when a trap is laid, is connected with the idea of tigers being animated by the souls of departed men.1 In other districts of the world, one of the most important cases connected with these is the worship paid by the North American Indian to his medicine-animal, of which he kills one specimen to preserve its skin, which thenceforth receives adoration and grants protection as a fetish.2 In South Africa, as has been already mentioned, the Zulus hold that divine ancestral shades are embodied in certain tame and harmless snakes, whom their human kinsfolk receive with kindly respect and propitiate with food.3 In West Africa, monkeys near a grave-yard are supposed to be animated by the spirits of the dead, and the general theory of sacred and worshipped crocodiles, snakes, birds, bats, elephants, hyænas, leopards, &c., is divided between the two great departments of the fetish-theory, in some cases the creature being the actual embodiment or personation of the spirit, and in other cases sacred to it or under its protection. Hardly any region of the world displays so perfectly as this the worship of serpents as fetish-animals endowed with high spiritual qualities, to kill one of whom would be an offence unpardonable. For a single description of negro ophiolatry, may be cited Bosman's description from Whydah in the Bight of Benin; here the highest order of deities were a kind of snakes which swarm in the villages, reigned over by that huge chief monster, uppermost and greatest and as it were the grandfather of all, who dwelt in his snake-house beneath a lofty tree, and there received the royal offerings of meat and drink, cattle and money and stuffs. So heartfelt was the veneration of the snakes, that the Dutchmen made it a

4

1 Marsden, 'Sumatra,' p. 292.

2 Loskiel, 'Ind. of N. A.' part i. p. 40; Catlin, 'N. A. Ind.' vol. i. p. 36; Schoolcraft, 'Tribes,' part i. p. 34, part v. p. 652; Waitz, vol. iii. P. 190.

3 See ante, p. 8; Callaway, 'Rel. of Amazulu,' p. 196.

4 Steinhauser, 'Religion des Negers,' l.c. p. 133. J. L. Wilson, 'W. Afr.' pp. 210, 218. Schlegel, 'Ewe-Sprache,' p. xv.

means of clearing their warehouses of tiresome visitors; as Bosman says, 'If we are ever tired with the natives of this country, and would fain be rid of them, we need only speak ill of the snake, at which they immediately stop their ears and run out of doors.'1 Lastly, among the Tatar tribes of Siberia, Castrén finds the explanation of the veneration which the nomade pays to certain animals, in a distinct fetish-theory which he thus sums up: 'Can he also contrive to propitiate the snake, bear, wolf, swan, and various other birds of the air and beasts of the field, he has in them good protectors, for in them are hidden mighty spirits.'

2

In the lower levels of civilization the social institution known as Totemism is of frequent occurrence. Its anthropological importance was especially brought into notice by J. F. McLennan, whose views as to an early totem-period of society have much influenced opinion since his time.3 The totemic tribe is divided into clans, the members of each clan connecting themselves with, calling themselves by the name of, and even deriving their mythic pedigree from some animal, plant, or thing, but most often an animal; these totem-clans are exogamous, marriage not being permissible within the clan, while permissible or obligatory between clan and clan. Thus among the Ojibwa Indians of North America, the names of such clan-animals, Bear, Wolf, Tortoise, Deer, Rabbit, &c., served to designate the intermarrying clans into which the tribes were divided, Indians being actually spoken of as bears, wolves, &c., and the figures of these animals indicating their clans in the native picture-writing. The Ojibwa word for such a clan-name has passed into English in the form 'totem,' and thus has become an accepted term among anthropologists to denote

See

1 Bosman, 'Guinea,' letter 19; in Pinkerton, vol. xvi. p. 499. Burton, 'Dahome,' ch. iv., xvii. An account of the Vaudoux serpent-worship still carried on among the negroes of Hayti, in 'Lippincott's Magazine,' Philadelphia, March, 1870.

2 Castrén, 'Finn. Myth.' p. 196, see 228.

3 J. F. McLennan in ‘Fortnightly Review,' 1869–70; reprinted in 'Studies in Ancient History,' 2nd series, pp. 117, 491.

similar clan-names customary over the world, this system of dividing tribes being called Totemism. Unfortunately for the study of the subject, John Long, the trader interpreter who introduced the Ojibwa word totem into Europe in 1791, does not seem to have grasped its meaning in the native law of marriage and clanship, but to have confused the totem-animal of the clan with the patron or guardian animal of the individual hunter, his manitu or medicine.'1 Even when the North American totem-clans came to be better understood as social institutions regulating marriage, the notion of the guardian spirit still clung to them. Sir George Grey, who knew of the American totem-clans from the Archæologia Americana,' put on record in 1841 a list of exogamous classes in West Australia, and mentioned the opinion frequently given by the natives as to the origin of these class-names, that they were derived from some animal or vegetable being very common in the district which the family inhabited, so that the name of this animal or vegetable came to be applied to the family. This seems so far valuable evidence, but Grey was evidently led by John Long's mistaken statement, which he quotes, to fall himself into the same confusion between the tribal name and the patron animal or vegetable, the ‘kobong' of his natives, which he regarded as a tribal totem.2 In Mr. J. G. Frazer's valuable collection of information on totemism,3 the use of the self-contradictory term 'individual totem' has unfortunately tended to perpetuate this confusion. In the present state of the problem of totemism, it would be premature to discuss at length its development and purpose. Mention may however be made of observations which tend to place it on a new footing, as being distinctly related to the transmigration of souls. In Melanesia men 1 John Long, 'Voyages and Travels of an Indian Interpreter,' London, 1791, p. 86. See pp. 233, 411 of present volume.

2 Grey, 'Journals of Expeditions in N. W. & W. Australia,' vol. ii. pp. 225-9; 'Archæologia Americana,' vol. ii. p. 109.

3 J. G. Frazer, 'Totemism,' p. 53; 'Golden Bough,' 2nd ed. vol. iii. pp. 419, 423.

may say that after death they will reappear for instance as sharks or bananas, and the family will acknowledge the kinship by feeding the sharks and abstaining from the bananas. It is not unreasonable that Dr. Codrington should suggest such practices as throwing light on the origin of totemism.1 The late investigations of Spencer and Gillen, conducted with scrupulous care in an almost untouched district of Central Australia, show totemism in the Arunta tribe, not as the means of regulating the intermarriage of clans, but as based on a native theory of the ancestry of the race, as descended from the Alcheringa, quasi-human animal or vegetable ancestors, whose souls are still reborn in human form in successive generations.2 This careful and definite account may be the starting-point of a new study. Savages would be alive to the absurdity of naming clans after animals in order to indicate a prohibition of marryingin, opposed to the habit of the animals themselves. Indeed, it seems more likely that such animal-names may have commonly belonged to inbred clans, before the rule of exogamy was developed. At present the plainest fact as to Totemism is its historical position as shown by its immense geographical distribution. Its presence in North America and Australia has been noticed. It extends its organization through the forestregion of South America from Guyana to Patagonia. Northward of Australia it is to be traced among the more unchanged of the Malay populations, who underneath foreign influence still keep remains of a totemic system like that of the American tribes. Thence we follow the totem-clan into India, when it appears among non-Aryan hill-tribes such as the Oraons and Mundas, who have clans named after Eel, Hawk, Heron, and so on, and must not kill or eat these creatures. North of the Himalaya it appears among Mongoloid tribes in their native low cultured state, such as the Yakuts with their intermarrying totem-clans Swan, Raven,

1 Codrington, 'Melanesians, pp. 32-3, 170.

2

Spencer and Gillen, 'Native Tribes of Central Australia,' 1899, pp. 73,

and the like. In Africa totemism appears in the Bantu district up to the West Coast. For example, the Bechuana are divided into Bakuena, men of the crocodile; Batlapi, of the fish; Balaung, of the lion; Bamorara, of the wild vine. A man does not eat his tribe-animal, or clothe himself in its skin, and if he must kill it as hurtful, the lion for instance, he asks pardon of it, and purifies himself from the sacrilege. These few instances illustrate the generalization that totemism in its complete form belongs to the savage and early barbaric stages of culture, only partial remains or survivals of it having lasted into the civilized period. Though appearing in all other quarters of the globe, it is interesting to notice that there is no distinct case of totemism found or recorded in Europe.1 1

The three motives of animal-worship which have been described, viz., direct worship of the animal for itself, indirect worship of it as a fetish acted through by a deity, and veneration for it as a totem or representative of a tribeancestor, no doubt account in no small measure for the phenomena of Zoolatry among the lower races, due allowance being also made for the effects of myth and symbolism, of which we may gain frequent glimpses. Notwithstanding the obscurity and complexity of the subject, a survey of Animal-worship as a whole may yet justify an ethnographic view of its place in the history of civilization. If we turn from its appearances among the less cultured races to notice the shapes in which it has held its place among peoples advanced to the stage of national organization and stereotyped religion, we shall find a reasonable cause for its new position in the theory of development and survival, whereby ideas at first belonging to savage theology have in part continued to spread and solidify in their original manner, while in part they have been changed to accommodate them to more advanced ideas, or have been defended from the attacks of reason by being set up as sacred mysteries. Ancient Egypt

1 General references in J. F. McLennan, 'Studies in Ancient History;' J. G. Frazer, 'Totemism.'