246 



NATURE 



[July 



12, 1 



total number of females of the same age — a proportion 

 which compares very favourably with that of widows and 

 spinsters in England. This wonderful power of rapid 

 recovery after decimation by famine or pestilence will be 

 fully exhibited in the tables given below. 



The registration of deaths was in regular operation 

 for several years, both in the North-West Provinces and 

 in Oudh, before the two were united under one administra- 

 tion in 1877. That of births was first introduced generally 

 in 1879, though it had been tentatively commenced in 

 municipalities and cantonments some time previously. 

 We have therefore now (February 1888) ten complete 

 years' death statistics for the united provinces, of a fairly 

 uniform degree of accuracy, and nine years' registers of 

 births, decidedly improving in completeness and accuracy 

 for the first five or six years. The births for the first year 

 of the ten — 1878 — may also be approximately arrived at by 

 a proportionate computation from those registered that 

 year in municipalities. The total number of births of 

 each of the ten years, was as 



Total 8,610,902 



An inspection of the last column shows that these 

 numbers require to be corrected, not only by an allowance 

 for general incompleteness of the records, but by a special 

 addition to counteract the tendency to omit females. 

 During the first seven years this tendency diminished as 

 registration improved, and the numbers of the two sexes 

 approximated more and more to equality ; but even 

 with the most intelligent and careful recording 

 agency, the true ratio between the sexes at birth 

 will never be attained in the records until the opinion 

 of the mass of the people on the relative values of 

 male and female life has undergone a complete altera- 

 tion. The ratios for the first seven or eight years inj 

 the table give a curve apparently asymptotic to a certaini 

 line, the ordinate of which would stand for the ratio 

 attainable by the greatest care in registration under the 

 present conditions. Representing the above ratios for 



b . 

 the first eight years by the formula, a + y, where / is 



counted in years from 1877, we find the ordinate of the 

 asymptote, rt, to be 108-57. In the provinces there are, 

 however, two districts in which the numbers born of the 

 two sexes invariably approach much more nearly to 

 equality. One is Garhwal, a Himalayan district inhabited 

 by an unsophisticated people who claim to be Rajputs, 

 but are probably of aboriginal descent, and who have 

 never come under Muhammadan influence in any way, or 

 acquired the custom of paying a heavy dowry with the 

 bride, which is the cause of female infanticide among 

 many of the higher castes. The other is Lalitpur, in the 

 extreme south, where the inhabitants are chiefly Chamdrs 

 and other low castes, who have never concealed their 

 women or practised infanticide, and amongst many of 

 whom the bridegroom's family pay for the bride. The 

 statistics for these two districts give a series of ratios 

 represented by a curve whose asymptote has an ordinate 

 of ioo'oo, or which points ultimately to exact equality 

 between the sexes. In like manner, if we select for each 

 year that district in which the recorded birth-rate was 



* Estimated from those registered in municipalities. 



highest, and where, therefore, the registration was pre- 

 sumably most complete, we get a curve pointing to an 

 ultimate ratio of 10278 males to 100 females. If we take 

 the mean of all three results, that for Garhwal and 

 Lalitpur being probably below the true average for the 

 whole population, we get 103-78 males to 100 females. 

 This comes very near the ratio for England, which, I 

 believe, is between 103 and 104, and is almost identical 

 with that deduced from the distribution of the population 

 according to age and sex at the last two censuses of the 

 North- West Provinces — namely, 103-75. It may therefore 

 be adopted as a close approximation to the truth, and it 

 shows that, in regard to the relative numbers of the 

 sexes, human nature is much the same in the East and 

 West, notwithstanding the deceptive appearance pre- 

 sented by unanalyzed statistics, as well as by public 

 gatherings in countries where respectable women seldom 

 venture out of doors. 



The numbers of females in the above table must there- 

 fore be all recast so as to give 103-78 males for every 100 

 females. 



This special inaccuracy in the birth tables being cor- 

 rected, there remains the general inaccuracy due to in- 

 completeness of the register, which is common to both 

 births and deaths, and has been estimated by Dr. Planck, 

 after careful and extended personal inquiry, at 20 per 

 cent, of the total, or one-fourth of the numbers recorded. 

 When both causes of error are allowed for, the total 

 number of births in each year will be as in the second 

 column of the next table. The third column gives the 

 recorded deaths, increased by 25 per cent, to make them 

 represent approximately the true mortality, and the last 

 shows the increase or decrease of population each year, 

 due to these causes. The figures in this column repre- 

 sent very fairly the total gain or loss of population, for the 

 number of emigrants is only three or four thousand 

 annually, and this loss is partly balanced by a return 

 migration, the numbers of which are not known. 



During the year of scarcity, 1878, and that of pesti- 

 lence, 1S79 — for the great epidemic of unprecedentedly 

 fatal malarial fever that year surely deserves the name 

 of pestilence — the net loss of population was over a 

 million ; but in the next three years this was fully 

 recovered, and in the succeeding years large numbers 

 were added to the population, especially in the healthy 

 year, 1883. Thus the net gain for the ten years, notwith- 

 standing famine and pestilence, was over two millions and 

 a half, an increase almost unprecedented since the first 

 census in i853,anddoubtless the result of an unusuallylong 

 succession of abundant harvests. Since 1 885, however, the 

 increase has grown less and less rapid ; and as another 

 srarcity is now nearly due, if any trust may be placed in 

 the average period of the recurrence of droughts in the 

 past, it seems likely that in the next two or three years 

 the increase may be temporarily stopped. 



With these figures, and the fixed point given by the 

 census of 188 1, it is possible to find the probable number 

 living at the commencement of each year from 1878 to 

 1888, and also the mean birth- and death-rate for each 

 year of the ten. The census was taken on the night of 

 February 17, 1881, and the total of the people numbered 



