In the census of 1881, the numbers would appear to have continued to diminish, the population being put at only 675; 382 males and 293 females.
In 1891, the number had risen to 736; 424 males and 312 females.
In the census of 1901, which was taken with especial care to record all the Todas, there were found to be 451 males and 354 females, making together 805.
The population as recorded in my genealogical tables compiled in 1902 was only 736; 419 males and 317 females. My numbers fall far short of those of the census taken a year previously. As I have already pointed out, my genealogies are untrustworthy as a record of the young children of the community now living, and it is possible also that I have omitted a certain number of women. The excess of men over women is distinctly greater in my figures than in the census of 1901, and this may be due to the fact that I failed to hear of a certain number of widows or unmarried women or girls. If so, it is probable that these defects are in the genealogies of the Teivaliol, and it is in them that the excess of men is greatest.
The earlier records of the population are certainly far below the mark. Captain Harkness, writing in 1832, estimates the attendance at a funeral at 300 men, nearly half that number of women, and about as many boys and girls. Those seen by Harkness may not have been all Todas, since Badagas and Kotas undoubtedly attend Toda funerals, but we may safely call this a total attendance of 500, which would show that the records of Hough in 1825 and of Birch in 1838 are far below the mark, and that Birch’s rough estimate of 800 is [[474]]probably far more nearly correct, and may even have been too small.
The records have probably been fairly complete since 1866, and if so, they show a falling off in population from this date till the 1881 census. It is, however, possible that the gradual increase in numbers during recent censuses has been due to the greater care taken at each succeeding census. Unsatisfactory as the records are, they seem to point to a diminution of population about the middle of the last century, which ceased between 1880 and 1890, since which time the population has probably increased.
Mr. R. C. Punnett[9] has analysed the data furnished by my genealogical records to ascertain the average size of the Toda family. He divided the families recorded in the genealogies into four groups: (A) those where the eldest child would in 1903 be over 90 years of age; (B) those where he would be between 60 and 90; (C) and (D) those where he would be between 30 and 60 and between 0 and 30 respectively. He has recorded the results for Tartharol and Teivaliol separately in the following table.
| Group. | Tartharol. | Teivaliol. | ||||
| No. of families. | Average size of family. | ♂s per 100 ♀s. | No. of families. | Average size of family. | ♂s per 100 ♀s. | |
| A | 9 | 3·0 [4·2] | 237·5 | 4 | 4·5 [6·0] | 200 |
| B | 49 | 4·1 [5·0] | 159·7 | 21 | 3·8 [5·4] | 259 |
| C | 87 | 3·3 [3·7] | 131·4 | 40 | 3·8 [5·0] | 202 |
| D | 104 | 2·5 [2·8] | 129·2 | 45 | 2·3 [2·9] | 171 |
The figures in square brackets give the average size of the family for each generation, making allowance for cases of female infanticide, which we shall see presently to be a Toda custom which is almost certainly diminishing in frequency.
The conclusion Mr. Punnett draws from this table is that there has been a marked decrease in fertility during the period covered by the genealogies.