suggestion was made that, a double census should be held, one enumeration in the
dry weather and another in the rains, but this would of course only give figures
for one year, which might be completely upset the following season.
In view of these facts it has been considered wise to attach no weight to
conclusions based upon the mortality rates of Bombay.
42. Errors in death-rates.-But the error in calculating death-rates upon
an unknown and fluctuating population is still further increased by the fact that
the crude death-rate takes no account of the number of sick people who leave
Bombay and die in other places; and also that in calculating the crude death-rate
no allowance is made for the age and sex distribution of the population. It is a
well known fact that persons of different ages die at different rates, mortality
being very high among young children, low among young adults and high again
among the old people. There are also slight differences in the death-rates of the
different sexes, but these are not so great in India as to be important. Towns
usually contain an excess of young adults whose death-rate is less than that of
the general population; and as a result the recorded death-rate of towns is
usually too low.
43. Various methods have been devised for correcting the errors in death-
rates arising from sex and age constitution of a population. The method most
generally used in England is based primarily upon the actual death-rate of
each sex at different ages throughout England and Wales. In use a hypothe-
tical death-rate is calculated for each large city, based on the number of deaths
that would have occurred in any given year if the population existing in the
city at different age periods had died at the same rates as persons of the same
age in the country as a whole. This hypothetical death-rate is called the
standard death-rate. The correction factor is obtained by dividing the actual
number of deaths in the year with the hypothetical total obtained under the
standard death-rate.
If we wish to compare the mortality rates of Bombay with those of the
whole Presidency, we must make a somewhat similar correction, and we find
that the standard death-rates for Bombay for the years 1907, 1908 and 1909
calculated on the assumption that the City population at various age periods
died each year at the same rate as those of the same age in the whole
Presidency were:-
1907 ... ... ...
281 per 1,000
1908 ... ... ...
224 " "
1909 ... ... ...
229 " "
The total number of deaths that would have occurred at these rates (on the
basis of the 1906 census) was-
1907 ... ... ... ...
1908 ... ... ... ...
1909 ... ... ... ...
And the actual number of deaths occurring each year divided by these figures
gives the correction factors of-
1403 for 1907
1741 for 1908
1566 for 1909
When corrected by means of these factors the death-rate for the City for the
three years becomes-

Crude death-rate as recorded in Bombay ... ...
Death-rate corrected for age distribution of population .
These figures are not put forward as being accurate, as neither the census
figures for the Presidency nor those for the City are correct, but they serve
to indicate the fallacy of basing conclusions regarding the health of Bombay