482 Mr Punnett, On the Proportion 



Several explanations have been put forward to account for this 

 phenomenon. It has been suggested that the preponderance of 

 living males in polyandrous societies is due to a great excess of 

 cf births resulting from defective nutrition, and, according to 

 Westermarck *, "much stress must be laid on the fact that 

 polyandry prevails chiefly in poor countries." The evidence^ by 

 which he supports his contention is however not very convincing, 

 and I have recently given figures I which do not strengthen the 

 hypothesis that sex determination depends upon nutrition. More- 

 over on this hypothesis it is not easy to account either for the 

 great fluctuations in the proportions of Js to $s found in 

 different generations (see below p. 483, Table II.), or for the fact 

 that the proportion of <?s among the Todas is enormously greater 

 than among the lowest classes of civilised peoples where the nutri- 

 tion is probably equally deficient. A second cause which has been 

 suggested to account for the excess of ,/s is in-breeding, and 

 Westermarck § quotes Dusing in favour of this hypothesis, and 

 supports it by cases where mixture of race tends to produce an 

 excess of $ s. Here again the information given is on the whole 

 somewhat too vague for a definite conclusion. Moreover it is not 

 supported by such detailed evidence as I have been able to collect. 

 On the islands of Mabuiag and Badu|| in Torres Straits, where 

 there exists a considerable amount of in-breeding, the ,/s pre- 

 ponderate only to a small extent, a state of things which holds 

 good for Murray Island also. The balance of evidence is on the 

 whole against either of the above hypotheses affording an adequate 

 explanation of the excessive preponderance of the £ sex among 

 the Todas. 



There yet remains for consideration a third factor, viz., female 

 infanticide. All who have studied the Todas are agreed upon the 

 frequency of this practice in earlier times. MarshalllF, writing in 

 1872, refers to the large amount of female infanticide in former 

 years but expresses his conviction that the practice had by that 

 time died out. Marshall's evidence is that of native assurance 

 only. Dr Rivers, who received the same assurance, is disinclined 

 to place much confidence in native veracity with reference to this 

 point, and in view of the lack of encouragement which the prac- 

 tice receives from the Indian Government this is not altogether 



* The History of Human Marriage, 3rd edition, 1901, p. 474. 



t Loc. cit. pp. 470 seq. X Proc. Gavib. Phil. Soc. Vol. xn. p. 262 seq. 



§ The History of Human Marriage, 3rd edition, p. 476. 



|| Proc. Gamb. Phil. Soc. Vol. xn. p. 271. Cf. also Rep. Cambridge Anthropo- 

 logical Expedition to Torres Straits, Vol. v. p. 233. Keference may be made here 

 to a recent paper by 0. Schultze in the Archiv f. Microscop. Anat. 1903. Working 

 with white mice Schultze found that neither starvation nor in-breeding affected 

 the proportion of the sexes. 



IF A Phrenologist among the Todas, 1873, p. 198. 



