrown off, whenever they desire it. It is rarely so strong, as to enable any of the aforesaid powers to draw a heavy tribute from them ; and some of the more warlike of the wandering tribes are much courted and caressed, — especially when their war services are re- quired. In general they claim an herditary right to the territories over which they roam, and pay but little heed *o the orders of either king, khan, or emperor. As ilready stated, these wandering people are of dif- ferent races ; in fact, they are of nearly all the varictiei 222 THE 1JKC0MANS. indigenous to the Asiatic continent ; and a whole cata logue of names might be given, of which Mongols, Tar tars, Turcomans, Usbecks, Kirghees, and Calmucks, are perhaps the most generally known. It has been also stated that in many points they are alike ; but there are also many important particulars in which they differ, — physical, moral, and intellectual. Some of the " hordes," or tribes, are purely pastoral in their mode of life, an^ of mild and hospital dispositions, exceedingly fond of strangers, and kind to such as come among them. there again are averse to all intercourse with others, than those of their own race and religion, and are shy, if not in- hospitable, when visited by strangers. But there is a class of a still less creditable character, — a large num- ber of tribes that are not only inhospitable, and hostile to strangers, but as ferocious and bloodthirsty as any savages in Africa, America, or the South-Sea Islands. As a fair specimen of this class we select the Turco- mans ; in fact, they may be regarded as its type ; and our description henceforward may be regarded as apply- ing particularly to these people. The country of the Turcomans will be found upon the map without difficulty ; but to define its exact boundary would be an impossibility, since none such exists. Were you to travel along the whole northern frontier of Persia, almost from the gates of Teheran to the eastern frontier of the kingdom, — or even farther towards Balk, — you would be pretty sure of hearing of Turcoman robbers, and in very great danger of being plundered by them,—- which last misfortune would be of less importance, as it would only be the prelude to your being either murdered on the spot, or carried o/f by them into captivity. In THE TURCOMANS. 223 making tins journey along the northern frontier of Per- sia, you would become acquainted with the whereabouts of the Turcoman hordes ; or rather you would discover that the whole north part of Persia, — a good broad band of it extending hundreds of miles into its interior, — if not absolutely in possession of the Turcomans, ia overrun and plundered by them at will. This, however, is not their home, — it is only their " stamping-ground," — the home of their victims. Their place of habitual residence lies further to the north, and is defined with tolerable accuracy by its having the whole eastern shora of the Caspian Sea for its western border, while the Amou River (the ancient Oxus) may be generally re- garded as the limit of their range towards the east. Some tribes go still further east than the Amou ; but those more particularly distinguished for their plunder- ing habits dwell within the limits described, — north of the Elburz Mountains, and on the great steppe of Kau- rezm, where they are contiguous to the Usbeck com- munity of Khiva. The whole of this immense territory, stretching from die eastern shore of the Caspian to the Amou and Aral Se&, may be characterized as a true desert. Here and there oases exist, but none of any importance, save the country of Khiva itself: and even that is but a mere irrigated strip, lying on both banks of the Oxus. In- deed, it is difficult to believe that this territory of Khiva, go insignificant in superficial extent, could have been the »eat of a powerful empire, as it once was. The desert, then, between the Caspian Sea and the Oxus River may be regarded as the true land of the Turcomans, and is usually known as T*j comania. It /•84 THE TURCOMANS. is to \yn rciwe.TkVred, however, that there are L*ome kindred i-*ioes noi included within the boundaries of Turcomania — for J.e Turkistan of the geographers is a country of much larger extent ; besides, an important division of the Turcoman races are settlers, or rather wanderers in Armenia. To Turcomania proper, then, and its inhabitants, we shall confine our remarks. We shall not stay to inquire into the origin of the people now called Turcomans. Were we to speculate upon that point, we should make but little progress in an account of their habits and mode of living. They are usually regarded as of Tartar origin, or of Usbeck origin, or of Mongolian race ; and in giving this ac- count of them, I am certain that I add very little to your knowledge of what they really are. The truth is, that the words Tartar and Mongol and some half-dozen othei titles, used in relation to the Asiatic races, are without any very definite signification, — simply because the relative distinctions of the different nations of that continent are very imperfectly known ; and learned eth- nologists are ever loath to a confession of limited knowl- edge. One of this class, Mr. Latham, — who requires only a few words of their language to decide categori- cally to what variety of the human race a people be- longs, — has unfortunately added to this confusion by pronouncing nearly everybody Mongolian: placing the proud turbaned Turk in juxtaposition with the squat and stunted Laplander ! Of course this is only bring- ing us back to the old idea, that all men are sprung from a single pair of first parents, — a doctrine, which, though popular, is difficult to reconcile with the rational knowledge derived from ethnological in estimation. THE TURCOMANS. 225 It matter & little to our present purpose from what anginal race the Turcoman has descended : whether he be a true Turk, as some regard him, or whether he is a descendant of the follows of the Great Khan of the Tartars. He possesses the Tartar physiognomy to a considerable extent — some of the tribes more than others being thus distinguished, — and high cheek-bones, liat noses, small oblique eyes, and scanty beards, are all characteristics that are very generally observed. Some of these peculiarities are more common among the women than the men — many of the latter being tall, 6tout, and well-made, while a large number may be seen who have the regular features of a Persian. Per- haps it would be safest to consider the present Turco- man tribes as not belonging to a pure stock, but rather an admixture of several ; and their habit of taking slaves from other nations, which has for a long time existed among them, would give probability to this idea. At all events, without some such hypothesis, it is diffi- cult to account for the wonderful variety, both in feature and form, that is found among them. Their complexion is swarthy, in some cases almost brown as that of an American Indian ; but constant exposure to the open air, in all sorts of weather, has much to do in darken- ing the hue of their skin. The newborn children are nearly as white as those of the Persians; and their young girls exnibit a ruddy brunette tint, which some consider even more pleasing than a perfectly white com- plexion. The costume of the Turcoman, like that of most Ori- ental nations, is rich and picturesque. The dress of the men varies according to rank. Some of the very pooref 226 THE TURCOMANS. people wear nothing but a short woollen tunic oi shirty with a pair of coarse woollen drawers. Others, in place of this shirt, are clad in a longer garment, a sort of robe or wrapper, like a gentleman's dressing-gown, made of camers-hair cloth, or some coarse brown woollen stuff. But the true Turcoman costume, and that worn by all who can afford it, consists of a garment of mixed silk and cotton, — the baronnee, — which descends below the knee, and though open in front, is made to button over the breast quite up to the neck. A gay sash around the waist adds to 'Le effect ; and below the skirt are seen trow r sers of cotton or even silk. Cloth wrappers around the legs serve in the place of boots or gaiters ; and on the feet are worn slippers of Persian fashion, with socks of soft Koordish leather. As the material of which the baronnee is made is of good quality — a mixture of silk and cotton — and as the fabric is always striped or checkered in colors of red, blue, purple, and green, the effect produced is that of a certain picturesqueness. The head-dress adds to this appearance — being a high fur cap, with truncated top, the fur being that beautiful kind obtained from the skins of the Astracau lamb, well known in commerce. These caps are of different colors, either black, red, or gray. Another style of head-dress much worn is a round-topped or helmet-shaped cap, made of quilted cotton-stuff j but this kind, although in use among the Turcomans, is a more characteristic costume of their enemies, the " Koords," who wear it universally. The " jubba" is a kind of robe generally "intended to go over the other garments, and is usually of woollen >r camers-hair cloth. It is also made like a dressing THE TURCOMANS. 227 gown, with wide sleeves, — tight, however, around the wrist. It is of ample dimensions, and one side is lapped over the other across the front, like a double-breasted coat. The " jubba " is essentially a national garment. The dress of the women is exceedingly picturesque. It is thus minutely described by a traveller : — " The head-dress of these women is singular enough : most of them wear a lofty cap, with a broad crown, resembling that of a soldier's cap called a shako. This is stuck upon the back of the head ; and over it is thrown a silk handkerchief of very brilliant colors, which covers the top, and falls down on each side like a veil. The front of this is covered with ornaments of silver and gold, in various shapes ; more frequently gold coins, mohrs, or tomauns, strung in rows, with silver bells or buttons, and chains depending from them ; hearts and other fanciful forms, with stones set in them. The whole gives rather the idea of gorgeous trappings for a horse, than ornaments for a female. "The frames of these monstrous caps ai v e made of light chips of wood, or split reeds, covered with cloth ; and when they do not wear these, they wrap a cloth around their heads in the same form ; and carelessly throw another, like a veil over it. The veil or curtain above spoken of covers the mouth ; descending to the breast. Earrings are worn in the ears ; and their lcng hair is divided, and plaited into four parts, disposed two on each side ; one of which falls down behind thu shoulders and one before, and both are strung with a profusion of gold ornaments, agates, cornelians, and other stones, according to the means and quality of the wearer. The rest of their dress consists of a long, loos* 228 THE TURCOMANS vest or shirt, with sleeves, which covers ths whole per* son down to the feet, and is open at the breast, in front, but buttons or ties close up to the neck : this is made of silk. 01 cotton-stuff, red, blue green, striped red, and yellow, checked, or various-colored : underneath this, are the zere-jameh, or drawers, also of silk or cotton ; and some wear a short peerahn or shirt of the same. This, I believe, is all ; but in the cold weather they wear, in addition, jubbas, or coats like those of the men, of striped stuff made of silk and cotton ; on their feet they generally wear slippers like those of the Persian women." The tents, or " portable houses " of the Turcomans — as their movable dwellings rather deserve to be called — differ from most structures of the kind in use else- where. They are thus described by the same intelligent traveller : — " The portable wooden houses of the Turcomans have been referred to by several writers ; but I am not aware that any exact description of their structure has been given. The frame is curiously constructed of light wood, disposed in laths of about an inch broad by three quarters thick, crossing one another diagonally, but at right angles, about a foot asunder, and pinned at each crossing with thongs of raw hide, so as to be movable ; and the whole framework may be closed up or opened in the manner of those toys for children that represent a company of soldiers, and close or expand at will, so as to form open or close column. " One or more pieces thus constructed being stretched out, surround a circular space of from fifteen .o twenty fset diameter ; and form the skeleton of the walls, — THE TURCOMANS. 229 which are made firm by bands of hair or woollen ropes, hitched round the end of each rod, to secure it in its po- sition. From the upper ends of these, rods c' a similar kind, bent near the wall end into somewhat lesc than a right angle, are so disposed that the longer portions slope to the centre, and being tied with ropes, «orm the frame- work of a roof. Over this is thrown a covering of black numud, leaving in the centre a large hole to give vent to the smoke, and light to the dwelling. Similar numuds are wrapped round the walls ; - and outside of these, to keep all tight, is bound another frame, formed of split reeds or cane, or of very light and tough wood, tied together with strong twine, the pieces being perpendicu- lar. This is itself secured by a strong, broad band of woven hair-stuff, which firmly unites. The large round opening at top is covered, as occasion requires, by a piece of numud, which is drawn off or on by a strong cord, like a curtain. If the wind be powerful, a stick ia placed to leeward, which supports the fabric. " In most of these houses they do not keep a carpel or numud constantly spread ; but the better classes us6 a carpet shaped somewhat in the form of a horseshoe, having the centre cut out for the fireplace, and the ends truncated, that those of inferior condition, or who do noi choose to take off their boots, may sit down upon th* ground. Upon this carpet they place one or two othei numuds, as may be required, for guests of distinction. TvTien they have women in the tent, a division of split reeds is made for their convenience ; but the richer people have a separate tent for their private apart- ments. " The fiirniture consists of little more than that of thf 230 THE TURCOMANS. camels and horses ; joals, or bags in which their go^dsi are packed, and which arfe often made of a very hand- some species of worsted velvet carpet, of Hen patterns the swords, guns, spears, bows and arrows, and dlbei implements of the family, with odds and ends of every description, may be seen hung on the ends of the wooden rods, which form very convenient pins for the purpose. Among some tribes all the domestic utensils are made of wood, — calleeoons, trays for presenting food, milk- vessels, &c. : among others, all these things are formed of clay or metal. Upon the black tops of the tents may frequently be seen large white masses of sour curd, expressed from buttermilk, and set to dry as future store ; this, broken down and mixed with water, forms a very pleasant acidulous drink, and is used as the basis of that intoxicating beverage called kimmiz. The most common and most refreshing drink which they offer to the weary and over-heated traveller in the forenoon is buttermilk, or sour curds and water ; and, indeed, a modification of this, with some other simple sherbets, are the only liquors presented at their meals. " Such are the wooden houses of the Turcomans, one of which just makes a camel's load. There are poorer ones, of a less artificial construction, the framework of which is formed of reeds. "The encampment is generally square, enclosing an open space, or forming a broad street, the houses being ranged on either side, with their doors towards each other. At these may always be seen the most pictu- resque groups, occupied with their various domestic duties, or smoking their simple wooden calleeoons. The more important encampments are often surrounded by THE TURCOMANS. 231 a fence of r *eds, which serve to protect the flocks frora petty thefts." It is now our place to inquire how the Turcomans occupy their time. We have already described them art a pastoral and nomadic people ; and, under ordinary circumstances, their employment consists in looking after their flocks. In a few of the more fertile oases they have habitations, or rather camps, of a more permanent character, where they cultivate a little corn or barley, to supply them with the material for bread ; but these set- tlements, if they deserve the name, are only exceptional ; and are used chiefly as a kind of head-quarters, where the women and property are kept, while the men them- selves are absent on their thieving expeditions. More generally their herds are kept on the move, and are driven from place to place at short intervals of a few weeks or even days. The striking and pitching of their tents gives them employment ; to which is added that of milking the cattle, and making the cheese and butter. The women, moreover, fill up their idle hours in weav- ing the coarse blankets, or " numuds," in plaiting mats, and manufacturing various articles of dress or household use. The more costly parts of their costume, however, are not of native manufacture : these are obtained by trade. The men alone look after the camels and horses, taking special care of the latter. Their flocks present a considerable variety of species. Besides horses, cattle, and sheep, they own manv camels, and they have no less than three distinct varieties of this valuable animal in their possession, — the dromedary with twe humps, and the common camel. The third sort is a cross breed — or "mule" — between these two. The 232 THE TUKCOMANS. dromedary is slightly made, and swifter than either of the others, but it is not so powerful as either . and being inferior as a beast of burden, is least cared for by the Turcomans. The one-humped camel is in more generaJ use, and a good one will carry a load of six or seven hundred pounds with ease. The mule camel is more powerful than either of its parents, and also more docile and capable of greater endurance. It grows to a very large size, but is low in proportion to its bulk, with stout, bony legs, and a large quantity of coarse, shaggy hair on its haunch, shoulders, neck, and even on the crown of its head, which gives it a strange, somewhat fantastic appearance. Its color varies from light gray to brown, though it is as often nearly black. This kind of camel will carry a load of from eight hundred to a thousand pounds. The Turcoman sheep are of the large-tailed breed, — their tails often attaining enormous dimensions. This variety of sheep is a true denizen of the desert, the fat tail being unquestionably a provision of nature against seasons of hunger, — just as in the single protuberance, or " hump," upon the camel. The horse of the Turcoman is the animal upon which he sets most value. The breed possessed by him is cele- brated over all Eastern Asia, as that of the Arab is in the West. They cannot be regarded, however, as hand- some horses, according to the true standard of " horae beauty ; " but the Turcomar cares less for this than for other good qualities. In point of speed and endurance) they are not excelled, if equalled, by the horses of any other country. Their size is that of the common English horse, but THE TURCOMANS 233 tbe/ aie veiy different in make. Their bodies are long in proportion to the bulk of carcass ; and they do not appear to possess sufficient compactness of frame. Their legs are also long, generally falling off in muscular de- velopment below the knee-joint ; and they would appear to an English jockey too narrow in the counter. They have also long necks, with large heavy heads. These are the points which are generally observed in the Tur- coman horses ; but it is to be remarked, that it is only when in an under-condition they look so ungraceful ; and in this condition their owners are accustomed to keep them, especially when they have any very heavy service to perform. Feeding produces a better shape, and brings them much nearer to the look of a well-bred English horse. Their powers of endurance are indeed, almost incredi- ble : when trained for a chappow, or plundering expedi- tion, they will carry their rider and provisions for seven or eight days together, at the rate of twenty or even thirty fursungs — that is, from eighty to one hundred miles — a day. Their mode of training is more like that of our pugilistic and pedestrian performers, than that adopted for race-horses. When any expedition of great length, and requiring the exertion of much speed, is in contemplation, they commence by running their horses every day for many miles together ; they feed them spar- ingly on barley alone, and pile numuds upon them at night to sweat them, until every particle of fat has buen removed, and the flesh becomes hard and tendonous. Of this they judge by the feel of the muscles, particu- larly on the crest, at the back of the neck, and on the haunches ; and when these are sufficiently firm and 234 THE TURCOMANS. hard, they say in praise of the animal, that " his flesh is marble." After this sort of training, the horse will proceed with expedition and perseverance, for almost any length of time, without either falling off in condition or knocking up, while horses that set out fat seldom sur- vive. They are taught a quick walk, a light trot, or a sort of amble, which carries the rider on easily, at the rate of six miles an hour ; but they will also go at a round canter, or gallop, for forty or fifty miles, without ever drawing bridle or showing the least symptom of fatigue. Their yaboos, or galloways, and large ponies are fully as remarkable, if not superior, to their horses, in their power of sustaining fatigue ; they are stout, compact, spirited beasts, without the fine blood of the larger breeds, but more within the reach of the poorer classes, and consequently used in by far greater numbers than the superior and more expensive horses. " It is a common practice of the Turcomans to teach their horses to fight with their heels, and thus assist their masters in the time of action. At the will of their riders they will run at and lay hold with their teeth of whatever man or animal may be before them. This acquirement is useful in the day of battle and plunder, for catching prisoners and stray cattle, but it at the same time renders them vicious and dangerous to be handled." In addition to the flocks and herds, the Turcomans possess a breed of very large fierce dogs, to assist them 11 keeping their cattle. These are also necessary as water -dogs, to protect the camp from thieves as well as more dangerous enemies to their peace ; and so weli trained are those faithful creatures, that it would to THE TURCOMANS. 235 impossible foi either friend or enemy to approach a Tur- coman camp without' the inmates being forewarned hi time Two or three of these dogs r„ay always be seen lying by the entrance of each tent ; and throughout the night several others keep sentry at the approaches to the camp. Other breeds of dogs owned by them are used for hunting, — for these wild wanderers sometimes devote their hours to the chase. They have two sorts, — a smooth-skinned dog, half hound half pointer, that hunts chiefly by the scent; and a greyhound, of great swift- ness, with a coat of long, silky hair, which they make use of in coursing, — hares and antelopes being their game. They have a mode of hunting — also practised by the Persians — which is peculiar. It should rather be termed hawking than hunting, as a hawk is employed for the purpose. It is a species of falcon denominated " goork," and is trained not only to dash at small game, such as partridges and bustards, but upon antelopes and even the wild ass that is found in plenty upon the plains of Turcomania. You will wonder how a bird, not larger than the common falcon, could capture such gamo as this ; but it will appear simple enough when tha method has been explained. The " goork " is trained to fly at the quadruped, and fix its claws in one particular place, — that is, upon the frontlet, just between the eyes'. When thus attached, the bird, instead of closing its wings and remaining at rest, keeps them constantly in motion, flapping them over the eyes of the quadruped. This it does, no doubt, to enable it to retain its perch ; while the unfortunate animal, thus assailed, kr.ows lot in what 236 THE TURCOMANS. direction to run, and is soon overtaken by the pai suing sportsmen, and either speared or shot with the bow and arrow. Wild boars are frequently hunted by the Turcomans and this, like everything else with these rude centaurs, :'iS performed on horseback. The bow and arrow is but a poor weapon when employed against the thick, tough hide o*f the Hyrcanian boar (for he is literally the Hyr- -»,anian boar), and of course the matchlock would be equally ineffective. How, then, does the Turcoman sportsman manage to bag this bristly game ? With all the ease in the world. It costs him only the effort of galloping his horse close up to the side of the boar after he has been brought to by the dogs, and then suddenly wheeling the steed. The latter, well trained to the task, without further prompting, goes through the rest of the performance, which consists in administering to the boar such a slap with his iron-shod heel, as to prostrate the porcine quadruped, often killing it on the instant ! Such employments and such diversions occupy only a small portion of the Turcoman's time. He follows another calling of a far less creditable character, which unfortunately he regards as the most honorable occupa- tion of his life. This is the calling of the robber. His pastoral pursuits are matters of only secondary con- sideration. He only looks to them as a means of sup- olying his daily wants, — his food and the more neces- sary portion of his clothing ; but he has other wants that may be deemed luxuries. He requires to keep up his stock of horses and camels, and wishes to increase them. He needs costly gear for his horse, and costly garments for himself, — and he is desirous of being possessed of THE TURCOMANS. 237 fine weapons, mch as spears, swords, bows, matchlocks, daggers, and pistols. His most effective weapons are the spear and sword, and these are the kinds he chiefly uses His spear consists of a steel head with four flutes, and edges very sharp, fixed upon a slender shaft of from eight to ten feet in length. In using it he couches it under the left arm, and directs it with the right hand, either straightforward, or to the right or left ; if to the right, the butt of the shaft lies across the hinder part of the saddle ; if to the left, the forepart of the spear rests on the horse's neck. The Turcomans manage their horses with the left hand, but most of these are so well broken as to obey the movement of the knee, or the impulse of the body. When cLose to their object, they frequently grasp the spear with both hands, to give greater effect to the thrust. The horse, spurred to the full speed of a charge, in this way, offers an attack no doubt very formidable iri appearance, but perhaps less really dangerous than the other, in which success de- pends so greatly on skill and address. The Turcomans are all sufficiently dexterous with the sword, which is almost universally formed in the curved Persian fashion, and very sharp ; they also wear a dagger at the waist- belt. Firearms are as yet little in use among them ; they possess a few, taken from the travellers they have plundered, and procure a few more occasionally from the Russians by the way of Bokara. Some use bows and arrows, but they are by no means so dexterous as their ancestors were in the handling of those weapons. Mounted, then, upon his matchless steed, and armed with spear and sword, the Turcoman goes forth to prac* 238 THE TURCOMANS. rise his favorite profession, — that of plunder. He doe* not go alone, nor with a small number of his comrades, either. The number depends altogether on the distance cr danger of the expedition ; and where these are con- sidered great, a troop of five hundred, or even a thou- sand, usually proceed together upon their errand. You will be inquiring to what point they direct them- selves, — east, west, north, or south ? That altogether depends upon who may be their enemies for the time, for along with their desire for booty, there is also mixed up something like a sentiment of hostility. In this respect; however, the Turcoman is a true Ishmaelite, and in lack of other victim he will not hesitate to plunder the people of a kindred race. Indeed, several of the Turcoman tribes have long been at war with one another; and their animosity is quite as deadly among themselves as when directed against strangers to their race. The butt, however, of most of the Turcoman expeditions is the northern part of Persia, — Korassan in particular. It is into this province that most of their great forays are directed, either against the peaceful citizens of the Per- sian towns and villages, or as often against the merchant caravans that are constantly passing between Teheran and the cities of the east, — Mushed, Balkh, Bokara, Herat, and Kelat. I have already stated that thes« forays are pushed far into the interior of Persia ; and the fact oi Persia permitting such a state of things to continue will perhaps surprise you ; but you would not be surprised were you better acquainted with the con- dition of that kingdom. From historic associations, you believe Persia to be a powerful nation ; and so it once \ms, both powerful and prosperous. That day is past. THE TURCOMANS. 23 & and at the present hour, this decaying monarchy is not only powerless to maintain order within its own borders, but is even threatened with annihilation from those very nomad races that have so often given laws to the great empires of Asia. Even at this moment, the more pow- erful Tartar Khans turn a longing look towards the tottering throne of Nadir Shah ; and he of Khiva has more than once made a feint at invasion. But the sub- ject is too extensive to be discussed here. It is only introduced to explain with what facility a few hundreds of Turcoman robbers can enter and harass the land. We find a parallel in many other parts of the world, — old as well as new. In the latter, the northern provinces of Mexico, and the southern countries of La Plata and Paraguay, are in just such a condition : the weak, worn- out descendants of the Spanish conquerors on one side, well representing the remnants of the race of Nadir Shah ; while, on the other, the Turcoman is type enough of the Red Indian. The comparison, however, is not just to the latter. He, at least, is possessed of courage and prowess ; while the Turcoman, notwithstanding his propensities for plunder, and the bloodthirsty ferocity of his character, is as arrant a coward as ever carried lance. Even the Persian can cope with him, when fairly matched ; and the merchant-caravans, — which are usually made up of true Turks, and other races possessing a little " pluck," are never attacked, unless when outnumbered in the ratio of three to one. For all this, the whole northern portion of the Per- sian kingdom is left to the mercy of these desert-robbers. The towns and villages have each their large fortress, into which the people retire whenever the plunderers 240 THE TURCOMANS. make their appearance, and there dwell till the lattei have ridden away, — driving off their flocks and herd* to the desert fastnesses. Even the poor farmer L obliged to build a fortress in the middle of his fields, to which he may retire upon the occasion of any sud- den alarm, and his laborers till the ground with thei 1 swords by their sides, and their matchlocks lying near ! These field fortresses of Korassan are altogether so curious, both as to construction and purpose, that we cannot pass them without a word of description. They are usually placed in some conspicuous place, at a con- venient distance from all parts of the cultivated tract They are built of mud, and raised to a height of fifteen or twenty feet, of a circular form, — bearing some re- semblance to the well-known round towers of Ireland. A small aperture is left open at the bottom, through which those seeking shelter may just squeeze their bodies, and this being barricaded inside, the defence is complete. From the top — which can be reached easily on the inside — the farmer and his laborers can use their matchlocks with effect ; but they are never called upon to do so, — as the cowardly freebooter takes good care to give the mud tower a wide birth. He has no weapons by which he might assail it ; and, moreover, he has no time for sieges : since an hour's delay might bring him into danger from the force that is fast ap- proaching. His only thought is to keep on his course, and sweep off such cattle, or make prisoners of such people as he may chance to find unwarned and un- armed. Now and then he ventures upon an attack — where there is much booty to tempt him, and but a weak force to defend it. His eiu mies, — the hafc** THE TUKCOMANS. 24J * Kuzzilbashes," as he calls the Persians, — if defeated, have no mercy to expect from him. All who resist are killed upon the spot, and often torture is the mods of their death; but if they can be made prisoners, the desert-robber prefers letting them live, as a captive is to him a more valuable consideration than the death of an enemy, ills prisoner, once secured, knows tol- erably well what is to follow. The first thing the Tur- coman does is to bind the victim's hands securely behind his back ; he then puts a long halter around his neck, attaching the other end of it to the tail of his horse, and in this fashion the homeward march commences. If the poor pedestrian does not keep pace with the horse, he knows what he may expect, — to be dragged at intervals along the ground, and perhaps torn to pieces upon the rocks. With this horrid fate before his fancy, he makes efforts almost superhuman to keep pace with the troop of his inhuman captors : though well aware that they are leading him off into a hopeless bondage. At night, his feet are also tied ; and, thrown down upon the earth, he is covered with a coarse "numud." Do not fancy that this is done to screen him from the cold : the object is very different indeed. The numud is placed over him in order that two of his captors may sleep upon its edges — one on each side of him — thus holding him down, and frustrating any chance of escape. On arriving at the robber-camp, the captive is not kept long in suspense as to his future fate. His owner — for he is now in reality a slave — wants a new ?word, or a piece of silken cloth, or a camel, or some othei article of luxury. That he can obtain either at Khiva w Bokara, in exchange for his stave ; and therefore th« 242 THE TURCOMANS. new captive — or captives, as the chance may be — is marched off to the ready market. This u no isolated nor rare incident. It is one of everyday occurrence ; And it is a noted fact, that of the three hundred thou sand people who constitute the subjects of the Khivan Khan, nearly one half are Persian slaves obtained from the robbers of Turcomania ! The political organization of the Turcomans is of the patriarchal character. From necessity they dwell in small communities that are termed " teers," the literal signification of which is "arrows," — though for what reason they are so styled does not appear. Perhaps it is on account of th* 3 rapidity of their movements : for, in hostile excursions, or moving from place to place, they proceed with a celerity that may be compared to arrows. Over each tribe or teer there is a chief, similar to the "sheik " of the Arab tribes, — and indeed, many of their customs offer a close analogy to those of the wandering Bedouins of Arabia and Egypt, and the Kabyles of Morocco and the Algerine provinces. The circumstances of life — almost alike to both — could not fail to pro- duce many striking resemblances. The Turcoman tribes, as already observed, frequently go to war with each other, but they oftener unite to rob the common enemy, — the caravan or the Persian village. In these mere plundering expeditions they go in such numbers as the case may require ; but when called forth to take side in anything like a national war, they can muster to the strength of many thousands ; and then indeed, they become terrible, — even to the most potent sovereigns of Central Asia, by whom much diplomacy I HE TUKCOMANS. 243 is -a vl ployed to enlist them on one side or the otter. It matters little to them what the cause be, — he who can promise them the largest booty in cattle or slaves is sure to have the help of their spears and swords. The Turcomans are not Pagans, — that is, they are not professedly so, — though, for all the regard which they pay to religious observances, they might as well be termed true Infidels. They profess a religion, however, and that is Mohametanisna in its worst and most bigoted form, — the " Sunnite." The Persians, as is well known, hold the milder Sheean doctrines; and as the votaries of the two, in most countries where both are practised, cordially hate each other, so it is between Turcomans and Persians. The former even scorn the Persian creed, calling its followers " Infidel " dogs, or Kuzzilbashes ; and this bigoted rancor gives them a sort of plausible excuse for the hostile attitude which they hold towards them. Ijjaking them upon the whole, the Turcomans may be looked upon as true savages, — savages dressed in folk instead of in skins. THE OTTOMACS, OR DIRT-EATERS On the banks of the Orinoco, a short distance above ihe point where that mighty river makes its second great sweep to the eastward, dwells a remarkable people, — a tribe of savages that, even among savages, are remark- able for many peculiar and singular customs. These are the Ottomacs. They have been long known, — and by the narratives of the early Spanish missionaries, rendered notorious, — on account of some curious habits ; but although the missionaries have resided among them, and endeavored to bring them within " sound of the bell," their efforts have met with a very partial and temporary success and at this present hour, the Ottomacs are as savage ir. their habits, and as singular in their customs, as the) were in the days of Columbus. The Ottomacs are neither a stunted nor yet a weak race of men. Their bodies are strong, and their arms and limbs stout and muscular ; but they are remarkably ill-featured, with an expression of countenance habitually stern and vindictive. Their costume is easily described, or rather cannot ot described a* all, since they have none. Both s^xes g« THE OTTOMACS. 245 entirely naked, — if we except a little belt of three or four inches in width, made from cotton or the bark of tree^, and called the guayuco, which they wear around the waist, — but even this is worn from no motives of modesty. What they regard in the light of a costume is a coat of paint, and about this they are as nice and particu- lar as a Parisian dandy. Talk about " blooming up " a faded belle for the ball-room, or the time spent by an exquisite in adjusting the tie of his cravat ! these are trifles when compared with the lengthy and elaborate toilette of an Ottomac lady or gentleman. The greater part of a day is often spent by them in a single dressing, with one or two helpers to assist in the operation ; and this is not a tattooing process, intended to last for a lifetime, but a costume certain to be disfigured, or entirely washed off, at the first exposure to a heavy shower of rain. Add to this, that the pigments which are used for the purpose are by no means easily ob- tained : the vegetable substances which furnish them are scarce in the Ottomac country ; and it costs one of these Indians the produce of several days of his labor to purchase sufficient paint to give his whole skin a single " coat." For this reason the Ottomac paints his body only on grand occasions, — contenting himself at ordinary times with merely staining his face and hair. When an Ottomac wishes to appear in " full dreas * he first gives himself a " priming " of red. This consists of the dye called " annotto," which is obtained from the fruit pulp of the Bixa orella?ia, and which the Indians knew how to prepare previous to their intercourse with Europeans. Over this red ground is then formed a lat 246 THE OTTOMACS, OR tice-work of lines of black, with a dot in the centre of every little square or diamond. The black dye is the "caruto," also a vegetable pigment, obtained from the Genipa Americana. If the gentleman be rich enough to possess a little "chica" which is a beautiful lake-coV ored red, — also the produce of a p*ant, — the Bignoni, chica, he will then feel all the ecstatic delight of a fash- ionable dandy who possesses a good wardrobe ; and, with half a pound of turtle-oil rubbed into his long black tresses, he will regard himself as dressed " within an inch of his life." It is not always, however, that he can afford the chica, — for it is one of the costliest materials of which a South American savage can manufacture his suit. The Ottomac takes far less trouble in the building of his house. Very often he builds none ; but when he wishes to guard his body from the rays of the sun, or the periodical rains, he constructs him a slight edifice — a mere hut — out of saplings or bamboos, with a thatch of palm-leaves. His arms consist of the universal bow and arrows, which he manages with much dexterity ; and he has also a harpoon which he employs in killing the ma- natee and the alligator. He has, besides, several other weapons, to aid him in the chase and fishing, — the latter of which forms his principal employment as well as his chief source of subsistence. The Ottomac belongs to one of those tribes of Indians termed by the Spanish missionaries Indios andantes, that is "wandering," or "vagabond Indians," who instead of remaining in fixed and permanent villages, roam about *Tom place to place, as necessity or inclination dictates DIRT-EATERS. 247 Perhaps this arises from the peculiarity of the country which they inhabit : for the Indios andantes do not live in the thick forests, but upon vast treeless savannas, which stretch along the Orinoco above its great bend In these tracts the "juvia" trees (bertholletia and lecy* (hys), which produce the delicious "Brazil-nuts" — and Other plants that supply the savage spontaneously with food, are sparsely found ; and as the savannas are an- nually inundated for several months, the Ottomac is forced, whether he will or no, to shift his quarters and try for subsistence elsewhere. When the inundations have subsided and the waters become settled enough to permit of fishing, the Ottomac " winter " is over, and he can obtain food in plenty from the alligators, the mana- tees, the turtles, the toninas or dolphins, and other large fish that frequent the great stream upon which he dwells. Of these the manatee is the most important in the eyes of the Ottomac — as it is the largest in size, and consequently furnishes him with the greatest amount of meat. This singular semi-cetaceous creature is almost too well known to require description. It is found in nearly all the large rivers of tropical America, where it feeds upon the grass and aquatic plants growing along their banks. It is known by various names, according to the place and people. The Spaniards call it vaca marina, or " sea-cow," and the Portuguese peixe hot, or " fish-ox," — both being appellations equally inappropriate, and having their origin in a slight resemblance which there exists between the animal's "countenance" and that of «tn ox. The West Indian ';ame is the one we have givei\ 248 THE OTTOMAOS, OR though the true orthography is manati, not manatee^ since the word is of Indian origin. Some writers deny this, alleging that it is a derivative from the Spanish word " mano," a hand, signifying, therefore, the fish with hands, — in allusion to the rudimentary hands which form one of its distinguishing characteristics. This is the account of the historian Oviedo, but another Spanish missionary, Father Gili, offers a more correct explana- tion of the name, — in fact, he proves, what is neither more nor less than the simple truth, that " manati " was the name given to this animal by the natives of Hayti and Cuba, — where a species is also found, — and the word has no reference whatever to the " hands " of the creature. The resemblance to the Spanish word which should signify " handed," is merely an accidental circum- stance ; and, as the acute Humboldt very justly remarks, according to the genius of the Spanish language, the word thus applied would have been written manudo, or manon, and not manati. The Indians have almost as many different names for this creature as there are rivers in which it is found ; but its appellation in the " lingo ageral " of the great Amazon valley, is "juarua." Among the Ottomacs it is called the " apoia." It may be safely affirmed that there are several species of this amphibious animal in the rivers of tropical America ; and possibly no one of them is id3ntical with that of the West Indies. All have hitherto been regarded as belonging to the same species, and described under the scientific title of 3fa- natus Americanus — a name given to the American manati, to distinguish it from the " lamantin " of Africa, and the "dugong" of me East Indian beas. But tbf DIRT-EATERS. 249 West Irrlian species appears to have certain character- istic differences, which shows that it is a separate one, or, at all events, a variety. It is of much larger size Aian tho»e of the South American rivers generally arc — though there also a large variety is found, but much rarer than those commonly captured by the fishermen. The West Indian manati has nails well developed upon the outer edge of its fins, or forearms ; while those on the oilier kinds are either not seen at all, or only in a very rudimentary state. That there are different spe- cies, may be deduced from the accounts of the natives, who employ themselves in its capture : and the obser vations of such people are usually more trustworthy than the speculations of learned anatomists. The Am- azon fishermen all agree in the belief that there are three kinds of manati in the Amazon and its numerous tributaries, that not only differ greatly in size — from seven to twenty feet long — and in weight, from four hundred to two thousand pounds, — but also in the color of their skin, and the shape of their tails and fins. The species found in the Orinoco, and called " apoia " by the Ottomacs, is usually about twelve feet in length, and weighs from five hundred to eight hundred pounds ; but now and then a much larger individual is captured, per- haps owing to greater age, or other accidental circum- stance. Humboldt heard of one that weighed eight thousand pounds ; and the French naturalist D'Orbigny speaks of one killed in the Bolivian waters of the Amazon that was twenty feet in length. This size is often attained by the Manatus Americanus of Cuba and Hayti. The manati is shaped sor\ewhat like a huge seal, and 250 THE OTTOMACS, OR has certain resemblances to a fish. Its body Li of ai oval oblong, with a large, flat, rounded tail, set horizon- tally, and which serves as a rudder to direct its course in the water. Just behind its shoulders appear, instead of fins, a pair of flippers, which have a certain resun- blance to hands set on to the body without arms. Of these it avails itself, when creeping out against the bank and the female also uses them in carrying her young. The mammas (for it must be remembered that this crea- ture is a mammiferous animal) are placed just below and behind the flippers. The muzzle is blunt, with thick lips, — the upper projecting several inches beyond the lower, and covered with a delicate epidermis : showing evidently that it avails itself of this prominence — which possesses a keen sense of touch — just as the elephant of his proboscis. The lips are covered with bristles, or beard, which impart a kind of human-like expression to the animal's countenance, — a circumstance more ob- servable in the " dugongs " of the Oriental waters. " Woman-fish," too, these have been called, and no doubt such creatures, along with the seals and walruses, have given rise to many a story of sirens and mermaids. The " cow-face," however, from which the manati obtains its Spanish and Portuguese epithets, is the most charac* teristic ; and in its food we find a still greater analogy to the bovine quadruped with which it is brought ib comparison. Beyond this the resemblance ceases. The body is that of a seal ; but instead of being covered with hair, as the cetaceous animal, the manati has a smooth skin that resembles india-rubber more than any thing else. A few short hairs are set here and there, but they are scarce observable. The color of the manati i« DniT-EATERS. 251 iJhat of lead, with a few mottlings of a pinkish- white hu€ upon the belly ; but in this respect there is no uniform- ity. Some are seen with the whole under- parts of a uniform cream-color. The lungs of this animal present a peculiarity worth? of being noted. They are very voluminous, — being sometimes three feet in length, and of such a porous and elastic nature as to be capab ] e of immense extension. When blown out, they present 4ie appearance of great swimming bladders ; and it is by means of this capacity for containing air that the manati is enabled to remain so long under water, — though, like the true cetacece, it requires to come at intervals to the surface to obtain breath. The flesh of the manati is eaten by all the tribes of Indians who can procure it, — though by some it is more highly esteemed than by others. It was once much relished in the colonial settlements of Guiana and the West Indies, and formed a considerable article of commerce ; but in these quarters manatis have grown scarce, — from the incessant persecution of the fisher- men. The flesh has been deemed unwl olesome by some, and apt to produce fevers ; but this is not the general opinion. It has a greater resemblance to pork than beef, — though it be the flesh of a cow, - — and is very savory when fresh, though neither is it bad eating when salted or dried in the sun. In this way it will keep for several months ; and it has always been a stock article with the monks of the South American missions, — who, in spite of its mammiferous character, find it convenient, during the days of Lent, to regard it as a JUh f The skin of the manati is of exceeding thick' 252 THE OTTOMACS, OR n*ss ; — on the back an inch and a half at least, thougr it becomes thinner as it approaches the under-parts of the body. It is cut into blips which serve various cur- poses, as for shields, cordage, and whips. " These whipg of manati leather," says Humboldt, " are a cruel instru- ment of punishment for the unhappy slaves, and even for the Indians of the missions, though, according to the laws, the latter ought to be treated as freemen." Another valuable commodity obtained from this animal is oil, known in the missions as manati-butter (manteca de manati). This is produced by the layer of pure fat, of an inch and a half in thickness, which, lying imme- diately under the skin, envelops the whole body of the animal. The oil is used for lamps in the mission churches ; but among the Indians themselves it is also employed in the cuisine, — as it has not that fetid smell peculiar to the oil of whales and salt-water cetaceae. The food of the manati is grass exclusively, which it finds on the banks of the lakes and rivers it frequents. Of this it will eat an enormous quantity ; and its usuai time of browsing is at night, — though this habit may have arisen from its observance of the fact, that night is the safest time to approach the shore. In those places, where is has been left undisturbed, it may be often seen browsing by day. I have been thus particular in my account of this animal, because it is more nearly connected with the history of Ottomac habits than perhaps that of any other tribe of South American Indians, — the Guamos alone excepted, who may themselves be regarded as merely a branch of the Ottomac family. Though, as already remarked, all the tribes who dwell upon manati DIRT- EATERS. 253 rivers pursue this creature and feed upon its fleiih, yet in no other part of South America is this species of fishery so extensively or so dexterously carried on as among the Ottoinacs and Guamos, — the reason being, that, amidst the great grassy savannas which charac- terize the Ottomac country, there are numerous stream3 and lagoons that are the favorite haunts of this her- bivorous animal. In one river in particular, so great a number are found that it has been distinguished by the appellation of the Rio de Manatis (river of manatis). The manati, when undisturbed, is gregarious in its habits, going in troops (or " herds," if we preserve the analogy) of greater or less numbers, and keeping the young " calves " in the centre, which the mothers guard with the tenderest affection. So attached are the parents to their young, that if the calf be taken, the mother can be easily approached ; and the devotion is reciprocated on the filial side ; since in cases where the mother has been captured and dragged ashore, the young one ha*, often been known to follow the lifeless body up to the very bank ! As the manati plays such an important part in the domestic economy of the Ottomacs, of course the cap- turing of this animal is carried on upon the grandest scale among these people, and, like the " harvest of turtle-eggs," hereafter to be described, the manati fishery has its particular season. Some writers have errone- ously stated this season as being the period of iniinda* tion, and when the water is at its maximum height. This is quite contrary to the truth ; since that period, both on the Amazon and Orinoco rivers, is just the time wheD all kinds of fishing is difficult and precarious 254 THE OTTOMACS, OR Then is the true winter, — the "blue months '' of *ue South American river Indians; and it is then, as \*iU presently be seen, that the Ottomac comes nearest tho point of starvation, — which he approaches every year of his life. There are manati and other kinds of fish taken at all times of the year ; but the true season of the manati- fishing is when the waters of the great flood have con- siderably subsided, and are still continuing to diminish rapidly. When the inundation is at its height, the manati passes out of the channel current of the great river, and in search of grass it finds its way into the lakes and surrounding marshes, remaining there to browse along their banks. When the flood is rapidly passing away from it, it begins to find itself a " little out of its ele- ment," and just then is the time when it is most easily captured. Sometimes the Indians assemble in a body with their canoes, forming a large fleet ; and. proceeding to the best haunts of the " cow-fish," cany on the fishery in a wholesale manner. The monks of the missions also head the tame tribes on these expeditions, — as they do when collecting the eggs of the turtle, — and a regular systematic course is carried on under the eye of dis- cipline and authority. A camp is formed at some con- rsnient place on the shore. Scaffolds are erected for sun-drying the flesh and skins ; and vessels and other utensils brought upon the ground to render the fat into oil. The manatis that have been captured are all brought in the canoes to this central point, and delivered up to be "jlmsed" cured, and cooked. There is the usuaJ assemblage of small traders from Angostura and other DIRT-EATERS. 255 porta on the lower Orinoco, who come to barter their Indian trinkets for the manteca de manati in the same manner &a it will presently be seen they trade for the manteca de tortngas. I need not add that this is a sea- eon of joy and festivity, like the wine-gatherings and harvest-homes of the European peasantry. The mode of capturing the manati is very similar to that employed by the Esquimaux in taking the seal, and which has been elsewhere described. There is not much danger in the fishery, for no creature could be moi^ harmless and inoffensive than this. It makes not the slightest attempt either at defence or retaliation, — though the accident sometimes occurs of a canoe be- ing swamped or drawn under water, — but this is noth- ing to the Ottomac Indian, who is almost as amphibious as the manati itself. At the proper hour the fisherman starts off in search of the manati. His fishing-boat is a canoe hollowed from a single trunk, of that kind usually styled a " dug- out." On perceiving the cow-fish resting upon the sur- face of the water, the Ottomac paddles towards it, ob- serving the greatest caution ; for although the organs of sight and hearing in this animal are, externally, but very little developed, it both hears and sees well; and the slightest suspicious noise would be a signal for it to dive under, and of course escape. When near enough to insure a good aim, the Ottomac hurls his harpoon into the animal's body ; which, after piercing the thick hide, sticks fast. To this harpoon a cord is attached, with a float, and the float remaining above water indicates the direction in which the wounded animal now endeavors to get off When it is tired of 256 THE OTTOMACS, OR struggling, thi> Indian regains the cord ; and taking it in, hand over hand, draws up his canoe to the side of the fish. If it be still too lively, he repeatedly strikes it with a spear ; but he does not aim to kill it outright until he has got it " aboard." Once there, he ends the creature's existence by driving a wooden plug into itt nostrils, which in a moment deprives it of life. The Ottomac now prepares himself to transport the carcass to his home ; or, if fishing in company, to the common rendezvous. Perhaps he has some distance to take it, and against a current ; and he finds it inconven- ient to tow such a heavy and cumbrous article. To remedy this inconvenience, he adopts the expedient already mentioned, of placing the carcass in his canoe. But how does he get it there ? How can a single Indian of ordinary strength raise a weight of a thousand pounds out of the water, and lift it over the gunwale of his un- steady craft ? It is in this that he exhibits great cun- ning and address : for instead of raising the carcass above the canoe, he sinks the canoe below the carcass, by first filling the vessel nearly full of water ; and then, after he has got his freight aboard, he bales out ' the water with his gourd-shell. He at length succeeds in adjusting his load, and then paddles homeward with his prize. On arriving at his village, — if it be to the village he takes it, — he is assisted in transporting the load by others of his tribe ; but he does not carry it to his own house, — for the Ottomacs are true socialists, and the produce both of the chase and the fishery is the common property of all. The chief of the village, seated in front of his hut, receives all that is brought home, and (lis* DIRT-EATERS. 257 tributes it out to the various heads of families, — giving to each in proportion to the number of mouths that are to be fed. The manati is flayed, — its thick hide, as already ob- served, serving for many useful purposes ; the strata of fat, or " blubber," which lies beneath is removed, to be converted into oil ; and finally, the flesh, which is es- teemed equal to pork, both in delicacy and flavor, is cut into thin slices, either to be broiled and eaten at the time, or to be preserved for a future occasion, not by salt, of which the Ottomac is entirely ignorant, but by drying in the sun and smoking over a slow fire. Fish and the flesh of the alligator are similarly " cured ; " and when the process is carefully done, both will keep for months. The alligator is captured in various ways : sometimes by a baited hook with a strong cord attached, — some- times he is killed by a stab of the harpoon-spear, and not unfrequently is he taken by a noose slipped over his paw, the Ottomac diving fearlessly under him and adjust- ing the snare. Some of the Indian tribes will not eat the musky flesh of the alligator ; but the Ottomacs are not thus particu- lar. Indeed, these people refuse scarce any article of food, however nasty or disagreeable ; and it is a saying among their neighbors — the Indians of other tribes — that " nothing is too loathsome for the stomach cf an Ottomac." Perhaps the saying will be considered as perfectly true when we come to describe a species of food which these people eat, and which, for a long time, has ren- dered them famous — or rather infamous — under the appellation of " dirt eaters." Of them it may literally 258 THE OTTOMACS, Oil be said that they a eat dirt," for such, in reality, is on* of their customs. This singular practice is chiefly reported to during those months in the year when the rivers swell to their greatest height, and continue full. At this time all fishing cea>es, and the Ottomac finds it difficult to obtain a sufficiency of food. To make up for the deficiency, he fills his stomach with a kind of unctuous clay, which he has already stored up for the emergency, and of which he eats: about a pound per diem ! It does not consti- tute his sole diet, but often for several days together it is the only food which passes his lips ! There is nothing nourishing in it, — that has been proved by analysis. It merely Jills the belly, — producing a satiety, or, at least, giving some sort of relief from the pangs of hunger. Nor has it been observed that the Ottomac grows thin or unhealthy on this unnatural viand : on the contrary, he is one of the most robust and healthy of American Indians. The earth which the Ottomac eats goes by the name of poya. He does not eat clay of every kind : only a peculiar sort which he finds upon the banks of streams. It is soft and smooth to the touch, and unctuous, like putty. In its natural state it is of a yellowish-gray color ; but, when hardened before the fire, it assumes a tinge of red, owing to the oxide of iron which is in it. It was for a long time believed that the Ottomac mixed this clay with cassava and turtle-oil, or some other sort of nutritive substance. Even Father Gumilla — who was credulous enough to believe almost any- thing — could not " swalltfw " the story of the clay in its natural state, bi t believed that it was prepared witl DIKT-EATERS. 259 vonie comoiiiation of farinha or fat. This, however, is Dot the case. It is a pure earth, containing (according to the analysis of Vauquelin) silex and alumina, with three or four per cent of lime ! This clay the Ottomac stores up, forming it into balls of several inches in diameter ; which, being slightly har- dened before the fire, he builds into little pyramids, just as cannon-balls are piled in an arsenal or fortress. When the Ottomac wishes to eat of the poya, he softens one of the balls by wetting it ; and then, scraping off as much as he may require for his meal, returns the poya to its place on the pyramid. The dirt-eating does not entirely end with the falling of the waters. The practice has begot a craving for it ; and the Ottomac is not contented without a little poya, even when more nutritious food may be obtained in abundance. This habit of eating earth is not exclusively Ottomac. Other kindred tribes indulge in it, though not to so great an extent ; and we find the same unnatural practice among the savages of New Caledonia and the Indian archipelago. It is also common on the west coast of Africa. Humboldt believed it to be exclusively a tropical habit. In this the great philosopher was in error, since it is known to be practised by some tribes of northern Indians on the frigid banks of the Mackenzie lliver. When the floods subside, as already stated, the Otto- mac lives better. Then he can obtain both fish and turtles in abundance. The former he captures, both with hooks and nets, or shoots with his arrows when they rise near the surface. Tl e turtles of the Ottomac rivers are of fwo kinds 260 THE OTTOMACS, OR the arau and terecay. The former is the one most sought after, as being by far the largest. It is nearly a yard across the back, and weighs from fifty to a hun dred pounds. It is a shy creature, and would be difficult to capture, were it not for a habit it has of raising its head above the surface of the water, and thus exposing the soft part of its throat to the Indian's arrow. Even then an arrow might fail to kill it ; but the Ottomac takos care to have the point well coated with curare poison, which in a few seconds does its work, and secures the death of the victim. The terecay is taken in a different and still more ingenious manner. This species, floating along the sur- face, or even when lying still, presents no mark at which a shaft can be aimed with the slightest chance of success. The sharpest arrow would glance off its flat shelly back as from a surface of steel. In order, therefore, to reach the vitals of his victim, the Indian adopts an expedient, in which he exhibits a dexterity and skill that are truly remarkable. He aims his shaft, not at the turtle, but up into the air, describing by its course a parabolic curve, and so calculating its velocity and direction that it will drop perpendicularly, point foremost, upon the back of the unsuspecting swimmer, and pierce through the shell right into the vital veins of its body ! It is rare that an Indian will fail in hitting such a mark; and, both on the Orinoco and Amazon, thousands of turtles are obtained in this manner. The great season of Ottomac festivity and rejoicing, however, is that of the cosecha de tortugas, or " turtle- crop." As has been already observed, in relation to the DIRT-EATERS 261 manati fishery, it is to him what the harvest-homt is t.c the nations of northern Europe, or the wine-gathering to those of the south ; for this is more truly the character of the cosecha. It is then that he is enabled, not only to procure a supply of turtle-oil with which to lubricate his hair and skin, but he obtains enough of this delicious grease wherewith to fry his dried slices of manati, and a surplus for sale to the turtle-traders from the Lower Orinoco. In this petty commerce no coin is required ; harpoon-spears, and arrow-heads of iron, rude knives, and hachets ; but, above all, a few cakes of aunotto chica, and cariUo, are bartered in exchange for the turtle-oil. The thick hide of the manati, — for making slave-whips, — the spotted skin of the jaguar, and some other pelts which the chase produces, are also items of his export trade. The pigments above mentioned have already been procured by the trader, as the export articles of com- merce of some other tribe. The turtle-oil is the product of the eggs of the larger species, — the arau, — known simply by the name tor- tuga, or turtle. The eggs of the terecay would serve] equally as well ; but, from a .inference in the habit of this animal, its eggs cannot be obtained in sufficient quantity for oil-making. There is no such thing as a grand " cosecha," or crop of them — for the creature is not gregarious, like its congener, but each female makes her nest apart from the others, in some solitary pla"e, and there brings forth her young brood. Not but that the nests of the terecay are also found and despoiled of their eggs, — but this only occurs at intervals; and as the contents of a single nest would not be sufficient foj 262 THE OTTOMACS, OK a "churning" no "butter" can be made of them. They are, therefore, gathered to be used only as eggs, and not as butter. The arau, on the other hand, although not gregarious under ordinary circumstance!, becomes pre-eminently so during the " laying season." Then all the turtles in the Orinoco and its tributaries collect into three or four vast gangs — numbering in all over a million of individuals — and proceed to certain points of rendezvous which they have been in the habit of visiting from time im- memorial. These common breeding-places are situated between the cataracts of the river and the great bend, where it meets the Apure ; and are simply broad beach- es of sand, rising with a gentle slope from the edge of the water, and extending for miles along the bank. There are some small rookeries on tributary streams, but the three most noted are upon the shores of the main river, between the points already indicated. That frequented by the Ottomacs is upon an island, at the mouth of the Uruana River, upon which these people principally dwell. The laying season of the arau turtle varies in the different rivers of tropical America, — occurring in the Amazon and its tributaries at a different period from that of the Orinoco. It is regulated by the rise, or rather the fall of the inundations ; and takes place when the waters, at their lowest stage, have laid bare the low sand-banks upon the shores. This occurs (in the Orinoco) in March, and early in this month the great assemblages are complete. For weeks before, the turtles are seen, in all parts of the river near the intended breeding-places, swimming about ou the siir* DIRT-EATERS. 263 face 4 or basking along the banks. As the sun grows strouger, the desire of depositing their eggs increases, — as though the heat had something to do with the ; * fecundation. For some time before the final action, the creatures may be seen ranged in a long line in front of /he breeding-place, with their heads and necks held high above the water; as if contemplating their intended nursery, and calculating the dangers to which they may be exposed. It is not without reason that they may dwell upon these. Along the beach stalks the lordly jaguar, waiting to make a meal of the first that may set his foot on terra firma, or to fill his stomach with the delicious "new-laid" eggs. The ugly alligator, too, is equally friand of a gigantic omelette ; a*nd not less so the "garzas" (white cranes), and the "zamuros" (black vultures), who hover in hundreds in the air. Here and there, too, may be observed an Indian sentinel, keeping as much as possible out of sight of the turtles themselves, but endeavoring to drive off all other ene- mies whose presence may give them fear. Should a canoe or boat appear upon the river, it is warned by these sentinels to keep well off from the phalanx of the turtles, — lest these should be disturbed or alarmed, — for the Indian well knows that if anything should occur to produce a panic among the araus, his cosecha would be very much shortened thereby. When at length the turtles have had sun enough to warm them to the work, they crawl out upon the dry Band-beach, and the laying commences. It is at night that the operation is carried on : for then their numer- ous enemies — especially the vultures — are less active, Each turtle scoops out a hole, of nearly a yard in diauw* 264 THE 01T0MACS, OR ter and depth ; and having therein deposited from fiftj to one hundred eg%*, it covers them up with the sand smoothing the surface, and treading it firmly down Sometimes the individuals are so crowded as to lay in one another's nests, breaking many of the eggs, and causing an inextricable confusion ; while the creaking noise of their shells rubbing against each other may be heard afar off, like the rushing of a cataract. Sometimes a number that have arrived late, or have been slow at their work, continue engaged in it till after daybreak; and even after the Indians have come upon the ground — whose presence they no longer regard. Impelled by the instinct of philoprogenitiveness, these "mad turtles," as the Indians call them, appear utterly regardless of danger, and make no effort to escape from it ; but are turned over on their backs, or killed upon the spot without difficulty. The beach being now deserted by the turtles, the egg- gatherers proceed to their work. As there are usually several tribes, who claim a share in the cosecha, the ground is measured out, and partitioned among them. The regularity with which the nests are placed, and the number of eggs in each being pretty nearly the same, an average estimate of the quantity under a given surface is easily made. By means of a pointed stick thrust into the sand, the outline of the deposit is ascertained — usually running along the beach in a strip cf about thirty yards in breadth. When the allotments are determined, the work of oil-making begins, — each tribe working by itself, and upon the social system. The covering of sand is re- nwred, and the eggs placed in baskets, which are then DIRT- EATERS. 265 emptied into large wooden troughs, as a common re- ceptacle. The canoes, drawn up 6n the sand, are fre- quently made to do duty as troughs. When a sufficient number of eggs have been thrown in, they are broken and pounded together, and whipped about, as if intended for a gigantic omelette. Water is added ; and then the mixture is put into large caldrons, and boiled until the oil comes to the top ; after which it is carefully skimmed off and poured into earthern jars (" botigas,") provided by the traders. It takes about two weeks to complete the operations, during which time many curious scenes occur. The sand swarms with young turtles about as big as a clol lar, which have been prematurely hatched ; and have contrived to crawl out of the shell. These are chased in all directions, and captured by the little naked Otto- macs, who devour them " body, bones, and all," with as much gusto as if they were gooseberries. The cranes and vultures, and young alligators too, take a part in this by-play — for the offspring of the poor arau has no end of enemies. When the oil is all boiled and bottled, the trader dis- plays his tempting wares, and makes the best market he can ; and the savage returns to his palm-hut village, — taking with him the articles of exchange and a few baskets of eggs, which he has reserved for his own eat- ing ; and so ends the cosecha de tortugas. It is in this season that the Ottomac indulges most in good living, and eats the smallest quantity of dirt. The waters afford him abundance of fish and turtle-flesh, beef from the sea-cow, and steaks from the tail of the alligator. He has his turtle and manati butter, in whicJf 265 THE OTTOMACS, OR to fry all these daities, and also to lubricate his hair and skin. He can dress, too, " within an inch of his life," having obtained for his oil a fresh supply of the precious pig- ments. He indulges, moreovei, in tits of intoxication, caused by a beverage made from maize or manioc root ; but oftener produced by a species of snuff which he in- hales into his nostrils. This is the niopo, manufactured from the leaves of a mimosa, and mixed with a kind of lime, which last is obtained by burning a shell of the genus helix, that is found in the waters of the Orinoco. The effect of the niopo resembles that produced by chew- ing betel, tobacco, opium, or the narcotic coca of Peru. When freely taken, a species of intoxication or rather mania is produced ; but this snuff and its effects are more minutely described elsewhere. It is here intro- duced because, in the case of the Ottomac, the drug often produces most baneful consequences. During the con- tinuance of his intoxication the Ottomac is quarrelsome and disorderly. He picks a hole in the coat of his neighbor ; but if there chance to be any " old sore h between him and a rival, the vindictive feeling is sure to exhibit itself on these occasions ; and not unfrequently ends in an encounter, causing the death of one or both of the combatants. These duels are not fought either with swords or pistols, knives, clubs, nor any similar weapons. The destruction of the victim is brought about in a very different manner ; and is the result of a very slight scratch which he has received during the fight from the nail of his antagonist. That a wound of so trifling a nature should prove mortal would be something very mysterious, did we not know that the nail whic^ DIRT-EATERS. 207 inflicted tlat scratch has been already enfiltrated with curare, — one of the deadliest of vegetable poisons, which the Ottomac understands how to prepare in its most potent and virulent form. Should it ever be your unfortunate fate therefore, tc get into a u scrimmage " with an Ottomac Indian, you must remember to keep clear of his " cl&ws " 1 THF COMANCHES, OR PMIRIE INDIANS Young reader, I need scarce tell you that the no blest of animals — the horse — is not indigenous t( America. You already know that when Columbus dis- covered the New World, no animal of the horse kind was found there ; and yet the geologist has proved in- contestably that at one time horses existed in the New World, — at a period too, geologically speaking, not very remote. The fossilized bones examined by one of the most accomplished of modern travellers — Dr. Darwin — establish this truth beyond a doubt. The horse that at present inhabits America, though not indigenous, has proved a flourishing exotic. Not only in a domestic state has he increased in numbers, out he has in many places escaped from the control of man, and' now runs wild upon the great plains both of North and South America. Although you may find in America almost every "treed" of horses known in Eu- rope., yet the great majority belong to two very distinct kinds. The iirst of these is the large English horse, in his different varieties, imported by the Anglo-Americans, and existing almost exclusively in the woodland territory of the United States. The second kind is the AndaJ* THE COMANCHES. 269 sian-Arab, — the horse of the Spanish conquerors, — a much smaller breed than the English-Arabian, but quite equal to him in mettle and beauty of form. It is the Andalusian horse that is found throughout all Spanish America, — it is he that has multiplied to such a won- derful extent, — it is he that has " run wild." That the horse in his normal state is a dweller upon open plains, is proved by his habits in America, — for in no part where the forest predominates is he found wild, — only upon the prairies of the north, and the llanos and pampas of the south, where a timbered tract forms the exception. He must have found these great steppes congenial to his natural disposition, — since, only a very short time after the arrival of the Spaniards in the New World, we find the horse a runaway from civilization, — not only existing in a wild state upon the prairies, but in posses- sion of many of the Indian tribes. It would be an interesting inquiry to trace the change of habits which the possession of the horse must have occasioned among these Arabs of the Western world. However hostile they may have been to his European rider, they must have welcomed the horse as a friend. No doubt they admired the bold, free spirit of the noble animal so analogous to their own nature. He and they soon became inseparable companions ; and have contin- ued so from that time to the present hour. Certain it is that the prairie, or "horse-Indians" of the present day, are in many respects essentially different from the staid and stoical sons of the forest so often depicted in romances ; and almost equally certain is it, that the pos- session of the horse has contributed much to bring about 270 THE COMAXCHES. OR th : s dissmiflarity. I: could no rwise. With the horse ne - were introduced, — n and , — new modes of thou* only the .: war its ne, — tc be played in an entirely different maimer. We shall no: k to inquire what these Indiana were when afoot It is our purpose only to describe what they are now that they are on horseback. lit- ay, may we say : :\ unless at this we may safely take it for granted they are upon the backs of their horses. — young and old of them, rich and poor. — for there is none of them so poor as not to be the master of a tt mustang * si In " Prairie-land " every tribe of Indians is in pos- session of the horse. On the north the frees, Ci and Blackfeet, the Si and Arapah on the plains of the Platte, the Kansas, and Osage, we find the Paw — all he Indians, West of the the Apa- che is mounted : so likewise the Utah, the Navajo, and the £ — the latter rather sparingly. - degree, possess this valuable animal ; but the true type of the u hor-e- Endian " is to be found in the Comanche, the lord of wide domain that extends from the Arkansas to the Bio Grand-. He it is who able to the frontier 3 of Texas, and equally harasses the Spanish settlements of Nc he it is who ear- linost into the heart i \ — even to the gates of the populous Dura: . Regarding the Comanche, thai pe ol th€ PRAIEIE INDIANS. 271 hor ; c-Tndians, we shall speak more particularly of him Allowing for some slight difference in the character of his climate and country, his habits and customs will be found not very dissimilar to those of the other tribes who make the prairie their home. To say that the Comanche is the finest horseman in the world would be to state what is not the fact. He is not more excellent in this accomplishment than his neighbor and bitter foeinan, the Pawnee, — no better than the " vaquero " of California, the " ranchero " of Mexico, the " llanero " of Venezuela, the " gaucho " of Buenos Ayres, and the horse-Indians of the u Gran Chaco " of Paraguay, of the Pampas, and Patagonia. He is equal, however, to any of these, and that is say- ing enough, — in a word, that he takes rank among the finest horsemen in the world. The Comanche is on horseback almost from the hour of infancy, — transferred, as it were, from his mother's arms to the withers of a mustang. "When able to walk, he is scarce allowed to practise this natural mode of progression, but performs all his movements on the back of a horse. A Comanche would no more think of making a journey afoot — even if it were only to the distance of a few hundred yards — than he would of crawling upon his hands and knees. The horse, ready saddled and bridled, stands ever near, — it differs little whether there is either saddle or bridle, — and flinging himself on the animal's back, or his neck, or his croup, or hanging suspended along his side, the Indian guides him to the destined spot, usually at a rapid gallop. It is of no consequence to the rider how fast the horse may be going : it will not hinder him from mounting of 272 THE COMANCHES, OR dismounting at will. At any time, by clutching tb* mane, he can spring upon the horse's shoulders, — just as may be often seen in the arena of the circus. The horse-Indian is a true type of the nomadic races, — a dweller in tents, which his four-footed associate enables him to transport from place to place with the utmost facility. Some of the tribes, however, and even some of the Comanches, have fixed residences, or " vil- lages," where at a certain season of the year they — 01 rather their women — cultivate the maize, the pumpkin, the melon, the calabash, and a few other species of plants, — all being vegetable products indigenous to their country. No doubt, before the arrival of Europeans, this cultivation was carried on more extensively than at present ; but the possession of the horse has enabled the prairie tribes to dispense with a calling which they cordially contemn : the calling of the husbandman. These misguided savages, one and all, regard agri- cultural pursuits as unworthy of men ; and wherever necessity compels them to practise them, the work falls to the lot of the women and slaves, — for be it known that the Comanche is a slave-owner ; and holds in bond- age not only Indians of other tribes, but also a large number of mestizoes and whites of the Spanish race, captured during many a sanguinary raid into the settle- ments of Mexico ! It would be easy to show that it is this false pride of being hunters and warriors, with lis associated aversion for an agricultural life, that has thinned the numbers of the Indian race — far more than any persecution they have endured at the hands of the white man. This it is that starves them, that makes unendurable neighbors of them, and has rendered PRAIRIE INDIANS. 27J ft necessary in acme instances to " civilize them off th« face ;>f the earth." But they are not yet all civilized from off the face of the earth ; nor is it their destiny to disappear so readi- ly as short-seeing prophets have declared. Their idle habits and internecine wars nave done much to thin their numbers, — far more than the white man's hos- tility, — but wherever the white man has stepped in and put a stop to their tribal contentions, — wherever lie has succeeded in conquering their aversion to indus- trial pursuits, — the Indian is found not only to hold his ground, but to increase rapidly in numbers. This is the case with many tribes, — Creeks, Choctaws, and Chero- kees, — so that I can promise you, young reader, that by the time you get to be an old man, there will be as many Indians in the world as upon that day when Columbus first set his foot upon " Cat " Island. You will be inquiring how the horse could render the prairie Indian more independent of agriculture ? The answer is simple. With this valuable auxiliary a new mode of subsistence was placed within his reach. An article of food, which he had hitherto been able to ob- tain only in a limited quantity, was now procurable in abundance, — the flesh of the buffalo. The prairies of North America have their own pecu- liarities. They are not stocked with large droves of ruminant animals, as the plains of Southern Africa, — where the simplest savage may easily obtain a dinner of flesh-meat. A few species of deer, thinly distributed, — all swift, shy animals, — the prong-horn antelope, still swifter and shier, — and the " big-horn," shiest of all, — were tts only ruminants of Prairie-land, with the ex* 274 THE COMANCHES, OR eeption of the groit bison, or buffalo, as lie is generally called. But even this last was not so easily captured in those days. The bison, though not a swift runner, is yet more than a match for the biped man ; and though the Indian might steal upon the great drove, and succeed in bringing down a few with Hs arrows, it was not always a sure game. Moreover, afoot, the hunter could not follow the buffalo in its grand migrations, — often ex- tending for hundreds of miles across plains, rivers, and ravines. Once mounted, the circumstances became changed. The Indian hunter could not only overtake the buffalo, but ride round him at will, and pursue him, 'f need be, to the most distant ^arts of Prairie-land. The result, therefore, of the introduction of the horse was a plentiful supply of buffalo-meat, or, when that failed, the flesh of the horse himself, — upon which two articles of diet the prairie Indian has almost exclusive- ly subsisted ever since. The Comanche has several modes of hunting the buffalo. If alone, and he wishes to make a grand coup, he will leave his horse at a distance, — the animal being trained to remain where his master has left him. The hunter then approaches the herd with great caution, keeping to leeward, — lest he might be " winded " by the old sentinel bulls who keep watch. Should there be no cover to shelter the approach of the hunter, the result would be that the bulls would discover him ; and, giving out their bellow of alarm, cause the others to scamper off. To guard against this, the Indian has already pre- pared himself by adopting a ruse, — which consists in disguising himself in the skin of a buffalo, horns and JKAIKIE INDIANS. 275 fill complete, and approaching the herd, us if he were some stray individual that had been left behind, and was just on the way to join its fellows. Even the motions of the buffalo, when browsing, are closely imitated by the red hunter ; and, unless the wind be in favor of his being scented by the bulls, this device will insure the success of a shot. Sometimes the skin of the large wlutish-gray wolf is used in this masquerade with equal success. This may appear singular, since the animal itself is one of the deadliest enemies of the buffalo : a large pack of them hanging on the skirts of every herd, and patiently waiting for an opportunity to attack it, But as this attack is only directed against the younger calves, — or some disabled or decrepit individual who may lag behind, — the strong and healthy ones have no. fear of the wolves, and permit them to squat upon the prairie within a few feet of where they are browsing ! Indeed, they could not hinder them, even if they wished : as the long-legged wolf in a few springs can easily get out of the way of the more clumsy ruminant ; and, therefore, does not dread the lowering frontlet of the most shaggy and ill-tempered bull in the herd. Of course the hunter, in the guise of a wolf, obtains the like privilege of close quarters ; and, when he has ar- rived at the proper distance for his purpose, he prepares himself for the work of destruction. The bow is the weapon he uses, — though the rifle is now a common weapon in the hands of many of the horse-Indians. But the bow is preferred for the species of " still hunt- ing " here described. The first crack of a rifle would scatter the gang, leaving the hunter perhaps only an empty gun for his pains; wr ile an arrow at such close 276 THE COMANCHES, OK quarters is equally as deadly in its effect ; and, being « silent weapon, no alarm is given to any of the buffaloes except that one which has felt the deadly shaft passing through its vitals. Often the animal thus shot — even when the wound is a mortal one — does not immediately fall ; but sinks gradually to the earth, as if lying down for a rest. Sometimes it gets only to its knees, and dies in this attitude ; at other times it remains a long while upon its legs, spreading its feet widely apart, as if to prop itself up, and then rocking from side to side like a ship in a ground-swell, till at last, weakened by loss of blood, it yields its body to the earth. Sometimes the struggles of a w^ounded individual cause the herd to " stampede/' and then the hunter has to content himself with what he may already have shot ; but not unfrequently the un- suspicious gang keeps the ground till the Indian has emptied his quiver. Nay, longer than that : for it often occurs that the disguised buffalo or wolf (as the case may be) approaches the bodies of those that have fallen, recovers some of his arrows, and uses them a second time with like deadly effect ! For this purpose it is his practice, if the aim and distance favor him, to send his shaft clear through the body of the bison, in order that the barb may not hinder it from being extracted on the other side ! This feat is by no means of uncommon occurrence among the buffalo-hunters of the prairies. Of course, a grand wholesale slaughter of the kind just described is not an every-day matter ; and can only be accomplished when the buffaloes are in a state of comparative rest, or browsing slowly More generally they detect the dangerous counterfeit in time to shv* PRAIRIE INDIANS. 277 theh skins ; or else keep moving too rapidly for the hunter to follow them on foot. His only resource, then, is to ride rapidly up on horseback, fire his arrows with- out dismounting, or strike the victim with his long lance while galloping side by side with it. If in this way he san obtain two or three fat cows, before his horse be« comes blown, or the herd scatters beyond his reach, he considers that he has had good success. But in this kind of chase the hunter is rarely alone : the whole tribe takes part in it ; and, mounted on their well-trained mustangs, often pursue the buffalo gangs for an hour or more, before the latter can get off and hide themselves in the distance, or behind the swells of the prairie. The clouds of dust raised in a melee of this kind often afford the buffalo a chance of escaping, — especially when they are running with the wind. A "buffalo surround" is effected by a large party of hunters riding to a great distance ; deploying them- selves into a circle around the herd ; and then galloping inward with loud yells. The buffaloes, thus attacked on all sides, become frightened and confused, and are easily driven into a close-packed mass, around the edges of which the mounted hunters wheel and deliver their arrows, or strike those that try to escape, with their long spears. Sometimes the infuriated bulls rush upon the horses, and gore them to death ; and the hunters, thus dismounted, often run a narrow risk of meeting with the same fate, — more than a risk, for not unfrequently they are killed outright. Often are they obliged to leap up on the croup of a companion's horse, to get out of the way of danger; and many instances are recorded where a horseman, by the stumbling of his horse, hai 278 THE COMANCHES, OR been pitched right into the thick of the herd, and has made his escape by mounting on the backs of the bulla themselves, and leaping from one to another until he has reached clear ground again. The buffalo is never captured in a " pound," as large mammalia are in many countries. He is too powerful a creature to be imprisoned by anything but the strongest stockade fence ; and for this the prairie country does not afford materials. A contrivance, however, of a some- what similar character is occasionally resorted to by va- rious tribes of Indians. When it is known that the buffaloes have become habituated to range in any part of the country, where the plain is intersected by deep ravines, — canons, or barrancas, as they are called, — then a grand battue is got up by driving the animals pellmell over the precipitous bluffs, which universally form the sides of these singular ravines. To guide the herd to the point where it is intended they should take the fatal leap, a singular contrivance is resorted to. This consists in placing two rows of objects — which appear to the buffalo to be human beings — in such a manner that one end of each row abuts upon the edge of the precipice, not very distant from the other, irhile the lines extend far out into the plain, until they have diverged into a wide and extensive funnel. It is simply the contrivance used for guiding animals into a pound ; but, instead of a pair of close log fences, the objects forming these rows 6tand at a considerable distance apart ; and, as already stated, appear to the not very discriminating eye of the buffalo to be human beings. They are in reality de- signed to resemble the human form in a rude fashion and the material out of which they are constructed ifl PRAIRIE INDIANS. 279 neither more nor less than the dung of the buffaloes themselves, — the bois dc vache, as it is called by the Canadian trappers, "who often warm their shins, and roast their buffalo ribs over a fire of this same ma- terial. The decoy being thus set, the mounted hunters next make a wide sweep around the prairie, — including in their deployment such gangs of buffaloes as may be browsing between their line and the mouth of the fun nel. At first the buffaloes are merely guided forward, or driven slowly and with caution, — as boys in snow- time often drive larks toward their snares. When the animals, however, have entered between the converging lines of mock men, a rush, accompanied by hideous yells, is made upon them from behind : the result of which is, that they are impelled forward in a headlong . course towards the precipice. The buffalo is, at best, but a half-blind creature. Through the long, shaggy locks hanging over his front- let he sees objects in -a dubious light, or not at all. He depends more on his scent than his sight ; but though he may scent a living enemy, the keenness of his organ does not warn him of the yawning chasm that opens before him, — not till it is too late to retire : for although be may perceive the fearful leap before taking it, and would willingly turn on his track, and refuse it, he finds it no longer possible to do so. In fact, he is not alk wed lime for reflection. The dense crowd presses from be- hind, and he is left no choice, except that of springing forward or suffering himself to be tumbled over upon his head. In either case it is his last leap ; and, fre« quently, the last of a whole crowd of his companions. 280 THE COMANCHF.S, OR With such persecutions, 1 need hardly say that the buffaloes are becoming scarcer every year ; and it ia predicted that at no distant period this really valuable mammal will be altogether extinct. At present their range is greatly contracted within the wide boundaries which it formerly occupied. Going west from the Mis- sissippi, — at any point below the mouth of the Missouri, — you will not meet with buffalo for the first three hundred miles ; and, though the herds formerly ranged to the south and west of the Rio Grande, the Comanches on the banks of that river no longer know the buffalo, except by their excursions to the grand prairie far to the north of their country. The Great Slave Lake is the northern terminus of the buffalo range ; and west- ward the chain of the Rocky Mountains ; but of late years stray herds have been observed at some points west of these, — impelled through the passes by the hunter-pressure of the horse-Indians from the eastward. Speculators have adopted several ingenious and plau- sible reasons to account for the diminution of the numbers of the buffalo. There is but one cause worth assigning, — a very simple one too, — the horse. With the disappearance of the buffalo, — or perhaps with the thinning of their numbers, — the prairie In- dians may be induced to throw aside their roving habits. This would be a happy result both for them and their neighbors ; though it is even doubtful whether it might follow from such a circumstance. No doubt some change would be effected in their mode of life ; but unfortu- nately these Bedouins of the Western world can live upon* the horse, even if the buffalo were entirely extir- pated. Even as it is. whole tribes of them subsist almosj PRAIRIE INDIANS. - 28a exclusively upon horse-flesh, which they esteem and rel* ish more than any other food. But this resource would, \p time, also fail them ; for they have not the economy to raise a sufficient supply for the demand that would occur were the buffaloes once out of the way : since the caballadas of wild mustangs are by no means so easy to capture as the "gangs" of unwieldy and lumbering buffaloes. It is to be hoped, however, that before the horse-In- dians have been put to this trial, the strong arm of civ- ilization shall be extended over them, and, withholding them from those predatory incursions, which they an- nually make into the Mexican settlements, will induce them to dismount, and turn peaceably to the tillage of the soil, — now so successfully practised by numerous tribes of their race, who dwell in fixed and flourishing homes upon the eastern border of the prairies. At this moment, however, the Comanches are in open hostility with the settlers of the Texan frontier. The lex talionis is in active operation while we write, and every mail brings the account of some sanguinary mas- sacre, or some act of terrible retaliation. The deeds of blood and savage cruelty practised alike by both sides — whites as well as Jbdians — have had their parallel, it is true, but they are not the less revolting to read about. The colonists have suffered much from these Ishmaelites of the West, — these lordly savages, who re- gard industry as a dishonorable calling ; and who fancy that their vast territory should remain an idle hunting- ground, or rather a fortress, to which they might betake themselves during their intervals of war and plundering. The colonists have a clear title to the land, — that title 282 THE COMANCHES, OR acknowledged by all right-thinking men, who behoTO the good of the majority must not be sacrificed to the obstinacy of the individual, or die minority, — that title which gives the right to remove the dwelling of the citizen, — his very castle, — rather than that the public way be impeded. All admit this right ; and just such a title has the Texan colonist to the soil of the Comanche. There may be guilt in the mode of establishing the claim, — there may have been scenes of cruelty, and blood unnecessarily spilt, — but it is some consolation to know that there has occurred nothing yet to parallel in cold-blooded atrocity the annals of Algiers, or the similar acts committed in Southern Africa. The crime of smoke-murder is yet peculiar to Peilisier and Pot- gieter. In their present outbreak, the Comanches have ex- hibited but a poor, short-sighted policy. They will find they have committed a grand error in mistaking the courageous colonists of Texas for the weak Mexicans, — with whom they have long been at war, and whom they have almost invariably conquered. The result is easily told : much blood may be shed on both sides, but it is sure to end as all such contests do ; and the Comanche, like the CaiFre, must " go to the wall." Perhaps it is better that things should be brought to a climax, — it will certainly be better for the wretched remnant of the Spano-Americans dwelling along the Comanche fron- tiers, — a race who for a hundred years have not known peace. As this long-standing hostility with the Mexican na- tion has been a predominant feature in the history of the Comanche Indian, it is necessary to give some ao PRAIKIE INDIANS. 283 count of how it is usually carried on. There was a time when the Spanish nation entertained the hope of Christianizing these rude savages, — that is, taming and training them to something of the condition to which they have brought the Aztec descendants of Montezuma, — a condition scarce differing from slavery itself. As no gold or silver mines had been discovered in Texas, it was not their intention to make mine-laborers of them ; but rather peons, or field-laborers, and tenders of cattle, — precisely as they had done, and were still doing, with the tribes of California. The soldier and the sword had proved a failure, — as in many other parts of Spanish America, — in fact, everywhere, except among the de- generated remnants of monarchical misrule found in Mexico, Bogota, and Peru. In these countries was encountered the debris of a declining civilization, and not, as is generally believed, the children of a progres- sive development ; and of course they gave way, — as the people of all corrupted monarchies must in the end. It was different with the " Indios bravos," or warrior tribes, still free and independent, — the so-called savages. Against these the soldier and the sword proved a com- plete failure ; and it therefore became necessary to use the other kind of conquering power, — the monk and his cross. Among the Comanches this kind of conquest had attained a certain amount of success. Mission-houses sprung up through the whole province of Texas, — the Comanche country, — though the new neophytes were not altogether Comanches, but rather Indians of other tribes who were less warlike. Many Comanches, how- ever, became converts ; and some of the " missiones * became establishments on a grand scale, — each having, 284 THK COMANCHES, OK according to Spanish missionary-fashion, its " presidio* or garrison of troops, to keep the new believers within sound of the bell, and to hunt and bring them back- whenever they endeavored to escape from that Christian vassalage for which they had too rashly exchanged their pagan freedom. All went well, so long as Spain was a power upon the earth, and the Mexican viceroyalty was rich enough to keep the presidios stocked with troopers. The monks led as jolly a life as their prototypes of " Bolton Abbey in the olden time." The neophytes were simply their slaves, receiving, in exchange for the sweat of their brow, baptism, absolution, little pewter crucifixes, and various like valuable commodities. But there came a time when they grew tired of the exchange, and longed for their old life of roving free- dom. Their brethren had obtained the horse ; and this was an additional attraction which a prairie life pre- sented. They grew tired of the petty tricks of the Christian superstition, — to their view less rational than their own, — they grew tired of the toil of constant work, the childlike chastisements inflicted, and sick of the sound of that ever-clanging clapper, — the bell. In fine, they made one desperate effort, and freed them- selves forever. The grand establishment of San Saba, on the river of the same name, fell first. The troops were abroad on some convert-hunting expedition. The Comanches entered the fort, — their tomahawks and war-clubs hid den under their great robes of buffalo-hide : the attack commenced, and ended only with the annihilation of the settlement. PRAIRIE INDIANS. 285 One monk alone escaped the slaughter — a man re- nowned for his holy zeal. He fled towards San Antonio, pursued by a savage band. A large river coursed across the route it was necessary for him to take ; but this did not intercept him : its waters opened for a moment, till the bottom was bare from bank to bank. He crossed without wetting his feet. The waves closed immediately behind him, offering an impassable barrier to his pur- suers, who could only vent their fury in idle curses But the monk could curse too. He had, perhaps, taken rfome lessons at the Vatican ; and, turning round, he anathematized every " mother's son " of the red-skinned savages. The wholesale excommunication produced a wonderful effect. Every one of the accursed fell back where he stood, and lay face upward upon the plain, dead as a post ! The monk, after baptizing the river " Brazos de Dios " (arm of God), continued his flight, and reached San Antonio in safety, — where he duly detailed his miraculous adventure to the credulous con- verts of Bejar, and the other missions. Such is the supposed origin of the name Brazos de Dios., which the second river in Texas bears to this day. It is to be remarked, however, that the river crossed by the monk was the present Colorado, not the Brazos : for, by a curious error of the colonists, the two rivers have made an exchange of titles ! The Comanches — freed from missionary rule, and now equal to their adversaries by possession of the horse — forthwith commenced their plundering expe- ditions ; and, with short intervals of truce, — periods en paz, — have continued them to the present hour. All Northern and Western Texas they soon recovered ; but 280 THE COMANCHES. OR they were not content with territory : they wanted horse? and cattle and chattels, and white wives and slaves ; and it would scarce be credited, were I to state the number of these they have taken within the last half-century. Nearly every year they have been in the habit of mak- ing an expedition to the Mexican settlements of the prov- inces Tamanlipas, New Leon, and Chihuahua, — every expedition a fresh conquest over their feeble and corrupt adversaries. On every occasion they have returned with booty, consisting of horses, cattle, sheep, household uten- sils, and, sad to relate, human captives. Women and children only do they bring back, — the men they kill upon sight. The children may be either male or female, — it matters not which, as these are to be adopted into their tribe, to become future warriors ; and, strange tc relate, many of these, when grown up, not only refuse to return to the land of their birth, but prove the most bitter and dangerous foes to the people from whom they have sprung ! Even the girls and women, after a period, become reconciled to their new home, and no longer de- sire to leave it. Some, when afterwards discovered and ransomed by their kindred, have refused to accept the conditions, but prefer to continue the savage career into which misfortune has introduced them ! Many a heart- rending scene has been the consequence of such appar- ently unnatural predilections. You would wonder why such a state of things has been so long submitted to by a civilized people ; but it is not so much to be wondered at. The selfishness that springs from constant revolutions has destroyed almost every sentiment of patriotism in the Mexican national heart; and, indeed, many of these captives are perhaps PRAIRIE INDIANS. 287 Dot much worse off under the guardiari3hip of the brave Comanches than they would have been, exposed to the petty tyranny and robber-rule that has so long existed in Mexico. Besides, it is doubtful whether the Mexican government, with all her united strergth, could retake them. The Comanche country is as inaccessible to a regular army as the territory of Timbuctoo ; and it will give even the powerful republic of the north no small trouble to reduce these red freebooters io subjection. Mexico had quite despaired of being able to make an effort ; and in the last treaty made between her and the United States, one of the articles was a special agree- ment on the part of the latter to restrain the Comanches from future forays into the Mexican states, and also cause them to deliver up the Mexican captives then ii> the hands of the Indians ! It was computed that their number at the time amounted to four thousand ! It is with regret I have to add, that these unfortunates are still held in bondage. The great republic, too busy with its own concerns, has not carried out the stipulations of the treaty ; and tli3 present Comanche war is but the result of tins criminal negligence. Had energetic measures been adopted at the close of the Mexico-American Avar, the Comanche would not now be harrying the settlers of Texas. To prove the incapacity of the Mexicans to deal with this warlike race, it only needs to consider the pres- ent condition of the northern Mexican states. One h.ilf the territory in that extensive region has returned to the condition of a desert. The isolated "ranchos" have been long since abandoned, — the fields are overgrown with weeds, — and the cattle have run wild, or beea 288 THE COMANCHES, OK earned off by the Comanches. Only the stronger set* tlements and large fortified haciendas any longer exist ; and many of these, too, have been deserted. Where children once played in the security of innocence, — where gayly-dressed cavaliers and elegant ladies amused themselves in the pleasant dia de campo, such scenes are no longer witnessed. The rancho is in ruins, — the door hangs upon its hinge, broken and battered, or has been torn off to feed the camp-fire of the savage ; the dwelling is empty and silent, except when the howling wolf or coyote wakes up the echoes of its walls. About ten years ago, the proud governor of the state of Chihuahua — one of the most energetic soldiers of the Mexican republic — had a son taken captive by the Comanches. Powerful though this man was, he knew it was idle to appeal to arms ; and was only too contented :o recover his child by paying a large ransom ! This fact, more than a volume of words, will illustrate the condition of unhappy Mexico. The Comanche leads a gay, merry life, — he is far from being the Indian of Cooper's description. In scarce- ly any respect does he resemble the sombre son of the forest. He is lively, talkative, and ever ready for a laugh. His butt is the Mexican presidio soldier, whom he holds in too just contempt. He is rarely without a meal. If the buffalo fails him, he can draw a steak from his spare horses, of which he possesses a large herd : besides, there are the wild mustangs, which he can capture on occasions. He has no work to do except war and hunting : at all other times he has slaves to wait upon him, and perform the domestic drudgery. When idle, he sometimes bestows great pains upon hi* PRAIRIE INDIANS. 289 die&s, — which is the usual deer-skin tunic of the prairie Indian, with moccasons and fringed leggings. Sometimes a head-dress of plumes is worn ; sometimes one of the skin of the buffalo's skull, with the horns left on ! The robe of buffalo pelt hangs from his shoulders, with all the grandeur of a toga ; but when he proceeds on a plundering expedition, all these fripperies are thrown aside, and his body appears naked from the waist to the ears. Then only the breech-clout is worn, with leggings and moccasons on his legs and feet. A coat of scarlet paint takes the place of the hunting-shirt, — in order to render his presence more terrific in the eyes of his enemy. It needs not this. Without any disguise, the sight of him is sufficiently horrifying, — sufficiently sugf gestive of " blood and murder." THE PEHUENCHES, OR PAMPAS INDIANS. The vast plain known as the " Pampas " is one of the largest tracts of level country upon the face of the earth. East and west it stretches from the mouth of the Rio de la Plata to the foot-hills of the Andes mountains. It is interrupted on the north by a series of mountains and hill country, that cross from the Andes to the Paraguay River, forming the Sierras of Mendoza, San Luis, and Cordova ; while its southern boundary is not so definite- ly marked, though it may be regarded as ending at the Rio Negro, where it meets, coming up from the south, the desert plains of Patagonia. Geologically, the Pampas (or plains, as the word sig- nifies, in the language of the Peruvian Indians) is an al- luvial formation, — the bed of an ancient sea - upheaved by some unknown cause to its present elevauon, which is not much above the ocean-level. It is not, therefore, a plateau or "table-land," but a vast natural meadow. The soil is in general of a red color, argillaceous in character, and at all points filkd with marine shells and other tes- timonies that the sea once rolled over it. It is in the Pampas formation that many of the fossil monsters have been found, — the gigantic megatherium, the colossal my* THE PEHUENCHES. 291 tocfon, and the giant armadillo (glyptodori), with many other creatures, of such dimensions as to make it a sub- ject of speculation how the earth could b«^e produced food enough for their maintenance. In giving to the Pampas the designation of a vent meadow, do not suffer yourself to be misled by this phrase, — which is here and elsewhere used in rather a loose and indefinite manner. Many large tracts in the Pampas country would correspond well enough to this definition, — both as regards their appearance and the character of the herbage which covers them ; but there are other parts which bear not the slightest resem- blance to a meadow. There are vast tracts thickly covered with tall thistles, — so tall as to reach to the head of a man mounted on horseback, and so thickly set, that neither man nor horse could enter them with- out a path being first cleared for them. Other extensive tracts are grown over with tall grass 60 rank as to resemble reeds or rushes more than grass ; and an equally extensive surface is timbered with small trees, standing thinly and without underwood, like the fruit-trees in an orchard. Again, there are wide mo- rasses and extensive lakes, many of them brackish, and some as salt as the sea itself. In addition to these, there are " salinas," or plains of salt, — the produce of salt lakes, whose waters have evaporated, leaving a stratum of pure salt often over a foot in thickness, and covering their beds to an extent of many square leagues. There are some parts, too, where the Pampas country assumes ^a sterile and stony character, — corresponding to that of the great desert of Patagonia. It is not cor- rect, therefore, to regard the Pampas as one unbroken 292 THE PEHUENCHES, OR tract of meadow. In one character alone is it uniform in being a country without mountains, — 01 any consider able elevations in the way of ridges or hill v., — though a few scattered sierras are found both on its northern and southern edges. The Thistle Pampas, as we take the liberty of naming them, constitute perhaps the most curious section of this great plain ; and not the less so that the " weed " which covers them is supposed not to be an indigenous pro- duction, but to have been carried there by the early colonists. About this, however, there is a difference of opinion. No matter whence sprung, the thistles have flourished luxuriantly, and at this day constitute a marked feature in the scenery of the Pampas. Their position is upon the eastern edge of the great plain, contiguous to the banks of the La Plata ; but from this river they extend backwards into the interior, at some points to the dis- tance of nearly two hundred miles. Over this vast sur- face they grow so thickly that, as already mentioned, it is not possible for either man or horse to make way through them. They can only be traversed by deviou? paths — already formed by constant use, and leading through narrow lanes or glades, where, for some rea- son, the thistles do not choose to grow. Otherwise they cannot be entered even by cattle. These will not, un- less compelled, attempt penetrating such an impervious thicket ; and if a herd driven along the paths should chance to be " stampeded " by any object of terror, and driven to take to the thistles, scarce a head of the whole flock can ever afterwards be recovered. Even the in- stincts of the dumb animals do not enable them to find their way out again ; and they usually perish, eithe? PAMPAS INDIANS. 293 from tlurst, 01 by the claws of the iierce pumas and juguars, which alone find themselves at home in the labyrinthine " car donates, " The little viscacha contrives to make its burrow among them, and must find subsist- ence by feeding upon their leaves and seed, since there is no other herbage upon the ground, — the well-armed thistle usurping the soil, and hindering the growth of any other plants. It may be proper to remark, how- ever, that there are two kinds of these plants, both of which cover large tracts of the plain. One is a tru( thistle, while the other is a weed of the artichoke family, called by the Spanish Americans "cardoon." It is a species of Cardunculus. The two do not mingle their stalks, though both form thickets in a similar mannei and often in the same tract of country. The cartoon is not so tall as the thistle ; and, being without spines, its " beds " are more easily penetrated ; though even among these, it would be easy enough to get entangled and lost. It is proper to remark here, that these thistle-thickets do not shut up the country all the year round. Only for a season, — from the time they have grown up and " shoot," till their tall ripened stalks wither and fall back to the earth, where they soon moulder into decay. The plains are then open and free to all creatures, — man among the rest, — and the Gaucho, with his herds of horses, horned cattle, and sheep, or the troops of roving Indians, spread over and take possession of them. The young thistles now present the appearance of a vast field of turnips; and their leaves, still tendei, are greedily devoured by both cattle and sheep. In this condition the Pampas thistles remain during their short 294 THE PEHUENCHES, OB winter ; but as spring returns, they once more " bristle * up, till, growing taller and stouter, they present a eh*- vaux-de-frise that at length expels all intruders from their domain. On the western selvage of this thistle tract lies the grass-covered section of the Pampas. It is much more extensive than that of the " cardonales," — having an average width of three hundred miles, and running longitudinally throughout he whole northern and south- srn extension of the Pampas. Its chief characteristic is a covering of coarse grass, — which at different sea- sons of the year is short or tall, green, brown, or yel- lowish, according to the different degrees of ripeness. When dry^ it is sometimes fired, — either by design or accident, — as are also the withered stems of the thistles ; and on these occasions a conflagration occurs, stupendous in its effects, — often extending over vast tracts, and reducing everything to black ashes. Nothing can be more melancholy to the eye than the aspect of a burnt pampa. The grass section is succeeded by that of the " open- ings," or scanty forests, already mentioned ; but the trees in many places are more closely set ; assuming the character of thickets, or "jungles." These tracts end among the spurs of the Andes, — which, at some points, are thrown out into the plain, but generally rise up from it abruptly and by a well-defined border. The marshes and bitter lakes above mentioned are the produce of numerous streams, which have their rise in the Great Cordillera of the Andes, and run eastward across the Pampas. A few of these, that trend in a southerly direction, reach the Atlantic by means of the PAMPAS INDIANS. 295 [W(\ ^reat outlets, — the " Colorado " and " Negro." All the others — and " their name is legion " — empty their waters into the morasses and lakes, or sink into the soil of the plains, at a greater or less distance from the Cor- dillera, according to the body of water they may carry down. Evaporation keeps up the equilibrium. Who are the dwellers upon the Pampas ? To whom does this vast pasture-ground belong ? Whose flocks and herds are they that browse upon it ? You will be told that the Pampas belong to the re- public of Buenos Ayres, or rather to the " States of the Argentine Confederation," — that they are inhabited by a class of citizens called " Gauchos," who are of Spanish race, and whose sole occupation is that of herdsmen, breeders of cattle and horses, — men famed for their skill as horsemen, and for their dexterity in the use of the " lazo " and " bolas," — two weapons borrowed from the aboriginal races. All this is but partially true. The proprietorship of this great plain was never actually in 7 the hands of the Buenos-Ayrean government, nor in those of their pre- decessors, — the Spaniards. Neither has ever owned it — either by conquest or otherwise — no further than by an empty boast of ownership ; for, from the day when they first set foot upon its borders to the present hour, neither has ever been able to cross it, or penetrate any great distance into it, without a grand army to back their progress. But their possession virtually ceased at the termination of each melancholy excursion ; and the land relapsed to its original owners. With the exception of some scanty strips along *ts borders, and some wider ranges, thinly o< cupied by the half-nomade Gauchos, the 296 THE PEHUENCHES, OR Pampas are in reality an Indian territory, us tt^y have always been ; and the claim of the white man is no more than nominal, — a mere title upon the map. It ia not the only vast expanse of Spanish American soil that never was Spanish. The true owners of the Pampas, then, are the red aborigines, — the Pampas Indians ; and to give some account of these is now our purpose. Forming so large an extent, it is not likely it should all belong to one united tribe, — that would at once elevate them into the character of a nation. But they are not united. On the contrary, they form several distinct associations, with an endless number of smaller subdivisions or communities, — just in the same way as it is among their prairie cousins of the north. They may all, however, be referred to four grand tribal asso- ciations or nationalities, — the Pehuenches, Puelches^ Picunches, and Ranqneles. Some add the Puilliches, who dwell on the southern rim of the Pampas ; but these, although they extend their excursions over a portion of the great plain, are different from the other Pampas Indians in many re- spects, — altogether a braver and better race of men, and partaking more of the character of the Patago- nians, — both in point of physique and morale, — of which tribes, indeed, they are evidently only a branch. In their dealings with white men, when fairly treated, these have exhibited the same noble bearing which char- acterizes the true Patagonian. I shall not, therefore, lower the standard — neither of their bodies nor their minds — by classing them among " Pampas Indians." Of these tribes — one and all of them — we have, PAMPAS INDIANS. 297 anfoitunately, a much less favorable impression ; and shall therefore be able to say but little to their credit. The different names are al] native. Puelches means the people living to the east, from "pud" east, and che, people. The Picunches derive this appellation, in a similar fashion, from "picun" signifying the north. The Pehuenches are the people of the pine-tree country, from "pehuen" the name for the celebrated "Chili pine" (Araucaria) ; and the Ranqueles are the men who dwell among the thistles, from ranquel, a thistle. These national appellations will give some idea of the locality which each tribe inhabits. The Ranqueles dwell, not among the thistles, — for that would be an unpleas- ant residence, even to a red-skin ; but along the western border of this tract. To the westward of them, and up into the clefts of the Cordilleras extends the country of the Pehuenches ; and northward of both lies the land of the Picunches. Their boundary in that direction should be the frontiers of the quasi-civilized provinces of San Luis and Cordova, but they are not ; for the Picunche can at will extend his plundering forays as far north as he pleases : even to dovetailing them into the similar excursions of his Guaycuru kinsmen from the " Gran Chaco" on the north. The Puelche territory is on the eastern side of the Pampas, and south from Buenos Ayres. At one time these people occupied the country to the banks of the La Plata ; and no doubt it was they who first met the Spaniards in hostile array. Even up to a late period their forays extended almost to Buenos Ayres itself; but Rosas, tyrant as he may have been, was nevertheless a true soldier, and in a grand military expedition again*! 298 THE PEHUENCHES, OR them swept their country, and inflicted such a terrible chastisement upon both them and the neighboring tribes^ as they had not suffered since the days of Mcndoza. The result ha^ been a retirement of the Puelche fron- tier to a much greater distance from Buenos Ayres ; but how long it may continue stationaiy is a question, — no longer than some strong arm — such as that of Rosas — is held threateningly over them. It is usual to inquire whence come a people ; and the question has been asked of the Pampas Indians. It is not difficult to answer. They came from the land of Arauco. Yes, they are the kindred of that famed peo- ple whom the Spaniards could never subdue, — even with all their strength put forth in the effort. They are near kindred too, — the Pehuenches especially, — whose country is only separated from that of the Araucanians by the great Cordillera of Chili ; and with whom, as well as the Spaniards on the Chilian side, they have constant and friendly intercourse. But it must be admitted, that the Araucanians have had far more than their just meed of praise. The ro- mantic stories, in that endless epic of the rhymer Ercilla, have crept into history ; and the credulous Molina has endorsed them : so that the true character of the Arau- canian Indian has never been understood. Brave he has shown himself, beyond doubt, in defending his country against Spanish aggression ; but so, too, has the Carib and Guaraon, — so, too, has the Comanche and Apache, the Yaqui of Sonora, the savage of the Mosquito shore, the Guaycuru of the Gran Chaco, and a score of other Indian tribes, — in whose territory the Spaniard has never dared to fix a settlement. Brave is the Amu- PAMPAS INDIANS. 299 canian; but, beyond this, he has few virtues indeed He is cruel in the extreme, — uncivil and selfish, — filthy and indolent, — a polygamist in the most approved fashiou, — a veiy tyrant over his own, — in short, tak- ing rank among the beastliest of semi-civilized savages, — for it may be here observed, that Jie is not exartly what is termed a savage : that is, he does not go naked, and sleep in the open air. On the contrary, he clothes himself in stuff of his own weaving, — or rather, that of his slave-wives, — and lives in a hut which they build *br him. He owns land, too, — beautiful fields, — of which he makes no use : except to browse a few horses, and sheep, and cattle. For the rest, he is too indolent to pursue agriculture ; and spends most of his time in drinking chicha, or tyrannizing over his wives. This is the heroic Araucanian who inhabits the plains and valleys of Southern Chili. Unfortunately, by passing to the other side of the Andes, he has not improved his manners. The air of the Pampas does not appear to be conducive to virtue ; and upon that side of the mountains it can scarce be said to exist, — even in the shape of personal courage. The men of the pines and thistles seem to have lost this quality, while passing through the snows of the Cordil- leras, or left it behind them, as they have also left the incipient civilization of their race. On the Pampas we find them once more in the character of the true savage : living by the chase or by plunder ; and bartering tlu produce of the latter for the trappings and trinkets of personal adornment, supplied them by the unprincipled white trader. Puelches and Picunches, Pehuenches and Ranqueles, all share this character alike, — all art treacherous, quarrelsome, and cowardly. 300 THE PEHUENCHES, OK But we shall now speak moie particularly of their customs and modes of life, and we may take the " pine people" as our text, — since these are supposed to be most nearly related to the true Araucanians, — and, in- deed, many of their " ways " are exactly the same as those of that " heroic nation." The " people of the pines " are of the ordinary stature of North- American Indians, or of Europeans ; and their natural color is a dark coppery hue. But it is not often you can see them in their natural color : for the Pampas Indians, like nearly all the aboriginal tribes, are " paint- ers." They have pigments of black and white, blue, red, and yellow, — all of which they obtain from dif- ferent colored stones, found in the streams of the Cordil- leras. " Yama," they call the black stone ; " colo," the red ; " palan," the white ; and " codin," the blue ; the yellow they obtain from a sort of argillaceous earth. The stones of each color they submit to a rubbing or grinding process, until a quantity of dust is produced ; which, being mixed with suet, constitutes the paint, ready for being laid on. The Pampas Indians do not confine themselves to any particular " escutcheon." In this respect their fancy is allowed a wide scope, and their fashions change. A face quite black, or red, is a commcn countenance among them ; and often may be seen a single band, of about two inches in width, extending from ear to ear across the eyes and nose. On war excursions they paint hideous figures : not only on their own faces and bodies, but on their trappings, and even upon the bodies of their horses, — aiming to render themselves as appalling as possible in the sight of their enemies. The same trick PAMPAS INDIANS. 301 h employ*! d by the warriors of the prairies, as well as in many other parts of the world. Under ordinary circumstances, the Pampas Indian is not a naked savage On the contrary, he is well clad ; and, so far from ob- taining the material of his garments from the looms of civilized nations, he weaves it for himself, — that is, his wives weave it ; and in such quantity that he has not only enough for his own " wear," but more than enough, a surplus for trade. The cloth is usually a stuff spun and woven from sheep's wool. It is coarse, but durable ; and in the shape of blankets or " ponchos," is eagerly purchased by the Spanish traders. Silver spurs, long, pointed knives, lance-heads, and a few other iron com- modities, constitute the articles of exchange, with various ornamental articles, as beads, rings, bracelets, and large- headed silver bodkins to fasten their cloaks around the shoulders of his " ladies." Nor is he contented with mere tinsel, as other savages are, — he can tell the difference between the real metal and the counterfeit, as well as the most expert assayer ; and if he should fancy to have a pair of silver spurs, not even a Jew pedler could put off upon him the plated " article." In this respect the Araucanian Indian has been distinguished, since his earliest intercourse with Europeans ; and his Pampas kindred are equally subtle in their apprecia- tion. The Pampas Indian, when well dressed, has a cloak upon his shoulders of the thick woollen stuff already described. It is usually woven in colors ; and is not unlike the " poncho " worn by the " gauchos " of Buenos Ayres, or the " serape " of the Mexicans. Besides the cloak, his dress consists of a mere skirt, — also of colored H02 THE PEHUENCHES, OR woollen stuff, being an oblong piece swathed around his loins, and reaching to the knee. A sash or belt — some* times elaborately ornamented — binds the cloth around the waist. Boots of a peculiar construction complete the costume. These are manufactured in a very simple manner. The fresh skin taken from a horse's hind leg is drawn on — just as if it were a stocking — until the heel rests in that part which covered the hock-joint of the original wearer. The superfluous portion is then trimmed to accommodate itself as a covering for the foot; and the boot is not only finished, but put on, — there to remain until it is worn out, and a new one required ! If it should be a little loose at first, that does not matter. The hot sun, combined with the warmth of the wearer's leg, soon contracts the hide, and brings it to " fit like a glove." The head is often left uncovered ; but as often a sort of skull-cap or helmet of horse-skin is worn ; and not unfrequently a high, conical hat of palm- fibre. This last is not a native production, but an im- portation of the traders. So also is a pah* of enormous rings of brass, which are worn in the ears ; and are as bulky as a pair of padlocks. In this costume, mounted on horseback with his long lance in hand, the Pampas Indian would be a picturesque object ; and really is so, when clean ; but that is only on the very rarest occa- sions, — only when he has donned a new suit. At all other times, not only his face and the skin of his body, but every rag upon his back, are covered with grease and filth, — so as to produce an effect rather " tatterde- malion " than picturesque. The " squaw " is costumed somewhat differently First she has a long "robe," which covers her from PAMPAS INDIANS. 303 neck to heels, leaving only her neck and arms bare. The robe is of red or blue woollen stuff of her own weaving. This garment is the " quedeto." A ijelt, embroidered with beads, called " quepique," holds it around the waist, by means of a large silver buckle. This belt is an article of first fashion. Over the shoul- ders hangs the "iquilla," which is a square piece of similar stuff, — but usually of a different dye ; and which is fastened in front by a pin with a large silver head, called the " tupo." The shock of thick, black hair — after having received the usual anointment of mare's tallow, the fashionable hair-oil of the Pampas Indians — is kept in its place by a sort of cap or coiffure, like a shallow dish inverted, and bristling all over with trader's beads. To this a little bell is fastened ; or sometimes a brace of them are worn as ear-rings. These tinkle so agreeably in the ears of the wearer, that she can scarce for a moment hold her head at rest, but keeps rocking it from side to side, as a Spanish coquette would play with her fan. In addition to this varied wardrobe, the Pampas belle carries a large stock of bijouterie, — such as beads and bangles upon her neck, rings and circlets upon her arms, ankles, and fingers ; and, to set her snaky locks in order, she separates them by means of a stiff brush, made from the fibrous roots of a reed. She is picturesque enough, but never pretty. Nature has given the Araucanian woman a plain face ; and all the adornment in the vorld cannot hide its homeliness. The Pchuenche builds no house. He is a true nom- ade, and dwells in a tent, though one of the rudest construction. As it differs entirely from the tent of 304 THE PEHUENCHES, OR the prairie Indians, it may be worth while deacrilv ing it. Its framework is of reeds, — of the same kind as are used for the long lances so often mentioned ; and which resemble bambusa canes. They grow in plenty throughout the Pampas, especially near the mountains* — where they form impenetrable thickets on the borders of the marshy lakes. Any other flexible poles will serve as well, when the canes are not " handy." The poles being procured, one is first bent into a semicircle, and in this shape both ends are stuck into the ground, so as to form an arch about three feet in height. This arch afterwards becomes the doorway or entrance to the tent. The remaining poles are attached to this first one at one end, and at right angles ; and being carried backward with a slight bend, their other ends are inserted into the turf. Tins forms the skeleton of the tent ; and its covering is a horse-skin, or rather a number of horse-skins stitched together, making a sort of large tarpaulin. The skins are sewed with the sinews of the horse or ox, — which are first chewed by the women, until their fibres become separated like hemp, and are afterwards spun by them into twine. The tent is not tall enough to admit of a man stand- ing erect ; and in it the Pehuenche crouches, whenever it snows, rains, or blows cold. He has sheep-skins spread to sleep upon, and other skins to serve as bed-clothes,— all in so filthy a condition, that but for the cold, he might find it far more comfortable to sleep in the open air. He never attempts to sweep out tins miserable lair ; but when the spot becomes very filthy, he " takes up his sticks " and shifts his penates to a fresh " location." He PAMPAS INDIANS. 305 fa generally, however, too indolent to make a " remove," ■*■ until the dirt has accumulated so as to " be in the way." The Pampas Indian is less of a hunter than most other tribes of savages. He has less need to be, — at least, in modern days ; for he is in possession of three kinds of valuable domestic animals, upon which he can subsist without hunting, — horses, horned cattle, and sheep. Of course, these are of colonial origin. He hunts, nevertheless, for amusement, and to vary his food. The larger ostrich (rhea Americana), the guanaco, and the great " gama " stag of the Pampas (cervus campes- tris) are his usual game. These he captures with the bolas, — which is his chief implement for the chase. In the flesh of the stag he may find a variety, but not a delicacy. Its venison would scarce tempt a Lucullian palate, — since even the hungriest Gaucho will not eat it. It is a large beast, often weighing above three hun- dred pounds ; and infecting the air with such a rank odor, that dogs decline to follow it in the chase. This odor is generated in a pair of glands situated near the eyes ; and it has the power of projecting it at will, — just as skunks and polecats when closely chased by an ene- my. If these glands are cut out immediately after the animal is killed, the flesh tastes well enough : otherwise it is too rank to be eatable. The Indians cure it of the " bad smell " by burying it for several days in the ground ; which has the effect of " sweetening " it, while at the same time it makes it more tender. But the Pampas Indian does not rely upon the chase for his subsistence. He is a small grazier in his way and is usually accompanied in his wanderings by a herd 306 THE PEHUENCHES, OR of homed cattle and sheep. He has also his stud dt horses ; which furnish the staple of his food, — for when- ever he hungers, a horse is " slaughtered." Strictly speaking, it is not a horse, for it is the mare f hat is used for this purpose. In no part of the Pampas region, — not even in the white settlement, — are the mares used for riding. It would be considered derogatory to the character of either Gaucho or Indian to mount a mare ; and these are kept only for breeding purposes. Not that the Indian is much of a horse-breeder. He keeps up his stock in quite another way, — by stealing. The same remark will apply to the mode by which he recruits his herds of horned cattle, and his flocks of sheep. The last he values only for their wool ; out of which his gar- ments are woven ; and which has replaced the scantier fleece of the vicuna and guanaco, — the material used by him in days gone by. From whom does he steal these valuable animals, — and hi such numbers as almost to subsist upon them? That is a question that can be easily answered ; though it is not exact language to say that he steals them. Rather say that he takes them, by main force and in open daylight, — takes them from the Creole Spaniard, — the Gaucho and estanciero. Nay, he does not con- tent himself always with four-footed plunder ; but often returns from his forays with a crowd of captives, — wo- men and children, with white skins and ruddy cheeks, — afterwards to be converted into his drudges and slaves- Not alone to the frontier does he extend these plundering expeditions ; but even into the heart of the Spanish set- tlements, — to the estancias of grandees, and the gates f*f fortified tewns ; and, strange as it may read, this con- PAMPAS INDIANS. 807 dition of things has been in existence, not for years, but, at intervals, extending over a century! But what may read stranger still — and I can vouch for it as true — is, that white men actually purchase this plunder from him, — not the human part of it, but the four-footed and the furniture, — fcr this, too, sometime? forms part of his booty. Yes, the surplus, of which th* Indian can make no use or cares nothing about, — more especially the large droves of fine horses, taken fron> the Spaniards of Buenos Ayres, — are driven through the passes of the Cordilleras, and sold to the Spaniard? of Chili ! the people of one province actually encour- aging the robbery of their kindred race in another' The very same condition of things exists in Nortb America. The Comanche steals, or rather takes, from the white settler of Tamaulipas and New Leon, — the Apache rieves from the white settler of Chihuahua and Sonora : both sell to the white settlers, who dwell along the banks of the Rio del Norte ! And all these settlers are of one race, — one country, — one kindred ! These things have hitherto been styled cosas de Mexico. Their signification may be extended to South America : since they are equally cosas de las Pampas. We are not permitted to doubt the truth of these ap- palling facts, — neither as regards the nefarious traffic, nor the captive women and children. At this very hour, not less than four thousand individuals of Spanish-Mexi- can race are held captives by the prairie tribes ; and when Rosas swept the Pampas, he released fifteen hun- dred of similar unfortunates from their worse than Egyp- tian taskmasters, — the Puelches ! With such facts as these before our eyes, who can 308 THE PEHUENCHES. doubt the decline of the Spanish power ? the dtter en. feeblement of that once noble race ? Who can contra- dict the hypothetical prophecy — more than once offered in these pages — that if the two races be left to them- selves, the aboriginal, before the lapse of a single cen- tury, will once more recover the soil ; and his haughty victor be swept from the face of the American conti- nent? Nor need such a change be too keenly regretted. The Spanish occupation of America has been an utter failure. It has served no high human purpose, but the contrary. It has only corrupted and encowardiced a once brave and noble race ; and, savage as may be the character of that which would supplant it, still that savage has within him the elements of a future civilization. Not so the Spaniard. The fire of his civilization has blared up with a high but fitful gleam. It has passed like the lightning's flash. Its sparks have fallen and died out, — never to be rekindled again. CHE YAMPARICOS, OR ROOT-DIGGERS. It id nv.w pretty generally known that there are many enter** m Noith America, — as wild, waste, and inhos- pitable us the lamed Sahara of Africa. These deserts oc- cupy a lar^e ponlon of the central regions of that great continent — extending, north and south, from Mexico to the shores of the Axetic Sea; and east and west for sev- eral hundred miles, on each side of the great vertebral chain of the Rocky Mountains. It is true that in the vast territory thus indicated, the desert is not contin- uous ; but it is equally true that the fertile stripes or valleys that intersect it, bear but a very small propor- tion to the whole surface. Many tracts are there, of larger area than all the Bntish Islands, where the desert is scarce varied by an oasis, and where the very rivers pursue their course amidst rocks and barren sands, with- out a blade of vegetation on their banks. Usually, how- ever, a narrow selvage of green — caused by the growth of cotton woods, willows, and a few humbler plants — de- motes the course of a stream, — a glad sight at all times e be greater or less. In fact, at many places, the widiii of the stream is no longer that of its ordinary channel ; but, on the contrary, a vast " freshet " or inun- dation, covering the country for hundreds of miles, — here flooding over immense marshes or grassy plains, and hiding them altogether, — there flowing among forests of tall trees, the tops of which alone project above the tumult of waters ! These inundations are peculiarly observable in the delta of the Orinoco, — where every year, in the months of July and August, the whole surface of the country becomes changed into a grand fresh-water sea: the tops of the trees alone rising above the flood, and proclaiming that there is land at the bottom. At this season the ordinary channels, or canos, would be obliterated ; and navigation through them become difficult or impossible, but for the tree-tops ; which, after the manner of u buoys " and signal-marks, serve to guide the pilots through the intricate mazes of the " bocas del Orinoco." Now it is this annual inundation, and the semi-sub- mergence of these trees under the flood, that has g/ven origin to the peculiar people of whom we are about to 6peak, — the Guaraons ; or, perhaps, we should rather say, from these causes have arisen their strange habik 546 THE GUARAONS, OR and modes of life which entitle tt.em to be consllered an "odd people." During the period of the inundation, if you should sail up the southern or principal caiio of the Orinoco,—* known as the " boca de navios," or " ships' mouth," — and keep your face to the northward, you would behold the singular spectacle of a forest growing out of the water ! In some places you would perceive single trees, with the upper portion of their straight, branchless trunks rising vertically above the surface, and crowned by about a dozen great fan-shaped leaves, radiating outwards from their summits. At other places, you would see many crowded together, their huge fronds meeting, and form- ing close clumps, or " water groves," whose deep-green color contrasts finely as it flings its reflection on the glis- tening surface below. Were it night, — and your course led you through one of the smaller cafios in the northern part of the delta, — you would behold a spectacle yet more singular, and more difficult to be explained ; a spectacle that astounded and almost terrified the bold navigators, who first ven- tured to explore these intricate coasts. You would not only perceive a forest, growing out of the water ; but, high up among the tops of the trees, you would behold blazing fires, — not the conflagration of the trees them- selves, as if the forest were in flames, — but fires regu- larly built, glowing as from so many furnaces, and cast- ing their red glare upwards upon the broad green leas-eSj and downwards upon the silvery surface of the water ! If you should chance to be near enough to these fires, you would see cooking utensils suspended over them human forms, both of men and women, seated or squat* PALM-DWELLERS. 347 ting around them ; other human forms, flitting like shad ows among the tops of the trees ; and down below, upon the surface of the water, a fleet of canoes (periaguas), fastened with their mooring-ropes to the trunks. All this would surprise you, — as it did the early navigators, — and, very naturally, you would inquire what it could mean. Fires apparently suspended in the air ! human beings moving about among the tops of the trees, talking laughing, and gesticulating ! in a word, acting just as any other savages would do, — for these human beings art savages, — amidst the tents of their encampment, or the houses of their village. In reality it is a village upon which you are gazing, — a village suspended in the air, — a village of the Guaraon Indians ! Let us approach nearer ; let us steal into this water- village — for it would not be always safe to enter it, except by stealth — and see how its singular habitations are constructed, as also in what way their occupants manage to get their living. The village under our ob- servation is now, — at the period of inundation, — nearly a hundred miles from shore, or from any dry land : it will be months before the waters can subside ; and, even then, the country around will partake more of the nature of a quagmire, than of firm soil ; impassable to any human being, — though not to a Guaraon, as we shall presently see. It is true, the canoes, already mentioned, might enable their owners to reach the firm shores be- yond the delta ; and so they do at times ; but it would be a voyage too long and too arduous to be made often, — as for the supply of food and other daily wants, — ' and it is not for this purpose the canoes are kept. No the&e Guaraons visit terra firma only at intervals ; and 348 THE GUARAONS, OR theu for purposes of trade with a portion of their own and other tribes who dwdl there ; but they permanently reside within the area of the inundated forests ; where they are independent, not only of foreign aggression, but also for their supply of all the necessaries of life. In these for- ests, whether flooded or not, they procure everything of which they stand in need, — they there find, to use an old-fashioned phrase, " meat, drink, washing, and lodg ing." In other words : were the inundation to continue forever, and were the Guaraons entirely prohibited from intercourse with the dry land, they could still find sub- sistence in this, their home upon the waters. Whence comes their subsistence ? No doubt you will say that fish is their food ; and drink, of course, they have in abundance ; but this would not be the true ex- planation. It is true they eat fish, and turtle, and the flesh of the manatee, or " fish-cow," — since the captur- ing of these aquatic creatures is one of the chief occupa- tions of the Guaraons, — but they are ofttimes entirely without such food ; for, it is to be observed, that, during the period of the inundations fish are not easily caught, sometimes not at all. At these times the Guaraons would starve — since, like all other savages, they are improvident — were it not that the singular region they inhabit supplies them with another article of food, — one that is inexhaustible. What is this food, and from whence derived ? It will scarce surprise you to hear that it is the produce of the trees already mentioned ; but perhaps you will deem it singular when I tell you that the trees of this great water- forest are all of one kind, — all of the same species,— *o that here we have the remarkable fact of a single PALM-DWELLERS. 349 species of vegetable, growing without care or cultivation, and supplying all the wants of man, — his food, clothing, fuel, utens Is, ropes, houses, and boats, — not even drink excepted, as will presently be seen. The name of this wonderful tree ? " Ita," the Gua- raons call it ; though it is more generally known as "morichi" among the Spanish inhabitants of the Ori- noco ; but I shall here give my young reader an account of it, from which he will learn something more than its name. The ltd is a true palm-tree, belonging to the genus mauritia ; and, I may remark, that notwithstanding the resemblance in sound, the name of the genus is not de- rived from the words " morichi," " murichi," or "muriti," all of which are different Indian appellations of this tree. Mauritia is simply a Latinized designation borrowed from the name of Prince Maurice of Nassau, in whose honor the genus was named. The resemblance, there- fore, is merely accidental. I may add, too, that there are many species of mauritia growing in different parts of tropical America, — some of them palms of large size, ' and towering height, with straight, smootli trunks ; while others are only tiny little trees, scarce taller than a man, and with their trunks thickly covered with conical pro- tuberances or spines. Some of them, moreover, affect a high, dry soil, be- yond the reach of floods ; while others do not prosper, except on tracts habitually marshy, or annually covered with inundations. Of these latter, the it a is perhaps the most conspicuous ; since we have already stated, that for nearly six months of the year it grows literally out of the water. 350 THE 3UARA0NS, OR Like all its congeners, the ita is a " fan-palm ; " that is, its leaves, instead of being pinnately divided, as in most species of palms, or altogether entire, as in some few, radiate from the midrib of the leaf-stalk, intc a broad palmated shape, bearing considerable resem- blance to a fan when opened to its full extent. At the tips these leaflets droop slightly, but at that end where they spring out of the midrib, they are stiff and rigid. TV* petiole, or leafstalk itself, is long, straight, and thick ; and where it clasps the stem or trunk, is swollen out to a foot in width, hollowed, or concave on the upper side. A full-grown leaf, with its petiole, is a wonderful object to look upon. The stalk is a solid beam full twelve feet in length, and the leaf has a diameter of nearly as much. Leaf and stalk together make a load, just as much as one man can carry upon his shoulders ! Set about a dozen of these enormous leaves on the summit of a tall cylindrical column of five feet in cir- cumference, and about one hundred in height, — place them with their stalks clasping or sheathing its top, — so that the spreading fans will point in every direction outwards, inclining slightly upwards; do this,- and you will have the great morichi palm. Perhaps, you may see the trunk swollen at its, middle or near the top, — so that its lower part is thinner than above, — but more often the huge stem is a perfect cylinder. Perhaps you may see several of the leaves drooping downward, as if threatening to fall from the tree ; you may even see them upon the ground where they have fallen, and a splendid ruin they appear. You may see again rising upward out of the very centre of the crown of foliage, a straight, thick-pointed column. This is the young leaf PALM-DWELLERS. 351 In process of development, — its tender leaflets yet un- opened, and closely clasped together. But the fervid tropical sun soon produces expansion ; and a new fan takes the place of the one that has served its time and fallen to the earth, — there to decay, or to be swept off by the flood of waters. Still more may be noticed, while regarding this noble palm. Out of that part of the trunk, — where it is embraced by the sheathing bases of the petioles, — at a certain season of the year, a large spathe will be seen to protrude itself, until it has attained a length of several feet. This spathe is a bract-like sheath, of an imperfect tubular form. It bursts open ; and then appears the huge spadix of flowers, of a whitish-green color, ar- ranged along the flower-stalk in rows, — pinnately. It will be observed, moreover, that these spadices are dif- ferent upon different trees ; for it must be remembered that the mauritia palm is dioecious, — that is, having the female flowers on one tree, and the male or staminif- erous flowers upon another. After the former have glowed for a time in the heat of the sun, and received the fertilizing pollen wafted to them by the breeze, — carried by bee or bird, or transported by some unknown and mysterious agency of nature, — the fruits take form and ripen. These, when fully ripe, have attained to the size of a small apple, and are of a very similar form. They are covered with small brown, smooth scales, — giving them somewhat the appearance of fir-cones, ex- cept that they are roundish instead of being cone-shaped. Underneath the scales there is a thinnish layer of pulp, and then the stone or nut A single spadix will carry carry several hundreds — thousands, I might say — of 352 THE GUARAONS, OR these nuts ; and the whole bunch is a load equal to tbe tjtrength of two ordinary men ! Such is the ita palm. Now for its uses, — the usea to which it is put by the Guaraons. When the Guaraon wishes to build himself a habita- tion, he does not begin by digging a foundation in the earth. In the spongy soil on which he stands, that would be absurd. At a few inches below the surface he would reach water ; and he might dig to a vast depth without finding firm ground. But he has no idea of laying a foundation upon the ground, or of building a house there. He knows that in a few weeks the river will be rising ; and would overtop his roof, however high he might make it. His foundation, therefore, in- stead of being laid in the ground, is placed far above it — just so far, that when the inundation is at its height the floor of his dwelling will be a foot or two above it. He does not take this height from guesswork. That wouTd be a perilous speculation. He is guided by cer- tain marks upon the trunks of palm-trees, — notches which he has himself made on the preceding year, or the natural watermark, which he is able to distinguish by certain appearances on the trees. This point once determined, he proceeds to the building of his house. A few trunks are selected, cut down, and then split into beams of sufficient length. Four fine trees, stand- ing in a quadrangle, have already been selected to form the corner-posts. In each of these, just above the watermark, is cut a deep notch with a horizontal base to serve as a rest for the cross-beams that are to form the foundation of the structure. Into these notches the beam-3 are hoisted, — by means of ropes, — and there PALM DWELLERS 358 lecarely ti

  See Also

  Gene Flow – The Genetic Variation Between Two Populations.In The Origin Of The Eukaryotes, EndosymbiosisWhat Does The Wide Variety Of Different Types Of Animal Eyes Represent?A Paleontologist Finds A New Tyrannosaur In A Site She Is Excavating, And Wishes To Date The Fossil. What Is The Most Likely Method She Will Use?

5. Byers_1999.pdf.txt - KU ScholarWorks - The University of Kansas

• ... bees are more likely to collect from Sphaeralcea (a colorful roadside wildflower) than from the very inconspicuous, low-growing M. leprosa. Thus it is possible ...

• Entomological Contributions in Memory of Byron A. Alexander Edited by George W. Byers, Robert H. Hagen, and Robert W, Brooks (9L ^ / The library ^-? rotr-^^- itive Zoology L- N^ y .', Harvard \ PUBLICATIONS OF THE NATURAL HISTORY MUSEUM, THE UNIVERSITY OF KANSAS The University of Kansas Publications, Museum of Natural History, beginning with Volume 1 in 1946, was discontinued with Volume 20 in 1971. Shorter research papers formerly published in the above series were published as The University of Kansas Natural History Museum Occasional Papers until Number 180 in Decem- ber 1996. The Miscellaneous Publications of The University of Kansas Natural His- tory Museum began with Number 1 in 1946 and ended with Number 68 in Febru- ary 1996. Monographs of The University of Kansas Natural History Museum were initiated in 1970 and discontinued with Number 8 in 1992. The University of Kan- sas Science Bulletin, beginning with Volume 1 in 1902, was discontinued with Vol- ume 55 in 1996. The foregoing publications are now combined in a new series entitled Scientific Papers, Natural History Museum, The University of Kansas, be- gun with Number 1 in 1997. Special Publicahons began in 1976 and continue as an outlet for longer contributions and are available by purchase only. All manuscripts are subject to critical review by intra- and extramural specialists; final acceptance is at the discretion of the editor. The publication is printed on acid-free paper. Publications are composed us- ing Microsoft Word" and Adobe PageMaker* on a Macintosh computer and are printed by The University of Kansas Printing Services. Inshtutional libraries interested in exchanging publicahons may obtain the Scienhfic papers. Natural History Museum, The University of Kansas, by address- ing the Exchange Librarian, The University of Kansas Libraries, Lawrence, Kansas 66045-2800, USA. Available back issues of The University of Kansas Science Bulle- tin may be purchased from the Library Sales Section, Retrieval Services Depart- ment, The University of Kansas Libraries, Lawrence, Kansas 66045-2800, USA. Avail- able issues of former publicaHon series, Scienhfic Papers, and Special Publicahons of the Natural History Museum can be purchased from the Office of Publicahons, Natural History Museum, The University of Kansas, Lawrence, Kansas 66045-2454, USA. Purchasing informahon can be obtained by calling (785) 864-4450, fax (785) 864-5335, or e-mail (kunhm@ukans.edu). VISA and MasterCard accepted; include expirahon date. Series Editor: William E. Duellman Printed BY The UNTVERsm- of Kansas Printing Services Lawrence, Kansas The UNrvERSiTi' of Kansas Natural History Museum Special Publication No. 24 17 December 1999 Entomological Contributions in Memory of Byron A. Alexander Edited by George W. Byers Professor Emeritus, Siiozi' Entomological Collection, Division of Entomology, Natural History Museum and Biodiversity Research Center, The University of Kansas, Lawrence, Kansas, 66045, USA Robert H. Hagen Adjunct Assistant Professor, Department of Entomology, The U)ui'ersity of Kansas, Lawrence, Kaiisas, 66045, USA AND Robert W. Brooks Collection Maiuiger, Snow Entomological Collection, Divsion of Entomology, Natural History Museum and Biodiversity Research Center, The University of Kansas, Lawrence, Kansas, 66045, USA Natural History Museum The University of Kansas Lawrence, Kansas c (1 SPECIAL PUBLICATIONS NATURAL HISTORY MUSEUM, THE UNIVERSITY OF KANSAS Series Editor: William E. Duellman MCZ LIBRARY FEB 1 6 2000 HARVARD UNlVERolTv Special Publication No. 24 pp. i-iv, 1-252 Published 10 December 1999 ISBN 0-89338-057-1 © 1999 BY Natural History Museum Dyche Hall The UNivERsirt' of ICansas Lawrence, Kansas 66045-2454 USA Printed by University of Kansas Printing Service Lawrence, Kansas CONTENTS Introduction George W. Byers, Robert H. Hageii, niui Clare T. Wiielliier 1 The Augochlorine Bee Genus Megaloptilla (Hymenoptera: Halictidae) Michael S. Engel and Robert W. Brooks 9 Redescription of Linsleyonides Skiles (Coleoptera: Cerambycidae) and Inclusion of Elaphidion pobtoricensis Fisher Steven W. Lingafelter 17 Anthidium oblongatum (Illiger): an Old World Bee (Hymenoptera: Megachilidae) New to North America, and New North American Records for Another Adventive Species, A. manicatum (L.) E. Richard Hocbeke and A. G. Wheeler, Jr 21 Distribution and ethology of Ceramius damarinus Turner (Hymenoptera: Vespidae: Masarinae) in Namibia Sarah K. Gess 25 The Ultrastructure of the Wall and Lining Epithelium of Glandular Pouches in Nomadine Bees (Hymenoptera: Apidae: Nomadinae) Bruce Cutler and Byron A. Alexander 33 Mating Behavior of Dianthidium curvatum (Hymenoptera: Megachilidae) at a Nest Aggregation Gail R. Michener and Charles D. Michener 37 Beflwior and Subcaste Specialization Among Workers of the Giant Tropical Ant, Paraponera clavata (Hymenoptera: Formicidae: Ponerinae) Rodney S. Hanley and James P. Lovett 45 The Female of Diadasia afflictula Cockerell Unveiled and Verified Using Molecular Markers (Hymenoptera: Apidae) Sedonia D. Sipes 51 Nesting Biology and Immature Stages of the South American Bee Genus Acamptopoeum (Hymenoptera: Andrenidae: Panurginae) Jerome G. Rozen, Jr. and D. Yanega 59 An Inventory Of Arthropod Fauna At Great Sand Dunes National Monument, Colorado Michael J. Weissmann and Boris C. Kondratieff 69 The Distribution of Halictus ligatus Say and H. poeyi Lep. (Hymenoptera: Halictidae) in North America Laurence Packer 81 Origins of Symbiosis: Phylogenetic Patterns of Social Insect Inquilinism in Cryptophagidae (Coleoptera: Cucujoidea) Richard A. B. Leschen 85 Timing of Mating Flights of Neotropical African and European Honey Bee Queens and Drones (Hymenoptera: Apidae) in Eastern Venezuela Gard W. Otis, Orley R. Taylor, Jr., Maria Spivak, Mark L. Winston, Susan J. Katz, and Penelope F. Kukuk 103 £/CKWORr/.4 (Apoidea: Halictidae), a New Genus of Bees from Mesoamerica Ronald J. McGinley Ill A New Bittacus from Southern Mexico (Mecoptera) George W. Byers 121 Resolving Conflict Between Morphological and Molecular Evidence for the Origin of Eusociality in the "Corbiculate" Bees: A Hypothesis-Testing Approach TedR. Schultz, Michael S. Engel, and Michael Prentice 125 Una Especie Nueva de Smeringodynerus (Hymenoptera: Vespidae: Eumeninae) de Mexico Alicia Rodriguez-Palafox 139 iii Behavior of a Cleptoparasitic Bee, Triepeolus distinctus (Hymenoptera: Nomadinae), Before Departing From the Nest of Its Host, Dieunomia triangulifera (Hymenoptera: Halictidae) Clare T. Wuclhier mid Mark S. Hi.wn 143 Description of Preimaginal Instars of Four Species of Elampini, with Some Notes on the Phylogenetic Importance of Larval Characters in this Tribe (H\'menoptera: Chrysididae) jose Tormos, Karl V. Krombeiu, Josep D. Asis, and Severiano F. Gayubo 151 Local Acoustics versus Host Plant Resources: Determinants of Calling Sites in a Tropical Katydid, Orophus conspersus (Orthoptera: Tettigoniidae) Sara L. Taliaferro, Danel Vickerman, Michael D. Greenfield 157 Revision of the Species of the Subgenera of Exoamlops/s Spinola, 1853, Occurring in South America. I: Diomalopsis MiCHENER & Moure, 1957 (Hymenoptera: Apidae), and a Revised Key to the Subgenera Eduardo A. B. Almeida and Fernando A. Silveira 167 New Species, Phylogenetic Placement, and Mammal Associations of Loberopsyllus (Languriidae: Xenoscelinae) Richard A. B. Leschen and James S. Ashe 171 Floating and Nest Adoption in a Nesting Aggregation of the Mexican Opuntia Bee Diadasia knabiana (Apidae) RudolfJander 179 Novel Use of Walking Trails b\ the Amazonian Bumble Bee, Bombus transversaiis (Hymenoptera: Apidae) Sydney A. Cameron, James B. Whitfield, Miles Cohen, and Natalie Thorp 187 Influence of Three Factors on African Honey Bee Swarms' Preference for Bait Hives in Mexico Carlos H. Vergara 195 Population Genetics of the Greater Wax Moth, Galleria mellonella L. (Lepidoptera: Pyralidae) Amy M. McMillan 201 A New Species of Diadasiopus (Acari: Acaridae) Associated with Diadasia chiliensis (Hymenoptera: Apidae) in Chile Barry M. OConnor and Gerald Daneshvar 207 Life on the Edge: Object Orientation and Aquatic Locomotion in Paederus arduus Sharp (Coleoptera: Staphylinidae) Sandra Haase-Statz and Elizabeth F. Smith 211 Odyneropsis, a Genus New to the United States, with Descriptions of Other New Cleptoparasitic Apidae (Hymenoptera: Apoidea) Terry Griswold and Frank D. Parker 217 Two New Genera of Pemphredonine Wasps from Australia (Hymenoptera: Apoidea, Crabronidae) Gabriel A. R. Melo and Ian Naumann 221 Reassessment of the Bee Genus Chaeturginus (Apoidea: Andrenidae, Panurginae), with the Description of a New Species From Southeastern Brazil Luisa Riiz and Gabriel A. R. Melo 231 Prolegs of Papilionini (Lepidoptera: Papilionidae): Alternative Solutions to the Problem of Attachment Robert H. Hagen 237 IV Bxjers, G. W., R. H. Hogcii, and R. W. Brooks (eds.). Entomological Contributions in Memory of Byron A. Alexander. University of Kansas Natural History Museum Special Publication 24. (1999) Pp. 1-S Byron A. Alexander (1952-1996) By George W. Byers\ Robert H. Hagen^ and Clare T. Wuellner'' BIOGRAPHY This collection of entomological contributions resulted from a feeling among his friends and colleagues that Byron Alexander deserved more than a memorial service and an obituary, following his tragically premature death on 30 November 1996. A biologist with broad interests and ex- periences, Byron had earned the friendship and respect of associates in a variety of fields and from institutions around the world, as indicated by the papers that make up this volume. Despite his many interests, Byron had worked mainly on the systematics, behavior, and morphology of bees and sphecid wasps. The published results of his re- search are listed at the end of this introduction. Reading this list, one is impressed by the amount of scholarly pro- ductivity in Byron's short professional career. Byron was an outstanding teacher, popular among his students not only for his knowledge of his subjects but also for his enthusiasm and his sense of humor His courses at the University of Kansas included insect classification, external morphology of insects, social insects and intro- ductory systematics. He also had taught two undergradu- ate biology courses and a summer field course in ento- mology; and he had helped teach two short courses at other institutions, one in Mexico, at the Centro de Ecologia, Hermosillo, the other in Honolulu, sponsored by the Uni- versity of Hawaii and the Bernice P. Bishop Museum. While a graduate student at Cornell University, he received an award for outstanding teaching. The amount of his teach- ing is the more remarkable because in addition to his ap- pointments in the departments of Entomology and of Sys- tematics and Ecology, Byron held a half-time curatorial position in the Snow Entomological Museum. Byron was an exceptionally gifted artist and was able to apply this talent to some of his publications, such as the "Comparative Morphology of the Female Reproductive System of Nomadine Bees" (Memoirs of the Entomologi- cal Society of Washington, 1996). During the summers in which he was employed by the National Park Service, Byron illustrated several brochures on wildlife in various parks. In 1984, he exhibited some of his drawings at a na- tional meeting of the Guild of Natural Science Illustrators, and four years later he again displayed drawings, this time for the Eastern Branch of the Entomological Society of America. He also illustrated a book on "The Natural His- tory and Behavior of North American Beewolves" (Philnnthiis wasps), by Howard Evans and Kevin O'Neill (1988; Fig. 1). His drawing of a sphecid wasp, Glenostictin pictifrons, appears on the cover of this volume, and has been used on a recruitment poster by the Department of Entomology at the University of Kansas. Byron Allen Alexander was born in Austin, Texas, on 14 April 1952, one of three sons of Harold and Betty Alexander. (In an earher obituary [Journal of Hymenoptera Research, 6: 186-189, 1997], his birthplace was incorrectly reported as El Paso.) The family moved to El Paso, where Byron later attended the University of Texas at El Paso, ' Snow Entomological Collection, Division of Entomology, Natural History Museum, University of Kansas, Lawrence, Kansas 66045. ^ Department of Entomology, University of Kansas, Lawrence, Kansas 66045. ' Department of Zoology, University of Texas, Austin, Texas 78712. The University of Kansas Natural History Museum Special Publication No. 24 graduating with highest honors in 1974. Interested in ani- mal behavior, Byron entered a graduate program in pri- matology at Stanford University, intending to work with Jane Goodall in Tanzania. However, in May, 1975, the re- search station at Gombe Reserve was attacked by rebels who kidnapped two American students. The field research program in primatology was terminated in response, and Byron left Stanford at the end of 1975. The next year he was able to join a Scottish research group studying chim- panzees in Senegal. After only 6 months, however, he be- came ill with hepatitis and had to return to the United States. Briefly in 1976, and again in 1978-1981, he was a seasonal park naturalist at Capitol Reef National Park in Utah, Great Sand Dunes National Monument in Colorado, and Tuzigoot National Monument in Arizona. While work- ing in the Great Sand Dunes, Byron became interested in entomology, particularly in the behavior of wasps, after meeting students of Prof. Howard Evans of Colorado State University. Byron enrolled at Colorado State to work with Evans and, in 1983, earned the M.S. degree there. He then entered Cornell University, where he studied behavior and systematics of Hymenoptera under Prof. George Eickwort. While at Cornell, Byron was awarded a National Science Foundation Graduate Fellowship, the John Henry Comstock Scholarship, and three other fellowships. He received the Ph.D. degree in 1989 and in the summer of that year joined the Department of Entomology at the University of Kansas as an assistant professor. In 1995, he was promoted to associate professor with tenure. During his professional career, Byron was a member of the Entomological Society of America (associate editor of the Thomas Say Pubhcations in Entomology, 1994-1996), the International Society of Hymenopterists, the Animal Behavior Society, the International Union for the Study of Social Insects, the Central States Entomological Society (president in 1995-1996; member of the editorial board of the Journal of the Kansas [Central States] Entomological Society, 1994-1996), the Society of Systematic Biology, the American Association for the Advancement of Science, and the Sigma Xi Scientific Research Society. George W. Byers BYRON AS NATURALIST AND SCIENTIST While still a sttident at the University of Texas at El Paso (UTEP), Byron met Art Metcalf, a professor in the Biology Department. They began a lifelong friendship, maintained through visits and correspondence. After Byron's death. Prof. Metcalf organized and transcribed the letters he had received from Byron over the years. The let- ters demonstrate Byron's skills as a keen observer of the natural and human environment, and vividly illustrate his development as an entomologist. Excerpts from some of these letters are included below, courtesy of Prof. Metcalf. When Byron left Stanford at the end of 1975, he was still determined to pursue field research in primatology. In the spring of 1976 he attended classes at UTEP, then took a summer job with the National Park Service while he considered his alternatives: (22 June 1976) ... I am presently working as a seasonal naturalist at Capitol Reef National Park [Utah], answering questions at the Visitor Center, leading conducted hikes, and giving evening pro- grams. . . . This park is relatively remote and obscure, so we don't have too many visitors very often, and the days of bad craziness are rare. However, one Memorial Day weekend was enough to persuade me that 1 wouldn't want to make a permanent career of working for the Park Service. One could get transferred to Yosemite. Nevertheless, there are abundant compensations to outweigh the tribulations of this job. This is without doubt the most beauti- ful place I have ever lived in. Just beyond my back fence is a de- lightful little creek patrolled by swallows and lined with a few magnificently gnarled and twisted, patriarchal cottonwoods. For a backdrop, there is a 900 foot wall of Wingate Sandstone. Everywhere the landscape is dominated by rock: awesome, immense, intricately sculptured and textured, brilliantly colored. An ideal place to regain, or lose, your sanity. Right now 1 am sit- ting amidst a forest of pinyon and juniper on Cedar Mesa, where the wind always blows. The sun is almost gone, its last rays light- ing the top of tarantula Mesa, gilding the strangely eroded forms of the Mancos Shale. Surely this is a good place to be, and at least for now past, present, and future are woven together in a single tapestry of timelessness. Byron arranged to join a research group based at St. Andrews University in Scotland that was studying chim- panzees in West Africa. He was able to pay for his passage across the Atlantic by working as a scientific illustrator for a UTEP herpetologist, and arrived at the field site in Senegal in the spring of 1977: (8 May 1977) The country around our camp at Mt. Assirik [Senegal] is very dry now, right at the end of the 7-month dry season. If this were the Southwest, Mt. Assirik would undoubt- edly be called a mesa. Its slopes are covered in woodland com- posed largely of Caesalpiniaceae (AfzeUa afrkana, Erythrophkum simi'tvlens) and Mimosaceae (Parkin Inglobosa), with an occasional baobab or rhum palm. The streams draining Mt. Assirik soon en- counter a layer of very hard, resistant rock (all rock around here, as well as dust, soil, or anything else to do with the earth, is called laterite), which forces them to cut steep, narrow gorges separated by broad stretches of flat land locally known as plateaux. The con- trast between the vegetation of the gorges and that of the plateaux is one of the most abrupt and striking I have ever seen. At this time of the year, the plateaux are almost completely barren, with only scattered stands of a shrubby tree called Comhretum (its fruits are like 4-winged saltbush) and occasional patches of grass which were missed by the fires. On the other hand, the gorges support a dense, luxurious, jungle-like vegetation called gallery forest, for they are the only places where there is a permanent source of sur- face water. The gallery forest is characterized by an abundance of vines, creepers, and strangling figs, as well as tall, stately trees with enormous buttressed trunks. At least during the dry season, it is also riddled with game trails linking the few precious remain- ing pools of water. So the chimps are living in an area with a wide variety of habi- tat types: woodland, forest, grassland, bamboo. Nevertheless, one Byron A. Alexander can't help feeling that it is marginal habitat, although that is noth- ing more than a subjective impression which is undoubtedly in- fluenced bv one's personal reaction to the country. It's damned hot here—last month the mean daily maximum temperature was 104 "F—and there are some annoying insects around, like sweat bees, tsetse flies, and vicious stinging bees. Consequently, one has to wear protective clothing: heavy duty long-sleeved shirts & trou- sers, a hat of some kind, canvas boots with gaiters. None of us is able to stay out in the field all day without being thoroughly dev- astated (or "knackered" as my English companions say) by the experience, so we tend to take a prolonged mid-day siesta. (4 June 1977) ... [Tambacoundal is crowded, filthy, ugly, reeking, hot, foreign, strange, threatening. Perhaps it is because I grew up on the border or something, but poverty, squalor, and overpopu- lation hold no exotic charm for me. Nor does the ravaged coun- tryside surrounding the "misty villages". Mist indeed! It's dust. Dust from the dry season, from the ceaseless pounding of hooves, from the relentless pressure of myriads of scrawny, undernour- ished cows, goats, sheep which immediately devour any blade of grass which is foolish, desperate, or unfortunate enough to at- tempt to raise its head above the parched ground. A depressingly familiar landscape, reminiscent of the country around Ojinaga [Mexico], or the Navajo reservation, or perhaps even the majority of the land surface inhabited by Homo sapiens. Still, it is true that Africa does have a special enchantment about it, an intangible essence, a special vitality which my humble pow- ers of description can never hope to evoke. Even the stench of the market place is at least ripe with the odors of life rather than stag- nation. And never before have 1 experienced such a vibrant, puls- ing, breathtaking symphony of colors. The design is bold and out- landish, without a trace of subtlety, every hue glowing with its utmost intensity, screaming reds, exploding yellows, piercing blues, exhilarating greens. My eyeballs are seared, my mind is reeling. Not at all like picking up a loaf of bread at your friendly neighborhood Safeway. The rains have come, and their arrival was indeed dramatic. As usual, the rains are said to be unusual this year. They are un- usual because they arrived suddenly, with 3 consecutive days of heavy rain totalling 110 mm. Those were awesome storms, quite reminiscent of a summer cloudburst in West Texas (or Kansas?). The change they wrought was abrupt and complete. Suddenly the air is clear and fresh, and all the subdued and dusty hues of the dry season have been washed away to reveal the inner fire of True Color which burns within all things. Already the plateaux are showing a faint flush of green, exquisite white lilies are erupt- ing from bare rock, shrubs are exploding into leaf or flower, ter- mites are swarming suicidally in mindless, milling hordes, the tse- tses are emboldened, the chimps have vanished without a trace. Life is asserting itself; the strange and improbable combinations of protons and electrons are whirKng feverishly in their frenzied, intricate, beautiful, irresistible dance. ! can't help joining in, al- though 1 have no idea what it's all about. Byron returned to the U. S. in October 1977. Having experienced first hand the difficulty of field research on great apes, he decided not to pursue an academic career in prima tology. However, he still hoped to return to West Africa, and applied for a position with the Peace Corps. While he waited, Byron supported himself by seasonal work with the Park Service and by freelance illustration jobs (Fig. 2). A chance encounter at Great Sand Dunes Na- tional Monument set him on the path towards entomol- ogy: (27 Sept 1978) An unexpected delight of the summer was meet- ing Darryl Gwynne, a British-born immigrant from Canada who is doing graduate work at Colorado State. The subject of Darryl's Eig. 1. Territorial defense by Philnitthus hiciuctus males. Great Sand Dunes National Monument. Byron used this as part of a composite illustration for The NnturnI History ami Behavior ofNortli Americnn Bcewolivs. (Howard E. Evans and Kevin M. O'Neill, Cornell University Press, Ithaca, NY, 1988.) studies is Pltilaiitlnis bicinctus, a congener of Tinbergen's European bee wolf. Darryl's enthusiasm for his wasps is infectious, and they are intriguing, if somewhat macabre, creatures. From a primatologist's perspective, there are also other attractive aspects of wasp research. Bumblebee wolves (as Darryl has christened his wasps) don't generally become active until around 9:00 a.m., and generally knock off in mid-afternoon. A generation only lasts about 6-8 weeks anyway. But those few short weeks are a time of intense activity for a multitude of individuals (enough to satisfy the most unyielding statistical demands). I doubt that I would have been able to discern much order in the frenzy of a P. bicinctus nesting area, but with the tutoring of Darryl's trained eye (this was his third year of observations) 1 was able to learn something of the bizarre life of Bumblebee Wolf Arroyo. The most conspicuous ac- tivity in the nesting area was the territorial behavior of the males, who displav an insane singleness of purpose in defending areas around female burrows. They will attack a pebble if you toss one through their territory (a handy trait if one wishes to census terri- torial males). Nor do thev hesitate to approach such predators as robber tlies, which makes them ridiculously vulnerable to preda- tion. But 1 suppose there is a limit to how many wasps a robber fly can suck dry before it becomes too bloated to fly. ... P. bicinctus is only one of many species of solitary wasps which thrive in the rich mosaic of sandy substrates here. Darryl's advisor is Howard Evans, who has gained some notoriety with his lifetime of work on the comparative ethology of various soli- tary wasps. He made a couple of visits to Great Sand Dunes this summer, and learned some interesting things. Just up the arroyo from Darryl's P. bicinctus nesting area is an aggregation of P. zebrntus burrows. In excavating one of these P. zebratus burrows, Evans found that the cells had been provisioned with males of P. bicinctus and P. zebrntus. So he suggested that P. zebratus be christened the bumblebee wolf wolf (or even the bumblebee wolf wolf wolf.). Another interesting observation of P. zebratus this summer was the discovery of 2 males simultaneously in copula with a single female. I can just imagine their respective sperm cells violently flagellating away at one another in a heated race to the ovum (or would it be ova? My knowledge of basic insect biology is still very shaky.) The University oe Kansas Natural History Museum Special Publication No. 24 Eventually, tired of waiting for the Peace Corps to act on his application, Byron turned toward other goals and decided to pursue graduate study in insect behavior: (29 January 1981 ) I've just returned from a trip to Colorado State and the University of Colorado. Both have good behavior pro- grams and both are encouraging about my chances of being ac- cepted as a graduate student, despite my somewhat erratic past. . .. I sat in on one day of a conference of the Entomological Soci- ety of America just after Thanksgiving. There was supposed to be a symposium on the sociobiologv of the social insects, but the main event seemed to be getting dnmk at the hospitality suites pro- vided by the chemical companies that manufacture pesticides. I also had my very first meal at a Japanese restaurant. Across the table, John Alcock, Randy Thornhill, and somebody from Utah State were scribbling mathematical models on a napkin and hav- ing an animated discussion about confidence of paternity being irrelevant to male paternal investment. Counter-intuitive stuff, quite risque intellectually. 1 learneci that chicken teriyaki is very tasty. (6 April 1981 ) ... I have decided to accept a T.A. that was offered to me by Colorado State. 1 think there are still some papers 1 have to sign to make it all official, but 1 have the impression that this is just routine procedure; and it's already too late to change mv mind. Not that I'm feeUng any inclination to do so, but there is a certain finaUty about signing documents. I'm going to be working under Howard Evans. I have 2 pos- sible research projects in mind: either a study of the nesting be- havior of Clypeadoii, a digger wasp which preys on harvester ants and isn't as common as one might expect for a wasp whose prey is so abundant; or a study of how parasitic miltogrammine flies recognize potential hosts, and how those hosts avoid their para- sites. The former is more specific and straightforward, and there- fore probably a better place for me to start. As it turns out, Howard suspects that miltogrammine flies are responsible for the relative scarcity of Oyyieadon, so the 2 problems are not unrelated. No doubt one could spend an entire lifetime studying the intricacies of the relationships of miltogrammine flies and sphecid wasps—but I'm not quite ready to make such a commitment myself. Fortunately, nobody has drawn up any contracts for me to sign. Byron completed his Master's thesis in the spring of 1983 (Alexander, 1985, 1986), then moved to Cornell Uni- versity to begin his Ph. D. work under George Eickwort: (12 March 1983) ... Fort Collins has been having a very mild winter, and now it appears to be over Grass is greening up all over town, trees are scattering pollen, green blades of daffodils and irises are rising, flickers are drumming, robins are trying out their familiar phrases, Canada geese are chasing each other across the pond down the road, honeybee workers are taking brief scout- ing trips, and the harvester ants are back at work on their mounds. It's difficult to concentrate on my thesis, but with an April 8 dead- line I haven't much time for Spring Fever 1 have at last received written comments from all my committee members, so I can spend Spring Break working on the final draft. If all goes well this week, perhaps 1 can allow myself a weekend trip to Great Sand Dunes. Then again, perhaps it will snow. I try to remind myself from time to time that this is Colorado. Next year I'll have to learn the vagaries of the climate of New York's Finger Lakes district. Cornell offered me a fellowship, which 1 was happy to accept. I hope they won't change their minds when I get there and it becomes evident that I'm not very Ivv-Leaguish (in fact, I'm not even sure how to spell it). On Monday evening I'll be going to my advisor's house to see some slides and find out what I'm getting mto. I have visions of warblers in Sapsucker Woods, but I suppose there will be more to it than that. The man who has agreed to be my advisor, George Eickwort, works on the behavior and svstematics of bees and mites, although he has a few students working on wasps and social spiders. (5 January 1984) ... I have been pleasantly surprised by how much I am enjoying Cornell. I'm devoting a lot of time and energy to filling in the substantial gaps in my understanding of basic insect biology (i.e. morphology, physiology, systematics, ecology), since there are good courses dealing with all these topics. At the same time, there is an entire department of Neurobiology and Behav- ior, with people working cm birds, mammals, fish, and even spi- ders. There is one graduate student studying grooming behavior in yellow baboons at Amboseh. So 1 have plenty of opportunity to keep in touch with ideas and observations in the field of behavior The entomology department has an active group of people inter- ested in systematics. 1 never encountered anything like this at Stanford or CSU (Howard Evans kept his taxonomy to himself). Consequently, I was quite baffled by Jim Carpenter, who was fin- ishing up his Ph.D. just as I arrived here, and who generously put me up in his house while I looked for a place to live. He wore T- shirts sporting such esoteric but obviously outrageous slogans as: "Cladistics-it's the real thing" and "Crush paraphyly under the iron heel of true science." (30 November 1 985) I had my first brief personal exposure to the mysteries of island biogeography in May Cornell University sent a small-scale collecting expedition to Puerto Rico. I was able to tag along as a field assistant on my major professor's grant, since classes were still in session, so that the undergraduate assistant he had hired for the summer wasn't yet available to work for him. Three members of our expedition were coleopterists so they spent their days breaking open logs and turning over rocks in the rain- drenched, mossy, pristine forests of the Central Highlands. How- ever, I was on the sweat bee team. We found the best nesting sites to be on Puerto Rico's public beaches (the name they used for these beaches was "balneario"). We were never arrested, or even questioned by the police, but I don't think our unusual behavior went unnoticed. For our part, we kept remarking that island fau- nas really are depauperate. It is easier to find sweat bees on a spring day in Ithaca than in Puerto Rico. The landscape looked and felt unmistakably tropical, but the much-vaunted diversity of the trop- ics was not there. The plants and animals we saw looked tropi- cal—bamboo, tree ferns, cecropia, flame trees, bananaquits, todies (not many of these), grackles, frigate birds, and the ubiquitous little treefrogs (called "coqui", which is a fair description of what they sound like). But there were lots of Neotropical things which were not on the islands—no monkeys, or motmots, or toucans, or armadillos, or stingless bees, or army ants or euglossine bees— the hst could go on and on. I realize that this isn't any surprise to a sophisticated biogeographer, but it's almost eerie to actually experience it. It's enough to make one believe in empty ecological niches. I've spent the summer and fall beginning to get acquainted with Nomada, the genus of bees whose phylogeny I'm trying to unravel. No doubt 1 would be more efficient if I knew what I was doing, but if I knew that would 1 be here? 1 was curious about whether a cleptoparasitic way of life would result in modifica- tions of the female reproductive tract, so I began dissecting bees this summer 1 found some unexpected and unexplained struc- tures—a pair of thin-walled sacs at the base of the reproductive tract. Whatever they're for, they seem to be present in all the mem- bers of the subfamily Nomadinae, and absent from other bees. 1 also diverted myself with trying to devise a technique for illus- trating these structures in a way that would show how very nearly transparent they are. Byron completed his Ph. D. dissertation in 1989 and moved to Kansas as an assistant professor in the Depart- ment of Entomology: (19 September 1989) Although my life has taken many strange turns, this is surely one of the strangest. I find myself on the fac- Byron A. Alexander ulty of the University of Kansas, in a tenure track position that is a joint appointment with the Snow Entomological Museum, the Entomology Dept. and the Systematics & Ecology Dept. Needless to say, I'm delighted to be here. After all, thev do have an excel- lent bee collection, which is not the sort of thing that universities routinely support. I also find Lawrence to be a very pleasant place to live. Ear nicer than, say, Chicago or Philadelphia where two other openings for an insect systematist were advertised this year. So I've been e.xtraordinarily lucky 1 had to really push to finish my thesis in time to start here for the fall semester, but at least it's over now and 1 won't ever have to do that again. Of course, it's out of the frying pan and into the fire, but 1 should be old enough to expect that by now. (9 October 1989) Yesterday was a glorious, sun-drenched golden autumn afternoon, so 1 made my first trip out to the Elint Hills. Lost somewhere in the fading memory of mv fast-receding youth is some mysterious influence that kindled in me a special enchant- ment with the prairie. And somehow the Flint Hills came to rep- resent the essence of the prairie, much as the view across the Snake River towards the Grand Tetons is the essence of the Rocky Moun- tains. 1 had never been in the Elint Hills before yesterday, yet some- how the landscape was exactly what 1 expected the prairie to look like, with gently undulating hills stretching out to eternity, an iso- lated stand of trees here and there, and a few distant cattle that could easily have been bison. And at this time of year, with the grasses taking on hues of brass and bronze and copper, the only animal sounds carried on the wind were the soft, patient, ageless songs of grasshoppers and katydids. Lots and lots of grasshop- pers. I only saw one kind of bee, a tiny exomalopsine (I think) that was visiting the handful of ground-hugging, inconspicuous white asters that were still in bloom. The towering sunflowers were dark, withered stalks, and the spikelets of the great prairie grasses, big and little bluestem and Indian grass, were beginning their piece- meal disintegration and dispersal. The gentle wind tugged at my weary, overworked spirit, reviving almost forgotten visions from my childhood imagination, wild west visions of pioneers, fur trad- ers. Pony Express riders, plains Indians, and before all of them, that strange, marvelous Pleistocene megafauna. Something about a wide-open landscape sets my spirit soaring. So I can tell I'm going to like it here. There was a retirement celebration for Michener last spring that brought in alumni from all over the world, and the returning pilgrims all expressed amazement at how much Lawrence has grown. Nevertheless, it still isn't a large city bv any means. A little larger than Ithaca, 1 think, but still basically a small, friendly town. When my parents were growing up in El Paso, it was about the size that Lawrence is now. It seems hard to believe doesn't it? But you've witnessed much of El Paso's cancerous expansion, so I guess you have to believe it. 1 hope a similar fate is not in store for Lawrence, but asne of the very most recent immigrants, I suppose I am in no position to complain about all the new people coming in (although 1 could point out that the last time they hired a bee systematist for the faculty here was 50 years ago). So I don't feel that bee systematists are a major threat to Lawrence's future, but perhaps that is just a self-serving post-hoc rationalization. It's much too late in the day for me to seriously ponder such profound moral issues. I need some sleep. In a 1995 Christmas letter to friends and family, Byron described some of his research: fossil-hunting in Montana with his student, Gabriel Melo, and wasp-watching near Lawrence, Kansas, with a visiting colleague, Josep Asis (Alexander and Asis, 1997): (22 December 1995) Eor me, the most enjoyable part of 1995 was the summer. The university only pays me for 9 months of the year, so 1 ha\'e more leeway about how I spend my summers, although it is expected that I will spend the time doing research. Because 1 actually enjoy doing research, I find this arrangement to be satis- Eig. 2. Ord Kangaroo Rat, Great Sand Dunes National Monu- ment, late 1978. Erom a notecard, part of a series Byron had printed from some of his drawings. factory. My summer research activities began with a quest for fos- sil insects in southern Montana. This was exciting, because at present there are (as far as I know) no insect fossils of Jurassic ageknown from North America. (T. D. A. Cockerell described one Ju- rassic fossil from New Mexico, but he was so unsure about its identity that he only guessed that it might be an immature insect, and he didn't even attempt to assign it to an order. In 1968, some- body presented a talk at the Geological Society of America and reported an assemblage of aquatic bugs in Jurassic deposits in New Mexico, but apparently no descriptions of these fossils were ever published.) The leader of our expedition was a vertebrate paleon- tologist whose real reason for exploring these deposits was the hope of finding some bird fossils. He has observed that sites with lots of insect fossils are the kinds of places where bird fossils are eventually found if you spend enough time looking. The two en- tomologists on the expedition (me and a graduate student of mine from Brazil) were hoping to find interesting Hymenoptera, be- cause the Jurassic is a time when we suspect that many important evolutionary changes were happening in the Hymenoptera. We ended up bringing home several thousand fossils, but none of them were Hymenoptera, and the overall di\-ersity was very low. The fossils were in a marine siltstone, and most of them seemed to have been transported for some distance before settling in the sedi- ments where they became fossils. Very few specimens had any legs or antennae, and the most common fossil was the ventral sur- face of an abdomen. We also found an occasional elytron. Our vertebrate paleontologist tells us that the matrix can' be chemi- cally dissolved away to reveal more detail in the fossils; so all that is needed now is the time to prepare a few thousand specimens and examine them in detail. I was planning to spend June and early July working with a Spanish colleague on a study of a species of wasp named Gk'iwstktin pictifroiif, which was nesting at a site about 11 miles South of Lawrence a couple of years ago. However, we had a very rainy spring in Northeast Kansas this year, and Gleiiostictin sim- ply never showed up. So we ended up studying a different spe- cies of solitary wasp, in the genus Ccrceris, which we selected for the simple reason that it was there. It was the most boring wasp I have ever watched. I don't mean for this to be taken as a criticism of the wasp, because I suspect that it actually does things that are at least as interesting as what any other wasp does. However, it does not do these things where one can easily observe them. If you sit among a group of nests (which are burrows in the ground), once every 30 minutes or so you will see a wasp leave her nest, or return to her nest carrying a paralyzed beetle. The wasp leaves the nest entrance open while she is away, which means that she can very quickly enter the nest when she returns with her prey So mostly you sit in the sun baking your brains, and this goes on all day long. Although a given individual wasp might be active for The University of Kansas Natural History Museum Special Publication No. 24 something like 4 to 6 hours a day, within the group of nests there are early risers and late risers, so that some wasp within the ag- gregation is likely to be active from around sunrise until a half hour or so after sunset. Despite the very slow pace of activity, we did learn some in- teresting things about these wasps; and of course there were other things to watch through the day as well. A couple of abandoned dogs befriended us, although thev eventually decided we were reallv boring companions. At the start of the field season our study site was a wheat field; but late one evening the field was harvested, the next day it was tilled, and the next day it was planted in soy- beans. I was quite impressed that wasps who had placed their nests in the wheat field were able to carry on with provisioning in spite of the Armageddon-like alterations of the world around them. Finally, a note on his Christmas card to Art Metcalf : (December, 1995) I realized how much my world view has come to be dominated by mellicentrism when 1 caught myself wonder- ing who pollinated the flowers that produced the berries that formed the centerpiece of your Christmas card. This is not to say that 1 have lost my appreciation for the berries themselves, as ob- jects pleasing to both the eye and the tongue. But I do wonder if the blossoms were visited just by honey bees, or by some spring Andrenas, or bumble bees, or halictids, or perhaps even flies or butterflies. Part of a mature understanding of the ways of bees is the realization that many of them are opportunistic thieves who simplv consume pollen and /or nectar without providing any ef- fective transfer of pollen. On the other hand, there are flowers (especiallv orchids) that exploit bees in equally opportunistic and one-sided ways. C'est la vie. Letters edited by Robert H. Hagen A TEACHER, MENTOR, AND FRIEND REMEMBERED Byron was an outstanding mentor and teacher who had a tremendous effect on the students with whom he worked. He was an exceptionally clear communicator, a master at relaying a great deal of information quickly and painlessly. Although not a dynamic speaker, he drew stu- dents in with the excitement that clearly boiled within his mind. His sense of wonder at the world around him stimu- lated and fostered curiosity in his students. He had the ability to present complex information in the most under- standable way, showing his classes how details fit into the larger framework. He seemed to sense the best pathway of explanation for students regardless of their prior edu- cation or experience. Byron had a large personal library, and he read extensively and widely. This gave him a great breadth and depth of knowledge from which to draw for his lectures. The subjects he taught were flawlessly woven into the greater fabric of science. Nurturing the minds of students was clearly Byron's passion and forte. Byron also taught in environments that required a less direct teaching style. He taught when he was an audience member in seminars, at informal discussions, and oral ex- ams. He formulated questions designed to clarify a par- ticular point for the speaker, or to lead the speaker down a pathway of thought she or he may have needed to con- sider. Byron had an impressive ability to use well-worded Fig. 3. Oak, Cornell University campus (see text). questions to allow the person to "discover" the point he thought needed to be addressed. Byron was an artist of unusual talent. He often said with a smile that he was an artist by temperament and a scientist by training. He expressed himself artistically in pencil, charcoal, pen and ink, and acrylic paint. Often us- ing what was at hand to render an image of what appealed to him, he once used a ballpoint pen to create a beautiful drawing of a oak tree on the Cornell campus (Fig 3). His subjects ranged from various animals to the interior of his student apartment to paintings of Texas landscapes that hang on the walls of his parents' home. His art was some- times three-dimensional. He constructed models of insect mouthparts or legs from paper for use in the classroom. If a student had difficulty finding a particular structure by use of the microscope, Byron would quickly sketch the object being magnified and then label it in his neat letter- ing. Those who knew Byron professionally have many fond memories of him. One of the great joys of being around Byron was experiencing his quick and bright wit. He en- livened any meeting, seminar, or social gathering he at- tended with his humor, making everyone within hearing distance laugh. Because he was so soft-spoken, if one was more than a few feet away from him, it was difficult to hear these good-natured witticisms. It was always a treat to sit close to him. In addition to conveying information, Byron was equally adept at another important component of teach- ing: he was an excellent mentor. He spoke of his students with pride. He assured them in word and deed that he had great confidence in their abilities. When conversing with students, he had a knack for seeing gaps in their knowledge, and skillfully pointed students in the direc- Byron A. Alexander tion they needed to go, giving boosts when needed. He was generous to a fauH with his time. Byron never hesi- tated to drop what he was doing to give his undivided attention to students. He never brushed off a student; his answers were always thoughtful and complete, usually leading to extended conversation about the topic at hand His mentor, the late George Eickwort, commented that Byron was unique in his ability to communicate. Byron could talk to anyone with ease, no matter what the person's age, position in life, education, or demeanor. He never spoke down (or up) to anyone. All who knew Byron knew him to be exceptionally humble and self-deprecating. Bright and talented as Byron was, he never made a com- ment that would behttle someone, though he surely could have. When he thought highly of someone, he would de- scribe him or her as a "fine person." Tliis being the scale that had meaning for him, I would have to say that Byron was the finest person 1 have ever known. Most important to me, Byron has influenced pro- foundly the way 1 interact with students. 1 always have been conscious of Byron's wisdom and influence, but fol- lowing his death 1 became aware of a cornucopia of les- sons Byron had taught me indirectly. He was (and contin- ues to be for me) an amazing teacher and mentor. Those of us who knew and admired him sincerely hope we can suc- cessfully pass to others the gift he gave so freely to us. It was truly a joy to be his student. Clare T. Wuellner Scientific Publications of Byron A. Alexander Kramer, H. D., B. A. Alexander, C. Clark, C. Busse and D. C. Riss. 1977. Empirical choice of sampling procedure for optimal research de- sign in the longitudinal study of primate behavior Piinmtcs 18; 825- 835. Alexander, B. A. 1985. Predator-prey interactions between the digger wasp Clypeadoti kfkijjctiis and the harvester ant Pogonomyrinex occiilentalis. Jounujlof Natural Histivy 19: m9-U54. Alexander, B. A. 1986. Alternative methods of nest provisioning in the digger wasp Clypeadon laticinctus (Hymenoptera: Sphecidae). jour- nal of the Kansas Entomological Society 59: 59-63. Henderson, C, J. Steiner and B. Alexander. 1986. Varroa jacobsoni life cycle. American Bee journal 126: 117, 119. Alexander, B. A. and J. G. Rozen, Jr 1987. Ovaries, ovarioles, and oocytes of parasitic bees. Pan-Pacific Entomologist 63: 155-164. Alexander, B. A. 1987. Eusociahty and parasitism in nest-provisioning insects. Page 387 in: Eder, J. and H. Rembold (eds.). Chemistry and Biology of Social Insects. Verlag J. Peperny, Munich. Kukuk, P. E, G. C. Eickwort, M. Raveret-Richter, B. Alexander, R. Gibson, R. A. Morse and F. Ratnieks. 1989. Importance of the sting in the evolution of sociality in the Hvmenoptera. Annals of the Entomologi- cal Society of America 82: 1-5. Alexander, B. A. 1990. A cladistic analysis of the nomadine bees (Hy- menoptera: Apoidea). Si/stematic Entomology 15: 121-152. Alexander, B. A. 1990. A preliminary phylogenetic analysis of sphecid wasps and bees. Sphecos 20: 7-16. Alexander, B. A. 1991. A phylogenetic analysis of honey bees (Hy- menoptera: Apidae: Apis). Pages 120-121 in: Veeresh, G. K., B. Mallik and C. A. Viraktamath (eds.). Social Insects and the Environment. Proceedings of the 11th lUSSI International Congress, Bangalore, India. Oxford and IBH Publ. Co., Pvt., Ltd. Alexander, B. A. 1991. Nomada phytogeny reconsidered (Hymenoptera: Anthophoridae). journal of Natural History 25: 315-330. Alexander, B. A. 1991. Phylogenetic analysis of the genus Apis (Hy- menoptera: Apidae). Annals of the Entomological Society of America 84:137-149. Alexander, B. A. 1991. A cladistic analysis of the genus Apis. Pages 1-28 in: Smith, D. R. R. (ed.). Diversity in the genus Apis. Westview Press, Boulder. Alexander, B. A. 1992.A cladistic analysis of the subfamily Philanthinae (Hymenoptera: Sphecidae). Systematic Entomology 17: 91-108. Alexander, B. A. 1992. An exploratory analysis of cladistic relationships within the superfamily Apoidea, with special reference to sphecid wasps (Hvmenoptera). fournal of Hymenoptera Research 1: 25-61. Alexander, B. A., D. Yanega and R. L. Minckley. 1993. Nesting biology of Glenostictia pictipvns (Smith) (Hymenoptera: Sphecidae: Bembicini). journal of the Kansas Entomological Society 66: 108-120. Kazenas, V. L. and B. A. Alexander. 1993. The nest, prey and larva of E}itomosericus kaufmani Radoszkowski (Hymenoptera: Sphecidae). journal of Hymenoptera Research 2: 221-225. Alexander, B. A. 1994. Species-groups and cladistic analysis of the cleptoparasitic bee genus Nomada (Hymenoptera: Apoidea). Uni- versity of Kansas Science Bulletin 55: 175-238. Alexander, B. A. and M. Schwarz. 1994. World catalog of the cleptoparasitic bee genus Nomada (Hymenoptera: Apoidea). Uni- versity of Kansas Science Bulletin 55: 239-268. Alexander, B. A. 1995. Description of the female of Nomada dreisbachorum Moalif (Hymenoptera: Apoidea: Nomadinae). Pan-Pacific Entomolo- gist 71: 130-132. Alexander, B. A. and C. D. Michener. 1995. Phylogenetic studies of the families of short-tongued bees (Hymenoptera: Apoidea). University of Kansas Science Bulletin 55: 377-424. Alexander, B. A. 1996. Comparative morphology of the female reproduc- tive system of nomadine bees. Memoirs of the Entomological Society of Washhigton, no. 17, pp. 14-35. Alexander, B. A. and J. D. Asis. 1997. Patterns of nest occupancy and provisioning in Cerceris rufopicta Smith (Hvmenoptera: Sphecidae). lournat of Insect Behavior 10: 871-893. Rozen, J. G., Jr., A. Roig-Alsina and B. A. Alexander. 1997. The cleptoparasitic bee genus Rhopalolenima, with reference to other Nomadinae (Apidae), and biology of its host Protodufourea (Halictidae: Rophitinae). American Museum Novitates No. 3194, pp. 1-28. Alexander, B. A., C. D. Michener, and A. L. Gardner 1998. Dasypodidae Borner, 1919 (Insecta, Hymenoptera): proposed emendation of spell- ing to Dasypodaidae, so removing the homonymy with Dasypodidae Gray, 1821 (Mammalia, Xenarthra). Case 3023. Bulle- tin of Zoological Nomenclature 55: 24-28. Cutler, B. and B. A. Alexander 1999. The ultrastructure of the wall and lining epithelium of glandular pouches in nomadine bees (Hy- menoptera: Apidae: Nomadinae). Special Publication No. 24, Natu- ral History Museum, The University of Kansas (this volume) Biiers, G. W., R. H. Hagen, ami R. W. Brooks (eds.). Entomological Contributions in Memory of Byron A. Alexander. Liniverfilii ofKnmns Natural Histon/ Museum Speciai Publication 24. (1999) Pp. 9-15 The Augochlorine Bee Genus Megaloptilla (Hymenoptera: Halictidae) By Michael S. Engel^ and Robert W. Brooks^ ABSTRACT The augochlorine genus-group Megnloptilln is given generic rank outside oi Megommntion and its position in augochlorine phylogeny briefly discussed. The group is diagnosed and the male is described for the first time. Two species are described, figured, and keyed for the genus; the type species Megnloptilln cnllopis (Vachal) new combination, and M. hyronclla new species. Megnloptilln cnllopis is found in Colombia and Peru, whereas M. bi/rouelln is known only from Panama. Keywords: Apoidea; Augochlorini; Neotropics; Taxonomy. INTRODUCTION Bees of the halictine tribe Augochlorini range from southern Canada to Argentina and Chile. At present, the tribe includes approximately 500 species classified into 38 extant genera and subgenera, along with three species in the fossil genus Oligochlorn. Although their range covers a greater portion of the western hemisphere, augochlorines are most diverse in South America. The augochlorine genus Megnloptilln was originally proposed by Moure and Hurd (1987) as a subgenus of Megommntion, a member of a clade of genera with a greatly narrowed labiomaxillary complex and a distinctly pointed galeal apex. Examination of the female lectotype desig- nated by these same authors for the type species, H. cnllopis Vachal, reveals an unmodified labiomaxillary complex, a broadly rounded galeal apex, and little affinity to Megommntion or any of its relatives. The male was un- known at the time of Moure and Hurd's work, but has recently been identified along with a second species of the genus. Megnloptilln is similar in general appearance to spe- cies of Cnennugochlorn, or even more strongly to species of Pnroxystoglossa, from both of which it is differentiated be- low. In a cladistic analysis of fhe augochlorine genera (Engel, 1998) Megnloptilln is part of a large clade contain- ing genera such as Augochlorn, Augochlorelln, Cerntalictus, Pnwxystoglossn, and Pereirnpis. This group is part of a larger clade containing genera such as Cnenntigochlom, Megnlopta, Thectoclilorn, and Chlewgella among others. These taxa are grouped on the combination of a disHnctly narrowed spicu- lum on S8 of males, an acute apex to the marginal cell (al- though this is reversed in 2 genera), poorly developed teeth on the labral margins (ultimately modified in Augochlora and Pereirnpis), a serrate inner hiiid tibial spur, and an en- tirely rimmed basitibial plate (reversed in 1 genus). Megnloptilln and Pnroxi/stoglossn fall outside of the eusocial group (composed of Augochlorelln, Pereirnpis, Cerntnlictus, and Augochlora) by lacking a carinate preoccipital ridge. Megnloptilln monophyly is supported by the uiimodified venter of the penis valve and the bilobed process on the apical margin of the male S7; while the species of Pnroxystoglossa are united by the obtuse epistomal sulcus, the truncated marginal cell apex, and the high projecting and strongly narrowed anterior mesoscutal border. The cladistic position of Megnloptilln among augochlorine gen- era will be treated in further detail in a forthcoming paper concerning the entire tribe (Engel, 1998). Acknowledgments MSE was supported by a National Science Founda- tion Predoctoral Fellowship and an Ernst Mayr Award from the Museum of Comparative Zoology, Harvard Univer- sity. MSE is sincerely thankful to Mme. ]. Casevitz- Weulersse (Museum National d'Histoire Naturelle, Paris), G. Else, and S. Lewis (both of the British Museum, Natu- ral History) for kindly hosting him during his stay at their respective institutions. We are indebted to G.W. Byers, B.N. Danforth, and CD. Michener for constructive criticisms on the manuscript which greatly improved the final ver- sion. This work is lovingly dedicated to the memory of Byron A. Alexander (1952-1996). Byron was a dear friend whose intellect was only surpassed by his kindness. This paper is contribution No. 3193 of the Snow Entomological Collection, University of Kansas Natural History Museum. MATERIAL AND METHODS Specimens of Megnloptilln were found in the following institutions and made available to us by the named indi- viduals: the Natural History Museum (British Museum), London, G. Else and S. Lewis (BMNH); Cornell Univer- sity Insect Collection, Ithaca, New York, J.K. Liebherr and ' Department of Entomology, Comstock Hall, Cornell University, Ithaca, New York 14853. Present address: Department of Entomology American Museum of Natural History, Central Park West at 79th St., New York, NY 10024. E-mail: engel@amnh.org - Snow Entomological Collection, Division of Entomology, Natural History Museum, University of Kansas, Lawrence, Kansas 66045. 10 The University of Kansas Natural History Museum Special Publication No. 24 R.E. Hoebeke (CUIC); Museum National d'Histoire Naturelle, Paris, J. Casevitz-Weulersse (MNHN); National Museum of Natural History, Smithsonian Institution, Washington, D.C., R.J. McGinley (USNM); Snow Entomo- logical Collection, Division of Entomology, Natural His- tory Museum, University of Kansas, Lawrence, Kansas (SEMC); and the Smithsonian Tropical Research Institute, Panama City, Panama, D. Roubik (STRI). In the descriptions the following abbreviations are used for morphological terms: F, flagellomere; S, sternum; T, tergum. All measurements were made using an ocular micrometer on a WILD-M5a microscope. Measures given in the descriptions are for the type specimens with ranges of variation for critical values given in parentheses. SYSTEMATICS Genus Megaloptilla Moure and Hurd, new rank Megommation {Megaloptilla) Moure and Hurd, 1987:241. Type species: Halidiis callopii Vachal, 1911, monobasic and original designation. Engel, 1998: 123 Megommatwn (Emgaloptilla) Moure and Hurd, 1987:vi. Lapsus calami for Megaloptilla Moure and Hurd, 1987. Diagnosis.—Members of Megaloptilla generally re- semble species of Caeiiaiigoclilora (s.str.) and Pawxi/stoglossa. It differs from the first genus by the more obtuse epistomal sulcus, transverse labral basal elevation, rounded preoc- cipital ridge, very slightly narrowed mesoscutum, and ser- rate inner hind tibial spur. From the latter it differs in the rounded preoccipital ridge, transverse labral basal eleva- tion, acute apex of the marginal cell, and very weakly nar- rowed mesoscutum (in Paroxystoglossa the mesoscutum is strongly narrowed, high, and projecting forward). Description.—The following description is based on the two presently included species: Female: Epistomal sulcus roughly orthogonal (Figs. 2, 11). Clypeus and supraclypeal area convex and slightly protuberant. Malar space short. Preoccipital ridge rounded. Inner margin of compound eye moderately emarginate; eyes moderately convergent below; eye hairs short. Ocelli of normal size, not greatly enlarged; interocellar furrow absent. Vertex normal, not greatly expanded posteriorly. Labral basal elevation transverse, protuberant; distal pro- cess narrowly triangular, teeth weak. Mandible strongly bidentate due to well-defined subapical tooth; supplemen- tary tooth on imier margin of mandible. Hypostomal ridge carinate, anterior angle rounded, not projecting posteri- orly beyond distal margin of head. Distal portion of max- illa normal; galeal apex lobed, inner strip with setae and cuticular markings, base of galea equal to base of stipes; galeal comb absent; maxillary palpus normal. Prementtim normal; salivary plate with V-shaped brace; segments 2+3 of labial palp longer than 1; glossa of moderate length, less than half length of prementum. Pronotal lateral angle ob- tuse, not produced; lateral ridge rounded; dorsal ridge strongly carinate, weakly lamellate in some areas. Mesoscutum slightly narrowed anteriorly, mesoscutal lip low and rounded. Tegula rounded. Propodeal triangle subequal to scutellum, longer than metanotum; propodeal dorsal ridge rounded; lateral ridge rounded, ridges slightly divergent; propodeum not narrowed posteriorly; pit of posterior face narrow. Marginal cell with acute apex (Fig. 1). Distal hamuli arranged 2-1-2 on hind wing margin. Anterior basitarsal brush present. Inner hind tibial spur serrate, serrations sharp (Fig. 14). Scopa formed of long, plumose hairs on hind femur and tibia (Figs. 4, 13). Basitibial plate narrowly rounded, well defined on all edges. Metasoma unmodified. Male: Mandible simple. Labrum without basal eleva- tion, without distal process. Scape of moderate length, reaching to lateral ocellus; F2 longer than Fl; antenna of moderate length, reaching back to scutellum. Metasoma oval, not elongated. Apical margins of S3-4 unmodified. Apical margin of S5 weakly emarginate; S6 more narrowly and deeply emarginate (Fig. 8). Male terminalia as in fig- ures 8-10. Identification.—The following key couplets are for Eickwort's (1969) keys to the genera and subgenera of Augochlorini. The male runs to Eickwort's couplet 22. This couplet should be entirely replaced by the version given below. Eickwort's version separates males of Con/niirella from Rhectomia. Coryniirella is a junior synonym of Rhectomia (Engel, 1995); therefore our modified couplet only separates Rhectomia from Megaloptilla. Couplet 13' is an additional couplet which should be inserted into his key to females. In the following couplets eye emargination is determined following Eickwort (1969: 377). A new key to the genera and subgenera of the Augochlorini is in preparation by one of us (Engel, 1998). Males 22 Mesoscutum weakly narrowed anteriorly; marginal cell apex acute; F2 longer than Fl, F2 as long as im- mediately following flagellomeres Megaloptilla — Mesoscutum broadly rounded anteriorly; marginal cell apex truncate; F2 as long as Fl, or if F2 slightly shorter than Fl, then F2 distinctly shorter than im- mediately following flagellomeres Rhectomia Females 13 Compound eyes moderately emarginate (w/1 < .11); propodeal triangle roughened, with striae or plicae; preoccipital ridge rounded or sharply angled 13' — Compound eyes deeply emarginate (w/1 > .125); propodeal triangle smooth or granular; preoccipital ridge rounded Corymira (s.str.) AUGOCHLORINE BeE GeNUS MeGALOPTILLA 11 13' Preoccipital ridge sharply angled; marginal cell apex truncate; mesoscutum strongly narrowed anteriorly, lip high and sharply angled; labral basal elevation orbiculate Panm/stglossa — Preoccipital ridge rounded; marginal cell apex acute; mesoscutum very slightly narrowed anteriorly, lip low and rounded; labral basal elevation transverse Mcgaloptilln Key to the Species of Megaloptilla 1 Pleura brown with metallic green reflections; hypoepimeron minutely punctured; Sc+R more strongly pigmented than other wing veins; propodeal dorsal ridge rounded; metasoma brown with red- copper reflections, integument smooth with minute punctures M. callopis — Pleura yellow-orange, without reflections; hypoepimeron smooth, without punctures; wing veins evenly pigmented, or at least Sc+R no more strongly pigmented than C; propodeal dorsal ridge with short medial carina; metasoma yellow-orange, integument imbricate M. In/wncUa Megaloptilla callopis (Vachal), new combination (Figs. 1-10, 15, 16) Halictus callopis Vachal, 1911:41. Oxyito^^Iossidin callopis (Vachal); Moure, 1944:69. Megoininalion {Megaloptilla) callopis (Vachal); Moure and Hurd, 1987:241, female lectotvpe designation. Lectotype.—PERU: Cuzco: Female (Figs. 15^16), Marcapata, Vachal collection (MNHN), examined by MSE. Additional material.—COLOMBIA: Ptituiiun/o: One female, Mocoa, 21 August 1978, M. Cooper, B.M. 1978-431 (BMNH). One female, Mocoa, c. 600 m, 31 May-7 June 1976, M. Cooper, B.M. 1976-354 (BMNH). One female, Mocoa, c. 600 m, 26 March-6 April 1976, M. Cooper, B.M. 1976-290 (BMNH). PERU: Cuzco: One female, Marcapata, Vachal collection (MNHN). Sim Martin: Two males, Rioja, 14 De- cember 1978, M. Cooper, B.M. 1979-20 (BMNH). Ucayali: One female, Boqueron [del Padre] Abad, 18 December 1961, J.M. Schunke, B.M. 1962-491 (BMNH). Diagnosis.—Mesoscutum brown with metallic green and copper reflections. Pleura brown with metallic green reflections. Hypoepimeron minutely punctured. Crossvein Ir-m distad Im-cu; Sc+R more strongly pigmented than other wing veins. Scopal hairs of hind tibia reaching to tibial apex (Fig. 4). Propodeal dorsal ridge rounded. Metasoma brown with red-copper reflections, integument smooth with minute punctures. Description.—The following description is based on the lectotype female and a male from the Natural History Museum, London: Female: Structure. Total body length 8.14 mm (7.9-8.54 mm); forewing 6.56 mm long (6.4-8.2 mm). Head longer than wide (length 2.18 mm, width 1.94 mm) (Fig. 2). Distal two thirds of clypeus projecting below lower tangent of compound eyes; supraclypeal area slightly shorter than clypeus. Frontal line carinate between antennae to half distance to median ocellus, becoming a weakly impressed line from that point on. Scape 0.92 mm long; pedicel about as long as wide, about as long as El; El as long as wide; F2 slightly longer than wide, longer than El. Distance from median ocellus to lateral ocellus 0.06 mm; between lateral ocelli 0.26 mm; lateral ocellus to compound eye 0.32 mm. Genal width roughly equal to that of compound eye in profile (Fig. 3). Intertegular distance 1.4 mm (1.4-1.7 mm). Scutellum 1.5 times longer than metanotum. Propodeal triangle almost as long as scutellum; dorsal ridge rounded. Basal vein distad cu-a by 3 times width of vein (Fig. 1); Ir- m distad Im-cu by width of vein; 2r-m distad 2m-cu by 2 times width of vein; 2r-m relatively straight. First submar- ginal cell longer than second and third submarginal cells combined; second only slightly narrowed anteriorly. Coloration and sculpturing. Mandible amber, except red-brown at apex and base. Labrum amber. Distal half of clypeus amber, remainder brown with a few metallic green- copper reflections; amber patch smooth, remainder of clypeus with sparse weak punctures separated by 1-3 puncture widths, integument between pimctures smooth (Fig. 2). Supraclypeal area brown with reflections as on basal half of clypeus, integument as on clypeus. Head brown with strong metallic green and weak copper reflec- tions; face densely punctured, punctures separated by less than their width, integument between punctures smooth. Scape amber, pedicel and flagellum light brown. Vertex with minute punctures separated by twice their width, in- tegument otherwise smooth. Gena and postgena weakly imbricate; postgena without metallic green reflections. Pronotum brown, amber at lobe and ventrally on lateral surfaces, a few weak metallic green reflections dorsally, integument smooth. Mesoscutum colored as face; minutely punctured, punctures separated by 2-5 puncture widths, integument otherwise smooth; median and parapsidal lines weakly impressed. Tegula amber and semi-translu- cent. Scutellum colored as mesoscutum, except punctures deeper and separated by 1-3 times puncture width. Metanotum as scutellum, copper reflections strong, minute punctures separated by 2-4 times puncture width, integu- ment weakly roughened. Pre-episternum with strong me- tallic green reflections; small punctures separated by a puncture width or less, integument otherwise smooth. Mesepisternum colored as pre-episternum; small punc- tures separated by a puncture width or less, except ven- trally punctures becoming more widely scattered. 12 The University of Kansas Natural History Museum Special Publication No. 24 S8 Figs. 1-10. Megalnptilla ailhfiif (Vachal). 1-Forewing venation. 2-Front view of female face. 3-Lateral \-iew of female head. 4-Hind leg. 5-Mesotibia. 6-Front view of male face. 7-Lateral view of male head. 8-Male sterna 5 and 6. 9-Male sterna 7 and 8. 10-Male genitalia; left half is dorsal, right half is ventral. AUGOCHLORINE BeE GeNUS MeGALOPTILLA 13 Hypoepimeron minutely piinctured, punctures separated by 1-3 times puncture width, integument otherwise smooth, colored as on remainder of mesepisternum. Metepisternum colored as on mesepisternum; punctured as on hypoepimeron, except punctures more widely scat- tered and weaker ventrally. Legs amber. Wing veins am- ber, except Sc+R brown. Lateral and posterior propodeal surfaces imbricate. Propodeal triangle brown with strong copper and weak green reflections; surface strongly im- bricate with reticulating rugae. Metasomal terga brown with strong red-copper reflections; integument smooth with minute punctures separated by 2-A times puncture width. Sterna brown and weakly imbricate. Pubescence. Pubescence pale to golden. Clypeus, supraclypeal area, and face with sparse simple hairs; face with additional short, suberect, plumose hairs nearly ob- scuring integument. Vertex and gena with scattered short, simple hairs. Postgena with widely scattered long, simple hairs; hairs on border with gena with a few short branches. Lateral surface or pronotvim with extremely short, fine, simple appressed hairs, not obscuring the surface. Mesoscutum with scattered, simple hairs. Scutellum as on mesoscutum, with additional longer, simple hairs, a few with short branches. Pre-episternal sulcus covered with short, suberect, plumose hairs. Lateral and posterior propodeal surfaces with scattered, long hairs each with a few branches; additional short, suberect, simple hairs, not obscuring surface. Tl-2 with sparse, simple hairs. T3-4 with hairs becoming more dense and longer. Sterna with scattered long, simple hairs. Male: As in the female with the following modifica- tions. Total body length 9.36 mm (9.3-9.36 mm); forewing length 7.04 mm (6.94-7.04 mm). Head longer than wide (length 2.24 mm, width 2.02 mm) (Fig. 6). Distal half of clypeus projecting below lower tangent of compound eyes (Fig. 7). Scape 0.64 mm long; pedicel as long as wide, longer than Fl; Fl wider than long, about half the length of F2; F2 longer than wide. Distance from median ocellus to lateral ocellus 0.08 mm; between lateral ocelli 0.26 mm; lateral ocellus to compound eye 0.24 mm. Intertegular distance 1.42 mm (1.40-1.42 mm). Male terminalia as in figures 8- 10. Coloration, sculpturing, and pubescence as in the fe- male, except for usual sex differences in pilosity. Remarks.—One male from the Natural History Mu- seum, London, has several mites clinging to the anterior surface of the first metasomal tergum. Megaloptilla byronella new species (Figs. 11-14) Holotype.—PANAMA: Darieii: Cana Biological Sta- tion, Serrania de Pirre, 1250 m, 7°45'18"N, 77°41'6"W, Figs. 11-14. Megaloplilln bynviclln n. sp. U-Front view of fe- male face. 12-Lateral view of female head. 13-Hind leg. 14-Iraier hind tibial spur. 4 June 1996, R.W. Brooks and J.S. Ashe, female, PANlAB96- 106, ex: flight intercept trap (SEMC). Paratypes.—PANAMA: Panama: Capira, Cerro Campana, 18 August 1982, D. Roubik, five females, com- iiig to cineole bait; 6 females, same except 8 September 1982, No. 41 (SEMC, CUIC, USNM, STRl). Diagnosis.—Mesoscutum black. Pleura yellow-or- ange, without metallic reflections. Hypoepimeron smooth and impunctate. Crossvein Ir-m confluent with Im-cu; Sc+R as pigmented as in C. Scopal hairs on hind tibia not reaching tibial apex (Fig. 13). Propodeal dorsal ridge with short medial carina. Metasoma yellow-orange, integument imbricate. 14 The University of Kansas Natural History Museum Special Publication No. 24 Figs. 15-16. Mcgaloptithi catlopis (Vachal), lectotype female. 15-Dorsal habitus. 16-Front view of face. Description.—The following description is based on the holotype female: Female: Structure. Total body length 10.64 mm (7.7- 13.8 mm); forewLng 8.48 mm long (6.6-8.5 mm). Head as long as wide (length, width 2.52 mm) (Fig. 11). Clypeus as long as broad, distal half projecting below lower tangent of compound eyes; supraclypeal area longer than wide. Frontal line carinate between antennae, becoming a mod- erately impressed line just above antennal sockets. Scape 1.18 mm long; pedicel about as long as Fl; Fl about as long as wide; F2 longer and wider than Fl, about as long as wide. Distance from median ocellus to lateral ocellus 0.06 mm; between lateral ocelli 0.32 mm; lateral ocellus to compound eye 0.28 mm. Genal width roughly equal to that of compound eye in profile (Fig. 12). Premental length 1.32 mm; width 0.32 mm. Glossal length 0.54 mm. Intertegular distance 1.92 mm (1.4-2.0 mm). Scutellum almost twice as long as metanotum. Propodeal triangle slightly shorter than scutellum; dorsal ridge weakly carinate medially, oth- erwise rounded. Basal vein distad cu-a by 2 times width of vein; Ir-m confluent with Im-cu; 2r-m distad 2m-cu by 4.5 times width of vein. First submarginal cell only slightly longer than second and third combined; second narrowed anteriorly. Coloration and sculpturing. Mandible yellow-orange, red-brown at apex. Clypeus yellow-orange, shallow punc- tures generally separated by more than a puncture width, integument between punctures smooth. Supraclypeal area dark brown, sometimes orange basally; sculptured as clypeus except punctures finer. Scape yellow, remainder AUGOCHLORINE BeE GeNUS MeGALOPTILLA 15 of antenna brown. Face dark brown with dull metallic green reflections; closely punctate, punctures minute and separated by width of puncture or less, integument be- tween smooth (Fig. 11). Vertex, gena and postgena dark brown; vertex and gena punctured as face, postgena im- bricate and impunctate. Prothorax yellow-orange, surface imbricate. Mesoscutum black with scattered minute punc- tures, integument smooth. Tegula light brown, weakly imbricate. Scutellum yellow-orange; sculptured as mesoscutum except punctures more dense. Metanotum brown and dull. Pleura largely yellow-orange with sur- face finely imbricate, becoming granular ventrally; hypoepimeron entirely smooth. Wings slightly yellowed; veins evenly pigmented and amber. Legs entirely yellow- orange. Propodeal triangle red-brown or brown, remain- der of area yellow-orange; triangle rugose, more finely so at center, rugosity not extending to margins of basal area, surface imbricate. Lateral surface of propodeum yellow- orange, posterior surface darkened; surface finely imbri- cate or granular. Metasoma yellow-orange, finely imbri- cate, with widely scattered weak punctures, except ante- rior surface of Tl smooth; dense, minute ptmctures on disks of Tl-2. Pubescence. Clypeus and supraclypeal area with scat- tered long, simple hairs. Face, vertex, and gena with short, golden, appresseci hairs, not obscuring surface; becoming plumose laterally on face, and below on gena (Fig. 11), in- termixed with scattered simple hairs. Postgena with long hairs sparse except row along hypostomal carina. Pronotal dorsal surface and pronotal lobes covered with dense pale hairs intermixed with scattered erect hairs. Mesoscutum with short, subappressed hairs along margins; slightly longer, erect hairs scattered across surface. Tegula with hairs on inner and anterior margins. Posterior half of scutel- lum with very long, orange hairs. Metanotum with long, plumose, yellow hairs and shorter, suberect, plumose hairs. Pleura with simple yellow hairs, a few with short branches; hairs sparse on hypoepimeron, hairs becoming longer ven- trally. Legs with golden pubescence. Lateral and posterior surfaces of propodeum with long, yellow hairs and layer of very short, simple hairs obscuring the surface. Anterior surface of Tl without pubescence; disks of Tl-2 with minute hairs intermixed with scattered, short, suberect, yellow hairs becoming progressively longer and more widely scattered on T3-4. T5 with similar hairs of moder- ate length and more numerous than on preceding terga. Sterna with long, erect, yellow hairs along apical margins. Male: Unknown. Etymology.—This species is named in honor of the late Byron A. Alexander. LITERATURE CITED Eickwort, G. C. 1969. A comparative morphological study and generic revision of the augochlorine bees (Hymenoptera: Halictidae). Uni- versity of Kansas Science Bulletin 48:325-524. Engel, M. S. 1995. The bee genus Rlwctomia (Hymenoptera: Halictidae): discovery of the male and two new species. Journal of the New York Entomological Society 103:302-310. Engel, M. S. 1998. Phytogeny, classification, and evolutionary ethology of the bee tribe Augochlorini (Hymenoptera: Halictidae). Ph. D. Dissertation, Cornell University, Ithaca, New York, \xii+306 pp. Moure, J. S. 1944. Abejas del Perii. Boletin del Museo de Historia Natu- ral, "Javier Prado" 8:67-75. Moure, J. S. and P. D. Hurd, Jr. 1987. An annotated catalog of halictid bees of the Western Hemisphere (Hymenoptera: Halictidae). Smithsonian Institution Press, Washington, D.C., vii -i- 405 pp. Vachal, J. 1911. Etude sur les Halictus d'Amerique (Hvm.). Miscellanea Entomologica 19:9-24, 41-56, 107-112. Bi/crs, G. W., R. H. Hagcit, and R. W. Brooks (cds.). Entomological Contributions in Memory of Byron A. Alexander. iiuiverfitii ofKnnsns Natural Hishvy Museum Special Publication 24. (J999) Pp. 17-20 Redescription of Linsleyonides Skiles (Coleoptera: Cerambycidae) and Inclusion of Elaphidion portoricensis Fisher By Steven W. Lingafelter^ ABSTRACT Elaphidion portoricaisis Fisher is transferred to Linslei/Piiides based on four hypothesized autapomorphies for the genus. Linsleyotiides is redescribed and diagnostic characters are illustrated. Differences between Liiislei/oiiides and the closely related Elaphidion are discussed. A key for the three species of Linsleyonides is presented. Kei/ words: Systematics, Taxonomy, West Indies, Longhorned Woodborers. INTRODUCTION Linsleyonides is a member of the tribe Elaphidiini. This West Indian genus was proposed by Skiles (1985) to ac- commodate 2 species: L. albomacidntus (Champlaiii & Knull, 1922) and L. chemsaki (Skiles, 1985). Elaphidion portoricensis Fisher is transferred to Linsleyonides because it possesses four character states shared by the other two species of Linsleyonides (hypothesized autapomorphies of the genus) which are not known from other elaphidiine taxa. Addi- tionally, L. portoricensis lacks the hypothesized synapomorphies of Elaphidion and other potentially closely related genera. Linsleyonides is redescribed below, and many of the diagnostic morphological features are illus- trated. Parenthetical notations are included to indicate hypothesized autapomorphies and deviations from or agreement with Skiles' 1985 description. Acknowledgments I thank Michael Ivie for bringing this misplaced taxon to my attention and providing me with many specimens for character analysis. His generosity in accommodating me on a research trip to Montana State University to study this and other West Indian elaphidiine taxa is very much appreciated. The habitus illustration was nicely rendered by Elizabeth Roberts. I thank Steve Ashe and Byron Alexander for my training and their support of my stud- ies on Cerambycidae while at the University of Kansas. I thank David Furth, Darlene Judd, Alexander Konstantinov, Allen Norrbom, Norman Woodley, and two anonymous reviewers for constructive comments on this manuscript. SYSTEMATICS Linsleyonides Skiles Linsleyonides Skiles, 1985: 316. Type species, Elaphidion alkvnacnlatuni Champlain & Knull, by original designation. Description. —Size: small to moderate (7-20 mm). Head: eye large and coarsely faceted, occupying more than 50% of the exposed region of the head when viewed laterally; distinct, rounded or triangular patches of dense, white or yellow, supraocular pubescence present (Figs. 1-3) (Autapomorphy); frontoclypeal margin arcuate with lat- eral pits present (first mentioned by Skiles, 1985) (Fig. 5) (Autapomorphy); mandible with a narrow incisor region (less than one-third width of base of mandible when viewed from mesal or biting surface) and an apical and basal indentation separated by an undentate plateau; ter- minal labial palpomere without digitiform sensillum; ter- minal maxillary palpomere expanded apically, apical width over half length with distinct, narrow digitiform sensillum (Fig. 6); antenna of female strongly spined apicomesally on antennomeres 3-6, weakly so on antennomere 7; antenna of male, strongly spined mesally on antennomeres 3-5, weakly so on antennomeres 6-7; an- tenna not spined laterally; antemiomeres gradually wid- ened at apices, particularly after antennomere 6; antenna without carina (Skiles, 1985 indicates antennae are partially, dorsally carinate, but my clearing of a specimen did not reveal this); anteiinomere three about two-thirds length of pronotum in male, slightly longer than half length of pronotum in female; terminal antennomere without pseudo-segmental constriction or setae. Prothorax: raised median callus and peripheral calli present on pronotum; procoxal cavities open posteriorly; prosternal intercoxal process only slightly expanded apically, gradually declivous posteriorly; lateral margin of procoxal cavity closed (trochantin hidden, propleuron and prosternum fused very close to coxa); pronotum without lateral tubercle or transverse ridges on pronotal disc; prosternum more heavily pubescent in female than male. Mesothorax: mesocoxal cavity closed or barely open laterally (mesepimeron contacting mesocoxae directly in some Systematic Entomology Laboratory, PSl, ARS, USDA, c/o U.S. National Museum of Natural History, MRC-168, Washington, D.C. 20560. E-mail: slingafelter@sel.barc.usda.gov 17 18 The University oe Kansas Natural History Museum Special Publication No. 24 C??9b*ris Fig.l. Habitus of Linsleyonidcs portoricensis (Fisher), female. Figs. 2—1. Habiti and diagnostic characters of Linslei/onides. 2-Habitus of L. chcfiisaki Skiles, male. 3-Habitus of L. alhomacidatus (Champlam & Knull), male. 4-Epipleural tooth on L. portcricensis. specimens, in others, closure of metasternum and mesos- ternum prevents this contact); anterior margin of mesos- temum as in Fig. 9; intercoxal process of mesostemum with small, indistinct lateral projection into acetabular excava- tion in mesocoxa; wide, truncate notch in mesosternal intercoxal process; anterior margin of mesonotum broadly rounded (Fig. 8); mesoprescutum (scutellum) with basal constriction and small apical notch on otherwise rounded posterior margin (Fig. 8); mesepisternal carina evenly ar- cuate (Fig. 7). Metntliorax: metasternal notch acute; metasternal sulcus incomplete, only attaining anterior one- third of metasternum; metepisternum with longitudinal keel positioned equidistant from dorsal and ventral mar- gin, more heavily sclerotized ventral to keel; metepisternal notch at posterior margin narrow and reaching approxi- mately half way to keel. Legs: mesal and lateral mesofemoral apices dentiform to weakly spinose; mesal and lateral metafemoral apices spinose; metafemur linear to slightly enlarged at middle; metafemur finely punctate; meso and metatibia with very reduced carina proximally, not visible distally; metacoxa with pronounced ridges on anterior face. Wings: elytron with scattered spots of dense white or yellow pubescence (Figs. 1-3) (Autapomorphy); elytral humerus with small, distinct tooth (Fig. 4) (Autapomorphy); elytra with strong apicolateral spine and dentiform sutural angle; hind wing MP-CuA incomplete, not contacting MPl+2; hind wing without CuAl+2. Diagnostic characters.—The hypothesized autapo- morphies for Liiislei/onides include the distinct postocular patches of pubescence as well as the small, dense, pubes- cent patches on the elytra (Figs. 1-3); the elytral humerus with a small epipleural tooth (but also present in some Eburiini, Fig. 4); and the arcuate frontoclypeal suture with ReDESCRIPTION of LiNSLEYONlDES 19 Figs. 5-9. Diagnostic cliaracters of Linslcyonidcf. 5-Head sliowing frontoclypeal pits. 6-Maxillary palpus showing medi- ally positioned digitiform sensillum. 7-Mesepisternum (anterior to the left) showing arcuate carina. 8-Mesonotum and scutellum (anterior to top) showing constricted scutellar base. 9-Mesoster- num showing sclerotized pattern on anterior margin. lateral pits (first discussed by Skiles, 1985) (Fig. 5). Other diagnostic characters not widely distributed in Elaphidiini include the incomplete metasternal sulcus; terminal antennomere without subapical setae and without pseudoantennomere constriction; and the sclerotization pattern of the anterior margin of the mesostemum (Fig. 9). Distribution and Diversity of Liusleyonides.—This attractive genus occurs in extreme southeastern United States and the West Indies; particularly southern Florida and Cuba (L. albomaculatus), Virgin Islands and Puerto Rico (L. portoricensis), and Jamaica (L. chemsaki). Discussion.—Linsleyonides and Elaphidkvi share the ba- sic form of the sclerotization of the anterior margin of the mesostemum, but iii Linsleyonides there is a posterior me- dial projection (see arrow, Fig. 9). Additionally, the mesofemoral and antennal spines in Linsleyonides are not as prominent as in Llaphidion, and Linsleyonides lacks the abruptly declivous prosternal intercoxal process charac- teristic of Elnpliidion. A phylogenetic analysis of Elaphidiini (Lingafelter, 1998) using implied weights (PIWE, Goloboff, 1993) showed Linsleyonides to be closely related to several gen- era including Elaphidion Audinet-Serville, Curtomeriis Stephens, and Lburia Lepeletier & Audinet-Serville. An equal weighting phylogenetic analysis of the same taxa in that study (Lingafelter, 1998) using PAUP (Swofford, 1991) showed Linsleyonides to be a sister taxon to other Elaphidion exemplars. These analyses used an exemplar approach and included L. portoricensis. Because the type species of the genus, L. nlboinaeulatus, and some potentially closely re- lated West Indian genera were not available for dissection and inclusion in that study, further analyses are required for a robust hypothesis of relationships among these closely related genera. Species Catalog of Linsleyonides Linsleyonides alhoinneulntus (Champlain and Knull), 1922; 146. Originally described as Elapliidion, transferred to Elaphidionoides by Linsley (1963), then placed in Linsleyonides by Skiles (1985). Designated as type species of Linsleyonides by Skiles (1985: 316). Type locality; Miami, Florida. Type deposition; Field Museum of Natural His- tory (Chicago, Illinois); not examined. Linsleyonides chemsaki Skiles, 1985: 317. Type locality: Hardwar Gap, Jamaica. Type deposition: Canadian Na- tional Collection, Agriculture Canada (Ottawa, Ontario); examined. Linsleyonides portoricensis (Fisher), 1932; 33. New Com- bination, transferred from Elaphidion. Type locality: Coamo Springs, Puerto Rico. Type deposition: American Museum of Natural History (New York, New York); examiiied. Key to species of Linsleyonides Postocular pubescence in large, rounded, contiguous patches (Fig. 2); patches of pubescence on head, pronotum, and elytra yellow; pronotal disc with four round patches, anterior two smaller than posterior two L. chemsaki Skiles Postocular pubescence in small, triangular, typically non-contiguous patches (Figs.l, 3); patches of pubes- cence on head, pronotum, and elytra white; pronotal disc with four or six patches, anterior two rounded and larger than the others 2 Each elytron with at least seven distinct, rounded patches of pubescence of differing sizes (Fig. 3) L. albomaculatus (Champlain & Knull) Each elytron with three triangular or irregularly shaped patches of pubescence (positioned basally, antemedially, and at posterior one-third) (Fig.l) L. portoricensis (Fisher) 20 The University of Kansas Natural History Museum Special Publication No. 24 LITERATURE CITED Champlain, A. B. and J. N. Knull. 1922. New North American Coleoptera. Entomological News 33: 144-149. Fisher, W. S. 1932. New West Indian cerambycid beetles. Proceedings of the United States NaHonal Museum 80(22): 1-93. Goioboff, P. A. 1993. Pee-Wee, Version 2.1. Program and documentation. New York, NY. Lingafelter, S. W. 1998. The genera of Elaphidiini Thomson, 1864 (Co- leoptera: Cerambycidae). Memoirs of the Entomological Society of Washington 20: 1-118. Linsley, E. G. 1963. The Cerambycidae of North America. Part IV. Tax- onomy and classification of the subfamily Cerambycinae, tribes Elaphidionini through Rhinotragini. University of California Pub- lications in Entomology 21: 1-165. Skiles, D. D. 1985. New genera and species of elaphidionine Cerambycidae (Coleoptera)froni North America and the West Indies. The Coleopterisfs Bulletin 39: 305-320. Swofford, D. L. 1991. Phylogenetic Analysis Using Parsimony (PAUP), version 3.0. Illinois Natural History Sur\'ey, Champaign. 6i/i7s, G. W., R. H. Hagcn, and R. W. Brooks (eds.), Entomological Contributions in Memory of Byron A. Alexander. iimvcrsity of Kimsa^ Natural History Miisciiiu Special Publication 24. (1999) Pp. 21-24 Anthidium oblongattim (Illiger): an Old World Bee (Hymenoptera: Megachilidae) New to North America, and New North American Records for Another Adventive Species, A. manicatiim (L.) By E. Richard Hoebeke' and A. G. Wheeler, Jr.^ ABSTRACT The palearctic wool-carder bee, Aitthidiiim oblongntiiiii (Illiger), is reported from North America for the first time. During 1994-1997, specimens were collected at several localities in Mary- land, New Jersey, New York, and Pennsylvania. A description, diagnosis, and illustrations are given, and its native geographic range and bionomics are summarized. New eastern U.S. and Canadian records are given for the adventive A. itianicatiiui (L.). Keywords: Hymenoptera; Megachilidae; Atithidiiitu obkmgatiiiii; Anthidium mauicatum; North America; Distribution; Bionomics. INTRODUCTION In the autumn of 1983, Byron Alexander —then a rela- tively youthful Texan who had received a Master's degree from Colorado State University under the masterful eye of Howard E. Evans—entered a Ph.D. program in ento- mology at Cornell Umversity, working with the late George Eickwort and pursuing research in the behavior and sys- tematics of Nomada bees. While at Cornell, 1 had the privi- lege to know Byron on both a personal and professional level. We had the opportunity to travel together, along with several others from Cornell, to Puerto Rico in 1985 and to collect insects for the university collection. And, of course, many exchanges—some humorous and others more sedate and scholarly—took place between Byron and me over the nearly 6 years he spent "high above Cayuga's waters" on East Hill in Ithaca. Byron was a unique individual, blessed with a witty sense of humor, and there was a certain charm about him. The entomological community at large lost an extremely talented member with the untimely and unan- ticipated death of Byron Alexander. He was an outstand- ing teacher, possessed rare artistic ability depicted in his many publications, and was a devoted student of insect natural history, but especially bee systematics, biology, and behavior. Because of his special penchant for bees, I take great pleasure in dedicating this paper — reporting on an exotic bee newly immigrant in North America —to the memory of Byron Alexander. He will be missed, but not forgotten! (E. R. Hoebeke) Thirty years have passed since the first report of an immigrant bee in the genus Aiithidiiiiii becoming estab- lished in North America. Jaycox (1967) recorded the oc- currence in the United States of the Old World bee, A. manicatum (L.), based on specimens reared from wooden trap nests in central New York in 1963. Anthidium uiaiticatum is the most widely distributed Anthidium in the world, occurring throughout Europe, the Mediterranean region bordering north Africa, and west- ern Asia. It has been accidentally introduced into Brazil, Argentina, Uruguay, and the Canary Islands (Pasteels, 1969; Schrottky, 1901; Moure and Urban, 1964; Lieftinck, 1958). In North America, this attractive, honey bee-sized species has been recorded from central New York (an area approximately 7,200 km- in Tompkins, Chemung, and Ontario counties) (Jaycox, 1967; Pechuman, 1967; Severinghaus et al., 1981) and from Ontario, Canada (Uni- versity of Guelph, Guelph, and Freelton) (Smith, 1991). On 17 October 1995, while approaching the entrance to his former workplace (Bureau of Plant Industry, Penn- sylvania Dept. of Agriculture, Harrisburg, PA), AGW ob- served and collected an interesting bee patrolling flowers of Russian sage (Perovslan artemesioides Boiss.) in a small garden. The specimen, a female, was identified by ERH as Anthidium oblongattim (Illiger), a species unknown in North America. Since then, additional males and females have been collected in 1996 and 1997 at several localities in a three-state area (Maryland, New York, and Pennsylvania). From June 1994 to June 1996, specimens of A. oblongatutn were frequently collected from flowers of various weeds at two urban restoration sites and landfills—one each in New Jersey and New York—as part of a study of wild bee pollinators of the weeds of landflUs and other post-indus- trial waste lands (M. Yurlina, in lift.). The above-mentioned records are cited below in Geographic Distribution and mapped in Figure 6. ' Department of Entomology Cornell University Ithaca, NY 14853. E-mail: erh2@cornell.edu - Department of Entomology, Clemson University Clemson, SC 26934. E-mail: awhlr@clemson.edu 21 22 The University of Kansas Natural History Museum Special Publication No. 24 X, Fig. 1. Anthidium oblongnttan. Female, dorsal habitus. Scale line = 5 mm. The greatest diversity of Anthidium (subg. Anthidium) in the U.S. is found in the West, where 25 species are re- corded (Hurd, 1979). Only two species oi Anthidium (subg. Anthidium) are known to occur in eastern North America— A. maculifrons Smith and A. psoraleae Robertson (Mitchell, 1962); however, neither of these species occurs in the North- east. Before the discovery of A. manicatum in New York, no Anthidium species were known from New England or the middle Atlantic states. Acknowledgments We thank Michael S. Engel (Cornell University) and Terry L. Griswold (Utah State University, Logan, UT) for their critical review of the manuscript. Mary E. YurlLna (Graduate Program in Ecology and Evolution, Cook Col- lege, Rutgers University, New Brunswick, NJ) kindly sup- plied us with records of A. obloiigntum taken from wild- flowers in landfills and other post-industrial sites in south- eastern New York and northern New Jersey. Some Penn- sylvania locality records for A. manicatum were provided by Ken Long (Carlisle, PA) and Karl Valley (Pennsylvania Dept. of Agriculture, Harrisburg). SYSTEMATICS Anthidiiuu oblongatum (Illiger) Diagnosis.—Anthidium oblpngatum, a member of the Old World subgenus Proanthidium (Michener and Griswold, 1994), differs from members of the nominate subgenus (Anthidium) by the presence of a small tooth at each side of the mesoscutellum (Fig. 4) and by the charac- Figs. 2-5. Anthidium ohiongatiini. 2-Female mandible. 3-Male mandible. 4-Mesoscutellum. 5-Terga 6 and 7, male. teristic bilobed tergum 7 of the male (Fig. 5) (Michener, 1948). Description.—Female: (Fig. 1) Total body length 8- 11 mm; forewing length 7-8 mm; black, with yellow macu- lahons; clypeus (except extreme apical margin and, in some specimens, a pair of median irregularly oval spots or macu- lations black), outer surface of mandibles, lateral facial maculations adjacent to clypeus, and pair of transverse, narrow bands on each side of vertex (sometimes narrowly connected along midline) yellow; tegulae with antero-lat- eral yellow maculation, small yellow maculation on each axilla, small yellow maculation on lateral margin of mesoscutellum and somehmes yellow spot or band on each side of median line of posterior margin of mesoscutellum; metasomal terga 1-2 with large, triangular, yellow macu- lahons on each side, narrowing considerably towards mid- line; terga 3-5 with large, transverse, yellow maculations on each side, broadly rounded towards midline; tergum 6 with large, round or quadrate, yellow maculation on each side. Face slightly longer than upper interorbital distance; lateral ocelli nearer margin of vertex than to eyes; genae nearly subequal to width of eyes; clypeus slightly convex, apical margin somewhat thickened on each side; mandibles Anthidilim obloncatum new to North America 23 Fig. 6. North American distribution of Antludiuin oblongaluiiL multidentate, with at least 10-11 strikingly dissimilar teeth along inner expanded margin (Fig. 2); vertex and genae bordering preoccipital area not carinate; wings subhyaline to smoky, veins piceous; legs mostly yellowish orange; coxae and trochanters black; front femora black, except apical 1/4 yellowish; middle and hind femora black except apical Vi yellowish; all tibiae and tarsi yellowish orange with dense, short, white-yellow pubescence; spurs testa- ceous. Body pubescence moderately short and thin, mostly pale or white, but rather long and dense over clypeus and face, between antennae, on thoracic venter and pleura, and lateral face of propodeum; ventral scopa golden yellow; punctures coarse, moderately dense, nearly contiguous over entire head and thorax; tegulae shining, with mix- ture of minute and coarse punctures, slightly separated; metasomal terga somewhat shining, punctures quite close but distinct, deep and separated by at least a puncture width, punctures coarser on anterior portion of basal ter- gum, slightly depressed apical rims of terga 1-5 somewhat more finely and closely pmictate, tergum 6 about twice as broad as its median length, broadly triangular, with small median notch on apical margin; lateral margins minutely crenulate. Male: In overall length and coloration extremely similar to female, differing chiefly in the following char- acters: mandibles tridentate along inner expanded mar- giii (Fig. 3), tergum 6 with curved spine at each posterolat- eral margin, and with robust, median projection on apical margin (Fig. 5), and tergum 7 deeply excised at middle and with broadly rounded lobes laterally (Fig. 5). Geographic Distribution This common palearctic species is found throughout most of southern and temperate Europe, and ranges north- ward to about 52° latitude (Warncke, 1980; Westrich, 1990); it also occurs in the Alps up to 1500 meters (Westrich, 1990). hi the United States, A. oblongatum has been collected at the following localities (Fig. 6): MARYLAND: Prince Georges Co., Beltsville, USDA Agric. Research Ctr.-West, 3 September 1997, AGW, ex Sediim spectabile (2, CUIC). NEW JERSEY: Hudson Co., Kearny (site of old landfill), 1 July 1996, M. E. Yurlina, ex Melilotus sp. (1, AMNH; 1, NPIC); 20 July 1995, M. E. Yurlina, ex Lotus corniculatus (2, AMNH; 1, NPIC). NEW YORK: Richmond Co., Staten Island, west shore (landfill complex), 26 June 1994, 17 and 25 July 1995, M. E. Yurlina, ex Lotus corniculatus (4, AMNH; 2, NPIC); 24 August 1995, M. E. Yurlina, ex Lythrum salicaria (1, AMNH; 1, NPIC); 25 June 1996, M. E. Yurlina, no host re- corded (1, AMNH); 26 August 1996, M. E. Yurlina, ex Plucliea purpnirascens (1, AMNH). Tompkins Co., Ithaca, Cornell Univ, A. D. White Gardens, 5, 15 September 1997, ERH, ex Sedum spcctabile (8, CUIC). PENNSYLVANIA: Dauphin Co., Harrisburg, 17 October 1995, 6 August 1996, AGW, ex Perovskia artemesioides (2, CUIC); Hershey, Hotel Hershey gardens, 10-11 August 1996, AGW, ex Sedum spectabile (16, CUIC; 2, PDA; 2, NPIC)). Lancaster Co., Elizabethtown, Masonic Homes, 11 August 1996, AGW, ex Sedum x 'Vera Jameson' (1, CUIC). Specimens are deposited in the following collections, as indicated by the acronym above: American Museum of Natural History New York, NY (AMNH), Cornell Univer- sity, Ithaca, NY (CUIC), Pennsylvania Department of Ag- riculture, Harrisburg, PA (PDA), and USDA National Pol- linating Insects Collection, Logan, UT (NPIC). Bionomics A univoltine species, /!>. oblongatum flies from mid-June to mid-August and is generally associated with xerophilic vegetation and inhabits mesophytic biotopes in its native habitat (Aliev, 1986). It o\erwLnters as a diapausing larva in a cocoon. Anthidium oblorigatum nests in rather dry, warm habitats and is commonly encountered among dry stone walls, in old vineyards, railroad embankments, weather- ing slopes and rocky outcroppings and ridges, as well as in cultivated and "unmanicured" rock gardens with an abundance of flowers (Westrich, 1990). Nests, composed 24 The University of Kansas Natural History Museum Special Publication No. 24 of up to eight cells, are constructed in hollowed-out cavi- ties in various substrates, including soil, among bricks, in rock outcrops, between layers of rocks, in rubbish, and oc- casionally in excavated stems of plants such as thistle and umbellifers (Westrich, 1990). Nesting materials consist of hairs from the following plants: Stacln/s germanica L. and S. byzantina L. (Lamiaceae), Verbasciim (Scrophulariaceae), and Helichrysum and Echinops ritro L.(Asteraceae). A polylectic species, A. oblongatuni obtains nectar and pollen from as many as eight plant families (Miiller, 1996), but mainly from the Crassulaceae (Sediini reflextim L., S. spurium Bieb., S. album L., S. acre L., and Seinpervivum arachnoideiiin L.), Fabaceae (=Leguminosae) (Lotus corniculatus L., Onobrychis viciaefolia Scop., Melilotus alba Desr., and M. officinalis (L.)), and Resedaceae (Reseda liitea L. and R. luteola L.) (Westrich, 1990). Its main pollen sources appear to be L. corniculatus, O. viciaefolia, and S. reflexuni (Westrich, 1990; Muller, 1996). Aliev (1986) also recorded A. oblongatuni from species of Carduus and Cichoriuni (Asteraceae) in Azerbaidjan, Caucasus Minor. At most U.S. collection sites, males and females of A. oblongatuni were captured while visiting flowers of Sedum spectabile Boreau (Crassulaceae). NEW EASTERN NORTH AMERICAN RECORDS FOR A. MANICATUM The geographic range of .4. numicatuni in eastern North America is apparently expanding since original discovery of the species in central New York in 1963. By 1991, A. ninnicatuni had been recorded from eastern Ontario (Smith, 1991), the first Canadian record for this immigrant bee. Examination of unidentified bees in the Cornell Univer- sity Insect Collection (CUIC) and continuing survey work by ERH and AGW in eastern North America have yielded the following new locality records for A. manicatuui: CANADA: Ontario: Niagara Falls, Niagara Parks Botani- cal Gardens, School of Horticulture, 28 August 1996, 16 September 1997, ERH, patrolling flowers of black hore- hound, Ballota nigra L. and other mint spp. (Lamiaceae) (6, CUIC). UNITEd'sTATES: New York: Erie Co., Tonawanda, 3 July 1992, ERH (1, CUIC). Monroe Co., Rochester, High- land Park, 16 September 1997, ERH, ex Salvia farinacea Benth. (6, CUIC). Onondaga Co., Syracuse, 8 July 1979, M. H. Evans (1, CUIC). Ontario Co., Canandaigua, 16 Septem- ber 1997, ERH, ex Salvia farinacea (3, CUIC); Geneva, 16 September 1997, ERH, ex Antirrhinum sp. (3, CUIC). Penn- sylvania: Cumberland Co., Carlisle, Dickinson College, 25- 28 August, 1-2 September 1997, J. K. Long, Jr, ex Salvia farinacea cv. Victoria (3, CUIC; 7, PDA); Mechanicsburg, 18 August 1990, J. K. Long, Jr. (1, CUIC; 1, PDA). Dauphin Co., Harrisburg, Dept. Agriculture building, 7 August 1996, AGW (in association with A. oblongatuni) (2, CUIC). Speci- mens are deposited in the CUIC and PDA, as indicated. LITERATURE CITED Aliev, H. A. 1986. On the fauna of the bee genus Antliiiiiiim Fabr. of the Azerbaidjan SSR, Caucasus. Spixiana 9: 271-274. Hurd, R D., jr. 1979. Family Megachilidae. Pp. 1981-2081, in K. V. Krombein, P D. Hurd, Jr., D. R. Smith, and B. D. Burks (eds.). Cata- log of Hi/iiicnoptcm in Atucricn North of Mexico. Volume 2; Apocrita (Aculeata). Smithsonian Institution Press, Washington, D.C. Jaycox, E. R. 1967. An adventive Antliidium in New York state (Hy- menoptera: Megachilidae). Journal of the Kansas Entomological Society 40: 124-126. Lieftinck, M. A.. 1958. Entomological results of the Finnish expedition to the Canary Islands 1947-1951. No. 16. Apreliminary account of the bees of the Canary Islands (Hym., Apoidea). Commentationes Biologicae. Societas Scientiarum Fennica 18: 1-34. Michener, C. D. 1948. The generic classification of the anthidiine bees. American Museum Novitates No. 1381: 1-29. Michener, C. D. and T. L. Griswold. 1994. The classification of Old World Anthidiini (Hvmenoptera, Megachilidae). University of Kansas Sci- ence Bulletin 55(9): 299-327. Mitchell, T. B. 1962. Bees of the Eastern United States. Volume il. North Carolina Agricultural Experiment Station, Technical Bulletin No. 152. 557 pp. Moure, J. S. and D. Urban. 1964. Revisao das especies Brasileiras do genero Anthidnim Fabricius, 1804. Anais II Congresso Latino-Americano de Zoologia (S. Paulo, 1962) 1: 93-114. Miiller, A. 1996. Host-plant specialization in western Palearctic anthidiine bees (Hymenoptera: Apoidea: Megachilidae). Ecological Mono- graphs 66(2): 235-257. Pasteels, J. J. 1969. La systematique generique et subgenerique des Anthidiinae (Hymenoptera, Apoidea, Megachilidae) de I'ancien monde. Memoires de la Societe Royale d'Entomologie de Belgique 31: 1-148. Pechuman, L. L. 1967. Observations of the behavior of the bee Antliitiitiin manicatum (L.). Journal of the New York Entomological Society 75: 68-73. Schrottky, C. 1901. Biologische Notizen solitarer Bienen von S. Paulo (Brasilien). Allgemeine Zeitschrift fljr Entomologie 6: 209-216. Severinghaus, L. L., B. H. Kurtak, and G. C. Eickwort. 1981. The repro- ductive behavior of AiUliidiuin manicatum (Hymenoptera: Megachilidae) and the significance of size for territorial males. Behavorial Ecology and Sociobiology 9: 51-58. Smith, 1. P. \99\ . Anthtdnim manicatum (Hymenoptera: Megachilidae), an interesting new Canadian record. Proceedings of the Entomologi- cal Society of Ontario 122: 105-108. Warncke, K. 1980. Die Bienengattung Antliidium Fabricius, 1804, in der Westpalaarktis und in turkestanischen Becken. Entomofauna 13: 341-376. Westrich, P. 1990. Die Wildbienen Baden-Wiirttembergs. Spezieller Teil: Die Gattungen und Arten. Verlag Eugen Ulmer. pp. 437-972. Byers, G. W., R. H. Hiigen, and R. W. Brooks (etis.), Entomological Contributions in Memory of Byron A. Alexander. University of Kansas Natural History Museum Special Publicntiou 24. (1999) Pp. 25-32 Distribution and Ethology of Ceramius damarinus Turner (Hymenoptera: Vespidae: Masarinae) in Namibia By Sarah K. Gess' ABSTRACT Ceramius dmuariuus Turner is the only species of Cerniiiiiis so far recorded from Namibia. Little was known of its distribution and nothing had been recorded concerning its ethology. The present paper greatly expands and discusses its known distribution, records for the first time its mating behaviour, flowers visited and provision, nesting situation and nest structure. C. damarimts is recorded from 17°52'S-25°24'S and 14°5rE-17°55'E, indicating an arid savanna distribution. Females collect water for nest construction from the water surface. Males patrol both pools of water and flowers. Matings were observed only on water. At the site where nestiiig was studied a wide variety of plants was in flower, but both females and males were visiting only Sestiviiim sesiwioides (Fenzl) Verde. (Aizoaceae), and pollen from provision obtained from a nest cell and a fully grown larva was solely from this species. However, at a very dry site, where there was no water and no nesting in progress, apparently recently emerged females and males were visiting flowers of Lamiaceae, and males in addition flowers of Aizoaceae, Molugiiiaceae, Zygophyllaceae and Acanthaceae, all ap- parently for nectar. Nesting is in horizontal ground. The nest consists of a vertical shaft surmounted by a short cylindrical turret and gives rise at its base to a whorl of short sloping lateral shafts each ending in an ovoid sloping cell. There is no "bulb" in the shaft and there is no construction of a mud-cell within the excavated cell. Keywords: Nests; Forage plants; Water collection; Mating; Conopid. INTRODUCTION CeraDiiiis dntnarimis, the only species of Cenmiiiis re- corded from Namibia, has been previously poorly known. The species was described by Turner (1935) from females and a male from Ongandjera and a male from Kamanyab, part of material collected by the staff of the South African Museum (Gess 1965). Recently this species was recorded flying abundantly on the road from Okaukuejo to Okondeka, 3.iv.l996, by D.W. and G.T.Gess (Gess et al. 1997). A sample of 12 females was taken. Since then 13 females from Khorixas and four females and one male from Okaukuejo collected by W. Pulawski have been submitted to F.W.Gess for determination. Amongst the targets of fieldwork by F.W. and S.K.Gess in Namibia in March-April 1997 was clarification of the distribution of C. damarinus and investigation of its flower associations, water collecting, mating behaviour and the architecture of its nests Acknowledgments The author would like to take the opportunity to ex- press her gratitude to Byron Alexander for the interest which he took in her research. Thanks are expressed to the following for much ap- preciated assistance: Coleen Mannheimer of the National Botanical Research Institute, National Herbarium of Namibia, Windhoek, for identifying some of the plant specimens and for confirming the author's identifications of others; The South African Foundation for Research De- velopment for a grant for fieldwork and publication costs. DISTRIBUTION Previously recorded and new collection records of C. damarinus are listed in Table 1 under the vegetation types of Giess (1971) and biomes of Rutherford and Westfall (1986) as adapted by Lovegrove (1993) (Figs. 1 and 2). Pre- vious records suggested a northern Namibian distribution: however, the new records of Gess and Gess show that C. damarinus probably occurs throughout the dry savanna and its desert margins, with a rainfall of 100-500 mm per an- num, to at least as far south as 25°24'S wherever water and suitable forage flowers are available. This gives this species a known distribution spanning almost 8 degrees latitude and reduces the apparent gap between it and its sister species in Ceraniius Group 4, C. be\jeri Brauns, from a remarkable 10 degrees to 5 degrees (Fig. 2). Albany Museum, Somerset Street, Grahamstown, 6140 South Africa. E-mail: amsg@warthog.ru.ac.za; S.Gess@ru.ac.za 25 26 Scientific Papers, Natural History Museum, The University of Kansas Table 1. Collection records of Ceramius damarinus Turner listed under the vegetation tv-pes of Giess (1971) (given by number and name, see also Fig. 1) and the biomes of Rutherford and Westfall (1986) as adapted by Lovegrove (1993). Vegetation Type Collection Record 5 - Mopane Savanna (falling within Nama Karoo and Savanna biomes) Ongandjera [17°52'S 15=59'E] (S.A.M. staff [Hesse], iii.1923) Kamanvab [19''35'S 14°51'E1 (S.A.M. staff [Hesse], iii.1925) Etosha National Park: between Anderson Gate and Okaukuejo, 19°13'S 15°55'E (sight record, FW. and S.K.Gess, 26.iii.1997) 15 km west of Khorixas [20°26'S 14''54'E] (W. Pulawski,4.iii.l990) 15.5 km bv road west of Khorixas, 20°26'S 14°54T (FW. and S.K.Gess, l.ivl997) 23km by road from Khorixas to Uis, 20°31'S 14°56'E (FW. and S.K.Gess, l.iv.l997) 10 - Saline desert with Dwarf Shrub Sa\'anna Fringe (falling within Savanna Biome) Okaukuejo to Okondeka [19°10'S 15°54'E to 18°59'S 15°52'E] (D.W. and G.TGess, 3.iv.l996) Okaukuejo [19°10'S 15°54'E] (W. Pulawski, 6.iii.l990) Between Okaukuejo and Halali at 19nO'S 15°58'E, 19°07'S 16°07'E, 19°03'S 16n4'E, 19°02'S 16°16'E and 19°00'S 16°23'E (sight records, FW.Gess and S.K.Gess, 26 and 27.iii.1997) 4 - Semi-desert and Savanna Transition (Escarpment Zone) (falling within Nama Karoo Biome) 12 km southwest of Usakos on the road to Swakopmund, 21°59'S 15°29'E (FW. and S.K.Gess, 22.iii.1997) 7 - Thornbush Savanna (falling within Savanna Biome) 30 km south of Omaruru on the road to Karibib, 2r41'S 15°59'E (FW. and S.K.Gess, 23 and 24.iii.1997) 9 - Dwarf Shrub Savanna (falling within Nama Karoo Biome) Nomtsas, 24°25'S 16''51'E (FW. and S.K.Gess, 18.iii.l997) 43 km and 97 km south of Mariental on road to Keetmanshoop, 24°58'S 17°55'E and 25"24'S 17°54'E (FW, and S.K.Gess, 3.iv.l997) ETHOLOGY Nesting areas and sites: Two nesting areas of C. damarinus were located. The first was in Dwarf Shrub Sa- vamia on the north ban]< of the upper reaches of the Fish River at Nomtsas. The nesting area borders tlie riverine bush, which is composed in the main of tall Acacia karoo Hayne (Fig. 7). The second was in Thornbush Savanna on the slope above a farm dam in the catchment of the Khan River, 30 km south of Omaruru beside the road to Karibib (Fig. 3). The sites in which the nests were aggregated were bare or very sparsely vegetated. All the nests were excavated in horizontal ground. Most nests were situated in expanses of level to gently sloping ground, but a few at the Nomtsas Fig. 1. Map of Namibia showing the distribution, based on collection records (Table 1 ), of Ceramius damarinus (dots) and the v'egetation types of Giess (1971): 1 = Northern Namib; 2 = Central Namib; 3 = Southern Namib; 4 = Semi-desert and Savanna Transi- tion (Escarpment Zone); 5 = Mopane Savanna; 6 = Mountain Sa- \anna and Karst\'eld; 7 = Thornbush Savanna; 8 = Highland Sa- vanna; 9 = Dwarf Shrub Savanna; 10 = Saline Desert with Dwarf Shrub Savanna Fringe; 11 = Tree Savanna and Woodland; 12 = Camelthorn Sa\'anna (Central Kalahari); 13 = Mixed Tree and Shrub Savanna (Southern Kalahari). nesting area had been excavated in level steps in water- cut banks of alluvial soil. At the second nesting area one of the aggregations was situated in a bare patch at the foot of a termitarium oi Macrotermes mossambicus (Hag.) (Fig. 11). In both areas the soil was sandy with sufficient clay to make it plastic when mixed with water. The soil at most sites was compacted to the degree that excavation of the nests required the use of a pick; however, some of the nests on banks on the fringes of the nesting area at Nomtsas were in weakly compacted alluvial soil. Water collection: Water is required by females for nest construction. Water collection was obser\'ed at all sites except that between Kliorixas and Uis, at which there was apparently no water. All water sources were pools in natu- rally occurring drainage channels (Figs. 3-6 and 8). Al- Distribution and Ethology of Ceramius damarinus Turner 27 28 Scientific Papers, Natural History Museum, The University of Kansas mt^i^mi-^'- ?fe.i L^^ Figs. 3-8. 3-Thombush Sa\'anna, thirty kilometres south of Omaruru (21 °41 'S 15°59'E), farm dam in the catchment of the Khan River 4- Water sources of Ceramius damarinm, pools in an otherwise dry bed of a river in Mopane Savanna, 1 5.5 km by road west of Khorixas, 20°26'S 14°54'E. Beyond the river banks the mopane trees are replaced with scattered low mopane bushes. 5-A water source of Ceramius damarimis in Mopane Savanna between Anderson Gate and Okaukuejo, 19°13'S 15°55'E, in the Etosha National Park. 6-A water source of Ceramius damarimis on the edge of the saline desert of the pan, between Okaukuejo and Halali, Etosha National Park. 7-Bridge over the Fish River at Nomtsas (24°25'S 16°51'E), area of bare alluvial soil in the foreground, pool of water in the middle distance, tall Acacia karoo Hayne along the river banks in the distance. 8-The water source of nesting Ceramius damarimis at Nomtsas (24°25'S 16°5rE). Distribution and Ethology of Ceramius damarinus Turner 29 r-^< ^dBg§ 30 Scientific Papers, Natural History Museum, The University of Kansas -\ - 'i Fig. 15. Plans of two nests of Ceramius daniarinus. A and B, vertical in a single plane. A-Nest in strongly compacted alluvial soil. B-Nest in strongly compacted alluvial soil overlain by loosely compacted alluvial soil. Dotted line indicates the level of soil change. C-Transverse plan, in a single plane at level of whorl of cells. At Nomtsas large numbers of females and males were visiting flowers. Although there was a wide variety of plants in flower, the only flowers visited by both males and females of C. damnrinits were the small, shallow, pur- plish-pink flowers of Sesiivium sesuvioides (Fenzl) Verde. (Aizoaceae) (Figs. 9 and 10). Provision was present in only one of the nests investi- gated at Nomtsas. It consisted of a firm white mass with separate loads imparting a "segmented" appearance. The pollen from this provision and from the gut of a fully fed larva from another cell was examined microscopically. All the pollen matched that of the S. sesuvioides on which fe- males were observed foraging. Individual pollen grains were 0.025 mm in diameter. The only other site at which flower visiting was ob- served was that between Khorixas and Uis, the only site at which there was no water. At this site, a patch of mixed flowers growing on the banks of a dry drainage channel, only one plant of S. sesinuoides was located. This plant was being visited by C. daiuarinus (2 males); however, the flow- ers of five other families, Limeum myosotis H. Walter and Gisekin africaiin (Lour.) Kuntze (Moluginaceae) (formerly included in Aizoaceae) (2 males and 1 male), Tribiiliis sp. (Zygophyllaceae) (1 male), Ocimuni americaniim L. and Leiicns pediuelii (Kuntze) Guerke (Lamiaceae) (11 males and 4 females, and 1 male, respectively) and a species of Acanthaceae (1 male) were also being visited. The predomi- nance of males suggests that these wasps were recently emerged and that the flowers were most probably being visited for nectar. As nests were not found in this area and therefore nest provision was not obtained, it is not known which flowers would be visited for provisioning. In the Etosha National Park, where large numbers of females and males were visiting water, none was seen to be visiting flowers; however, it was noted that only pools in close proximity to patches of pink-flowered Aizoaceae were favoured. Sheltering and sleeping: Towards the end of the after- noon, activity at the water, on the forage plants and in the nesting area ceases. At Nomtsas both actively worked on, turreted nests and turretless, apparently old disused bur- rows were investigated in the late afternoon. Each turreted nest contained only a single female. One old nest burrow, lacking a turret, sheltered four males. Nest guarding: No instances of nest guarding were observed. During the heat of the day, when foraging and water collection were in full swing, only actively nesting females were present at nesting sites. No males were present and therefore no guarding of nests by males, as has been recorded for Ceramius bicolor (Thunberg) (Gess and Gess, 1986), was taking place. Nest: The nest (Fig. 15) consists of a sub-vertical bur- row surmounted by a short (up to 7mm high) cylindrical, vertical to sloping mud turret (Figs. 12, 14 and 15) with an inner diameter of 4 mm, equal to that of the shaft, and an outer diameter of 5 mm. Distribution and Ethology of Ceramius damarinus Turner 31 Most nests were in the initial stages of excavation, con- sisting of a turret and a sub-vertical shaft, but no cells. However, of eight turreted nests of 70 mm or more in depth three contained cells. At its maximun depth the main shaft curves outwards to end in a sloping cell 5 mm in diameter at its widest. Two of the nests contained one cell each and one three cells. Each cell terminates a short lateral shaft and all radiate out in a single whorl, that is, all at the same depth (Fig. 15). Nest depth varies considerably, the depths of the vertical shafts of the three nests with cells having been 85, 190 and 215 mm. The walls of the excavated cells had clearly been stablized, compacted and smoothed with the use of water. Cell contents: The cell in one of the smgle-celled nests was newly constructed and empty, and that in the other contained only an egg. Of the cells in the three-celled nest one was open and contained an egg and provision, and the other two had been sealed and the lateral shafts filled with soil and sealed off from the vertical shaft. One of these sealed cells contained provision and a beetle larva, which had presumably devoured the wasp's egg or larva, and the other contained a fully fed C. divuariinis larva. Parasite: The dead, dry, brittle remains of a female C. damarinus were found in a nest which was clearly no longer being worked. The abdomen contained a foreign puparium from which a female conopid later emerged. DISCUSSION The construction of an entrance turret is common to all species of Ceramius for which nesting is known. The use of water as a bonding agent is furthermore common to all southern African species. However, the nests of C. damarinus are unlike those of the other five southern Afri- can Ceramius species groups in that the diameter of the vertical shaft is constant along its entire length, that is there is no portion widened to form a "bulb." Lack of a "bulb" has been recorded, however, for a palaearctic species, C. tuberculifer Saussure (Mauss, 1996). The arrangement of the cells in a single whorl radiat- ing out sub-horizontally from the base of the vertical shaft further distinguishes the nests from those of the other southern African species groups —the cells of groups 2 and 6 being grouped to one side of the shaft, those of Group 3 being positioned sub-vertically beneath the base of the vertical shaft, and those of groups 5 and 8 not being all at one level, with those of Group 5 in addition all being above the base of the shaft. The cell resembles that of Ceramius Group 8, in that the walls are stablized, compacted and smoothed with the use of water, and in the lack of a constructed earthen cell within the excavated cell. Ln this it differs from all the other southern African Ceramius and the palaearctic C, tuberculifer, which do not smooth the walls of the excavated cell but do construct a mud cell within it (Gess, 1996; Mauss, 1996). Gess (1996) constructed a key to the nests of Ceramius species groups 2, 3, 5, 6 and 8. By including Group 4 as represented by C. damarinus, this becomes a key to all the southern African species groups: 1 Excavated cells not containing constructed cells .... 2 — Excavated cells containing constructed cells 3 2 "Bulb" present in vertical shaft Group 8 — "Bulb" absent Group 4 3 No cell terminating main shaft Group 5 — Cell terminating main shaft 4 4 Cells subvertical Group 3 — Cells subhorizontal 5 5 "Bulb" short, bottom end well above level of cells .. Group 2 — "Bulb" long, bottom end level with cells .... Group 6 Gess (1996) stated that pollen from provision obtained from 14 Ceramius species in southern Africa was for each species derived from a single plant family, which indicates that the genus Ceramius is markedly oligolectic and makes it possible to recognize clear associations. She furthermore demonstrated that all species in a group, or in the case of Group 2 a subgroup, specialize in a single plant family and that visits to flowers of more than one plant family, even for nectar, are infrequent. At the time the only flower visiting record for Group 4 was of C. beyeri visiting Aizoaceae. It was therefore predicted that C. damarinus would also specialize iii Aizoaceae. The observations and the analysis of pollen from provision taken from nest cells at Nomtsas were supportive. That such a wide range of plant families were being visited at the site between Khorixas and Uis came as a disconcerting surprise! It is not known which of these plants would have been favoured for pollen collection as observations of the wasps on the flowers suggested that they were imbibing nectar and no nests were located at this site. As noted above only one plant of Aizoaceae was located at this site, so it could be that a shortage of the favoured forage plant resulted in opportunistic behaviour with regard at least to nectar col- lection. This would be in keeping with the records of C. lichtensteinii (Klug) which provisions with Aizoaceae, tak- ing nectar from Blepharis (Acanthaceae) in Eastern Cape when the flowers of Aizoaceae are not available. It is of interest that subsequent to the publication of Gess (1996) Mauss (1996) has recorded a wide range of forage plants for C. tuberculifer and has furthermore found that the cell provision of this species is of mixed plant fam- 32 Scientific Papers, Natural History Museum, The University of Kansas ily provenance. This would indicate that, though in south- em Africa species of Cerauiius are markedly oligolectic and even when visiting flowers for nectar almost always visit the same flowers as they visit for pollen, marked oligolecty in Cernmiiis is not necessarily the rule. Finally, of particular interest are the records of C. damarinus visiting Lamiaceae (=Labiatae). Gess (1996) noted that flowers of this family of plants, though favoured by some masarines in Europe, had not been recorded even as a casually visited nectar plant of any masarine in south- ern Africa. In this connection it is of interest that Mauss (1996) found Teiicriiiin (Lamiaceae) a favoured forage plant of C. tiiberciilifer. LITERATURE CITED Gess, F.W. 1965. Contributions to the knowledge of the South Al^rican species of the genus Ccrnniius Latreille (Hymenoptera: Masaridae). Annals of the South African Museum 48: 219-231. Gess, S.K. 1996. The Pcllcii Wnsfis. ecology/ and naturnl histon/ oftltc Mnsariuae. Cambridge, Mass.: Harvard University Press. Gess, F.W. and Gess, S.K. 1986. Ethological notes on Ccramius bicoior (Thunberg), C. clypecihis Richards, C. nigiipcnuis Saussure and C. socitis Turner (Hymenoptera: Masaridae) in the Western Cape Prov- ince of South Africa. Aiinals of the Cape Provincial Museums (Natu- ral History) 16: 161-178. Gess, S.K., Gess, F.W. and Gess, R.W. 1997. Update on the flower associa- tions of southern African Masarinae with notes on the nesting of Masariiw strucki Gess and Cchnites gariepeusis Gess (Hymenoptera: Vespidae: Masarinae). Journal of Hymenoptera Research 6: 75-91. Giess, W. 1971. A preliminary vegetation map of South West Africa. Dinteria 4: 1-114. Lovegrove, B. 1993. The Living Deserts of Soiitliern Africn. Cape Town: Fernwood Press. Mauss, V. 1966. Contribution to the bionomics of Ceramius tuberciilifer Saussure (Hymenoptera, Vespidae, Masarinae). Journal of Hy- menoptera Research 5: 22-37. Rutherford, M.C. and Westfall, R.H. 1986. Biomes of Southern Africa— an objective categorization. Memoirs of the Botanical Survey of South Africa 54: 1-98. Turner, R.E. 1935. Notes on the masarid wasps of the genus Cemniius. Annals and Magazine of Natural History (10) 15: 290-299. Bi/ers, G. W., R. H. Hagcii. ami R. W. Brooks (eds.), Entomological Contributions in Memory of Byron A. Alexander. University of Kansas Natural History Museum Special Publication 24. (1999) Pp. 33-36 The Ultrastructure of the Wall and Lining Epithelium of Glandular Pouches in Nomadine Bees (Hymenoptera: Apidae: Nomadinae) By Bruce Cutler' and Byron A. Alexander* ABSTRACT The microstructure of the wall and lining epithelium of the female glandular pouches of two species of nomadine bees {Noiuada crcssoiiii and Tricpcoliis distiiictiis) was studied. The wall consists of an inner electron-lucent layer and a very thin electron-dense outer layer. The inside of the outer layer may exhibit thicker focal densities. Wall thickness varies from 0.1-2 mm. The lining epithe- lium forms a monolayer of morphologically secretory cells, with vacuoles, endoplasmic reticulum and vesicles and with numerous villi at the apical border. One region showed fibrous material extending from the villi to the outer layer of the wall. The cells of N. cressoiiii presented a degenerate appearance which may suggest a senescence cycle for these cells. Keywords: Nouiadn; Triepeohis; Transmission electron microscopy; Female reproductive system. INTRODUCTION The gross morphology of the female reproductive sys- tem of nomadine bees has been described and illustrated by Alexander (1996). These pouches were first described by Dufour (1841). Similar appearing but probably non- homologous structures are known in Ichneumonidae (Pampel, 1914; Robertson, 1968) and in Eurytomidae (James, 1926). In nomadine bees the pouches are fluid con- taining and vary greatly in size. In the two genera stud- ied, Noinadn and Tncpcolus, the pouches are relatively large. Nomadine bees are cleptoparasitic on other bees. It has been suggested thu the fluid may enhance egg survival by waterproofing and by inhibiting attack by soil microbes since the eggs are in direct soil contact. Another suggested function is that of chemical masking of the egg so that it is hidden from the host. However, the above hypotheses are purely conjectural and we do not have definitive studies delineating the function of the secretions of the pouches, see Alexander (1996) for a review of these topics. Alexander (1996) referred to unpublished studies elu- cidating ultrastructural details of the lining epithelium of the glandular pouches. This paper illustrates and describes these features. Acknowledgments I (B. C.) wish to dedicate this paper to my late col- league and co-author, Byron Alexander, a fine scientist and a wonderful field companion. I also wish to thank Robert Minckley, Auburn University, Alabama, for information and fruitful discussion. MATERIALS AND METHODS Female specimens of Noninda cressonii Robertson (Nomadini) and Triepeolus distiiicttis (Cresson) (Epeolini) were collected in Douglas Co., Kansas. The internal repro- ductive system was removed while the bees were im- mersed in 2.5% glutaraldehyde in 0.1 M, 7.3 pH cacody- late buffer at room temperature. The organs in fixative were placed in a refrigerator at 4°C. After 2-3 hours the original fixative was replaced by fresh fixative at 4°C. After about 40 hours the specimens were rinsed three times for 10 min- utes each in the buffer without glutaraldehyde. Specimens were fixed again in 1% OsO^ in the same buffer for 2 hours at 4°C. This was followed by two rinses in buffer, 10 min- utes each, and one 15 minute rinse each sequentially in 30%, 50% and 80% ethanol. Specimens were kept in 80% ethanol overnight at 4°C, then processed through higher graded ethanol solutions to acetone and ultimately to Embed 812 epoxy resin (Electron Microscopy Sciences); the resin was then baked overnight at 65°C. Sections were cut with a diamond knife and examined with a JEOL 1200 ExII transmission electron microscope. RESULTS Because of differences in detail, the ultrastructure of the pouch walls and epithelium will be described for each species separately. In both species the epithelium is a uni- cellular layer, and the wall is electron-lucent except at the outer extremity where it may exhibit focal densities. ' Electron Microscopy Laboratory and Department of Entomology, University of Kansas, Lawrence, KS 66045-2106. E-mail: bcutler@alive.bio.ukans.edu * Deceased 33 34 The University of Kansas Natural History Museum Special Publication No. 24 > :Slfct,^»L ** a^'^K'X'i f^-:,^i^^4' -^^ -Ji I ' /'-"I •,->.' Figs.1-4. 1-Pouch wall and lining epithelium of Nomada cressonii: W = wall; arrowheads = focal densities. Scale bar = 2 |am. 2-Pouch wall and adjoining epithelial region of Triepeoliis distinctus: W = wall. Scale bar = 0.5 |am. 3-Pouch wall of T. distinctiis showing focal densities (arrowheads). Scale bar = 0.5 \iTn. 4-Pouch wall of T. distinctus showing fibrils extending from epithelium to outer wall. Scale bar = 0.2 |im. In N. cressonii the wall \'aries from about 0.5-2 |im thick. Three layers can be discerned; the layer adjoining the cells is thickest and is electron-lucent. It is followed by a dense layer with focal denser areas. The outermost layer is very thin, a few tens of nanometers, and is the densest (Fig. 1). The cells exhibit distinct cell membranes, but nuclei are not apparent. The cytoplasm is highly vacuolated with large granular vesicles. The general appearance suggests cell degeneration (Fig. 1). The pouch wall of T. distitictiis is much thinner, rang- ing from about 0.1-1 ^m in thickness. For most of its length in the sections examiiied it appears as an inner lucent layer and a denser outer layer about 50 nanometers thick (Fig. 2). Ln one small region focal densities on the inside of the outer layer, similar to those seen in N. cressonii, appear (Fig. 3). In another small region fine fibrils can be seen extend- ing from the villi to the outer layer of the wall (Fig. 4). The epithelium exhibits distinct cell membranes, nuclei with dispersed chromatin, abundant vacuoles, a few large granular vesicles, abundant mitochrondria, smooth endo- plasmic reticulum, and an apical border with many con- voluted villi (Fig. 5). Golgi complexes can be seen close to nuclei (Fig. 6). At a few places on the inner cell border, small muscle fibril bundles were noted (Fig. 7). The gen- eral appearance is that of a metabolically active cell layer. DISCUSSION Since these are the first ultrastructural observations of glandular pouch structure, no direct comparisons are pos- sible. However, comparisons can be made to known secre- tory cells. As an example, in the numbers of mitochon- dria, nuclear appearance and intimate folding of the dis- tal epithelial border, the pouch bears a resemblance to the cells of the anterior silk gland of Bombyx mori (Akai, 1984). Muscle bundles are found associated with secretory epi- thelium (e.g., Happ, 1984) in insects. Noirot and Quennedey (1974) classified insect epidermal gland cells based on predominantly morphological characters. The lining epithelial cells of the pouches bear closest resem- Ultrastructure of Glandular Pouches in Nomadine Bees 35 o-%o Figs. 5-7. 5-View of epithelium and pouch wall of T. distmctus. See text for details. Scale bar = 2 |im. 6-Golgi complex (arrowhead), vacuoles (v) and vesicles (arrows) in epithelium of T. distincfui^. Scale bar = 1 |am. 7-Muscle fibril bundle adjoining inner layer of epithelium in T. iHstiitcttis. Scale bar = 0.2 nm. blance to Noirot and Quennedeys' (1974) class I cells. These cells are the simplest type, a "thickening of the epidermis" with elaboration of apical microvilli, and numerous mito- chondria, vacuoles and vesicles. It should be noted that Noirot and Queneddey (1974) were discussing epidermis specifically, not epithelium in general. Because the origin of the lining epithelium is not known, it may be prema- ture to attempt to fit it into a classification scheme for epi- dermal cells. The fibrils extending from the tips of the villi (Fig. 4) to the outer wall imply that the epithelial cells are actively involved in wall formation. Thus, there is circumstantial evidence for a secretory function for these cells. The difference in the morphology between the two genera is difficult to explain. While in Triepeolus the epi- thelium shows well preserved organelles, many of those in the Noniadn epithelium seem to be degenerate. It is pos- sible that degeneration of these cells is a normal process. As the adult bee ages, the epithelium completes its func- tions and dies. Thus, the difference may be the result of the examined individuals of Triepeolus being younger than those of Nouiada. The small number of specimens exam- 36 The University of Kansas Natural History Museum Special Publication No. 24 ined, two of each species, precludes any close examina- tion of individual differences, but the morphology was con- sistent between the two specimens of each. Alexander (1996) refers to the granular appearance of the pouch surface, attributing this to invaginations of the epithelial cell layer. Electron microscopic examination shows that the wall itself has greater relief than the under- lying epithelium. LITERATURE CITED Akai, H. 1984. The ultrastructure and functions of the silk gland cells of Bmnbyx mori. Pp. 323-364, in R. C. King, and H. Akai (eds.). Insect Ultrastructure. New York:Plenum. Alexander, B. A. 1996. Comparative morphology of the female reproduc- tive system of nomadine bees (Hymenoptera:Apidae:Nomadine). Memoirs of the Entomological Society of Washington 17:14-35. Dufour, L. 1841. Recherches anatomiques et physiologiques sur les Orthopteres, les Hymenopteres et les Neuropteres. Academie des Sciences, Paris. Memoires presentes par divers savants. Sciences Mathematiques et physiques. Series 2, Volume 7:265-647. Happ, G. M. 1984. Structure and development of male accessory glands in insects. Pp. 365-396, in R. C. King, and H. Akai (eds.). Insect Ul- trastructure. New York:Plenum. James, H. C. 1926. The anatomy of a British phytophagous chalcidoid of the genus Harmolita (Isosoma). Proceedings of the Zoological Soci- ety of London 1926:75-182. Noirot, C, and Quennedey, A. 1974. Fine structure of insect epidermal glands. Annual Review of Entomology 19:61-80. Pampel, W. 1914. Die weiblichen Geschlectsorgane der Ichneumoniden. Zeitschrift fur Wissenschaftliche Zoologie 108:290-357. Robertson, P. L. 1968. A morphological and functional study of the venom apparatus in representatives of some major groups of Hymenoptera. Australian Journal of Zoology 16:133-166. Bi/crs, G. W., R. H. Hagen, and R. W. Brooks (eds.), Entomological Contributions in Memory of Byron A. Alexander. llnwenitij of Kansas Natural Histon/ Museum Special Publication 24. (1999) Pp. 37^3 Mating Behavior oi Dianthidium ciirvatum (Hymenoptera: Megachilidae) at a Nest Aggregation By Gail R. Michener^ and Charles D. Michener- ABSTRACT Interactions between males and females of the megachilid bee Dianihidium curvaium Cockerell were observed during 152.4 h over 4 years at a nest aggregation in a sandy bank in southern Alberta, Canada. Both sexes appeared in late June or early July and disappeared in early to mid-September. Females commonly lived for >3 weeks, some for 7 weeks. Male-female contacts at the nest aggregation ranged from momentary to 201 s; when the numerous contacts of <5 s were excluded, average duration was 59.1 s (» = 80). Bees of both sexes mated multiple times. Males mated repeatedly, often within minutes and sometimes with the same female. Aged females and recently mated females remained attractive to males. Such behaviors suggest a reproductive advantage to the male that is the female's last mate before oviposition. Males did not preferentially seek contact with pollen-carrying females, but copulations tended to last longer with females carrying pollen than with females engaged in any other activity. Although the nest aggregation was a site at which females were located depend- ably, orJy a few males sought females there. The nest aggregation may not be the preferred encounter site for mating because the females most frequently encountered there are closing cells or constructing new cells, so are a day or more away from their next oviposition. Key Words: Dianthidium; Longevity; Multiple mating; Sex ratio; Sexual interactions. INTRODUCTION Male bees do not search randomly for females, but instead concentrate their activity in locations where recep- tive females are most likely to be found. Eickwort and Ginsberg (1980) identified two classes of encounter sites: landmarks containing no food or nesting resources and resource-based sites containing flowers, nest material sources, or nest sites required by females. The distribution of such resources can influence the mate-locating behav- ior of males (Alcock, 1980; Eickwort and Ginsberg, 1980). When patchy resource distribution results in aggregation of females, territoriality by males is predicted, particularly if the number of competitors is low. The frequency with which females mate in a lifetime is another factor that affects mate-locating behavior of male bees (Alcock et al., 1978; Alcock, 1980; Eickwort and Ginsberg, 1980). Single mating by females favors males that locate virgin females, either as they emerge from the natal nest or when seeking their first nectar meal, whereas mul- tiple mating potentially favors the last male to mate be- fore oviposition, especially if sperm displacement occurs. Males could locate polyandrous females shortly before they oviposit either by seeking females at pollen sources or by mating with returning foragers at the nest site. Although females of polyandrous species remain receptive, they may exercise choice among males. The greater complexity of the endophallus in polyandrous than monandrous species of bees suggests that genital morphology is influenced by sexual selection (Roig-Alsina, 1993). Females of the megachilid Dianthidium curvatum sayi Cockerell mate more than once (Custer and Hicks, 1927) and males have a complex multi-lobed endophallus with sclerotized and spiculated regions (Roig-Alsina, 1993). Females nest in aggregations to which they repeatedly re- turn with resin, pebbles, and pollen to build and provi- sion cells. Females gather pollen (and presumably nectar) from several genera of Compositae and resin from Heliatithus (Custer and Hicks, 1927). D. curvatum copulates both on flowers and at nest aggregations (Custer and Hicks, 1927), but little is known about the role of the nest aggre- gation as an encounter site. We observed male-female in- teractions at such an aggregation to determine whether nest aggregations serve as principal encounter sites, whether males are territorial at nest aggregations, and whether males preferentially mate with females that are about to oviposit. Additionally, we report information on the occurrence of multiple mating and the duration of copulatory contacts. Acknowledgments We thank Sarah Hughes, Stacey Barrett-Dowey, Alex Shearing, and Conley Chee for assistance in making ob- servations. Manuscript preparation was facilitated by Vir- ginia Ashlock. ' Department of Biological Sciences, University of Lethbridge, Lethbridge, Alberta, TIK 3M4, CANADA. E-mail: michener@uleth.ca - Snow Entomological Collection, Division of Entomology, Natural History Museum, University of Kansas, Lawrence, Kansas, 66045. 37 38 The University of Kansas Natural History Museum Special Publication No. 24 NESTING SITE AND METHODS From 1988 to 1992, we observed D. curvatiim sayi nest- ing in an area of <1 m= in a vertical, bare, south-facing, sandy bank above the steep slopes that margin the north side of the Oldman River, 2 km south and 5 km east of Picture Butte, County of Lethbridge, Alberta, Canada (49°51'N, 112°42'W). Our site is located near the northern limit of the known range of the subspecies sayi, which ex- tends south to Arizona and Texas. Numerous horizontal burrows of unknown origin in the sandy bank were used by female D. curvatiwi as nesting sites. Each year we in- spected the site on or before 24 June, then revisited at 2- to 3-day intervals to establish the commencement of activity. In 1988 we initiated systematic observations 7 days after we found the nest aggregation, and in the subsequent 3 years we began obser\'ations as soon as 2-5 females were constructing cells. Approximate final dates of activity by bees were determined by visiting the site at 4- to 10-day intervals during September and early October. We recorded the activity of females, the presence of males, and the occurrence of interactions between males and females at the nest aggregation during 152.4 h of ob- servation over 4 years. We obser\'ed, usually for 60 min per day, on 34 days (30.2 h) from 1 July to 15 September 1988, on 19 days (18.4 h) from 29 June to 4 September 1989, on 9 days (12.3 h) from 8 July to 14 September 1990, and on 49 days (91.5 h) from 11 July to 15 September 1991. In all years, most observations were made in the early after- noon (1201-1500 h MDT; 56% of 152.4 h), with the remain- ing obser\'ations distributed between morning (19%) and late afternoon (25%). On 3 days (20-22 July) in 1991 we observed bees throughout the entire daily activity period (approximately 0900-1900 h MDT). In 1991 and 1996, we made a few additional observations at a larger aggrega- tion in a similar sandy bank about 55 m away. For each male-female interaction, we attempted to record the following: identity of the female (based either on a color mark painted on the thorax or on the nest en- trance used); whether the female was approaching or leav- ing the nest aggregation; the stage in the female's nesting cycle (provisioning pollen or collecting nest materials such as resin and pebbles); the location of the pair during con- tact (female's nest entrance or elsewhere at the sand bank); the duration of physical contact (timed to the nearest sec- ond using a watch); and the subsequent behavior of the female (enter her nest or depart from the nest aggrega- tion). Incidental records of male-female interactions, such a those occurring on flowers or at the nest aggregation outside the observation periods, were included in calcula- tions of durations of contacts. RESULTS Seasonality and Longevity Diaiithidiiiiii curvaium in southern Alberta was active for about 10 weeks each year, typically from late June to mid-September (Table 1). Males of D. curvatuni first ap- peared at about the same time as females, but usually dis- appeared in September a few days earlier than females. Numbers of females at the nest aggregation increased dur- ing July, reaching a maximum in late July-early August (Table 1). Females with unworn or little-worn wings were captured in each of the 3 years in wtiich we inspected bees during early to mid-August, and we caught a female ex- hibiting only slight wing wear on 5 September 1988. Thus, emergence of females spanned >4 weeks. Limited infor- mation on wing wear of males indicated that males emerged until at least late July. Based on recaptures of individually marked bees in 1991, females sometimes lived >5 weeks. For example, 8 females survived for 39-50 days (average = 44 davs) and an additional 6 females had minimum lifespans of 22-34 days. Two males, one with slightly worn and the other with moderately worn wings on first capture, sur^'ived at least 9 and 13 days, respectively, after marking. In all years, the extent of wing wear of males generally fell within the range of wing wear of females captured on the same day suggesting that, if the wings of males wear at similar rates to those of females, males may live as long as females. Size and Sex Ratio Males and females of D. curvatum were of similar size. For bees collected on flowers, mostly Lygodesmia, within 100 m of the nest aggregation in 1978, 1988, and 1994, mean ± SD head widths were 3.18 ± 0.14 mm for 18 males and 3.11 ± 0.14 mm for 21 females {t = 1.63, P > 0.10). Based on captures of bees arriving at the aggregation, females substantially outnumbered males (Table 2). Cap- tures at another aggregahon 55 m from our observation site further confirmed our impression that males were not numerous at nest aggregations (Table 2). However, the population sex ratio is not so strongly biased toward fe- males as observations at the nest aggregation imply (Table 2). We collected 10 males and 8 females on flowers on 24 June 1988, and, from the small sample of cells we removed for rearing, 9 males and 3 females emerged. The Nest Aggregation as a Resource for Males To assess the value of the nest aggregation as a mating resource for males, we determined the number of females using the nest aggregation (Table 1) and the frequency with which they visited their burrows. Females almost always Mating Behavior of Dianthidium curvatum 39 Table 1. Active season of D. curvatum in southern Alberta. Year Table 2. Numbers of males and females of D. curoatum captured at nest aggregations and reared from cells. 1988 1989 1990 1991 1992 First sighting Female Male Last sighting Female Male Peak nesting activity'^ 24 June 27 June 29 June 8 July 26 June 24 June [2 July]" [10 July!" 11 July (11 Julv]" 15 Sept 4 Sept 14 Sept 15 Sept 3 Oct 15 Sept —b 5 Sept 2 Sept 19 Sept Date 25 July 40 The University of Kansas Natural History Museum Special Publication No. 24 resting and basking 25 times, hovering 25 times, and pur- suing or interacting with females 15 times. Frequent rest- ing might indicate that the male was aged, but an alter- nate interpretation is that the male was protecting the site against conspecific males. The next day, we saw a chase, the only male-male interaction we witnessed during the study. Despite many hours spent searching for bees at flow- ers and potential resin sources, we rarely observed copu- lations away from the nest aggregation; those we did see were on flowers. We detected no tendency for females of D. ciirvatiii}! to congregate at sites other than vertical sand banks used for nesting or for males to congregate at any location. Interactions between Males and Females Male-female interactions occurred throughout the day, from as early as 0900 to as late as 1720 h MDT, and through- out the active season. The earliest date on which we ob- served sexual contact was 25 June 1988 and involved a fe- male collecting pollen from Li/godesniia. At the nest aggre- gation, the earliest sexual contact was observed on 27 June 1988 and the latest on 5 September 1990. We detected no obvious differences between males and females in the times of day or weather conditions in which they were active. At the nest aggregation, sexual contacts were initiated only by males. Males seemed to rely on visual rather than olfactory orientation to the female; they did not preferen- tially move upwind toward females, but tracked them vi- sually regardless of wind direction. The efforts of males to grasp and copulate with females at the nest aggregation formed a continuous series connecting the following three categories: (a) the male chased a female in flight or at- tempted to pounce on a female, usually as she approached or entered her nest, but she eluded contact by flying away or by quickly entering her nest; sometimes the male fol- lowed the female into her nest burrow but emerged within a few seconds; (b) the male pounced and made brief physi- cal contact with the female; (c) the male pounced, retained his grasp, and attempted to copulate. Males of D. curoatnm pounced on females from above. The male aligned his body with hers and he grasped the female with his legs, thus restraining her and sometimes holding her wings down. As described by Frohlich and Parker (1985) for D. ulkei (Cresson), the male's position was far back on the female, with his head above the posterior end of her thorax or base of her abdomen. He then quickly curved his curled abdomen over and behind the apex of the female's abdomen. The male's body pulsated spas- modically, without production of any sound audible to us, and the pair separated without warning. Because we were unable to determine when insemination occurred, we use 35-1 30- I 25- 8 20- o 53 15- n i 10- z 5- 0- o -^ r o o CD CT> O II o o o in 00 00 CM in Duration of sexual contacts (s) Fig. 1. Durationsof sexual contacts for D. nirpnfKHi. Sample size = 80 contacts in which the male retained hold of the female for >5 s. Mean duration = 59.1 s. the term "contact" to denote that the male made physical contact with the female without implying that insemina- tion occurred. The duration of male-female contacts at the nest ag- gregation was extremely variable, from <1 s to 201 s. Most physical contacts (148/247) were brief (<5 s) and probably did not involve sperm transfer. Sometimes the female ap- peared to break away after 1-2 s but, in general, we were unable to determine whether contacts were brief because the female managed to escape before the male attained a firm grasp or because the male rejected the female. Of 99 contacts with durations of >5 s, we knew both initiation and termination times for 80 (Fig. 1). Although most of these timed contacts (63%) lasted <60 s, a substantial pro- portion (11%) lasted >120 s. Mean ± SD duration of the 80 timed contacts was 59.1 ± 45.7 s. Sexual contacts at the nest aggregation occurred pre- dominantly at nest entrances to which females were re- turning or, less commonly, from which females were leav- ing. Of 183 sexual contacts for which we noted the female's location, 155 occurred at the female's nest entrance with an arriving female, 7 at the female's nest entrance with a departing female, 9 below the bank on a sandy patch where females collected pebbles and chaff, and 12 on the bank face. Arriving females seemed easier for males to grasp because they slowed down and maneuvered to alight at the nest entrance, whereas departing females typically burst out of their burrows and flew away rapidly, without Mating Behavior of Dianthidium curvatum 41 orientation flights. When a male grasped an arriving fe- male, the contact usually occurred with the female par- tially in her burrow and the head of the male either in the entrance or against the sand bank above it. Females en- tered the burrow immediately after the pair separated or they made a brief flight of <60 s before entering the bur- row; in no case did we see a female drop her load, whether resin, pebble, chaff, or pollen, when contacted by a male. If an arriving female evaded contact by flying off, she re- turned within a minute or two, still carrying her load. Departing females sometimes withdrew into their burrows and briefly delayed departure if a male was hovering nearby. Mate Preferences Because we could usually identify the load carried by the female as she returned to her nest, we were able to assess whether males preferentially copulated with females that were provisioning cells with pollen, anci therefore presumably shortly to lay an egg. We determined the avail- ability of pollen-carrying females on the basis of return trips during 24 h of observation on 18 days between 25 July and 14 August, 1991. Pollen collection occurred throughout the day and females commonly required more than 1 day to provision a single cell; the nesting stages of different females within the aggregation were not synchro- nized. We considered each return trip by a female to be a potential opporh.mity for a male to pounce on the female as she maneuvered into her nest burrow. Most return trips involved females engaged in cell construction, not pollen collection; females returned carrying resin on 22.9%, pebbles or chaff on 47.5%, and pollen on 29.6% of trips (» = 1647 trips by 16 females on which the female's load could be identified). Of 68 approaches (11 chases, 13 attempted contacts, 18 contacts of <5 s, 26 contacts of >5 s) by males to 15 of these females, the female was carrying resin on 41.2%, pebbles or chaff on 35.3%, and pollen on 23.5%, in- dicating that resin-carrylng females were approached more often than expected given the frequency with which they arrived at the aggregation (x' = 12.14, df= 2, P < 0.01). When only contacts of >5 s were considered {ii = 26 contacts in- volving 13 females), the bias towards resin-carrying fe- males was still apparent; females were carrying resin on 46.1%, pebbles or chaff on 30.8%, and pollen on 23.1% of such contacts (x-= 7.84, df= 2, P < 0.02). Although pollen-carrying females were not ap- proached preferentially by males, contacts of known du- ration (h = 24) tended to last longer (Mann-Whitney U = 25, P = 0.05) with females that were carrying pollen (mean ± SD = 119 ± 68 s, range 19-201 s, n = 6 contacts involving 6 females) than with females returning with material for cell construction (58 ± 38 s, range 10-141 s, n = 18 contacts involving 9 females). Multiple Mating Males of D. curvatum engaged in numerous sexual contacts, often rapidly switching from female to female and never guarding females. For example, during a 9-min period a marked male copulated for 124 s with one female (provisioning pollen), grabbed a second female (initiating a new cell) but lost contact when they fell from the bank face, chased a third female (collecting pebbles) as she en- tered her burrow and shortly thereafter retained contact with this female for 49 s, then attempted to pounce on the second female again. During 7.9 h of intense observation and 3.4 h of miscellaneous observation (while remarking bees) on 8 consecutive days 3-10 August 1991, this marked male made 21 long (>5 s) contacts (average duration of 15 timed contacts = 66 s) with 11 females and 21 brief con- tacts (<5 s) with 7 of these females and 3 other females. Thus, this male was known to have contacted 14 of the 16 females active at the site 3-10 August. On the last day he was identifiable, 16 August, the marked male engaged in a 30-s sexual contact, indicating that males can mate over at least a 2-week period. Both previously mated and old females remained at- tractive to males. For example, a female was engaged in contacts of 82 and 31 s in a 10-min interval on 29 July 1991 (when she was 1 week old), a 201-s contact on 3 August, a 137-s contact on 13 August, and a 95-s contact on 16 Au- gust. Another female was involved in sexual contacts of 112 and 141 s with a marked male when she was 4 weeks old; 2 days later, she was contacted within an 11-min pe- riod by an uiunarked male for 8 s, by the marked male for 30 s, and by the unmarked male again for 57 s. The oldest females we saw engaged in sexual contacts were 6 weeks old (a 10-s contact on 1 September 1991) and 7 weeks old (an 8-s contact on 25 August 1991). Durations of physical contacts showed no seasonal trend (r = -0.13, P > 0.35, n = 52 contacts of >5-s duration between 25 July and 1 Sep- tember 1991), indicating that aging of bees did not signifi- cantly affect length of sexual contacts. A potential cost to females of continued attractiveness to males is repeated interruption of cell preparation and provisioning. During observations made at the nest ag- gregation 25 July-14 August 1991 (see previous section), females returning with resin were twice as likely to be approached by males (7.4% of 377 resiii trips) as females returning with other nest material (3.1% of 782 trips with pebbles or chaff) or with pollen (3.3% of 488 pollen trips). Of approaches that resulted in contacts of >5 s, returning females were interrupted on 3.2% of resin trips, 1.0% of nest material trips, and 1.2% of pollen trips. 42 The University of Kansas Natural History Museum Special Publication No. 24 DISCUSSION Females of Diniitliidiuiii citrontinu mate more than once, so males presumably gain no reproductive advantage from finding virgin females. Males do not emerge noticeably earlier than females and, as with Anthidielliim, Anthidiuin, and Callantliidiuui (Alcock, 1977), males do not patrol nest aggregations from which virgins emerge. The tendency for males of D. curvatum to contact females repeatedly, even those with which they have recently copulated, indicates that males do not discriminate against recently mated fe- males. If eggs are fertilized preferentially by sperm from the last copulation, as postulated for other anthidiines (e.g., Alcock et al., 1977; Alcock, 1980), every male should at- tempt to be the last one to mate with a female before she lays each of her eggs. A male could achieve this status by guarding the female imtil she lays her next egg, or he could increase his odds of being the last male by mating selec- tively with females that are provisioning cells rather than those that are closing completed cells or constructing new cells. We saw no tendency for males of D. ciirvatiiin to guard females at the nest aggregation or to make contact prefer- entially with females returning with pollen and therefore soon to lay an egg. Instead, males contacted females car- rying resin more often than expected, perhaps because fe- males with a large mass of resin in the mandibles were less agile and more easily grabbed and restrained by males. Given that 70% of incoming trips involved females en- gaged in nest construction, another option for a male to enhance his reproductive success would be to mate longer with those females that are most likely to oviposit soon, assuming that more sperm are transferred during longer copulations. Indeed, we found that copulations with pollen-carrying females tended to last longer than those with females engaged in other nesting activities. Because this pattern would also result if females were less resistant to mating when egg laying was imminent and because we could not ascertain which sex tended to terminate sexual contacts, we cannot assess whether the longer copulations with pollen-carrying females resulted from male choice or female choice. At the nest aggregation, males of D. curvatum were more successful at pouncing on arriving than departing females. This pattern differs from the congener D. Iieterulkei Schwarz for which mating usually involves departing fe- males (Clement, 1976). Females of D. curvatum were vis- ible as they approached the nest aggregation, and they hesi- tated briefly as they maneuvered into their burrows; how- ever, because of their aerial agility, females often evaded males. Females experienced sexual contacts of >5 s dura- tion on <4% of return trips to the nest aggregation. Though such interruptions were short and infrequent, the evasive behavior of females and the brevity of most sexual con- tacts suggest that females at the nest aggregation attempted to avoid copulations. Whereas females of most bee spe- cies are monandrous (Eickwort and Ginsberg, 1980), pro- longed attractiveness to males in other species suggests that the costs of frequent interruption are either minimal or offset by some advantage from polyandry and multiple insemination. Sexual contacts of D. curvatum last longer than those of Authidium and some other Dianthidium, but are shorter than those of Anthidiellum. Average duration of copula- tions is <30 s for Anthidium manicatum (L.) (Severinghaus et al., 1981), A. maculosum Cresson (Alcock et al., 1977), and A. septemspinosum Lepeletier (Sugiura, 1991). Copula- tions are brief for D. Iieterulkei (16-32 s; Clement, 1976), somewhat longer for D. ulkei (0.5-3 min; Frohlich and Parker, 1985), and substantially longer for Authidicllum uotatum (Latreille) and A. perplexum Smith (average dura- tions of 3.4 min and 2.9 min, respectively; Turrell, 1976). Custer and Hicks (1927) reported an average copulation time for D. curvatum of 59.7 s for 7 matings, virtually iden- tical to the 59.1-s average duration we obtained for 80 sexual contacts. Extrapolation from our observation that one male made 21 sexual contacts, with an average duration of 66 s per contact, in 11.3 h of observation suggests that males copulate hundreds of times during their lifetime. The pat- tern of numerous sexual contacts, sometimes with inter- vals of <10 min between consecutive contacts, raises the question of what proportion of sexual contacts by male D. curvatum result in sperm transfer. The complex endophallus of D. curvatum and other anthidiines prob- ably represents an evolutionary adaptation of males to polyandrous mating by females, possibly as a result of in- tersexual selection by female choice or intrasexual selec- tion by sperm competition (Roig-Alsina, 1993). Males may have been selected for an ability to transfer sperm when- ever mating opportunities occur. However, the sperm sup- ply at any time must be finite, so males might be expected to exercise some mate discrimination. Because males of D. curvatum seemingly pounced on any female they were able to approach, discrimination, if it occurred, took place after contacting the female. Males of D. curvatum in southern Alberta encountered females both at flowers and at nest aggregations, but the nest aggregation did not serve as a principal encounter site for sexual interactions. Eickwort and Ginsberg (1980) proposed that territoriality should arise when encounter sites that reliably contain receptive females are localized and defensible. Territoriality and male-male aggression are pronounced in many anthidiines (Alcock et al., 1977; Jaycox, 1967; Severinghaus et al., 1981; Sugiura, 1991), in- cluding the congener D. ulkei (Frohlich and Parker, 1985), Mating Behavior of Dianthidium curvatum 43 but we saw only one possible instance of territorial de- fence by male D. curvatum at the nest aggregation. Because the nest aggregation was used as a mating resource by only a few males, we conclude that most males sought females elsewhere, presumably at flowers. Failure to use nest ag- gregations as principal encounter sites for mating also oc- curs in some other species of bees (Eickwort and Ginsberg 1980), implying that some advantage must accrue to males that locate females elsewhere than the nest aggregation. Although females can be dependably located at the nest aggregation, the area may not be a resource worth patrol- ling or defending because the females most frequently en- countered there are closing cells or constructing new cells and are likely to copulate with other males before they next oviposit. LITERATURE CITED Alcock, J. 1977. Patrolling and mating by males of CnUmiihidium ilhistre. Southwestern Naturalist 22; 554-557. Alcock, J. 1980. Natural selection and the mating systems of solitary bees. American Scientist 68: 146-153. Alcock, J., E. M. Barrows, G. Gordh, L. J. Hubbard, L. Kirkendall, D. W. Pyle, T. L. Ponder, and F. G. Zalom. 1978. The ecology and evolu- tion of male reproductive behaviour in the bees and wasps. Zoo- logical Journal of the Linnean Society 64: 293-325. Alcock, J., G. C. Eickwort, and K. R. Eickwort. 1977. The reproductive behavior of Anthiiiium mnadosum (Hymenoptera: Megachilidae) and the evolutionary significance of multiple copulations by females. Behavioral Ecology and Sociobiology 2: 385-396. Clement, S. L. 1976. The biology of Dianthidium heterulkei hetcrulkci Schwarz, with a description of the larva (Hymenoptera: Megachilidae). Wasmann Journal of Biology 34: 9-22. Custer, C. P., and C. H. Hicks. 1927. Nesting habits of some anthidiine bees. Biological Bulletin 52: 258-277. Eickwort, G. C, and H. S. Ginsberg. 1980. Foraging and mating beha\ior in Apoidea. Annual Review of Entomology 25: 421-446. Fischer, R. L. 1951 . Observations on the nesting habits of megachilid bees. Journal of the Kansas Entomological Society 24: 46-50. Frohlich, D. R., and F. D. Parker 1985. Obser\'ations on the nest-building and reproductive behavior of a resin-gathering bee: Diauthidiuiii idkei (Hymenoptera: Megachilidae). Annals of the Entomological Soci- ety of America 78: 804-810. Hicks, C. H. 1926. Nesting habits and parasites of certain bees of Boulder County, Colorado. University of Colorado Studies 15: 217-252. Jaycox, E. R. 1967. Territorial behavior among males of Aiitliidiuin hcinningeiise (Hymenoptera: Megachilidae). Journal of the Kansas Entomological Society 40: 565-570. Roig-Alsina, A. 1993. The evolution of the apoid endophallus, its phylo- genetic implications, and functional significance of the genital cap- sule (Hymenoptera, Apoidea). BoUettino di Zoologia 60: 169-183. Severinghaus, L. L., B. H. Kurtak, and G. C. Eickwort. 1981. The repro- ductive behavior of Aiitliidiuiii manicatum (Hymenoptera: Megachilidae) and the significance of size for territorial males. Be- havioral Ecology and Sociobiology 9: 51-58. Sugiura, N. 1991. Male territoriality and mating tactics in the wool-carder bee, Anthidiuni scptcnispinosun: Lepeletier (Hymenoptera: Megachilidae). Journal of Ethology [Kyoto] 9: 95-103. Turrell, M. J. 1976. Observations on the mating behavior of AnthidieHum notatiiiii and AtithidicUiitn pcrplcxinn. Florida Entomologist 59: 55- 61. Byers, G. W., R. H. Hagen, and R. W. Brooks (eds.), Entomological Contributions in Memory of Byron A. Alexander. Uniz'ersih/ of Kansas Natural History Museum Special Publication 24. (1999) Pp. 45-50 Behavior and Subcaste Specialization Among Workers of the Giant Tropical Ant, Paraponera clavata (Hymenoptera: Formicidae: Ponerinae) By Rodney S. Hanley^ and James P. Lovett- ABSTRACT This study addresses two questions regarding the natural history and behavior of Paraponera clavata (Fabricius), the giant tropical or Bala ant and the largest ant in Central America: (1) what are their foraging habits; and (2) are foraging workers specialized to particular food types? Our study was conducted at La Selva Biological Station, Costa Rica, in June 1996. Natural history observa- tions and feeding trials of workers of P. clavata show that foraging workers are divided into two subcastes based on specialization to food types (fluid or prey). Additionally, we found that foragers are prima- rily nocturnal in their activity patterns, recruit to food sources, probably lay down individual phero- mone trails, exhibit a curious falling behavior, produce stridulatory sounds, and exhibit the "social bucket" technique of fluid transfer. Keyuvrds: Paraponera clavata; Ponerinae; Formicidae; Behavior; Natural history; La Selva, Costa Rica; Feeding trials. INTRODUCTION Social insect colonies are composed of functionally different castes within groups of same-sex individuals. The social hymenopteran society consists of male and female reproductives and a worker class of sterile females. Work- ers may be differentiated, depending on species, into subcaste-groups of individuals that perform similar types of labor (Waddington, 1988). Indi\'iduals in each subcaste typically have similar morphology that distinguishes them from members of other castes (Wilson, 1953). Ants have the greatest worker differentiation among the Hy- menoptera, with some species having two or three mor- phological subcastes (Oster & Wilson, 1978), although many ant species contain monomorphic workers. Our study addresses two questions regarding the natu- ral history and behavior of the giant tropical or Bala ant, Paraponera clavata (Fabricius). First, what are the foraging habits of these ants, including time of peak activity? It is well known that the temporal patterns of foraging frequen- cies in these ants can vary greatly (see McClusky & Brown, 1972; Hermann, 1973, 1975; Young & Hermann, 1980; Janzen & Carroll, 1983), and have been reported to be in- fluenced by weather conditions (Baader, 1996). These stud- ies highlighted several interesting features of the natural history of P. chwata; however, few details were provided. Second, are foraging workers of P. clavata specialized to particular types of food items? Several authors (Hermann & Blum, 1966; Hermann & Douglas, 1976; Janzen & Carroll, 1983) observed foraging workers rettirn- ing to the nest carrying medium-sized dead insects. None of these observations indicates whether the ants are preda- tors, scavengers, or both. Furthermore, workers are known to return to their nest frequently carrying a large drop of fluid (presumably nectar, honevdew, or sap) between their mandibles (personal observations; Young & Hermann, 1980; Bennett & Breed, 1985). Breed & Harrison (1988) reported allometric "growth" among workers of P. clavata that correlated with task. Guards and foragers were significantly larger in overall size and smaller in ovary size than brood carriers and ants collected within the nests. However, Breed & Harrison (1988) treated foraging workers as one inclusive group and made no attempt to differentiate workers foraging for fluid or prey. Paraponera clavata, the largest ant in Central America at approximately 20-35 mm in length, is dark reddish brown to black, with stiff bristly hairs and a solid build. This giant tropical ant prefers to nest at the base of large trees, especially Pentaclethra niacroloba, and typical colonies are composed of 700-1000 individuals. Recent studies have shown a considerable diversity in patterns of labor allocation in ponerine ants. Fresneau & Dupuy (1988), Fresneau et al. (1982), Pratt (1994), and Pratt et al. (1994) found that some ponerines exhibit tem- poral polyethism typical of species of higher subfamilies, while Traniello (1978) reported a complete lack of tempo- ' Snow Entomological Collection, Di\ision of Entomology, Natural History Museum, University of Kansas, Lawrence, KS 66045. E-mail: rshanley@falcon.cc.ukans.edu. - Department of Entomology, Haworth Hall, University of Kansas, Lawrence, KS 66045. E-mail: ilovett@falcon.cc.ukans.edu. 45 46 The University of Kansas Natural History Museum Special Publication No. 24 ral division of labor for Ambhjopoiic pnllipes (Haldeman). However, Lachaud et al. (1988) detected rough behavioral castes in A. pallipes similar to those reported in other ponerines. Acknowledgments We thank Michael Greenfield and Charles D. Michener for their help during the course of this study, and Daphne Fautin and Richard A. B. Leschen for their critical reviev^s of this manuscript. This is contribution no. 3214 from the Division of Entomology, Natural History Museum, Uni- versity of Kansas, Lawrence. We give special thanks to the late Dr. Byron Alexander whose encouragement and help made this study possible. We considered Byron as both mentor and friend. Byron will be missed. MATERIALS AND METHODS This study was conducted as part the University of Kansas, Department of Entomology Summer Field Course, from 30 May-16]une 1996, and was divided into two parts: (1) natural history observations, and (2) feeding trials. The study site was the Holdridge Arboretum at La Selva Bio- logical Station of the Organization for Tropical Studies, Heredia Province, Costa Rica. A colony of P. clavata was observed at the base of a large canopy tree (Protinm panamense, Family Burseraceae) from 8-13 June 1996. Natural history information, including patterns of daily activity and types of food items, was obtained by observing the colony continuously over a 24 h period, and intermittently over the following 5 d. All ants leaving the colony were counted and their time of departtire recorded. Returning foragers were lightly dusted with a fluorescent dye for identification purposes. Pink dye was used on those returning with prey items and orange was used for those with fluid. All ants returning with food items were counted and their time of return recorded. It was noted whether they had been previously dusted, and a variety of other natural history observations were recorded. Trials were performed to test for specialization to food types among foraging workers. We attached feeding sta- tions at a height of 2 m on the tree and observed them during peak activity times over 3 d, once in the morning and once in the afternoon. Each station consisted of one dish containing cotton soaked with maple syrup and an- other containing tima fish or live termites. All ants forag- ing at the stations were dusted with a fluorescent dye de- pending on the food items being taken. Ants foraging at the stations were counted, and it was noted whether they had been previously dusted. Other natural history obser- vations were recorded. Chi-square contingency analyses were used to examine these foraging data. RESULTS Natural History Observations The observed colony of Paraponera clavata had two openings 10 cm apart at the base of the tree. These were approximately 6 cm x 2 cm, slightly raised, and connected by smooth-walled tunnels just below the soil surface. Foraging workers of P. clavata were primarily noctur- nal in their activity patterns, with little activity between 1000 and 1600 h (Fig. 1). Between 1500 and 1800 h (before dusk) a large number of foragers left the nest and disap- peared into the canopy. Foraging occurred throughout the night with several peaks in returns and exits from the nest. The greatest mass return to the nest occurred between 0700 h and 0800 h. Workers exiting from or returning to the nest used one of two primary trails on the trunk of the tree. Foragers not using one of these trails would search up and around the tree to locate one of them. As returning work- ers approached the nest entrance (at approximately 2 m up the tree), they would accelerate until reaching the nest opening. Occasionally, workers were observed carrying dead ants and heavily sclerotized insect parts from the nest up into the canopy, presumably to a refuse site. Ponerine ants have been reported to lay down indi- vidual pheromone trails (Holldobler & Wilson, 1990), in- cluding P. clavata (Breed & Harrison, 1987; Breed et al., 1987). In our observations, workers would rarely follow one another in a line, but would remain restricted to what appeared to be primary trails. Workers appeared to move independently of each other while antennating the sur- face of the trails, often passing one another. During and after rain showers, workers appeared to have difficulty locating a primary trail on the trunk of the tree. The ants would initiate search behavior and, upon finding a trail, would quickly move in a more directed fashion. About 1 h after the rain showers ended, the ants once again moved "normally" along the trails. We interpret these behaviors as the ants utilizing pheromone trails in their movements; however, further work is needed for confirmation. Perhaps the most curious behavior we observed was individuals falling out of the canopy onto the forest floor. We first suspected this behavior when we observed a few foragers approaching the nest on the ground; most ants traveled only up and down the tree. We then heard noises that sounded like small pellets hitting the leaf litter. We always found a worker ant on the ground in the vicinity of what we perceived to be the sound's origin. Many of these ants held struggling insects (usually other ants) in their mandibles; others held nothing. After hitting the for- est floor, the ants would slowly navigate to the base of tree. Generally, they did not return directly to the nest en- trance; rather, they traveled up and around the tree until Behavior and Subcaste Specialization in Paraponera 47 120 T -0- Leaving Colony -- Returning to Colony Fig. 1. they encountered one of the primary trails. They would then move down this trail into the nest. This behavior has not been reported for other ponerine ants, and we observed no marked individuals performing this behavior Perhaps this activity is performed exclusively by non-foraging workers such as guards or soldiers who were defending foraging columns and food sources. It is likely that this behavior relies on the use of visual land- marks as described by Baader (1996) for P. clavata. We also heard foraging workers producing squeaking sounds upon dishirbance. Direct physical contact elicited this response as well as merely passing a probe or hand over an individual. The function of this behavior is un- known— perhaps serving as a warning to potential preda- tors or used as an alarm call to recruit assistance. These sounds were likely produced through stridulation of the file, located on the mid-dorsal region of abdominal seg- ment IV, and the scraper, located on the posterior margin of abdominal tergite III as described by Giovannotti (1996). Returning foragers were observed carrying one of two primary types of food in their mandibles: arthropod prey items such as termites, or fluid, presumably water or nec- tar. Most of the observed arthropod prey items were dead and chewed up, which made identification difficult. How- ever, a significant portion appeared to be coleopteran and lepidopteran larvae, of which several were still alive. Time of Day 24 h activity pattern of Paraponera clavata colony. Fluid foragers carried droplets suspended between their mandibles, which made these ants quite easy to rec- ognize. Ants not only transported fluid droplets; they also transferred droplets to other individuals. Holldobler & Wilson (1990) called this behavior the "social bucket" tech- nique of liquid food transmission. We observed individu- als with large fluid loads moving from side to side on the trunk of tree. Another forager would approach and wave its head up and down. The solicitor would rapidly antennate the head of the donor, and about half of the drop would be slowly transferred between the mandibles. Af- ter the transfer, both ants returned to the nest. Foraging Specializations To determine whether foragers of P. clavata are spe- cialized on one type of food, we recorded what returning foragers were carrying and whether they had previously carried that type of food item. We did this over a 24 h ob- servation period and during subsequent food-choice tri- als. A returner was defined in our study as a marked ant that returned to the nest with a food item. A switcher was defined as an ant returning with a food item for which it was not originally marked. In those workers returning with fluid, only 3 switchers out of 148 returners occurred over a 24 h period (Fig. 2). In those workers rehirning with prey 48 The University of Kansas Natural History Museum Special Publication No. 24 Fluid (1st Record) Fluid (Returner) tV) Behavior and Subcaste Specialization in Paraponera 49 No termites were observed carried from the nest into the canopy as had been observed with the pieces of tuna. Chi-square contingency analysis of the trial at 0730 h did not support forager specialization (P > 0.5). However, the results of the afternoon trial clearly illustrate special- ization of foragers to one food type (P < 0.001). DISCUSSION some sort of temporal polyethism. Further replicating stud- ies dealing with natural history of P. clavata would be quite useful in resolving this problem. Some evidence reported by other researchers on Pnmponcra indicates that these ants exhibit allometry associated with task (Breed & Harrison, 1988). If this hypothesis is further substantiated, then Paraponera may indeed exhibit permanent subcaste spe- cializations. Our natural history observations of Parapwiicra clavata confirm many obser\'ations made by previous authors. Our observations indicate that foragers of P. clavata are prima- rily nocturnal m their activity patterns with a discernable change in activities correlated with rain showers. It is well known, however, that the temporal patterns of foraging frequencies in these ants can vary greatly (see Hermann, 1975; Young & Hermann, 1980) and have been reported to be Influenced by weather conditions (see Baader, 1996). In addition, our interpretation of the behavior of foragers of P. clavata is that they lay individual pheromone trails that are used by recruited ants to locate food resources. This behavior has been previously reported for P. clavata by Breed & Harrison (1987) and Breed et al. (1987), and our observations agree with their observations. Additionally, individuals of P. clavata are known to produce sound through stridulation (Janzen and Carroll, 1983; Giovannotti, 1996). In our observations, foragers of P. clavata produced squeaking sounds upon disturbance, which we interpret as a warning function of some kind; however, this conclusion remains to be confirmed. It has been well documented in the literature that foragers of P. clavata carry one of two types of food items in their man- dibles: arthropod prey items and fluid, presumably nec- tar. Our observations indicate that foragers are active predators of other insects and utilize graded recruitment when a concentrated source of prey is found. In addition to previously reported observations for P. clavata, we report for the first time additional natural his- tory observations. Individuals of P. clavata were observed falling from the canopy onto the forest floor, typically hold- ing struggling ants of other species. We interpret these in- dividuals to be guards or soldiers likely defending forag- ing columns and food sources. In addition, foragers of P. clavata were observed to transfer fluid droplets to other foragers (i.e., the "social bucket" technique of fluid trans- fer). Presumably, the efficiency of colony pro\'isionmg is greatly enhanced through this behavior. We have shown that foragers of Paraponera clavata are divided into subcastes based on their specialization to for- age for either arthropod prey items or fluid. What we can- not answer at this time is whether these ants exhibit per- manent subcaste specializations or whether these tasks are LITERATURE CITED Baader, A. P. 1996. The significance of visual landmarks for navigation of the giant tropical ant, Paraponera clavata (Formicidae, Ponerinae). Insectes Sociaux 43: 435^50. Bennett, B. & M. D. Breed. 1985. On the association between PeiUaclethra macroloba (Mimosaceae) and Paraponera clavata (Hymenoptera: Formicidae) colonies. Biotropica 17: 252-255. Breed, M. D. & J. M. Harrison. 1987. Individually discriminable recruit- ment trails in a ponerine ant. Insectes Sociaux 34: 222-226. Breed, M. D. & J. M. Harrison. 1988. Worker size, development and divi- sion of labor in the giant tropical ant, Paraponera clavata (Hy- menoptera: Formicidae). Journal of the Kansas Entomological Soci- ety 61: 285-291. Breed, M. D., J. H. Fewell, A. J. Moore, & K. R. Williams. 1987. Graded recruitment in a ponerine ant. Behavioral Ecology and Sociobiol- ogy 20: 407-411. Fresneau, D. & P. Dupuy. 1988. A study of polvethism in a ponerine ant: Neopoiiera apicalis (Hymenoptera: Formicidae). Animal Behavior 36: 1389-1399. Fresneau, D., J. Garcia Perez, & P. Jaisson. 1982. Evolution of polyethism in ants: observational results and theories. Pp. 129-155, hi P. Jaisson (ed.). Social Insects of the Tropics. Vol. 1. Paris: Presses de I'Universite de Paris XIIl. Giovannotti, M. 1996. The stridulatory organ of five ponerine species, a SEM study (Hymenoptera, Formicidae). Fragmenta entomologica, Roma 28: 157-165. Hermann, H. R. 1973. Formation of preforage aggregations in ponerine ants (Hymenoptera: Formicidae), a possible step toward group raid- ing. Journal of the Georgia Entomological Society 8: 185-186. Hermann, H. R. 1975. Crepuscular and nocturnal activities of Paraponera clavata (Hvmenoptera: Formicidae: Ponerinae). Entomological News 86: 94-98.' Hermann, H. R., & M. S. Blum. 1966. The morphology and histology of the hymenopterous poison apparatus. 1. Paraponera clavata (Formicidae). Annals of the Entomological Society of America 59: 397-i07. Hermann, H. R., & M. Douglas. 1976. Sensory structure of the venom apparatus of a primitive ant species (Hymenoptera: Formicidae). Annals of the Entomological Society of America 69: 681-686. Holldobler, B., & E. O. Wilson. 1990. The Ants. Cambridge, MA: Belknap Press/Harvard University Press. 732 pp. Janzen, D. H. & C. R. Carroll. 1983. Paraponera clavata. Pp. 752-753, in D. H. Janzen (ed.), Costa Rican Natural tiistory. Chicago, IL: Universib,' of Chicago Press. Lachaud, J. P., D. Fresneau, & B. Corbara. 1988. Mise en evidence de sous- castes comportementales chez Ainbh/opone pallipes. Actes des Colloques Insectes Sociaux 4: 141-147. McCluskey, E. S., & W. L. Brown, Jr. 1972. Rhythms and other biology of the giant tropical ant Paraponera. Psyche 79: 335-347. Oster G. F., & Wilson, E. O. 1978. Caste and Ecologi/ hi the Social Insects. Princeton, NJ: Princeton University Press. 352 pp. 50 The University of Kansas Natural History Museum Special Publication No. 24 Pratt, S. C. 1994. Ecology and behavior of Gnamfitogenys horni (Formicidae: havior in honey bees. Pp. 385-417, in R. L. Jeanne (ed.), Interindwidual Ponerinae). Insectes Sociaux 41: 255-262. Behavioral Variability in Social Insects. Boulder, CO: Westview Press. Pratt, S. C, N. F. Carlin, & P. Calabi. 1994. Division of labor in Ponera Wilson, E. O. 1953. The origin and evolution of polymorphism in ants. pennsylvannica (Formicidae: Ponerinae). Insectes Sociaux 41: 43-61. Quarterly Review of Biology 28: 136-156. TraniellcJ. FA. 1978. Caste in a primitive ant: absence of age polyethism Young, A. M. & H. R. Hermarm. 1980. Notes on foraging of the giant in Ambli/opone. Science 202: 770-772. tropical ant Paraponera clavata. Journal of the Kansas Entomological Waddington, K. D. 1988. Body size, individual behavior, and social be- Society 53: 35-55. Bi/ers, G. W.. R. H. Hagen, and R. W. Brooks (eds.). Entomological Contributions in Memory of Byron A. Alexander. University of Kansas Natural Histon/ Museum Special Publication 24. (1999) Pp. 51-57 The Female of Diadasia afflicUila Cockerell Unveiled and Verified Using Molecular Markers (Hymenoptera: Apidae) By Sedonia D. Sipes^ ABSTRACT I provide a description of the female of Diadasia afflictula Ckll., a species originally de- scribed from a male holotype, with which no female was previously associated. I describe characters and provide a key to distinguish D. afflictula females from morphologically similar congeners. To test the identity of the female D. afflictula independently from morphology I sequenced 379 nucleotides of the mitochondrial gene Cytochrome Oxidase I in both a male and a female of D. afflictula, and in males and females of four morphologically similar species. The sequence data support the identity of the D. afflictula female, but raise questions concerning the identity of specimens previously determined to be D, martialis Timb. I discuss the use of molecular data to distinguish morphologically similar species and to associate sexes of the same species correctly. Keywords: Solitary bee; Taxonomy; DNA sequence; Species identification. INTRODUCTION Diadasia Patton (Hymenoptera: Apidae) is a genus of about 55 species in arid regions of western North America and South America. Twenty-eight species of these solitary mostly gregarious bees occur in western North America and Central America. Diadasia is particularly interesting not only because its members are putative floral special- ists, but also because different species or groups of species use pollen from distantly related plant families (Linsley and MacSwain, 1957; Adlakha, 1969). Of the 28 North American species, most (16 spp.) specialize on members of the Malvaceae, one species each specializes on Clarkia (Onagraceae), Convolvulus (Convolvulaceae), and Helianthus (Asteraceae), and five species visit Opuntia (Cactaceae). Four species have been so rarely collected that insufficient floral records exist to characterize their host associations. One species that has rarely been collected is Diadasia afflictula Ckll. This bee was described by Cockerell (1910) from a male holotype from Dona Ana Co., New Mexico, without associated female specimens. Timberlake (1941) included D. afflictula in his key to most North American Diadasia males, and Adlaklia (1969) included it m his sys- tematic revision of Diadasia north of Mexico, but still no female specimen had been associated with the males of this species. In August, 1997, 1 collected 5 male D. afflictula from Hidalgo Co. , New Mexico, and 3 females which I determined to be D. afflictula based on their morphologi- cal similarities and geographic association with males, and their distinctiveness from females of other Diadasia spe- cies. Here I provide a description of the female of D. afflictula and a key to separate females of this species from other Diadasia species. I also provide molecular evidence supporting the identification of these female specimens as D. afflictula and discuss the use of molecular data in spe- cies delineation within this genus. Molecular and allozyme data have been used to aid in species identification and delineation in hymenopterans as well as other insects. Allozyme data have been used to differentiate closely related and cryptic species of bees (Scholl et al., 1990; Packer et al.,1992; Blanchetot and Packer, 1992; Carman and Packer, 1997), and DNA restriction frag- ment length polymorphisms (RFLPs) have been used to distinguish the morphologically similar European and African honeybee subspecies (e.g.,McMichael and Hall, 1996). DNA sequence data have been used to taxonomi- cally associate different generations of aphids that alter- nate between host plants (Stem et al.,1997; Aoki et al, 1997). I propose that similar techniques will be useful in species identification in bees in situations where one sex lacks dis- tinguishing characteristics, or where the sexes of a species cannot otherwise be associated with confidence. Both situ- ations are commonly encountered in the genus Diadasia. Acknowledgments This paper is dedicated to Dr. Byron Alexander, who served briefly on my graduate committee, and with whom I had numerous insightful conversations about the ecol- ogy, evolution and systematics of Diadasia. Paul Wolf (USU), Terry Griswold (USDA/ARS), Carol von Dohlen (USU), and Vincent Tepedino (USDA/ARS) gave advice and suggestions. C. Michener, R. Brooks, B. Danforth, and J. Rozen provided locality information helpful in field col- Department of Biology, Utah State Uiuversity, Logan, UT 84322-5305. E-maii: ssipes@biology.usu.edu 51 52 The University of Kansas Natural History Museum Special Publication No. 24 lections. Two reviewers gave suggestions which improved the manuscript. Financial support for this study was pro- vided by the USDA/ARS Bee Biology and Systematics Laboratory, Logan, UT, by the USU Ecology Center, by the AMNH Theodore Roosevelt Memorial Fund, and by NSF Doctoral Dissertation Improvement Grant no. DEB- 9701190. METHODS Morphological Description 1 examined three females that 1 collected and an addi- tional 8 females found in undetermined material from the U.S. National Museum of Natural History. Body measure- ments for the morphological description of D. afflictiila and comparison to other species were made following Adlaklia (1969): Head width was measured at the outer orbits of the eyes in the middle of the face. Thorax width was mea- sured between the outer edges of the tegulae. Forewing length was measured from the base of the wing to the apex. Measurements were made to the nearest 0.05 mm. To in- sure that my measurements were comparable to Adlakha's, I chose a sample of 10 females from the U.S. Pollinating Insect Collection, Logan, UT (USPIC), of D. consociata Timb. and D. lutzi Ckll. (species similar in size and coloration to D. afflictiila) and made the following measurements: head width, thorax width, body length, forewing length. My measurements were all within the ranges reported by Adlakha. In particular, my means for head and thorax widths for these two species (measurements which I use to distinguish D. afflictiila from several other species) were almost identical to Adlakha's. The metasomal tergites have been abbreviated Tl, T2, etc. Molecular Analysis To test the association of male and female D. afflictula independently from morphology, 1 compared their DNA sequences from a fragment of the mitochondrial gene Cy- tochrome Oxidase subunit I (CO-I). This gene has been used to examine phylogenetic relationships of insects, in- cluding Hymenoptera, at various taxonomic levels (e.g., Dowton and Austin, 1997; Pedersen, 1996; Stern et al., 1997). CO-1 has a highly conserved amino acid sequence. 1 expected this gene to exhibit little intraspecific variation in Diadasia based on patterns observed in other insect taxa. No intraspecific variation was observed in 5 species of Bombiis and 6 species of Psithyriis examined by Pedersen (1996), and less than 1% sequence divergence has been found within some species of aphids (Stern et al., 1997). 1 chose one male and one female of D. afflictiila for se- quencing. Additionally, 1 chose one male and one female from each of four species morphologically similar to D. afflictiila: D. martialis Timb., D. lutzi, D. pahiianiiu Timb. and Table 1. Primers used for PCR and sequencing. Locations are based on positions in Apii luellifera mitochondrial genome (Crozier and Crozier, 1993). Designation Sequence Location (5'-3') C1-I-175P (alias Ron) 5' ggatcacctgatatagcattccc 3' 2049 - 2071 C1-N-366T' (alias Barbara) 5' ccacaaatttctgaacattgacca 3' 4213 - 4191 C1-N-2418-D'' 5' atgaagtattaaaatttcgatcaa 3' 2442 - 2419 ' Designed by Harrison et al. (see Simon et al., 1994). '' Designed from Diadnsia , Ptilollirix, and Apis sequences by S. Sipes. D. coiiiociata. An additional male D. martialis was later se- quenced to further evaluate the unexpectedly high varia- tion observed within this species. For comparison, 1 also sequenced one specimen of a morphologically dissimilar species, D. enavata Cr. All specimens had been preserved in 100% ethanol after field collection. I identified these specimens using Timberlake's (1941) and Adlakha's (1969) keys to the North American Diadasia, and Adlakha's de- scriptions and illustrations of male genitalia. DNA was extracted from the thorax (excluding legs and wings) of each specimen, following standard proto- cols (Sambrook et al., 1989). Body parts not used in the laboratory analysis are accessioned and deposited in the collection of the USPIC (accession information available upon request). 1 used the polymerase chain reaction (PCR) to amplify CO-1, using primers Cl-J-1751 (Ron) and Cl- N-3661 (Barbara) (Harrison et al., reviewed in Simon et al., 1994) (Table 1). Amplifications were performed in 100 ml volumes using 50-100 ng of template DNA, 1.5 mM MgCk, Ix Biolase'^' reaction buffer, 1.25 mM each dNTP 0.25 mM each primer, and 5 units of Biolase^'^' Taq DNA polymerase. Thermocycle conditions were as follows: hot start at 94 °C for 5 minutes followed by addition of poly- merase, denature at 94 °C for 1:30 minutes, anneal at 53 °C for 1 :30 minutes, extension at 72 °C for 5 minutes, 30 cycles total, final extension at 72 °C for 10 minutes. The PCR prod- ucts were purified using Wizard® PCR preps (Promega, Madison Wl). The first 379 bases of the amplified fragment were sequenced directly using an AB1-373A DNA se- quencer and the ABl Prism^"^ dye terminator cycle sequenc- ing core kit. Sequences were obtained for both strands of the fragment, using the forward PCR primer (Cl-J-1751) and the internal primer C1-N-2417-D (Table 1). Sequences are deposited in GenBank (consecutive accession numbers AF039090 through AF039101). Sequences were aligned using Sequencher 3.0 and checked \'isually (the fragment sequenced is entirely pro- tein coding, with no deletions or insertions among the taxa examined). 1 used the test version 4.0d59 of PAUP (writ- ten by David L. Swofford) to generate a matrix of pairwise genetic distances (total number of observed pairwise nucle- DiADASIA AFFLICTLILA UnVEILED 53 otide differences), and to perform a UPGMA cluster analy- sis based on these genetic distances. I included a sequence from Apis nicUifera, the honeybee (Crozier and Crozier, 1993; GenBank accession L06178) in the analysis for align- ment and comparative purposes. RESULTS Diadasia afflictula Cockerell D. afflichiln Cockerell, 1910. Ann. Mag. Nat. Hist. London 5: 366. male. Male described in detail in Adlakha (1969). Female: body length 6.15 to 8.6mm (mean = 7.82); head width 2.5 to 2.9 mm (mean 2.7); thorax width 2.4 to 2.8 mm (mean = 2.65); ratio head width to thorax width 0.98 to 1.06 (mean 1.02); forewing length 5.2 to 6.7 (mean 5.98); abdomen width 2.9 to 3.25 mm (mean 3.05); N = 11; in- tegument dark brown to black (slightly reddish in one specimen); head with pale hairs; clypeus with coarse pimc- tures separated by 2-3 puncture diameters; mid tlagellomeres of antenna wider than long; imier orbits of eyes divergent above. Thorax covered with pale hairs; scutum finely punc- tured, slightly shiny between punctures; propodeal enclo- sure polished; tibial spurs not strongly hooked, only very slightly curved at apex; hairs of fore femur sparse and evenly distributed; hairs on ventral side of hind femur sparse, of variable lengths, pale, branched; dorsal side of hind femur polished, nude; scopal hairs of tibia long, pale, branched. Abdomen with hairs short and tightly appressed on all terga; Tl with pale hairs, appressed on entire dorsal face up to edge of anterior basin of tergite; T2 with some brown hairs on disk, often interspersed with pale hairs, and with pale hairs at very base; apical band of thick, pale hairs, usually slightly widened medially; T3, T4 with dark brown hairs on basal half and well defined, wide apical band of thick, pale hairs, usually widened medially; T5 with dark brown hair at base, band of white hairs on disk, apical hairs light brown to ochraceous. Species Delineation Females of several North American Diadasia species possess distinctive characteristics that allow easy identifi- cation. These include D. iiigrifwiis Cr. (black pubescence and process on front femur), D. bitubercnlata Cr (large size and dark scopa), D. angusticeps Timb. (inner eye orbits parallel rather than divergent), and D. tiibcrcidifroiis Timb. (large polished tubercles on frons). Females of the three North American species within the subgenus Dasiapis (D. olivacea Cr., D. ocliraceae Ckll., and D. tropicalis Ckll.) have a yellow spot at the base of the mandible, and rounded meso- and metatarsal claws (Snelling, 1994). Based on morphology, three species groups can be recognized from the remaining North American Diadasia species; within these groups females of different species may be more dif- ficult to distinguish. Females of all of the species known or suspected to use cactus pollen (D. aiistralis Cr., D. riiicoiiis Ckll., D. opuntiae Ckll., D. piercei Ckll., D.friesei Ckll., and D. knabiaim Ckll.) share a combination of characters that separate them from females of other species; these include strongly hooked hind tibial spurs and unbranched hairs on the fore basitarsi. Of the remaining species, which in- clude mostly Malvaceae specialists, two main groups can be recognized by color of pubescence: those whose females have only pale hairs on the metasomal tergites, and those with both pale and dark hairs. D. afflictula Cr. is within this latter group, which also includes D. afflicta, D. consociatn, D. hitzi, D. martialis, D. iiitidifwns Ckll., and D. pabnaruin. The following discussion refers only to the fe- males of species in this latter group. Although D. afflictula overlaps somewhat in body length with D. martialis, D. afflicta, D. iiitidifrons, and D. pahnaruni, the widths of head and thorax of D. afflictula do not overlap with those of the latter species. D. afflictula dif- fers from D. afflicta in having pale scopal hairs; the latter has dark brown scopal hairs. D. afflictula differs from all these in having more appressed pubescence on the metasomal tergites, and in having a head width to thorax width >1 (<0.9 in the other species). Additionally, in D. afflictula, the hairs of the fore femur are sparse and evenly distributed. Ln D. afflicta, D. consociata, D. lutzi, D. martialis, D. Iiitidifrons and D. palmarum, the hairs of the fore femur form a dense brush basally. D. iiitidifrons differs from D. afflictula as well as D. afflicta, D. lutzi, D. martialis, and D. palmarum in having a dull rather than shiny propodeal enclosure. D. afflictula is similar in size and hair coloration to D. consociata and (some) D. lutzi. Again, metasomal pubes- cence distinguishes D. afflictula from D. lutzi. D. lutzi has much sparser and less appressed hair on tergites than D. afflictula. Additionally, D. afflictula has a head width to tho- rax width >1, while in D. lutzi this ratio is <0.9. This latter characteristic is useful in identifying worn specimens. D. lutzi appears more than once in both Timberlake's and Adlakha's keys; apparently, some individuals have only pale hairs on the tergites, and some have dark hairs. D. consociata is similar to D. afflictula with respect to the head to thorax width ratio, but differs in two other characters: the mid tibial spurs are strongly hooked in D. consociata but not in D. afflictula. Additionally, in D. consociata, the frontal line below the median ocellus is a sharp groove flanked by slight but obvious convexities; this feature is lacking in D. afflictula. The integument of D. 54 The University of Kansas Natural History Museum Special Publication No. 24 consociata is often (but not always) reddish, as opposed to dark brown or black of most Diadnsin species. The integu- ment of D. afflictula is usually dark, but one specimen ex- amined had slightly reddish integument. D. consociata has the propodeal enclosure shining to slightly dull; the propodeal enclosure of D. afflictula is always shining. The areas in the U.S. where D. afflictula has been col- lected (Hidalgo and Dona Ana counties, NM) lie outside the known ranges of most of the six morphologically simi- lar species just discussed. Of these, only D. lutzi occurs in southwestern New Mexico (Adlakha, 1969). Key The following key distinguishes females of D. ajflictula from the other females in the complex defined above and takes the place of couplets 20-23 in Timberlake's (1941) key to female Diadasia of North America. I have attempted to reduce this section's dependence on subtle and variable differences in color of metasomal pubescence . My modi- fication can also be used with Adlakha's (1969) unpub- lished key to female Diadasia of North America, replacing couplets 21 to 25. I leave D. lutzi and D. palniarum unre- solved in the final couplet because, in my experience, the characteristics used by both Adlakha and Timberlake to separate these two are not reliable. Additionally, the char- acteristics used to identify D. martialis are those reported by Timberlake and Adlakha; based on molecular data pre- sented below, these characteristics need to be reevaluated. (Potential solutions to these and other problems of species delineation are discussed later.) 20 T1-T5 with strongly appressed hairs (hair on dorsal side of Tl appressed to edge of anterior basin); head as wide or wider than thorax (ratio of head width: tho- rax width usually > 1.0; range 0.98-1.05); small bees (thorax width <3 mm) 21 Hair not strongly appressed dorsally at base of Tl; head narrower than thorax (ratio of head width: thorax width <0.9); bees of variable size 22 21 Mid tibial spurs strongly hooked; frontal line of head be- low median ocellus a sharp groove, flanked by con- vexities; fore femur with thick brush of hair basally D. consociata Timb. Mid tibial spurs not hooked; head lacking both the sharp groove below median ocellus and flanking convexi- ties; fore femur without basal hair brush; hairs sparse and everily distributed on fore femur D. afflictula Ckll. 22 Propodeal enclosure dull; pubescence of thorax and metasoma rich ochraceous overall, except for brown hair at the bases of T3-T5 (and occasionally T2) D. nitidifrons Ckll. Propodeal enclosure polished; pubescence of thorax and metasoma white to pale ochraceous, with dark brown hair at the bases of T2 to T5, or T3 to T5 23 23 T5 with predominantly dark hairs and only a few pale hairs interspersed; scopal hairs dark brown . D. afflicta Cr. T5 with distinct fringe of white hairs on disk; scopal hairs pale (grey-brown in some D. lutzi, but still not as dark as in D. afflicta) 24 24 T2 with abundant dark hairs on basal half; T2-T4 with dark hairs basally, and with thick, well-defined apical bands of pale hairs, widened medially D. martialis Timb. T2 with either pale hairs only, or with few brown hairs on disk, interspersed with pale hairs; T2-T4 with dark hairs basally, and with apical bands of pale hairs, not well defined, sometimes widened medially. D. lutzi Ckll. (in part) and D. palmaruiii Timb. Molecular Analysis Among the Diadasia individuals sequenced, 104 of the 379 nucleotides (27%) were variable. Of these 104 variable sites, 29% were at first codon positions, 10% were at sec- ond positions, and 61% were at third positions. Of the 126 amino acids encoded by this fragment, 34 varied among the Diadasia species examined. The sequence data support the association of the D. afflictula male and female: the 379 nucleotides sequenced were identical in the male and female, and these two dif- fered from the other morphologically similar species at 40 or more nucleotide positions (Table 2, Fig. 1). Similarly, D. consociata, D. lutzi, and D. pahnaruni exhibited much lower (or no) within-species variation than among-species varia- tion. Two males of D. martialis had identical sequences, but differed significantly from a female of D. martialis. For example, the female of D. martialis differed from the males at 18 nucleotide positions, almost as much as this speci- men differed from D. consociata (19 positions) (Table 2). The morphologically dissimilar D. enavata differed from the other Diadasia examined by 51 to 63 positions, almost as much as Apis mellifcra differed from Diadasia (65- 77 positions). It is probable that Apis and Diadasia have diverged so much that third codon positions in this frag- ment have become saturated with substitutions. DlADASlA AFFLICTULA UnVEILED 55 Table 2. Matrix of pairwise absolute genetic distances (number of pairwise variable nucleotide positions) calculated from the 379 nucleotides se- quenced from 12 individuals representing 6 species of Diadasia and Apis mellifera (m = male; f = female). Intraspecific distances are in bold face. 1 2 3 4 5 6 7 8 9 10 11 12 13 D. ajflictula (m) D. afflkhda (f) D. consociata (f) D. consociata (m) D. lutzi (f) D. lutzi (m) D. nmrtinlis (f) D. inartialis (m) D. martialis (m) D. pabnarwn (f) D. palmarum (m) D. enavata A. mellifera 40 40 51 50 44 42 42 47 46 56 67 40 40 51 50 44 42 42 47 46 56 67 26 27 19 13 13 39 38 60 65 26 27 19 13 13 39 38 60 65 4 27 23 23 45 44 63 66 27 24 24 45 44 61 67 18 18 44 43 58 66 37 38 61 63 10 11 12 13 37 38 61 63 1 52 76 51 75 77 DISCUSSION Biology of D. afflictula Our knowledge of the geographic range of D. afflictula is incomplete. Specimens have been collected rarely in Hidalgo and Dona Ana counties in New Mexico; however, the species has also been reported in Mexico (Ayala et al., 1993), where it may be more common. Possibly its restricted occurrence in the extreme southern U.S. represents the northern limit of a more widespread distribution. D. afflictula has been collected from Malvella leyrosa (Ortega) Krapov. (=Sida hederacea (Hooker) A. Gray) and Sphaeralcea spp. Most of the 16 species of North American Diadasia that specialize on Malvaceae prefer Sphaeralcea, but at least one other species, D. consociata, uses M. leprosa as its primary pollen host (Linsley et al., 1952). It is not clear which of these two host taxa is most important to D. afflictula. Most of the specimens I examined were collected from Sphaeralcea, suggesting that it is the most commonly used host. However, collectors of bees are more likely to collect from Sphaeralcea (a colorful roadside wildflower) than from the very inconspicuous, low-growing M. leprosa. Thus it is possible that D. afflictula only occasionally visits Sphaeralcea, but that these floral records are over-repre- sented in collections. Use of Molecular Characters in Species Delineation in Diadasia Males of Diadasia species usually possess distinguish- ing characters of the 6th and 7th sterna, and of the genita- lia. However, females of some species are difficult to dis- tinguish. There are three species groups for which deter- mination of females is especially problematic: a group of large species that specialize on cactus (D. australis, D. rincoiiis, D. opuntiae, D. piercei, and D. friesei), some of the small Malvaceae specialists in which females have pale hairs only on the metasomal tergites (D. diminuta, D. vallicola Timb., D. mexicana Timb., and D. lutzi (in part)), and the Malvaceae specialists with some dark hairs on metasomal tergites, discussed above. Keys currently used to identify females of Diadasia rely heavily on differences in the color and arrangement of pubescence (Timberlake, 1941; Adlakha, 1969). However, states for many of these characters vary within species and overlap among species. Moreover, these characters are easily lost as pubescence wears off during the life span of the bee. To complicate matters further, species using the same host plants are often sympatric on a very small spatial scale. For example, males of at least 8 species of Diadasia have been collected from a single site in Mohave Co., AZ, from Sphaeralcea (Sipes, Tepedino, Griswold, unpublished) so geographic association with males may not provide any clues concerning the identity of females. Moreover, differ- ent species of Diadasia may nest in close proximity or in mixed aggregations (Eickwort et al, 1977; Neff et al., 1982; Ordway, 1984), calling into queshon even the association of individuals collected from nesting sites. Molecular data provide an additional, independent means to solve (or at least gain insight into) species delin- eation problems. Here, I have used DNA sequence data to provide support to the identity of females of D. afflictula, for which no description has previously been published. The UPGMA phenogram also supports the correct asso- ciation of both sexes of D. consociata, D. lutzi, and D. palmarum, but calls into question the idenhty of one or more of the specimens that I identified as D. martialis: the DNA sequence of the female I determined as D. martialis did not match that of D. martialis males nor that of any other 56 The University of Kansas Natural History Museum Special Publication No. 24 22.3 6.7 5.0 3.4 7.6 2.0 6.5 2.8 6.5 9.3 10.7 19.8 29.0 34.0 afflictula-male 2.0 afflictula-female consociata-female consociata-male martialis-male martialis-male martialis-female lutzi-female 2.0 0.5 0.5 lutzi-male palmarum-female palmarum-male enavata A.meilifera Fig. L UPGMA phenogram based on the matrix of absolute distances in Table 2. Numbers on phenogram are branch lengths. The phenogram does not necessarily reflect phylogenetic relationships among the species. DlADASIA AFFLICTULA UnVEILED 57 morphologically similar species (Fig. 1). The three D. nmrtinlis specimens were collected from three geographi- cally proximate sites in Mohave Co., Arizona. The large genetic distance between the D. martialis female and males suggests either that the female is of a different species than the male, or that D. lunrtialis exhibits a much higher amount of intraspecific variation in this highly conserved gene than the other species examined here. A larger sample of indi- viduals will have to be examined and sequenced to test these hypotheses. I am currently sequencing CO-I and other genes in Diadasia for use in phylogenetic reconstruc- tion as well as for use in reevaluating the morphological characters that distinguish closely related species. LITERATURE CITED Adlakha, R. L. 1969. A systematic re\'ision of the bee genus Diadasia in America north of Mexico (Hymenoptera: Anthophoridae). Ph.D. dis- sertation. University of California, Davis, CA. Avala, R., T.L. Griswold, and S.H. Bullock. 1993. The native bees of Mexico. In T.P Ramamoorthy, R. Bye, A. Lot and J. Fa (eds.). Biological Diivr- sity of Mexico: Origins and Distribution. New York, NY: Oxford Uni- versity Press. Aoki, S., CD. von Dohlen, U. Kurosu, and H. Ishikawa. 1997. Migration to roots by first-instar nymphs, and not by alates, in the gall aphid Ch/iicsinithin caiiaticnsis. Naturwissenchaften 84: 35-36. Blanchetot, A., and L. Packer. 1992. Genetic variabilitv in the social bee Lasioglossuni marginatum and a cryptic undescribed sibling species, as detected by DNA fingerprinting and allozyme electrophoresis. Insect Molecular Biology 1: 89-97. Carmen, G.M. and L. Packer. 1997. A cryptic species allied to Halictus ligatus Say (Hymenoptera: HaUctidae) detected by allozyme elec- trophoresis. Journal of the Kansas Entomological Society 69, suppl. (1996): 169-176. CockereU, T.D.A. 1910. Descriptions and records of bees, xxvii. Annals and Magazine of Natural History, London 5: 361-369. Crozier, R.H. and Y.C. Crozier. 1993. The mitochondrial genome of the honeybee Ap'is metlifera: Complete sequence and genome organiza- tion. Genetics 133:97-117. Dovfton, M. and A.D. Austin. 1997. Evidence for AT-transversion bias in wasp (Hymenoptera: Symphyta) mitochondrial genes and its im- plications for the origins of parasitism. Journal of Molecular Evolu- tion 44: 398-405. Eickwort, G.C., K.R. Eickwort, and E.G. Linsley. 1977. Observations on nest aggregations of the bees Diadasia olii'aceac and D. diminuta (Hy- menoptera: Anthophoridae). Journal of the Kansas Entomological Society 50: 1-16. Linsley, E. G. and J.W. MacSwain. 1957. The nesting habits, flower rela- tionships, and parasites of some North American species of Diadasia (Hymenoptera: Anthophoridae). Wasmann Journal of Biology 15: 199-235. Linsley, E.G. , J.W. MacSwain, and R.H. Smith. 1952. The bionomics of Diadasia consociata Timberlake and some biological relationships of Emphorine and Anthophorine bees. University of California Publi- cations in Entomology 9: 267-290. McMichael, M. and H.G. Hall. 1996. DNA RFLPs at a highly polymor- phic locus distinguish European and African subspecies of the hon- eybee Apis melliffra L. and suggest geographical origins of New World honey bees. Molecular Ecology 5: 403—416. Neff, J.L., B.B. Simpson, and L.J. Dorr. 1982. The nesting biology of Diadasia afflicta Cress. (Hymenoptera: Anthophoridae). Journal of the Kan- sas Entomological Society 55: 499-518. Ordway, E. 1984. Aspects of the nesting behavior and nest structure of Diadasia opnintiac Ckll. (Hymenoptera: Anthophoridae). Journal of the Kansas Entomological Society 57: 216-230. Packer, L., C. Plateaux-Quenu, and R.E. Owen. 1992. Electrophoretic evi- dence that Lasioglossuiu (Evylaeus) mediterraneum (Bliithgen) is a spe- cies distinct from L. (E.) laticcps (Schenck) (Hymenoptera, Halictidae), with notes on its phylogenetic position. Canadian Entomologist 124: 371-380. Pedersen, B.V. 1996. A phylogenetic analysis of cuckoo bumblebees (Psithyrus, Lepeletier) and bumblebees {Bombiis, Latreille) inferred from sequences of the mitochondrial gene cytochrome oxidase I. Molecular Phylogenetics and Evolution 5: 289-297. Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular cloning, 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. Scholl, A., E. Obrecht, and R.E. Owen. 1990. The genetic relationship be- tween Bombiis moderatus Cresson and the Bombus lucorum auct. spe- cies complex (Hymenoptera: Apidae). Canadian Journal of Zool- ogy 68: 2264-2268. Simon, C, E FraH, A. Beckenbach, B. Crespi, H. Liu, and P Flook. 1994. Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reac- tion primers. Annals of the Entomological Society of America 87: 651-701. Snelling, R. R. 1994. Diadasia, subgenus Dasiapis, in North America (Hy- menoptera: Anthophoridae). Contributions in Science, no. 448: 1-8. Natural History Museum of Los Angeles County. Stem, D.L., S. Aoki, and U. Kurosu. 1997. Determining aphid taxonomic affinities and life cycles with molecular data: a case study of the tribe Cerataphidini (Hormaphididae; Aphidoidea: Hemiptera). Sys- tematic Entomology 22: 81-96. Timberlake, P.H. 1941. Synoptic table of North American species of Diadasia (Hymenoptera: Apoidea). Bulletin of the Brooklyn Ento- mological Society 36: 1-11. APPENDIX Specimens Examined D. afflictula Ckll.: Hidalgo Co., NM, 3 km east of Animas on hwy 9, 3r57'25"N,108°43'2"W, 22 August, 1997, on Sphaeralcea, 1 female, S. Sipes, USPIC, Logan. Hidalgo Co., NM, 1 km north of Animas, 3r56'56"N,108M8'43"W, 21 August, 1997, on Malvella leprosa, 4 males, 1 female, S. Sipes, USPIC, Logan. Same locality and date, from pan traps,l male, 1 female, S. Sipes, USPIC, Logan. Hidalgo Co., NM, 2 mi southwest of Rodeo, 18 August 1977, on Sphaeralcea, 1 male, 4 females, R.J. McGinley. USNM. Hidalgo Co., NM, 3 mi west of Rodeo, 17 August 1977, on Sphaeralcea, 1 female, R.J. McGinley USNM. Hidalgo Co., NM, 20 mi south ofAnimas, 6 April 1977, on Sphaeralcea, 2 females, R.J. McGinley USNM. Hidalgo Co., NM, 10 mi west of Animas, 20 August 1977, on Sphaeralcea, 1 female, R.J. McGinley. USNM. Bi/ers, G. W., R. H. Hagen, and R. W. Brooks (eds.). Entomological Contributions in Memory of Byron A. Alexander. University of Kansas Natural History Museum Special Publication 24. (1999) Pp. 59-67 Nesting Biology and Immature Stages of the South American Bee Genus Acamptopoeum (Hymenoptera: Andrenidae: Panurginae) By Jerome G. Rozen, Jr./ and D. Yanega^ ABSTRACT Information on the nesting biology of Acnmptopwcuin prinii (Holmberg), A. siibmetallictim (Spinola), and A. argentinuni (Friese) is given, including the following: nesting site characteristics; nest architecture; brood-cell orientation, structure, and lining; description of provisions; egg placement; larval activity; and voltinism. The nesting biology of this genus does not vary significantly from that of Calliopsis. Nesting biology of other genera in the Calliopsini is insufficiently known for comparisons to be made. Males and females of at least Acampiopoeiim siibmetallicimi fly in copula, a behavior that has been observed in certain other Panurginae and warrants further study. The mature larva and pupa of Acnmptopoeum prinii are taxonomically described and illustrated, as is the egg of A. siibinctnlliciini, the larva of which is also described. The mature larvae of Acaitipitopoeimi are similar to, but distinct from, those of Calliopsis. Preliminary observations on larval representatives of other Calliopsini (Arhysosage , Spinoliella, and Callonychium) indicate that they may be similar to one another but quite different from Aaviiptopocuiii and Calliopsis. This suggests that lar\'ae of all calliopsine genera should be compared and analyzed to illuminate the phylogenetic relationships within the tribe. Kex/words: Nest architecture; Chile; Argentina. INTRODUCTION The discovery of a nest of Acntnptopoeiiin prinii (Holmberg) in Brazil in January 1997 by the second au- thor (D.Y.) occasioned the following descriptive notes on its structure and contents and has led to the descriptions of the mature larva and pupa of the species by the first author (J.G.R.). J.G.R. briefly studied the nest of A. snbi)ictalliciini (Spinola) and collected nest components and some immatures in Chile in 1994. He also made brief notes on the nest of A. argentiniim (Friese) in Argentina in 1989. The information on these three species is presented here so that it can eventually shed light on the phylogenetic relationships and taxonomy of the Panurginae, especially the Calliopsini. The genus Acainptopoewn consists of seven or eight, moderate-sized species all of which occur in South America, with the possible exception of A. maculatum (Smith) (Ruz, 1991; Shinn, 1967). It is of phylogenetic in- terest because Ruz (1986) at first postulated that it was the basal clade in the New World tribe Calliopsini. Later Ruz (1991) proposed that it was a sister group only to Calliopsis sensu lato and that together they were a sister group to the three other calliopsine genera (Arhysosage, Spinoliella, and Callonychiinii), all from South America. We follow Ruz (1991 ) by adopting her classification of Calliopsis. Although Calliopsis is the only calliopsine genus known from both North and South America, none of its eleven subgenera is amphitropical. The only original published data on nesting habits for Acamptopoeum were given by Claude-Joseph (1926) for the Chilean A. siibiiietalliciiiii (referred to by Rozen, 1967). He described and illustrated its mature larva in the same pa- per. His account of the lar\'a was interpreted subsequently by Michener (1953), who did not ha\'e an opportuiiity to examine specimens. Specimens (immatures and adults) of Acamptopoeum prinii and A. siibmetalliciim from this study are in the col- lections of the American Museum of Natural History, as are samples of cells and cell closures of A. siibmetallicimi and A. argentiniim. Acknowledgments Work by D.Y. in Brazil was supported by a visiting research grant [301019/96-7 (NV)] from "Conselho Nacional de Desenvolimento Cientifico e Tecnologico" (CNPq). Field collecting in Chile that led to the discovery of the nests of Acamptopoeum siibmetalliciim was sponsored by National Science Foundation grant BSR-9024566 to the American Museum of Natural History, Dr. Norman I. Platnick, PI. We dedicate this paper to the memory of our friend Byron A. Alexander, thoughtful scholar whose contribu- tions to entomology will long be admired. ' Department of Entomology, American Museum of Natural History, New York, NY 10024. E-mail: rozen@amnh.org - Depto. Biologia Geral, Instituto de Ciencias Biologicas Universidade Federal de Minas Gerais, Belo Horizonte, MG, BRAZIL. Present address: Department of Entomology, Entomology Research Museum, University of California, Riverside, CA 92521. E-mail: dyanega@pop.ucr.edu 59 60 The University of Kansas Natural History Museum Special Publication No. 24 Table 1. Statistics and cell contents of nest of Acamptopocum prinii. , — ENTHANCE Figs. 1-2. Diagram of nest architecture of Aavnptopocmn prinii; for explanation, see Table 1 and text. 1-Top view. 2-Side view. Scale (=1 cm) refers to both figures. NESTING BIOLOGY Acaniptopoeiim prinii The study site was in a nearly horizontal, bare trail of hard-packed dirt, roughly 3 m wide, on the periphery of the campus of the Universidade Federal de Minas Gerais (UFMG), in Belo Horizonte, Minas Gerais, Brazil. The nest entrance, unshaded during most of the day, was approxi- mately 30 cm from the nearest trailside vegetation, which at that point was low and herbaceous. D.Y. found the nest on January 22, 1997, when he observed a male returning to the same spot several times within a short period, act- ing as if searching the ground. D.Y. excavated the nest a week later. The nest entrance was closed with loose soil and sur- rounded by a very slight trace of a tumulus. It is uncertain to what depth the closure penetrated, but the tunnel (Fig. 2) below the entrance was vertical for only a short distance (approximately 2 cm) before branching laterally, and the branches did not appear to be soil-filled. The branches meandered downward, as indicated in Figs. 1, 2. The rela- tive ages of the cells (as determined by the Lmmatures, Table 1) suggest that the three branches (A-C) of the nest were excavated and provisioned sequentially, presumably by the single female, which was found at the end of the new- est branch (C). Probably considerably more time elapsed between construction of tunnels B and C than A and B, given the disparity in developmental stages (Table 1). The Branch Cell Lateral Cell depth length (cm) (mm) Cell length X diameter (mm) Contents A A A B B B C C 10 4.5 5 6 6.5 7 7.5 8 5 5.5 24 12 10 18 11 16 10 13 21 (curved) 10.5 X 5 male pupa (damaged) 10x5 male pupa 10 X 5 male pupa n.a. postdefecating larva (starting ecdysis) 10 X 5 predefecating larva 10 X 5 defecating larva 9x5 predefecating larva n.a. predefecating larva 9x5 egg and provision sphere n.a. provision sphere burrow diameter was relatively uniform, approximately 5+1 mm, and the three burrows ended blindly, not in cells. Each terminated at the following depths: A-7.5 cm; B-9 cm; C-6.5 cm. The entire complex of cells spread over roughly 18 cm E-W and 6 cm N-S. Order of cell construc- tion and closure in a branch does not appear to be uni- form from one branch to the next; Ln A the youngest indi- vidual was the lowest, but in B younger individuals were both above and below the oldest. Ten cells were encountered, all closed, and none as yet vacated. They were at the end of soil-filled laterals (mostly straight and horizontal, one curving laterally) of roughly 1-2 cell-lengths. Well-defined, spiral cell closures were not detected, although they may have been obscured by fine, loose soil particles or damaged during excavation. Cells were subhorizontal, with the far end slightly lower and, in side view, almost symmetrically elliptical, with rounded rear ends. Cell walls, not noticeably harder than the sub- strate, were smooth but slightly rougher near the closure and exhibited a thin, faintly shiny, transparent inner coat- ing that was not separable from the surrounding soil. Stored food masses were located close to the far end of the cell and were almost perfect spheres, roughly half a cell-width in diameter (one measured 2.6mm in diameter). They were composed of a mixture of large spherical pol- len grains and much smaller spherical pollen grains with diameters one-half to one-third of those of the large grains. They formed a semisolid matrix that showed no trace of coating when fresh, and, when broken apart later in alco- hol, revealed no traces of any sort of special outer coating, film, or membrane under microscopic examination. When stored in alcohol, these food masses retained their shape and did not break or fall apart. Indeed, the individual pol- len grains continued to adhere to one another even when Bee Genus Acamptopoeum 61 the provisions were broken apart with forceps. These facts suggest that clumping of pollen grains was not caused by nectar (which would have been dissolved in the alcohol) nor by any sort of material coating the provisions. Egg placement was visible in only one cell, in which the egg may have been displaced slightly from normal orientation (the egg itself was damaged when the cell was opened). The egg, white and moderately elongate, appar- ently was placed dorsally (or possibly slightly toward the rear of the cell) on the provision mass, with the egg's long axis aligned along the cell axis. The three pupae in the nest were all males, but no other nests were found to establish whether order of production of the two sexes is consis- tently protandrous. In the cells containing pupae and the postdefecating larva, feces were found on the rear ceiling. The defecating larva, when preserved, was thought to be a postdefecating form because the body was slender, the paired dorsal abdominal tubercles (Fig. 4) were more pronounced than those of predefecating larvae (Fig. 5), and pupal tissue was already becoming organized within the larval integument. However, when dissected and partly cleared in an aqueous solution of sodium hydroxide, the midgut still contained some ingested pollen. The external shape of the larva was probably similar to that of a postdefecating larva before it conforms to the shape of the developing pupa. However, because pupal tissue was vis- ible in predefecating larvae as well as in the defecating larva, a postdefecating body form unmodified by the pu- pal physiognomy must be a fleeting phenomenon when the larva does not enter diapause. A postdefecating larva preserved as it was starting ecdysis revealed how the pupa was encased in the larval integument immediately after the integument split along the dorsal midline and before the pupa started to emerge. The curled pupal antennae occupied nearly the entire lar- val head capsule, the pupal head was contained in the lar- val prothorax, the pupal prothorax and mesoscutum oc- cupied the dorsum of the larval mesothorax, and the pu- pal mesoscutellum was accommodated by the larval met- athorax. The large, paired, pupal mesoscutellar tubercles occupied the paired dorsal metathoracic tubercles of the larva. The presence of pupae in the nest and the absence of quiescent, diapausing larvae indicated that this species has more than one generation a year. Acamptopoeum submetallicum A nesting site of this species was discovered near the top of Cuesta Gavilolen, northeast of Los Vilos, Choapa Province, Chile, on October 16, 1994. J.G.R. observed males flying in long sweeps over the mostly barren nesting area and males and females flying in copula.' In one case the male was seen to disengage only when the female de- scended into her burrow. The nesting area was in the nearly level, unshaded surface at the bottom of a small ravine; the day was partly sunny and cool (less than 65 °F except when the sun was out). Nests were widely scattered and five of them were excavated. The nesting season was just beginning, as indicated by the absence of older larvae and the reduced number of cells in some nests. All nest entrances possessed abundant tumuli, and main burrows were plugged with soil at the surface. Three of the five nests each contained two adult females, and one of these nests contained a single cell. Another nest also consisted of a single cell, two nests con- tained 4-5 cells, and no data were recorded on the last nest except to note that it too was closed with soil at the sur- face. Multi-celled nests seemed to consist of a single bur- row from which laterals radiated, as opposed to the branch- ing configuration of the nest of Acamptopoeum prinii. The unbranching nests of A. submetallicum may have been an artifact created by examining nests in early construction. However, Claude-Joseph's (1926: p. 207) depiction of the nest also showed a single descending main tunnel. Main tunnels meandered downward, as was also the case for the branches of Acamptopoeum prinii. Two were approximately 5 mm in diameter near the surface, and one was 4 mm wide at the surface but opened to 6mm at lower depths. Cells were arranged singly, each with its own lateral, soil-filled after closure, as was the case for Acamptopoeum prinii. They were essentially symmetrical around their long axis. Cell lengths ranged from 10 to 11 mm (N=2); maxi- mum cell diameter was 6.5-7.0 mm (N=5); minimum di- ameter of cell entrance was 4.0^.5 mm (N=4). Cell orien- tation was nearly horizontal with the rear slightly lower than the front end. The cell wall consisted of soil not dis- cernibly different from that of the substrate except for be- ing more consolidated and therefore tending to remain intact when excavated by j.G.R. The hard cell wall con- trasts with that of A. prinii, but detection of a hard cell wall may depend on soil moisture and nature of substrate as much as on manipulation of the wall material by the fe- male during construction. The inner cell surface was mostly smooth but somewhat rougher next to the cell closure. The smooth surface was dark and possessed a shiny, transpar- ent lining when first exposed. The lining remained faintly shiny when dry and was water retardant when tested. The rough surface near the closure lacked the lining (or the ' J. G. R. again observed pairs of Acamptopoeum submetallicum flying in copula, at 5 km. W. Monte Patria, Limari Province, Chile, on October 6, 1997. Here two couples were visiting a yellow composite. The female of one was carrying pollen on her hind legs, an indication that females forage while in copula. 62 The University of Kansas Natural History Museum Special Publication No. 24 APEX SALIVARY SWELLING ' Figs. 3-9. 3-Egg of Acamptopoeum subinctnlUcuin, lateral view, anterior end at left. 4-9. Mature lar\'a of Acatnptopoeum prinii. 4-Defecat- ing larva, lateral view. 5-Predefecating larva, lateral view. 6-Defecating larva, frontal view, enlarged, showing pigmentation pattern on right. 7-Head, lateral view. 8-Head, frontal view. 9-Right mandible, ventral and inner views. Scales (=1 mm) refer to Figs. 3-5, respectively. lining may have been fenestrated) and was water absor- bent when tested. A similar situation is reported below for the cell surfaces of A. argentinum. Cell depths ranged from 5 to 9 cm, essentially the same as those of Acamptopoeiim priiiii. Cell closures were a well-formed, slightly concave spiral of about 4 coils on the inside, much as reported by Claude-Joseph (1926). Hence, this feature appears to con- trast with the inner surface of the closure of Acamptopoeiim prinii. One preserved closure, with a center thickness of about 1 mm, revealed a more or less smooth, evenly con- cave outer surface, against which the female had piled the fill of the lateral. As in Acampiopoeum prinii and A. argentinum, larval provisions were perfect spheres, 3.9-4.4 mm (N=2) in di- ameter, apparently considerably larger than those of A. prinii. These spheres appeared to have a thin, transparent coating in contrast to those of A. prinii and A. argentinum. The reported difference is significant since coated provi- sions have been reported for many Calliopsini and some Perditini. Fresh provisions of all species should be re-ex- amined carefully to confirm this feature. Bee Genus Acamptopoeum 63 Two eggs (Fig. 3) preserved in Kahle's solution were nearly white and possessed a clear, smooth, transparent chorion. Strongly curved, they were 2.5 mm long and 0.5- 0.6 mm in maximum diameter, with a round anterior end, an elongate, parallel-sided midsection, and a rather short, rapidly narrowing posterior end. The fact that one of the eggs had a mass of pollen adhering to the posterior end suggests that the egg was attached to the pollen mass by this end. A single larva, described below and reared in the labo- ratory, was preserved as a postdefecating form on Novem- ber 9, 1994. This information, as well as Claude-Joseph's (1926: p. 207) account, suggests that Acaiiiptopoeiim siibmetallicum, in contrast to A. prhiii, has a single genera- tion per year. Acamptopoeum argentinum A single nest of this species was discovered and exca- vated at San Fernando, Catamarca Province, Argentina, on November 3, 1989, by J.G.R. It was found at the side of an unpaved, essentially horizontal, unshaded roadway, where the soil was sandy. The soil was powdery and dry on the surface but moist and firmer 2-3 cm below. The single female was seen disappearing into the pow- dery surface. No nest entrance was discernible, and the main burrow was not detected, perhaps because it had been soil-filled. Laterals to closed cells were soil-filled; one leading to an open cell was 4.0 mm in diameter except just before the cell mouth where it narrowed to 3.5 mm. All four cells were arranged singly at a depth of 7-8 cm and were subhorizontal, tipping to the rear by about 10°. Cell dimensions were: maximum diameter 5.8-6.2mm (N=2); length 10.2 mm (N=l). A cell wall differentiated from the substrate was not certainly detected. Tlie inner surface of the cell was very smooth over the rear two-thirds and less so over the front one-third. When tested with water droplets, the smooth area was highly water retardant and the front, somewhat rougher surface was immediately absorbent. However, at the time of excavation, the entire surface appeared shiny. Cell closures were a moderately concave spiral on the inside of approximately three coils and a relatively smooth concave surface on the outside, probably with a center thickness similar to that oi Acamptopoeum siibmctnlllciim. Provisions were spheres 4.0—4.2 mm (N=3) in diam- eter that rested on the floor of the cell. They were quite firm and not coated by a waterproof lining. When broken apart, one consisted of yellow pollen toward the outside and orange pollen inside, suggesting that two kinds of plants had been visited by the foraging female. One cell contained a strongly arched, whitish egg with a shiny, clear chorion. It was 2.2 mm long and 0.55 mm in maximum diameter, with one end (presumably front) round, and the other more tapering. IMMATURE STAGES Mature Larv.4 of Acamptopoeum prinu Diagnosis.—(Figs. 4-9) See Diagnosis of Acamptopoeum submetaUicum, below. This description follows the format of that of the larva of Aiithemurgus pnssifloiae Roberston (Neff and Rozen, 1995). Head.—(Figs. 7, 8) Integument of head capsule with scattered, inconspicuous sensilla, which were not obvi- ously setiform; head capsule, maxillary and labial palpi, and, apparently, lower end of salivary swelling (see Re- marks, below) faintly pigmented; mandibles and their at- tachments to subgenal ridge more darkly pigmented than elsewhere. Head moderately small in relation to rest of body; head capsule much wider than maximum length from vertex to lower clypeal margin. Anterior tentorial arms slender but present; posterior arms present, arising from deeply in- flexed posterior pits; tentorial bridge and dorsal arms (of defecating larva) absent, presumably because individual was nearing ecdysis; anterior and posterior tentorial pits in normal position; postoccipital ridge normal in position, thin, almost absent at midline of head (perhaps as result of approaching ecdysis); median longitudinal thickening of head capsule absent; hypostomal ridge well developed with ramus connecting to postoccipital ridge because of deep inflection of head capsule at posterior tentorial pit; pleurostomal ridge moderately developed; epistomal ridge well developed between anterior mandibular articulation and anterior tentorial pit, absent between anterior tento- rial pits. Parietal bands indistinct, scarcely noticeable. An- tennal prominences moderately low; antennal papilla and disc small; papilla short, distinctly shorter than its basal diameter, with three sensilla. Vertex in lateral view (Fig. 7) evenly rounded; frontoclypeal area normal in length. La- brum normal in size and shape for Panurginae, without sclerite; paired labral tubercles arising from disc, moder- ate in size; epipharyngeal surface strongly spiculate on sides but nonspiculate medially. Mandible slender, gradually tapering apically in adoral view (Fig. 9); dorsal surface spiculate; outer surface with inconspicuous seta-bearing irregularity (too indistinct to be called a tubercle) one-third distance to apex; apex at- tenuate with row of denticles along upper apical edge, without enlarged subapical tooth; ventral apical edge with- out denticles; cusp weakly produced adorally, not pro- duced ventrally, with numerous, evenly spaced denticles. 64 The University of Kansas Natural History Museum Special Publication No. 24 but without large tooth. Labiomaxillary region (Fig. 7) moderately recessed and fused as in other Panurginae. Maxilla in lateral view (Fig. 7) projecting moderately be- yond labial apex; cardo a vague, unpigmented cuticular thickening; stipes and articulating arm of stipes not evi- dent; palpus moderate in size, equal to or somewhat smaller than labral tubercle, directed forward (rather than down-turned), with dorsal surface curving downward no more than ventral surface curving upward; dorsal surface of maxilla and palpus spiculate, spicules large. Labium weakly divided into prementum and postmentum; premental sclerite not evident; in lateral view (Fig. 7), la- bial apex clearly recessed compared to apices of maxillary palpi; labial palpus moderate in size but conspicuously smaller than maxillary palpus. Salivary opening in curved groove that extends to hypopharyngeal groove; salivary swelling extending no farther than apex of labium on ei- ther side. Hypopharyngeal groove deep; hypopharynx strongly spiculate dorsally. Body.—Integument without obvious setae but ante- rior paired dorsal tubercles with a few microscopic, setiform sensilla; body spiculation extensive in many ar- eas, but paired dorsal tubercles of thorax and first abdomi- nal segment nonspiculate; other paired dorsal tubercles with apices finely, densely spiculate; venter of body in- conspicuously spiculate by comparison with many dorsal areas; integument of paired dorsal thoracic tubercles and those of first abdominal segment pigmented about as darkly as head capsule (Fig. 6). Body form (Fig. 4) of def- ecating larva moderately robust; that of predefecating larva (Fig. 5) somewhat more so; intrasegmental lines not evi- dent; lateral prothoracic swelling absent (this feature should be evaluated on postdefecating larva); thoracic seg- ments and first abdominal segment of both defecating and predefecating larvae with paired dorsal tubercles; paired dorsal tubercles evident on abdominal segments II-VII of defecating larva (Fig. 4) but absent or nearly so on predefecating larva (Fig. 5); anterior body tubercles tend- ing to be conical but posterior ones more or less transverse (but not pronouncedly so as in Halictinae); prothoracic dorsal tubercles large, with apex of each directed forward (Figs. 4, 5) and laterad (Fig. 6); bases of pro- and mesotho- racic tubercles not certainly forming pockets, as described for Antliemurgus (but this character perhaps better evalu- ated on postdefecating larva); bases of meso- and metatho- racic paired dorsal tubercles apparently not forming pock- ets; dorsal tubercles of abdominal segment VIII of defecat- ing larva faint, those of segments IX and X absent; pleural region not produced; venter of most abdominal segments slightly produced on each side; abdominal segment X at- tached dorsally to IX on predefecating larva (Fig. 5), not as distinctly so on defecating larva (Fig. 4). Spiracles small, unpigmented, subequal in size; peritreme present; atrium projecting slightly beyond body wall, with rim; atrial wall with faint ridges parallel with atrial rim; subatrium of moderate length, with about ten chambers. Gender char- acters not detected. Material Studied.—One defecating larva, three predefecating larvae, and one postdefecating larva start- ing ecdysis, Belo Horizonte, UFMG, Minas Gerais, Brazil, 1-29-1997 (D. Yanega). Remarks.—Problems exist in our understanding of the anatomical variation in the labiomaxillary/hypopharyn- geal regions of bee larvae. The salivary swelling was so named and first identified by Rozen (1993) with reference to Euherbstia (Andreninae) and later referred to as a sali- vary tubercle in Ancylandrena (Andreninae) (Rozen, 1994) and Neffapis (Panurginae) (Rozen and Ruz, 1995). How- ever, in another sense the swelling had been identified earlier by McGinley (1981 : p. 25, character 79, state 3) when he described the salivary opening as being strongly upcurved. The area enclosed by the upcurved opening and the hypopharyngeal groove is the salivary swelling. He pointed out that the strongly upcurved salivary opening was possibly an andrenid autapomorphy. If this proves to be true, then the swelling is also an andrenid autapomorphy, although the extent of projection of the swelling varies from taxon to taxon within the family. The salivary swelling appears to be a different feature from the salivary plate, a term used by McGinley (1981: p. 26, character 83) for a structure in certain cocoon-spinning bee larvae. Both features are associated with the salivary opening. However, the former is circumscribed by the upcurved opening and the hypopharyngeal groove, and the opening itself does not project because the larva is noncocoon-spinning. In the latter, the plate surrounds the salivary opening, which is projecting, and is found only on cocoon-spinning larvae of the Melittidae"', Megano- miidae\ and the Diphaglossinae, as reported by McGinley (1981). It is unclear whether or not the salivary plate of the diphaglossines is homologous with that of the melittids and meganomiids. McGinley (1981: character 87) referred to a sclerotized ramus associated with the hypopharyngeal groove (p. 27) and later to a hypopharyngeal groove that is sclerotized laterally (pp. 59-60) (presumably references to the same structure). Rozen and Michener (1988) interpreted this scle- rotized structure to be the articulating arm of the stipital sclerite, actually a part of the maxilla. Sensu Alexander and Michener (1995). Bee GENnjs Acamptopoeum 65 POSTDEFECATING LaRVA OF ACAMPTOPOEUM SUBMETALLICUM Pupa of Acamptopoeum prinh Diagnosis.—The single, laboratory-reared, postdefecating larva described here was shrunken and perhaps deformed, presumably because it was reared un- der unnatural conditions. Furthermore, the body contents did not completely dissolve when treated repeatedly with an aqueous sodium hydroxide solution, perhaps because the larva was infected by fungal hyphae. However, some of its features can be compared with those of Acamptopoeum prinii. The only characteristic identified to separate the larva of this species from larvae of i4. prbiii is the presence of teeth along the lower apical edge of the mandible (but see Rozen [1958: p. 25] for a discussion of the taxonomic reliability of such a larval feature). Perhaps the significant point of the following description is the similarity of the larvae of the two species. See section of Taxonomic and Phylogenetic Interpre- tations, below, for characters that distinguish mature lar- vae of Acamptopoeum from those of other Calliopsini. Head.—As described for Acamptopoeum prinii except for following: Head size relative to body size uncertain because of shrunken condition of body. Condition of in- flection of posterior pits, tentorial bridge, hypostomal ridge, pleurostomal ridge, and epistomal ridge unknown. Labral tubercles evident. Mandible with ventral apical edge bearing distinct denticles in addition to those on upper apical edge; cuspal teeth apparently greatly reduced com- pared with those of A. prinii. Labiomaxillary region includ- ing salivary opening and hypopharynx greatly distorted. Body.—As described for Acamptopoeum prinii except for following: Body spiculation pattern unknown. Body form of postdefecating larva presumably moderately ro- bust but larva somewhat shrunken because reared under laboratory conditions; thoracic segments and abdominal segments I-VII with paired dorsal tubercles; those of ab- dominal segments VIII low but evident (all dorsal body tubercles may have been accentuated because of shrunken body); prothoracic dorsal tubercles large with apex of each directed forward and laterad as in /I. prinii, but apex not so strongly produced as in that species; bases of pro- and mesothoracic and meso- and metathoracic paired dorsal tubercles apparently not forming pockets; dorsal tubercles of segments IX and X absent; pleural regions produced, apparently owing to shrunken condition of body; abdomi- nal segment X attached somewhat dorsally to IX as seen in lateral view. Material Studied.—One postdefecating larva, 33 km NE Los Vilos, Choapa Province, Chile, collected X-1 6-1994 as intermediate larva, preserved XI-7-1994 as postdefecating larva (Rozen, Quinter, and Ascher). The main purpose of the following description is to record the distinctive pupal characteristic oiAcamptopoeum prinii for future systematic analysis. The diagnosis may enable this pupa to be distinguished from known pupae of other panurgines, as listed in Neff and Rozen (1995). Diagnosis.—(Figs. 10-12) The pupa oi Acamptopoeum prinii possesses the following combination of diagnostic features (Figs. 10, 11): tegular tubercles present; paired scutal tubercles distinct; hind tibia with basal tubercle on outer surface. It therefore can be distinguished from known pupae of the following panurgine genera, which do not possess one, two, or all three of these characteristics: Perdita, Callonychium, Heterosarus, Melitturga, Anthemurgus, Protandrena, Panurginus, Psaenythia, and Liphantlms. These same features will distinguish it from some but not all sub- genera of Calliopsis sensu lato. The irregular surface of the upper part of the pupal compound eye of A. prinii as seen in frontal view (Fig. 12) (contrasting with an evenly curved surface) has not been reported for any other panurgine. Description.—Length 6.9-7.4 mm; body without setae. Head.—Scape unmodified except for small apical pro- jection (tubercle) next to compound eye; pedicel with small tubercle next to eye. Mandible without tubercle, but ma- lar area with small tubercle immediately below compound eye (no doubt homologous with tubercle incorrectly called "genal tubercle" in Neff and Rozen, 1995: fig. 21). Vertex with conspicuous rounded median tubercle on each side (see Remarks for explanation of terminology) near lateral ocellus; median ocellar tubercle virtually absent; lateral vertical tubercle and upper frontal tubercle low, scarcely noticeable, verruculose. Compound eye with upper sur- face slightly irregular as seen in frontal view (Fig. 12). Mesosoma.—Lateral angles of pronotum scarcely evi- dent; lateral lobes of pronotum pronounced, tuberculate, best viewed dorsally (Fig. 10); mesoscutum with pair of moderate-sized, sharply erect, paramedian tubercles; mesoscutellum with pair of large ttibercles; axilla produced as low mound; metanotum strongly produced medially; mesepisternum without tubercle. Tegula with low, distinct tubercle. Front and hind wings each with basal swelling, best seen in dorsal view (Fig. 10). Each coxa with small, acute, ventral tubercle; each trochanter with moderate- sized, acute, apical tubercle; anterior femur somewhat swollen ventrally at base, mid- and hind femora less swol- len ventrally at base; fore- and hind tibiae, but not mid tibia, with outer apical projection; hind tibia with distinct acute basal tubercle on outer surface. Metasoma.—Terga 1-VI with transverse row of acute, moderate-sized tubercles, most of which are apically 66 The University of Kansas Natural History Museum Special Publication No. 24 PAIRED MEDIAN VERTICAL TUBERCLE LATERAL LOBE OF PRONOTUM-, MESOSCUTAL TUBERCLE TEGULAR TUBERCLE AXILLA MESOSCUTAL TUBERCLE MESOSCUTELLAR TUBERCLE — SWELLING OF METANOTUM LATERAL TUBERCLE VERTICAL TUBERCLE HIND TIBIA IRREGULARITY PAIRED MEDIAN VERTICAL TUBERCLES COMPOUND EYE- TEQULAR TUBERCLE MESOSCUTELLAR TUBERCLE APICAL PROJECTION 11 TERMINAL SPINE Figs. 10-12. Pupa of Acninptopoeuin prinii. 10-Anterior part of body, dorsal view. 11-Entire body, lateral view. 12-Head (partial), enlarged, frontal view. Scale (=1 mm) refers to Figs. 10 and 11. pointed (see enlargement, Fig. 11). Sterna without tu- bercles. Terminal spine, short, apically rounded. Material Studied.—Two male pupae, Belo Horizonte, UFMG, Minas Gerais, Brazil, 1-29-1997 (D. Yanega). Remarks.—Tubercles on the vertex and upper frons are a common, but not universal, feature among the Panurginae. The position of these tubercles appears to be constant, suggesting that, where present, they can be ho- mologized from one genus to the next even though the extent of their expression varies. Names applied to them were adopted from Neff and Rozen (1995: fig. 20). taxonomic and phylogenetic interpretations Information presented here, as well as that provided by Claude-Joseph (1926), indicates that Acninptopoeuin is a shallow-nesting genus, examined species of which exca- vate meandering tunnels and laterals that lead to usually single, subhorizontal cells. Entrances to active nests are kept clogged and sometimes are completely hidden by loose surface soil. Females waterproof most of the inner surface of brood cells, presumably with a secretion. Provi- sions are shaped into a sphere that rests on the cell floor, and an arched egg is deposited on the top of the sphere. Laterals are soil-filled after cell closure. Young larvae re- main stationary on their provisions while they feed, and later reorient to rest on their dorsa. At the end of feeding. larvae deposit feces on the rear ceiling of the cell. In all of these respects, the nesting biology of the genus is similar to what is known about the genus Calliopsis sensu lato (Danforth, 1990; Rozen, 1958, 1963, 1967 [and references therein, p. 4], 1970; Rust, 1988; Shinn, 1967). The only published information on nesting biology about other genera of the Calliopsini was given by Claude- Joseph (1926) for Spinoliella. This accoimt is brief, and details given do not present features that distinguish the genus from either Calliopsis or Acamptopoeum on the basis of nesting biology. Hence, additional studies of Spinoliella are needed as are published investigations on the nesting biology of Arhysosage and Callonychium. After compara- tive data are available, thev can be interpreted phyloge- netically. Aside from aspects of nesting biology, mating behav- ior for one species of Acamptopoeum is known. Flying in copula, observed in A. subnietallicum, has also been noted for some subgenera of Calliopsis as well as for some spe- cies of Callonychium, Arhysosage, and Perditini (Rozen, 1958, 1963, 1970, and unpublished observations; Shinn, 1967). Further data regarding mating behavior should be gath- ered for analysis. Mature larvae oi Acamptopoeum obviously share many characters with those of Calliopsis (see McGinley, 1989, for references) but apparently can be distinguished on the basis of the large prothoracic tubercles, apices of which are di- Bee Genus Acamptopoeum 67 rected forward and laterad (Figs. 4-6), and on the basis of the faint pigmentation of the integument of all thoracic dorsal tubercles and those of the first abdominal segment. The more-or-less transverse shape of the dorsal tubercles toward the posterior part of the abdomen may also be dis- tinctive. Mature larvae of only Spinoliella of the other calliopsine genera have been illustrated and described (Claude-Joseph, 1926). However, J.G.R. has collected mature larvae of Spinoliella, Arhysosage, and Calloin/chium although they have yet to be studied. Larvae of these three genera have highly modified dorsal prothoracic tubercles and some- what modified mesothoracic tubercles, and the other body tubercles are conical, slender, and tapering (as suggested by Claude-Joseph's illustration (1926: fig. 59). Hence, lar- vae of these genera are quite unlike those of both Acaiiiptopoeiun and Cnlliopsis. A phylogenetic analysis of all genera of the Calliopsini based on larval characters should be revealing, after available specimens are stud- ied. LITERATURE CITED Alexander, B. A., and C. D. Michener. 1995. Phylogenetic studies of the families of short-tongued bees (Hymenoptera: Apoidea). Univer- sity of Kansas Science Bulletin 55: 377-424. Claude-Joseph, F. 1926. Recherches biologiques sur les Hymenopteres du Chili (MelHferes). Annates des Sciences Naturelles, Zoologie, Series 10, 9: 113-268. lin French] Danforth, B. N. 1990. Provisioning behavior and the estimate of invest- ment ratios in a solitary bee, Calliopsif {Hypoiitacrotern) pcrsimilis (Cockerell) (Hymenoptera: Andrenidae). Behavioral Ecology and Sociobiology 27: 159-168. McGinley, R. J. 1981. Systematics of the Colletidae based on mature lar- vae with phenetic analysis of apoid lar\ae (Hymenoptera: Apoidea). Uruversity of California Publications in Entomology 91; 307 pp. McGinley, R. J. 1989. A catalog and review of immature Apoidea (Hy- menoptera). Smithsonian Contributions to Zoology 494: 24 pp. Michener, C. D. 1953. Comparative morphology and systematic studies of bee larvae with a key to the families of hymenopterous larvae. University of Kansas Science Bulletin 15: 987-1102. Neff, J. L., and J. G. Rozen, Jr. 1995. Foraging and nesting biology of the bee Anthemurgus pasfiflorae (Hymenoptera: Apoidea), descriptions of its immature stages, and observations on its floral host (Passifloraceae). American Museum Novitates 3138: 19 pp. Rozen, J. G., Jr. 1958. Monographic study of the genus Nomadopsis (Hy- menoptera: Andrenidae). University of California Publications in Entomology 15: 202 pp. Rozen, J. G., Jr. 1963. Notes on the biology of Nonindopfts, with descrip- tions of four new species (Apoidea, Andrenidae). American Mu- seum Novitates 2142: 17 pp, Rozen, J. G., Jr. 1967. Review of the biology of panurgine bees, with ob- ser\'ations on North American forms (Hymenoptera, Andrenidae). American Museum Novitates 2297: 44 pp. Rozen, J. G., Jr. 1970. Biology and immature stages of the panurgine bee genera Hypamacrotera and Psaenythia (Hymenoptera, Apoidea). American Museum Novitates 2416: 16 pp. Rozen, J. G., Jr. 1993. Phylogenetic relationships of Euherbstia with other short-tongued bees (Hymenoptera: Apoidea). American Museum Novitates 3060: 17 pp. ' Rozen, J. G., Jr. 1994. Biologies of the bee genera Anci/landrena (Andrenidae: Andreninae) and HexLyeolus ( Apidae: Nomadinae) and phylogenetic relationships oi Aucylmuirena based on its mature larva (Hymenoptera: Apoidea). American Museum Novitates 3108: 19 pp. Rozen, J. G., Jr, and C. D. Michener. 1988. Nests and immature stages of the bee Paratetrapedia su'aiitsoiinc (Hymenoptera: Anthophoridae). American Museum Novitates 2909: 13 pp. Rozen, J. G., Jr., and L. Ruz. 1995. South American panurgine bees (Andrenidae: Panurginae), Part II. Adults, immature stages, and biology of Neffnpis longilingua, a new genus and species with an elon- gate glossa. American Museum Novitates 3136: 15 pp. Rust, R. W. 1988. Biology of Nomadopsis larreae (Hymenoptera: Andrenidae), with an analysis of yearly appearance. Annals of the Entomological Society of America 81: 99-104. Ruz, L. 1986. Classification and phylogenetic relationships of the panurgine bees (Hymenoptera-Andrenidae). Doctoral Dissertation, University of Kansas, 312 pp. Ruz, L. 1991. Classification and phylogenetic relationships of the panurgine bees: the Calliopsini and allies (Hymenoptera: Andrenidae). University of Kansas Science Bulletin 54: 209-256. Shinn, A. F. 1967. A revision of the bee genus Calliopsis and the biology and ecology of C. nndrcniforniis (Hymenoptera, Andrenidae). Uni- versity of Kansas Science Bulletin 66: 753-936. Byers, G. W., R. H. Hagett, and R. W. Brooks (t'rfsJ, Entomological Contributions in Memory of Byron A. Alexander. University of Kansas Natural Histon/ Museum Special Publication 24. (1999) Pp. 69-80 An Inventory Of Arthropod Fauna At Great Sand Dunes National Monument, Colorado By Michael J. Weissmann' and Boris C. Kondratieff' ABSTRACT Over 900 species of arthropods are listed from Great Sand Dunes National Monument, Alamosa and Saguache counties, Colorado. The list is compiled from specimens collected by more than 30 researchers over a period of six decades. The inventory includes four species that are appar- ently endemic to the Monument and surrounding habitat. Keywords: Arthropod; Biodiversity; Colorado; Great Sand Dunes National Monument. INTRODUCTION Byron A. Alexander was a seasonal park naturalist at Great Sand Dunes National Monument in the late 1970's. It was here that his interest in entomology was inspired through his contact with Howard Evans and his students who were studying the behavior of wasps in and around the dunes. The illustrations of plant and insect life accom- panying this inventory were created by Byron Alexander while he was working at the Monument, and are repro- duced here with permission from Great Sand Dunes Na- tional Monument. Great Sand Dunes National Monument is located in Alamosa and Saguache Counties in south central Colorado. The dunes, which cover 101 km- on the east side of the San Luis Valley, are pushed up against the Sangre de Cristo Mountains. Prevailing winds cross over the mountains to the northeast by way of two passes (Mosca Pass and Medano Pass) and deposit the sand load in dunes tower- ing to more than 200 m above the valley floor. Less than 35 cm of precipitation fall on the dunes each year, and the water quickly percolates down from the surface. Daytime summer surface temperatures have been recorded in ex- cess of 55 °C. At elevations above 2400 m, the Great Sand Dunes are also subject to cold winter temperatures and cool evening summer temperatures. Strong winds push sand across the surface and obscure most foot prints daily. The geology and hydrology of the dune mass is not fully understood and has been the subject of extensive research in recent years (for a review of the current knowledge of the geology, hydrology, biology, and anthropology of Great Sand Dunes National Monument, see Schenk, in press). Understanding the hydrology of the dunes has been cru- cial because of the recent threat of intense withdrawal of surface and ground water for agriculture and other con- sumptive uses (U.S. Bureau of Reclamation, 1987). Plant life on the main dune mass is sparse, consisting principally of scurfpea, Psoralidium lanceolatum (Pursh) Rydberg (Fig. 1), Indian ricegrass, Oryzopsis hymenoides (Roemer and Schultes) Ricker (Fig. 2), blowout grass, Redfieldia flexuosa (Thurber) Vassey (Fig. 3), skeletonweed, Lygodesmia jimcen (Pursh) Don (Fig. 4), and prairie sun- flower, Hcliaiitlius pctiolaris Nuttall (Fig. 5). Most of the plant life is clustered, forming vegetation "islands," mainly in the interdune areas. Despite its seemingly barren landscape. Great Sand Dunes National Monument supports an abundance of arthropods. Several researchers have studied arthropods at the Monument, but no work has been done to combine the efforts of these endeavors into one comprehensive spe- cies list. We initiated this study in August 1990 to create an inventory of the arthropods known from the Great Sand Dunes. The entomological collections of Great Sand Dunes National Monument (GRSA), the University of Colorado Museum in Boulder (UCM), and the C.P. Gillette Entomo- logical Museum at Colorado State University in Fort Collins (CSU) contain numerous specimens previously collected within and around the Monument. These speci- mens were located, recorded, and identified as specifically as possible by taxonomic specialists. This cumulative col- lection exists thanks to the efforts of a number of research- ers, including (specimen locations in parentheses): B. Rotger, 1942-1977 (UCM and published records in Rotger, 1944); T.P MasHn, 1949 (UCM); R.E. Gregg, 1949-1954 (UCM, published records in Gregg, 1963); H.G. Rodeck, 1954-1955 (UCM); FM. Brown, 1954-1981 (GRSA, UCM); H.E. and M.A. Evans, 1954-1983 (GRSA, CSU, and pub- lished records in Evans, 1966, and in Evans and O'Neill, 1988); J.C. Daniel et al, 1959 (GRSA); E.R. Tinkham, 1960 (published records in Tinkham, 1962); J. Brookhart, 1965 (UCM, published records in Brookhart, 1972); J. Stanatov, Department of Bioagricultural Sciences and Pest Management, Colorado State University, Fort Collins, CO 80523. 69 70 The University of Kansas Natural History Museum Special Publication No. 24 Figs. 1-6. 1-Scurfpea, Psornlidium Innccolalum (Pursh) Rydberg. 2- Indian ricegrass, Oryzopsis Inimenoidcs (Roemer and Schultes) Ricker. In- set; Indian ricegrass flower. 3-Blowout grass, Redfieldia flexiwsa (Thurber)Vassey. Inset: Blowout grass flower 4-Skeletonweed, Lygodesmia juncen (Pursh) Don. Inset: Skeletonweed petal and fruit. 5-Prairie sun- flower, HcUivitluis petiolaris Nuttall. Inset: Prairie sunflower seedhead and seed. 6-Rabbitbrush, Chn/sothnniuns iinu^cosus (Pallas) Britton. a: Rabbitbrush in bloom, b: Rabbitbrush in fruit. Inset: Rabbitbrush flower. 1971 (GRSA); P. Eades, 1974 (GRSA, CSU); C.A. Triplehorn, 1974 (published records in Triplehorn, 1964); K.M. O'Neill 1974-1980 (GRSA, CSU, and published records in O'Neill, 1983a and 1983b, and O'Neill and Evans, 1981); W. Rubink, 1975 (GRSA, CSU); D. Gwynne, 1976-1979 (GRSA, CSU, and published records in Gwynne, 1978, 1980, and 1981; Gwynne and Evans, 1975; Gwynne and Hosteller, 1978; Gwynne and O'Neill, 1980; and Ahlbrandt et al., 1978); S. Condie, 1977 (GRSA, CSU); G.H. Kemper, 1977-1980 (GRSA, UCM); D. Huffman, 1978 (GRSA); S. Vertrees, 1978 (GRSA); W. Soenger, 1978-1980 (GRSA, CSU); B.A. Alexander, 1978-1995 (GRSA, CSU); S. Wingate, 1980 (GRSA); J. Crock, 1981 (CSU); L.D. Zuckerman, 1981-1982 (GRSA), T.R Sluss, 1983 (GRSA and records in Sluss, 1986a and 1986b); M. Kippenhan, 1990 (CSU); W.Cranshaw, 1991 (CSU); D. & M. Leatherman, 1991 (CSU). Figs. 7-10. 7-Great Sand Dunes tiger beetle, Cicindcln thcalwa Rotger. a: Adult, b: Larva. 8-Giant sand treader cricket, Diulunibncnetes gignnteii? Tinkham. 9-Wolf spider, Geoh/cosa rafaelana Chamberlain. 10- Grasshopper, Triinerotropis sp. Inset: in flight. Several species have been added to the inventory list based on collections from August 1990 through Septem- ber 1995. Insects were sampled from several habitats in Great Sand Dunes National Monument. These included the open dune mass in several different areas, at different times of year, and during different times of the day and night. Many interdune areas, dominated by grass and scurfpea plants, were also sampled. Off the dunes, rabbit- brush, Chn/sothammis museosus (Pallas) Britton (Fig. 6), and other scrub zones were sampled, as was the area around the campground and in the pinon/juniper zones to the east (Pinus ediilis Engelmann and Juniperus commiDiis Linnaeus). Some sampling was also done at Denton Spring southeast of the dune mass and at Sand Creek on the west side of the Monument. Collection techniques included sweep netting, black- light traps, and mercury-vapor lamps. All specimens col- lected or examined were recorded in two notebooks of data sheets, copies of which have been deposited at the mu- seum at GRSA. All specimens collected during 1990-1995 have been catalogued into the National Park Service Au- tomated National Catalog System (ANCS) and deposited at GRSA, CSU, or UCM. Great Sand Dunes Arthropod Inventory 71 The updated species list which follows is intended to provide baseline information on species diversity of the region. More than 900 species of arthropods have been re- corded to date from the Monument. Among the new find- ings are four undescribed species. Two of these are anthicid beetles (Aiubli/derus n.spp.), which were found walking on the dunes at mid-day, with surface temperatures in excess of 40 °C. There is also a new robber fly (ProctacaiitJnis n.sp.), among the largest insects in the Monument and distinctly different from two closely related species also recorded from the area. An undescribed species of geometrid moth (Pwchoerodcs n.sp.) has been found, also recorded from other places in the western U.S. but not yet formally de- scribed in the literature. Unexpected records include a species of noctuid moth, Schinia avemensis (Dyar), previously known only from Manitoba, Canada, the caterpillar of which probably feeds on the flowerheads of the prairie sunflower, and a species of curtonotid fly, Curtonotum helvum (Loew), previously unknown west of the Mississippi River, collected at lights and on the Cleome flowers at Sand Creek. The current species list represents only a small per- centage, perhaps as little as 25%, of the total number of arthropods which occur within the Monument boundaries. Other sampling techniques, such as pit traps. Malaise traps, and Berlese funnels were not employed and would likely vield additional species. Additionally, historically there has been some bias in the collecting at the Monument. The large numbers of sphecid wasps and noctuid moths recorded are partially due to the habitat type, and partially due to collecting bias of previous researchers—Howard E. Evans, an authority on the behavior of sphecid wasps, and the late F. Martin Brown, a lepidopterist with a fondness for light trapping. Many of the smaller insects were likely overlooked, as were those with very short life cycles or that emerge for only a short time during the season. Other major collections which a,re known to have speci- mens from the Monument include the American Museum of Natural History in New York, and the U.S. National Museum of Natural History (Smithsonian) in Washington, D.C. Both collections have moths (Paul A. Opler and F. Martin Brown, personal communications) which were not examined during this study. Of special interest are the species that are apparently endemic to Great Sand Dunes National Monument and the San Luis Valley. This includes the darkling beetle ("cir- cus beetle"), Elcodes liirtipcuuis Triplehorn; the Great Sand Dunes tiger beetle, Cicindela thentina Rotger (Fig. 7); and the two new species of anthicid beetles, Ambh/dcrus spp., which are as yet undescribed and have not been found elsewhere to date. The giant sand treader cricket, Dniliiuibaenetes gigniiteiis Tinkham (Fig. 8), once believed to occur only at the Monument, has since been found in northern New Mexico and southern Utah (Theodore J. Cohn, personal communication). All of the above species were seen regularly during the study and are therefore not considered locally rare. They are threatened only to the extent that they have a limited distribution. Their survival is probably dependent upon the survival of the dunes habitat itself. These species are good candidates for population monitoring to deter- mine human impact on the dunes, or to monitor overuse of the resource. A drop in their population levels might indicate overuse of the Monument. Additionally, popula- tions of the species in relatively isolated areas should be compared to populations in heavily used areas of the Monument. Further study is necessary to describe the natu- ral history of these species. The sand dunes at the Monument are a unique habi- tat. The dunes appear to be barren but actually are the habitat for a large number of resident insects. Almost one- fourth of the species recorded are considered residents of the sand and the grass/pea interdune areas. The others li\'e in the surrounding vegetated areas, and may come to the sand for feeding or are blown onto the sand by the strong winds. With daytime summer surface temperatures surpassing 55 °C, the dunes are very inhospitable to in- sects that are not adapted to survive such temperatures. Insects blown onto the sand from other places usually die and are eaten by the scavengers. Many of the insects which are resident on the dunes, including the darkling beetles and sand treader crickets, are mainly scavengers and eat the dead and dying insects blown in from outside. There are also many sand mites which feed on the corpses. Preda- tors such as the resident wolf spiders (Geolycosa rafaelana Chamberlain) (Fig. 9) and the tiger beetles also take ad- vantage of the weakened insects, the spiders hunting by night and the tiger beetles by day. Primary production (from plants) is limited to the grass/scurfpea interdune areas. Herbivores, such as the plant bugs {Li/giis spp.) and grasshoppers (Trimerotropis spp.) (Fig. 10), are generally restricted to these areas. Some additional information about the arthropods of the Monument is found in Weissmann etal. (1993). Acknowledgments This research was made possible through the assis- tance of numerous individuals and institutions. Co-inves- tigators on the research were Linda P. Clement (National Park Service) and Karolyn Darrow (University of Colo- rado Museum). Field assistance was provided by Judith Schaeffer, Dr. Edwin Licht, Dr. Boyce Drummond, and Rachel Williams. The staff of Great Sand Dunes National Monument, under the leadership of Superintendent Bill Wellman, provided valuable logistical support and lodg- 72 The University of Kansas Natural History Museum Special Publication No. 24 ing throughout this study. This research was funded in part by the Rocky Mountain Region of the National Park Ser- vice and a 1991 Colorado Natural History Small Grant from the Colorado Natural Areas Program of the Colorado Di- vision of Parks and Outdoor Recreation (Project Number GRSA-R91-0145; Cooperative Agreement Number CA 1200-9-0004). Additional support was provided by a grant from the Orthopterists' Society Research Fund in 1994. The following specialists provided assistance with arthropod identifications: R.W. Brooks, J.M. Carpenter, S.M. Clark, TJ. Cohn, R. Davidson, B. Drummond, S. Elias, H.E. Evans, M.A. Evans, K. Hagen, D.J. Heffern, R.L. Hoffman, EX. Hovore, M. Kippenhan, P.K. Lago, U.N. Lanham, D. Larson, E.L. Licht, W.N. Mathis, C.R. Nelson, G.H. Nelson, M.W. Nielson, PA. Opler, R.S. Peigler, J.T Polhemus, M.D. Schwartz, D.B. Thomas, C.A. Triplehom, D. Weissman, J.L. Welch, and F. Werner. SPECIES LIST2'3 HEXAPODA MICROCORYPHIA Machilidae (unidentified) ephemeroptera Baetidae B«t'fis trkaudatus Dodds Heptageniidae Cinygmula sp. Epeorus longimanus (Eaton) LeptopWebiidae Paraleptophlebia debilis (Walker) Ephemerellidae Ephemeretta infrequens McDunnough ODONATA Aeshnidae Aeshna constricta Say Aeshna palmata Hagen Libellulidae Sympetriim corruptiim (Hagen) Sympetrum occidentale Bartenev Lestidae testes congener Hagen Lesles dryas Kirby testes tmguicidatus Hagen PHASMIDA Heteronemiidae Parabacilhis coloradus (Scudder) ORTHOPTERA Acrididae Acrolophitus hirtipes (Say) Amphitornus coloradus (Thomas) Arphia pseudonietana (Thomas) Camnula peUucida Scudder Circotettix rabuta rabula Rehn and Hebard CordUlacris occipitalis (Thomas) Cratypedes neglectus (Thomas) Dissosteira Carolina (Linnaeus) Hadrotettix trifasciatus (Say) Hesperotettix speciosus (Scudder) Melanophis bowditchi Scudder Melanoplus bowditchi canus Hebard Melanophis femur-rubrian (DeGeer) Melanophis packardii packardii Scudder Melanoplus spp. Mestobregma plattei corrugata (Scudder) Psoloessa dehcatula (Scudder) Spharagemon campestris (McNeill) Spharagemon collare (Scudder) Trimerotropis agrestis McNeill Trimerotropis cincta (Thomas) Trimerotropis fratercula McNeill Trimerotropis marttima (Harris) Trimerotropis verruculatus suffusa Scudder Xanthippus montanus (Thomas) Gryllidae Oecanthus quadripunctatus BeutenmuUer Rhaphidophoridae Ceuthophilus utahensis Thomas Ceuthophilus spp. Daihinibaenetes giganteus Tinkham Stenopelmatidae Stenopelmatus sp. Tettigoniidae Anabrus sirnpitex Haldeman Conocephalus fasciatus (DeGeer) MANTODEA Mantidae Yersiniops solitarium (Scudder) PLECOPTERA Pteronarcyidae Pteronarcella badia (Hagen) Nemouridae Malenka coloradensis (Banks) Perlodidae Isoperla quinquepunctata Banks Chloroperlodidae Plumiperla diversa (Prison) Siveltsa hvnba (Banks) HEMIPTERA Pentatomidae Banasa sordida (Uhler) Chtorochroa granulosa Uhler Chlorochroa sayi (Stal) Chlorochroa uhlcri (Stal) Murgantia histrionica (Hahn) Perillus exaptus (Say) Scutelleridae Amaurochrous cinctipes (Say) (unidentified) Cydnidae Microporus obliquus Uhler Coreidae ^ Apparently endemic species = *** ' Species names in brackets [ 1 are questionable records (no specimen available) or recorded from just outside the Monument boundary but not specifically within the Monument. Chelinidea vittiger Uhler Leptoglossus clypealis Heidemann Leptoglossus occidentalis Heidemann Alydidae Ali/dus ptuto Uhler Ah/dus spp. Megalotomus quiiujuespinosus (Say) Rhopalidae Harmostes sp, Liorhyssus sp. (unidentified) Lygaeidae U/gaeus reclivatus Say Neocoryphus lateralis (Say) Nysius sp. Peritrechus fraternus Uhler ? Xyonyshis c.f. californiciis Stal Great Sand Dunes Arthropod Inventory 73 Phymatidae Phymata sp. Reduviidae Arilus cristatus Linnaeus Fitchia aptera Stal Sinea diademci (Fabricius) Nabidae Nahis sp. Miridae Atractotomus striacolor (Knight) Hadronema picta Uhler ? Ilnacora chlons (Uhler) Litomiris debilis (Uhler) Lcpidea sp. Lygidea rubecula (Uhler) Lygus atriflavus Knight Lygus elisus VanDuzee Lygus lineolaris (Pahsot) Lygus nubilaius Knight Lygus shulii Knight Orthotylus vindis VanDuzee Phytocoris conudus Knight Phylocoris consors VanDuzee Phytocoris heidenuvim Reuter Phytocoris lnop^s Uhler Phytocoris simutatus Knight Phytocoris imlidus Reuter Polymerus bnlli Knight Stenodema pitosipes Kelton Tuyiiocoris agilis (Uhler) ? Gerridae (unidentified) Corixidae Ceiiocorixa sp. Craptocorixn ahdomiimhs (Say) Sigarn alternatn (Say) Notonectidae Notonecta ktrbyi Hungerford Notonecta undulata Say HOMOPTERA Aphididae (unidentified) Delphacidae (unidentified) Membracidae Publilia modesta Uhler (unidentified) Cicadellidae Aceratogallia arida Oman Aceratogallia uhleri (VanDuzee) Aceratogallia sp. Athysanella occideutalis Baker Idiocerus snowi Gillette and Baker Kybos sp. Laei'icephalus parbulus (Gillette) Oncometopia lateralis Fabricius (unidentified) Dictyopharidae Scolops sp. NEUROPTERA Myrmeleontidae Brachynemurus nigrilabris Hagen Brachynemurus peregrnius (Hagen) Brachynemurus sackeui (Hagen) Myrmeleon invnaculatus DeGeer Chrysopidae Chrysopa coloradensis Banks Chrysopa oculata Say Chrysoperla sp. Eremochrysa sp. Hemerobiidae Hemerobius sp. Micromus sp. Raphidiidae Rjjf>hidia sp. COLEOPTERA Carabidae Agotium placidum (Say) Amara imputicticollis Say Bembidion graplucum Casey Calosoma obsoletum Say Elaphrus lecontei Crotch Euryderus grossus (Say) Geopinus incrassatus (Dejean) Harpalus amputatus Say Harpalus erraticus Say Harpmlus paratus Casey Harpalus spp. Omophron tessetlatum Say Cicindelidae Cicindela formosa Say Cicindela hirticollis shelfordi Graves Cicindela leugi Horn Cicindela oregona guttifera LeConte Cicindela punctulata Ohvier Cicindela repanda Dejean ***Cicindela theatina Rotger Dytiscidae Agabus lutosus LeConte Colymbetes exaratus incognitus LeConte Dytiscus sp. Graplwderes occidentalis Horn Hydroporus sp. (-oilis complex) Hygrotus impressopunctatus (Schaller) Hygrotus infuscatus (Sharp) Hygrotus rnasculinus (Crotch) Hygrotus tumuiwentris (Fall) Hygrotus sp. Ilybius fraterculus LeConte Laccoplnlus maculosus decipieiis LeConte Lwdessus obscurellus (LeConte) Stictotarsus striatellus (LeConte) Rhanlus sp. Gyrinidae Gyrinus sp. Hydrophilidae Cercyon sp. Hydrochus sp. Spltaeridium scarabaeoides (Linnaeus) Tropisternus sp. Histeridae Hypocaccus n.sp. Saprinus lugens Erichson Spilodiscus sp. (unidentified) Staphylinidae Aleochara bimaculata (Gravenhorst) Creophilus maxillosus Linnaeus Silphidae Heterosilplm ramosa (Say) Nicrophorus guttula Motschulsky Nicrophorus marginatus Fabricius Lucanidae Pseudolucanus mazama LeConte Scarabaeidae Aphodnis sp. Dtplotaxis helfragei Fall Eucanthus impressus Howden Glaresis ecostata Fall Ligyrus gibbosus (DeGeer) Phyllophaga fimbripes (LeConte) Phyllopihaga sp. Polyphylla decimtineata (Say) 74 The University of Kansas Natural History Museum Special Publication No. 24 Serial alternata LeConte Sericn bruncri Dawson Serktt pwcula Casey Serial sp. (prob. anthracima LeConte) Trichictiiuis assiinilis (Kirby) Tro.v scviorne LeConte Drvopidae Heliclnis striatiis LeConte Elnnidae Hetcrlimnius corpulcnia (LeConte) Optiosennis divergeiis (LeConte) Buprestidae Dicercn tenebrica (Kirby) Melaiwphila gentilis LeConte Lampyridae (unidentified) Cantharidae Cliaulwguatlius scutcUaris LeConte Dermestidae Trcgodcrnm sp. (prob. aiigustuin (Solier)) (unidentified) Cleridae Enoderus moestus (Klug) Pln/llohaenus sp. Tricliodes ornatiis ornatiia Say Melyridae CoHops bipiinctatus Say (unidentified) Nitidulidae Carpiopliilus sp. Phalacridae Phnlacrus sp. Coccinellidae Anatii hxontei Casey Coccinelh monticolci Mulsant CoccincHa fepyteinpunctntn Linnaeus Cocciiuila transversoguttntn riciuudsoni Brown Hippodnmia casei/i Johnson Hipp'odainici amvergens Guerin-Meneville Hippodntniii quinqucsigimta (Kirby) Hippodamin tredecimpitnctnta tribalis (Say) Myzin interrupta (Casey) Tenebrionidae Botbrotes plumbeus (LeConte) Coiiiontis obcsa LeConte Eleodes nciiticaudus LeConte Eleodes brunnipes Casev Eleodes caudiferus LeConte Eleodes extricatus (Say) ***Eleodes birtipeunis Triplehorn Eleodes longieolUs LeConte Eleodes obscurus dispvrsus LeConte Eleodes pimelioides Mannerheim Eleodes snoui Blaisdell Eleodes spoiisus LeConte Eleodes trieostatiis (Say) Embiiphion eontiisiim LeConte Embaphion glnbrum Blaisdell Eusatlits retieulntus Say Helops sp. Telnbif nspvrn Casey Mordellidae (unidentified) Meloidae Epicnuta sp. Lytta inittalli (Say) Nenwgnntlui sp. Pyrotii bilineata Horn Tricrnnia stnnsbunti Haldeman (unidentified) Anthicidae ***Amblyderus n.sp.l ***Amblyderus n.sp.l AnthicKS liitutentHs Casey Anthiciis sp. Notoxus sp. Cerambycidae Anoplodeni canadensis (Olivier) Arhopalus rusticus tnontanus (LeConte) Batyle ignieoUis (Say) Balyle sutiirnlis penrsalli (Bland) Cortodera longicoruis (Kirby) Crossidius coralinus jocosus (Horn) Crossidius hirtipes unckhami Casey Crossidius pulcbellus (LeConte) Grammoptera sidmrgentata (Kirby) MoneUemii npprcssum LeConte Monocliamus clnmator (LeConte) Monocbamus sciitellatus (Say) Pachyta lamed liturnta Kirby Prionus califor)iicus Motschulsky Prionus emarginatus Say Prionus integer LeConte Typocerus baltentus Horn Xylotrechus undulatus (Say) Chrysomelidae Altica bimarginntn Say Altica sp. Cryptocepbnlus spp. Disonycha alternata (lUiger) Disonyclui tatifrons Schaeffer Disonycha sp. Galeruca costatissima Blake ? Macrohaltica sp. Microhopala excavata cyanea (Say) Pacln/brachis sp. Phyllotreta spp. Saxinis saucia LeConte Tricholochamaea sp, ? Trirhabdn Iczvisii Crotch Trirhabda nitidicollis LeConte Zygograniinn conjuncta Rogers Curculionidae Epimecbus sp. (unidentified) TRICHOPTERA Hydropsychidae Arctopsyche grandis Banks Chcumatopsycbe sp. Glossosomatidae Agapetus boulderensis Milne Glossosoma sp. Hydroptilidae Hydroptila sp. Rhyacophilidae Rhyacoplnln coloradensis Banks Brachvcentridae Bracbycenlrus aniericanits (Banks) Limnephilidae Linmepbilus sp. LEPIDOPTERA Oecophoridae Agoiwpterix sp. ? Gelechiidae (unidentified) Tortricidae Acleris sp. ? Argyrotaenia coloradana (Fernald) Doritbia semicircidana (Fernald) Eucosma crambitana Walsingham Eucosma fernaldana (Grote) Eucosma nr. ridingsana (Robinson) Eucosma sp. Syndemis sp. [or Pandemis sp.] Xenotenvia pallorann (Robinson) Great Sand Dunes Arthropod Inventory 75 Hesperiidae En/nnis iceliis (Scudder and Burgess) Hesperiti comma Colorado (Scudder) ? Hesperia comma ocltracea Lindsey Hesperia nevada (Scudder) Hesperia uncas Edwards Oarisma garita (Reakirt) Pyrgus communis (Grote) Pyrgus xautlius Edwards Yvretta rlicsits (Edwards) Papilionidae Papilio rutulus (Linnaeus) Parnnssiiis phoehus pseudorotgcri Edwards Pieridae Artogeia rapae (Linnaeus) Colias eun/themc Boisduval Eurema nicippe (Cramer) Neophasia menapia (Felder and Felder) Pontia beckerii (Edwards) Pontia protodice (Boisduval and LeConte) Lycaenidae Agriades franklinii rusticus (Edwards) Callophn/s apama iwmoperplcxa Barnes and Benjamin Cekstrina ladon cinerca (Edwards) Chakeria rubida sirius (Edwards) Eupjhihtes ritn ccloradensis (Mattoni) Hemiiirgus isola alec (Edwards) karicia acmon lutzi dos Passos karicia icarioides lycea (Edwards) bicisalia niphon iiiphoti (Hiibner) Leptotes marina (Reakirt) Lycaeides melissa melissa (Edwards) Mitoura siva siva (Edwards) Satyrinm behrii crossi (Field) Strynwn melinus franki Field Tharsaka arota scIicUbadii Tilden Riodinidae Apodemia nwrmo mornio Felder and Felder Nymphalidae Basikrchia weidemei/ert n'eideineycri (Edwards) Clunidn/as acastus (Edwards) Euptoieta claudia (Cramer) Nymphalis antiopa (Linnaeus) Phyciodes campestris camillus Edwards Pokdn/as arachne arachne (Edwards) Polygonia hyhs (Edwards) Speyeria aphrodite ethne (Hemming) [Vanessa atahnta atalanta (Linnaeus)] Vanessa cardui (Linnaeus) Satyridae Cem/onis meadii alamosa Emmel and Emmel Cercyonis oetiis charon (Edwards) Coenonympha ochracea Edwards CyUofisis pertepida dorolhea (Nabokov) Neominois ridingsii ridingsii (Edwards) Oeneis chn/xus chn/xus (Doubledav and Hewitson) Danaidae Danaus gilipptis Cramer Danaiis pkxippus (Linnaeus) Pvralidae (unidentified) Geometridae Caripeta aequnliaria Grote Caripeta interalbkans Warren Cheteoscelis bistriarta (Packard) Chlorosea nevadaria Packard Enypia griseata Grossbeck Eriplatymetra coloradaria (Grote and Robinson) Eupithecia antkaria Walker Hydriomena tnorosata Barnes and McDunnough Hydriomena perfracta centrnlis McDunnough Hydriomena simikris Hulst Iridopsis enmscidata (Dvar) Itame bitactata (Walker) Itame decorata (Hulst) 7fn»it' flnvicnrin (Packard) Metanema inatomaria Guenee Perizoma cuslodialn (Guenee) Pero behrensaria (Packard) Plataca triUnearia (Packard) Prionomelin spmlodea (Hulst) Prochoerodes forficaria (Guenee) Prochoerodes truxaliata (Guenee) Prochoerodes n.sp. [nr. amplicinerarin (Pearsall)] Scelidacantha triseriata (Packard) Semiotliisa niibicuktn (Packard) Semiothisa stdvnininta (Packard) Semiothisa sp. Stamnoctenis morrisata (Hulst) Stamnodes formosata (Strecker) Syncldora aerata liqnoraria Guenee Lasiocampidae Epicnaptera americana (Harris) Ghveria arizonensis Packard Makcosoma californka Packard Saturniidae Antheraea poh/phemus (Cramer) Coloradia doris Barnes HemUeiica nuttalli (Strecker) Hyalophora ghnvri (Strecker) Sphingidae Heniaris senta (Strecker) Hi//t's linenta (Fabricius) Smeriulhiis cerisyi Kirby Sphinx dollii Neumoegen Arctiidae Eilema bicokr Grote Lophocampa ingcns (Edwards) Lophocampa macidata Harris Turuptiana pernmculnta (Packard) Lymantriidae Dasychira sp. ? Noctuidae Abagrotis discoidalis (Grote) Abagrotis reedi Buckett Agroperina conradi (Grote) Agrotis ipisikn (Hufnagel) Anathix aggressa (Smith) Andropolia diversiWieata (Grote) Apamea occidens (Gtote) Aseptis fumosa (Grote) Bracln/hmia popidi (Strecker) Catocala grotinna Bailey Catocak hermia Edwards Copablepharon absiditm (Harvey) Copablepharon grande (Strecker) Copablepharon sp. Crassivesica bocha (Morrison) Crymodes devastator (Brace) Cucidlia sp. Drasteria mirifica klotsi Richards [Enargia decolor (Walker)] Enargia infnmata (Grote) Enrols prncfixa (Morrison) Euxoa albipcnnis (Grote) Euxoa anrulenta (Smith) Euxoa auxiliaris (Grote) Euxon brcvipennis Smith Euxoa cicalricosa (Grote and Robinson) Euxoa divergens (Walker) Euxoa messoria (Harris) Euxon uuinaltonis (Grote) 76 The University of Kansas Natural History Museum Special Publication No. 24 EuxM moereus (Grote) ? Euxoa obt'liscoiiies (Guenee) Eiixoa ohlongiftigmn (Smith) ? lor olivnlis (Grote)] Euxoa qitadridcntatn (Grote and Robinson) Euxoa ridingsiana (Grote) Euxoa scandenf (Riley) Euxoa stigmatalts (Smith) Euxoa tessellata (Harris) ? Euxoa sp. Helicoverpa zea (Boddie) Hemieuxoa rudcns (Harvey) Homohadena fifia Dvar LKauobia Ulacina (Harvey) Lacuiipolia uaevia (Smith) Lacmipolia oUvacea mcgnrena (Smith) LacinipoUa umbrosa (Smith) LacinipoUa vktna (Grote) LacmipoUa sp. Litholomia napaea (Morrison) Oligia sp. Oncocnemis haltcata Smith Oncocnemis Colorado Smith Oncocneinif lia\/esi Grote Oncocnemis homogena Grote Oncocnemis tricolor Smith Peridroma saucia (Hubner) Platyperigea camina Smith Platyperigea extima (Walker) Platyperigea meralis (Morrison) Ponometta sutrix (Grote) [Protagrotis niveiz'enosa (Grote)] Protogygta sp. Protortlwdes utaliensis (Smith) Protortliodes sp. (prob. al!(Ssp.(Eichoff 1866) Bombus impatiens Smith (Plath 1922) Bombus ruderarius Miiller (.=derhamellus) (Bold 1856) Bomhus sp. (Seidlitz 1869-70) Bomhus sp. (Donisthrope 1906) Bomhus atratus Franklin (Roubik & Wheeler 1982) Bombus ephipiatus Say (Chavarria :994b) Bombus puUntus Franklin (Chavarria 1994a) Bombus (Pyrohomhus) rohustus Smith (R. Brooks & C. Marshall, pers. obs.). Crematognster subdentata Mayr (Reitter 1889) Vespa crahro Linnaeus (Tuck 1897) Anthophora sp. (Falcoz 1929) Lasius fuUginosus (Latreille) (Palm 1953 in Horion 1960) Lasius brunneus Latreille (Koch 1989) Bombus pennsylvanicus (DeGeer) {=americanorunt) (Prison 1926) Bombus ephipiatus Say (Grouvelle 1911b) Erichson) Bombus agrorum (Buckle 1900) Bombus horatorum (Dollman 1912) Bombus terrestris (Buckle 1900) Bombiis sp. (Tuck 1896) Vespula vulgaris (Tuck 1897) Formica exsecta Nvlander (Falcoz 1929; Kieseritzkv & Reichardt 1936) Osmia sp. (Falcoz 1929) Table 1 (Continued) Cryptophagid Symbiosis Table 1 (Continued) 89 Taxon Interaction N=in nest/P=phoresy Host & Reference C. fiifcicoruis Sturm C. intermedins Bruce C. lahilus Erichson N N N N N N C. lycciperdi (Scopoli) N N Lasius hrunncus (Koch 1989) Lasius fiiligmosus (Koch 1989) Wspiisp, (Koch 1989) Lasius bnmueus (Horion 1960) Lasius brunncus (Koch 1989) Myrmica ruginotiis Nylander (Koch 1989) Bomhus lardarius (Horion 1960) Tree-living wasps (Coombs & Woodroffe 1955) C. pallidus Sturm N 90 The University of Kansas Natural History Museum Special Publication No. 24 Table 1 (Continued) Taxon Interaction N=in nest/P=phoresy Host & Reference A. punctithorax Reitter N Vespn crabro Linnaeus (Johnson 1993) A. pusilla N Forniicn nifn (Motschulsky 1844) A. rubricollis Bris. de Barne\'ille N L(7s/»s flni'iis Fabricius (Johnson 1993) A. testacen Stephens {=ruficorms Marsham) N Vespula imlgnn-. (Tuck 1896) Epliisteinus sp. (Central America) N Formicidae (C. Johnson, pers. com.) HYPOCOPRINl Hypocoprus formketcrum Motschulsky N Formica rufn (Motschulsky 1844) H. latridioides Motschulsky N Formica exsecia (Kieseritzkv & Reichardt 1936) N Formica fuliginosa (Motschulsky 1844) N Foniiictt fusca Linnaeus (Motschulsky 1844) N Formica rufa (Chaudoir 1845; Kieseritzky & Reichardt 1936) N Formica sp. (Crowson 1955; Colin Johnson pers. com.) H. tenuis Casey N Formica sp. (J.L. Carr, pers. com.) * Larval and adult record. ** Records for B. venustus may be B. hiimilif Illiger or B. siihterrnneiis (Linnaeus). '*' Probably a junior synonym of H. latridioiiiea Motschulsky. t A junior synonym of Formica gagates Dallatorre or f. picea Nylander (Bolton 1995). a well-defined peripheral rim and serves as an evapora- tive surface for secretions released from cuticular glandu- lar ducts. Although the angularity is a landmark structure for the recognition of the Cn/ptopihagns group by beetle tax- onomists, it varies considerably among the taxa of this group (it is reduced in Anthewplmgiis, Catopocliwtiis [Fig. lb], some Cn/ptophagus, and Spaniophaenus) , and is present in two genera outside the Ciyptophagus group (Coombs and Woodroffe, 1955; Leschen, 1996). Members of the Cn/ptophagus group are 1.9-5.2 mm in length, with some species of Antheropihagiis attaining the largest size among cryptophagid species. Antherophagus Dejean.—Antherophagus contains 13 species distributed mainly in the Holarctic with some spe- cies in both the Old and New World tropics. These are well known inquilines of Bombus bees and have been collected in the nests as larvae and adults. Adults are typically golden brown or tan in color with a vestiture of short sparse setae. These beetles are sexually dimorphic; females are unmodified while males have a clypeal notch and com- pact antennomeres. There are records of phoresy for tropi- cal and temperate species, adults of which have been found attached to mouthparts, legs or antennae by their man- dibles. Some species of tropical Antherophagus have a re- duced number of eye facets and are flightless (hind wings are reduced or vestigial), such as A. ludekingi Grouvelle (Java) and A. ruficoniis Grouvelle (South America). Al- though the life history of Antherophagus species has not been fully documented, it is known that adults of holarctic species (e.g., Wheeler, 1919, and Prison, 1921) wait for for- aging Bombus at flowers, attach to the body and are car- ried to the nest where, presumably, mating and oviposi- tion ensues. Catopochrohis Reitter.—(Fig. IB). Catopochrotus con- tains a single species distributed in the Caucasus region of southeastern Europe and was originally described in a separate monotypic family (Reitter, 1889). As the name C. crematogastri Reitter impUes, specimens have been collected from nests of ants in the genus Crematogaster . Adults are brown with well developed, suberect setae. The body is limuloid with the head retracted into the prothorax, and the legs are somewhat flattened and concealed beneath the body. The antenna is in the form of an incrassate club with compact antennomeres. The hind wings are well devel- oped. Nothing has been recorded on the life history of this species and the larva is unknown. There are only a few published records for this species in ant nests; most of the specimens examined by Leschen (1996) were pinned with their host ant species. Cn/ptophagus Herbst.—(Fig. 1 A). Cri/ptophagus is one of the most diverse cryptophagid genera and contains more than 200 described species distributed throughout the world. Most species have been described from the Holarc- tic, although there are some species that are widely dis- tributed stored-grain pests, and a few species occur in northern tropical regions. The hind wings are well devel- oped in most species but there are many continental and island species in which they are reduced or completely absent. Some species of Cryptopliagus have been collected from social insect nests; however, because of their occur- rence in other habitats as adults and larvae (Hinton, 1945; Horion, 1960; Koch, 1989), these records may be inciden- tal and indicative of species that are habitat generalists. Therefore, in this paper, contrary to Leschen (1996), mem- bers of Cryptophagus are considered free-living. Mi/rmcdophila Bousquet.—Myrmedopliila contains a single species, M. aiiwricainis (LeConte), distributed in western North America. Adults are red with sparse pu- bescence. The head is somewhat retracted into the protho- rax, and the tarsi and antennomeres are slightly compact. The hind wings are fully developed. This species most closely resembles species in the genera Cn/ptophngus and Cryptophagid Symbiosis 91 Spnviiif and has been regarded as a member of Cryptophagus bv Ljubarsky (1992). Nothing has been recorded about its Ufe history, although specimens have been collected from the thatched mounds of Formica ants. The larva is un- known. Spauiophaetms Reitter.—Spauiopliaeuus contains three species distributed in southern Europe (Otero and Diaz Pazos, 1995). Spaniophaenus have been collected as adults from termite and ant nests. Adults are brown with well developed suberect setae. The head is retracted into the prothorax and the hind wings are well developed or re- duced. Nothing has been recorded on the life history of Spaiiiophaeiiiis, although specimens have been collected from insect nests and under stones (Ljubarsky, 1992), sug- gesting that members of this genus may be facultative in- quilines. The larva is unknown. Spavins Motschulsky.—Spaz'iiis contains a single palaearctic species. Adults are red with sparse pubescence. This species most closely resembles M. anicricaiiiis (Bousquet, 1989). The head is retracted into the prothorax, and the tarsi and antennomeres are slightly compact. The hind wings are well developed. Adult and larval speci- mens are commonly collected from the thatched mounds of Foruiica ants and the larva is describeci (Eichelbaum, 1927). HYPOCOPRINI The tribe Hypocoprini (Atomariinae), containing three genera, is perhaps the most enigmatic group of cryptophagids. They are minute (0.8-2.2 mm), and the oc- currence of two genera {Alfieriella Wittmer and Ann/dwpn Reitter) in drier habitats and the presence of several unique morphological features not present in other cryptophagids (e.g., lack of a pronotal bead, presternum long in front of procoxae) led Leschen (1996) to doubt their inclusion in the family. One genus, below, is a putative inquiline. Hypocopnis Motschulsky.—Members of Hypocoprus occur in the Holarctic and have been collected in ant nests (Formica) and in leaf litter (Crowson, 1955; Leschen, 1996), and it is not clear if the species (or all populations) are true inquilines. For example, many specimens of the North American species H. tenuis Casey have been collected in Formica nests, while the Old World H. lathridioides Motschulsky may be free-living or found in ant nests. Hind wings in these species are well developed. The larva is unknown. PHYLOGENETIC METHODS Evolutionary studies based on phylogenetic informa- tion are only as good as the phylogenies they are based upon. That is, if any phylogenetic hypothesis is spurious Cryptophaginae Atomariinae Cryptosomatulini Atomariini HypocopriniCryptophagini Caenoscelini Cryptafricini Fig. 2. Phylogenetic hypothesis for the beetle family Cryptophagidae. Numbers of genera are indicated in triangles representing higher ta\a. The Cn/ptophnguf group is a member of the tribe Cryptophagini. then any study using it can be misleading. The phyloge- netic relationships of the genera of Cryptophagidae were recently analyzed by Leschen (1996), and the phylogenies produced in that study are used here. Phylogenetic analy- ses based on parsimony and successive approximations character weighting (Farris, 1969; see also Carpenter, 1988, 1994 ) produced several competing cladograms. The main reason that many competing parsimonious constructions were produced, especially with regard to the tribe Cryptophagini, is many of the branches are supported by relatively few (one to three) synapomorphies. These par- simonious reconstructions are supported by clear synapomorphies that can be challenged by subsequent phylogenetic studies based on additional morphological, behavioral or molecular characters. A strict consensus tree of one set of trees produced by successive approximations character weighting is shown in Fig. 2. Some characters (diet and inquilinism, hind wing reduction or loss) used in my 1996 study and examined here were removed from the data matrix, and the analyses were rerun (Leschen, 1996) because of ambiguities in the coding of the character states. The different phylogenetic hypotheses resulting from these analyses for the Cryptophagus group are dis- cussed in detail elsewhere (Leschen, 1996). The study of character evolution requires that traits of interest be mapped onto terminal taxa to infer ancestral character states. Several statistical tests have been devel- oped for analyzing characters within the context of phy- logenies (Harvey and Pagel, 1991; Brooks and McLennan, 1991; Westneat, 1995), and there is some debate over the null models and their assumptions (Wenzel and Carpen- ter, 1994; Maddison, 1994; Wenzel, 1997). Homoplasy meth- ods (Pagel, 1994) require that the traits in question occur 92 The University of Kansas Natural History Museum Special Publication No. 24 .^ J? Jc / / / ^ cf 4^ t • 27.5 J Bombus symbiosis I Phoresy 2 • 22.5 I® 1 • 27.5 / J^ 2*29 . • 27.0 0*25 2*25 3*27 ® / j9 I I 1 n I I I IJ 8 • 21 1 • 31 4 • 31 |>^;^^^ Reversal to free living 2 '31 Ant symbiosis 2 • 27.5 2 • 27.5 l ! i:i: i :i Saprophagy P7771 Red color I I Gold color reduction Limuloid body form 1 • 27.5 Fig. 3. Phylogenetic reconstruction of the Cryptophagus group (excluding Micramhe and Henotimcrphus) based on Leschen (1996). Char- acters are mapped onto phylogeny using DELTRAN and hind wing reduction is polymorphic for all terminals. Numbers on branches are: branch length • no. of glandular ducts. as multiple independent origins on the branches of recon- structed phylogenies. The frequency of convergent featT.ires as they appear in the trees is the basis for statistical analy- ses. In contrast, homology approaches (Coddington, 1994; Wenzel and Carpenter, 1994) are not statistical in nature but aim to understand character evolution as logical de- ductions of character changes based on cladograms. Al- though homoplasy and homology methods may demon- strate character correlation either statistically or logically, they may fail in explaining true adaptation which requires additional information (i.e., population stn.icture, environ- mental or developmental data) external to tree topology (Coddington, 1988; Leroi et al., 1994; Wenzel and Carpen- ter, 1994). I use a homology approach in this paper to de- termine the origin and nature of shifts to symbiosis be- cause actual (rather than simulated) distributions of data on phylogenetic trees better reflect the origins of symbio- sis. Moreover, the traits of interest are relatively rare in Cryptophagidae, and there are many phylogenetic hypoth- eses to choose among for the relationships among mem- bers of the Cryptophagus group making statistical analyses cumbersome. All possible parsimonious character mappings onto the cladogram shown in Fig. 3 (also see Leschen, 1996, Fig. 3) were investigated for determining character evolution (Maddison, 1994) by using the Equivocal Cycling option in MacClade (Maddison and Maddison, 1992). Both ACCTRAN and DELTRAN optimizations (Maddison et al., 1984) were used, and different resolutions of polytomous reconstructions (where necessary) were exam- Cryptophagid Symbiosis 93 ined to construct character state graphs of inquihne host use. Total numbers of glandular ducts were mapped as continuously varying characters using maximum linear parsimony (Maddison and Maddison, 1992). Branch length was determined by counting the number of unambiguous character changes (Leschen, 1996, Fig. 3). Polymorphic inquiline associations were coded in the data matrix as monomorphic units (Nixon and Davis, 1991) that represent total variation seen in the genus Spaniophaeiiiis. This genus contains individuals that are free-living or occur with two social insect hosts. It is repre- sented by three terminal taxa and the trichotomy is re- solved arbitrarily (not shown in figures). Hind wing loss or reduction, which is polymorphic in some taxa, was treated as a fixed polymorphic character state (1/0) for terminal taxa. CHARACTER INTERPRETATION AND PATTERNS OF SYMBIOSIS Different hypotheses about character state construc- tion can lead to different interpretations of the evolution of that character (Pogue and Mickevich, 1990; Wilkinson, 1995). This problem is most difficult to resolve with eco- logical characters such as inquilinism because definitions are often axiomatic (Miller and Wenzel, 1995). Character states of incjuilinism (as complex multistate characters) can be interpreted differently based on ecological context, taxo- nomic rank, and independence among character states re- sulting in three biologically explicit interpretations of ho- mology. Ecological iuquiliuism is defined as cryptophagid symbiosis with ants, bees, and termites. These associations are assumed to be ecologically similar but may not be bio- logically independent (see below). This character consists of two states: free-living (0) and symbiotic (1). There are a variety of arguments against using taxon-based definitions for comparative studies of character evolution and spe- cies diversity (e.g., Doyle and Donoghue, 1993). Advocates of the phylogenetic approach strictly limit comparisons to monophyletic groups because taxonomic ranks are arbi- trary assignments for categories that may or may not re- flect monophyly. Biases due to rank of the inquiline host, as a character mapped onto a cladogram, may also affect evolutionary interpretation. There are two definitions used here for understanding the nature of symbiosis that reflect monophyly of the host taxa. Taxon-spccific inquiluusni (TSI) is defined as symbiosis only with ants (Formicidae), or bees (Apidae), or termites (Isoptera). These associations, as char- acter states, represent independent character states equiva- lent to family- or ordinal-level monophyletic taxa. This character consists of four character states: free-living (0) or ant (1), bee (2), and termite associations (3). This char- acter state interpretation most closely resembles the model Termites -<- Ants Fig. 4. A 4-step model for inquiline characters states based on a taxonomic-specific interpretation (TSI) for homology and on tree in Fig. 3. used by Leschen (1996) for his character number 114. Hosl- specific iiiquilinisni (HSl) is defined as cryptophagid sym- biosis with a specific genus or species of host. This defini- tion, therefore, considers that each specific association is biologically independent. This character consists of six character states: free living (0) or Anncanthoiermes (1), Crematogaster (2), Bombiis (3), Formica (4), and Messor (5) associations. It is unequivocal that the association with Formica is not homologous between members of Hypocoprus and the Cn/ptophagiis group because these taxa are distantly related and isolated in the cryptophagid phylogeny (Fig. 2). Char- acter state transformations within the Cryptophagus group are considered in the following discussion. Based on character optimizations, ecological inquilinism represents a 3-step model of character transi- tion by ACCTRAN (one gain at A in Fig. 3 and two losses; one for Cryptoplmgiis and the other within Spaniophaemis) and DELTRAN (two gains, one for Autherophagus and at D in Fig. 3, and one loss within Spaniophaenus). The num- ber of colonization events may be underestimated due to the fact that cryptophagid inquilines use a variety of so- cial insects as hosts and that each symbiotic association may have originated independently. A TSI definition of symbiosis produces a 4-step model (see character state graph in Fig. 4) of character change where bee inquilinism evolved once in Autherophagus, ant inquilinism evolved once atA in Fig. 3, termite inquilinism evolved once within Spauiophaeuus, and there was a single reversal back to free- living in Spaniophaouis from an ancestral ant association. When symbiosis is based on an HSI interpretation (not shown), inquilinism evolved one or two times in Formica , and once in Bombus, Crematogaster, Acauthotermes, and Messor. The shift from Formica (in the ancestor of Spavius + Mi/rmedophila using DELTRAN optimization) to Crematogaster (in Catopochrotus) is evidence for a shift that 94 The University of Kansas Natural History Museum Special Publication No. 24 Free living ® Phoresy/Bombus > Ants A Termites Free living lor 2, © lor 2 Ants Phoresy/Bombus Termites / .-^ ^^ c / / =^ termite symbiosis / d-" 4^ 5>^ // Cryptophagid Symbiosis 95 by two iterations, and a strict consensus of these is shown for the Cn/ptopliagiis group in Fig. 5B. Note that the rela- tionships shown in Fig. 3 are similar to that shown in Fig. 5B if two internodes are collapsed and the tree is rooted at Mi/nncdophila. Hypotheses about the evolution of inquilinism based on the newly produced SAW trees are as follows: Bombiis/ phoretic symbiosis evolved once, while nonphoretic sym- biosis arose once with termites and once or twice with ants, while reversals to free-living occurred once or twice from ant associations (Fig. 5C). In all reconstructions Boinbus associations originated once and did not occur within lin- eages of ant-symbiotic lineages, indicating that the two forms of inc]uilinism (nonphoretic and phoretic) are indeed independently derived. This pattern is also reflected in other phylogenies proposed by Leschen (1996). Choosing this 4-step model over previous models may be desirable because the additional information regarding phoresy has been used to illuminate an additional path to the evolu- tion of symbiosis. However, this model assumes that all members of Ajitliewp^Imgiis are phoretic (which may not be true; see Leschen, 1996) and the remaining symbionts fly directly to insect nests. The phoretic /nonphoretic hypoth- esis is equally parsimonious to the TSl interpretation, and these together provide more parsimonious interpretations for character evolution compared to the model based on an HSl definition. HOST USE AND THE ORIGIN OF SYMBIOSIS As is true for any affirmation of host relationship, lar- val associations, repeated collections and behavioral ob- servations must be made to substantiate a symbiotic asso- ciation. In most cases, these criteria have not been met in Cryptophagidae. Additional field data and direct behav- ioral observations are necessary to determine the nature of host associations. This is especially true for SpiDiiophmniis and Hypocopnis, for which there is very little known about natural history. None of the Cryptophagus species are true inquilines because these have been reported as stored prod- uct pests and in other habitats (Hinton, 1945; Horion,1960; Koch, 1989). Moreover, larval records for some of the spe- cies of Cn/ptoplmgiis (Table 1) were made in association with habitats other than social insect nests (see Leschen, 1996). Current data reveal two classes of inquilines: obligate or facultative symbionts. Facultative symbionts are those taxa containing a mixture of species or populations that are free-living or symbiotic. Obligate inquilines are those taxa that contain populations that occur exclusively in as- sociation with social insect nests. While this simple scheme is useful, there are some limitations and phylogenetic im- plications for these classes. For example, if a facultative taxon contains more than two recognizable monophyletic taxa, and the internal phylogenetic relationships are un- known, then the true phylogenetic pattern of svmbiosis may be enigmatic. At the generic level, Hypocopnis and Spaniophaeniis may be considered facultative inquilines because they contain a mixture of species or populations some of which are free-living and others symbiotic. The remaining taxa mentioned in the review are considered true inquilines. Despite different theories about character transforma- tion, inc]uilinism in its various forms has evolved in mem- bers of the Cryptoplingiis group, suggesting that this group is prone to some level of inquiline diversification. More- over, because different interpretations of character states may increase the number of "origins," this pattern sup- ports Wilson's (1971) view that some groups are more prone to inquilinism than others. Although causal factors promoting symbiosis in the Cryptophagus group remain obscure, a better understanding of specific host- or habi- tat-seeking behaviors may be useful to determine proxi- mate mechanisms. The nature of the shift from free-living to inquilinism may be revealed by comparing the microhabitats present in social insect nests to those microhabitats utilized by ancestral free-living and derived inquiline cryptophagids. Many Boiubiis bees build their nests in pre-existing rodent burrows (Michener, 1974), while Acanthotermes stores caches of plant foods (Artemev and Zhuzhikov, 1968), Messor (harvester ants) are seed gatherers, and Formica ants construct mounds of thatch-debris (Wilson and Holldobler, 1990). Incidental associations of other free-living cryptophagid species with social insects may be based on collections made from abandoned nests or active nests with vacant portions that have decayed. These microhabitats promote fungal growth and could have attracted ances- tral free-living cryptophagids that were mycophagous (a primitive behavior for the family) to abundant food re- sources present in social insect nests, facilitating the shift to inquilinism. Similarity among fungus-promoting micro- habitats outside and in social insect nests supports a habi- tat similarity hypothesis for the evolution of inquilinism in Cryptophagidae. The evolution of host-use in phytophagous and my- cophagous insects can be understood by examining host suitability, preference, and encounter-frequency (Jeanike, 1990). It has been argued by Wilson (1971) that large popu- lations of social insects may promote the evolution of sym- biosis by creating abundant microhabitats in ecological communities. Formica ant colonies, for example, are very common in holarctic ecosystems and may have served as attractive microhabitats to the ancestors of cryptophagid inquilines. Therefore both habitat similarity and ecologi- cal opportunity may explain the occurrence of 96 The University of Kansas Natural History Museum Special Publication No. 24 cryptophagids in social insect nests as implied by Wilson (1971) and Kistner (1979) for other groups of inquilines. Shifts among unrelated hosts supports the hypothesis that similar ecologies, rather than genealogy, has influenced the origin of symbiosis in Cryptophagidae with social insects. DIET AND MORPHOLOGICAL EVOLUTION ASSOCIATED WITH SYMBIOSIS Character evolution of inquilines of the Cn/pHopluii^us group is considered in detail in this section. Character state reconstructions are mapped onto the resolved phylogeny shown in Fig. 3. General classifications of inquiline behavior consider diet as one important indicator of behavioral integration into the host nest (Wasmann, 1894; Wilson, 1971). Liquid food obtained by trophallaxis between host and inquiline and food gleaned from host bodies are thought to be some of the important feeding mechanisms used by fully-inte- grated symbionts (Kistner, 1982; Holldobler and Wilson, 1990). Although liquid diets are difficult to observe because gut squashes appear empty (Leschen, 1993), the guts of cryptophagid inquilines always appear to contain some material. This suggests that cryptophagid symbionts may not be fully integrated into their host nests to the degree seen in other symbiotic beetles that interact directly with their hosts. Empirical observations on symbiont and host interactions are necessary to support this presumption. Leschen (1996) considered Antliewphagiis a pollen feeder because he observed guts of larvae and adults col- lected from Bowbns nests are filled with pollen. However, this pollen originally collected by foraging Boinhns is re- leased as feces of the bee larvae among nest debris (Scott, 1920; Erison, 1921; Plath, 1922). Therefore, the diet of adults and larvae of Aiithewpliagus in the nest is considered here as saprophagy, while the diet of adults collected on flow- ers remains unknown. Based on gut analysis, mycopha- gous and saprophagous diets occur in cryptophagid in- quilines, while mycophagy is a primitive feature for Cryptophagidae (Leschen, 1996). Saprophagy arose at least three times independently in the Cryptophngiis group (four gains or three gains and one reversal) and does not appear to be consistently correlated with shifts to symbiosis be- cause of a single reversal to mycophagy (i.e.. Spavins), and some free-living taxa are saprophagous (Leschen, 1996: Mnionomidius Reitter and Striatocryptiis Leschen). Cryptophagid inquilines that have shifted from mycophagy to saprophagy in association with social in- sect nests may be ingesting suitable foods other than fungi (dead and moribund insects, debris in waste heaps, etc.). There are, however, biases associated with small sample sizes in the number of guts examined due to availability of specimens for dissection (Leschen, 1996). Saprophagy, if viewed as a specialization in Cryptophagini, needs to be documented further. There are several morphological features that are of- ten observed in inquiline lineages (Kistner, 1982). The fact that these characters (head retracted into the prothorax, compact anteraiomeres, body color, etc.) do not consistently occur together in every inquiline lineage or species sug- gests that natural selection for these traits differs, depend- ing on underlying phylogenetic history, developmental constraints, and specific host ecology and behavior. There are several unique features in some cryptophagid genera that could be correlated with inquilinism (e.g., clypeal notch of male Antherophagus); however, proving that these unique characters are true adaptations to symbiosis may be difficult (Coddington, 1988) without a better under- standing of function (Lauder, 1981). The majority of cryptophagid species are brown in color, in contrast, AiitJwrophagus, MyrnwdopMa, and Spmviiis are red or gold and have a reduced vestiture of sparse se- tae. Red coloration has evolved once (DELTRAN optimi- zation) or twice (ACCTRAN optimization) and gold col- oration is unique to Aiitlwwphngiis spp. The coloration and vestiture in these beetles may be linked to chemical sys- tems useful in obtaining specific colony odors or cuticular hydrocarbons of the host (Howard et al., 1980). Many in- quilines that are known to have active hormonal systems also have characteristic cuticular modifications (Seevers, 1965; Kistner, 1982). Fig. 3 shows that the Fonuicn inquilines Myrmedophila and Spavins are the only species that are red (but see Ljubarsky, 1992) and are separated by one step. This suggests that inquilines of Foriuica ants either share similar selective regimes for red coloration by convergence or common ancestry (homology). Members of Hypocopriis differ in color from their analogous counterparts Mynnedophila and Spavins which refutes an adaptive col- oration hypothesis for all species associated with Formica. Moreover, different phylogenetic results would also refute a homology hypothesis for red coloration in Myrmedophila and Spavins (e.g., see Fig. 5B). Hind wing loss or reduction is a feature often discussed in the context of inquilinism (Seevers, 1965), especially in some beetle groups where dispersal to new nests occurs only by phoresy (e.g., Roubik and Wheeler, 1982). Wing reduction has at least four independent origins and is poly- morphic for the terminal taxa of the Cn/ptophagns group. This character is probably not directly related to inquilinism in cryptophagids because it has evolved re- peatedly in many free-living cryptophagine species and genera (Leschen, 1996: Henoticns Thomson, Micrambe, Miiionomidins, and Mnioticus Scott). Therefore, it is likely that wing loss is a function of resource abundance or habi- tat stability (Roff, 1990; Thayer, 1992) and not directly re- Cryptophagid Symbiosis 97 lated to inquilinism per se. Inquilinism may be correlated with character evolution in Autlicwplmgiis where wing loss and eye reduction covary in tropical species that are strictly phoretic (Leschen, 1996). Additional behavioral data and species-level phylogenies are needed to fully understand the correlation of these traits. Many inquiline beetles have special glands and tri- chomes that function in appeasement, behavioral duping of the host, and chemical mimicry (Kistner, 1979; Steidle and Dettner, 1993; Holldobler and Wilson, 1990). Some- times glandular systems already present in free-living an- cestors may become more complex or simplified in sym- bionts (Steidle and Dettner, 1993). The simple cuticular glandular ducts of Cryptophagini are present at various positions on the body (Leschen, 1996). Curiously, the glan- dular ducts of cryptophagid inquilines have not under- gone any of the complex changes in reservoir and deliv- ery systems that have occurred in the glandular systems of some other inquilines (i.e., staphylinids, histerids, and others) in which duct openings may be furnished with elaborate trichomes for chemical release. If anything, some of the cryptophagine inquilines have a poorly-developed pronotal angularity and reduction also occurs in the lat- eral teeth on the pronotal carina which may be associatecl with symbiosis. However, angularity-reduction occurs in various free-living species of Cryptopliagiis and Micnvnbe. The overall cuticular morphology suggests that the chemi- cal system in Cryptophagini was not co-opted for use in symbiosis, although detailed histological and chemical studies may show otherwise. The total number of glandu- lar ducts present throughout the body in the Cn/ptophagits group varies from 20-31 (Leschen, 1996). This range does not appear to change drastically in phylogeny even though the ranges of glandular duct numbers in the symbiotic genera AiitJiewphagiis (20-25) and Cntopochrotus (21) are below the mean in Cryptophagini (26.9). If a causal rela- tionship truly exists, the number of glandular ducts must be shown to have some performance advantage, such as being linked to specific changes in behavior, or in response to differences in chemical ecology. Inquilines are regarded as some of the strangest insect forms ever to have evolved, and it is conceivable that over the evolutionary course of an association with social in- sect hosts radical morphologies emerged as a response to tighter links to the host through behavioral interactions. Many authors have called attention to the highly modi- fied body forms in inquiline groups that are purportedly used to dupe or mimic hosts to gain access into colonies and protect the inquiline from aggressive attacks from its host or visual predators. In beetles, especially staphylin- ids, there are two forms present (Seevers, 1965; Wilson, 1971; Kistner, 1979), one of which is present in Crvptophagidae. A "mimetic" form converges on the body plan of the host to the extent that only a trained eye can discriminate host from beetle in the field. This body form does not exist in Cryptophagidae. Another form found in many beetle inquilines is a limuloid body form, which is present in the cryptophagid genus Cafopodvotus (Fig. IB). The limuloid body form conceals the appendages below by an expansion of lateral portions of the body. Limuloidy has evolved repeatedly in several lineages of Coleoptera and other insects and it is thought to protect the insect from dorsal attacks during aggressive interac- tions between incjuiline and host (Kistner, 1979; Holldobler and Wilson, 1990). Among the morphological "grades" of inquilinism discussed by other authors, limuloidy is con- sidered as one of the most advanced. Two phylogenetic hypotheses can be made about the origin of this body form. Limuloidy may evolve within an inquiline lineage either at the origin of symbiosis or some time thereafter. A grade of morphologies is evidence supporting the hypothesis that modified body forms evolved subsecjuent to the origin of symbiosis. Discrimination among these hypotheses can be made by examining where Catopochwtiis occurs relative to other inquilines in the cryptophagid tree. Catopochwtiis crematognstri is consistently placed as one of the most de- rived members (with its sister taxon Spniiiophneinis) of the Cn/ptophagiis group, suggesting that limuloidy may not necessarily be an important pre-adaptation for the evolu- tion of inquilinism to occur. Moreover, the similarity be- tween the remaining obligate inquilines with Cn/ptophagiis supports the view that over time more "adaptive" and di- vergent morphologies evolve as lineages remain inte- grated. This is supported by the number of character changes present on branches of the phylogenetic tree (Fig. 3), where Catopocluvtus has eight terminal changes (four losses and four gains—see Leschen, 1996) and the branch lengths of the remaining terminal taxa and ancestors of the Cn/ptofihagiis group range from one to four. This ap- parent latent shift in the acceleration rate of morphologi- cal change has occurred after the origin of symbiosis. The morphological evolution in Catopochwtiis, however, may be a specific response to its host and not at all related to the general phenomenon of inquilinism. This may also be true for the phoretic genus Anthcwphagus, which also has a number of unique features that may be related to inquilinism. Also, comparisons among branch lengths are biased with respect to taxonomic rank (comparisons made among genera) and do not consider character changes that are associated with cladogenetic events occurring within each genus (Doyle and Donoghue, 1993). RATES OF SPECIATION AND BIOGEOGRAPHY Changes in the speciation rate among groups of mono- phyletic taxa is perhaps the most intriguing and least un- 98 The University of Kansas Natural History Museum Special Publication No. 24 derstood evolutionary question concerning biologists (Sanderson and Donoghue, 1995). While some explana- tions emphasize factors intrinsic to each clade, such as key innovations correlated with entry into new adaptive zones (Simpson, 1944, 1953; Leim, 1974; Ricklefs and Schluter, 1993), or differences of niche breadth (Vrba, 1988; Eldredge, 1989), other explanations emphasize extrinsic factors such as climatic and geological changes, which may also create differences in diversity among monophyletic groups of taxa (Nelson and Platnick, 1981; Wiley, 1981; Cracraft, 1985). Paramount to understanding phylogenetic patterns are well-resolved phylogenies for hosts and inquilines, well- documented data on social insect hosts, and estimates of rates of divergence (either geologic or molecular-based). Cryptophagid fossils are rare; the best preserveci fauna that has probably been correctly identified are those specimens recorded from Baltic amber and dated as Oligocene (Spahr, 1981). In this fauna there are several genera, including Anthcropha^iis, Cr\/ptopliagiis, Micrmube and Spavins, sug- gesting that the ecological associations for the Ciyptophagiis group are at least 30 my old. The evolution of inquilinism in Cryptophagidae did not accompany major radiations (number of species per monophyletic lineage) as it has in other groups containing inquilines, such as Staphylinidae or Histeridae. None of the inquilines has a high number of species relative to the diverse taxa Cn/ptoplmgtis and Micrambc with 200 and 80 species, respectively (Leschen, 1996). This pattern may be related to relatively recent evolution of the majority of in- quilines subsequent to the branching event separating Cryptophagus from other taxa in many of the phylogenetic reconstructions (Fig. 3). On the other hand, Aiithewp^Imgiis is more diverse than the remaining genera of cryptophagid inquilines and is also a relatively basal member in the Cri/ptopliagiis group. The number of species in Antherophagiis may be the result of an early divergence relative to other members of the Cryptophagus group rather than an association with social insects. It is curious that despite an estimated historical association for 30 my there has been a relatively slow rate of speciation in inquiline lineages. By comparing the phylogenies of several staphylinid inquiline groups, Kistner (1979) elucidated four phyloge- netic patterns of host use and speciation rates. Kistner's major points (patterns) are: 1) speciation in inquilines oc- curs faster than in the host lineage; 2) host shifts or trans- fers in inquiline lineages appear to be associated with slower speciation rates while 3) lineages that maintain host fidelity have higher rates of speciation; and 4) "incomplete" host specificity or broad host use may increase rates of speciation. In all cases the number of cryptophagid in- quiline species is far less than the number of species con- tained in the host genera (e.g., compare the number of cryptophagid species to the number of host species in Michener [1974] for bees and Holldobler and Wilson [1990] for ants) which seems to refute Kistner's first point. On the other hand, a relatively slow rate of evolution may be related to the amount of host-shifting present in the Cryptophagus group. For Antherophagiis, current data for host-use (Table 1) suggests that some species are general- ists on Bombus species while maintaining an exclusive as- sociation with the genus. Clearly, the data for Cryptophagidae do not verify all of Kistner's (1979) hy- potheses, but his study may be relevant only to those groups of Staphylinidae he studied. Cryptophagid inquilinism appears to be limited to those groups occurring in the Holarctic, and while there is a growing body of literature about the biogeographical affinities in this geographical region (Enghoff, 1995) few phylogenetic studies refer to symbiotic associations span- ning this area (although insect/host-plant examples exist, see Moran, 1982; Mitter and Farrell, 1991). There are sev- eral symbionts limited to Formica in the Holarctic (Kistner, 1982; Wilson, 1971), and Wasmann (1906) suggested that the three staphylinid genera in the Lomochusina originated in Europe and later evolved with their respective Formica hosts in the Holarctic. There may have been a vicariant event marking the separation of Spiaviiis and Myrmedophila (at least in trees showing that these may be closely related) consistent with the separation of North America from Eurasia that is also present in lomochusine staphylinids (if the association with Formica is primitive). This pattern may also hold true for the association of Hypocopriis with Formica. Host fidelity, therefore, is preserved in phyloge- netic history that includes large-scale vicariant events in Formica symbionts. Another case in host fidelity is Antherophagiis, which is completely sympatric with its host Bombus. Most Bombus species are found in the Holarctic region (Michener, 1974), although there are some species occurring in tropical ar- eas in the Old World and New World that represent sub- sequent dispersal, assuming that Bombus originated in the Holarctic. If Antherophagiis has tracked the evolution of Bombus bees, then species phylogenies of both Antiierophagiis and Bombus may be concordant. No clear pattern exists for host associations when Antlieropliagus species are mapped onto the preliminary phylogeny of Bombus provided by Williams (1985). Fully resolved phy- logenies for Antherophagiis and Bombus are necessary to address this hypothesis. In conclusion, knowledge about the natural history and evolution of symbiosis in Cryptophagidae is in its in- fancy. Behavioral observations are certainly necessary, es- Cryptophagid Symbiosis 99 pecially of larval associations that indicate oviposition pref- erences. Because of alternative phylogenies for the Ciyptophagiis group (Leschen, 1996) that provide many equally probable hypotheses for patterns of character evo- lution, choosing among these hypotheses is difficult. Leschen (1996) acknowledged that the Ciyptophagiis group is a systematic mess. A complete understanding of the phylogenetic relationships of the group hinges on a com- prehensive consideration of the morphological variation within the genus Cryptophagus. LITERATURE CITED Akre, R. D. and C. W. Rettenmeyer. 1966. Behavior of Staphylinidae as- sociated with army ants (Formicidae: Ecitonini). lournal of the Kan- sas Entomological Society 39: 745-782. Artemev, M.M. and D.P. Zhuzhikov. 1968. Food collection and feeding in Anncantliotcnnefnhni;criniuif Jacobs. (Isoptera, Hodotermitidae). Pp. 74-78, in Luppova, A.N. (ed.) Tcniiitci^ and Control Measures Against Them. Proceedings of the 2nd AU-Union Conference on the Study of Termites and Their Control, [in Russian) Bernard, F. 1968. Les fourmis (Hi/menoptera Forniicidae) d'Europe occidentalc et septeutrionale (Favme de 'Europe et du Bassin Mediterraneen, no. 3). Paris, Masson et Cie. 411 pp. Bold, T.J. 1856. Note on Antherophagtis nigricornis and Bomhus sylvartini. Zoologist 14; 5003. Bolton, B. 1995. A New General Catalogue of the Ants of the World. Cam- bridge, Harvard University Press. 504 pp. Bousquet, Y. 1989. A review of the North American genera of Cryptophaginae (Coleoptera: Cryptophagidae). Coleopterists Bul- letin 43: 1-17. Brooks, D.R. and D.H. McLennan 1991. PIn/logeny, Eeology, and Belnroior: a Researcli Program in Comparative Biology. Chicago, University of Chicago Press. 434 pp. Bugnion, E. 1869-70. (Letter to E. Ferris on Antherophai;us). L'Abielle 7: 26-27. Buckle, C. W. 1900. Beetles collected in Lough Foyle District, Cos. Donegal and Derry. Irish Naturalist 9: 2-11. Carpenter, J. M. 1988. Choosing among multiple equally parsimonious cladograms. Cladistics 4: 291-296. Carpenter, J. M. 1994. Successive weighting, reliabilitv and evidence. Cla- distics 10: 215-220. Chaudoir, M. 1845. Notices entomologiques^sur le gouvernement et la ville de Kiew. Bulletin de la Societe Imperiale des Naturalistes de Moscou 18: 158-213. Chavarria, G. 1994a. Phoresy on a neotropical bumblebee (Hymenoptera: Apidae) bv Antherophas;us (Coleoptera: Cryptophagidae). Psyche 101:109-111. Chavarria, G. 1994b. First neotropical mutualistic associations in bumble- bee nests (Hymenoptera: Apidae). Psyche 101: 112-118. Coddington, J. A. 1988. Cladistic tests of adaptational hypotheses. Cla- distics 4: 30-22. Coddington, J. A. 1994. The roles of homology and convergence studies in studies of adaptation. Pp. 53-78, in Eggleton, P. and R. I. Vane- Wright (eds.)Plniloge)uiies and Ecolog]/ . New York, Academic Press. Coombs, C. W. and G. E. Woodroffe. 1955. A revision of the British spe- cies o( Cryptophagus (Herbst) (Coleoptera: Cryptophagidae). Trans- actions of the Royal Entomological Society, London 106: 237-282. Cottam, R. 1909. Antherophagus pallens in Derbyshire, Lanes. Naturalist 2: 266. Cracraft, J. 1985. Biological diversification and its causes. Annals of the Missouri Botanical Garden 72: 794-822. Crowson, R. A. 1955. The Natural Classification of the Families of Coleoptera. London, Nathaniel Lloyd and Co. 187 pp. Crowson, R.A. 1980. On amphipolar distribution patterns in some cool climate groups of Coleoptera. Entomologia Generalis 6: 281-292. Cumber, R.A. 1949. Humble-bee parasites and commensals found within a thirty mile radius of London. Proceedings of the Entomological Society of London (ser. A) 24: 119-127. de Pinna, M.C.C. 1991. Concepts and tests of homology in the cladistic paradigm. Cladistics 7: 367-394. Dettner, K. and C. Leipert. 1994. Chemical mimicry and camouflage. Annual Review of Entomology 39: 129-154. DoUman, H. 1912. Coleoptera taken at Ditchling, Sussex, during 1911. The Entomologists Record 24: 20-23. Donisthorpe, H. 1904. Coleoptera in the New Forest. The Entomologists Record 16: 325-326. Donisthorpe, H. 1 909. On the colonisation of new nests of ants by myrme- cophilous Coleoptera. Transactions of the Entomological Society of London 1909: 413-29. Donisthorpe, H. 1920. The phoresy of Antherophagus. Entomologists Record 32; 181-187. Doyle, J. A. and M.J. Donoghue. 1993. Phylogenies and angiosperm di- versification. Paleobiology 19: 141-167. Eichelbaum, F. 1907. Die Larven von Cis festivus Panz. und von Emphylus glaher Gyll. Zeitschrift fiir wissenschaftliche Insekten-Biologie 3: 25- 30 Eichoff, 0. 1866. Sammelberichte. Berliner Entomologische Zeitschrift 10; 293-295. Eldredge, N. 1989. Macroevolutionary Dynamics: Species, Niches, and Adap- tive Peaks. New York, McGraw-Hill. 226 pp. Enghoff, H. 1995. Historical biogeography of the Holarctic: area relation- ships, ancestral areas, and dispersal of non-marine animals. Cladis- tics 11: 223-263. Falcoz, L. 1929. Tableaux analytiques des coleopteres de la faune Franco- Rhenane (France, Hollande, Belgique, Region rhenane, Valais); Famille 33: Cryptophagidae. Miscellanea Entomologique 31-32: 1- 197. Farris, J.S. 1969. A successive approximations approach to character wieghting. Systematic Zoology 18; 374-385. von Frisch, K. 1952. Hummeln als unfriewillige Transportflieger. Natur und Volk82: 171-174. Prison, TH. 1921. Antherophagus ochraceus Melsh. in the nests of bumble- bees. The American Naturalist 45; 188-192. Prison, TH. 1926. Contribution to the knowledge of the interactions of the bumblebees of Illinois with their animate environments. An- nals of the Entomological Society of America 29; 226-230. Gorham, H. S. 1869. Leptinus in bee's nests. Entomologists Monthly Maga- zine 6; 89. Grandi, G. 1936. Contributi all conoscanza degli Imenotteri Aculeati. XV. BoUettino dell'Instituto di Entomologia dell R. Universita degli Studi di Bologna 8:27-121 + 4 pits. Grouvelle, A. 19113. Description d'un Antherophagus de Java et de la larve (Coleoptera: Cryptophagidae). Notes Leyden Museum 33: 117-120. Grouvelle, A. 1911b. Coleopteres cryptophagides recueilles au Mexique par M.L. Diguet dans les nids du Bonthus ephippiatus Say Bulletin Museum Histoire Naturelle, Paris 1 9 1 1 : 97-1 00. Harvey, P.H. and M.D. Pagel. 1991. The Comparatwe Method ni Evolution- ary Biology. Oxford, Oxford, University Press. 239 pp. Hempel, C. G. 1966. Philosophy of Science. Foundations of Philosophy Se- ries. Toronto, Prentice Hall. 116 pp. Hinton, H. E. 1945. A Monograph of the Beetles Associated with Stored Prod- ucts. Vol. I. British Museum (Natural History), London. 443 pp. 100 The University of Kansas Natural History Museum Special Publication No. 24 HoUdobler, B. and E. O. Wilson. 1990. The Ants. Cambridge, Harvard University Press. 732 pp. Horion, A. 1960. Faunistik der mitteleuropiiischen Kafer. Band VII: Clavicornia. Teil 2. (Sphaeritidae bis Phalacridae). Teredilia, Coccinellidae. Uberlingen, A. Feyel, xiii + 375 pp. Howard, R. W., C.A. McDaniel, and G.J. Blomquist. 1980. Chemical mim- icry as an integrating mechanism: cuticular hydrocarbons of a termitophile and its its host. Science 210: 431-433. Husband, R.W. and T.M. Brown. 1976. Insects associated with Michigan bumblebees (Bombus). The Great Lakes Entomologist 9: 57-62. Jeanike, J. 1990. Host specialization in phytophagous insects. Annual Re- view of Ecology and Systematics 21: 243-273. Johnson, C. 1993. Pwvisuvml atlas of the Cryptophaguiae-Atcmarimae (Co- U'opHera) of Britain and Ireland. Huntington, Biological Records Cen- tre. 91 pp. Joy, N. H. 1932. A prnetical handbook of British beetles. 2 vols. London, Witherby. 622 pp. Kato, M., S. Salmah, and T Nagamitsu. 1992. Colony cycle and foraging activity of a tropical-montane bumblebee, Bombus rufipes (Hy- menoptera, Apidae) in Southeast Asia. Japanese Journal of Ento- mology 60: 765-776. Kieseritzky, V. and A. Reichardt. 1936. Zur Frage iiber die Selbstandigkeit der Familie Catopochrotidae, im Zusammenhang mit der Entdeckung einer neuen Gattimg. Trudy Zoologicheskogo Instituta Akademii Nauk SSSR 3: 693-697 |in Russian and Latin]. Kistner, D. H. 1979. Social and evolutionary significance of social insect symbionts. Pp. 339^13, in H. R. Hermann (ed.). Social Insects, Vol. 1. New York, Academic Press. Kistner, D. H. 1982. The social insects' bestiary. Pp. 361-423, in H. R. Hermann (ed.). Social Insects, Vol. III. New York, Academic Press. Koch, K. 1989. Die Kafer Mitteleuropas, Okologie, Band 2. Krefeld, Goecke and Evers. 382 pp. Lauder, G. V. 1981. Form and function: structural analysis in evolution- ary morphology. Paleobiology 7: 430^142. Leim, K. F. 1974. Evolutionary strategies and morphological innovations: Cichlid pharyngeal jaws. Systematic Zoology 22: 425^141. Leroi, A.M, M. R. Rose, G.V Lauder 1994. What does the comparative method reveal about adaptation. American Naturalist 143: 381^02. Leschen, R. A. B. 1993. Evolutionary patterns of feeding in selected Staphylinoidea (Coleoptera): shifts among food textures. Pp. 59- 104, II! C.W. Schaefer and R. A, B. Leschen (eds.). Functional Mor- phology of Insect Feeding. Thomas Say Publications in Entomology: Proceedings of the Entomological Society of America. Leschen, R.A.B. 1996. Phylogeny and revision of the genera of Cryptophagidae (Coleoptera: Cucujoidea). University of Kansas Science Bulletin 55: 549-634. Ljubarsky, G. Y. 1992. On the taxonomical status of the genera Myrmedophila Bousquet, 1989 and Catopochrotides Kies. and Rchdt., 1936 (Coleoptera, Cryptophagidae). Entomologicheskoye Obozrenie 71: 595-598 [in Russian; translation in Entomological Review 72: 28-31,1993). Maddison, D.R. 1994. Phylogenetic methods for inferring the evolution- ary history and processes of change in discretely valued characters. Annual Review of Entomology 39: 267-292. Maddison, W. P. 1990. A method for testing the correlated evolution of two binary characters: are gains or losses concentrated on a certain branch of a phylogenetic tree? Evolution 44: 539-557. Maddison, W. P and D. R. Maddison. 1992. MacClade: Analysis of Phylog- eny and Character Evolution, version 3. Sunderland, Sinauer. Maddison, W. P., Donoghue, M. J., and D. R. Maddison. 1984. Outgroup analysis and parsimony. Systematic Zoology 33: 83-103. Meer Mohr, J. C. van der and M. A. Lieftinck. 1947 (1945). Over de biologie van Antherophagus ludekingi Grouv. (Col.) in hommelnesten (Bombus Latr.) op Sumatra. Tijdschrift voor Entomologie 88: 207-214. Michener, C. D. 1974. The Social Behavior of Bees: A Comparative Study. Cam- bridge, Harvard University Press, xii + 404 pp. Mickevich, M. F, and S. J. Weller. 1990. Evolutionary character analysis: tracing character change on a cladogram. Cladistics 6: 137-170. Miller, J.S. and J.W. Wenzel. 1995. Ecological characters and phylogeny. Annual Review of Entomology 40: 389-415. Mitter C. M. and B. Farrell. 1991. Macroevolutionary aspects of insect- plant relationships. Pp. 35-78, in E.A. Bernays (ed.). Insect/plant In- teractions, vol. 3. Boca Raton, Florida, CRC Press. Moran, N. 1989. A48-million-year-old aphid—host plant association and complex life cycle: biogeographic evidence. Science 245: 173-175. Motschulsky, V. 1844. Bemerkungen zu dem im V'™Bande der Zeitschrift fiir die Entomologie p. 192 von H."" Msrkel gegebenen "Beitrsge zur Kenntniss der unter Ameisen Lehenden Insekten." Bulletin de la Societe Imperiale des Naturalistes de Moscou 17: 812-823. Nixon, K. C. and J. I. Davis. 1991. Polymorphic taxa, missing values and cladistic analysis. Cladistics 7 :233-241. Otero, J. C. and J. A. Diaz-Pazos. 1995. Revision of the genus Spaniopbaenus Reitter, 1875 (Coleoptera: Cryptophagidae). Koleopterologische Rundschau 65: 183-188. Packard, A. S. 1864. The bumble-bees of New England and their para- sites, with notes of a new species of Antheropihorabia. and a new ge- nus of Proctotrupidae. Proceedings of the Essex Institvite 4: 107- 140. Pagel, M. 1994. The adaptationist wager. Pp. 29-51, in Eggleton, P and R.I. Vane-Wright (eds.). Phytogenetics and Ecalogi/. New York, Aca- demic Press. Pogue, M. C. and M. F Mickevich. 1990. Character definitions and char- acter state delineation: the bete noire of phylogenetic inference. Cla- distics 6: 319-361. Perris, E. 1869-1870. Notes on Antherophagus nigricornis. L'Abeille 7: 9- 10,25-27. Perris, M. E. 1876 (1875). Larves de Coleopteres. Annales de la Societe Linneenne de Lyon. 590 pp. -i- 14 pis. Plath, O.E. 1922. Notes on the nesting habits of several North American bumblebees. Psyche 29: 189-202. Reitter, E. 1875. Revision der europaischen Cryptophagiden. Deutsch Entomologische Zeitschrift 19: 1-86. Reitter, E. 1889. Zwei neue Coleopteren-Gattungen aus Transkaukasien. Wiener Entomologishe Zeitung 8: 289-292, + pit. IV. Ricklefs, R.E. and D. Schluter. Species Diversity in Ecological Communities: Historical and Geographical Perspectives. Chicago, The University of Chicago Press. 414 pp. Roff D.K. 1 990. The evolution of flightlessness in insects. Ecological Mono- graphs 60: 389-421 . Roubik, D W. and Q. D. Wheeler. 1982. Flightless beetles and stingless bees: phoresy of scotocryptine beetles (Leiodidae) on their meliponine hosts (Apidae). Journal of the Kansas Entomologocal Society 55: 125-135. Sanderson, M. J. and M. J. Donoghue 1994. Shifts in diversification rate with the origin of Angiosperms. Science 264: 1590-1593. Schwarz, E.A. and H. Ulke. 1891. [Notes on Emphylus americanus]. Pro- ceedings of the Entomological Society of Washington 11: 227. Scott, H. 1920. Notes on the biology of some inquilines and parasites in a nest of Bombus derhainellus Kirby; with a description of the larva and pupa of Epuraea depressa Illig. (= aistiva Auctt.: Coleoptera, Nitidulidae). Transactions of the Entomological Society of London 1920: 99-127. Seevers, C.H. 1957. A monograph on the termitophilous Staphylinidae (Coleoptera). Fieldiana: Zoology 40: 1-334. Seevers, C.H. 1965. The systematics, evolution and zoogeography of sta- phylinid beetles associated with army ants (Coleoptera, Cryptophagid Symbiosis 101 Staphylinidae). Fieldiana: Zoology 47: 139-351. Seidlitz, G. 1869-70. (Note on Anthcropha<^us nigricornis). L'Abielle 7: 62- 63. Simpson, G.G. 1944. Tempo and Mode in Evolution. New York, Columbia University Press. 247 pp. Simpson, G.G. 1953. Tlie Mnjor Features of Evolution. New York, Columbia University Press. 434 pp. Smith, J.B. 1909. Annual report of the New Jersey State Museum, includ- ing reports on the Insects of New Jersey. Trenton, MacCrellish and Quigley Smith, F. 1855. Catalogue of the British Hymenoptera, Part I. Apidae - bees. London, Taylor and Francis. 252 pp. + 10 pits. Spahr, U. 1981. Systematischer Katalog der Bernstein- und Kopal- Kafer (Coleoptera). Stuttgarter Beitrage Naturkunde (Ser B; Geologie und Palaontologie) 80: 1-107. Steidle, J.L.M. and K. Dettner. 1993. Chemistry and morphology of the tergal gland of freeliving Aleocharinae (Staphylinidae, Coleoptera). Systematic Entomology 18: 149-168. Thayer, M.K. 1992. Discovery of sexual wing dimorphism in Staphylinidae (Coleoptera): "Omalium" flavidum, and a discussion of wing dimor- phism in insects. Journal of the New York Entomological Society 100: 540-573. Trautmann, G. 1915. Ein aussert seltener Fund: Anthewpliagus. Latr. ni^^ricornis, Fabr., an einer lebenden Hummel. Internationale Entomologische Zeitshcrift Guben 9: 59. Tuck, W. H. 1896. Inquiline and other inhabitants in nests of Aculeate Hymenoptera. Entomologists Monthly Magazine 32: 153-155. Tuck, W. H. 1897. Coleoptera, etc., in nests of Aculeate Hymenoptera. Entomologists Monthly Magazine 33: 58-60. Vrba, E.S. 1980. Evolution, species and fossils: How does life evolve? South African Journal of Science 66: 61-84. Wasmann, E. 1894. Kritisches Verzeichniss der myrmecophilen und termitophilen Arthropoden. Mit der Lebensweise und mit Beschreihungen neuer Arten. Berlin, Felix L. Dames. 231 pp. Wasmann, E. 1906. Beispiele rezenter Artenbildung bei Ameisengasten und Termitengasten. Biologisches Zentralblatt 26: 565-580. Wenzel, J.W. and J.M. Carpenter. 1994. Comparmg methods: adaptive traits and tests of adaptation. Pp. 79-101, m Eggleton, P. and R. I. Vane-Wright (eds.), Phylogenetics and Ecologi/. New York, Academic Press. Wenzel, J.W. 1997. When is a phylogenetic test good enough? Pp. 31-45, /); Grandcolas, P. (ed.). Tlie Origin of Biodiversity in Insects . Memoires Museum Histoire Naturelle no. 173. Wheeler, W. M. 1919. The phoresy oi Antlierophagus. Psyche 26: 145-152. Wheeler, W.M. 1910. Ants: Their Structure, Development and Behavior. New York, Columbia University Press, xxv + 663 pp. Westneat, M. 1995. Feeding, function, and phytogeny: analysis of histori- cal biomechanics in labrid fishes using comparative methods. Sys- tematic Biology 44: 361-383. Wiley, E. 0. 1981 . Phylogenetics: The Tlieory and Practice ofPln/logenetic Sys- tematics. New York, J. Wiley and Sons. 439 pp. Williams, P. H. 1985. A preliminary cladistic investigation of relation- ships among bumble bees (Hymenoptera, Apidae). Systematic En- tomology 10: 239-255. Wilkinson, M. 1995. A comparison of two methods of character construc- tion. Cladistics 11: 297-308. Wilson, E.O. 1971. The Insect Societies. Cambridge, Harvard University Press. 548 pp. Byers, G. W., R. H. Hagen, and R. W. Brooks ieds.), Entomological Contributions in Memory of Byron A. Alexander. University of Kansas Natural Historx/ Museum Special Publication 24. (1999) Pp. 103-109 Timing of Mating Flights of Neotropical African and European Honey Bee Queens and Drones (Hymenoptera: Apidae) in Eastern Venezuela By Gard W. Otis'-, Orley R. Taylor, Jr.\ Marla Spivak'', Mark L. Winston^ Susan J. Katz", and Penelope F. Kukuk'' ABSTRACT Mating flight characteristics of reproductives of neotropical African and European honey bees (Apis mellifera L.) were studied in eastern Venezuela. For queens, mean exit times of all flights and midpoints of mating flights differed significantly between the subspecies, with the midpoints of mat- ing flights by European queens occurring an average of 47 min earlier in the afternoon. Queens of the two subspecies did not differ in other aspects of their orientation and mating flights. Flights of Euro- pean drones also occurred significantly earlier than those of neotropical African drones. Mating flights of queens and drones of the same subspecies were approximately synchronous. These results are suf- ficient to explain the weak positive assortative mating that has been reported previously. Ke\/ Words: Neotropical African bees; Apis luellifcrn; Assortative mating; Drone; Honey bee; Queen; Mating. INTRODUCTION The introduction of African honey bees (Apis tiiellifera sciitellata) to South America has resulted in the remark- able colonization by their descendants of a vast area from Argentina to the southern United States. Prior to 1956, only honey bees of European origin (EHBs) had been imported into Latin America (Winston, 1992; Rinderer et al., 1993). However, as the introduced African bees became estab- lished, the characteristics of feral and managed honey bees suggested that their drones predominated in matings with both managed EHBs and feral neotropical African honey bee (NAHB") queens (Nogueira-Neto, 1964; Hellmich et al., 1988; Taylor and Rowell, 1988; Hall, 1990; Taylor et al, 1991; Echazarreta, 1993; Taylor, in press), although more slowly or to a lesser extent in areas with high populations of EHBs (Rinderer et al, 1991; Quezada-Euan and Hinsull, 1995). These studies suggest that NAHB drones have a mating advantage. Factors which may contribute to this advantage include greater production of drones by NAHB colonies (Rinderer et al., 1987; Spivak, 1992; Echazarreta, 1993), seasonal differences in the production of drones (Echazarreta, 1993), longer and/or more mating flights by NAHB drones (Echazarreta, 1993), possible suppression of drone production in EHB colonies by NAHB drone para- sitism (Rinderer et al., 1985; Rinderer and Hellmich, 1991), and differences in the spatial (Rowell and Taylor, 1988; Taylor and Rowell, 1988) and temporal (Hellmich and Collins, 1990; Hellmich et al., 1991; Collins and Mbaya, 1994) distributions of queens and drones of each subspe- cies at the time of mating. In contrast, Kerr and Bueno (1970) reported no mat- ing advantage of NAHB drones when NAHB and EHB queens mated in an isolated site containing equal num- bers of NAHB and EHB drones. They found that Italian queens mated with Italian drones (64.8% of matings), and NAHB queens with NAHB drones (58.5% of matings), more often than would be expected if mating were ran- dom. Their results clearly indicated weak positive assor- tative mating, but the behavior underlying the phenom- enon was not evident. Based on several sets of unpublished observations (some reported here), Taylor (1985) stated that there is a difference in the timing of mating flights of EHB and NAHB drones. Hellmich and associates confirmed that EHB drones generally flew earlier in the day than did NAHB drones in western Venezuela (Hellmich, 1987; Hellmich and Collins, 1990; Hellmich et al, 1991). They reported a difference of 17-19 min in mean flight times. Department of Environmental Biology, University of Guelph, Guelph, ONT, NIG 2W1, CANADA. E-mail: gotis@evbhort.uoguelph.ca Address for correspondence. Department of Entomology, University of Kansas, Lawrence, KS 66045. Department of Entomology, University of Minnesota, St. Paul, MN 55108. Department of Biological Sciences, Simon Fraser University, Burnaby, BC, V5A 1S6, CANADA. Division of Biological Sciences, University of Montana, Missoula, MT 59812. We use the term "African" honey bee to refer to A. in. scutellnta imported from South Africa and Tanzania. Following terminology suggested in Page's (1989) title, we have called the descendents of African bees introduced into the New World "neotropical African" honey bees (NAHBs). For convenience, we refer to NAHBs and European honey bees (EHBs) as "subspecies." 103 104 The University of Kansas Natural History Museum Special Publication No. 24 with the difference greater for immature than for mature drones. In southern Texas, all drones flew much later in the day in mid-summer than in spring, and the differences between the two subspecies were less pronounced (Collins and Mbaya, 1994). Preliminary studies by one of us (PFK) as early as 1977 indicated a subspecific difference in the timing of queen mating flights as well. Partial temporal segregation of the mating flights of EHB and NAHB drones combined with the synchronization of queen and drone flights within each subspecies might explain the results of Kerr and Bueno (1970). The only published data for queens indicated that their flight distributions differ significantly between the two subspecies, but also documented unex- pectedly poor synchrony between EHB queens and drones (Hellmich and Colhns, 1990). In this study we compare the characteristics of mating flights of queens and drones of EHBs and NAHBs, pro- vide further documentation of the temporal segregation between the subspecies, and discuss the importance of our results with respect to the assortative mating reported by Kerr and Bueno (1970) and the "Africanization" of honey bees in tropical America. Acknowledgments This paper is dedicated in memory of our colleague and friend, Byron Alexander. Byron's unselfish assistance to his students and colleagues was highly valued and will be missed. We wish to thank the Ministerio de Agricultura y Cria de Venezuela for providing lodging, a vehicle, equipment, and some of the bees, and Ricardo Gomez for his assis- tance and hospitality. Nestor Alvarez, Abbie Zeltzer, and Gary Pundt assisted with the field work. Don Hamilton assisted with figure preparation and Debbie MacDonald typed earlier versions of this manuscript and the tables. We appreciate the constructive feedback of Dawn Bakker, Judy Wearing-Wilde, and four anonymous reviewers. This research was supported by USDA Cooperative Agreement 58-7B30-8-7 and Research Agreement 12-14-7001-363 (to ORT), and a grant from Idaho State University (to MLW). MATERIALS AND METHODS Observations of mating flights were made at a field station maintained by the Ministerio de Agricultura y Cria de Venezuela, at Laguna Grande, 16 km E of Maturin, Monagas, Venezuela (9°47'N, 63°4'W). Data for queens were obtained from 29 April to 21 July 1980. Daylength during this period ranged from 12h 25 min to 12h 39 min. Temperatures during the flight times were 26-35 °C, winds were generally light (<2m/sec), and cloud cover ranged from clear to overcast in the afternoons. Observations on several days toward the end of this period were interrupted by rains. Several procedural steps assured that queens of both subspecies were observed under similar conditions. All queens being observed at any one time were reared to- gether using standard queen-rearing techniques. Conse- quently, the resulting virgin queens were approximately the same age (although NAHB queens develop slightly more rapidly; DeGrandi-Hoffmann et al., 1998). Five sets of these queens were established sequentially in mating nuclei. Each set consisted of two NAHB and two EHB colo- nies observed simultaneously by two observers. The EHB queens (n=10) were daughters of Italian stocks (8 cjueens from two different commercial produc- ers) and a Carniolan stock (2 queens) imported from the USA. The virgin NAHB queens (n=10) were reared from a stock established from feral colonies that exhibited worker cell size (diameter: 4.6-5.0 mm; see summaries by Rinderer et al., 1986, and Spivak et al., 1988), morphological traits (small size, all black abdominal tip), and behavioral char- acteristics (e.g., rapid movement on combs) typical of NAHBs. The NAHB colonies were maintained in a site >60 km from the nearest apiaries that may have contained EHBs, in order to minimize possible influences of EHBs on our NAHB stock. Queen cells or virgin queens <48 h post adult emer- gence were introduced into 5-frame nucleus colonies of 3,000-5,000 bees of the same race. The virgin queens ei- ther eclosed directly into the colonies or were allowed to emerge into tubular wire cages from which they were re- leased 1-2 d after emergence. The entrance of each hive was reduced and fitted with a clear plastic tube 2.5 cm in diameter affixed to a landing platform 10 cm wide which facilitated observations of queens. Beginning when queens were <5 d old, and continuing until queen flights had ceased for at least 2 d, we observed each colony from 13:00- 17:00 h (except during rains) and recorded all flights. Be- cause queen flights were not restricted in any way, con- tinuous observation was required, which limited the num- ber of queens for which we could obtain data. Following the convention established by others (e.g., Oertel, 1940; Roberts, 1944; Ruttner, 1985), we distinguished mating flights either by the presence of a "mating sign" attached to the queen or by a flight duration of >10 min. In our study, only 2 of the 26 queens that returned with a mating sign took flights that lasted less than 10 min, and only one flight (18 min duration) was scored as a mating flight in the absence of a mating sign. For some variables the num- ber of observations per queen was unequal; in these in- stances we computed averages for each queen prior to cal- culating the overall means and Student's lvalues. The two values for sample size we report indicate the number of queens and the total number of flights that provided in- formation for the analysis. Mating Flights of African and European Honey Bee Queens 105 Table 1. Comparison of mating and non-mating flights of European (EHB) and neotropical African (NAHB) queens in eastern Venezuela. Data represent means ± standard deviations; sample sizes in parentheses indicate the number of queens for which data were obtained and the total number of flights. Statistical probabilities represent the results of Student's f-tests. 106 The University of Kansas Natural History Museum Special Publication No. 24 Ol2 UJ CL CXI LU Z O 6 O LU O 2 LU ^0 DRONE FLIGHTS —EUROPEAN N = 223 •--•AFRICAN . N = 1,613 •—m^.^ QUEEN FLIGHTS EUR IN = 17) • • • AFR|N = 10| J I I L _L 1:35 2:05 2:35 3:05 3:35 TIME 4:05 4:35 5:05 Fig. 1. Temporal distributions of mating flights by queens and drones of European and neotropical African honey bees near Maturin, Venezuela. The drone flight distribution (above) was obtained by combining the data for all marked drones of known race trapped between 1-20 January 1980, in several drone congregation areas 0.6_3.4 km from the experimental apiary. Mean capture times were 13:11 h (EHB) and 15:35 h (NAHB). The midpoints of each queen mating flight obtained between 29 April-21 July 1980, are plotted (below). Mean queen flight times are 15:11 h (EHB) and 15:58 h (NAHB). The time scale applies to distributions for both drones and queens. The pattern of drone flights was similar on all days of observation. Drones of both subspecies often started and stopped flying at approximately the same times. However, the number of EHB drones at DCAs increased more rap- idly and peaked earlier than for NAHB drones, resulting in differences of 21-38 min in the mean times of drone flights. The top part of Fig. 1 demonstrates this partial seg- regation of the drone flight distributions of EHB and NAHB drones (data from drones trapped in DCAs between 1-20 January 1980). Overlying this general pattern were sea- sonal differences in the timing of drone flights (Table 2). Mean times of flights in January and June occurred at about the same hour relatively early in the day. In contrast, in March/ April, during the height of the hot, windy dry sea- son, the distributions of drone flights shifted to 30-50 min later in the day. Within each subspecies, queen and drone mating flights were synchronized. For EHBs, the mean times of departure of queens ( = 15:03 h) and drones ( = 14:57 h, June data) from hives were very similar Similar synchrony was observed for NAHB reproductives (queen departure: = 15:47 h; drone departure: = 15:35 h). Note that data for drones may include orientation flights which tend to oc- cur earlier in the flight period. Mating Flights of African and European Honey Bee Queens 107 DISCUSSION We observed a substantial temporal difference (47 min) between the subspecies in the mean midpoint of queen mating flights. Similar differences between EHB and NAHB queens were recorded by Hellmich and Collins (1990) in western Venezuela during the dry season (56 min; calculated from data in their figure) and by one of us (PFK, unpublished observations) in French Guiana during the dry season (42 min). Taken together, these results demon- strate differences in the temporal pattern of mating behav- ior by the two types of queens and their colonies at each of the sites. The timing of queen mating flights also varies substantially as a function of seasonal environmental con- ditions. For example, the mean mating flight times (solar times) of queens we recorded in eastern Venezuela during the early rainy season (EHB: 14:59 h; NAHB: 15:46 h; daylength 12 h 38 min) were approximately 40 min ear- lier than those recorded by Hellmich and Collins (1990) in the dry season in western Venezuela (EHB: 15:34 h; NAHB: 16:31 h; daylength 12 h 3 min). (For this comparison only, local times were converted to solar times by subtracting 4 min for each 1° west of longitude 60°W). Mating flights of EHB drones occur earlier in the day than those of NAHB drones, as also shown by Hellmich and Collins (1990), Hellmich et al. (1991), and Collins and Mbaya (1994). However, the magnitude of the difference in mean flight times recorded for drones of the two sub- species is less than that for queens, particularly under windier (dry season, Venezuela) and /or hotter (summer, Texas) climatic conditions (21-38 min. Table 2; 19 min, Hellmich et al, 1991; 17 min, calculated from Hellmich and Collins, 1990, Fig. 1; 18 min, calculated from partial March data of Collins and Mbaya, 1994; 5 min, calculated from June/July data of Collins and Mbaya, 1994). Some of the influences of daylength, temperature, winds, overcast skies and rain on mating flights are discussed by Alber et al. (1955), Taber (1964), Verbeek (1976), Fletcher and Tribe (1977), Jung (1981), and Lensky and Demter (1985). During our period of queen observation, mating flights of queens and drones of the same subspecies were syn- chronized. Others have commented on the coincidence of queen and drone mating flights within populations (re- viewed by Koeniger, 1991; see also Verma et al., 1990; Yoshida et al., 1994; and Yoshida, 1995). Such synchrony is to be expected because selection should favor queens that take mating flights at times of high drone abundance, re- sulting in fewer or shorter mating flights and therefore lower rates of predation as a consequence (Hellmich and Collins, 1990). Because synchrony should result from se- lection on both sexes, males should be selected to take mating flights when queens are available for mating. Our data from the early wet season showing synchrony within each subspecies in queen and drone flights do not agree with the dry season observations of Hellmich and Collins (1990). There are no obvious explanations for this discrep- ancy between our results and theirs, or for the lack of syn- chrony in European queen and drone flights that they re- ported. Partial temporal segregation to the extent we observed, combined with synchronization of queen and drone flights within each subspecies and with equal numbers of EHB and NAHB drones, can account for the positive assorta- tive mating demonstrated by Kerr and Bueno (1970). A change in any of these three parameters will influence the degree to which one subspecies is favored in mating. Sev- eral other variables could influence this phenomenon, but our research suggests that they are relatively unimportant. For example, there is no obvious spatial separation ofEHB and NAHB drones; when trapping drones, drones of both subspecies were either present or absent at any particular location, and their proportions were approximately the same at different DCAs (ORT and GWO, unpublished data). There is no apparent preference of drones for c]ueens of their own subspecies; drones of each subspecies mated with plastic tubes containing confined queens that were suspended in a OCA (Taylor, 1984b) at the same frequency as they were trapped in flight (Taylor, 1984a) in the same DCA (ORT, unpublished observations). No one has evalu- ated mate choice by honey bee queens. The results of Kerr and Bueno (1970) relate to a popu- lation-level phenomenon. They estimated mating frequen- cies (64.8% EHB X EHB matings; 58.5% AHB x AHB matings) based on characteristics of worker progeny of many queens. In that study, the Italian queens mated with an average of 5.3 males. Most of the EHB queens they stud- ied mated with both EHB and NAHB drones. Relatively few EHB queens would be predicted to mate exclusively with EHB (e.g., 0.648'-'- 10.0%) or AHB (0.352" = 0.4%) drones (although they obtained both these extreme results). Examination of Figure 1 clarifies this phenomenon. Euro- pean queens flying early would encounter a preponder- ance of European drones; late-flying queens would encoun- ter predominantly Africanized drones. On average, how- ever, the partial temporal separation of flights by Euro- pean and Africanized reproductives would result in posi- tive assortative mating, but only to a relatively small ex- tent because the distributions overlap broadly. Unlike the experimental conditions of Kerr and Bueno (1970), NAHB drones available for mating usually greatly outnumber EHB drones in tropical regions. The factors that influence the relative proportions of NAHB and EHB drones encountered by queens (e.g., higher rates of drone production, earlier drone production, more mating flights, longer mating flights, and greater population size of 108 The University of Kansas Natural History Museum Special Publication No. 24 NAHBs; Rinderer et al., 1987; Winston, 1992; Echazarreta, 1993) influence matings of queens much more than the minor temporal difference between the subspecies in the timing of mating. In this regard we reached the same con- clusion as Hellmich et al. (1991). However, we do not agree with the simplified prediction of Hellmich et al. (1991 ) that mixed mating between EHBs and NAHBs will necessarily lead to gene flow, the complete mixing of their genomes over time, and eventual homogenization of their mating flight distributions. The historical evidence from tropical regions (Sheppard et al, 1991a; Hall, 1992) and studies showing hybrid dysfunction (Harrison and Hall, 1993) cio not favor the racial mixture hypothesis. More likely out- comes are the complete replacement of one race by the other in most regions (Smith, 1991; Hall, 1992) and the es- tablishment of relatively permanent hybrid zones (Sheppard et al., 1991b; Taylor, in press). LITERATURE CITED Alber, M., R. Jordan, F. Ruttner, and H. Ruttner. 1955. Von der Paarung der Honigbienen. Zeitschrift fiir Bienenforschung 3: 1-28. [German, German summary] Collins, A. M. and J. S. K. Mbaya. 1994. Drone flight times in South Texas: AHB and EHB. American Bee Journal 134: 830. DeGrandi-Hoffmann, G., J. C. Watkins, A. M. Collins, G. M. Loper, J. H. Martin, M. C. Arias, and W. S. Sheppard. 1998. Queen development time as a factor in the Africanization of European honey bee (Hy- menoptera: Apidae) populations. Annals of the Entomological So- ciety of America 91: 52-58. Echazarreta, C. M. 1993. The Biology and Reproductive Behnviour of Africanized and European Honeij Bee Drones (Apis mellifera L.) in Tropi- cal Yucatan Mexico. Ph.D. Thesis, University of Wales College of Cardiff, UK. 258 pp. Fletcher, D. J. C. and G. D. Tribe. 1977. Natural emergency queen rearing by the African bee A. m. adansonii and its relevance for successful queen production by beekeepers. II. Pp. 161-168 in D. J. C. Fletcher (ed.), African Bees: Taxonomy, Biology and Economic Use. Pretoria, Apimondia. Hall, H. G. 1990. Parental analysis of introgressive hybridization between African and European honev bees using nuclear DNA RFLPs. Ge- netics 125: 611-621. Hall, H. G. 1992. DNA studies reveal processes involved in the spread of New World African honey bees. Florida Entomologist 75: 51-59. Harrison, J. F. and H. G. Hall. 1993. African-European honeybee hybrids have low nonintermediate metabolic capacities. Nature 363: 258- 260. Hellmich, R. L. 1987. Flight time differences between Africanized and European drones: implications for controlling matings. American Bee Journal 127: 846. Hellmich, R. L. II and A. M. Collins. 1990. A nuptial disadvantage for European honey bees in Venezuela. Pp. 141-142, in G. K. Vereesh, B. MaUik, and C. A. Viraktamath (eds.). Social Insects and the Envi- ronment, Proceedings of the 11th International Congress of the In- ternational Union for the Study Social Insects, 1990. New Delhi, Oxford and IBH. Hellmich, R. L. II, A. M. Collins, R. G. Danka, and T. E. Rinderer. 1988. Influencing matings of European honey bee queens in areas with Africanized honey bees (Hymenoptera: Apidae). Journal of Eco- nomic Entomology 81: 796-799. Hellmich, R. L. II, T E. Rinderer, R. G. Danka, A. M. Collins, and D. L. Boykin. 1991. Right times of Africanized and European honey bee drones (Hymenoptera; Apidae). Journal of Economic Entomology 84:61-64.' Jung, G. 1981. Climate and mating. Bienenvater 102: 71-74 (Apicultural Abstract #840/82). [In German] Kerr, W. E. and D. Bueno. 1970. Natural crossing between Apis mellifera adansonii and Apis mellifera ligustica. Evolution 24: 145-148. Koeniger, G. 1991. Diversity in /Ip/s mating systems. Pp. 199-212, /» D. R. Smith (ed.). Diversity in the Genus Apis. Boulder, Westview Press. Lensky, Y. and M. Demter. 1985. Mating flights of the queen honey bee CApis mellifera) in a subtropical climate. Comparative Biochemistry and Physiology 81 A: 222-241. Nogueira-Neto, P. 1964. The spread of a fierce African bee in Brazil. Bee World 45: 119-127. Oertel, E. 1940. Mating flights of queen bees. Gleanings in Bee Culture 68: 292-293, 332-333. Page, R. E., Jr. 1989. Neotropical African bees. Nature 339: 181-182. Quezada-Euan, J. J. G. and S. M. HinsuII. 1995. Evidence of continued European morphometries and mtDNA in feral colonies of honey bees [Apis mellifera) from the Yucatan Peninsula, Mexico. Journal of Apicultural Research 34: 161-166. Rinderer, T E., A. M. Collins, R. L. Hellmich II, and R. G. Danka. 1987. Differential drone production by Africanized and European honey bees. Apidologie 18: 61-67. Rinderer, T. E. and R. L. Hellmich II. 1 991 . The processes of Africanization. Pp. 95-117, in M. Spivak, D. J. C. Hetcher, and M. D. Breed (eds.). The "African" Honey Bee. Boulder, Westview Press. Rinderer, T E., R. L. Hellmich II, R. G. Danka, and A. M. Collins. 1985. Male reproductive parasitism: a factor in the Africanization of Eu- ropean honey bee populations. Science 228: 1119-1121. Rinderer, T E., B. P Oldroyd, and W. S. Sheppard. 1993. Africanized bees in the U.S. Scientific American 269 (6): 84-90. Rinderer, T. E., J. A. Stelzer, B. P Oldroyd, S. M. Buco, and W. L. Rubink. 1 991 . Hybridization between European and Africanized honey bees in the neotropical Yucatan Peninsula. Science 253: 309-311. Rinderer, T E., H. A. Sylvester, M. A. Brown, J. D. Villa, D. Pesante, and A. M. Collins. 1986. Field and simplified techniques for identifying Africanized and European honey bees. Apidologie 17: 33-48. Roberts, W. C. 1944. Multiple mating of queen bees proved by progeny and flight tests. Gleanings in Bee Culture 72: 255-259, 303. Rowell, G. A. and O. R. Taylor, Jr. 1988. Some computer simulations us- ing the neutral mating model for the honey bee. Apis mellifera. Pp. 184-192, m G. R. Needham, R. E. Page, M. Delfinado-Baker, and C. E. Bowman (eds.), Afi'icanized Honey Bees and Bee Mites. Chichester, Ellis Horwood, Ltd. Ruttner, F. 1985. Reproductive behaviour in honey bees. Pp. 225-236, in B. HoUdobler and M. Lindauer (eds.). Experimental Behavioral Ecol- ogii and Sociobiolog}). Sunderland, Sinauer Associates, Inc. Sheppard, W. S., A. E. E. Soares, D. Dejong, and H. Shimanuki. 1991a. Hybrid status of honey bee populations near the historic origin of Africanization in Brazil. Apidologie 22: 643-652. Sheppard, W. S., T E. Rinderer, J. A. Mazzoli, J. A. Stelzer, and H. Shimanuki. 1991b. Gene flow between African- and European-de- rived honey bee populations in Argentina. Nature 349: 782-784. Smith, D. R. 1991 . African bees in the Americas: insights from biogeogra- phy and genetics. Trends in Ecology and Evolution 6:17-21. Spivak, M. 1992. The relative success of Africanized and European honey- bees over a range of life-zones in Costa Rica. Journal of Applied Ecology 29: 150-162. Spivak, M., T. Ranker, O. R. Taylor, Jr., W. Taylor, and L. Davis. 1988. Discrimination of Africanized honey bees using behavior, cell size, morphometries, and a newly discovered isozyme polymorphism.Pp. Mating Flights of African and European Honey Bee Queens 109 313-324, (/; G. R. Needham, R. E. Page, Jr., M. Delfinado-Baker, and C. E. Bowman (eds.), Africaiuzed Honc\/ Bct's and Bee Mitef. Chichester, Ellis Horwood, Ltd. Taber, S. 1964. Factors influencing the circadian rhythm of drone honey bees.Annals of the Entomological Society of America 57: 769-775. Tavlor, O. R., Jr 1984a. An aerial trap for collecting drone honey bees in congregation areas. Journal of Apicultural Research 23: 18-20. Taylor, O. R., Jr. 1 984b. A mating tube for studymg attractiveness of queen honey bees and mating behaviour of drones. Journal of Apicultural Research 23: 21-24. Taylor, O. R., Jr. 1985. African bees: potential impact in the United States. Bulletin of the Entomological Society of America 31: 15-24. Taylor, O. R., Jr. In press. Displacement of European honeybee {Apis mcUifera) subspecies by an invading African subspecies in the Ameri- cas. In L. J. Connor and R. Hoopingarner (eds.). Beekeeping in the 21st Century. Cheshire, CT, Wicwas Press. Taylor, O. R., A. Delgado, and F. Brizuela. 1991. Rapid loss of European traits from feral neotropical African honey bee populations. Ameri- can Bee Journal 131: 783-784. Taylor, O. R., Jr. and G. A. Rowell. 1988. Drone abundance, queen flight distance, and the neutral mating model for the honey bee. Apis mellifera. Pp. 173-183, i» G. R. Needham, R. E. Page, Jr., M. Delfinado- Baker, and C. E. Bowman (eds.). Africanized Honey Bees and Bee Mites. Chichester, Ellis Horwood, Ltd. Verbeek, B. 1976. Untersuchungen der Ausflugsaktivitat von jungen Bienenkoniginnen unter Festlands und Inselbedingungen Mittels lichtelektronischer Uberwachung. Apidologie 7: 151-168. [German, German summary! Verma, L. R., B. S. Rana, and V. K. Mattu. 1990. Mating behaviour of Apis cerana Fabr. and Apis mellifera Linn, in the Simla Hills of the north- west Himalayas. Journal of Insect Science 3: 130-132. Winston, M. L. 1992. The biology and management of Africanized honey bees. Annual Review of Entomology 37: 173-193. Yoshida, T. 1995. Comparative studies on the mating system of Japanese honeybee, Afiis cerana japonica Radoszkowski and European honey- bee Apis mellifera L (Hymenoptera: Apidae). Bulletin of the Faculty of Agriculture, Tamagawa University 35: 159-208. Yoshida, T, J. Saito, and N. Kajigaya. 1994. The mating flight times of native Apis cerana japionica Radoszkowski and introduced Apis mellifera L. in sympatric conditions. Apidologie 25: 353-360. Si/i'/s, G. W., R. H. Hagcn, ami R. W. Brooks tefs.). Entomological Contributions in Memory of Byron A. Alexander. Uiiiivrsitii ofKausaf Nntural History Mtiseuin Special Publication 24. (1999) Pp. lU-UO Eickwortia (Apoidea: Halictidae), a New Genus of Bees from Mesoamerica By Ronald J. McGinley^ ABSTRACT A new generic name, Eickwortia, is proposed for Halictiis tii/cteris Vachal, and a new species of this taxon, E. alexmidcri, is described. Eickioortin has been rarely collected and is known from only 69 specimens taken at high elevations from central Mexico to Costa Rica. Morphological charac- teristics and limited biological data suggest these bees may be social, and possibly wood-nesting. Suggestions for future collecting and research priorities are offered. Keywords: Eickwortia; Apoidea; Halictidae; Bees; Lasioglossiini. INTRODUCTION Vachal (1904) described Halictiis iii/ctcris based on a syntype series of eight females collected at Orizaba, Veracruz, Mexico, by Biart in 1862. It has remained enig- matic for nearly a century. J.S. Moure visited the Paris Museum, where the type series is deposited, and wrote (Moure and Hurd, 1987): "One female from Orizaba (Biart, 1862), Mexico was labeled as the lectoholotype in March, 1958, by one of us (Moure) and is now so designated. Dr. George C. Eickwort, who has examined the lectoholotype, informs us that this species appears to represent a new genus near Neocorynura." Eickwort kindly provided me with a copy of his 1975 notes from the Paris Museum, which include the following entry: "H. nycteris V. female holotype + 4 females, n.g. near Neocorynura. homotype Berkeley." Eickwort later recognized (pers. comm., 1986) that H. nycteris was not an augochlorine but still thought it represented a new genus of Halictini, near Evylaeus. Eickwort and I had planned to co-designate a new genus for H. nycteris, but his untimely passing in 1994 prevented this action. It is in his honor that I propose this new ge- neric name. In addition, a new species of this taxon is de- scribed in honor of the late Byron A. Alexander. Acknowledgments Posthumous thanks go to George Eickwort who shared his opinions with me regarding the systematic status of this taxon. Eickwort also assembled a valuable collection of specimens, including a specimen compared to the type series. After Dr. Eickwort's death, Michael S. Engel and Douglas A. Yanega alerted me to the existence of this ho- motype in the Cornell University Collection. Dr. Yanega also kindly sent me SEM photographs of pollen samples taken from three specimens of £. nycteris, forwarded speci- mens collected by the Programa Cooperativo sobre la Apifauna Mexicana (PCAM), and provided locality data from two specimens he identified at the Illinois Natural History Survey. Charles D. Michener also provided local- ' Department of Entomology, MRC-188, National Museum of Natural H E-mail: mcginley.ron@nmnh.si.edu ity data from specimens in the KU Collection. Joan W. Nowicke verified that the pollen samples taken by Dr. Yanega were most likely from three different plant fami- lies. Elaine R.S. Hodges was responsible for the excellent illustrations, and Maureen ]. Mello assisted with SEM and GIS work and processed the return ofborrowed specimens. Ricardo Ayala helped clarify various locality data. The manuscript was greatly improved by review comments from Robert W. Brooks, George W. Byers and CD. Michener. I would like to thank the following institutions and curators who arranged for loans of specimens, especially Janine Casevitz-Weulersse who processed the loan of type material from the Paris Museum. CAS California Academy of Sciences, San Fran- cisco (W.J. Pulawski) CU Cornell University, Ithaca, New York (B.N. Danforth, M.S. Engel, G.C. Eickwort) INHS Illinois Natural History Survey, Urbana (W.E. LaBerge, D.A. Yanega) LACM Los Angeles County Museum of Natural History, California (R.R. Snelling) KU University of Kansas, Lawrence (B.A. Alexander, R.W. Brooks, CD. Michener) PARIS Museum National d'Histoire Naturelle, Paris (J. Casevitz-Weulersse) PCAM Programa Cooperativo sobre la Apifauna Mexicana (specimens loaned by D. Yanega) UCB University of California, Berkeley (H. V. Daly) UNAM Universidad Nacional Autonoma de Mexico, Mexico City (H. Brailovsky) UNAMC Estacion de Biologia Chamela, Universidad Nacional Autonoma de Mexico (R. Ayala) USU Utah State University, Logan (T.L. Griswold) istory, Smithsonian Institution, Washington, DC 20560. Ill 112 The University of Kansas Natural History Museum Special Publication No. 24 Fig. 1, Eickivortia nycteris, female; insert showing detail of bidentate mandible. Scale = 1mm Eickwortia, new genus As recognized herein, Eickwortia includes two rarely collected species: £. lu/cferis (Vachal) (Fig. 1), known only from 60 females and 8 males, ranging from Nayarit and San Luis Potosi, south to Nicaragua (Fig. 2); and E. alexandcri, new species, known only from one female from Costa Rica. Both species ha\'e been collected only at high elevations (E. mjcteris: 823-2200 m; £. alexanderi: 1510 m). Diagnosis.—Eickivortia females can be differentiated from other Halictidae by the Dialictiis/Evylaeus-iype forewing venation, i.e, second transverse cubital vein weaker than the first cubital vein (more like the third cubital vein; see McGinley, 1986, Fig. 80), combined with the strongly bidentate mandibles of the females (Figs. 6, 12). The mandibles of Dialictus and Evylaeiis females have small subapical teeth. Also helpful in diagnosis are the darkly infuscated forewings (anterior third infuscated in both sexes of £. mjcteris; entirely infuscated in the female of £. alexanderi). Only the male of £. mjcteris is currently known (see following species diagnosis). Both sexes key out to Lasioglossum (sensu lato) in Michener et al. (1994); females run to Evylaeus in their subgeneric key. Eickwortia nycteris (Vachal), new combination Figures 1-11, 15, 16 Hnlictiis nycteris Vachal, 1904:119 [female syntype series]; Moure and Hurd, 1987:308 [lectotype designation, taxonomic status]. Type Material.—The lectotype female specimen col- lected in 1862 is in good condition, missing only its right hind leg. It is pinned on a rectangular piece of white paper and is labeled: "Museum Paris [,] Mexique Orizaba [,] Biart 1862/[circular green piece of paper, folded in half]/TYPE [red label] /nycteris 9 Vach [handwritten, presumably by Vachal] /Halictus nycteris Vach." [in different handwrit- ing]. Moure (Moure and Hurd, 1987) indicated that he la- beled one female of the syntype series "... as the lectoholotype" in March, 1958. This label is no longer as- sociated with the specimen and I have attached the fol- lowing label: "LECTOTYPE [,] Halichis nycteris Vachal, des.[ignated by] Moure & Hurd, 1987." The four known paralectotypes have been labeled: "PARALECTOTYPE [,] Halictus nycteris Vachal [,] des.[ignated by] Moure & Hurd, 1987" [red label]. Vachal's original description indicated he examined eight females. In 1975, Eickwort examined only the lectotype and four paralectotypes which I have EicKWORTiA, New Genus of Halictine Bee 113 Fig. 2. Distribution of Eickwortia (dots = £. nyctcrif, circled dot = E. alexauderi). also seen. The whereabouts of the three other type speci- mens is unknown to me. Etymology.—The generic name, Eickwortia, in honor of the late George C. Eickwort, a premier student of bees and outstanding educator. Vachal's specific epithet, nycteris, is apparently derived from the Greek, in/ktos = "night," or mjkieris = "bat," presumably a reference to the dark coloration and infuscated forewings of this species. Diagnosis.—Both sexes of E. nycteris superficially re- semble Neocoryiium (AugochlorLni) due to the basal con- striction of tergum I. However, they have the diagnostic features of Halictini, i.e., females lack a median cleft on tergum V, and males have the hind basitarsus broadly ar- ticulated with the second tarsomere. The conspicuously infuscated anterior third of the forewing surface, strongly bidentate mandibles, basally constricted tergum I, com- bined with the forewing venation characteristic of Dialictus and Evylaeus, i.e., second transverse cubital vein weaker than first (more like third) differentiate females of Eickwortia nycteris from all other known New World halictids. Males can be recognized by their highly con- stricted abdomen, infuscated wings, and in particular, their elongate and conspicuously slender legs. Description.—(follows format of McGinley, 1986) Female: (1) Length approximately 6.0-11.0 mm; (2) wing length 6.78-9.41 mm; (3) abdominal width 2.18-3.14 mm. [Measurements were taken from what appeared to be the smallest and largest specimens; only ranges are given because of the great size variation in this species.] Structure. (4) Head broad, slightly wider than long (Fig. 5); length/width ratio, x = 0.91, n = 3. (5) Gena, at midpoint, slightly wider than eye width (eye/genal ratio = 0.89) to much wider than eye in large-headed forms (eye/ genal ratio = 0.61 ), [see Nesting Biology section, below, for discussion of "large-headed" terminology]; (6) gena rounded posteriorly to angulate in large-headed forms. (7) Supraclypeal area evenly rounded, (8) weakly protuber- ant. (9) Clypeus much broader than long, width/length ratio = 3.0-3.75; clypeus projecting approximately 0.75 its length below lower margin of eyes; (10) surface weakly convex to flat in large-headed forms, shallowly depressed apically; (11) clypeal surface usually with shallow median longitudinal sulcation, best developed in large-headed forms. (12) Frontal carina present, extending at least to midpoint between antennae and median ocellus. (13) Dis- tance between lateral ocellus and eye less than distance between lateral ocellus and hind margin of vertex, espe- cially so in large-headed forms (ocular-ocellar space ap- proximately 3.5-3.8 lateral ocellar diameter); (14) distance between lateral ocellus and eye greatly exceeding distance between lateral ocelli (ratio = 1.5-2.0 in large-headed forms); (15) lateral ocelli joined above by weak impressed line. (16) Compound eyes slightly converging below, to parallel in large-headed forms. (17) Hypostomal carina extremely well developed; (18) anterior angle broadly rounded, (20) anterior carina nearly perpendicular to lon- gitudinal carina. (21) Scape not quite reaching top of ver- tex; (22) pedicel longer than wide, slightly shorter than flagellomere 1; (23) flagellomere 1 subequal in length to flagellomere 2. (24) Labrum with basal area and distal pro- cess; (25) basal elevation well developed (unlike Lasioglossitm sensu stricto, elevation is broader than long); (26) basal lateral depressions absent; (27) distal keel ex- tremely narrow, becoming slightly broader apically as seen in frontal view (unlike Lasioglossiiin sensu stricto, keel is vertically constricted near basal third, appearing bilobed in lateral view); (28) distal lateral projections absent; (29) fimbrial setae acutely pointed. (30) Mouthparts not unusu- ally modified or elongate; (31) mandible strongly bidentate (Fig. 6). (32) Pronotal lateral angle forming sharply pointed, projecting right angle; (33) pronotal lateral ridge complete; (34) lower portion of lateral ridge sharply edged; (34a) pronotal lobe narrowly rounded and projecting, conspicu- ously pointed and projecting in large-headed forms. (35) Mesoscutal anterior edge weakly bilobed, (36) strongly elevated from pronotum; (37) median mesocutal line mod- erately well impressed to about half length of mesoscutum; (38) parapsidal lines approximately 0.30 the length of mesoscutum. (39) Median scutellar impression virtually absent. (40) Dorsal surface of propodeum about 0.75 the length of scutellum and approximately 1.4 times the length of metanotum, (41) weakly depressed centrally, (42) pos- 114 The University of Kansas Natural History Museum Special Publication No. 24 Figs. 3-4. Eickimrtia uyctcris, male. 3-Abdomen, dorsal view. 4-LateraI view. terior margin rounded; (43) propodeal triangle weakly developed, median V-shaped area absent, lateral rims of propodeal dorsal surface absent; (44) lateral propodeal carinae weakly developed, extendiiig no more than one- third height of posterior surface. (45) Inner hind tibial spur strongly pectinate, with four to five teeth (Fig. 8). (46) Lat- eral edge of metasomal tergum 11 straight. Sculpture. (47) Face somewhat shiny, (48) area be- tween ocelli and antennae finely granulate with conspicu- ous punctures thrtiughout, punctures nearly contiguous above, becoming less dense near antemiae where punc- tures are separated by 2-3 times their diameters. (49) Ver- tex near eye and (30) behind ocelli granulate with fine punctures separated by 1-2 times their diameters. (51) Supraclypeal area dull, densely granulate with obscure widely spaced punctures. (53) Clypeus granulate with widely spaced punctures separated by 3-5 times their di- ameters, apical quarter of clypeus usually less granulate, somewhat polished. (55) Hypostoma finely striolate. (56) Mesoscutum somewhat dull; (57) surface granulate throughout and doubly-punctate: fine punctures separated by 1-2 times their diameters, and larger, conspicuous widely scattered punctures separated by 5-10 times their diameters. (58) Scutellum and (59) metanotum granulate with widely scattered punctures. (60) Pre-episternum roughly striate; (61) hypoepimeral area finely striate, mesepisternum strongly striate, nonpunctate; (62) metepisternum strongly striate. (63, 64) Dorsal surface of propodeum dull, microscopically granulate, usually finely striolate (Figs. 1, 7; 1 examined one specimen from Michoacan that had surface nearly smooth, with striae confined to anterior edge of propodeum). (65) Metasomal tergum I constricted basally (Fig. 1) with distinct bicon- vexities apically (terga Il-lll also with less conspicuous con- vexities); (66) surface granulate throughout with fine widely spaced punctures. Coloration. (67) Head, thorax black, abdomen black with dark brown tones. (68) Clypeus without maculation. (69) Flagellum black dorsally, brown ventrally. (70) Tegula dark brown. (71) Wing membrane strongly infuscated on anterior one-third of surface (Fig. 1); veins and stigma brown. (72) Legs brown. Vestiture. (73) Pubescence of head between vertex and antennae weakly plumose; (74) hairs yellowish brown. (75) Pubescence of thorax white to yellowish white; (76) mesocutal hairs sparse, widely spaced. (77) Hind tibial hairs dark brown to black dorsally, nearly white ventrally. (78) Anterior hairs of metasomal tergum 1 and (79) basal hair band of tergum 11 white. (80) Acarinarium (glabrous area surrounded by elongate hair fringe) on anterior sur- face of tergum I absent. (81) Basal hair bands on terga III- IV usually absent (hair band on tergum III weakly devel- oped in 5 specimens, and strongly developed in one speci- men from Guerrero). Male: As described for female except: (1) Length 8.5- EicKWORTiA, New Genus of Halictine Bee 115 Figs. 5-8. Eickwortia nyctens, female. 5-Head. 6-Mandible. 7-Fropodeum, dorsal view. 8-lnner hind tibial spur. 10.0 mm (x = 9.0, n = 5); (2) wing length 2.2-2.6 mm (x = 2.5, n = 5); (3) abdominal width 1.6-1.8 mm (x = 1.7, n = 5). Structure. (4) Head length and width subequal, or nearly so; length/width ratio, x = 0.96-1.0, n = 3. (5) Gena, at midpoint, subequal in length to eye, (6) rounded poste- riorly. (9) Clypeus approximately twice as wide as long, projecting approximately 0.75 its length below lower mar- gin of eyes; (10) surface broadly rounded and somewhat protruding (unlike broad, flat clypeus of female); (11) clypeal surface without shallow median longitudinal sulcation. (13) Distance between lateral ocellus and eye subequal to distance between lateral ocellus and hind margin of vertex; (14) distance between lateral ocellus and eye only slightly exceeding distance between lateral ocelli (ratio = approximately 1.3); (16) Compound eyes converg- ing below. (17) Hypostomal carina well developed (less so than in females); (18) anterior angle very narrowly rounded, (20) anterior carina forming acute angle with lon- gitudinal carina. (21) Scape short, extending to midpoint between antemial base and median ocellus; (22) pedicel wider than long, half length of flagellomere 1; (23) flagellomere 1 half as long as flagellomere 2. (24) Labrum without basal area and distal process; (25) basal elevation absent; (27) distal keel absent. (31) Mandible very short, barely reaching opposing clypeal angle, simple, subapical tooth absent. (32) Pronotal lateral angle well developed but not as sharply acute as in females; (34a) pronotal lobe narrowly rounded but not projecting or conspicuously pointed as in females. (45) Inner hind tibial spur weakly serrate, not pectinate. (46) Lateral edge of metasomal ter- gum II weakly concave. Sculpture. (48) Area between ocelli and antennae densely punctate throughout, punctures contiguous. (49) Vertex near eye and (50) behind ocelli irregularly striolate, nonpunctate. (51) Supraclypeal area dull, densely granu- late, nonpunctate. (53) Clypeus weakly granulate, some- what shiny, with widely spaced punctures separated by 2-3 times their diameters. (55) Hypostoma weakly striolate. (56) Mesoscutum, (58) scutellum and (59) metanotum 116 The University of Kansas Natural History Museum Special Publication No. 24 11 Figs. 9-11. Scale = 1mm. Eickivortici nycteris, male terminalia. 9-Genitalia, ventral view. 10-Genitalia, dorsal view. 11-Stema VII-VIII, ventral view. somewhat dull, surface densely and roughly granulate throughout, with minute nearly contiguous punctures. (60) Pre-episternum and (61) hypoepimeral area ruguloso- striolate; (62) metepistemum more distinctly striolate. (65) Metasomal tergum I constricted basally, with inconspicu- ous biconvexities apically (much weaker than those of fe- males). Coloration. (67) Head and thorax black, abdomen black dorsally, sterna, lateral edges of terga and basal quar- ter of tergum I orange-brown. (68) Apical half of clypeus with broad yellow maculation. (69) Flagellum black dor- sally, orange-brown ventrally. (70) Tegula brown to yel- lowish brown. (71) Wing membrane strongly infuscated on anterior one-third of surface (similar to female. Fig. 1); veins and stigma brown. (72) Femur dark brown to black with distal quarter orange-brown; tibiae light brown to black dorsally, orange-brown ventrally; tarsi light brown to orange-brown. Vestiture. (81) Basal hair bands on terga III-IV present or absent (present in approximately half the specimens examined). (82) Sternum IV with short, adpressed hairs on apical edge; (83) sternum V with moderately elongate fringe of hairs on apical edge (not forming a distinct pat- tern). Terminalia. Sterna VII-VIII as in Fig. 11; (84) Sternum Vll with elongate, slender median process, lateral lobes broad; (85) sternum VIII broad, without median process. Genitalia as in Figs. 9, 10; (86) gonobase short; (87) EicKwoRTiA, New Genus of Halictine Bee 117 gonostylus moderately elongate, rounded apically; (88) retrorse membranous lobe present, moderately short, slen- der, narrowly rounded apically; (90) volsella with promi- nent lateral lobe. Specimens Examined.—68 (60 females, 8 males). EL SALVADOR. Mt. San Salvador, 8 Jul 1963, M.E. Irwin, D.Q. Cavagnaro (1 female; UCB). GUATEMALA. Escuintla, 6.3 mi NE [13°42'N, 89°12'W], 1 Aug 1966, Univ. Kans. Mex. Expedition (1 male; KU). Finca San Rafael, Sacatpequez, 6900 ft [14°20'N, 90°31'W], 11 Jun 1948, sweeping path in woods, CNHM Guatemala Expedition (1948), R.F. Mitchell (1 female; INHS). Zacapa, 3.5 km S.E. La Union, 1500 m [14°58'N, 89°32'W], 23 June 1993, R. Brooks, J. Ashe #084 (1 female; KU). MEXICO. Chiapas: Lomata, 5 Mar 1953, R.C. Bechtel, E.I. Schlinger, Halictus nycteris Vach, homotype 1975, det. G.C. Eickwort (1 female; UCB). Sibakte 'el, Tenejapa, 5500 ft, 6-8 Aug 1966 [16°49'N, 92°31'W], D.E. Breedlove, J. Emmel (1 female; CAS). Municipio Zinacantan, Paraje Navenchauk, 2194 m, 3 Aug 1976, D.E. & J.A. Breedlove (1 female; CAS). Municipio Zinacantan, Paraje Vobits, 1371 m, 18 Aug 1976, D.E. & J.A. Breedlove (1 female; CAS). Guerrero: Taxco, 7 km E, 1560 m, 29 Oct 1991 [18°33'N, 99°36'W], [ex] Difourca along small dirt path, big patch fl along rd, R. Ayala (1 female; UNAMC). Hidalgo: Chapulhuacan [21°10'N, 98°54'W], 19 Jun 1941, H.S. Dybas (1 female; INHS). Chapulhuacan, 2.4 mi S (Hwy. 85), 2700 ft, 12 Jul 1973, R.R. Snelling, T.W. Taylor (4 females; LACM). Jacala, 8 mi S (La Placita, Hwy 85), 5400 ft, 13 July 1973, Ipomoea, R.R. Snelling, T.W. Taylor (2 females; LACM). Jacala, 10.6 km N, Hwy 85 (km 192) [21°01'N, 99°11'W], 1620 m, 11 Jul 1990, R.L. Minckley (1 female; KU). Jacala, 38 mi NE, 3100 ft, 10 Jul 1961, Univ Kans. Mex. Expedi- tion. Jacala, 32 mi NE, 3950 ft, 10 Jul 1961, on flowers of Bidens, Univ. Kans. Mex. Expedition, (6 females, 2 males; KU). Otongo, 10 km E, 1110 m [20°59'N, 98°42'W], 10 Nov 1991, manganese mine area, composites along rd, R. Ayala, C. Everaert (1 female, 1 male; UNAM). Tenango de Doria, Cerro El Cirio [20°19'24"N, 98°11'52"W], 26 Mar 1994, 13:00 hrs, L. Godinez, LG-980 (1 female; UNAM). Tenango de Doria, La Colonia, 8 Sep 1993, 14:15 hrs, L. Godinez, LG- 818 (1 female; UNAM). Tenango de Doria, El Damo [20°19'30"N, 98°13'38"W], 3 Nov 1993, 10:00 hrs, L. Godinez, LG-761 (1 male; UNAM). Tenango de Doria, El Texme, 1250 m, 18 Apr 1994, 15:30 hrs, J.L. Salinas, JL-106 (1 male; UNAM). Tenango de Doria, El Texme, 1250 m, 10 Aug 1993, 15:30 hrs, L. Godinez, no. 834 (1 female; UNAM). Tenango de Doria, Camino a El Texme, 1250 m, 8 Oct 1993, 11:45, 14:20, 15:50 hrs, L. Godinez, LG-828, 833, 835 (3 fe- males; UNAM). S of Tamazunchale, Hwy 85 (km 239) [21°16'N, 98°47'W], 1050 m, 10 Jul 1990, 1. Yarom (1 female; Figs. 12-14. Eicku'ortui alexnudcri, female. 12-Head. Propodeum, dorsal view. 14-Inner hind tibial spur. 13- KU). Tlanchinol [20°59'04"N, 98°38'13"W], 1600 m, 13 Sep 1993, 16:15 hrs, L. Godinez, LG-896 (1 female; UNAM). Tlanchinol, [20°59'04"N, 98°38'13"W], 1600 m, 30 Apr 1994, 11:30 hrs, L. Godinez, LG-1002 (1 male; UNAM). Xochicoatlan [20^48'N, 98°40'W], 18 Mar 1979, en madera de Pinits, M.A. Moron (1 female; UNAM: Mexico City). 19.5 km S San Luis Potosi-Hidalgo border on Hwy 85, 1200 m, 11 Jul 1990, D. Conlon, R.L. Minckley (2 females; KU). 118 The University of Kansas Natural History Museum Special Publication No. 24 20 18 16 females males Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Fig. 15. Eickn'ortia nycteris, flight records. Hwy 85 (km 232), 1200 m, 11 Jul 1990, 1. Yarom (2 females; KU; one used for habitus illustration by E. Hodges, and is so labeled). Mexico: Chalma [18°55'N, 99°26'W], 1700 m, 26 Nov 1990, L. Godinez, no. 629 (1 female; PCAM). Chalma [18°55'N, 99°26'W], 1700 m, 26 Nov 1990, on Simsin nmplexicaulis, L. Godinez, no. 625 (1 female; PCAM). Ixtapan de la Sal, 9 mi N [labeled as from Guerrero], [18°50'N, 99°41'W], 16 Aug 1981, J. Chemsak, A&M Michelbacher (1 female; UCB). Michoacan: Patzcuaro, 8 kmS [19°31'N, 100°36'W], 2200 m, 27 Oct 1987, L. Godinez, LG-205 (1 female; KU). Nayarit: Santa Isabella [Santa Isabel?, 21°10'N, 104°37'W], 9 mi NW, 10 Mar 1972, P. Parker, D. Miller, Utali State Univ. Intermountain Insect Survey (3 females; USU). Oaxaca: Guelatao [17°18'N, 96°29'W], 20 Feb 1991, on Bidens pilosa, L. Godinez, no. 663 (2 males; UNAM). Puebla: Cuetzalan, 3 mi SW (N of Zacapoaxtla) [20°02'N, 97°31'W], 4100 ft, 19 Jun 1961, Univ. Kans. Mex. Expedition (2 females; KU). Huauchinango, 8 mi E [20°11'N, 98°03'W], 4050 ft, 21 Aug 1962, Univ Kans. Mex. Expedition (1 female; KU). Teziutlan, 5 m NE [19°49'N, 97°21'W], 4700 ft, 27 Jun 1953, Univ Kans. Mex. Expedition (1 female; KU). Teziutlan, 5 m NE, 1700 ft, 27 Jun 1953, Univ. Kans. Mex. Expedition (1 male; KU). Teziutlan, 8.5 mi NE, 4800 ft, 13 Aug 1969, Univ Kans. Mex. Expedition [1 female; KU]. Queretaro: Jalpan, 43 km E [21°14'N, 99°29'W], 1100 m, 17 Aug 1987, on Compositae, D. Yanega, (1 female; KU). Jalpan, 43 km E, 1500 m, 24 Aug 1988, Bidetis, D. Yanega (1 female; KU). San Luis Potosi: Platanito, 141 km W, 900 m, 27 Jul 1990, W. Bell, I. Yarom (1 female; KU). Veracruz: Coscomatepec, Rio Jamapa NE of Coscomatepec [19°04'N, 97°02'W], 4300 ft. 8 Aug 1969, Univ. Kans. Mex. Expedition [1 female; KU]. Jalapa, 2 mi NW, 17 Aug 1959, L.A. Stange, A.S. Menke (1 female; LACM). Las Vigas, 12 km SE [19°36'N, 97°08'W], 2070 m, 14-15 Jul 1974, J.A. Chemsak, E. & J. Linsley J. Powell (1 female; UCB). Orizaba [18°51'N, 97°06'W], 1862, Biart [one female lectotype, four female paralectotypes; PARIS]. Xico, Texolo Falls [19°25'N, 97°00'W], 11 Jan 1982, in ground [incorrectly labeled as being "in ground," R. Brooks pers. comm.j, B.H. Smith [1 female; KU]. NICARAGUA. Santa Maria de Ostuma [13°00'N, 86°00'W], Nov 1959, N.L.H. Krauss (1 female; CU). Eickivortia alexanderi, new species Figures 12-14 Type Material.—The female holotype of £. alexanderi is deposited in the Snow Entomological Collection (Uni- versity of Kansas). It is labeled: "Costa Rica, Heredia [,] 8.7 km N. Varablanca [10°00'N, 84°07'W], 1510 m, July 18, 1964 [,] M.G. Naumann coll. HOLOTYPE [,] Eickwortia alexanderi [,] R.J. McGmley" [red label]. The head has been glued to the thorax but the specimen is otherwise in excel- lent condition. Etymology.—This species is named in honor of the late Byron A. Alexander, a valued colleague and former stu- dent of George Eickwort's, whose brilliant young career was tragically cut short in 1996. Description.—Female: as described for E. nycteris ex- cept for the following: (1) Length approximately 9.0 mm; (2) wing length 7.26 mm; (3) abdominal width 2.84 mm. EicKWORTiA, New Genus of Halictine Bee 119 in CD i-c\j'^LncDr^cqa)co : T^ c\icMCNic\ioJc\ic\ic\i Intermandibular width (mm) Fig. 16. Histogram showing size variation in intermandibular widths for 51 females of E, m/ctcris. Structure. (4) Head broad (Fig. 12), not appearing rounded due to compound eyes converging ventrally; length /width ratio = 0.90. (5) Gena, at midpoint, wider than eye width (eye/genal ratio = 0.64), (6) rounded pos- teriorly (9) Clypeus broader than long, width/length ra- tio = 2.75 (not as broad as in £. mfcteris); (10) surface weakly convex, not depressed apically; (11) clypeal surface with- out shallow median longitudinal sulcation. (13) Distance between lateral ocellus and eye slightly greater than dis- tance between lateral ocellus and hind margin of vertex (ocular-ocellar space approximately 2.3 times lateral ocel- lar diameter); (14) distance between lateral ocellus and eye exceeding distance between lateral ocelli (ratio = 1.5). (16) Compound eyes converging below. (24-29) Labral charac- ters were not examined because only one specimen, the holotype, is currently known. (31) Mandibles strongly bidentate as in E. mjcteris (Fig. 12). (33) Pronotal lateral ridge incomplete, weakly developed; (34) lower portion of lateral ridge rounded, indistinct; (34a) pronotal lobe narrowly rounded and projecting but not as conspicuous as in £. uyctcris. (35) Mesoscutal anterior edge rounded, not bilobed; (37) median mesocutal line weakly impressed. (41) Dorsal surface of propodeum depressed centrally (Fig. 13), lateral rims of propodeal dorsal surface well devel- oped with conspicuously elevated, narrowly rounded edges; (44) lateral propodeal carinae moderately devel- oped, extending one-half distance of posterior surface. (45) hiner hind tibial spur similar to that of £. in/cteris (Fig. 14). (46) Lateral edge of metasomal tergum II very weakly con- cave. Sculpture. (48) Punctures in area between ocelli and antennae separated by 2-3 diameter widths. (49) Vertex near eye and (50) behind ocelli sculpKired as between ocelli and antermae. (51) Supraclypeal area somewhat shiny granulate. (53) Clypeus finely alveolated, apical quarter of clypeus not distinctly different from basal area. (56) Mesoscutum somewhat shiny; (57) surface smoothly alveolated throughout, simply punctate (not doubly-punc- tate as in E. m/ctcris), punctures separated by 1-5 times their diameters. (58) Scutellum sculptured as mesoscutum but punctures more dense medially; (59) metanotum densely punctate with small nearly contiguous punctures. (60) Pre-episternum, (61) hypoepimeral area, mesepisternum, and (62) metepisternum all distinctly stri- ate (not as conspicuous as in E. in/cteris). (63, 64) Dorsal surface of propodeum somewhat shiny, not granulate, dis- tinctly and regularly striolate to posterior rim (Fig. 13). (65) Metasomal tergum I normal (not narrowed basally, with- out apical biconvexities); (66) surface microscopically striolate throughout with extremely fine widely spaced punctures. Coloration. (71) Membrane of forewing entirely Lnfuscated. 120 The University of Kansas Natural History Museum Special Publication No. 24 Vestiture. (75) Pubescence of thorax dark browrr ex- cept for white hair patch on edge of pronotal lobe and sur- rounding area of pre-episternum. (77) Hind tibial hairs all dark brown. (78) Anterior hairs of metasomal tergum I dark brown, very sparse; (79) basal hair band of tergum II ab- sent. (81) Basal hair bands absent from terga III-IV. DISCUSSION Flight Data Females of £. m/cteris appear to be active throughout the year and have been collected from January through November (Fig. 15). The peak collections in July may sim- ply reflect when bee specialists are most active in the field. Detailed studies of individual populations are needed to determine seasonal patterns. Males of £. m/cteris have been collected from February to November. The one known specimen of E. alexanderi was collected in July. Floral Asscxtiations Unfortunately only 12 (20%) of the 60 known females of E. nycteris are associated with floral data. Nine of these are records from Asteraceae. Seven of these specimens (five with pollen in their scopa) were taken from Bidens (Asteraceae) and one from Simsin arnplexicnulis (Asteraceae). Two females with pollen were recorded from "Compositae." Douglas Yanega provided SEM photo- graphs of pollen taken from three specimens of E. nycteris. As Dr. Yanega suggested (pers. comm., 1996), and was verified by Joan Nowicke (NMNH palynologist), one type of pollen is that of Asteraceae and the other two are most likely from two different plant families. While this would indicate £. nycteris is probably polylectic (a floral general- ist), the known floral associations would suggest that flow- ers of Asteraceae would be a good place to look for addi- tional specimens of Eickwortia. Nesting Biology The great range in female body size which appears to be correlated with head width, genal length, and poste- rior genal projection (a syndrome similar to that of the primitively eusocial Halictiis ligatiis), suggests that £. nycteris may be social to some degree. "Large-headed" fe- males are often twice the size of most smaller females, as is reflected in the range of intermandibular distances mea- sured for nearly all known specimens (Fig. 16). The nest- ing biology of this species should be considered a primary research target for bee specialists, especially for those in- terested in halictine social behavior. Of additional biological interest are the strongly bidentate mandibles of females of both species of Eickxmrtia. Females of Lasioglossiim (sensu lato) are (with few exceptions) ground-nesting bees that have mandibles with small, subapical teeth. The strongly bidentate man- dibles of Eickwortia suggest these species possibly nest in wood. The first specimen of E. nycteris I examined was associated with the handwritten note "... en madera de Pinus" [specimen from Xochicoatlan, Hidalgo; UNAM collection]. Future Collecting and Research Priorities Suggestions for future work on Eickwortia include (in no particular order): 1) Collection of additional specimens of both species with emphasis on high altitude oak/pine forests (823 meters and above) and with a focus on Asteraceae as host plants (especially Bidens). Discovery of males of £. alexanderi would be an excellent find. With the intensive survey work being undertaken in Costa Rica, e.g., INBIO, it would be surprising not to see additional specimens of this species in the future. 2) Because the size variation among females of E. nycteris suggests this species may be social to some de- gree, knowledge of its nesting biology would be of great interest. The strongly bidentate mandibles of the females and limited biological data reviewed above would indi- cate that bee specialists should consider searching for nest- ing activity associated with wood. 3) The phylogenetic relationships of this genus to other halictines need to be explored, based on morphological and molecular data. Because Eickwortia appears related to New World Evylaeus, a large and taxonomically difficult group currently being revised (McGinley, in prep.), explo- ration of phylogenetic relationships based on morphologi- cal characters is deferred for an anticipated study of halictine higher classification. LITERATURE CITED McGinley, R. J. 1986. Studies of Halictinae (Apoidea: Halictidae), I: Revi- sion of New World Lasioglossiim Curtis. Smithsonian Contributions to Zoology 439, vi + 294 pp" Michener, C. D., R. I. McGinley and B. N. Danforth. 1994. Tlie Bee Genera ofNorth and Central America (Hynienoptera: Apmtdea). Smithsonian In- stitution Press: Washington, D. C, viii + 209 pp. Moure, ]. S. and P D. Hurd, Jr. 1987. An Annotated Catalog oftlie Halietid Bees of the Western tieniispliere (Hynienoptera: Halictidae). Smithsonian Institution Press: Washington, D. C, vii + 405 pp. Vachal, J. 1904. Etude sur les Halictus d'Amerique (Hym.). Miscellanea Entomologica , Narbonne 12: 113-128. Bi/t'rs, G, W., R. H. Hagcn, and R. W. Brooks (eds.), Entomological Contributtons in Memory of Byron A. Alexander. Liiiiivrsihi of Knnsnf Naturnl History Museum Specinl Publiaition 24. (1999) Pp. 121-123 A New Bittacus from Southern Mexico (Mecoptera)^ By George W. Byers^ ABSTRACT Bittacus alexamieri, n. sp., is described from southern Puebla and compared with similar regional species. Taxonomically useful characters of both sexes are illustrated. Kex/imrds: Hanging-flies; Scorpion-flies. INTRODUCTION Four species of Bittacus have already been described from Mexico: B. mexicamis Klug in central Mexico; B. hauksi Esben-Petersen, widespread in Mexico and Central America; B. si/lvaticus Byers in central Veracruz; and B. peniiisularis Byers in southern Baja California. Others have been found but await description. The species described here was found, almost accidentally, when I was search- ing for the elusive Eremobittacus (Byers, 1997), near Petlalcingo, in southern Puebla. Acknowledgments 1 thank Robert W. Brooks, who accompanied Byron Alexander during the study of bees in northern Mexico, and who described to me Byron's goals on that trip. Thanks also to Steve Ashe, Wes Bicha, and Don Webb for helpful comments on an earlier version of this paper, to A. R. Thornhill, who accompanied me on the 1972 trip to Mexico, and to Cynthia Woods and Sharon Lee Green for entering this paper into computer. SYSTEMATICS Bittacus alexanderi, new species Description based on 10 males, 24 females, pinned. Head.—Dorsum and frons unevenly sordid brown, ocellar prominence darker brown to almost black. Lateral ocelh nearly twice diameter of median ocellus; males (and a few females) with two short, black setae close behind median ocellus. Rostrum dark yellowish brown to reddish brown, clypeus glossy; maxillary palps dark brown except nearly bare terminal segment yellowish brown to brown; labial palps yellowish brown. Antennal scape and pedicel brown, flagellum light brown with 21-22 flagellomeres. Thorax.—Pronotum brown, with two large, black se- tae at each side on anterior margin, each arising from wid- ened, slightly raised prominence on margin; posterior margin with one seta at each side. Mesonotum and metanotum unevenly brown; large black seta on prescutal prominence of mesonotum above and before base of fore wing, another on anepisternum anterior to and slightly lower than wing attachment; pair of small setae on mesonotal scutellum in some males. Pleural surfaces, coxae and mera unevenly sordid light brown, with sparse, yel- lowish setae, longest and most dense on coxae; single black seta on mesepimeron, one on anterior coxa (most speci- mens) and one on middle coxa, 3 or 4 on posterolateral surface of hind coxa. Femora, tibiae and tarsi dark yellow- ish brown, femora darkest; femoral setae sparse, long, black; shorter tibial and tarsal setae black; tibial spurs long and slender, spurs of hind tibia nearly as long as basitarsus. Hind femora enlarged in both sexes, about 1.5 times di- ameter of other femora. Wing membrane strongly tinged with brown; pterostigma hght brown, scarcely evident; veins light brown. Scv shortly before level of Frs, 2 Pcv, apical cross- vein present but not close to end of 1 A. Fore wing with 6- 7 black setae along posterior margin near base; 3-7 such setae on hind wing. Abdomen of male.—Terga unevenly yellowish brown to brown; tergum 2 with 6-9 large, black setae along each lateral margin. Sterna similarly colored, narrow, anterior ones usually drawn upward and concealed in lateral as- pect by terga in dried specimens. Terga 3, 4 and 5 each nearly as long as 6 and 7 together Segment 8 very short, tergum partly concealed dorsally by tergum 7. Posterior edges of sterna 7-9 thin. Epiandrial appendages of tergum 9 (Figs. 1,2) elongate, generally parallel-sided, rounded at apex, with long bordering setae; spinose projection on mesal surface of dorsal margin and group of 5-6 thick, black spines directed mesad from ventral margin near base. Basistyles short, subrectangular in lateral aspect, with con- centration of stiff setae on posterior surface. Dististyles small, blunt-tipped. Cerci (Fig. 1) longer than epiandrial appendages, of nearly uniform diameter throughout, with dorsal setae on apical half recurved. Aedeagus widened Contribution Number 1399 from the Snow Entomological Collection, Museum of Natural History The University of Kansas. Snow Entomological Collection, Division of Entomology, Natural History Museum, University of Kansas, Lawrence, KS 66045. 121 122 The University of Kansas Natural History Museum Special Publication No. 24 ^ ^S>yA^-2'-''jiiyp^jfMj 1 mm Figs. 1-3. BithKiis nkxaiidcri, new species. 1-Terminal abdominal segments of male, right lateral aspect; aed — aedeagus, bs—basistyle, ce—cercus, ea—epiandrial appendage (ninth abdominal tergum), pr—proctiger, s—sternum, t — tergum. 2-Right epiandrial appendage of male, dorsal aspect; 7t —abdominal tergum 7, concealing median portion of tergum 8. 3-Terminal abdominal segments of female, left lateral aspect; ce — cercus, s —sternum, sgp—subgenital plate, t — tergum. Scale: all figures. near base, then reduced to thin, non-coiled, sclerotized fila- ment. Proctiger two-branched, upper branch strongly scle- rotized, arched dorsad, without setae; lower branch semi- membranous, with pale dorsal setae. Abdomen of female.—Terga unevenly brown, darker toward posterior end except segments 8-11 again paler; 5-8 black setae along each lateral margin of second ter- gum. Sterna long, narrow, brown. Sternum 8 narrowly but completely divided along ventral mid-line, notched dor- sally at each side around spiracle. Subgenital plates (Fig. 3) with numerous strong setae near posterior margin. In dried specimens, tenth segment recessed withm ninth; cerci and parts of segment 11 protruding (Fig. 3). Eggs subcuboidal, with six impressed sides. Measurements.—Body length, male, 19.3-21.8 mm. (holotype 20.0 mm.); female, 16.5-20.0 mm. (allotype 18.2 mm.). Fore wing, male, 22.0-24.0 mm. (holotype 23.2 mm.); female 22.0-23.6 mm. (allotype 23.5 mm.). Antenna (both sexes) 9.9-11.2 mm. Types. — Holotype, male, allotype and 4 male, 14 fe- male paratypes collected 3 miles (4.8 km.) northwest of Petlalcingo, Puebla, Mexico, on 29 August 1972, by G.W. Byers and R. Thornhill (GWB field catalogue: Puebla no. 33); 2 male, 5 female paratypes, same locality, 5 September 1972 (GWB cat.: Puebla no. 34); in the Snow Entomologi- cal Division, Natural History Museum, University of Kan- sas, Lawrence, Kansas 66045; 2 male, 3 female paratypes collected 3 miles north of Petlalcingo, 14 August 1993, and 1 male, 1 female at same locality, 15 August 1993, by Wes and Fred Bicha, in Wes Bicha collection. The type locality is along a dry stream bed at a bridge on Highway 190, 4.8 km. northwest of the junction of the highway and a short side road leading into the village of Petlalcingo; elevation 1400 m. The bittacids were in the shade provided by low herbaceous plants a meter or less in height; they were not found beneath larger acacia-like trees. The habitat was sur- prisingly dry for bittacids, the ground bare, dry and stony between plants. This was also the habitat of the rare Eremobittacus, found by other collectors many years ear- lier The Bicha collections were made in low growth "in the shade of fence row trees" (including acacias) at the edge of a corn field left dry and barren by mid-August. There was no stream anywhere nearby. Discussion.—In its coloration and general appearance, Bittacus alcxividcri resembles both B. baiiksi Esben-Petersen and B. spatulatus Byers (a species currently known only from Costa Rica and Nicaragua). Both these species have elongate cerci and long, approximately parallel-sided epiandrial appendages in the male. Males of B. alcxaiideri can be easily recognized, however, by the blunt, spinose projection from the inner dorsal margin of each epiandrial appendage. Females of B. alexandcri can be recognized by the char- acteristic shape of the eighth abdominal sternum and at- tached subgenital plates. However, the exact outlines of these structures are not easily seen in females that have become somewhat deformed in the drying process. This species is named in memory of Dr. Byron A. Alexander (1952-1996), with whom 1 enjoyed a close work- ing relationship for several years. Byron was not just a col- New Bittacus from Mexico 123 league in the Department of Entomology and in the Snow Entomological Museum. He was, in a sense, my replace- ment, holding a curatorial position in the Museum and teaching some of the courses that 1 had taught for many years. He was interested in Mexican insects, particularly bees, and had studied the plant associations of bees in Chihuahua, Coahuila and Sonora. He had been investi- gating the resources used by bees of the genus Dindasia to learn whether plants chosen as nectar sources by these bees correlated with the supposed phylogeny within the genus. LITERATURE CITED Byers, G. W. 1997. Four puzzling new species of Mecoptera. Proceedings of the Entomological Society of Washington 99: 681-692. Byers, G. W., R. H. Hngen, ami R. W. Brooks (eds.), Entomological Contributions in Memory of Byron A. Alexander. Liniveriitii of Kanscif Natural History Museum Special Publication 24. (1999) Pp. 125-138 Resolving Conflict Between Morphological and Molecular Evidence for the Origin of Eusociality in the "Corbiculate" Bees (Hymenoptera: Apidae): A Hypothesis-Testing Approach By Ted R. Schultz'--, Michael S. Engel^ and Michael Prentice "^ ABSTRACT A hypothesis-testing approach is employed in an attempt to resolve conflicting evidence in two data sets, one morphological and one molecular, for the origin of eusociality in the "corbiculate" bee tribes Apini, Bombini, Euglossini, and Meliponini. After determining that both data sets contain high levels of character congruence (i.e., "phylogenetic signal"), four statistical tests are used to deter- mine whether each data set is significantly able to distinguish between three pairs of alternative hy- potheses. Based on the results of these tests, neither data set is able to distinguish between single (favored by the morphological data) vs. dual (favored by the rDNA data) origins for "general" eusociality, present in Bombini, Meliponini, and Apini. In contrast, the two data sets significantly fa- vor opposite scenarios for single (morphology, >95% confidence) vs. dual (rDNA, >90% confidence) origins for "advanced" eusociality, present in Apini and Meliponini. When the morphological and rDNA data are combined into a single data set, the resulting tree favors a dual origin for general eusociality and a single origin for advanced eusociality; however, internal character conflict elimi- nates all power to significantly distinguish this result from alternative hypotheses. We conclude that, despite strong conflict between the two data sets, there is no compelling evi- dence for rejecting Darwin's (1859) logical "null" hypotheses of a single origin for general eusociality in the group (Bombini + Meliponini + Apini) and a single origin of advanced eusociality in the group (Meliponini + Apini). A serious problem with the reliability of the gl-statistic as a measure of phyloge- netic signal is noted. Keywords: Hymenoptera; Apoidea; Phylogeny; Eusociality; rDNA. INTRODUCTION '^ ^^^ monophyly of each of the tribes (Michener, 1990, as subfamilies), no agreement exists on the phylogenetic re- The four tribes comprising the "corbiculate" bees rep- lationships among the tribes. Considering the enormous resent the full spectrum of social evolution, ranging from interest in the group, this situation is unfortunate, and re- solitary and communal behavior (Euglossini) to primitive suits from conflicting evidence from numerous morpho- eusociality (Bombini) to advanced eusociality (Meliponini logical and molecular studies. and Apini). For this reason, as well as for other behavioral j^g jj-ibe Apini, or honey bees, comprises one extant elaborations such as the waggle dance, worker policing, genus. Apis. All Apis species occupy the "advanced and queen control, the corbiculate bees are of consider- eusocial" behavioral grade {scnsii Michener, 1969: possess- able interest to students of animal behavior One species jj^g reproductive division of labor, overlapping genera- in particular. Apis mellifem (Apini), has been the subject of tions, and morphologically distinct castes), construct large more biological research than any other insect, primarily ^^x combs, and use the dance language to communicate because of its economic importance. fooj location to nestmates (von Frisch, 1967). Apis currently Although the monophyly of the corbiculate bees is well consists of seven species {A. niidreniformis, A. cerana, A. supported ("Apidae" sensuSakagami and Michener, 1987; dorsata, A. floren, A. kosclicvtiikovi, A. iiigwciiictn. and A. "apine clade" sensu Roig-Alsina and Michener, 1993), as meUifera)'" , though some populations of the "giant honey ' Department of Entomology, National Museum of Natural History, MRC 165, Smithsonian Institution, Washington, DC 20560. E-mail: schultz@onyx.si.edu - Address for correspondence. ' Department of Entomology, Comstock Hall, Cornell University, Ithaca, NY 14853. Present address: Department of Entomology, American Museum of Natural History, Central Park West at 79th Street, New York, NY 10024. ' Department of Environmental Science, Policy, and Management, Wellman Hall, University of California, Berkeley, CA 94720. ^ Recently a new species of honey bee, A. nuluensis, has been proposed (Tingek et al., 1996). Through the courtesy of G. W. Otis one of us (MSE) has been able to examine representatives of these Asian bees. They are in actuality only a distinctive variant or subspecies of A. cerana (synonymy by MSE). Refer to Engel (1999b) for current honey bee classification. 125 126 The University of Kansas Natural History Museum Special Publication No. 24 Michener, 1974; Michener, 1990 Euglossini Bombini Meliponini Apini Apini Sheppard & McPheron, 1991: Cameron, 1993 Euglossini Bombini Meliponini Euglossini Bombini Meliponini Meliponini Apini Euglossini lb. \ 1c. Apini Bombini Michener, 1944; Winston & Prentice, 1 991 ; Chavarria & Carpenter, Michener, 1 977; Kimsey, 1 984; 1994; Prentice & Daly, in prep. Sakagami & Maeta, 1984 Euglossini Meliponini Bombini Plant & Paulus, 1987 Apini Apini Meliponini Euglossini Bombini 1d. Cameron, 1991 Pereira-Martins &Kerr, 1991 Euglossini Apini Bombini Meliponini Euglossini Meliponini Bombini Apini Fig. 1. Previously proposed phylogenies for the corbiculate bee tribes. bee" A. dorsata may represent distinct species. Diversity in Apis is discussed in Ruttner (1988) and Smith (1991). Re- cent phylogenetic analyses of the genus inckide Alexander (1991a,b), Garnery et al. (1991), Cameron et al. (1992), Willis et al. (1992), Engel and Schultz (1997), and Engel (1998a, 1999b). Crozier and Crozier (1993) have published the com- plete sequence and organization of the mitochondrial ge- nome for A. mellifera. The Meliponini (stingless bees) are a diverse group of 21 genera with worldwide distribution. All occupy the "advanced eusocial" behavioral grade. The oldest known fossil bee, Trigona prisca, is a meliponine remarkably simi- lar to present-day members of the same genus (Michener and Grimaldi, 1988a,b). Wille (1979) presented a phyloge- netic hypothesis of the genera and subgenera. Cladistic analyses have been conducted by Michener (1990, for all 21 genera) and Camargo and Pedro (1992a). Camargo and Pedro (1992b) review the systematics, phylogeny, and bio- geography of the stingless bees. The Bombini, or bumble bees, encompass two extant genera, Bombus and Psitbynis. All Bombus species are "primitively" eusocial (i.e., eusocial but lacking morpho- logically distinct castes), while Psithynis species are social parasites of Bombus species. Bombus consists of a plethora of subgenera, reviewed by Richards (1968), with recent additions listed by Michener (1990). Based on cladistic studies of morphological and molecular data (Ito and Sakagami, 1985; Pamilo et al, 1981; Pamilo et al., 1987; Pedersen, 1996; Williams, 1985, 1991), Bombus is apparently paraphyletic with respect to Psithi/rus. The tribe Euglossini, or orchid bees, consists of five genera of neotropical bees, many of which are large and brightly metallic colored. Euglossines possess extremely long tongues, in some species trailing between their hind legs in flight, and many male euglossines are known to seek out orchids in order to collect oils as possible ingredi- ents in female attractants. No euglossines are eusocial, though some species of Euglossa and Eulaema are commu- nal. Two genera, Aglne and Exaerete, are cleptoparasitic. Kimsey (1987) and Engel (1999a) present genus-level phy- logenetic analyses of the tribe. In recent years, 8 of the possible 15 rooted trees for the four corbiculate tribes have been proposed, each based on a particular morphological or molecular data set, and each with its own implications for the evolution of eusocial be- havior (Figure 1). Earlier morphological studies (Michener, 1944; Michener, 1974), as well as more recent ones (Michener, 1990; Prentice, 1991; Prentice and Daly, in prep.), have supported a sister-group relationship between the Apini and Meliponini and thus a single origin for advanced eusocial behavior, in spite of disagreement over other de- tails of tree topology (Figs, la, lb). However, morphologi- CORBICULATE BeE PhYLOGENY 127 cal studies based on other characters (Winston and Michener, 1977; Kimsey, 1984; Plant and Paulus, 1987) sug- gest that Apini and Meliponini are not sister taxa, and thus support a dual origin for advanced eusociality (Figs. Ic, If), as do studies focusing on cytological data (Mello and Kerr, 1984; updated in Kerr, 1987) and nest architecture (Pereira-Martins and Kerr, 1991) (Figs. Ig, Ih). A dual ori- gin is also implied by a molecular phylogeny of 16s mito- chondrial rDNA, although different most-parsimonious tree topologies are reported in separate publications dis- cussing the same data (Cameron, 1991, 1993) (Figs. Id, le). One of these topologies (Cameron, 1993) is also supported by a study of the large rDNA subunit gene (Sheppard and McPheron, 1991). A recent "total evidence" analysis incor- porating data from a wide variety of studies carried out at different taxonomic levels also supports a "single-origin" scenario (Chavarria and Carpenter, 1994). The fact that the various data sets for the corbiculate tribes disagree in their phylogenetic implications demon- strates that high levels of homoplasy are present within at least some of them, i.e., a high proportion of presumed morphological or molecular homologies are actually cases of nonhomologous resemblance due to convergence, par- allelism, or reversal to an ancestral state. Acknowledgments We thank T.D. Seeley, G. Eickwort, and the Cornell "Social Insects Seminar Group" for stimulating this study. B.A. Alexander, S.A. Cameron, J.M. Carpenter, G. Chavarria, B.N. Danforth, CD. Michener, J.K. Liebherr, and T.D. Seeley critically read and commented on earlier ver- sions of the manuscript, resulting in substantial improve- ments. Additional thanks to: R.B. Cocroft and M.J. McDonald for stimulating discussions regarding this ma- terial, S. Rehner for aid in the sequence alignment with the Lnsergt'iie software package, D.W. Roubik for meliponine identifications, J. Sullivan for access to an un- published manuscript, D. Swofford for permission to use results obtained with PAUP* 4, and J. Wilgenbusch for advice on maximum-likelihood analysis. None of these generous souls is responsible for our errors, conclusions, or opinions. This study is dedicated to the memory of two great bee systematists and morphologists: George C. Eickwort (1940-1994), who encouraged us to pursue this study, and Byron A. Alexander (1952-1996), who read and commented on an earher version. We cherish their friendship, we honor their scientific contributions, and we mourn their loss. METHODS Choice of Data Sets Two data sets bearing on the tribal relationships of the corbiculate bees were analyzed, one consisting of morpho- logical characters, the other of molecular characters. These two data sets have the virtue of sharing all 16 terminal taxa in common (Appendix 1). Molecular data are Cameron's (1991, 1993) mitochondrial 16s rDNA sequences (Genbank accession numbers L22891-L22906) consisting of 495 nucleotide sites, of which 171 are parsimony-infor- mative. The morphological data, consisting of 25 charac- ters, are drawn from the matrices of Prentice and Daly (in prep., summarized in Prentice, 1991) and Michener (1990). In accord with usual morphological systematic procedure, character homologies were defined with the goal of pro- viding independent items of evidence on phylogeny. Al- though these characters were originally coded using the corbiculate tribes as terminal taxa (Prentice, 1991), we have specifically examined the 16 species analyzed here to con- firm the accuracy of the character states. In order to avoid any vulnerability to the claim that we have employed a circular argument in reconstructing social evolution, the character of social behavior was excluded from the analy- ses described below. We note, however, that inclusion or exclusion of this character makes no difference in the test results and that the issue of whether to include or exclude characters of interest from phylogenetic analyses is a mat- ter of legitimate debate (e.g., Wenzel, 1997; Zrzavy, 1997, Luckow and Bruneau, 1997). All other potentially useful tribal-level morphologi- cal and molecular data sets were conservatively rejected for use in this study due to a lack of terminal taxa shared in common. Use of such data sets could be justified only under the unwarranted assumption that all species drawn from within the same tribe will be identical in unexamined character states and, when this assumption is not met, could lead to unrealistic reconstructions at ancestral nodes. For instance, although Sheppard and McPheron's (1991) rDNA sequence data include one representative from each of the four tribes as well as an outgroup species in A)ithophora, in no case is the species the same as one present in the studies included here, with the exception of Apis mclUfcva. Data from a variety of other studies were unus- able because they are concerned with relationships within particular tribes and include no character information rel- evant at the tribal level [e.g., Apini: Lindauer (1956, 1961), Kreil (1975), Sakai et al. (1986), Alexander (1991a,b), Garnery et al. (1991), Willis et al. (1992); Bombini: Will- iams (1985, 1991); Euglossini: Kimsey (1987); Meliponini: Michener (1990), later revised by Camargo and Pedro (1992a)]. Alignment The 16s rDNA data set of Cameron (1991, 1993) con- sists of 495 nucleotide sites and 16 species (Appendix 1). Initially, various experimental alignments were performed using the computer programs Malign (Wheeler and 128 The University of Kansas Natural History Museum Special Publication No. 24 a. Morphology 600- 500- 400 300 200 100 L. OCM-^ CDCO OOJ'^CDCOOCJ b. rDNA 600 -.- CD ,- CD 1- CTl O C\J CD lO CO O) O) oj O^ c. Combined Fig. 2. The distribution of tree lengths (DTL's) for the mor- phology data set, the 16s rDNA data set, and for the data set pro- duced by their combination, based on 10,000 trees generated by the "random trees" command in PAllP 3.1 (Swofford, 1991) (see text). Number of trees is indicated on the y-axis, tree length on the X-axis. Hillis (1991) and Huelsenbeck (1991) have suggested that left-skewness, which is obvious in these three DTL's, indicates that the data contain strong phylogenetic signal. Gladstein, 1993), GCG (Genetics Computer Group, 1991), Seqiicncher 3.0 (Gene Codes Corp., 1995), and Megalign (DNASTAR, 1992). Depending upon the parameter settings employed, varying numbers of insertions and deletions were required by these programs; however, all produced essentially the same results, i.e., identical tree topologies and nearly identical tree lengths. Subsequently, alignment by eye was attempted and was judged to be trivial. As submitted to Genbank, there is no length variation across the 16 sequences, and thus no hypothetical insertions or deletions are necessarily required in the few highly vari- able regions. The alignment requiring no insertions and deletions was favored and was used in all phylogenetic analyses described below. Phylogenetic Analyses Parsimony analyses were carried out on the morpho- logical data set, the rDNA data set, and the data set result- ing from combining the morphological and DNA charac- ters (the "combined" data set) using PAUP* 4.0d56 (Swofford, 1997) with 10 random-addition heuristic searches and TBR branch-swapping. Bootstrap frequen- cies for branches (Felsenstein, 1985b) were likewise ob- tained using PAUP* 4.0d56 with 1000 pseudoreplicates and 10 random-addition TBR branch-swapping heuristic searches per replicate, and with autapomorphies excluded. Decay indices (a.k.a. Bremer support values or support indices), i.e., the number of extra steps required by the minimal tree in which a particular group does not appear (Bremer, 1988), were obtained by analyzing data with and without non-monophyly constraints in PAUP* 4.0d56. Maximum-likelihood (ML) analyses of the rDNA data set were also carried out in PAUP* 4.0d56 under various con- straints in order to conduct the ML Kishinio-Hasegawa (ML K-H) tests described below, and are summarized in Appendix 2. Legitimate disagreement exists over the rela- tive merits of the parsimony vs. theML criterion. Although this study largely relies on parsimony, ML K-H tests of the rDNA data were conducted in order to avoid any vulner- ability to the criticism that considerably different signifi- cance values might have been obtained under the ML cri- terion than those that were obtained under the parsimony criterion. Statistics and Tests Consistency index (C.I.), retention index (R.I.), and per- mutation tail probability (PTP) statistics were calculated in PAUP* 4.0d56 (Swofford, 1997); distribution of tree length (DTL) statistics were calculated in PAUP 3.1 (Swofford, 1991). PTP values were derived from 1000 heu- ristic-search permutation runs using TBR branch-swap- ping. Gl values were generated using the "random trees" CORBICULATE BeE PhYLOGENY 129 command to generate a distribution of tree lengths for 10,000 randomly selected trees. Tests for significance of the gl values (Hillis, 1991; Donoghue et al., 1992) were car- ried out for the three data sets by generating 100 permuted matrices in Data Ra)idomiscr (Trueman, 1992), then gener- ating a distribution of 10,000 trees for each using the "ran- dom trees" command in PAUP 3.1 (Swofford, 1991). In a test of the usefulness of the gl statistic, described below, 50 permuted matrices were generated and analyzed in the same way. The actual gl value is considered to indicate significant signal if it is more negative than some prede- termined fraction (e.g., 0.95) of the gl values for the per- muted matrices. For reasons given below, however, the gl value may be a poor measure of phylogenetic DTL skew. The partition homogeneity test, also known as the in- congruence length difference test (Farris et al., 1995), quan- tifies the between-data-set incongruence as the additional treelength required when two data sets are combined and analyzed as compared to the sum of the treelengths re- quired when the two data sets are analyzed separately. Whether this number is significantly large is determined by finding its position in the distribution of such numbers obtained from analyses of data sets of the same size con- structed from characters randomly sampled from the com- bined data set. We applied the partition homogeneity test to the morphological and rDNA data using PAUP* 4.0d56 (Swofford, 1997) with 1000 replicates, 10 random-addition TBR branch-swapping heuristic searches per replicate. The three hypothesis tests (tree comparisons) summa- rized in Figs. 4, 5, and 6 are concerned only with tribal- level relationships. For this reason, comparison trees were generated to minimize conflict below the tribal level, which would artificially increase the apparent conflict between hypotheses for a given data set. For each comparison and each data set, the alternative trees are the most parsimoni- ous topologies that conform to the tribal-level hypotheses being compared, generated using the minimum necessary constraints in PAUP* 4.0d56 with lO random-addition heu- ristic searches and TBR branch-swapping. In the parsimony analyses these constraints never included the monophyly of the corbiculate bees nor did it include the monophyly of the four tribes, i.e., these features were in all cases natu- ral outcomes of the various analyses. In contrast, the maxi- mum-likelihood analyses included these constraints (Ap- pendix 2). The four statistical tests employed are not without problems, some of which are pointed out in the descrip- tions below. One such problem is that all three are a priori tests (Kishino and Hasegawa, 1989; Swofford et al, 1996); however, if it is accepted that alternative hypotheses about the evolution of eusociality in the corbiculate bees have been framed a priori, this unavoidable problem will affect only our comparisons of the morphology and the rDNA trees (Fig. 4). In any case, these are the best tests currently available for determining whether a given data set is able to distinguish significantly between two alternative topo- logical hypotheses, and we have employed them here as gross indicators of strength of support (i.e., of character congruence) for one hypothesis relative to an alternative. Given this perspective, our emphasis is less on the precise P-values of the outcomes and more on the relative differ- ences in performance of the morphological data set vs. the rDNA data set. The four tests are: Wilcoxon's sigiH'd-rank test (WSR): The WSR test (Templeton, 1983; Felsenstein, 1985a) compares the num- ber of extra steps (positive and negative) required for each character on one tree relative to the number required on an alternative tree and ranks these numbers according to absolute magnitude. Following the ranking process, the positive and negative values are restored and the separate sums of the positive and negative ranks are compared us- ing a table of significance values (Rohlf and Sokal, 1995: Table V), with the expectation that when the trees explain the data equally well, the difference in sums of positive and negative ranks will not significantly depart from 0. WSR tests were carried out using the "Compare trees" com- mand in MacClade 3.06 (Maddison and Maddision, 1992) to obtain extra steps across trees; they were also calculated (with additional precision) in PAUP* 4.0d56. The WSR test is formally carried out as a one-tailed test; however, fol- lowing the recommendation of Felsenstein (1985a; see also Larson, 1994, and Mason-Gamer and Kellogg, 1996), the more conservative two-tailed probabilities are used here. Compare 2 trees test (C2T): Implemented in PAUP* 4.0d56 (Swofford, 1997), this test is a modification of the T- PTP test (Faith, 1991) and uses as a test statistic the differ- ence in lengths obtained when a given data set is con- strained to fit two alternative tree topologies. The signifi- cance of this length difference is judged by determining its position in a distribution of such values derived from a series of pseudoreplicates generated by permutations of both data sets (Archie, 1989; Faith, 1991). C2T tests were generated using the "compare 2 trees" command in PAUP* 4.0d54 (Swofford, 1997) with 1000 branch-and-bound pseudoreplicates. Significance was evaluated under a con- servative two-tailed criterion (Engel and Schultz, 1997) rather than under the one-tailed criterion implemented in PAUP* 4.0d56, which avoids a major problem of this test reported by Mason-Gamer and Kellogg (1996). The related T-PTP test has been criticized for a number of reasons (Car- penter 1992; Swofford et al, 1996), but, as discussed by Engel and Schultz (1997, p. 53), we have attempted to com- pensate for these problems in the version of the C2T test used here. 130 The University of Kansas Natural History Museum Special Publication No. 24 a. Morphology c. Combined (Morphology + rDNA) Fig. 3. The results of three unweighted parsimony analyses. Numbers preceding and following the colons indicate bootstrap frequencies and decay indices, respectively, for corresponding branches. Bootstrap frequencies are based on 1000 pseudoreplicates in PAllP* 4.0d54 (Swofford, 1997) using TBR branch-swapping, 10 random-addition heuristic searches per replicate, and with autapomorphies e.xcluded. (a.) The single most parsimonious tree resulting from the analysis of the morphology data set. Length = 33; C.l. = 0.818; R.I. = 0.953. (b.) The single most parsimonious tree resulting from an analysis of the 16s rDNA data (Cameron 1991, 1993). Length = 517; C.I. = 0.449; R.l. = 0.549. (c.)The single most parsimonious tree resulting from the combination of the rDNA and morphology data sets. Length = 563; C.I. = 0.460; R.I. = 0.599. Kishino-Hasegawa pamiiietric tests (MP K-H and ML K- H): Kishino and Hasegawa (1989) proposed two tests, one under the parsimony criterion (MP H-K) and one under the maximum-hkehhood criterion (ML K-H), that deter- mine the significance of the length/hkehhood difference separating two trees based upon a calculated variance about that difference. MP K-H and ML K-H tests were car- ried out in PAUP* 4.0d54 (Swofford, 1997). One case, that in which the rDNA data were analyzed with the constraint (Apis -I- Meliponini) (Fig. 6b), resulted in five ecjually par- simonious trees, all with the same tribal-level topol- ogy and differing only in the within-tribe topologies of the Apini and Meliponini. Because the purpose of these tests is to determine whether there is overlap in the "error ranges" of support for competing hypotheses. Fig. 6b ap- propriately reports the least significant WSR and MP K-H test values found in comparisons of the rDNA most-parsi- monious tree and each of the five equally parsimonious constraint trees. For ML K-H calculations, trees were gen- erated with maximum-likelihood searches via a procedure designed to minimize likelihood differences due to topo- logical incongruence below the tribal level (Appendix 2). RESULTS AND DISCUSSION Phylogenetic Analyses A parsimony analysis of the morphological data re- sulted in a single most-parsimonious tree ("MPT") (Fig. 3a; length = 33, C.I. = 0.818, R.I. = 0.953) that is identical at the tribal level with those of Michener (1944), Prentice (1991), Chavarria and Carpenter (1994), and Prentice and Daly (in prep.) (Figure lb). As noted above, no character describing eusociality was included in this analysis; how- ever, when such a character (State 1: eusociality absent. State 2: "primitive" eusociality present. State 3: "advanced" eusociality present) is included, the same tree topology is obtained (length = 35, C.I. = 0.829, R.I. = 0.955, whether or not the states are ordered). A parsimony analysis of the rDNA data set resulted in a single MPT (Fig. 3b; length = 517, C.I. = 0.449, R.I. = 0.549) that is identical to that pre- sented by Cameron (1991; our Fig. Id). Constraining the topology to conform to Cameron's (1993) tree (our Fig. le) required 7 additional steps. A parsimony analysis of the data set composed of the combined morphological and molecular characters resulted in a single MPT (Fig. 3c; length = 563, C.I. = 0.460, R.l. = 0.599) that is identical to that of Michener (1974) and Michener (1990) (Fig. la). Overall Character Congruence ("Phylogenetic Signal") A number of statistics have been proposed for assess- ing overall levels of character congruence ("phylogenetic signal") in data sets, some of them controversial. Four such measures were used to evaluate the morphological and rDNA data sets: 1) the consistency index (C.I.) (Kluge and Farris, 1969), 2) the retention index (R.l.) (Farris, 1989), 3) the PTP test (Archie, 1989; Faith and Cranston, 1991), and 4) the distribution of tree lengths (DTD (Hillis, 1991; Huelsenbeck, 1991). As indicated in Table 1, the C.I. for the rDNA data (0.449) is considerably worse than the ex- pected C.I. (0.634) for a study with 14 taxa, based on the survey of Sanderson and Donoghue(1989). In contrast, the C.I. for the morphological data set (0.818) far exceeds this amount. No similar criterion is available for the R.I. mea- CORBICULATE BeE PhYLOGENY 131 Table 1. Various measures of overall character congruence ("phyloge- netic signal") for the morphology and lbs rDNA data sets for the corbiculate bees, considered separately and when they are combined and analyzed as a single matrix. According to a survey of 60 cladistic analy- ses by Sanderson and Donoghue (1989), the expected C.I. for 16 taxa is 0.634; the values for the rDNA data set and for the combined data set fall well below this expectation. The negative gl values for both the rDNA and the combined data sets indicate left-skewness and hence phyloge- netic signal according to the criterion of Hillis (1991) and Huelsenbeck (1991 ); however, exploration of a recommended significance test (Hillis, 1991; Donoghue et al., 1992) indicates that the gl statistic may be a poor indicator of phylogenetic signal (see text). FTP tests (Faith and Cranston, 1991) indicate significant phylogenetic signal in all three data sets. C.I. R.I. gl FTP "P" value Morphology 132 The University of Kansas Natural History Museum Special Publication No. 24 Euglossini Bombini Meliponini Morphology Apini L = 33 a. Morphology Data Set 18 steps (55% MPT length) WSR: P < 0002 • C2T: P = 001 • MP K-H P < 0.0001 • L = 51 L = 531 b. rDNA Data Set 1 4 steps (3% MPT lengtti) WSR: P = 00449* C2T P = 0.082 MP K-H: P = 0-0303 ' ML K-H P = 0599 *- L = 517 c. Morphology -i- rDNA Data Set 4 steps L = 564 -*- (< 1% MPT lengtti) WSR: P = 0.6534 C2T: P = 776 MP K-H P = 6068 L = 568 rDNA Euglossini Apini Meliponini Bombini Fig. 4. Statistical tests of alternative hypotheses of tribal-level relationships in the corbiculate bees. WSR = Wilco-xon's signed-rank test, C2T = Compare 2 trees test, MP K-H = parsimony-based Kishino-Hasegawa parametric test, ML K-H = likelihood-based Kishino-Hasegawa parametric test. (Left) the single most-parsimonious tree (MPT) resulting from unweighted parsimony analysis of the morphology data set. (Right) The MPT resulting from unweighted parsimony analysis of the rDNA data set. Asterisks (*) indicate significant power to distinguish between the alternative hypotheses at the 95^''t level. See text for description of tests. The ability of the morphology data set to distinguish between the two topologies is significant near the limits of resolution for all three tests, whereas the ability of the molecular data set to distinguish between the two topologies is non-significant at the 95^7, confidence level in two out of four tests. The combined data set is unable to distinguish between the two topologies by any of the three test criteria, indicating serious erosion of character support for tribal-level groupings due to the union of two significantly conflicting data sets. (Note that neither topology represents the most parsimonious solution for the combined data set.) index = 1 ), with one exception: support for the sister-group relationship (Apini + Mehponini) is remarkably high (boot- strap frequency = 100%, decay index = 14 = 42% of MPT length). Bootstrap and decay index branch support for the monophyly of each of the tribes Euglossini, Meliponini, and Apini in the rDNA tree (Fig. 3b) is quite high (boot- strap frequency > 95%, decay index > 10). However, sup- port for tribal-level relationships is much lower: The sis- ter-group relationship (Euglossini + Apini) is supported by a bootstrap frequency of 72% and a decay index of 3 (0.6% of MPT length); the sister-group relationship of Bombini -i- Meliponini is supported by a bootstrap fre- quency of 75%' and a decay index of 5 (1 %- of MPT length). This decrease in support at the tribal level may indicate a limit in the usefulness of the 16s rDNA gene sequence char- acters due to saturation (multiple "hits" at informative sites) at deeper nodes within the tree. The group (Apini -t- Meliponini), present on the mor- phology MPT (Fig. 3a), also appears on the combined-data MPT (Fig. 3c). However, due to conflict between the mor- phological and molecular data sets, bootstrap/decay in- dex support for this group is eroded from 100%/14 in the morphology MPT to 84% /5 in the combined-data MPT. Hypothesis Tests Even when it can be shown, e.g., with the partition homogeneity (ILD) test, that significant overall disagree- ment exists between two data sets, the localization of that disagreement to various groups within alternative topolo- gies remains a complex problem. Tltis problem can be made more manageable by addressing the ability of the indi- vidual data sets to discriminate between narrowly defined alternative hypotheses. We have chosen to adopt this hy- pothesis-testing approach to focus on what is arguably the greatest single biological implication of the phylogeny of the corbiculate bees, the evolution of eusociality within the tribe. Hypothesis Test 1 : Morphology MPT Topology vs. rDNA MPT Topology When the morphological data are constrained to fit the tribal-level topology favored by the rDNA data, a length difference of 18 extra steps (55% of the MPT length) is required (Fig. 4a). This length difference is significant to the limits of resolution in two of the three tests employed, indicating that the morphological data very strongly fa- vor the MPT over the alternative topology. In light of the boostrap frequency and decay index corresponding to the branch supporting (Apini + Meliponini) on the morphol- ogy tree (Fig. 3a), this is perhaps unsurprising. CORBICULATE BeE PhYLOGENY 133 "General" Eusociality Single vs. Multiple Origins Euglossini a. Morphology 1 step L = 33 L = 517 134 The University of Kansas Natural History Museum Special Publication No. 24 1 = 33 "Advanced" Eusociality: Single vs. Multiple Origins Euglossini a. Morphology Bombini . . . 14 steps * (42% MPT length) Meliponini WSR: P = 0.0061 * C2T: P = 0.028 * MP K-H: P = 0.0006 Apini Euglossini Apini b. rDNA 1 2 steps 1=517 Meliponini (2% MPT length) WSR: P = 0.0897 C2T P = 0,082 MP K-H: P = 00513 ML K-H: P = 0.0486 ' fyleliponini Apini Euglossini Bombini Euglossini Bombini Apini Meliponini Euglossini Bombini rDNA + Morphology _^ 5 steps L = 563 Meliponini Apini (< 1% MPT length) WSR: P = 0.5575 C2T P = 0744 MP K-H: P = 0.5016 Apini Meliponini Fig. 6. The ability of the three data sets to distinguish between alternate hypotheses for the origin of the "advanced" eusocial behavior present in the Apini and Meliponini. See Figure 4 cap- tion for test abbreviations. Tick-marks represent origins or losses of advanced eusociality; asterisks {*) indicate significant power to distinguish between the alternative hvpotheses at the 95"c level. Tree-lengths are the results of unweighted parsimony analyses, (a.) The most parsimonious tree for the morphologv data set fa- vors a single origin of advanced eusociality; a dual origm (or, al- ternatively a gain and a loss) requires 14 extra steps, judged to be highly significant in all three tests, (b.) The most parsimonious tree for the 16s rDNA data supports a dual parallel origin of ad- \'anced eusociality. A single origin requires a tree 12 steps longer, a difference found to be non-significant at the 95'7c confidence level by all three parsimony-based tests. The topologv difference was found significant at the 95% level, however, under the ML K-H test (see text), (c.) When the data are combined and analyzed, the resulting MPT favors a single origin of advanced eusociality. How- ever, the unweighted length difference between this topology and the alternati\'e is now only 5 steps, found to be nonsignificant by all three tests. fidence level in three of the four tests employed; however, all tests found significance at the 90% level. Under the maximum-likelihood criterion, the ML K-H test found a significant difference in character support for the two trees at the 95% level (Fig. 6b). Not surprisingly, as in both previous hypothesis tests, the combined data are again unable to distinguish signifi- cantly between the two alternatives (Fig. 6c). Based on these results, we conclude that the morpho- logical data very strongly favor the grouping (Apini -i- Meliponini), whereas the rDNA data strongly (but less strongly) oppose it, favoring instead the grouping (Bombini + Meliponini). Based on the results of Hypothesis Test 2, above, the position of the Apini outside of (Bombini -i- Meliponini) plays little or no role in the preference of the rDNA data for one topology over the other, i.e., the group- ing (Euglossini -i- Apini) lacks significant support. CONCLUSION There is clearly strong conflict between the morpho- logical and rDNA data sets, and this conflict appears to exceed what might be expected from random noise. It might be conjectured that this conflict is due to rampant, concerted parallelism in one or the other data set, e.g., con- vergence in multiple details of worker morphology in the Meliponini and Apini, or parallel, site-specific base changes in the 16s rDNA molecule of the Apini and Euglossini. In the absence of evidence for such phenomena, this conflict will only be resolved by new character data appropriate to the problem. Based on the foregoing tests, however, we conclude that neither data set provides a significant rea- son for rejecting the group (Bombini + Meliponini -i- Apini) (Figs. 5a and 5b), thus allowing to stand the null hypoth- esis of a single origin for general eusociality within the corbiculate bees. A single origin (as opposed to multiple origins) is certainly the appropriate null hypothesis for further inquiry, based both upon Hennig's (1966: 121-122) "auxihary principle" and upon the highly conservative (i.e., uniformly fixed) distribution of this character across spe- cies within the three tribes (Schultz et al., 1996). The origin of so-called "advanced" eusociality is more problematic, however. While the morphological data strongly support a single origin (Fig. 6a), the rDNA data also provide strong (if somewhat more equivocal) evidence in favor of either a dual origin or a single origin with a subsequent loss of morphological castes in the Bombini (Fig. 6b). Based on morphological and behavioral characters, Darwin (1859: 224-235) proposed that the stingless bees (Meliponini) are intermediate between the more primitive bumble bees (Bombini) and the more derived honey bees (Apini). In spite of the conflict between the data sets and in the absence of additional data, we conclude that the preponderance of evidence favors Darwin's implied to- pology (our Fig. 3a) and its implication of a single origin for general eusociality and a single origin for "advanced" eusociaUty. This evidence includes: 1) Judging from the C.I. value, the rDNA data set, which contradicts (Apini -i- Mehponini), has a demonstra- bly higher noise level than does the much smaller mor- phological data set (Table 1 ). That this noise is concentrated CORBICULATE BeE PhYLOGENY 135 at more ancient phylogenetic levels is supported by the relatively higher bootstrap values on more recent branches (Fig. 3b), suggesting that the average rate of evolution in 16s rDNA characters mav be too rapid to preserve infor- mation on tribal-level relationships. The presence of strong signal on recent (within-tribe) branches is supported by perfect congruence between the rDNA data and a mor- phological data set specific for the genus Apis (Engel and Schultz, 1997). Genes better suited for resolving tribal-level relationships within the corbiculate bees might include elongation factor 1-alpha (Friedlander et al., 1992; Fried- lander et al. 1994; T. Schultz, unpublished data) and dopa decarboxylase (Fang et al., 1997). 2) As indicated both by the bootstrap frequency (100%), decay index (14 steps) (Fig. 3a), and highly significant re- sults of the three statistical tests (P > 0.97) (Fig. 6a), the grouping (Meliponini + Apini) is well supported when the morphological data set is analyzed separately, whereas the alternative tribal grouping of (Bombini + Meliponini) is considerably less well supported when the rDNA data set is analyzed separately (bootstrap frequency = 75%, decay index = 5; 0.90 < P < 0.95) (Figs. 3b and 6b)'. 3) When the data are combined and analyzed, the re- sulting topology continues to support the grouping (Apini -I- Meliponini) with a reasonably high bootstrap value of 85%, despite the presence of strong within-data-set con- flict. Further, the combined data provide only minimal rea- son for opposing a single origin of general eusociality, as a tree of only one additional step is required to unite (Bombini + Apini -i- Meliponini), a step that is provided by the inclusion in the analysis of an ordered multistate char- acter "eusociality absent/eusociality present/advanced eusociality present." It bears repeating, however, that char- acter conflict is so high in the combined data set that it is unable to distinguish between any of the hypotheses tested at reasonable significance levels (Figs. 3c, 4c, 5c, and 6c). 4) The grouping (Apini + Meliponini) accords with the fact that fossil species of Elcctrapis are intermediate in some character states for both tribes (Zeuner and Marmiiig, 1976; Prentice, 1991; Engel, 1998b, unpublished data), a condi- tion that might be expected in a species little diverged from the most recent common ancestor of (Apini + Meliponini). 5) The sister-group status of the Apini and Meliponini is also supported by a tribal-level analysis of the corbiculate bees (Chavarria and Carpenter, 1994) that adopts a "total evidence" approach in attempting to incorporate data from a wide variety of studies carried out at different taxonomic levels. 6) Unless the branch lengths connecting tribal ances- tral nodes and terminal taxa within the tribes Bombini, Apini, and Meliponini are vastly shorter than branch lengths connecting tribal-level ancestral nodes with each other, the characters of eusociality and "advanced" eusociality are distributed across taxa in a pattern that is quite inconsistent with multiple parallel origins (Schultz et al., 1996). Rather, the observed fixation of eusociality across entire tribes of species within the corbiculate bees, as well as its fixed absence in the Euglossini and the two outgroup genera, suggest extreme conservatism in transi- tions between the absence of eusociality, the presence of general eusociality, and the presence of advanced eusocialitv. This conservatism, combined with the principle of parsimony, suggests that Darwin's (1859) grouping (Euglossini (Bombini (Meliponini -i- Apini))) constitutes the logical null hypothesis for corbiculate bee phylogeny that has yet to be definitively rejected by the available data. LITERATURE CITED Alexander, B. A. 1991a. Phylogenetic analysis of the genus Apis (Hy- menoptera: Apidae). Annals of the Entomological Society of America 84: 137-149. Alexander, B. A. 1991b. A cladistic analysis of the genus Apis. Pp. 1-28, in D. R. Smith (ed.). Diversity in the genus Apis. Oxford: Westview Press. Archie, J. W. 1989. A randomization test for phylogenetic information in systematic data. Systematic Zoology 38: 239-252. Bremer, K. 1988. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795-803. Camargo, J. M. F. and S. R. M. Pedro. 1992a. Sistematica de Meliponinae (Hymenoptera, Apidae): Sobre a polaridade e significado de alguns caracteres morfologicos. Pages 45-49, in C. d. Cruz Landim and J. C. Netto Ceds.), Anais do encontro brasileiro sobre biologia de abelhas e outros insetos socials. Naturalia (No. especial). Camargo, ]. M. F. and S. R. M. Pedro. 1992b. Systematics, phylogeny, and biogeography of the Meliponinae (Hymenoptera, Apidae): a mini- re\-iew. Apidologie 23: 509-522. Cameron, S. A. 1991 . A new tribal phylogeny of the Apidae inferred from mitochondrial DNA sequences. Pages 71-87, in D. R. Smith (ed.). Diversity iii llic genus Apis. Oxford: Westview Press. Cameron, S. A. 1993. Multiple origins of advanced eusociality in bees inferred from mitochondrial DNA sequences. Proceedings of the National Academy of Science USA 90: 8687-8691 . Cameron, S. A., J. N. Derr, A. D. Austin, J. B. Woolley, and R. A. Wharton. 1992. The application of nucleotide sequence data to phylogeny of the Hymenoptera: A review, [ournal of Hymenoptera Research 1: 63-79^ Carpenter, J. M. 1992. Random cladistics. Cladistics 8: 147-153. Carpenter, J. M. and K. C. Nixon. 1997 (1996). On simultaneous analysis. Cladistics 12: 221-241. Chavarria, G. and J. M. Carpenter. 1994. "Total evidence" and the evolu- tion of highly social bees. Cladistics 10: 229-258. Crozier, R. H. and Y. C. Crozier. 1993. The mitochondrial genome of the honeybee Apis mellifenv. complete sequence and genome organiza- tion. Genetics 133: 97-117. Darwin, C. 1859. On the Origin of Species by Means of Natural Selection or the Preservation ofFavoureit Races in the Struggle for Life. London: John Murray. DNASTAR Inc. 1992. Lasergene software, Madison, Wl. Donoghue, M. J., R. G. Olmstead, J. R Smith, and ). D. Palmer 1992. Phy- logenetic relationships of Dipsacales based on rbcL sequences. An- nals of the Missouri Botanical Garden 79: 333-345. Engel, M. S. 1998a. Fossil honey bees and evolution in the genus Apis 136 The University of Kansas Natural History Museum Special Publication No. 24 (Hymenoptera: Apidae). Apidologie 29: 265-28L Engel, M. S. 1998b. A new species of the Baltic amber bee genus Elcctrnpis (Hymenoptera: Apidae). Journal of Hymenoptera Research 7: 94- 101. Engel, M. S. 1999a. The first fossil Ei/g/osa? and phylogeny of the orchid bees (Hymenoptera: Apidae; Euglossini). American Museum Novitates (in press). Engel, M. S. 1999b. The taxonomy of recent and fossil honey bees (Hy- menoptera: Apidae; Apis ). Journal of Hymenoptera Research 8 (in press). EngeL M. S. and T R. Schultz. 1997. Phylogeny and behavior in honey bees (Hymenoptera: Apidae). Annals of the Entomological Society of America 90: 43-53. Faith, D. P. 1991. Cladistic permutation tests for monophyly and nonmonophyly. Systematic Zoology 40: 366-375. Faith, D. P. and P. S. Cranston. 1991. Could a cladogram this short have arisen by chance alone? On permutation tests for cladistic struc- ture. Cladistics 7: 1-28. Fang, Q. Q., S. Cho, J. C. Regier, C. Mitter, M. Matthews, R. W. Poole, T. P. Friedlander, and S. Zhao. 1997. A new nuclear gene for insect phylogenetics: Dopa decarboxylase is informative of relationships within Heliothinae (Lepidoptera: Noctuidae). Systematic Biology 46:269-283. Farris, J. S. 1989. The retention index and the rescaled consistency index. Cladistics 5: 417-^19. Farris, J. S., M. Kallersjo, A. G. Kluge, and C. Bult. 1994. Permutations. Cladistics 10: 65-76. Farris, J. S., M. Kallersjo, A. G. Kluge, and C. Bult. 1995. Constructing a significance test for congruence. Systematic Biology 44: 570-572. Felsenstein, J. 1985a. Confidence limits on phylogenies with a molecular clock. Systematic Zoology 34: 152-161. Felsenstein, J. 1985b. Confidence limits on phylogenies: an approach us- ing the bootstrap. Evolution 39: 783-791. Friedlander, T P., J. C. Regier, and C. Mitter. 1992. Nuclear gene sequences for higher level phylogenetic analysis: 14 promising candidates. Systematic Biology 41 : 483-490. Friedlander, T. P., J. C. Regier, and C. Mitter 1994. Phylogenetic informa- tion content of five nuclear gene sequences in animals: Initial as- sessments of character sets from concordance and divergence stud- ies. Systematic Biology 43: 511-525. Frisch, K. von. 1967. The Dance Language and Orientation of Bees. Cam- bridge, Massachusetts: Harvard University Press. Gamery, L., D. Vautrin, J. M. Comuet, and M. Solignac. 1991. Phyloge- netic relationships in the genus Apis inferred from mitochondrial DNA sequence data. Apidologie 22: 87-92. Gene Codes Corporation. 1995. Sequcnchcr version 3.0. P.O. Box 3369, Ann Arbor, Michigan 48106. Genetics Computer Group. 1991. Program Manual for the GCC Package, version 7, April 1991, 575 Science Drive, Madison, Wisconsin, USA 53711. Hasegawa, M., H. Kishino, and T. Yano. 1985. Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 21: 160-174. Hennig, W. 1979. Phylogenetic Systcmatics. Urbana: University of Illinois Press. Hillis, D. M. 1991. Discriminating between phylogenetic signal and ran- dom noise in DNA sequences. Pp. 278-294, ;'); M. M. Miyamoto and J. Cracraft (eds.) Phylogenetic Analysis ofDNA Sequences. New York: Oxford University Press. Huelsenbeck, J. P. 1991. Tree-length distribution skewness: an indicator of phylogenetic information. Systematic Zoology 40: 257-270. Ito, M. and S. F. Sakagami. 1985. Possible synapomorphies of the para- sitic bumblebees (Psitliyrus) with some nonparasitic bumblebees (Bonihus) (Hymenoptera: Apidae). Sociobiology 10; 105-119. Kallersjo, M. J., J. S. Farris, A. G. Kluge, and C. Bult. 1992. Skewness and permutation. Cladistics 8: 275-287. Kerr, W. E. 1987. Sex determination in bees. XVII. Systems of caste deter- mination in the Apinae, Meliponinae and Bombinae and their phylogenetical implications. Revista Brasileira de Genetica 10: 685- 694. Kimsey, L. S. 1984. A re-evaluation of the phylogenetic relationships in the Apidae (Hymenoptera). Systematic Entomology 9: 435-441. Kimsey, L. S. 1987. Generic relationships within the Euglossini (Hy- menoptera: Apidae). Systematic Entomology 12: 63-72. Kishino, H. and M. Hasegawa. 1989. Evaluation of the maximum likeli- hood estimate of the evolutionary tree topologies from DNA se- quence data, and the branching order in Hominoidea. Journal of Molecular Evolution 29: 170-179. Kluge, A. G. and J. S. Farris. 1969. Quantitative phyletics and the evolu- tion of anurans. Systematic Zoology 18: 1-32. Kreil, G. 1975. The structure of Apis dorsata melittin: phylogenetic rela- tionships between honeybees as deduced from sequence data. FEBS Letters 54:100-102. Larson, A. 1994. The comparison of morphological and molecular data in phylogenetic systematics. Pp. 371-390, In B. Schierwater, B. Streit, G. P. Wagner, and R. DeSalle (eds.). Molecular Ecology and Evolution: Approaches and Applications. Basel: Birkhauser Verlag. Lindauer, M. 1956. Uber die Verstandigung bei indischen Bienen. Zeitschrift fiir Vergleichende Physiologic 38: 521-557. Lindauer, M. 1961. Communication Among Social Bees. Cambridge, Mas- sachusetts: Harvard University Press. Luckow, M. and A. Bnmeau. 1997. Circularity and independence in phy- logenetic tests of ecological hypotheses. Cladistics 13; 145-151. Maddison, W. P. and D. R. Maddison. 1992. MacClade: Analysis of Phylog- eny and Character Evolution, Version 3.06. Computer program and book. Sunderland, Massachusetts: Sinauer. Mason-Gamer, R. J. and E. A. Kellogg. 1996. Testing for phylogenetic con- flict among molecular data sets in the tribe Tritaceae (Gramineae). Systematic Biology 45: 524-545. Mello, M. L. S. and W. E. Kerr. 1984. Histochemistry of salivary gland and Malpighian tubule secretion contributing to the cocoon in Plebeia droryana and Scaptotrigona postica. Zoologischer Anzeiger (Jena) 213: 177-189. Michener, C. D. 1944. Comparative external morphology, phylogeny, and a classification of the bees (Hymenoptera). Bulletin of the American Museum of Natural History 82; 151-326. Michener, C. D. 1969. Comparative social behavior of bees. Annual Re- view of Entomology 14: 299-342. Michener, C. D. 1974. The Social Behavior of the Bees: A Comparative Study. Cambridge, Massachusetts: Harvard University Press. Michener, C. D. 1990. Classification of the Apidae (Hymenoptera). Uni- versity of Kansas Science Bulletin 54: 75-164. Michener, C. D. and D. A. Grimaldi. 1988a. A Trigona from late Creta- ceous amber of New Jersey (Hymenoptera: Apidae: Meliponinae). American Museum Novitates 2917: 1-10. Michener, C. D. and D. A. Grimaldi. 1988b. The oldest fossil bee: Apoid history, evolutionary stasis, and antiquity of social behavior Pro- ceedings of the National Acadademy of Science USA 85: 6424-6426. Pamilo, P., A. Pekkarinen, and S. Varvio-Aho. 1981. Phylogenetic rela- tionships and the origin of social parasitism in Vespidae and in Bombus and Psithyrus as revealed by enzyme genes. Pp. 37-^8, in P. E. Howse and J. L. Clement (eds.), Biosystcinatics of Social Insects. London: Academic Press. Pamilo, P., A. Pekkarinen, and S. Varvio-Aho. 1987. Clustering of bumble- bee subgenera based on interspecific genetic relationships (Hy- menoptera, Apidae: Bombus and Psithyrus). Annals Zoologici Fennici CORBICULATE BeE PhYLOGENY 137 24: 19-27. Pederson, B. V. 1996. A phylogenetic analysis of cuckoo bumblobees iPfitln/rus, Lepeletier) and bumblebees {Bcmtnif, Latreille) inferred from sequences of the mitochondrial gene cytochrome oxidase I. Molecular Phylogenetics and Evolution 5: 289-297. Pereira-Martins, S. R. and W. E. Kerr. 1991. Biologia de Eulaemn uiiiritn. 3. Inferencias evolutivas. Papeis Avulsos de Zoologia (Sao Paulo) 37: 245-250. Plant, J. D. and H. F. Paulus. 1987. Comparative morphology of the postmentum of bees (Hymenoptera: Apoidea) with special remarks on the evolution of the lorum. Zeitschrift fiir Zoologische Systematik und Evolutionforschung 25: 81-103. Prentice, M. 1991. Morphological analysis of the tribes of Apidae. Pp. 51- 69, in D. R. Smith (ed.). Diversity in the genus Apis. Oxford: Westview Press. Richards, O. W. 1968. The subgeneric divisions of the genus Bottibus Latreille (Hymenoptera: Apidae). Bulletin of the British Museum of Natural History (Entomology) 22: 209-276. Rohlf, E ). and R. R. Sokal. 1995. Statistical Tables (Third Edition). New York: W.H. Freeman. Roig-Alsina, A. and C. D. Michener 1993. Studies of the phvlogeny and classification of long-tongued bees (Hymenoptera: Apoidea). Uni- versity of Kansas Science Bulletin 55: 123-162. Ruttner, F. 1988. Biogeography and Taxonomy ofHoneybees. Berlin: Springer- Verlag. Sakagami, S. F. and Y. Maeta. 1984. Multifemale nests and rudimentary castes in the normally solitary bee Ceratinn jnponica (Hymenoptera; Xylocopinae). Journal of the Kansas Entomological Society 57: 639- 656. Sakagami, S. F. and C. D. Michener. 1987. Tribes of Xylocopinae and ori- gin of the Apidae (Hymenoptera: Apoidea). Annals of the Entomo- logical Society of America 80: 439^50. Sakai, S., H. Hoshiba, and E. Hoshiba. 1986. Integrated taxonomic infor- mation on honey bees. Proceedings of the 30th International Con- gress of Apiculture, Nagoya, Japan., Apimondia (1986): 134-139. Sanderson, M.J. and M.J. Donoghue. 1989. Patterns of variation in levels of homoplasy Evolution 43: 1781-1795. Schultz, T. R., R. B. Cocroft, and G. A. Churchill. 1996. The reconstruction of ancestral character states. Evolution 50: 504-511. Sheppard, W. S. and B. A. McPheron. 1991. Ribosomal DNA diversity in Apidae. Pages 89-102, in D. R. Smith (ed.). Diversity in the genus Apis. Oxford: Westview Press. Smith, D. R. 1991. Diversity in the Genus Apis. Oxford: Westview Press. Sokal, R. R. and Rohlf, E J. 1995. Biometry (Third Edition). New York: W. H. Freeman. Swofford, D. L. 1991. PAUP, Phylogenetic Analysis Using Parsimony, ver- sion 3.1. Champaign, Illinois: Illinois Natural History Survey. Swofford, D. L. 1997. PAUP*, Phylogenetic Analysis Using Parsimony, ver- sion 4.0d56 (unpublished test version, used with author's permis- sion). D. L. Swofford, Smithsonian Institution, Washington, DC. Swofford, D. L., G. J. Olsen, P J. Waddell, and D. M. Hillis. 1996. Phylo- genetic inference. Pp. 407-514, in D. M. Hillis, C. Moritz, and B. K. Mable (eds.). Molecular Systematics (Second Edition). Sunderland, Massachusetts: Sinauer Associates. Swofford, D. L., J. L. Thome, J. Felsenstein, and B. M. Wiegmann. 1996. The topology-dependent permutation test for monophyly does not test for monophylv. Systematic Biology 45: 575-579. Templeton, A. R. 1983. Phylogenetic inference from restriction endonu- clease cleavage site maps with particular reference to the evolution of humans and apes. Evolution 37: 221-244. Tingek, S., G. Koeniger, and N. Koeniger. 1996. Description of a new cav- ity nesting species of Apns {Apis luiluensis n. sp.) from Sabah, Borneo, with notes on its occurrence and reproductive biology. Senckenbergiana Biologica 76: 115-119. Trueman, J. 1992. Data Randomiser, Version 3. Computer program and documentation. Distributed by J. Trueman, School of Life Sciences, Australian National University, Canberra, Australia. Wenzel, J. W. 1997. When is a phylogenetic test good enough? Pp. 31-45, in P. Grandcolas (ed.). The Origin of Biodiversity in Insects: Phyloge- netic Tests of Evolutionary Scenarios, Memoir Museum National d'Histoire Naturelle 173.' Wheeler, W. and D. Gladstein. 1993. Malign, Version 1.91. Computer pro- gram and documentation, distributed by the authors, American Museum of Natural History, New York. Wille, A. 1979. Phytogeny and relationships among the genera and sub- genera of the stingless bees (Meliponinae) of the world. Revista de Biologia Tropical 28: 41-64. Williams, P. H. 1985. A preliminary cladistic investigation of relation- ships among the bumble bees (Hymenoptera, Apidae). Systematic Entomology 10: 239-255. Williams, P H. 1991. The bumble bees of the Kashmir Himalaya (Hy- menoptera—Apidae, Bombini). Bulletin of the British Museum of Natural History (Entomology) 60: 1-204. Willis, L. G., M. L. Winston, and B. M. Honda. 1992. Phylogenetic rela- tionships in the honeybee (Genus Apis) as determined by the se- quence of the cytochrome oxidase II region of mitochondrial DNA. Molecular Phylogenetics and Evolution 1: 169-178. Winston, M. L. and C. D. Michener. 1977. Dual origin of highly social behavior among bees. Proceedings of the National Academy of Sci- ence USA 74: 1135-1137. Zeuner, F. E. and F. J. Manning. 1976. A monograph on fossil bees (Hy- menoptera: Apoidea). Bulletin of the British Museum of Natural History (Geology) 27: 155-268 Zrzavy, J. 1997. Phylogenetics and ecology: All characters should be in- cluded in the cladistic analysis. Oikos 80: 186-192. APPENDIX 1 Taxa Represented in Phylogenetic Analyses Outgroups: Exoneura hicolor Smith, Xylocopa virginica (Linnaeus). Euglossini: Eulaema polychroma (Mocsary), Eufriesia caerulescens (Lepeletier). Bombini: Bombus pennsylvanicus (De Geer), Bombus avinovielhis Skorikov, PsitJn/rus variabilis (Cresson). Meliponini: Trigona necrophaga Camargo & Roubik, Trigona nnizoetisis Schwarz, Scnptotrigona hiteipennis (Friese), Melipona coiupressipes (Fabricius). Note: revised iden- tifications of r. )U'crophaga and T. muzoensis (T. "hypogea" and T. "pallens" of Cameron, 1991, 1993) are due to D. Roubik (pers. comm.). Apini: Apiis cerana Fabricius, Apis koschevnikovi Buttel-Reepen, Apis mcllifcra Linnaeus, Apis florca Fabricius, Apis dorsata Fabricius. Morphological Data Matrix Characters 1 1111111112 22222 1234 5 67890 1234567890 12345Taxa 0000 00 00 00 0000000 00 00 20000000 00 0000010100 0000 Exoneura bicolor Xylocopa virginica EUGLOSSINI Eulaemn polychroma 2000000000 0001111100 00021 Eufi-icsia caerulescens 2000000000 0001111100 00021 BOMBINI Bombus pennsylvanicus 2 00010 000 010111110 00011 Bombus avinovielhis 2 00010000 OlOlllUOO 00011 Psithyrus varjabdis 2 10 10111110 10 138 The University of Kansas Natural History Museum Special Publication No. 24 MELIPONINI Trigona necrophaga 1110 111111 Trigona muzoensis 1110111111 ScapHotrigona luteipennis 1110111111 Melipona compressipes 1110111111 APINI Apis cernna Apis koschevnikoi'i Apis meUifera Afiis florea Apis dorsata 1 1 1 1 n 1 1 1 1 1111111111 1111111111 1111111111 1111111111 1110000011 1110 11 1 1 1000001 1 1 1 1000001 1 1111110011 1111110011 1111110 11 1111110 11 1111110 11 1101 1101 1101 1101 1101 1101 1101 1101 1101 Morphological character descriptions 1: Mandibular grooves. (0) Present. (1) Almost or completely absent. (2) Very well developed. 2: Hypopharyngeal plate. (0) Sensory lobes elongate. (1 ) Sensory lobes short, transverse. 3: Stipites. (0) Posteriorlv produced into a dorsal flange overlapping cardines laterally in repose. (1) Not produced posteriorly as a flange. 4: Basistipital process. (0) Normally developed. (1) Reduced. 5: "Basigaleal bar". (0) Not differentiated. (1) Differentiated as a dis- tinct process. 6: Postmentum. (0) Continuous. (1 ) Divided into mentum and lorum. 7: Prosternal constriction. (0) Absent. (1) Present. 8: Prosternal apophyseal pit. (0) Present. (1) Absent. 9: Prosternal setae. (0) Present. (1) Absent. 10: Basisternum. (0) Normally developed. (1) Enlarged. 11: Antero-lateral mesoscutal process. (0) With parascutal carina present. (1) With parascutal carina absent. 12: Metapleural ridge. (0) Extending to postero-lateral corner of mesopleuron. (1) Curved before postero-lateral corner of mesopleuron. 13: Metatibial spurs. (0) Present. ( 1 ) Absent. 14: Auricle. (0) Absent. (1) Present. 15: Strigilis. (0) Without anterior velum. (1) With anterior velum. 16: Stigma. (0) Large. (1 ) Small. 17: Jugal lobe. (0) Present. (1 ) Reduced or absent. 18: Alar papillae. (0) Absent. (1) Present. 19: Gonobase. (0) Normally developed. (1) Reduced or absent. 20: SVIl and SVlll of male. (0) Normally developed. (1) Reduced or absent. 21: Cuticle. (0) Even. (1) With very uneven patches of darker cuticle. 22: Larval food. (0) Not highly supplemented with pharyngeal gland secretions. (1) Llighly supplemented with pharyngeal gland se- cretions. 23: First recurrent vein (Ir-m). (0) Longer, oblique, not angulate or mod- erately so. (1) Short and angulate. 24: Arolia. (0) Present. (1) Greatly reduced. (2) Absent. 25: Corbicula. (0) Absent. (1) Present. APPENDIX 2 Maximum Likelihood Kishino-Hasegawa Test (ML K-H) In order to minimize likelihood differences (and correspondingly mis- leading hypothesis-test results) due to conflict below the tribal level, it was necessary to generate maximum-likelihood (ML) trees for the ML K-FI tests. (We refer the reader to the caveats in the text regarding our use of the ML criterion.) These were constructed as follows: First, the most-parsimonious tree (MPT) for the DNA data (the "DNAMPT," Fig. 3b) was evaluated under eight models involving all combinations of val- ues for the general time-reversible and "HKY85" (Hasegawa et al, 1985) substitution models, rate heterogeneity across sites, and proportion of invariable sites (Swofford et al., 1996). Likelihood-ratio tests of the re- sults (Sokal and Rohlf, 1995) indicated that the most parameter-rich model supplied a significantly better explanation of the data than any of the others (-In likelihood = 3009.1184; comparison with competing less-pa- rameter-rich model: c=|,|= 5.3914; 0.025 < P < 0.01). The parameters supplied by this model were then employed in two ML searches with 10 random-addition heurisfic-search replicates and TBR branch-swapping: (1) an unconstrained search and (2) a constrained search in which the monophyly of the corbiculate bees and the mono- phyly of each of the tribes was enforced. Based on the ML K-H test, nei- ther of the resulting trees was significantly better at explaining the data than the DNAMPT (P = 0.2846 and P = 0.2947, respectively). The tree resulting from the unconstrained search contained questionable features, including a paraphyletic outgroup (caused by (X. virginicn + Euglossini -I- Apini)) and a paraphyleHc Bomlnis; the tree resulting from the con- strained search was identical in topology with the DNAMPT at the tribal level, but disagreed in certain features of within-tribe topologies in the Apini and the Meliponini. The latter tree (the ML "minimal constraint" tree) was then evaluated under the eight models described above, with likelihood-ratio tests indicating that the data were best explained by a general fime-reversible substitution model assuming gamma-distributed among-site rate variation and no sites invariant ("GTR + G") rather than by the most parameter-rich model (-In likelihood = 3002.5343; c-^ = 3.4364;0.05\ Laportea aestuans (Urticaceae) -^jL- Solanum lancaeifolium (single plant; Solanaceae) Fig. L Distribution of plant species in a representative 5 m x 5 m section of the study area. Brachiaria fasciculntn (Gramineae) predominates in the unshaded area, but 2% is Cyathula achyraiithoides (Gramineae). the above patterns in a habitat with high plant species di- versity and complete ground cover. Confronted with a multitude of potential host plants, females might be less restricted in their movements as they forage. Thus, it may be more critical for males to remain at a site that is suitable for broadcasting calls rather than on preferred host plants. Ideally, a superior broadcasting site would be on or adja- cent to preferred host plants and within the home ranges of one or more females. Some song characteristics of neotropical Tettigoniidae further suggest that the acoustic properties of their calling sites may be more important than in temperate species. The songs of neotropical katydids are usually brief and sporadic (Belwood, 1990). Call length is normally less than 1 second, and calls may be repeated only following many seconds to several minutes of silence. These call features may represent adaptations for avoiding foliage-gleaning bats (Belwood, 1990), a major guild of insectivores in the neotropics (Heller, 1995). Sound frequencies are generally high, and they are ultrasonic (>20 kHz) in many species (Morris et al. 1994). Neotropical katydids also move very little (generally <1 m) during a 24-hour period (Belwood, 1990). Restricted movement coupled with brief, sporadic calls could make location of males by females rather diffi- cult (Belwood, 1990). High frequency calls, while more readily localized than low frequency ones (Michelsen and Larsen, 1985), would not propagate well over long dis- tances (Griffin, 1971). Thus, features of the calling site that enhance song broadcasting (see Stephen and Hartley, 1991 ) may be more critical than in other orthopterans. We studied the determinants of caUing sites in a bush katydid, Orophus conspersiis Bruner von Wattenwyl (Tettigoniidae: Phaneropterinae), at La Selva Biological Station in the Caribbean lowlands of Costa Rica. We in- vestigated whether males of O. conspersiis called from sites that (1) were dispersed in accordance with attraction to- ward or repulsion by con- or hetero-specific calling neigh- bors, (2) were situated on or near preferred host plants, or (3) afforded superior broadcasting characteristics. Acknowledgments This paper is dedicated to the memory of Byron Alexander for his devotion to teaching, scholarship, and fellowship. We would like to thank the Organization for Tropical Studies and its La Selva Biological Station for per- mission to use their facilities and reserve and for logistic assistance with this project. The Tinker Fund (Latin Ameri- can Studies), the Department of Entomology, and the Hungerford Fund (Dept. of Entomology) of the Univer- sity of Kansas provided financial support. We thank Or- lando Vargas and G. Stevens for their help in plant identi- fication and several anonymous referees for valuable criti- cism of the manuscript. This paper reports a project con- ducted during Field Entomology 1996, Biology 789 (Uni- versity of Kansas). METHODS Study Site We studied O. conspersiis within a 30 x 5 m plot that included various woody and herbaceous plants and grasses. This site was a patch of weedy growth along the edge of an area originally cleared for agriculture. The plot was gridded by flagging every 2 m along its length and every 1 m along its width. We created narrow paths along the transverse grid lines so that observations could be made without excessive disturbance of the animals. A vegeta- Determinants of Katydid Calling Sites 159 0.02 0.04 0.06 Time (seconds) 0.08 0.10 c > b 5.0 10.0 15.0 Frequency (kHz) 20.0 Fig. 2. Calling song of male Owphus conspcrsus; (a) oscillo- gram, (b) frequency spectrogram. The y-axis is scaled linearly in both (a) and (b). Refer to methods section for equipment. tion map was generated, and the percent cover of all plant species was estimated within each 2 x 1 m section (Fig. 1). Twenty-nine plant species in 18 families were found within the plot. Voucher specimens were collected and deposited in the Herbarium of the University of Kansas Natural His- tory Museum. Natural History of Orophus conspersus Orophiis conspcrsus is a folivorous katydid found throughout the lowlands of Central America (Rentz, 1983; Nickle, 1992). These katydids may breed continuously throughout the year, although the greatest numbers of adults appear during the rainy season (Rentz, 1983). They are excellent leaf mimics when at rest; both sexes exist in 2.0 Time (seconds) 4.0 c a> a) > 5.0 10.0 15.0 Frequency (l 0.05) as indicated by Mann- Whitney U-tests, which examined individual SPL readings at each height and location. Q. CO 70 60 50 ab i K3 Non-broadcast site ^ s Playback sites Fig. 9. Sound pressure levels (SPL) of recorded calls of Orophus consptersus broadcast from a height of 1.1 m at each of 8 different locations in the plot; 6 broadcast sites and 2 non-broad- cast sites. SPLs were measured at a 1 m horizontal distance from the broadcasting loudspeaker. Bars indicate means of at least 6 measurements taken in each of the 4 cardinal directions; means were calculated by converting dB to Pa, averaging the Pa values, and reconverting the averages to dB. Vertical lines indicate ranges. Bars with the same letters are not significantly different from each other (P > 0.05) as indicated by Mann-Whitney U-tests, which ex- amined individual SPL readings at each site. census nights, and the average over all seven nights was 0.77 (Fig. 4). Feeding Preferences Ten of the 24 plant species tested were neither eaten nor palpated by O. conspersus (Table 1). Of the remaining 14 species, only seven were consumed over significantly longer intervals than the ten uneaten species (Fig. 6). O. conspersus did not feed on the tough, fibrous blades of Cyperaceae or Gramineae, but they did eat bracts under seeds or soft seed coats. They did not feed on any parts of Solanaceae (Fig. 6). We observed no correlation between percent cover of a plant species and preference for it in feeding trials (Fig. 6). More importantly, there was no correlation between the popularity of a plant species as a calling site and the pref- erence for it in feeding trials (Fig. 7). For example, H. pa- tens served as a most popular calling site, but O. coiisp>ersiis ate only its young foliage. On the plant, young leaves of H. patens occur only at the apex, whereas O. conspersus was never observed there. Further, only 32% of the broadcast sites were located within a 1.0-m radius of a plant species preferred in feeding trials, and only 19% of the broadcast sites were near preferred plant species other than H. pa- tens. Song Propagation We found that the song of O. conspersus was propa- gated more effectively in open areas and higher above the ground. At an elevation of 0.1 m, playback stimuli were attenuated to a significantly greater extent within the veg- etation of the plot than in adjacent open areas (Fig. 8). And within the vegetation of the plot, attenuation was signifi- cantly greater at 0.1 m than at 0.8 or 1.6 m, but no differ- ence was noted between 0.8 and 1.6 m (Fig. 8). Attenua- tion of test sounds appeared the same whether broadcast from locations used as calling sites or from other locations at comparable sites (Fig. 9). DISCUSSION Calling males in various Orthoptera aggregate at valu- able resources (Campbell and Clarke, 1971; Shaw et al., 1981; Shelly et al., 1987), while others exhibit regular dis- persion patterns (Meixner and Shaw, 1979; Bailey and Thiele, 1983; Schatral and Yeoh, 1990; Rheinlaender and Romer, 1990), presumably as a consequence of territorial defense (Greenfield, 1997) and/or to maxiniize rates of encounter with females (Arak et al., 1990). Nonetheless, we found that coefficients of dispersion of O. conspersus on successive census nights ranged between aggregated Determinants of Katydid Calling Sites 165 and uniform values. These fluctuations probably reflect small sample sizes of individuals and cannot be regarded as conclusive, yet they are consistent with an overall pat- tern of random dispersion. Neither aggregated nor uni- form dispersion would be expected in O. consperstis, as preferred food resources do not form discrete patches, and there exists no evidence for territorial or aggressive be- havior Does the distribution of valuable food resources, or preferred host plants, influence selection of calling sites in O. avispcrsiis? Little information is available on feeding preferences in Tettigoniidae; e.g., neotropical Phaneropterinae are assumed to be rather generalist folivores (Rentz, 1983; Nickle, 1992). Our study indicated, however, that O. co)ispei-stis readily consumed leaves of fewer than 20% of the plant species available in its habitat and harbored no special preference for the more common species. Despite this level of oligophagy, though, we found no evidence that the fine-scale (within habitat) selection of a calling site was influenced by its value as a food re- source. And the one plant species preferred in feeding tri- als that was also a popular calling site, H. pnteiis, is un- likely to serve as a food source in the field. Local acoustics appear to represent the strongest in- fluence on selection of calling sites in O. couspcrsiis, but the impact of this factor too appears limited. The song propagation characteristics of sites near the ground within the vegetation of the plot, where the insects rest during the day, are clearly inferior. Thus, the elevation of calling sites above the ground affords males reduced attenuation of their songs and the potential for advertising to a greater number of receptive females (Paul and Walker, 1979). Re- verberation, and consequently distortion of critical tem- poral features of the song, may also be reduced at these elevated calling sites. Our findings also show that calling at elevations above those typically used would not improve a male's broadcasting ability. Use of such elevated sites also may render him less easily localized by females ap- proaching via walking (see Walker, 1983). Nonetheless, the specific sites selected by calling males are not acoustically distinguished from general locations at their elevation. Rather, males generally select sites on the tallest vegeta- tion, even though they do not avail themselves of the full height of these plants. Possibly, these plants represent vi- sual "beacons" for males and females during crepuscular movements. Additionally, their stems and foliage may serve as superior platforms for calling males. We observed no interest by the many bats at our field plot in O. couspcrsiis or in loudspeaker-broadcast songs of O. conspcrsits, and therefore we do not suggest that the fine-scale selection of calling sites is influenced by any local protection from bats or other phonotactic natural enemies. Our study considered various social and physical fac- tors that may influence where males advertise within their preferred habitat. We found that other than selecting broad- cast sites on the tallest vegetation, males of Orophus coiispersiis appeared to be randomly dispersed within the plot. Neither proximity to preferred food sources nor su- periority of broadcast quality were obvious factors influ- encing advertisement locations of males. Future studies should further examine the orientation of males toward taller plants and their selection of particular sites on those plants. Oviposition substrates, not considered in our study, may also play a role in determining males' calling sites. During our feeding trials, two different females were ob- served attempting to oviposit into leaves of Hcliconia imbricata and Axonopus compressiis, respectively (see Rentz, 1983). If plant species and leaves vary in their quality as potential oviposition substrates, males might preferentially call from the more desirable sites. This would be particu- larlv likely if (last male) sperm precedence and the oppor- tunity to mate with a female just prior to oviposition exist. LITERATURE CITED Alexander, R.D. 1975. Natural selection and specialized chorusing be- havior in acoustical insects. Pp. 35-77, in D. Pimentel (ed.), Insects, Science, and Society. Academic Press, New York. Arak, A., T. Eiriksson, and T. Radesater. 1990. The adaptive significance of acoustic spacing in malehushcrickets Tettigonia viridissinm: a per- turbation experiment. Behavioral Ecology and Sociobiology 26: 1- 7. Bailey, W.J. and D.C.F Rentz. 1990. The Tettigoniidae: Biolog]/. Systenmtics, and Evolution. Crawford House Press. Bathurst, Australia, 395 pp. Bailey, W.J. and D.R. Thiele. 1983. Male spacing behavior in the Tettigoniidae: an experimental approach. Pp. 163-184, in D.T. Gwynne and G.K. Morris (eds.) Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects. Westview Press, Boulder, CO. Belwood, J.J. 1990. Anti-predator defenses and ecology of neotropical forest katydids, especially the Pseudophyllinae. Pp. 8-26, in W.J. Bailey and D.C.R Rentz (eds.). The Tettigoniidae: Biology, Systenmtics, and Evolution. Crawford House Press, Bathurst, Australia. Campbell, D.J, and P.J. Clarke. 1971. Nearest neighbor tests of signifi- cance for nonrandomness in the spatial distribution of singing crick- ets Teleogryllus convnodus (Walker). Animal Behaviour 19: 750-756. Ellis, PE., D.B. Carlisle, and D.J. Osborne. 1965. Desert locusts: sexual maturation delayed by feeding on senescent vegetation. Science 149: 546-547. Feaver, M. 1977. Aspects of behavioral ecology of three species of Orchelimum (Orthoptera: Tettigoniidae). Ph. D. dissertation. Univer- sity of Michigan, Ann Arbor. 199pp. Greenfield, M. D. 1988. Interspecific acoustic interactions among katy- dids (Neonconoccphalus): inhibition-induced shifts in diel periodic- ity. Animal Behaviour 36: 684-695. Greenfield, M.D. 1997. Sexual selection in resource defense polygyny: lessons from terrritorial grasshoppers. Pp. 75-88, in: Choe, J.C. and B.J. Crespi (eds.), Tlie Evolution ofMatnig Systems in Insects and Atacli- nids. Cambridge University Press, Cambridge, U.K. Greenfield, M.D., T.E. Shelly and A, Gonzalez-Coloma. 1989. Territory selection in a desert grasshopper: the maximization of conversion efficiency on a chemically defended shrvib. Journal of Animal Ecol- ogy 58: 761-771. 166 The University of Kansas Natural History Museum Special Publication No. 24 Griffin, D.R. 197L The importance of atmospheric attenuation for the echolocation of bats (Chiroptera). Animal Behaviour 19: 55-6L Heller, K.-G. 1995. Acoustic signalling in palaeotropical bushcrickets (Orthoptera: Tettigonioidea: Pseudophyllidae): does predation pres- sure by eavesdropping enemies differ in the Palaeo- and Neotropics? Journal of Zoology (London) 237: 469^85. Meixner, A.J. and K.C. Shaw. 1979. Spacing and movement of singing Neoconocephalus nehrciscensis males (Tettigoniidae: Copophorinae). Annals of the Entomological Society of America 72: 602-606. Michelsen.A. andO.N. Larsen. 1985. Hearing and sound. Pp. 495-556, ih G.A. Kerkut and L.I. Gilbert (eds.), Compreliensive Insect Pliysiologi/, Biochemistry and Pharmacology. Vol. 6. Pergamon Press, Oxford, U.K. Morris, G.K., A.C. Mason, R Wall, and J.J. Bellwood. 1994. High ultra- sonic and tremulation signals in neotropical katydids {Orthoptera: Tettigoniidae). Journal of Zoology 233(1): 129-163. Nickle, D.A. 1992. Katydids of Panama (Orthoptera: Tettigoniidae). Pp. 142-184, in D. Quintero and A. Aiello (eds.). Insects of Panama and Mesoamerica, Selected Studies. Oxford University Press, London. Osborne, D.J. 1973. Mutual regulation of growth and development in plants and insects. Pp. 33-42, in H.F. Van Emden (ed.), Insect/Plant Relationships. Blackwell, Oxford. Paul, R., and T.J. Walker. 1979. Arboreal singing in a burrowing cricket, Anurogn/llus arboreus. Journal of Comparative PhvsiologyA 132: 217- 223. Rentz, D.C.F. 1983. Orcphus conspersus (Esperanza, Bush Katydid). Pp. 749-750, m D.H. Janzen,(ed.), Costa Rican Natural History. Univer- sity of Chicago Press, Chicago, IL. RheiiJaender, J., and Romer, H. (1990) Acoustic cues for sound localiza- tion and spacing in orthopteran insects. Pp. 248-264, in W.J. Bailey and D.C.F. Rentz (eds.). The Tettigoniidae: Biology, Systematics, and Evolution. Crawford House Press, Bathurst, Australia. Samways, M.J. 1977. Bush cricket interspecific acoustic interactions in the field (Orthoptera, Tettigoniidae). Journal of Natural Historv 11: 155-168. Schatral, A. 1990. Interspecific acoustic behaviour among bushcrickets. Pp. 152-165, in W.J. Bailey and D.C.F Rentz (eds.). The Tettigoniidae: Biology, Si/stetnatics, and Ei'olution. Crawford House Press, Bathurst, Australia. Schatral, A. and P. Yeoh. 1990. Spatial distribution, calling and interspe- cific acoustic interactions in two species of the Australian tettigoniid genus Tymphanophora (Orthoptera: Tettigoniidae). Journal of Zool- ogy 22L 375-390. Shaw, K.C, R.C. North, and A.J. Meixner. 1981. Movement and spacing of singing Amhiycoryplia parvipennis males. Annals of the Entomo- logical Society of America 74: 436^144. Shaw, K.C, R.J. Bitzer, and R.C. North. 1982. Spacing and movement of Neoconocephalus cnsiger males (Conocephalinae: Tettigoniidae). Jour- nal of the Kansas Entomological Society 55(3): 581-592. Shelly, T.E., M.D. Greenfield, and K.R. Downum. 1987. Variation in host plant quality: influences on the mating system of a desert grass- hopper. Animal Behaviour 35: 1200-1209. Sokal, R.R. and FJ. Rohlf. 1981. Biometry: The Principles and Practice of Statistics in Biological Research. 2nd eci. W.H. Freeman and Company, New York. 859 pp. Spooner, J.D. 1968. Pair-forming acoustic svstems of phaneropterine ka- tvdids (Orthoptera, Tettigoniidae). Animal Behaviour 16:197-212. Stephen, R.O. and J.C Hartley. 1991. The transmission of bush-cricket calls in natural environments. Journal of Experimental Biology 155: 227-244. Walker, T.J. 1983. Mating modes and female choice in shorttailed crickets {Anurogr\/llus arboreus) Pp. 240-267, in D.T Gwynne and G.K. Mor- ris (eds.), Orthopteran Mating Systems: Sexual Competition in a Di- verse Group of Insects. Westview Press, Boulder, CO. Byers, G. W., R. H. Hngeii, and R. W. Brook (cds.). Entomological Contributions in Memory of Byron A. Alexander. University of Knmas Natural Histoiy Mufcum Special PiiMicatiou 24. (1999) Pp. 167-170 Revision of the Species of the Subgenera of Exomalopsis Spinola, 1853, Occurring in South America. I: Diomalopsis Michener & Moure, 1957 (Hymenoptera: Apidae), and a Revised Key to the Subgenera By Eduardo A. B. Almeida^ and Fernando A. Silveira' ABSTRACT Two species are considered in Exomalopsis (Diomalopsis): E. hicelhdaris, the type species from southeastern Brazil, and £. alexaiuieri, described here as new, from Paraguay. Characters tradi- tionally used to define Diomalopsis are discussed and new synapomorphies presented. A revised key for the subgenera of Exomalopsis is provided to incorporate characters recently discovered. Kex/words: Taxonomy; Solitary bee; Exomalopsini; Exomalopsis; Diomalopsis; Apidae. INTRODUCTION The scope of Exomalopsis was recently changed by Silveira (1995) who considered Autlioplionila Cockerell, 1897, and Anthophorisca Michener & Moure, 1957, as sub- genera of a separate genus, Aiithophoriila. The remaining species of Exomalopsis were then distributed in four sub- genera: Exomalopsis s.str. (including Megomalopsis Michener & Moure, 1957), Phanomalopsis and Diomalopsis (both de- scribed by Michener & Moure, 1957) and Stilbomalopsis Silveira, 1995. This is the first of three papers dealing with the species of the subgenera of Exomalopsis occurring in South America. In the next two we will treat Phanomalopsis and Exomalopsis s.str. Diomalopsis was described by Michener & Moure (1957) for a single species, Exomalopsis hicclliilaris, from southern Brazil. Here we discuss some characters used by them to define the group; new synapomorphies are indi- cated for the subgenus; and a new species is described. Additionally, a revised key for the subgenera of Exomalopsis is provided. Acknowledgments This paper was prepared with the support of the Bra- zilian council for development of science and technology (CNPq), which granted fellowships to both authors. We are indebted to TL.Griswold, of the USDA-ARS Bee Biol- ogy and Systematics Laboratory and Utah State Univer- sity at Logan, to J.M.F.Camargo, of the Universidade de Sao Paulo at Ribeirao Preto, and to J.S. Moure, from the Universidade Federal do Parana, Curitiba, for the loan of the material examined. D.A.Yanega corrected English mis- takes in the manuscript; and G.W.Fernandes kindly al- lowed us to use his photographic equipment for the prepa- ration of figures. The support by the late B.A.Alexander and C.D.Michener to F.A.S. during his stay at the Univer- sity of Kansas was a stimulus for undertaking this work. MATERIAL AND METHODS Morphological terminology used here is mainly that of Michener (1944), with the following additions: The word vertex is used in reference to the topmost area of the head between the lateral ocelli and the eye. Following Silveira (1995), we refer to the occipital carina and occipital fringe of previous authors as the postocellar ridge and postocellar fringe, respectively. The band of erect plumose hairs along the posterior margin of the scutellum is the sciitellar fringe. The patch of erect plumose hairs on the mid metanotum, contiguous with the scutellar fringe, is the metanotal tuft. Metasomal terga and sterna are referred to, respectively, as T-1, 1-1, etc., and S-1, S-2, etc. In referring to regions of the antenna, it is assumed that the antenna is extended so that its long axis is perpendicular to the plane of the face and parallel to the long axis of the bee's body. Size of punctures is expressed in an absolute but sub- jective scale representing a size class (minute, very fine, moderately coarse and coarse). Density of punctures, how- ever, is expressed relative to the size of punctures (num- ber of puncture diameters between the margins of two clos- est punctures). Thus, distance between dense, fine punc- tures is actually smaller than that between dense, coarse punctures. RESULTS Subgenus Dioiiinlopsis Michener & Moure Diomalopsis Michener & Moure, 1957:431. Type spe- cies: Exomalopsis bicellidaris Michener & Motire, 1957 (origi- nal designation). Laboratorio de Sistematica e Ecologia de Abelhas, Departamento de Zoologia, ICB, Universidade Federal de Minas Gerais. C.P. 486. 31.270-901 Belo Horizonte, MG, BRAZIL. E-mail: ealmeida@mono.icb.ufmg.br 167 168 The University of Kansas Natural History Museum Special Publication No. 24 Fig. 1. Apical fimbriae of T-5 and T-6 of female of (a) E. (Dionmlopsis) akxaiidcri (paratype) and (b) £. (Phanomalopsis) snowi. Scale line = 0.2 mm. In their description of the subgenus, Michener & Moure called the attention to the following characters which distinguish this group from E.wiimlopsis s.str.: 1) head short behind ocelli but without evidence of a postocellar ridge (preoccipital carina in their paper); 2) paraocular carina present; 3) pterostigma longer than length of mar- ginal cell on wing margin, over five times as long as prestigma and much wider than latter; 4) two submarginal cells; 5) seventh sternum of male with broadly truncate distal process bearing thickened or peg-like hairs; eighth sternum with postapodemal part large, truncate apically, strongly constricted basally, hairy; 6) dorsal gonocoxal bridge short. They observed that the lack of a preoccipital carina and presence of paraocular carinae resemble char- acters oi Anthophornln, Aiithoplioriscn and some species of Phanomalopsis, and that characters 3, 5 and 6 above were "completely unlike those of any other Exomalopsis" known to them. The paraocular carinae, however, are also present at least in several species of the subgenera Exomalopsis s.str. and Phanomalopsis (sensu Silveira,1995), being absent only in the species of Stilhonialopsis Silveira examined by us. Fig. 2. Hamuli of right wing of female of (a) E. < Dioiimlopi^is) nh'xanderi (paratype) and of left wing of female of (b) E. (Dioiunlcpsis) hicellularis and (c) E. (Phiwoynalop'sis) snoiui. Scale line = 0.1 mm. In Silveira (1995), Dionmlopsis was defined by the fol- lowing unique apomorphies (only one species was avail- able then): Disc of S-7 of male subtriangular, with basilateral expansions (Fig.Sf in Silveira, 1995; his charac- ter 60-6); disc of S-8 of male with a median longitudinal carina for almost all the sternal length (Fig.9g in Silveira, 1995; his character 66-1—also present, among the Exomalopsini, in Ercmapis Ogloblin and Anthopiiorula) and apical process of S-8 of male long with a single, broad, flat, hairy lobe, separated from disc by long, strongly con- stricted stalk (character 67-9). With the addition of another species, three characters can be established as synapomorphies for the subgenus: 1) two submarginal cells (highly variable in other groups, but consistently three in all other subgenera of Exomalopsis); 2) the apical fimbriae of T-5 and T-6 of females, the hairs of which are dense, long and have straight apices (in other groups the fimbriae are generally made of thin, short, and curved hairs; Fig.l) and 3) the arrangement of the hamuli on the margin of the posterior wing. In all other subgenera of Exomalopsis, the hamuli are closely and evenly spaced; in Diomalopsis, they are characteristically arranged (Fig. 2) with a set of three, close hamuli at either end of the row, and one or two be- tween them. Exomalopsis (Diomalopsis) hicellularis Michener & Moure Fig.2b Exomalopsis (Diomalopsis) hicellularis Michener & Moure, 1957:449-450. Species of Exomalopsis (Diomalopsis) 169 Michener & Moure (1957) gave an extensive descrip- tion for both sexes of thiis species. Holotype male.—Curitiba, Parana, Brazil; September, 1943 [R.B.Lange]. Deposited at the Departamento de Zoologia, Universidade Federal do Parana, Curitiba, Bra- zil (Moure's collection) (Curitiba is at 25° 20' S, 49° 10' W). Paratypes. —One female and one male, same locality, date and collector as holotype. The female is deposited with the holotype, the male in the Snow Entomological Divi- sion of the Natural History Museum, University of Kan- sas, Lawrence, Kansas. One female paratype, same local- ity, November 1, 1956 [C.D.Michener & R.B.Lange], in the collection of C.A.Campos Seabra, which is being moved to the Museu Nacional, Rio de Janeiro, Brazil. Besides the types at Lawrence and Curitiba, the fol- lowing material was examined: two males deposited in the Departamento de Zoologia, Universidade Federal do Parana, Curitiba, Brazil (Moure's collection), one from Curitiba, Parana, Brazil; 950 m.a.s.l., September 27, 1955 [Michener & Lange] and the other from Guarapuava, Parana, Brazil; 1120 m.a.s.l; September 7, 1955 [Michener & Moure] (Guarapuava is at 25°20'S, SrSO'W). Three fe- males (collected on July 20, 1992, May 2 and October 13, 1993) and one male (collected on May 10, 1992) deposited at the Departamento de Biologia, Universidade de Sao Paulo, Ribeirao Preto, Brazil (Camargo's collection), from Salesopolis, Sao Paulo, Brazil; Estac^ao Ecologica da Boraceia, 800-900 m.a.s.l.; 23°32'S, 45°51'W [Wolfgang Wilms]. Exomalopsis (Diomalopsis) alexatideri n.sp. Figs, la, 2a This species can best be distinguished from £. biceUularis by its smaller size (length less than 5 mm in alcxniuicri and more than 6 mm in biceUularis), the brown- ish tegula (in biceUularis it is black), the pale yellow hairs on the mesosoma and the whitish tomentum on the paraocular areas (both black in biceUularis). Female.—Body color: Black, except as follows: apical margin of clypeus dark ferruginous; pregradular region of S-2 to S-4 ferruginous; tarsi, tibial spurs and strigilis light ferruginous; apical half of mandible, flagellum dorsally, pedicel, scape, tegula and wing veins dark brown; tibiae, femora, trochanters, disc of S-1 and S-2 brownish black; flagellum ventrally and pterostigma light brown; wings hyaline with brownish tint. Pubescence: Pale yellow, except: scopa fuscous medi- ally and posteriorly on tibia and posteriorly on basitarsus; whitish on paraocular areas, clypeus, genae, coxae, tro- chanters, femora, propodeum, mesepisterna, venter of mesosoma, laterally on appressed hairs of T-1 and T-2, apical bands of T-3 and T-4, metasomal sterna; apical fim- briae of T-5 and T-6 black, depending on light incidence. Hairs semierect, forming apical fringe on labrum; on clypeus semierect and fine, homogeneously distributed; on supraclypeal area semidecunibent; inferiorly on paraocular area decumbent; on the rest of frons semierect and more densely plumose; on gena, semierect and fine, becoming more erect away from eye; on postocellar fringe short, erect, and densely plumose; on pronotum long erect and densely plumose; on disc of mesoscutum semidecumbent and densely plumose, longer on anterior third; on anterior margin of scutellum short and plumose; scutellar fringe and metanotal tuft long, erect, and densely plumose; on anterior third of propodeum short and semidecumbent, decumbent on lateral areas; on mesepisternum semidecumbent, long, and plumose; lat- erally on marginal areas of T-1 to T-3 forming appressed patches; on posterior edges of T-3 to T-5 forming ill-de- fined bands. Denser on apex than on base of forewing. Punctures: On labrum coarse and dense (one diameter or less apart from each other), leaving a median longitudi- nal band and two apico-lateral areas smooth; on clypeus moderately coarse and sparse (one to two diameters apart), intermixed with finer punctures; on supraclypeal area moderately fine and sparse (one to three diameters apart), becoming denser toward supra-antennal area and leaving a median longitudinal band smooth; on front and vertex fine and sparse (two to four diameters apart); on gena fine and very sparse (three to five diameters apart), becoming denser toward vertex; behind ocelli cciarse and dense (one or less diameter apart); between ocelli fine and sparse (two to four dianieters); on disc of mesoscutum moderately fine and sparse (one to three diameters), becoming denser on posterior third medially; on scutellum moderately fine and sparse (one to three diameters apart) on anterior margin, disc impunctate; on lateral portions of metanotum and on mesepisternum, minute and sparse (two to five diameters apart); on propodeum fine and sparse (two to four diam- eters apart) on anterior third, impunctate posteriorly; dor- sal surface of T-1 mostly shiny and impunctate, but mod- erately fine and sparse (one to three diameters apart) on lateral areas and along transverse carina; on T-2 fine and sparse (two to three diameters apart) laterally and anteri- orly, very minute and sparse (more than five diameters apart) elsewhere; on T-3, and T-4 fine and sparse (two to four diameters apart); on T-5 fine and very dense (less than one diameter apart). Structure: Labrum trapezoidal, disc plane, lateral parts folded back at right angle to disc; clypeus gently convex, disc plane, apical margin delimited by a strongly punc- tured transverse line; frontal sulcus short and ill-defined; hamuli, seven per wing and unevenly spaced; disc of T-1 one-fourth of dorsal surface of tergum. 170 The University of Kansas Natural History Museum Special Publication No. 24 Measurements ofholotype (mm): Approximate length of body = 4.8; of forewing = 4.4. Length and width of head = 1.42, 1.81. Maximum, inferior and superior distance be- tween eyes = 1.06, 1.03, 0.97. Interocellar and ocellar-ocu- lar distances = 0.31, 0.23. Diameters of mid and lateral ocelli = 0.14, 0.13. Length and diameter of scape = 0.49, 0.13. Length of pedicel, T', 2"^, 3"^ and terminal flagellomeres = 0.14, 0.18, 0.11, 0.11, 0.25. Diameter of 5"' flagellomere = 0.13. Length and width of mesoscutum = 1 .06, 1 .53. Length and width of prestigma = 0.14, 0.11. Length and width of pterostigma = 0.76, 0.23. Length and width of niarginal cell (measured on wing margin) = 0.63, 0.36. Male.—Unknown. Holotype female.—Paraguay, San Pedro Cororo, Rio Ypane; Malaise trap; Xl-27/30-1983 [M. Wasbauer]. De- posited at the ARS-USDA Bee Biology & Systematics Labo- ratory, Utah State University, Logan, Utah, U.S.A. (No geo- graphic coordinates were found for the type locality. The Ypane river flows between 56°W and 57°30'W and between 23°S and 23°30'S.) Paratype females. —Same locality, date and collector as holotype. Three deposited with the holotype; one de- posited in the Snow Entomological Division of the Natu- ral History Museum, University of Kansas, Lawrence, Kansas; one in the collection of the Departamento de Zoologia da Universidade Federal de Minas Gerais, Belo Horizonte, Brazil. Key to Subgenera of Exomalopsis In his paper, Silveira (1995) included a key for the sub- genera of Exomalopsis in which the subgenera Exomalopsis, Phanomalopsis and Diomalopsis were separated in a triplet. This was done because the only character available to dis- tinguish clearly the female Diomalopsis was that it possesses two instead of three submarginal cells. Since this charac- ter is highly variable among bees, it was felt that no em- phasis should be given to it in a key. Now, two additional unique characters have been found for the female Diomalopsis, and the examination of thousands of speci- mens of all other subgenera of Exomalopsis suggested that none of them includes species with two submarginal cells. For this reason, a revised key for the subgenera of Exomalopsis is provided below: 1. Vertex in frontal view convex; lateral ocelli bellow level of summit of head; paraocular carina absent; mar- ginal zone of T-1 and T-2 of female smooth and gla- brous; white, dense, apical fascia present on T-2 to T- 4 of female, sometimes interrupted medially; apical process of S-7 of male present as narrow, transverse sclerite fused laterally to arms of disc; apical process of S-8 of male a single bare lobe Stilbomalopsis —Vertex in frontal view straight; lateral ocelli at least partly above level of summit of head; paraocular carina present; marginal zone of T-1 and/or T-2 of female punctate and pilose; apical fascia absent or present on T-2 to T-4 of female; apical process of S-7 of male absent or complex and with two free basi-lateral lobes under ventral surface; apical process of S-8 of male bearing two apical arms (short or long); // a single lobe, the lobe is hairy. 2 2. Vertex of female between ocellus and eye strongly exca- vated; postocellar ridge present, sometimes limited to portion just to sides of lateral ocelli; S-6 of male with median elevated area that broadens toward apex of sternum, forming a carina or spine at each side. ExomalopKis s.str. — Vertex of female between ocellus and eye not excavated or only gently excavated; postocellar ridge absent; S-6 of male entirely flat 3 3. Three submarginal cells; hamuli evenly spaced (Fig. 2c); vertex of female between ocellus and eye gently ex- cavated; pre-marginal line on T-1 of female de- pressed, forming transverse sulcus; apical fringe of T-5 and T-6 thin (sometimes dense), hairs short, their apices arched (Fig.lb); apical process of S-8 of male bearing a pair of arms; S-7 and S-8 of male without peg-like setae Phanomalopsis —Two submarginal cells; hamuli unevenly spaced (Fig.2a,b); vertex of female between ocellus and eye convex; pre-marginal line on T-1 of female not de- pressed; apical fringe of T-5 and T-6 dense, hairs long, their apices straight (Fig. la); apical process of S-8 of male a single bare lobe; S-7 and S-8 of male with peg- like setae Diomalopsis LITERATURE CITED Michener, C. D. 1944. Comparative external morphology, phytogeny, and a classification of the bees (Hymenoptera). Bulletin of the American Museum of Natural History 82: 151-326. Michener, C. D. and J. S. Moure. 1957. A study of the classification of the more primitive non-parasitic anthophorine bees (Hymenoptera, Apoidea). Bulletin of the American Museum of Natural History 112: 395-452. Silveira, F. A. 1995. Phylogenetic relationships and classification of Exomalopsini with a new tribe Teratognathini (Hymenoptera: Apoidea). The University of Kansas Science Bulletin 55: 425^54. Bycis, G. W., R. H. Hngcn, and R. W. Brooks (eds.), Entomological Contributions in Memory of Byron A. Alexander. Uiiiivrgity of Kansas Nntiirnl Histon/ Museum Spccinl Publicnlio}i 24. (1999) Pp. 171-177 New Species, Phylogenetic Placement, and Mammal Associations of Loberopsylhis (Languriidae: Xenoscelinae)^ By Richard A. B. Leschen- and James S. Ashe^ ABSTRACT The phylogenetic position of the genus LobcwpsifUus Martinez and Barrera is consid- ered, and two new species are described: L. ociilatns n. sp. from southern Mexico and L. explaiiatus n. sp. from Costa Rica. A key to the species, an expanded definition of the genus that includes a free- living species, and a review of the biology and the cricetine rodent hosts are included. We hypothesize that phoresy on mammals arose subsequent to the origin of wing reduction in Loberopsylhis. Keywords: Languriidae; Phylogenetic placement; Biology; Mammal hosts; New species; Loberopsyllus. INTRODUCTION Though many species of beetles are associated with mammal nests, there are only a few taxa that are phoretic on their mammal hosts. These are: platypsylline Leiodidae on beavers, mice, shrews and moles (Wood, 1965), amblyopinine Staphylinidae on rodents and opossums (Seevers, 1955; Ashe and Timm, 1987; Ashe et al., 1996), dung beetles on monkeys, wallabies, and sloths (Halffter and Matthews, 1966; Matthews, 1972; Ratcliffe, 1980; Lawrence and Britton, 1991), and the languriid genus Loberopsylhis on rodents (Martinez and Barrera, 1966; Barrera, 1969). Byron Alexander would have been amazed if shown the extraordinary photographs of Loberopsyllus hanging like beads on the rump of a mouse (Barrera, 1969). So it is in his honor that we include this paper in his memorial volume, describe two new species and provide additional host information and comments on the biology of Loberopsyllus. Moreover, the discovery of a "free-living" species from montane Mexico provides fresh insight into character evolution and the systematic placement of Loberopsyllus in Languriidae. Specimens are deposited in the following museums: Cornell University Insect Museum, Ithaca (CUIC); Field Museum of Natural History, Chicago (FMNH); Florida State Collection of Arthropods, Gainesville (FSCA); Mon- tana State University Insect Collection, Bozeman (MONT); R. A. B. Leschen Collection (RALC); Snow Entomological Collection, KU Natural History Museum, University of Kansas, Lawrence (SEMC); T. Lanzewizki Collection, Marburg, Germany (TLAN); Universidad de Costa Rica, San Pedro (UCRS); Zoology Institute, Lund University, Lund (LUND). Several inquiries to obtain type specimens of L. tmiibi and L. Imlffteri from Escuela Nacional de Ciencias Biologicas, Mexico City, Mexico, were unsuccessful; how- ever, a paratype of L. frnubi was located in FSCA. Acknowledgments We thank Roy Danielsson (LUND), Mike Ivie (MONT), Horst Korn, Thomas Lanzewizki, Al Newton (FMNH), Paul Skelley (FSCA) and Quentin Wheeler (CUIC) for loans or gifts of material and Bob Timm (KUNHM, University of Kansas) for comments on mam- mal taxonomy. Sara Taliaferro arranged the figures and Rod Hanley provided a review of the manuscript. Travel to FSCA for RABL was provided by NSF grant DEB-9222863. Genus Loberopsyllus Martinez and Barrera Loberopsylhis Martinez and Barrera, 1966: 11. Type species: Loberopsyllus traubi Martinez and Barrera, 1966: 11 {original designation). Diagnosis.—Dorsal setae sparse; elytral punctation confused; antennal insertions hidden in dorsal view; subocular glandular duct present; procoxal cavities closed externally (or nearly so); mesosternum and mesepimeron fused; broad single knob articulation of meso- and metast- ernum present; posterior condyles of metasternum well- developed; intercoxal process of ventrite I broad; elytral base with thick bead. Description.—Length 1.40-2.94 mm. Body unicolorous, red-brown, dark brown, or black; glabrous or subglabrous; form parallel sided and moderately flat- tened to convex; dorsal setae sparsely distributed, simple, short, appressed. Dorsal and ventral surfaces with microsculpture. Punctation of elytron fine and confused, scutellary striole absent. Head with margin of clypeus Contribution no. 3219 from the Snow Entomological Collection, Museum of Natural History, The University of Kansas. Manaaki Whenua Landcare Research, Private Bag 92 170, 120 Mt Albert Road, Auckland, NEW ZEALAND. E-mail: leschenr@landcare.ci.nz Snow Entomological Collection, Division of Entomology, Natural History Museum, University of Kansas, Lawrence, KS 66045. E-mail: ashe@falcon.cc.ukans.edu 171 172 The University of Kansas Natural History Museum Special Publication No. 24 straight; vertexal line absent; frontoclypeal suture present; subgenal spine present; gular sutures absent; subocular glandular ducts present, arising from below eye, not ex- tending into prothorax. Antennal insertion hidden in dor- sal view; antennomeres of club relatively compact. Ante- rior portion of gula with a distinct, but not well-impressed, straight line. Eyes (Figs. 19, 20) present or absent; when present, with ocular setae. Supraocular line (Figs. 19, 20) and labral bead present, reduced or absent. Pronotum (Figs. 16-18) more or less parallel sided; anterior margin not in- vaginated; sides not explanate; lateral margins smooth; basal pronotal impressions present but poorly developed. Pronotal lateral bead without glandular ducts. Posterior margin of prosternal process emarginate. Procoxal cavi- ties closed externally (or nearly so). Mesosternum and mesepimeron (Figs. 3, 11 ) fused; mesepimeron without pit. Mesosternum (Figs. 3, 11) with or without delimited coxal rests; with broad single knob articulation of the meso- and metasternum (Fig. 3). Metasternum (Figs. 3, 11) with me- dian longitudinal line; subcoxal lines absent; posterior condyles of metasternum well-developed. Metendosternite reduced; anterior tendons and lateral processes absent. Abdominal ventrites I and II connate; ventrite 1 (Figs. 3, 11) with or without subcoxal lines, intercoxal process broad. Abdominal support structures present. Abdominal terg- ite VII hidden beneath elytra. Scutellum (Figs. 5, 12) poorly developed and slightly visible in dorsal view. Elytral base with thickhead; elytra fused in apterous forms. Hind wing reduced to narrow strip or completely absent; submedial fleck weakly developed in brachypterous form. Tibia more or less parallel-sided; outer margin without spines or crenulations; inner margin without tubercles. Tirsomeres slightly lobed below or strongly dilated with modified se- tae in apterous forms (pro- and mesotarsomeres more strongly dilated than metatarsomeres). Male.—Without stridulatory file on head. Venter with- out glandular pores. Aedeagus (Figs. 7, 15) with two elon- gate struts. Spiculum ventrale asymmetrical (Fig. 13). Protarsomeres l-III with modified setae. Female.—Without stridulatory file on head. Stylus inserted subapically on coxite (Figs. 6, 14). Spermatheca c- shaped. Phylogenetic position.—We include L. ociilntus in Loberopsi/lliis because it shares several diagnostic charac- ters with other members; mainly, the mesosternum and mesepimeron fused, meso-metasternum with a single and broad internal articulation, posterior condyles of metast- ernum well developed, and basal margin of elytra with thick bead (Figs. 3, 11, 16-18). Other characters concordant with these are the reduced scutellum (Figs. 5, 12), microsculpture distinct on the body, and the dorsal sur- face glabrous or subglabrous. The genus Lobcwpsifllus was originally described as a member of Cryptophagidae by Martinez and Barrera (1966) but was later transferred to Pharaxonothini (Xenoscelinae) of the Languriidae by Sen Gupta (1968), based on its ex- ternally open procoxal cavities, absence of subcoxal lines on ventrite 1 (Fig. 11), simple tarsomeres, and relatively short trochanters. Among these characters, that of the tro- chanter is based on relative length, which is not clearly defined, and is not considered further here. Though the remaining characters are certainly present in many Pharaxonothini, they are also present in some members of the subfamilies Cryptophilinae, Languriinae, and Setariolinae. This set of characters, therefore, does not con- clusively confirm the placement of Loberopsi/llus in Pharaxonothini. Moreover, the presence of subcoxal lines on ventrite I in L. ociilatus (Fig. 3) is contradictory evidence for placement of the genus in Pharaxonothini (Sen Gupta and Crowson, 1971). Two characters suggest that Loberopsi/Uus may be in- cluded either in Cryptophilinae or Toraminae. These taxa have a broad intercoxal process of ventrite 1 as shown in Figs. 3, 11. Outside of these subfamilies it is present in the genus Paphezia Zablotny and Leschen (1996) which is pro- visionally included in Xenoscelinae. Most Cryptophilinae and all Toraminae have a double knob articulation of the meso-metasternal junction (dicondylic), whereas all Xenoscelinae have a button-like articulation (monocondylic). However, some species of Cn/ptopliilus Reitter and all Xcnosceliuus Grouvelle (Cryptopliilinae) lack condylic articulations that resemble the flattened or broad articulation present in LoberopsyUus (Figs. 3, 11). The posterior condyles of the metasternum in Lohcwpsyllus (Figs. 3, 11) are homologous to areas that sur- round the socket for reception of the intercoxal process of ventrite 1 in other languriid taxa. These areas are typically less developed in other languriids but have the striations that are also present in LoberopsyUus. Posterior condyles are also well developed in Xenoscelimis and somewhat developed in the brachypterous genus Lobosteniuni Reitter (Toraminae; see Leschen, 1997). We conclude that LoberopsyUus is doubtfully placed in Xenoscelinae based on contradictory evidence provided above. Because few cladograms exist for languriids (and the related erotylids) it is difficult to determine whether these characters are primitive or derived, and conclusive subfamilial placement of this genus must await further study. Distribution.—LoberopsyUus is distributed in southern Mexico and Costa Rica. Included species.—LoberopsyUus halffteri Martinez and Barrera, L. explanatus new species, L. oculatus new spe- cies, and L. tnnibi Martinez and Barrera. LOBEROPSYLLUS, NeW SpECIES AND MaMMAL HoSTS 173 Key to the Species of Lohcwps\/Ilus (Fig. 3) with suhcoxal hnes. Hind wing reduced to narrow strip, submedial fleck present. Tarsomeres not markedly 1 . Eye present, supraocular line present (Fig. 19), color black dilated, without modified setae. or dark brown L. ociilatiis - . , , . , j n n n j i jMale.—Internal sac and tlagellum well developed. Eye absent, supraocular line reduced or absent (Fig. 20), subequal in length to aedeagal struts (Fig. 7). color red-brown or brown 2 r t r- i •. i i- • iFemale.—Gonocoxites setose along entire apical mar- 2. Elytral margin distinctly explanate (Fig. 17) gjj, (pjg 5) ''^'' "" "^ Comments.—Loheropsyllus oculatus can easily be dis- Elytral margin not explanate (Fig. 18) 3 tinguished from the remaining species of Loheropsi/llus by 3. Lateral pronotal margins moderately convex and wid- the characters listed in the diagnosis. est at middle, length of aedeagal struts equal to that Etymology.—The name is derived from the Latin word of tegmen L. traiibi ociilatns, meaning having eyes. Lateral pronotal margins parallel; length of aedeagal struts Holotype.—Mexico, Oaxaca, 82 km N. Oaxaca City, almost 2 times that of tegmen L. hnlffteri. hwy. 175, 2900 m, 29.1X.1990, leg R. Baranowski, sifting litter, mixed oak forest (LUND). Loheropsyllus oculatus, new species Paratypes.-Mexico, Oaxaca: 5 specimens, same data (Figs.1-7, 16,19) as holotype (LUND, RALC, SEMC); 3 specimens, same but Diagnosis.-Eyes present; procoxal rests present; ^4 km S. Valle Nacional, 2850 m, 4 IX.1990, pine-oak bo- ventrite 1 with subcoxal lines; margin of elytron not real forest (LUND); 1 specimen (slide-mounted), same but ^^ j^j^^j^ ^ ^ 21 km N. Villa Diaz Ordaz, 3100m, 7,IX.1990, boreal forest exp ana e. ^ (RALC); 2 specimens, 2 mi. S. Cerro Felon, 8-9000 ft.. Description.—Length 1.90-2.08 mm (x = 1.95, N = 8). 3 ix.1982, M. 1. Ivie (MONT). Color of body brown (teneral), dark brown or black; an- tennae, mouthparts and legs dark or light brown. Average width of puncture 0.002 mm; punctation moderately dense dorsally, moderately sparse ventrally, punctures separated by 1-3 diameters. Microsculpture of fine points present on head, pronotum and venter. Body setae fine and sparse, decumbent. Head moderately punctate, punctures of ver- tex separated by about 1-3 diameters; poorly-developed alveolate microsculpture on lateral margin behind eye. Eye (Fig. 19) poorly developed, coarsely faceted, 3^ facets at greatest length. Supraocular line (Fig. 19) present. Antenna (Fig. 2) relatively long, extending to posterior margin of pronotum; relative lengths of antennomeres 1.6 : 2 : 1.3:1.3:1.3:1 : 1 : 1.3 : 1.6:2:2.3. Mandible (Fig. 1) with apex curved; mola spinose. Mentum carinate at middle (Fig. 4). Pronotum (Fig. 16) widest at middle, about 0.70 times as long as wide (pronotal length/maximum pronotal width = 0.69-0.70, x = 0.70); depth = 0.40-0.58, x = 0.51 mm; punctation dense, punctures of disc separated by about 1-2 diameters. Prosternal process slightly emargin- ate at middle. Elytra about 1.39 times as long as wide (elytral length/maximum elytral width - 1.37-1.40, x = 1.39) and 2.38 times as long as pronotum (elytral length/ pronotal length = 2.39-2.41, x = 2.40); glabrous, punctures not strongly impressed, separated by 1-3 diameters. Epipleuron distinct to level of ventrite IV. Mesosternum (Fig. 3) with procoxal rests; areolate punctures present. Mesonotum (Fig. 5) with moderately developed scutellum, fused solidly to sclerotized metanotum. Abdomen with internal support structures on ventrites 11-lV; ventrite 1 Loheropsyllus explanatus, new species (Figs. 8-15, 17, 20) Diagnosis.—Eyes absent; procoxal rests absent; ventrite 1 without subcoxal lines; margin of elytron explana