C 57 ] 
It is ufual with me, to note the yearly differences 
of mortality, which I could not do here, on account 
of the iregularity, but only in decad 21 , &c. By the 
way, I muff give a reafon why I fum up the column 
of lives, which I confider as fo many annual ex- 
pofures, and this as the total of the lives, each ex- 
pofed to the chance of mortality for one year; (i. e.) 
406 in the firff, 402 in the fecond, &c. and 388^ 
in the ten years; and, upon the whole, 38 deaths. 
Exp. Die. Exp. Die. 
And thence 3885 : 38 :: 10,000 : 97.812; which 
laid term is the proper date or degree of mortality 
for that decad, and 9.78 a mean thereof, at an aver- 
age. 
It is generally acknowleged, that fome one be- 
tween the 10th and 20th is the healthieft year, i. e. 
the year in which feweft would die out of 1000, 
and the annual degree (18) of mortality fhould in- 
creafe (fwifter or flower as it happens) from thence 
to the end of life. But how is fuch year to be found 
among the irregularities of the firft of thefe three 
decads ? Or how fhall we look upon 10. it. 14. iy 
as a due progreffion in the fecond ? And if the num- 
bers 16. 19. 20 do go on increafing in the third, 
why does the degree of mortality go back to 17, 
then forward to 20, then back to 17 again, and 
forward to 2 1 per mille ? in fuch a manner, that one 
out of 40.8 fhould die at the age of 36, and but one 
out of 58.6 at the age of 37 ; and again, one out of 48 
fhould die at the age of 38, and but one out of 56.4 
at the age of 39 ? Is not this reprefenting the 37th 
i *8) A miftaken inference from this, fee p. 65. note (3c). 
Vol. LII. I year / 
