488 Mr Punnett, On the Proportion of the Sexes, etc. 



actual source of error. On drawing up a list of the population 

 from my genealogical tables, I find that there are fewer children 

 between birth and five years of age than between five and ten, or 

 between ten and fifteen, years of age. Further, my total popu- 

 lation as calculated from the tables falls short of the number 

 obtained in the official census of 1901. My record of adults and 

 married children is almost certainly complete, and the discrepancy 

 between the official record and my own is probably due to de- 

 ficiencies in my record of young children. 



This error would only affect the latest of Mr Punnett's groups, 

 and he has for other reasons given figures in his paper which 

 enable different generations to be compared when children under 

 ten years are omitted. In these figures this source of error will 

 have been excluded. If the error is present in the genealogies, 

 its correction would accentuate the tendency to increase of popu- 

 lation which Mr Punnett believes to exist. If the conditions are 

 as Mr Punnett believes, the Todas have gone through a period of 

 decreasing fertility from which they are now emerging and this 

 would be in agreement with what seems to be happening else- 

 where among communities of low culture. It seems probable that 

 on the introduction of the various influences comprised under the 

 term " civilization " population falls, to rise again when the com- 

 munity becomes adapted to the new conditions. 



Records of the Toda population, official or semi-official, have 

 been taken since the year 1812, but all the records before the year 

 1866 are certainly untrustworthy. In 1812 Keys estimated the 

 number of Todas as only 179, and a census in 1856 only gave 

 316 individuals. In 1866 the number had gone up to 706*. At 

 the census of 1871 the population was found to be 683; in 1881, 

 675; in 1891, 739; and in 1901, 805. It is probable that the 

 increase is largely apparent and due to the greater completeness 

 with which the record has been made at each succeeding census, 

 but the falling off in 1871 and 1881 is perhaps significant, and is 

 consistent with the figures derived from the genealogical record. 



The various records have all shown a preponderance of males, 

 though in the estimate of Ouchterlony-f" made in 1847 the differ- 

 ence is very slight. In 1871 the numbers were about 400 males 

 and 283 females J, and in 1901 451 males and 354 females. These 

 records show that the increase during the last thirty years has 

 been relatively greater among the females but not to any very 

 marked extent. Here again the official records show a general 

 agreement with the figures deduced from the genealogical record. 



* Grigg, Manual of the Niligiri District, 1880, pp. 25, et seq. t Ibid., p. 27. 



X Various errors were made in compiling the statistics of this census of the 

 Toda population and in the corrected figures given by Brecks (Tribes of the Niligiri 

 Hills, 1873, p. 5), the numbers of each sex are not given. 



