54 Annual Address. [Feb. 



strongest and most warlike men. Exogamous groups thus strengthened 

 would tend, as time went on, to ' eat up,' as the Zulus used to say, their 

 endogamous neighbours or at any rate to deprive them of the pick of 

 their marriageable girls ; and the custom of exogamy would spread, 

 partly by imitation, and partly by the extinction of the groups which 

 did not practise it. 



The fact that we cannot say how people came to vary in this 

 particular fashion is not necessarily fatal to the hypothesis put forward. 

 In the case of animals other than man, we do not call in question the 

 doctrine of natural selection because we cannot divine the precise cause 

 which gave rise to some beneficial variation. It is enough that variations 

 do occur, and that the beneficial ones tend to be transmitted. If howevei- 

 an attempt must be made to pierce the veil which shuts off from our 

 view the ages of pre-historic evolution, it does not seem unreasonable to 

 suppose that here and there some half-accidental circumstance, such as 

 the transmission of a physical defect or an hereditary disease, may have 

 given primitive man a sort of warning and thus have induced the 

 particular kind of variation which his circumstances required. Con- 

 quest again may have produced the same effect by bringing about a 

 beneficial mixture of stocks, though it is a little difficult to see, as Mr. 

 Lang pointed out long ago, why the possession of foreign women should 

 have disinclined people to marry the women of their own group. At 

 the same time it is conceivable that the impulse may have been set 

 goitig by some tribe from which all its marriageable women had been 

 raided, and which was thus driven by necessity to start raiding on its 

 own account. I have elsewhere given instances, drawn from the 

 Kandhs and Nagas, which lend themselves to this view ; but I am not 

 sure that we need travel beyond the tendency to accidental variation 

 which appears in all living organisms and may be assumed to have 

 shaped the evolution of primitive man. 



I may now draw together the loose ends o*f thoughts which the 

 proposal of the British Association have suggested. Speaking as 

 President of this Society, and leaving out of consideration for the 

 moment those questions of ways and means which have to be taken 

 into ace <unt when one quits the domain of the ideal, 1 would formulate 

 the requirements of ethnological research in India at the present time 

 as follows : — 



First, we want the ethnographic survey, which has been carried 

 out in Bengal and the N.-W.P., to be extended on similar lines to the 

 rest of India. This ought be done within a reasonable time and cannot 

 be postponed indefinitely. 



Secondly, we want a record of the physical characters of typical 



