484 Mr Punnett, On the Proportion 



present day*. To return however to the main features of Table II., 

 since neither the country nor the marriage customs of the Todas 

 have changed appreciably during the last half century, we can 

 explain these figures only by supposing that there has been a 

 steady falling off in the practice of $ infanticide during recent 

 years. Even in the last generation (D) the preponderance of Js 

 is very considerable, and lends support to the view that the 

 practice is still prevalent, though not nearly to the same extent 

 as in former years f. 



The preservation of a larger proportion of $ infants should 

 naturally lead to a larger family, and it is somewhat unexpected 

 to find so marked a decrease in fertility of late years, especially 

 when the masking effect of $ infanticide in earlier generations is 

 taken into account. This effect I have endeavoured to eliminate 

 by taking the number of Js as a basis and estimating the size of 

 the family in each case on the supposition that there is numerical 

 equality among the sexes at birth. The figures so obtained are 

 enclosed in square brackets in Table II. and placed beside the rest 

 for purposes of comparison. They serve to emphasize the very 

 marked decrease in fertility which has occurred in recent years. 



Before proceeding to discuss the cause of this progressive 

 sterility, we may pause for a moment to consider some of the 

 statistics collected by Marshall. On page 197 of his book on 

 the Todas this author has given a table showing the number of 

 </■ and $ offspring belonging to females of different ages at the 



* In connection with this argument my friend Dr C. S. Myers has made the very 

 pertinent criticism that in shortening the period taken by 10 years I may be actually 

 selecting for sterility. The two following considerations however seem to me to tell 

 against this possibility : (a) When group C of the Tatharol tribe is treated in the 

 same way, and those families selected in which the age of the eldest member is not 

 greater than 60 and of the youngest not less than 40, the average size of the family 

 (out of a total of 32) works out to 3-4, and on the <? basis (see above) to 4"0. Making 

 allowance for irregularities due to the small number of families these figures are 

 very near to the 3*3 and 3-7 of Table II. There is certainly no evidence of the 

 selection of sterile families by this method. (/3) Though it is not possible to give 

 actual figures there seems little doubt but that the fertility span is considerably less 

 than 20 years. It is therefore unlikely that reduction of the period considered from 

 30 to 20 years implies any selection for sterility. 



Here it may be mentioned that the figures given in this paper for the average 

 size of the Toda family and for the fertile period do not agree with those given by 

 Marshall. This author reckons the average size of the family (excluding sterile 

 marriages) at 6*7, and the fertile period of the women at 19 '6 years (p. 122). These 

 figures are however based on the progeny of 17 women only (Table VII. p. 120), and 

 doubtless they present an inadequate picture of the real state of things. This is at 

 once apparent in the results which Marshall obtained when he attempted to forecast 

 the growth of the Toda population from his data. Taking the number of the Todas 

 as 713 in 1870 he estimated that it would in 1902 become 2852 (p. 107), an estimate 

 more than thrice what Dr Kivers actually found in that year. There can be little 

 doubt therefore that the data given in this paper are to be preferred to those 

 of Marshall. 



f The possibility of j infanticide is of course conceivable, but as I know of 

 nothing to support it, there is not much point in dwelling upon it. 



