?12 It might be supposed at first sight that plague which broke out at the end of 1896 was solely responsible for the sudden alteration in the character of the seasonal mortality, but it cannot be entirely ascribed to this cause, as a similar alteration in the character of the seasonal mortality took place in 1864-1866 and again in 1876. In 1863 51 per cent. of deaths were recorded in the period June to December, but two years later, co-incident with an increased mortality totalling nearly 10,000, the deaths occurring in the dry season nearly doubled in number and only 37 per cent. of the total took place in the period June to November. In 1894 only 50 per cent. of the total deaths took place in the dry season, but during 1895 and the following years a rapid increase took place in the deaths recorded at that time of the year, and in 1903 over 70 per cent. of the total deaths took place in that period. Since then there has been a gradual return to normal again and in 1910 over 45 per cent. of the total deaths took place in the period June to November. 32. An examination of the return of male and female deaths for a number of past years supports the view that, co-incident with the advent of plague in Bombay increased migration has taken place, leading to an excess of males in the population and causing a complete alteration in the ratio of male to female deaths recorded in the City. Sex Mortality.-The following table shows the ratio of female to male deaths recorded in the City from 1886 to 1909:- Year Female deaths to l,000 male deaths. Year.. Female deaths to l,000 male deaths. 1886 ... ... 931 1898 692 1887 ... ... 910 1899 693 1888 ... ... 872 1900 746 1889 ... ... 892 1901 675 1890 ... ... 893 1902 737 1891 ... ... 925 1903 745 1892 ... ... 951 1904 747 1893 ... ... 866 1905 763 1894 ... ... 903 1906 787 1895 ... ... 855 1907 761 1896 ... ... 826 1908 768 1897... ... 747 1909 756 It will be noticed that the decrease in the proportion of female to male deaths, which took place in 1896, became greater in subsequent years, the maximum variation being in 1901, the year of the Census. From 1901 until 1906 the ratio of female deaths to male gradually increased again, but since that year there has been a slight decline. As a general rule, when dealing with large populations, there is only a slight difference between the death-rates of males and females living under the same conditions; and the only explanation that can be given of the variations shown in the above table is a frequent alteration in the proportion of the two sexes in the population. 33. Effect of Migration upon Mortality.-If we were dealing with a permanent population of known number it would be reasonable to assume that an increase or decrease in the total number of deaths occurring from year to year was due either to a greater or a lessened prevalence of disease; and that this in turn was directly related to the climatic, economic or other con- ditions obtaining at different times. But the factors which determine the variations in the total mortality of large cities in India are usually of a complex nature. In Bombay migration of the population is a factor of tremendous importance both as regards the real and the apparent health conditions of the City. As has already been mentioned, the population of Bombay is augmented each dry season by many thousands of persons most of whom leave the City again before the onset of the south-west monsoon. There are strong grounds for believing that this periodical movement affects at least