1900.] Annual Address. 53 



supposes that by performing the most elaborate parody of the demeanour 

 of certun animals a man can really cause them to increase and multiply. 

 In India, on the other hand, our totemistic people have got rid of all 

 such antics, if indeed they ever practised them, and retain only the 

 unquestionably effective factor in the system, the rule that a man may 

 not marry a woman of his own totem. They have, it is true, also the 

 rule that people may not eat, injure or make use of their totems, but 

 this prohibition is relatively weak, and in some cases the totems are 

 articles such as rice and salt which the members of the totem-kin could 

 hardly do without. 



Given then a state of things such as this, that tribes which are in 

 no way moribund or degenerate, but on the contrary extremely full of 

 life, retain the effective part of an archaic usage along with traces of 

 its ineffective parts, may we not reasonably conclude that this effective 

 part, which has stood the wear and tear of ages and contributed to 

 the evolution of the tribe, furnishes the clue to the real origin of the 

 usage itself? Assume this to be so and totemism at once wheels into 

 line and takes the place, which it appears clearly to occupy in India, 

 of a form of exogamy. The particular form presents no great difficulty. 

 Primitive men are like children : the}' are constantly saying to them- 

 selves " Let's pretend," and a favourite and wide spread form of the 

 game is to pretend to be animals. Only they play it in earnest, and 

 very grim earnest it sometimes is, as one discovers when one has to 

 administer a district where people believe that men can transform 

 themselves into animals at will, or can be so transformed by the agency 

 of witchcraft. 



It will be asked, what then is the origin of exogamy ? Here again 

 I think the Indian evidence suggests an answer. Just as the special 

 phenomenon of totemism may be explained by reference to the general 

 law of exogamy, so exogniny itself may be traced to the still more 

 general law of natural selection. Nor need we strain the law. We 

 know that there is a tendency in individuals or groups of individuals to 

 vary tlieir habits; and that useful variations tend to be preserved and 

 ultimately transmitted. Now suppose that in a primitive community, 

 such as the Naga Khel or the Kandh Gochi, the men happened to vary 

 in the direction of taking their wives from some other community, and 

 that this infusion of fresh blood proved advantageous to the group. 

 The oiiginal instinct would then be stimulated by heredity, and the 

 element of sexual selection would in course of time come into play. For 

 an exogamous groups would have a larger choice of women than an 

 endogamous one, and would thus get finer women, who again, in the 

 course of the primitive struggle for wives, would be appropriated by the 



