the dog would be in hunting, and how pleasant, I suppose, the horse would be to ride; and then, without more ado, deliberately set to work to domesticate the animals. The early history of man's first faithful comrade, the dog, escapes our ken; but not so with the horse. It is as certain as things of this kind can be, that the primitive Indo-European reared droves of tame or half-tame horses for generations, if not centuries, before it ever occurred to him to ride or drive them, and this fact, inexplicable on the ordinary theory, confirms our hypothesis. To sum up, the cause which our hypothesis postulates, namely, that man spared and protected certain animals without any thought of making economic use of them, is a vera causa, for men do so treat their totem animals. That animals worshipped as totems do become tame, is also matter of fact. In Shark's Bay "the natives there never kill them [kites], and they are so tame that they will perch on the shoulders of the women and eat from their hands." 2 Further, our hypothesis accounts for all the facts, especially for such survivals as the lingering reluctance of civilised man to eat the flesh of certain animals. It also accounts for savages making pets. It is the tameness of the totem animal which suggests the idea of taming other creatures. Again, it alone supplies a motive strong enough to restrain the savage from recklessly devouring or destroying (instead of breeding from) the animals he caught or tamed. Finally, it admits of verification; for if it can be shown that not merely is the treatment of totem animals such as would naturally result in the taming of those that were domesticable, but that some domestic animals were actually totems, all the verification that can be required will be forthcoming. This will be seen to be the case with cattle in Egypt, and probably elsewhere also.

It seems, then, if the above argument commends itself to the reader, that totemism, and totemism alone, could have led to that "substitution of an artificial for a natural basis of subsistence" which consisted in the domestication of plants and animals, and which constituted the advance from savagery

1 Schrader, Prehistoric Antiquities, 263; and Hehn, Kulturpflanzen und Hausthiere, 19 ff.

2 Woodfield, ap. Galton, op. cit. 251,

to civilisation. But totemism did not universally lead to civilisation, or invariably develop into a higher form of religion. On the contrary, the civilised and civilising peoples are in the minority, and totemism still exists.

Now, if we consider the geographical distribution of totemism, we find that the two countries in which it is (or was at the time of the discovery of those countries) most marked are Australia and North America; while the peoples in which its traces are hardest to find are the Semitic and the Indo-European. If, again, we consider the geographical distribution of those species of animals which are capable of domestication and on the domestication of which the possibility of civilisation depended, we shall find that "the greatest number belonged to the Old World, those of America were fewer, and Australia had none at all "1; indeed, of the three species occurring in America (reindeer, llama, and paco), none come into account in this argument, for they are outside the totem-area of North America. It will scarcely be considered a merely fortuitous coincidencehowever we may explain it-that the two areas in which totemism lasted longest and flourished most are precisely those in which there are no domesticable animals. Nor is it a merely accidental occurrence that the peoples who have most. completely thrown off totemism, are precisely those which have by the domestication of plants and animals attained to civilisation. The inference is that the domestication to which totemism inevitably leads (when there are any animals capable of domestication) is fatal to totemism.

The fundamental principle of totemism is the alliance of a clan with an animal species, and when the clan ceases to exist as a social organisation the alliance is dissolved also. But with the transition from a nomad to a settled form of life, which the domestication of plants and animals entails, the tie of blood-relationship, indispensable to the existence of a wandering tribe, is no longer necessary to the existence of the community: local association and the bond of neighbourhood take its place, for the restriction of civic and political rights to the actual descendants of the original clan is inconsistent with the expansion of the community. By

1 Payne, 283,

this expansion of society beyond the narrow bounds of bloodrelationship, the germ of higher religious belief which totemism envelops is enabled also to burst its sheath, and man's conception of the deity sloughs off the totem-god. But though totemism perished in the very process of producing the advance from savagery to civilisation, still even in the civilisation of the Old World survivals of the system may be traced.

"For the Egyptians totemism may be regarded as certain." Egypt was divided into nomes or districts, in each of which a different animal was revered by the inhabitants. It was not an individual animal, but the whole species which was thus reverenced, and it was by all the inhabitants of the nome that it was revered. The lives of such animals were sacred, each in its own nome, and their flesh might not be consumed as food by the inhabitants of that nome. The god of the district manifested himself in the species sacred to that district. But this is not a survival of totemism. It is totemism, the thing itself.

No one, however, alleges that the religion of Egypt never got beyond totemism. On the contrary, we can see side by side with it in Egypt many of the stages and processes by which religion gradually divested itself of this its first protecting envelope, just as we may see sedimentary rocks by the side of the igneous rocks from which they are derived. Indeed, even in the lowest stratum of Egyptian totemism we may detect signs if not evidence of the disintegrating process: the bond of kinship, the tie of blood is relaxed. It is to be presumed that the inhabitants of a nome did not for ever continue to be blood-relations of one another, as they probably were when first they settled in the district; and the belief that the sacred species of animal was one blood and one flesh with the human tribe also faded. But though the blood-tie which held the human clansmen together, and which also bound the human clansmen to the animal, was relaxed and faded away from memory, the effects which it produced continued to exist. Thus, the sacred animal, whether it was still believed to be a blood-relation or not, received the same obsequies and was mummified in the

1 Frazer, 94.

same way as man; and the killing by one nome of an animal sacred to another was avenged in effect, if not consciously, in the spirit with which the blood-feud was exacted on behalf of a slaughtered kinsman.

Another and a further stage of development is reached, when one particular specimen of the species is selected as being the one which the deity has chosen to abide in, as, for instance, the calf marked by twenty-nine particular signs which showed that the Calf-god Apis was present in him. On the one hand, the concentration of veneration on an individual would tend to withdraw sanctity from the rest of the species, and the result might easily be a final separation of the animal-god from the animal species. On the other hand, that in Egypt at anyrate the worship of an individual animal, such as the Apis-calf, is the outcome of totemism, is plain from two things: first, the rest of the species did continue to be sacred-eating cow's flesh was as abhorrent to the Egyptians as cannibalism—and, next, "when the sacred animal died, the god as such did not die with him, but at once became incorporated in another animal resembling the first," evidently, as in Samoa, when an owl died, "this was not the death of the god; he was supposed to be yet alive and incarnate in all the owls in existence." 2

That, in spite of the ties which bound him to the rest of his species, the animal-god did shake off his humbler relations, and came to be worshipped in his higher aspect exclusively, is certain; and the process was facilitated by the dissolution of the bonds which tied down his worship to one particular nome. Apis, e.g., came to be worshipped all over Egypt. But the fact that his cult was originally local, not universal, is shown by the circumstance that his calf, wherever in Egypt it appeared, was taken to Memphis and kept there. Thus not only was the individual animal exalted above the rest of his species, but the god that dwelt in him was far removed from all his worshippers, except those who dwelt in the immediate neighbourhood of his animal manifestation. Thus he gained in magnificence both ways, and in both ways the associations which bound him to his animal form and 1 Wiedemann, Die Religion der alten Aegypter, 96.

2 Turner, Samoa, 21, sec above, p. 101.

origin were weakened.

The universalising of a cult is due to political causes: the political ascendancy of the nome from which the reigning dynasty derived would be marked by the extension of the local worship. The synoikismos which makes a nation also makes a pantheon.

But these causes are external, social, and political, not religious. They may and did loosen bonds which checked the progress of religion, but they were not themselves the force which was struggling to get free. But it is to the action of that force that we must attribute the dissociation of the god from his animal form, and his gradual appearance in human shape, which took place in Egypt. Here, however, our immediate concern is not to explain how this force operated, but to point out that the totemism which, as we have seen, demonstrably existed in Egypt, along with other higher elements of religion, did eventually become refined into a pantheon of anthropomorphic gods. "It is remarkable," says Dr. Wiedemann,1 “in view of the important part which the sacred animal plays in cultus, how relatively seldom it is portrayed. For a thousand representations of the gods, scarcely one will be found of an animal. The god appears either in human form, or as a man with the head of the animal sacred to him." Now, whether the mischbild of an animal-headed man was intended to intimate the idea that the god was of the same flesh with both his human kin and his animal kind, or is due to purely graphic considerations, as Dr. Wiedemann, who does not believe in totemism, is inclined to think, the fact remains that in the nome where a certain species was sacred the god is represented as a man with the head of that animal. In Mendes, e.g., the goat was sacred and the god goat-headed. And as for the great gods universally worshipped in the Egyptian religion, as Mr. Lang says, "it is always in a town where a certain animal is locally revered that the human-shaped god wearing the head of the same animal finds the centre and chief holy place of his worship." 2 The last stage is reached when the god casts aside his animal garb altogether, and the animal is thought and spoken of as being sacred to him, but has no other or more intimate relation with him.

1 Op. cit. p. 97.

2 Lang, Myth, Ritual, and Religion, ii. 104.