Encyclopædia Britannica, Ninth Edition/Totemism

by James George Frazer

TOTEMISM.A totem is a class of material objects which a savage regards with superstitious respect, believing that there exists between him and every member of the class an intimate and altogether special relation. The name is derived from an Ojibway (Chippeway) word which was first introduced into literature, so far as appears, by J. Long, an Indian interpreter of last century, who spelt it totam.^[1] The connexion between a man and his totem is mutually beneficent: the totem protects the man, and the man shows his respect for the totem in various ways, by not killing it if it be an animal, and not cutting or gathering it if it be a plant. As distinguished from a fetich, a totem is never an isolated individual, but always a class of objects, generally a species of animals or of plants, more rarely a class of inanimate natural objects, very rarely a class of artificial objects.

Kinds of totems.Considered in relation to men, totems are of at least three kinds:—(1) the clan totem, common to a whole clan, and passing by inheritance from generation to generation; (2) the sex totem, common either to all the males or to all the females of a tribe, to the exclusion in either case of the other sex; (3) the individual totem, belonging to a single individual and not passing to his descendants. Other kinds of totems exist and will be noticed, but they may perhaps be regarded as varieties of the clan totem. The latter is by far the most important of all; and where we speak of totems or totemism without qualification the reference is always to the clan totem.

Clan Totem.The Clan Totem.—The clan totem is reverenced by a body of men and women who call themselves by the name of the totem, believe themselves to be of one blood, descendants of a common ancestor, and are bound together by common obligations to each other and by a common faith in the totem. Totemism is thus both a religious and a social system. In its religious aspect it consists of the relations of mutual respect and protection between a man and his totem; in its social aspect it consists of the relations of the clansmen to each other and to men of other clans. In the later history of totemism these two sides, the religious and the social, tend to part company; the social system sometimes survives the religious; and, on the other hand, religion sometimes bears traces of totemism in countries where the social system based on totemism has disappeared. We begin with the religious side.

Totemism as a Religion.Totemism as a Religion, or the Relation between a Man and his Totem.—The members of a totem clan call themselves by the name of their totem, and commonly believe themselves to be actually descended from it.

Descent from totem.Thus the Turtle clan of the Iroquois are descended from a fat turtle, which, burdened by the weight of its shell in walking, contrived by great exertions to throw it off, and thereafter gradually developed into a man.^[2] The Cray-Fish clan of the Choctaws were originally cray-fish and lived underground, coming up occasionally through the mud to the surface. Once a party of Choctaws smoked them out, and, treating them kindly, taught them the Choctaw language, taught them to walk on two legs, made them cut off their toe nails and pluck the hair from their bodies, after which they adopted them into the tribe. But the rest of their kindred, the cray-fish, are still living underground.^[3] The Osages are descended from a male snail and a female beaver. The snail burst his shell, developed arms, feet, and legs, and became a fine tall man; afterwards he married the beaver maid.^[4] Some of the clans of western Australia are descended from ducks, swans, and other waterfowl.^[5] In Senegambia each family or clan is descended from an animal (hippopotamus, scorpion, &c.) with which it counts kindred.^[6]

Somewhat different are the myths in which a human ancestress is said to have given birth to an animal of the totem species. Thus the Snake clan among the Moquis of Arizona are descended from a woman who gave birth to snakes.^[7] The Bakalai in western equatorial Africa believe that their women once gave birth to the totem animals; one woman brought forth a calf, others a crocodile, hippopotamus, monkey, boa, and wild pig.^[8]

Respect shown to totem.Believing himself to be descended from, and therefore akin to, his totem, the savage naturally treats it with respect. If it is an animal he will not, as a rule, kill nor eat it. In the Mount Gambier tribe (South Australia) "a man does not kill or use as food any of the animals of the same subdivision with himself, excepting when hunger compels; and then they express sorrow for having to eat their wingong (friends) or tumanang (their flesh). When using the last word they touch their breasts, to indicate the close relationship, meaning almost a part of themselves.

To illustrate:—One day one of the blacks killed a crow. Three or four days afterwards a Boortwa (crow) named Larry died. He had been ailing for some days, but the killing of his wingong hastened his death.^[9] The tribes about the Gulf of Carpentaria greatly reverence their totems: if any one were to kill the totem animal in presence of the man whose totem it was, the latter would say, "What for you kill that fellow? that my father!" or "That brother belonging to me you have killed; why did you do it?"^[10] Sir George Grey says of the western Australian tribes that a man will never kill an animal of his kobong (totem) species if he finds it asleep; "indeed, he always kills it reluctantly, and never without affording it a chance to escape. This arises from the family belief that some one individual of the species is their nearest friend, to kill whom would be a great crime, and to be carefully avoided."^[11] Amongst the Indians of British Columbia a man will never kill his totem animal; if he sees another do it, he will hide his face for shame, and afterwards demand compensation for the act. Whenever one of these Indians exhibits his totem badge (as by painting it on his forehead), all persons of the same totem are bound to do honour to it by casting property before it.^[12] The Damaras in South Africa are divided into totem clans, called "eandas"; and according to the clan to which they belong they refuse to partake, e.g., of an ox marked with black, white, or red spots, or of a sheep without horns, or of draught oxen. Some of them will not even touch vessels in which such food has been cooked, and avoid even the smoke of the fire which has been used to cook it.^[13] The negroes of Senegambia do not eat their totems.^[14] The Mundas (or Mundaris) and Oraons in Bengal, who are divided into exogamous totem clans, will not kill or eat the totem animals which give their names to the clans.^[15] A remarkable feature of some of these Oraon totems is that they are not whole animals, but parts of animals, as the head of a tortoise, the stomach of a pig. In such cases (which are not confined to Bengal) it is of course not the whole animal, but only the special part, that the clansmen are forbidden to eat. Such totems may be distinguished as split totems. Split totems.The Jagannáthi Kumhár in Bengal abstain from killing or injuring the totems of their respective clans, and they bow to their totems when they meet them.^[16]

Plant totems. When the totem is a plant the rules are such as these. A native of western Australia, whose totem is a vegetable, "may not gather it under certain circumstances and at a particular period of the year."^[17] An Oraon clan, whose totem is the kujrar tree, will not eat the oil of that tree, nor sit in its shade.^[18] The Red Maize clan of the Omahas will not eat red maize. Those of the people of Ambon and Uliase who are descended from trees may not use these trees for firewood.

Totem taboos.The rules not to kill or eat the totem are not the only taboos; the clansmen are often forbidden to touch the totem or any part of it, sometimes even to look at it.

Thus the Elk clan of the Omahas neither eat the flesh nor touch any part of the male elk.^[19] The Deer-Head clan of the Omahas may not touch the skin of any animal of the deer family, nor wear moccasins of deer skin, nor use the fat of the deer for hair-oil; but they may eat the flesh of deer.^[20] Of the totem clans in Bengal it is said that they "are prohibited from killing, eating, cutting, burning, carrying, using, &c.," the totem.^[21] The Bechuanas in South Africa, who have a well-developed totem system, may not eat nor clothe themselves in the skin of the totem animal.^[22] They even avoid, at least in some cases, to look at the totem. Thus to a man of the Bakuena (Bakwain) or Crocodile clan, it is "hateful and unlucky" to meet or gaze on a crocodile; the sight is thought to cause inflammation of the eyes.

Totem kept in captivity.Sometimes the totem animal is fed or even kept alive in captivity. Among the mountaineers of Formosa each clan or village keeps its totem (serpent, leopard, &c.) in a cage.^[23] A Samoan clan whose totem was the eel used to present the first fruits of the taro plantations to the eels.^[24] Amongst the Narrinyeri in South Australia men of the Snake clan sometimes catch snakes, pull out their teeth or sew up their mouths, and keep them as pets.^[25] In a Pigeon clan of Samoa a pigeon was carefully kept and fed.^[26] Amongst the Kalang in Java, whose totem is the red dog, each family as a rule keeps one of these animals, which they will on no account allow to be struck or ill-used by any one.^[27]

Totem buried and mourned.The dead totem is mourned for and buried like a dead clansman. In Samoa, if a man of the Owl totem found a dead owl by the road side, he would sit down and weep over it and beat his forehead with stones till the blood flowed. The bird would then be wrapped up and buried with as much ceremony as if it had been a human being. "This, however, was not the death of the god. He was supposed to be yet alive, and incarnate in all the owls in existence."^[28] The generalization here implied is characteristic of totemism; it is not merely an individual but the species that is reverenced. The Wanika in eastern Africa look on the hyæna as one of their ancestors, and the death of an hyæna is mourned by the whole people; the mourning for a chief is said to be as nothing compared to the mourning for an hyæna.^[29] A tribe of southern Arabia used to bury a dead gazelle wherever they found one, and the whole tribe mourned for it seven days.^[30] A Californian tribe which reverenced the buzzard held an annual festival at which the chief ceremony was the killing of a buzzard without losing a drop of its blood. It was then skinned, the feathers were preserved to make a sacred dress for the medicine-man, and the body was buried in holy ground amid the lamentations of the old women, who mourned as for the loss of a relative or friend.^[31]

Totem not referred to by name.As some totem clans avoid looking at their totem, so others are careful not to speak of it by its proper name, but use descriptive epithets instead. The three totems of the Delawares—the wolf, turtle, and turkey—were referred to respectively as "round foot," "crawler," and "not chewing," the last referring to the bird's habit of swallowing its food; and the clans called themselves, not Wolves, Turtles, and Turkeys, but "Round Feet," "Crawlers" and "Those who do not chew."^[32] The Bear clan of the Ottawas called themselves not Bears but Big Feet.^[33] The object of these circumlocutions is probably to give no offence to the worshipful animal.

Consequences of disrespect to totem.The penalties supposed to be incurred by acting disrespectfully to the totem are various. The Bakalai think that if a man were to eat his totem the women of his clan would miscarry and give birth to animals of the totem kind, or die of an awful disease.^[34] The Elk clan among the Omahas believe that if any clansman were to touch any part of the male elk, or eat its flesh or the flesh of the male deer, he would break out in boils and white spots in different parts of the body.^[35] The Red Maize subclan of the Omahas believe that, if they were to eat of the red maize, they would have running sores all round their mouth.^[36] And in general the Omahas believe that to eat of the totem, even in ignorance, would cause sickness, not only to the eater, but also to his wife and children.^[37] The worshippers of the Syrian goddess, whose creed was saturated with totemism, believed that if they ate a sprat or an anchovy their whole bodies would break out in ulcers, their legs would waste away, and their liver melt, or that their belly and legs would swell up.^[38]

The Samoans thought it death to injure or eat their totems. The totem was supposed to take up his abode in the sinner's body, and there to gender the very thing which he had eaten till it caused his death.^[39]

Samoan mode of appeasing totem.Thus if a Turtle man ate of a turtle he grew very ill, and the voice of the turtle was heard in his inside saying, " He ate me; I am killing him."^[40] In such cases, however, the Samoans had a mode of appeasing the angry totem. The offender himself or one of his clan was wrapped in leaves and laid in an unheated oven, as if he were about to be baked. Thus if amongst the Cuttle-Fish clan a visitor had caught a cuttle-fish and cooked it, or if a Cuttle-Fish man had been present at the eating of a cuttle-fish, the Cuttle-Fish clan met and chose a man or woman who went through the pretence of being baked. Otherwise a cuttle-fish would grow in the stomach of some of the clan and be their death.^[41]

Australian food taboos. In Australia, also, the punishment for eating the totem appears to have been sickness or death.^[42] But it is not merely the totem which is tabooed to the Australians, they have, besides, a very elaborate code of food prohibitions, which vary chiefly with age, being on the whole strictest and most extensive at puberty, and gradually relaxing with advancing years. Thus young men are for bidden to eat the emu; if they ate it, it is thought that they would be afflicted with sores all over their bodies.^[43]

TOTEMISM 469 The relation between a man and his totem is one of mutual help and protection. If the man respects and cares for the totem, he expects that the totem will do the same by him. In Senegambia the totems, when they are dangerous animals, will not hurt their clansmen; e.g., men of the Scorpion clan affirm that scorpions (of a very deadly kind) will run over their bodies without biting them. 1 A Snake clan (Ophiogenes) in Asia Minor, believing that they were descended from snakes, and that snakes were their kinsmen, submitted to a practical test the claims of any man amongst them whom they suspected of being no true clansman. They made a snake bite him; if he survived, he was a true clansman ; if he died, he was not. 2 The Psylli, a Snake clan in Africa, had a similar test of kinship ; they exposed their new-born children to snakes, and if the snakes left them unharmed or only bit without killing them, the children were legitimate ; otherwise they were bastards. 3 In Senegambia, at the present day, a python is expected to visit every child of the Python clan within eight days after birth. 4 Other totem clans regard a man who has been bitten by the totem, even though he survives, as disowned by the totem, and therefore they expel him from the clan. Among the Crocodile clan of the Bechuanas, if a man has been bitten by a crocodile, or merely had water splashed over him by a crocodile s tail, he is expelled the clan. 5 But it is not enough that the totem should merely abstain from injuring, he must positively benefit the men who put their faith in him. The Snake clan (Ophiogenes) of Asia Minor believed that if they were bitten by an adder they had only to put a snake to the wound and their totem would suck out the poison and soothe away the inflammation and the pain. 6 Hence Omaha medicine men, in curing the sick, imitate the action and voice of their (individual) totem. 7 Members of the Serpent clan in Senegambia profess to heal by their touch persons who have been bitten by serpents. 8 A similar profession was made in antiquity by Snake clans in Africa, Cyprus, and Italy. 9 Again, the totem gives his clansmen important informa tion by means of omens. In the Coast Murring tribe of New South Wales each man s totem warned him of com ing danger ; if his totem was a kangaroo, a kangaroo would warn him against his foes. 10 The Samoan totems gave omens to their clansmen. Thus, if an owl flew before the Owl clan, as they marched to war, it was a signal to go on ; but if it flew across their path, or back wards, it was a sign to retreat. 11 Some kept a tame owl on purpose to give omens in war. 12 When the conduct of the totem is not all that his clansmen could desire, they have various ways of putting pressure on him. Thus, in harvest time, when the birds eat the corn, the Small Bird clan of the Omahas take some corn which they chew and spit over the field. This is thought to keep the birds from the crops. 13 If worms infest the corn the Reptile clan of the Omahas catch some of them and pound them up with some grains of corn which have been heated. They make a soup of the mixture and eat it, believing that the corn will not be infested again, at least for that year. 14 During a fog the men of the Turtle subclan of the Omahas used to draw 1 Revue cV Ethnographic, iii. p. 396. 2 Varro in Priscian, x. 32, vol. i. p. 524, ed. Keil. For the snake descent of the clan, see Strabo, xiii. 1, 14; Julian, N. A., xii. 39. 3 Varro, loc. cit.; Pliny, N. H., vii. 14. Pliny has got it wrong end on. He says that if the snakes did not leave the children they were bastards. We may safely correct his statement by Varro s. 4 Revue d Ethnographie, iii. p. 397. 5 Livingstone, South Africa, p. 255. 6 Strabo, xiii. 1, 14. 7 James, Expedition to the Rocky Mountains, i. p. 247. 8 Revue d Ethnographie, iii. p. 396. 9 Pliny, JT. H., xxviii. 30. 10 J. A. I., xiii. 195 n, xvi. 46. 1 Turner, Samoa, 21, 24, 60. M Ibid., 25 sq. 13 Third Report, p. 238 sq. The idea perhaps is that the birds eat in the persons of their clansmen, and give tangible evidence that they have eaten their fill. " Third Rep., 248. the figure of a turtle on the ground with its face to the south. On the head, tail, middle of the back, and on each leg were placed small pieces of a red breech-cloth with some tobacco. This was thought to make the fog disappear. 15 In order, apparently, to put himself more fully under Man the protection of the totem, the clansman is in the habit assimi - of assimilating himself to the totem by dressing in the skin j^g elf or other part of the totem animal, arranging his hair and to totem. mutilating his body so as to resemble the totem, and repre senting it on his body by cicatrices, tattooing, or paint. Among the Thlinkets on solemn occasions, such as dances, memorial festivals, and burials, individuals often appear disguised in the full form of their totem animals ; and, as a rule, each clans man carries at least an easily recognizable part of his totem with him. 16 Amongst the Omahas, the smaller boys of the Black Shoulder (Buffalo) clan wear two locks of hair in imitation of horns. 17 The Small Bird clan of the Omahas " leave a little hair in front, over the forehead, for a bill, and some at the back of the head, for the bird s tail, with much over each ear for the wings." w The Turtle subclan of the Omahas " cut off all the hair from a boy s head, except six locks ; two are left on each side, one over the forehead, and one hanging down the back in imitation of the legs, head, and tail of a turtle. " 19 The practice of knocking out the upper front teeth at puberty, which prevails in Australia and elsewhere, is, or was once, probably an imitation of the totem. The Batoka in Africa who adopt this practice say that they do so in order to be like oxen, while those who retain their teeth are like zebras. 20 The Haidas of Queen Charlotte Islands are universally tattooed, Tattoo- the design being in all cases the totem, executed in a conventional ing. style. When several families of different totems live together in the same large house, a Haida chief will have all their totems tattooed on his person. 21 Tribes in South America are especially distinguished by their tattoo marks, but whether these are totem marks is not said. 22 The Australians do not tattoo but raise cicatrices ; in some tribes these cicatrices are arranged in patterns which serve as the tribal badges, consisting of lines, dots, circles, semicircles, &c. 23 According to one authority, these Australian tribal badges are sometimes representations of the totem. 24 Again, the totem is sometimes painted on the person of the clans man. This, as we have seen (p. 468), is sometimes done by the Indians of British Columbia. Among the Hurons (Wyandots) each clan has a distinctive mode of painting the face ; and, at least in the case of the chiefs at installation, this painting represents the totem. 25 Among the Moquis the representatives of the clans at foot-races, dances, &c. , have each a conventional representation of his totem blazoned on breast or back. 26 The clansman also affixes his totem mark as a signature to treaties and other documents, 27 and paints or carves it on his weapons, hut, canoe, &c. The identification of a man with his totem appears further to have been the object of various ceremonies observed at birth, marriage, death, and on other occasions. Birth Ceremonies. On the fifth day after birth a child Birth of the Deer-Head clan of the Omahas is painted with red cere : spots on its back, in imitation of a fawn, and red stripes mi are painted on the child s arms and chest. All the Deer- Head men present at the ceremony make red spots on their chests. 28 When a South Slavonian woman has given birth to a child, an old woman runs out of the house and calls out, " A she-wolf has littered a he- wolf," and the child is drawn through a wolfskin, as if to simulate actual birth from a wolf. Further, a piece of the eye and heart of a 5 Third Report, 240. 16 Holmberg in Acta Soc. Scient. Fennicae, iv. 293 sq., 328 ; Petroff, Report on Population, Industries, and Resources of Alaska, p. 166. 17 Third Rep., 229. 18 Ibid., 238. 19 Ibid., 240. 20 Livingstone, South Africa, p. 532. 21 Geolog. Surv. of Canada, Rep. for 1878-79, pp. 108s, 135s ; Smithsonian Contrib. to Knowl., vol. xxi. No. 267, p. 3 sq. ; Nature, 20th January 1887, p. 285 ; Fourth Annual Report of the Bureau of Ethnology, Washington, 1886, p. 67 sq. 22 Martius, Zur Ethnographie America s, zumal Brasiliens, p. 55. 23 Brough Smyth, Aborigines of Victoria, i. p. xli. sq., 295, ii. 313 ; Eyre, Jour., ii. 333, 335; Ridley, Kamilaroi, p. 140; Jour, and Proc. R. Soc. N. S. Wales, 1882, p. 201. 24 Mr Chatfield, in Fison and Howitt, Kamilaroi and Kurnai, p. 66 n. On tattooing in connexion with totemism, see Haberlandt in Mittheil. der anthrop. Gesell. in Wien, xv. (1885) p. [58] sq. 25 First Rep., pp. 62, 64. M Bourke, Snake Dance, p. 229. 27 Heckewelder, Indian Nations, p. 247. 28 Third Rep., p. 245 sq. 470 TOTEMISM wolf are sewed into the child s shirt, or hung round its neck ; and, if several children of the family have died before, it is called Wolf. The reason assigned for some of these customs is that the witches who devour children will not attack a wolf. 1 In other words, the human child is disguised as a wolf to cheat its supernatural foes. The same desire for protection against supernatural danger may be the motive of similar totemic customs, if not of totemism in general. Marriage Marriage Ceremonies. Among the Kalang of Java, cere- whose totem is the red dog, bride and bridegroom before monies. marr i a g e are rubbed with the ashes of a red dog s bones. 2 Among the Transylvanian Gipsies, bride and bridegroom are rubbed with a weasel skin. 3 The sacred goatskin (aegis) which the priestess of Athene took to newly married women may have been used for this purpose. 4 At Rome bride and bridegroom sat down on the skin of the sheep which had been sacrificed on the occasion. 5 An Italian bride smeared the doorposts of her new home with wolf s fat. 6 It is difficult to separate from totemism the custom observed by totem clans in Bengal of marrying the bride and bridegroom to trees before they are married to each other. The bride touches with red lead (a common marriage ceremony) a mahwa tree, clasps it in her arms, and is tied to it. The bridegroom goes through a like ceremony with a mango tree. 7 Death Death Ceremonies. In death, too, the clansman seeks cere- to become one with his totem. Amongst some totem monies. c j ans j^ j g an ar ticle of faith that, as the clan sprang from the totem, so each clansman at death reassumes the totem form. Thus the Moquis, believing that the ancestors of the clans were respectively rattlesnakes, deer, bears, sand, water, tobacco, &c., think that at death each man, accord ing to his clan, is changed into a rattlesnake, a deer, &c. 8 Amongst the Black Shoulder (Buffalo) clan of the Omahas a dying clansman was wrapped in a buffalo robe with the hair out, his face was painted with the clan mark, and his friends addressed him thus : " You are going to the animals (the buffaloes). You are going to rejoin your ancestors. You are going, or your four souls are going, to the four winds. Be strong." 9 Cere- Ceremonies at Puberty. The attainment of puberty is monies at celebrated by savages with ceremonies some of which seem puberty. j. Q ^ Q <ji rec t;iy connected with totemism. The Australian rites of initiation at puberty include the raising of these scars on the persons of the clansmen and clanswomen which serve as tribal badges or actually depict the totem. They also include those mutilations of the person by knocking out teeth, &c., which we have seen reason to suppose are meant to assimilate the man to his totem. At one stage of these Australian rites a number of men appear on the scene howling and running on all fours in imitation of the dingo or native Australian dog ; at last the leader jumps up, clasps his hands, and shouts the totem name "wild dog." 1 " The Coast Murring tribe in New South Wales had an initiatory cere mony at which the totem name "brown snake" was shouted, and a medicine-man produced a live brown snake out of his mouth. 11 As the fundamental rules of totem societies are rules regulating social intercourse, perhaps these pantomimes were intended to supply the youths with a symbolic language by means of which they might communicate with persons speaking different languages, 1 Krauss, Sitte und Branch der Siidslaven, p. 541 sq. 2 Baffles, Hist, of Java, i. 328. On rubbing with ashes as a religious ceremony, cf. Spencer, De Legibus Hebrseorum Ritualibus, vol. ii. diss. iii. lib. iii. cap. 1. 3 Original- Mittheil. aus der ethnolog. Abtheil. der kimigl. Museen Kit Berlin, i. p. 156. * Suidas, s.v. atyis. 5 Servius on Virgil, ^En., iv. 374; Festus, s.v. Innelle. 6 Pliny, Nat. Hist., xxviii. 142. 7 Dalton, Ethn. of Bengal, 194 (Mundas), 319 (Kurmis). Among the Mundas, both bride and bridegroom are sometimes married to mango trees. For Kurmi totems, see As. Quart. Rev., July 1886, p. 77. 8 Schoolcraft, Ind. Tr., iv. 86. 9 Third Re})., p. 229. 10 J. A. I., xiii. 450. " Ibid., xvi. p. 43. and thus ascertain whether they belonged to clans with which marriage was allowed. The totem clans of the Bechuanas have each its special dance or pantomime, and when they wish to ascertain a stranger s clan they ask him, " What do you dance ? " 12 We find elsewhere that dancing has been used as a means of sexual selection. But in some cases these dances seem to be purely re ligious. At their initiatory rites the Yuin tribe in New South Wales mould figures of the totems in earth and dance before them, and a medicine-man brings up out of his inside the " magic " appropriate to the totem before which he stands : before the figure of the porcupine he brings up a stuff like chalk, before the kangaroo a stuff like glass, &c. 13 Again, it is at initiation that the youth is solemnly forbidden to eat of certain foods ; but, as the list of foods prohibited to youths at puberty both in Australia and America extends far beyond the simple totem, it would seem that we are here in contact with those unknown general ideas of the savage, whereof totemism is only a special product. Thus the Narrinyeri youth at initiation are forbidden to eat twenty different kinds of game, besides any food belonging to women. If they eat of these forbidden foods it is thought they will grow ugly. 14 In the Mycoolon tribe, near the Gulf of Carpentaria, the youth at initiation is forbidden to eat of eagle- hawk and its young, native companion and its young, some snakes, turtles, ant-eaters, and emu eggs. 15 The Kurnai youth is not allowed to eat the female of any animal, nor the emu, nor the porcupine. He becomes free by having the fat of the animal smeared on his face. 16 On the other hand, it is said that "initiation confers many privileges on the 3 r ouths, as they are now allowed to eat many articles of food which were previously forbidden to them." 1 Thus in New South Wales before initiation a boy may eat only the females of the animals which he catches ; but after initiation (which, however, may not be complete for several years) he may eat whatever he finds. 18 In North America the Creek youths at puberty were forbidden for twelve months to eat of young bucks, turkey-cocks, fowls, pease, and salt. 19 These ceremonies seem also to be meant to admit the Admis- youth into the life of the clan, and hence of the totem, sion to The latter appears to be the meaning of a Carib ceremony, llfe of in which the father of the youth took a live bird of prey, of a particular species, and beat his son with it till the bird was dead and its head crushed, thus transferring the life and spirit of the martial bird to the future warrior. Further, he scarified his son all over, rubbed the juices of the bird into the wounds, and gave him the bird s heart to eat. 20 Amongst some Australian tribes the youth at initia tion is smeared with blood drawn from the arms either of aged men or of all the men present, and he even receives the blood to drink. Amongst some tribes on the Darling this tribal blood is his only food for two days. Among some tribes the youths at initiation sleep on the graves of their ancestors, in order to absorb their virtues. 21 It is, however, a very notable fact that the initiation of an Australian youth is said to be conducted, not by men of the same totem, but by men of that portion of the tribe into which he may marry. 22 In some of the Victorian tribes no person related to the youth by blood can interfere or assist in his initiation. 23 Whether this is true of all tribes and of all the rites at initiation does not appear. Connected with totemism is also the Australian cere- Resurrec- mony at initiation of pretending to recall a dead man to tlon - life by the utterance of his totem name. An old man lies 12 Livingstone, South Africa, p. 13 ; J. Mackenzie, Ten Years North of the Orange River, p. 391, cf. p. 135 n. ; J. A. I., xvi. 83. 13 Jour, and Proc. R. Soc. N. S. Wales, 1882, p. 206. 14 Nat. Tribes of S. Austral., p. 17. 15 J. A. I., xiii. p. 295. 16 Hid., xiv. p. 316. 17 Ibid., 360. 18 Jour, and Proc. R. Soc. JV. S. Wales, 1882, pp. 208. 19 Gatschet, Migration Legend of the Creek Indians, i. p. 185. 20 Eochefort, Hist. nat. et mor. des lies Antilles (Rotterdam, 1666), p. 556; Du Tertre, Histoire genfrale des Antilles, vol. ii. p. 377. 21 Jour, and Proc. R. Soc. N. S. Wales, 1882, p. 172. 22 Howitt inJ.A.L, xiii. 458. 23 Dawson, Australian Aborigines, p. 30. 471 down in a grave and is covered up lightly with earth ; but at the mention of his totem name he starts up to life. 1 New Sometimes it is believed that the youth himself is killed birth, by a being called Thuremlui, who cuts him up, restores him to life, and knocks out a tooth. 2 Here the idea seems to be that of a second birth, or the beginning of a new life for the novice ; hence he receives a new name at the time when he is circumcised, or the tooth knocked out, or the blood of the kin poured on him. 3 Amongst the Indians of Virginia and the Quojas in Africa, the youths after initiation pretended to forget the whole of their former lives (parents, language, customs, &c.). and had to learn everything over again like new-born babes. 4 A Wolf clan in Texas used to dress up in wolf skins and run about on all fours, howling and mimicking wolves; at last they scratched up a living clansman, who had been buried on purpose, and, putting a bow and arrows in his hands, bade him do as the wolves do rob, kill, and murder. 5 This may have been an initiatory ceremony, revealing to the novice in pantomime the double origin of the clan from wolves and from the ground. For it is a common belief with totem clans that they issued originally from the ground. Piacular Connected with this mimic death and revival of a clans- sacrifice man appear to be the real death and supposed revival of f totem. ^ totem itself. We have seen that some Calif ornian Indians killed the buzzard, and then buried and mourned over it like a clansman. But it was believed that, as often as the bird was killed, it was made alive again. Much the same idea appears in a Zuni ceremony described by an eyewitness, Mr Gushing. He tells how a procession of fifty men set off for the spirit-land, or (as the Zunis call it) " the home of our others," and returned after four days, each man bearing a basket full of living, squirming turtles. One turtle was brought to the house where Mr Gushing was staying, and it was welcomed with divine honours. It was addressed as, " Ah ! my poor dear lost child or parent, my sister or brother to have been ! Who knows which ? May be my own great great grandfather or mother ? " Nevertheless, next day it was killed and its flesh and bones deposited in the river, that it might " return once more to eternal life among its comrades in the dark waters of the lake of the dead." The idea that the turtle was dead was repudiated with passionate sorrow; it had only, they said, " changed houses and gone to live for ever in the home of our lost others. " 6 The mean ing of such ceremonies is not clear. Perhaps, as has been suggested, 7 they are piacular sacrifices, in which the god dies for his people. This is borne out by the curses with which the Egyptians loaded the head of the slain bull. 8 Sex Sex Totems. In Australia (but, so far as is known at totems. p res ent, nowhere else) each of the sexes has, at least in some tribes, its special sacred animal, whose name each individual of the sex bears, regarding the animal as his or her brother or sister respectively, not killing it nor suffer ing the opposite sex to kill it. These sacred animals therefore answer strictly to the definition of totems. Thus amoiigst the Kurnai all the men were called Yeerung (Emu-Wren) and all the women Djeetgun (Superb "Warbler). The birds called Yeerung were the "brothers" of the men, and the birds called Djeetgun were the women s "sisters." If the men killed an emu-wren they were attacked by the women, if the women killed a superb warbler they were assailed by the men. Yeerung and Djeetgun were the mythical ancestors of the Kurnai. 9 1 J. A. I., xiii. 453 sq. 2 76., xiv. 358. 3 Angas, i. 115 ; Brough Smyth, i. 75 n; J. A. L, xiv. 357, 359; Nat. Tr. of S. Austr., pp. 232, 269. 4 R. Beverley, History of Virginia (London, 1722), p. 177 sq. ; Dapper, Description de VAfrique, p. 268. 5 Schoolcraft, Ind. Tr., v. 683. 6 Mr Cashing in Century Magazine, May 1883. 7 See SACRIFICE, vol. xxi. p. 137. 8 Herod., ii. 39. 8 Fison and Howitt, 194, 201 sq., 215, 235. The Eulin tribe in Victoria, in addition to sixteen clan totems, has two pairs of sex totems : one pair (the emu-wren and superb warbler) is identical with the Kurnai pair ; the other pair is the bat (male totem) and the small night jar (female totem). The latter pair extends to the extreme north-western confines of Vic toria as the "man s brother" and the "woman s sister." 10 The Ta-ta-thi group of tribes in New South Wales, in addition to regu lar clan totems, has a pair of sex totems, the bat for men and a small owl for women ; men and women address each other as Owls and Bats ; and there is a fight if a woman kills a bat or a man kills a small owl. 11 Of some Victorian tribes it is said that " the common bat belongs to the men, who protect it against injury, even to the half killing of their wives for its sake. The fern owl, or large goat sucker, belongs to the women, and, although a bird of evil omen, creating terror at night by its cry, it is jealously protected by them. If a man kills one, they are as much enraged as if it was one of their children, and will strike him with their long poles." 12 The sex totem seems to be still more sacred than the clan totem ; for men who do not object to other people killing their clan totem will fiercely defend their sex totem against any attempt of the opposite sex to injure it. 13 Individual Totems. It is not only the clans and the Indi- sexes that have totems ; individuals also have their own vidual special totems, i.e., classes of objects (generally species of animals), which they regard as related to themselves by those ties of mutual respect and protection which are characteristic of totemism. This relationship, however, in the case of the individual totem, begins and ends with the individual man, and is not, like the clan totem, transmitted by inheritance. The evidence for the existence of indi vidual totems in Australia, though conclusive, is very scanty. In North America it is abundant. In Australia we hear of a medicine-man whose clan totem through his mother was kangaroo, but whose "secret" (i.e., individual) totem was the tiger-snake. Snakes of that species, therefore, would not hurt him. 14 An Australian seems usually to get his individual totem by dreaming that he has been transformed into an animal of the species. Thus a man dreamed three times he was a kangaroo ; hence he became one of the kangaroo kindred, and might not eat any part of a kangaroo on which there was blood ; he might not even carry home one on which there was blood. He might eat cooked kangaroo ; but, if he were to eat the meat with the blood on it, the spirits would no longer take him up aloft 15 In America the individual totem is usually the first animal of which a youth dreams during the long and generally solitary fasts which American Indians observe at puberty. He kills the animal or bird of which he dreams, and henceforward wears its skin or feathers, or some part of them, as an amulet, especially on the war-path and in hunting. 16 A man may even (though this seems exceptional) acquire several totems in this way ; thus an Ottawa medicine-man had for his individual totems the tortoise, swan, woodpecker, and crow, because he had dreamed of them all in his fast at puberty. The respect paid to the individual totem varies in different tribes. Among the Slave, Hare, and Dogrib Indians a man may not eat, skin, nor if possible kill his individual totem, which, in these tribes is said to be always a carnivorous animal. Each man carries with him a picture of his totem (bought of a trader) ; when he is unsuccessful in the chase, he pulls out the picture, smokes to it, and makes it a speech. 17 The Indians of Canada changed their okki or manitoo (indivi dual totem) if they had reason to be dissatisfied with it ; their women had also their okkis or manitoos, but did not pay so much heed to them as did the men. They tattooed their individual totems on their persons. 18 Amongst the Indians of San Juan Capistrano, a figure of the individual totem, which was acquired as usual by fasting, was moulded in a paste made of crushed herbs on the right arm of the novice. Fire was then set to it, and thus the figure of the totem was burned into the flesh. 19 Sometimes the individual totem is not acquired by the individual himself at puberty, but is 10 J. A. L, xv. p. 416 ; cf. xii. p. 507. 11 Ibid., xiv. 350. ^ Dawson, Australian Aborigines, p. 52. 13 J. A. L, xiv. p. 350. 14 Ibid., xvi. p. 50. 15 Ibid., 45. 16 Catlin, JV. Amer. Indians, i. p. 36 sq. ; Schoolcraft, Ind. Tr., v. p. 196 ; Id., Amer. Ind., p. 213 ; Sproat, Scenes and Studies of Savage Life, p. 173 sq. ; Bancroft, i. 283 sq. ; Id., iii. 156; Mayne, Brit. Columb., p. 302; P. Jones, Hist. Ojebway Ind., p. 87 sq., &c. 17 Annual Report of the Smithsonian Institution for 1866, p. 307. 18 Charlevoix, Hist, de la Nouv. Fr., vi. 67 sq. The word okki is Huron ; manitoo is Algonkin (ibid. ; Sagard, Le grand Voyage du pays des Hurons, p. 231). 19 Boscana in A. Eobinson s Life in California, pp. 270 sq., 273; Bancroft, i. 414, iii. 167 sq. 472 TOTEMISM fixed for him independently of his will at birth. Thus among the tribes of the isthmus of Tehuantepec, when a woman was about to be confined, the relations assembled in the hut and drew on the floor figures of different animals, rubbing each one out as soon as it was finished. This went on till the child was born, and the figure that then remained sketched on the ground was the child s tona or totem. When he grew older the child procured his totem animal and took care of it, believing that his life was bound up with the animal s, and that when it died he too must die. 1 Similarly in Samoa, at child-birth the help of several gods " was invoked in succession, and the one who happened to be addressed at the moment of the birth was the infant s totem. These " gods " were dogs, eels, sharks, lizards, &c. A Samoan had no objection to eat another man s "god" ; but to eat his own would have been death or injury to him. 2 Sometimes the okkis or mauitoos acquired by dreams are not totems but fetiches, being not classes of objects but individual objects, such as a particular tree, rock, knife, pipe, &c. 3 Besides the clan totem, sex totem, and individual totem, there are (as has been indicated) some other kinds or varieties of totems ; but the consideration of them had better be deferred till after the consideration of the social organization based on totemism. Blood Social Aspect of Totemism, or the relation of the men of feud. a totem to each other and to men of other totems. (1) All the members of a totem clan regard each other as kins men or brothers and sisters, and are bound to help and protect each other. The totem bond is stronger than the bond of blood or family in the modern sense. This is ex pressly stated of the clans of western Australia and of north-western America, 4 and is probably true of all societies where totemism exists in full force. Hence in totem tribes every local group, being necessarily composed (owing to exogamy) of members of at least two totem clans, is liable to be dissolved at any moment into its totem elements by the outbreak of a blood feud, in which husband and wife must always (if the feud is between their clans) be arrayed on opposite sides, and in which the children will be arrayed against either their father or their mother, according as de scent is traced through the mother or through the father. 5 In blood feud the whole clan of the aggressor is responsible for his deed, and the whole clan of the aggrieved is entitled to satisfaction. 6 Nowhere perhaps is this solidarity carried farther than among the Goajiros in Colombia, South Ame rica. The Goajiros are divided into some twenty to thirty totem clans, with descent in the female line ; and amongst them, if a man happens to cut himself with his own knife, to fall off his horse, or to injure himself in any way, his family on the mother s side immediately demand payment as blood-money from him. " Being of their blood, he is not allowed to spill it without paying for it." His father s family also demands compensation, but not so much. 7 Exo- (2) Exogamy. Persons of the same totem may not gamy, marry or have sexual intercourse with each other. The Navajos believe that if they married within the clan " their bones would dry up and they would die." 8 But the penalty for infringing this fundamental law is not merely natural; the clan steps in and punishes the offenders. In Australia the regular penalty for sexual intercourse with a person of a forbidden clan is death. It matters not whether the woman be of the same local group or has been captured in war from another tribe ; a man of the wrong 1 Bancroft, i. 661. 2 Turner, Samoa, 17. 3 Lafitau, Mceursdes Sauvages Ameriquains, i. 370 sq.; Charlevoix, Hist, de la Nouv. Fr., vi. 68 ; Kohl, Kitchi Garni, i. 85 sq. 4 Grey, Jour., ii. 231 ; Report of the Smithsonian Inst. for 1866, p. 315; Petroff, Rep. on Alaska, p. 165. Other authorities speak to the superiority of the totem bond over the tribal bond (Morgan, League of the Iroquois, p. 82; Mayne, Brit. Columb., p. 257; American Antiquarian, ii. p. 109). 5 Grey, Journals, ii. 230, 238 sq.; Smithsonian Rep., loc. cit. 6 Fison and Howitt, 156 sq., 216 sq. Sometimes the two clans meet and settle it by single combat between picked champions (Jour, and Proc. R. Soc. &. S. Wales, 1882, p. 226). 7 Simons in Proc. R. Geogr. Soc., Dec. 1885, p. 789 sq. 8 Bourke, Snake Dance of the Moquis of Arizona, p. 279. clan who uses her as his wife is hunted down and killed by his clansmen, and so is the woman ; though in some cases, if they suc ceed in eluding capture for a certain time, the offence may be con doned. In the Ta-ta-thi tribe, New South Wales, m the rare cases which occur, the man is killed but the woman is only beaten or speared, or both, till she is nearly dead ; the reason given for not actually killing her being that she was probably coerced. Even in casual amours the clan prohibitions are strictly observed ; any violations of these prohibitions "are regarded with the utmost abhorrence and are punished by death." 9 An important exception to these rules, if it is correctly reported, is that of the Port Lincoln tribe, which is divided into two clans, Mattiri and Karraru, and it is said that though persons of the same clan never marry, yet " they do not seem to consider less virtuous connexions between parties of the same class [clan] incestuous." 10 Again, of the tribes on the lower Murray, lower Darling, &c., it is said that though the slight est blood relationship is with them a bar to marriage, yet in their sexual intercourse they are perfectly free, and incest of every grade continually occurs. 11 In America the Algonkins consider it highly criminal for a man to marry a woman of the same totem as himself, and they tell of cases where men, for breaking this rule, have been put to death by their nearest relations. 12 In some tribes the marriage prohibition only extends to Phratries. a man s own totem clan ; he may marry a woman of any totem but his own. This is the case with the Haidas of the Queen Charlotte Islands, 13 and, so far as appears, the Narrinyeri in South Australia, 14 and the western Aus tralian tribes described by Sir George Grey. 15 Oftener, however, the prohibition includes several clans, in none of which is a man allowed to marry. For such an exogamous group of clans within the tribe it is convenient to have a name ; we shall therefore call it a phratry (L. H. Morgan), defining it as an exogamous division intermediate between the tribe and the clan. The evidence goes to show that in many cases it was originally a totem clan which has undergone subdivision. The Choctaws, for example, were divided into two phratries, American each of which included four clans; marriage was prohibited be- phratries. tween members of the same phratry, but members of either phratry could marry into any clan of the other. 16 The Seneca tribe of the Iroquois was divided into two phratries, each including four clans, the Bear, Wolf, Beaver, and Turtle clans forming one phratry, and the Deer, Snipe, Heron, and Hawk clans forming the other. Originally, as among the Choctaws, marriage was prohibited within the phratry but was permitted with any of the clans of the other phratry ; the prohibition, however, has now broken down, and a Seneca may marry a woman of any clan but his own. Hence phratries, in our sense, no longer exist among the Scnecas, though the organization survives for certain religious and social purposes. 17 The phratries of the Thlinkets and the Jlohegans deserve especial attention, because each phratry bears a name which is also the name of one of the clans included in it. The Thlinkets are divided as follows : Raven phratry, with clans Haven, Frog, Goose, Sea- Lion, Owl, Salmon ; Wolf phratry, with clans Wolf, Bear, Eagle, Whale, Shark, Auk. Members of the Raven phratry must marry members of the Wolf phratry, and vice versa. iS Considering the prominent parts played in Thlinket mythology by the ancestors of the two phratries, and considering that the names of the phratries are also names of clans, it seems probable that the Raven and Wolf were the two original clans of the Thlinkets, which afterwards by subdivision became phratries. This was the opinion of the Russian missionary Veniaminoff, the best early authority on the tribe. 19 Still more clearly do the Mohegan phratries appear to have been formed by subdivision from clans. They are as follows : 20 Wolf phratry, with clans Wolf, Bear, Dog, Opossum; Turtle 9 Howitt in Rep. of Smithsonian Inst. for 1883, p. 804 ; Fison and Howitt, pp. 64-67, 289, 344 sq.; J. A. I., xiv. p. 351 sq. 10 Nat. Tr. of S. Australia, p. 222. 11 Jour, and Proc. R. Soc. N. S. Wales, 1883, p. 24; Transactions of the Royal Society of Victoria, vi. p. 16. 12 James in Tanner s Narr., p. 313. 13 Geol. Sur. of Canada, Rep. for 1878-79, p. 134s. 14 Nat. Tr. of S. Austr., p. 12; J. A. /., xii. p. 46. 15 Grey, Jour., ii. p. 226. 16 Arch&ologia Americana, Trans, and Collect. Americ. Antiq. Soc., vol. ii. p. 109; Morgan, A. S., pp. 99, 162. 17 Morgan, op. cit., pp. 90, 94 sq. 18 A. Krause, Die Tlinkit-Indianer, 112, 220; Holmberg, op. cit., 293, 313; Pinart in Bull. Soc. Anthrop. Paris, 7th Nov. 1872, p. 792 sq. ; PetrofF, Rep. on Alaska, p. 165 sq. 19 Petroff, op. cit., p. 166. 20 Morgan, p. 174. phratry, with clans Little Turtle, Mud Turtle, Great Turtle, Yellow Eel; Turkey phratry, with clans Turkey, Crane, Chicken. Here we are almost forced to conclude that the Turtle phratry was origin ally a Turtle clan which subdivided into a number of clans, each of which took the name of a particular kind of turtle, while the Yellow Eel clau may have been a later subdivision. Thus we get a probable explanation of the origin of split totems; they seem to have arisen by the segmentation of a single original clan, which had a whole animal for its totem, into a number of clans, each of which took the name either of a part of the original animal or of a subspecies of it. "We may conjecture that this was the origin of the Grey Wolf and Yellow Wolf and Great Turtle and Little Turtle clans of the Tuscarora-Iroquois;^[44] the Black Eagle and White Eagle and the Deer and Deer-Tail clans of the Raws;^[45] and of the Highland Turtle (striped), Highland Turtle (black), Mud Turtle, and Smooth Large Turtle clans of the Wyandots (Hurons).^[46] Warren actually states that the numerous Bear clan of the Ojibways was formerly subdivided into subclans, each of which took for its totem some part of the Bear's body (head, foot, ribs, &c.), but that these have now merged into two, the Common Bear and the Grizzly Bear.^[47] The subdivision of the Turtle (Tortoise) clan, which on this hypo thesis has taken place among the Tuscarora-Iroquois, is nascent among the Onondaga-Iroquois, for among them "the name of this clan is Hahnowa, which is the general word for tortoise; but the clan is divided into two septs or subdivisions, the Hanyatengona, or Great Tortoise, and the Nikahnowaksa, or Little Tortoise, which together are held to constitute but one clan. "^[48]

On the other hand, fusion of clans is known to have taken place, as among the Haidas, where the Black Bear and Fin-Whale clans have united;^[49] and the same thing has happened to some extent among the Omahas and Osages.^[50]

Australian phratries.In Australia the phratries are still more important than in America. Messrs Howitt and Fison, who have done so much to advance our knowledge of the social system of the Australian aborigines, have given to these exogamous divisions the name of classes; but the term is objection able, because it fails to convey (1) that these divisions are kinship divisions, and (2) that they are intermediate divisions; whereas the Greek term phratry conveys both these meanings, and is therefore appropriate. We have seen examples of Australian tribes in which members of any clan are free to marry members of any clan but their own; but such tribes appear to be excep tional. Often an Australian tribe is divided into two (exogamous) phratries, each of which includes under it a number of totem clans; and of tener still there are subphratries interposed between the phratry and the clans, each phratry including two subphratries, and the subphratries including totem clans. We will take examples of the former and simpler organization first.

The Turra tribe in Yorke Peninsula, South Australia, is divided into two phratries, Wiltu (Eaglehawk) and Miilta (Seal). The Eaglehawk phratry includes ten totem clans (Wombat, Wallaby, Kangaroo, Iguana, Wombat-Snake, Bandicoot, Black Bandicoot, Crow, Rock Wallaby, and Emu); and the Seal phratry includes six (Wild Goose, Butterfish, Mullet, Schnapper, Shark, and Salmon). The phratries are of course exogamous, but (as with the Choctaws, Mohegan, and, so far as appears, all the American phratries) any clan of the one phratry may intermarry with any clan of the other phratry^[51]. But the typical Australian tribe is divided into two exogamous phratries; each of these phratries is subdivided into two subphratries; and these subphratries are subdivided into an indefinite number of totem clans. The phratries being exogamous, it follows that their subdivisions (the subphratries and clans) are so also. The well-known Kamilaroi tribe in New South Wales will serve as an example. Its subdivisions are as follows:^[52]—

Phratries. Subphratries. Totem Clans. Dilbi. Kupathin. Muri.^[53] Kubi. Ipai. Kumbo. Kangaroo, Opossum, Bandicoot, Padimelon, Iguana, Black Duck, Eaglehawk. Scrub Turkey, Yellow-Fish, Honey-Fish, Bream. Emu, Cai pet-Snake, Black Snake, Ked Kangaroo, Honey, Walieroo, Frog, Cod-Fish.

In such tribes the freedom of marriage is still more curtailed. A subphratry is not free to marry into either snbphratry of the other phratry; each subphratry is restricted in its choice of partners to one subphratry of the other phratry; Muri can only marry Kumbo, and vice versa; Kubi can only marry Ipai, and vice versa. Hence (supposing the tribe to be equally distributed between the phratries and subphratries), whereas under the two phratry and clan system a man is free to choose a wife from half the women of the tribe, under the phratry, subphratry, and clan system he is restricted in his choice to one quarter of the women.

Equivalence of Australian tribal divisions.A remarkable feature of the Australian social organization is that divisions of one tribe have their recognized equivalents in other tribes, whose languages, including the names for the tribal divisions, are quite different. A native who travelled far and wide through Australia stated that " he was furnished with temporary wives by the various tribes with whom he sojourned in his travels; that his right to these women was recognized as a matter of course; and that he could always ascertain whether they belonged to the division into which he could legally marry, though the places were 1000 miles apart, and the languages quite different. "^[54] Again, it is said that " in cases of distant tribes it can be shown that the class divisions correspond with each other, as for instance in the classes of the Flinders river and Mitchell river tribes; and these tribes are separated by 400 miles of country, and by many intervening tribes. But, for all that, class corresponds to class in fact and in meaning and in privileges, although the name may be quite different and the totems of each dissimilar."^[55] Particular information, however, as to the equivalent divisions is very scanty.^[56] This systematic correspondence between the intermarrying divisions of distinct and distant tribes, with the rights which it conveys to the members of these divisions, points to sexual communism on a scale to which there is perhaps no parallel elsewhere, certainly not in North America, where marriage is always within the tribe, though outside the clan.^[57] But even in Australia a man is always bound to marry within a certain kinship group; that group may extend across the whole of Australia, but nevertheless it is exactly limited and defined. If endogamy is used in the sense of prohibition to marry outside of a certain kinship group, whether that group be exclusive of, inclusive of, or identical with the man's own group, then marriage among the totem societies of Australia, America, and India is both exogamous and endogamous; a man is forbidden to marry either within his own clan or outside of a certain kinship group.^[58]

Rules of descent.(3) Rules of Descent.—In a large majority of the totem tribes at present known to us in Australia and North America descent is in the female line; i.e., the children belong to the totem clan of their mother, not to that of their father. In Australia the proportion of tribes with female to those with male descent is as four to one; in America it is between three and two to one.

As to the totem tribes of Africa, descent among the Damaras is in the female line,^[59] and there are traces of female kin among the Bechuanas.^[60] Among the Bakalai property descends in the male line, but this is not a conclusive proof that descent is so reckoned; all the clans in the neighbourhood of the Bakalai have female descent both for blood and property.^[61] In Bengal, where there is a considerable body of totem tribes, Mr Risley says that after careful search he and his coadjutors have found no tribe with female descent, and only a single trace of it in one.^[62] Among the totem tribes of Bengal descent is male.^[63] In Assam the exogamous totem clans of the Kasias have female descent,^[64] as also have the exogamous clans of the Garos, but it does not appear whether their clans are totem clans, though some of their legends point to totemism.^[65]

Indirect descent.In the Australian tribal organization of two phratries, four subphratries, and totem clans, there occurs a peculiar form of descent of which no plausible explanation has yet been offered. It seems that in all tribes thus organized the children are born into the subphratry neither of their father nor of their mother, and that descent in such cases is either female or male, according as the subphratry into which the children are born is the companion subphratry of their mother's or of their father's subphratry. In the former case we have what may be called indirect female descent; in the latter, indirect male descent. But it is only in the subphratry that descent is thus indirect. In the totem clan it is always direct; the child belongs to the clan either of its mother or of its father. Thus, in the typical Australian organization, descent, whether female or male, is direct in the phratry, indirect in the sub phratry, and direct in the clan.

To take examples, the following is the scheme of descent, so far as the phratries and subphratries are concerned, in the Kamilaroi:—

Phratries.		Male.	Marries	Children are
Dilbi.	1	Muri.	Kumbo.	Ipai.
		Kubi.	Ipai.	Kumbo.
Kupathin.	2	Ipai.	Kubi.	Muri.
		Kumbo.	Muri.	Kubi.

This is an example of indirect female descent, because the children belong to the companion subphratry of their mother, not to the companion subphratry of their father. But in the totems the female descent is direct; e.g., if the father is Muri-Kangaroo and the mother is Kumbo-Emu, the children will be Ipai-Emu; if the mother is Kumbo-Bandicoot, the children will be Ipai-Bandicoot.^[66]

The following is the scheme of descent in the Kiabara tribe:^[67]—

Phratries.		Male.	Marries	Children are
Dilebi.	1	Baring.	Bundah.	Turowine.
		Turowine.	Bulcoin.	Baring.
Cubatine.	2	Bulcoin.	Turowine.	Bundah.
		Bundah	Baring.	Bulcoin.

This is an example of indirect male descent, because the children belong to the companion subphratry of their father, not to the companion subphratry of their mother. We have no information as to the totems, but on the analogy of indirect female descent we should expect them to be taken from the father. This at any rate is true of a large tribe or group of tribes to the south of the Gulf of Carpentaria; their rules of marriage and descent, so far as concerns the subphratries, are like those of the Kiabara, and the totems (which at the lower Leichhardt river are the names of fish) are inherited from father to son.^[68]

In some Australian tribes sons take their totems from their father and daughters from their mother. Thus the Dieri in South Australia are divided into two phratries, each of which includes under it sixteen totem clans, (Caterpillar, Mullet, Dog, Rat, Kangaroo, Frog, Crow, &c.);^[69] and if a Dog man marries a Rat woman, the sons of this marriage are Dogs and the daughters are Rats.^[70] The Ikula (Morning Star) tribe, at the head of the Great Australian Bight, has, with certain exceptions, the same rule of descent.^[71]

Transformation from female to male descent.Besides the tribes whose line of descent is definitely fixed in the female or male line, or, as with the Dieri and Ikula, half-way between the two, there are a number of tribes among whom a child may be entered in either his mother's or his father's clan. Among the Haidas, children regularly belong to the totem clan of their mother; but in very exceptional cases, when the clan of the father is reduced in numbers, the newly-born child may be given to the father's sister to suckle. It is then spoken of as belonging to the paternal aunt, and is counted to its father's clan.^[72] Among the Delawares descent is regularly in the female line; but it is possible to transfer a child to its father's clan by giving it one of the names which are appropriated to the father's clan.^[73] In the Hervey Islands, South Pacific, the parents settled beforehand whether the child should belong to the father's or mother's clan. The father usually had the preference; but sometimes, when the father's clan was one which was bound to furnish human victims from its ranks, the mother had it adopted into her clan by having the name of her totem pronounced over it.^[74] In Samoa at the birth of a child the father's totem was usually prayed to first; but if the birth was tedious, the mother's totem was invoked; and whichever happened to be invoked at the moment of birth was the child's totem for life.^[75]

Arrangement of clans on the march, &c.When a North American tribe is on the march, the members of each totem clan camp together, and the clans are arranged in a fixed order in camp, the whole tribe being arranged in a great circle or in several concentric circles.^[76] When the tribe lives in settled villages or towns, each clan has its separate ward.^[77] The clans of the Osages are divided into war clans and peace clans; when they are out on the buffalo hunt, they camp on opposite sides of the tribal circle; and the peace clans are not allowed to take animal life of any kind; they must therefore live on vegetables unless they can obtain meat in exchange for vegetables from the war clans.^[78] Members of the same clan are buried together and apart from those of other clans; hence the remains of husband and wife, belonging as they do to separate clans, do not rest together.^[79] It is remarkable that among the Thlinkets the body must always be carried to the funeral pyre and burned by men of another totem,^[80] and the presents distributed on these occasions by the representatives of the deceased must always be made to men of a different clan.^[81]

Phratric and subphratric totems.Here we must revert to the religious side of totemism, in order to consider some facts which have emerged from the study of its social aspect. We have seen that some phratries, both in America and Australia, bear the names of animals;^[82] and in the case of the Thlinkets and Mohegans we have seen reason to believe that the animals which give their names to the phratries were once clan totems. The same seems to hold of the names of the Australian phratries, Eaglehawk, Crow, and Seal, or at least of Eaglehawk and Crow, for these are clan totems in other tribes, and are, besides, important figures in Australian mythology. Indeed, there appears to be direct evidence that both the phratries and subphratries actually retain, at least in some tribes, their totems. Thus the Port Mackay tribe in Queensland is divided into two phratries, Yungaru and Wutaru, with subphratries Gurgela, Burbia, Wungo, and Kubera; and the Yungaru phratry has for its totem the alligator, and Wutaru the kangaroo,^[83] while the sub phratries have for their totems the emu (or the carpet snake), iguana, opossum, and kangaroo (or scrub turkey).^[84] TOTEMISM 475 As the subphratries of this tribe are said to be equivalent to the subphratries of the Kamilaroi, it seems to follow that the subphratries of the Kamilaroi (Muri, Kubi, Ipai, and Kumbo) have or once had totems also. Hence it ap pears that in tribes organized in phratries, subphratries, and clans each man has three totems his phratry totem, his subphratry totem, and his clan totem. If we add a sex totem and an individual totem, each man in the typical Australian tribe has five distinct kinds of totems. What degree of allegiance he owes to his subphratry totem and phratry totem respectively we are not told ; indeed, the very existence of such totems, as distinct from clan totems, appears to have been generally overlooked. But we may suppose that the totem bond diminishes in strength in proportion to its extension; that therefore the clan totem is the primary tie, of which the subphratry and phratry totems are successively weakened repetitions. In these totems superposed on totems may perhaps be discerned a rudimentary classification of natural objects under heads which bear a certain resemblance to genera, species, &c. This classification is by some Australian tribes extended so as to include the whole of nature. Thus the Port Mackay tribe in Queensland (see above) divides all nature between the phratries ; the wind belongs to one phratry and the rain to another ; the sun is Wutaru and the moon is Yungaru ; the stars, trees, and plants are also divided between the phratries. 1 As the totem of Wutaru is kangaroo and of Yungaru alligator, this is equivalent to making the sun a kangaroo and the moon an alligator. The Mount Gambier tribe in South Australia is divided into two phratries (Kumi and Kroki), which again are subdivided into totem clans. Everything in nature belongs to a totem clan, thus : 2 Phratries. Totem Clans. Including Kumi. J Kroki. I 1. JIula= Fish-Hawk. 2. Parangal= Pelican. 3. Wa = Crow. 4. WIla= Black Cockatoo. 5. Karato=A harmless Snake. 1. Werio= Tea-Tree. 2. Murna=An edible Root. 3. Kara al = Black crest less Cock atoo. Smoke, honeysuckle, trees, <fcc. /Dogs, blackwood trees, fire, frost 1 (fern.) / Rain, thunder, lightning, winter,

hail, clouds, &c. 

Stars, moon, <fec. /Fish, stringy-bark trees, seals,

eels, <fcc. 

Ducks, wallabies, owls, cray-flsh, &c. {Bustards, quails, dolvich (a small kangaroo). Kangaroo, she-oak trees, summer, sun, autumn (fern.), wind (fern.). With reference to this classification Mr D. S. Stewart, the authority for it, says, " I have tried in vain to find some reason for the arrangement. I asked, To what division does a bullock belong ? After a pause came the answer, It eats grass : it is Boortwerio. I then said, A cray-fish does not eat grass; why is it Boortwerio ? Then came the standing reason for all puzzling questions : That is what our fathers said it was. " 3 The natural objects thus classed under and sharing the respect due to the totem may be conveniently called, as Mr Howitt proposes, 4 subtotems. Again, the "Wotjoballuk tribe in north-western Victoria has a system of subtotems, thus : B Phratries. Totem Clans. Subtotems. Krokitch. -j Gamutch. < 1. Hot Wind. 2. White crestless Cockatoo. 3. Belonging to the Sun. 4. Deaf Adder. 5. Black Cockatoo. 6. Pelican. Each totem has subordinate to it a number of objects, animal or vegetable, e.g., kangaroo, red gum-tree, <fec. Do. Of the subtotems in this tribe Mr Howitt says, " They appear to me to be totems in a state of development. Hot wind has at least five of them, white cockatoo has seventeen, and so on for the others. That these subtotems are now in process of gaining a sort of inde pendence may be shown by the following instance : a man who is Krokitch-Wartwut (hot wind) claimed to own all the five subtotems of hot wind (three snakes and two birds), yet of these there was one which he specially claimed as belonging to him, namely, Moiwuk (carpet-snake). Thus his totem, hot wind, seems to have been in process of subdivision into minor totems, and this man s 1 Brough Smyth, i. 91; Fison and Howitt, 168; cf. J. A. I., xiii. 300. 2 Fison and Howitt, loc. cit. 8 Fison and Howitt, 169. 4 In Smithson. Rep. for 1883, p. 818. 5 Ibid. division might have become hot wind carpet-snake had not civilization rudely stopped the process by almost extinguishing the tribe." Geographical Diffusion of Totemism. In Australia Diffusion totemism is almost universal. 6 In North America it may of totem- be roughly said to prevail, or have prevailed, among all the mn * tribes east of the Rocky Mountains, 7 and among all the Indian (but not the Eskimo) tribes on the north-west coast as far south as the United States frontier. On the other hand, highly competent authorities have failed to find it among the tribes of western Washington, north-western Oregon, and California. 8 In Panama it exists apparently among the Guaymies : each tribe, family, and individual has a guardian animal, the most prevalent being a kind of parrot. 9 In South America totemism is found among the Goajiros on the borders of Colombia and Venezuela, 10 the Arawaks in Guiana, 11 the Bosch negroes also in Guiana, 12 and the Patagonians. 13 Finding it at such distant points of the continent, we should expect it to be widely prevalent ; but, with our meagre knowledge of the South American Indians, this is merely conjecture. The aborigines of Peru and the Salivas on the Orinoco believed in the descent of their tribes from animals, plants, and natural objects, such as the sun and earth ; u but this, though a presumption, is not a proof of totemism. In Africa totemism prevails in Senegambia, among the Bakalai on the equator, on the Gold Coast and in Ashantee, and among the Damaras and Bechuanas in southern Africa. 15 There are traces of totemism elsewhere in Africa. In east ern Africa the Gallas are divided into two exogamous sec tions, and have certain forbidden foods. 16 In Abyssinia certain districts or families will not eat of certain animals or parts of animals. 17 The territory of the Hovas in Mada gascar is divided and subdivided into districts, the names of the subdivisions referring " rather to clans and divi sions of people than to place." One of these names is " the powerful bird," i.e., either the eagle or the vulture. The same clan is found occupying separate districts. 18 One Madagascar tribe regard a species of lemur as " an embodi ment of the spirit of their ancestors, and therefore they look with horror upon killing them." Other Malagasy tribes and families refrain from eating pigs and goats; others will not eat certain vegetables nor even allow them to be carried into their houses. 19 The only occasion when the Sakalava tribe in Madagascar kill a bull is at the cir cumcision of a child, who is placed on the bull s back during the customary invocation. 20 In Bengal, as we have seen, there are numerous totem tribes among the non-Aryan races. In Siberia the Yakuts 6 Perhaps the only known exceptions are the Kurnai in eastern and the Gournditch-mora in western Victoria. For the latter see Fison and Howitt, p. 275. 7 Gatschet, Migration Legend of the Creek Indians, 153 ; H. Hale, The Iroquois Book of Rites, p. 51. 8 George Gibbs in Contrib. to N. American Ethnol., i. 184; S. Powers, Tribes of Calif., 5. 9 A. Pinart in Revue d Ethnographic, vi. p. 36. 10 Simons in Proc. R. Geog. Soc., Dec. 1885, pp. 786, 796. 11 Brett, Ind. Tribes of Guiana, 98 ; Im Thurn, Among the Indians of Guiana, 175 sq. 12 Crevaux, Voyages dans VAmZrique du Sud, p. 59. 13 Falkner, Descr. of Patagonia, 114. 14 Garcilasso de la Vega, Royal Commentaries of the Incas, pt. i. bk. i. chs. 9, 10, 11, 18 ; Gumilla, Hist, de VOrenoque, i. 175 sq. 13 Revue d Ethnologic, iii. 396 sq., v. 81; A. B. Ellis, The Tshi- speaking People of the Gold Coast, p. 204 sq. ; Bowdich, Mission to Ashantee, ed. 1873, p. 216; Du Chaillu, Equal. Afr., 308 sq. , Id., Journey to AsJiango Land, 427, 429 ; 0. J. Anderson, Lake Ngami, 221 sq. ; Livingstone, Travels in S. Africa, 13 ; Casalis, The Basutos, 211 ; J. Mackenzie, Ten Years North of the Orange River, 393 ; J. A. I., xvi. 83 sq. 16 Charles New, Life, Wanderings, &c., in Eastern Africa, 272, 274. 17 Mansfield Parkyns, Life in Abyssinia, 293 ; Tr. Ethnol. Soc., new series, vi. 292. 18 Ellis, Hist, of Madagascar, i. 87. 19 Folk-Lore Record, ii. 22, 30. w Ibid., iv. 45. are divided into totem clans; the clansmen will not kill their totems (the swan, goose, raven, &C.);^[85] and the clans are exogamous.^[86] The Altaians, also in Siberia, are divided into twenty-four clans, which, though interfused with each other, retain strongly the clan feeling; the clans are exo gamous; each has its own patron divinity and religious ceremonies; and the only two names of clans of these and kindred tribes of which the meanings are given are names of animals.^[87] Totemism exists among the mountaineers of Formosa, ^[88] and there are traces of it in China.^[89] In Polynesia it existed, as we have seen, in Samoa. In Melanesia it appears in Fiji,^[90] the New Hebrides,^[91] and the Solomon Islands.^[92]. Amongst the Dyaks there are traces of totemism in the prohibition of the flesh of certain animals to certain tribes, respect for certain plants, &c.^[93] It exists in the islands of Ambon, Uliase, Leti, Moa, Lakor, Keisar (Makisar), Wetar, and the Aaru and Babar archipelagos.^[94] In the Philippine Islands there are traces of it in the reverence for certain animals, the belief that the souls of ancestors dwell in trees, &c.^[95]

With regard to ancient nations, totemism may be re garded as certain for the Egyptians, and highly probable for the Semites,^[96] Greeks, and Latins. If proved for one Aryan people, it might be regarded as proved for all; since totemism could scarcely have been developed by any one Aryan branch after the dispersion, and there is no evi dence or probability that it ever was borrowed. Prof. Sayce finds totemism among the ancient Babylonians, but his evidence is not conclusive.^[97]

No satisfactory explanation of the origin of totemism has yet been given. Mr Herbert Spencer finds the origin of totemism in a "misinterpretation of nicknames": savages first named themselves after natural objects, and then, confusing these objects with their ancestors of the same names, reverenced them as they already reverenced their ancestors.^[98] But this view attributes to verbal mis understandings far more influence than, in spite of the so-called comparative mythology, they ever seem to have exercised.

Literature.—Apart from the original authorities, the literature on totemism is very scanty. The importance of totemism for the early history of society was first recognized by Mr J. F. M'Lennan in papers published in the Fortnightly Review (Oct. and Nov. 1869, Feb. 1870). The subject has since been treated of by E. B. Tylor, Early History of Mankind, p. 284 sq.; Sir John Lubbock, Origin of Civilization, 260 sq.; A. Lang, Custom and Myth, p. 260, &c.; Id., Myth, Ritual, and Religion, i. p. 58 sq., &c.; E. Clodd, Myths and Dreams, p. 99 sq.; W. Robertson Smith, Kinship and Marriage in Early Arabia. See also Sacrifice, vol. xxi. p. 135. For fuller details, see J. G. Frazer, Totemism (Edinburgh, 1887). (J. G. FR.)

---

1. Voyages and Travels of an Indian Interpreter, p. 86, 1791.
2. Second Annual Report of the Bureau of Ethnology, Washington, 1883, p. 77.
3. Catlin, North American Indians, ii. p. 128.
4. Schoolcraft, The American Indians, p. 95 sq.; Lewis and Clarke, Travels to the Source of the Missouri River, London, 1815, i. p. 12.
5. Sir George Grey, Vocabulary of Dialects of S. W. Australia.
6. Revue d'Ethnographie, iii. p. 396, v. p. 81.
7. Bourke, Snake Dance of the Moquis of Arizona, p. 177.
8. Du Chaillu, Explorations in Equatorial Africa, p. 308.
9. Stewart in Fison and Howitt, Kamilaroi and Kurnei, p. 169.
10. Jour. Anthrop. Inst., xiii. p. 300.
11. Grey, Journals of Two Expeditions in North-West and Western Australia, ii. p. 228.
12. R. C. Mayne, British Columbia, p. 258.
13. C. J. Anderson, Lake Ngami, 222 sq.
14. Rev. d'Ethn., iii. 396.
15. Dalton in Trans. Ethnolog. Soc., new series, vi. p. 36; Id., Ethnol. of Bengal, pp. 189, 254; As. Quart. Rev., July 1886, p. 76.
16. As. Quart. Rev., July 1886, p. 79.
17. Grey, Journals, ii. 228 sq.
18. Dalton, Ethn. of Bengal, 254; Id., Trans. Ethnol. Soc., vi. 36.
19. E. James, Expedition from Pittsburgh to the Rocky Mountains, ii. p. 47; Third Rep. Bur. Ethnol., p. 225.
20. James, loc. cit.; Third Rep., 245.
21. As. Quart. Rev., July 1886.
22. Casalis, The Basutos, 211.
23. Verhandl. der Berliner Gesell. f. Anthropologie, 1882, p. (62).
24. Turner, Samoa, p. 71.
25. Native Tribes of S. Australia, p. 63.
26. Turner, op. cit., p. 64.
27. Raffles, Hist, of Java, i. p. 328, ed. 1817.
28. Turner, op. cit., p. 21, cf. 26, 60 sq.
29. Charles New, Life, Wanderings, &c., in Eastern Africa, p. 122.
30. Robertson Smith, Kinship and Marriage in Early Arabia, p. 195.
31. Boscana, in Alfred Robinson's Life in California, p. 291 sq.; Bancroft, Native Races of the Pacific States, iii. p. 168.
32. Brinton, The Lenape and their Legends, p. 39; Morgan, Anc. Soc., p. 171; Heckewelder, p. 247.
33. See Acad., 27th Sept. 1884, p. 203.
34. Du Chaillu, Equat. Afr., p. 309.
35. Third Rep., 225.
36. Ibid., 231.
37. James, Exped. to the Rocky Mountains, ii. p. 50.
38. Plutarch, De Superst., 10; Selden, De Dis Syris, p. 269 sq., Leipsic, 1668.
39. Turner, Samoa, p. 17 sq.
40. Ibid., p. 50.
41. Turner, Samoa, p. 31 sq.
42. J. A. I., xiii. p. 192.
43. T. L. Mitchell, Three Expeditions into the Interior of Eastern Australia, ii. p. 341.
44. Morgan, op. cit., p. 73.
45. Morgan, p. 156.
46. First Rep., p. 59.
47. Collections of the Minnesota Historical Society, v. p. 49.
48. H. Hale, The Iroquois Book of Rites, p. 53 sq.
49. Geol. Surv. of Canada, Rep. for 1878-79, p. 134B.
50. Third Rep., p. 235; American Naturalist, xviii. p. 114.
51. Fison and Howitt, p. 285.
52. J. A. I., xii. 500.
53. Corresponding female forms are made by adding tha to these male names: Muri—Mathu (for Muritha), Kubi—Kubitha, &c.
54. Fison and Howitt, p. 53 sq.; cf. Brough Smyth, i. p. 91.
55. J. A. I., xiii. p. 300.
56. For a few particulars see Fison and Howitt, 38, 40; Brough Smyth, ii. 288; J. A. I., xiii. 304, 306, 346, xiv. 348 sq., 351.
57. First Rep., p. 63. Between North-American tribes "there were no intermarriages, no social intercourse, no intermingling of any kind, except that of mortal strife " (Dodge, Our Wild Indians, p. 45).
58. Cf. First Rep., loc. cit.; As. Quart. Rev., July 1886, p. 89 sq.
59. Anderson, Lake Ngami, p. 221.
60. Casalis, The Basutos, p. 179 sq.
61. Du Chaillu, Journey to Ashango Land, 429; Id., Equat. Afr., 308 sq.
62. As. Quart. Rev., July 1886, p. 94.
63. As. Quart. Rev., July 1886, p. 94.
64. Dalton, Ethn. of Beng., p. 56 sq.; W. W. Hunter, Statistical Account of Assam, ii. p. 217 sq.
65. Dalton, op. cit., 60, 63; Hunter, op. cit., ii. 154 sq.
66. Fison and Howitt, p. 37 sq.; J. A. I., xiii. 335, 341, 344.
67. J. A. I., xiii. 336, 341.
68. Ibid., xii 504
69. Ibid., xii. 500.
70. Letter of Mr S. Gason to the present writer.
71. J. A. I., xii. 509.
72. Geol. Surv. of Canada, Rep. for 1878-79, p. 134B.
73. Morgan, A. S., p. 172 sq.
74. Gill, Myths and Songs of the South Pacific, p. 36.
75. Turner, Samoa, p. 78 sq. The child might thus be transferred to a clan which was that neither of his father nor of his mother.
76. First Rep., 64; Third Rep., 219; Amer. Naturalist, xviii. 113.
77. Gatschet, Migration Legend of the Creek Indians, 154; Bourke, Snake Dance, 229; Acad., 27th Sept. 1884, p. 203.
78. Rev. J. Owen Dorsey in American Naturalist, xviii. p. 113.
79. Adair, Hist. Amer. Ind., 183 sq.; Morgan, A. S., 83 sq.; Brinton, The Lenape and their Legends, 54; Id., Myths of the New World, 87 n; A. Hodgson, Letters from North America, i. p. 259; Dalton, Eth. of Beng., 56; cf. Robertson Smith, Kinship and Marriage in Early Arabia, 315 sq.
80. Holmberg, op. cit., 324.
81. Krause, Die Tlinkit-Indianer, 223.
82. As among the Chickasas, Thlinkets, and Mohegans in America, and the Turra, Ngarego, and Theddora tribes in Australia. The subphratries of the Kiabara also bear animal names.
83. Fison and Howitt, 38 sq., 40.
84. Fison and Howitt, p. 41. The totems of the phratries and subphratries are given by different authorities, who write the native names of the subphratries differently. But they seem to be speaking of the same tribe; at least Mr Fison understands them so.
85. Strahlenberg, Description of the North and Eastern Parts of Europe and Asia, London, 1738, p. 383.
86. Middendorf, Siber. Reise, p. 72, quoted by Lubbock, Origin of Civilization, p. 135. The present writer has been unable to find the passage of Middendorf referred to.
87. W. Radloff, Aus Siberien, i. 216, 258. The Ostiaks, also in Siberia, are divided into exogamous clans, and they reverence the bear (Castren, Vorlesungen über die Altaischen Völker, 107, 115, 117). This, however, by no means amounts to a proof of totemism.
88. Verhandl. d. Berl. Gesell. Anthropologie, &c., 1882, p. (62).
89. Morgan, A. S., p. 364 sq. One of the aboriginal tribes of China worships the image of a dog (Gray, China, ii. 306).
90. Williams, Fiji and the Fijians, ed. I860, i. 219 sq.
91. Turner, Samoa, 334.
92. Fison and Howitt, p. 37 n
93. Low, Sarawak, 265 sq., 272-274, 306; St John, Life in the Forests of the Far East, i. 186 sq., 203; cf. Wilken in Ind. Gids, June 1884, p. 988 sq.; Ausland, 16th June 1884, p. 470.
94. Riedel, De sluik- en kroesharige rassen tusschen Papua en Selebes, pp. 32, 61, 253, 334, 341, 376 sq., 414, 432.
95. Blumentritt, Der Ahnencultus und die religiösen Anschauungen der Malaien des Philippines-Archipel, 159 sq.
96. See W. R. Smith, Kinship and Marriage in Early Arabia.
97. A. H. Sayce, The Religion of the Ancient Babylonians, p. 279 sq.
98. Spencer, Principles of Sociology, i. 367.