848 
DE. TAEE OX THE COXSTEECTIOX OE LIFE-TABLES. 
now out of this number (384) at that age three die when the moitaKty is This 
accounts for the difference of -00768 and -00771 ; the former occuriing in a year out of 
a unit of life of which the waste is not replaced. 
From these considerations it may be inferred that if is known, may be deduced 
from it upon the hypothesis of equal decrements through the year by the foi-mula 
Thns m,. being -0077072, we have = -99232 * as before. 
The Xpao previous method is 1-9966528, and by this method it is the same. B\ 
either of the methods the value of may be deduced for the subsequent ages, and 
will be obtained. These values are here given, and it will 
he seen that the results by the two methods are nearly identical at all ages, except the 
two last, when the observations themselves become less exact. 
Females. 
Age (x). 
K2m 
20 
1-9966528 
r-9966527 
30 
•9960967 
•9960967 
40 
•9966263 
•9956264 
60 
•9946669 
•9946676 
60 
•9902049 
•9902073 
70 
•9773538 
•9773557 t 
80 
•9463182 
•9462643 1 
90 
•8809176 
•8801776 
It will be observed that the fraction approximates to \~-m as m becomes 
less ; for upon developing it into a series, = 1 -- m + taking in 
infinitely small, the terms after the two first may be neglected. 
The values of Wo, m, m, may be obtained by the method akeady described. But 
it rarely happens that the population living at each year of age is acciu-ately enumeratec 
at the Census ; and besides inaccuracies of statement, the numbers Imng at each of the 
early years of age fluctuate considerably, so that the numbers of children lining of each 
-year of age in 1861 do not represent the average numbers living of those ages in the 
five years 1849 to 1853, for instance. 
The following method is less exceptionable. It may be assumed for ths ^urpobc 
(1) that the births registered in the year 1848 represent the births in that year- ; (2) that 
the births are equally distributed over the years in which they occim, and consequent!) 
* OT at tlie precise age 20 is nearly -00765. The increase in this mortahty from the age 20 to 20h 
the middle of the year of age 20 to 21 is obtained by adding i\r, as above given, to that is, to the 
log of (OTi 94 +»i 2 oi)l ; ■»i 2 o— "0077072 
\?Wi94 3'8835130 
O OO3380-1 
Xniio 3" 886899-4 
