[ III J 
aftions (vol. LVII. p. 46 ij, of the i?7creafe and 
mortality of the inhabitants of the if and of Madeira. 
In this ifland, it feems, the weddings have been 
to the births,, for 8 years, from 1759 to 1766, as 10, 
to 46.8 j and to the burials as 10 to 27.5. Double 
thefe proportions, therefore, or the proportion of 20 
to 46.8, and of 20 to 27.5 are the proportions of the 
number marrying annually, to the number born and. 
the number dying. Let r marriage in 10 be a 
2d or 3d marriage on the iide of either the man or 
the woman j and 10 marriages will imply 19 indi- 
viduals who have grown up to maturity, and lived to 
marry once or oftener; and the proportion of the 
number marrying annually the firft time, to the num- 
ber dying annually, will be 19 to 27.5, or near 3 to 
4. It may feem to follow from hence, that in this 
ifland near three-fourths of thofe who die have been 
* married, and, confequently, that not many more than 
a quarter of the inhabitants die in childhood and celi- 
bacy } and this would be a juft conclulion were there, 
no increafc, or had the births and burials been equal. 
But it muft be remembered, that the general effedtof 
an increafe, while it is going on. in a country,, is ta 
render the proportion of perfons marrying annually 
to the annual deaths greater^ and to the annual births* 
lefs than the true proportion marrying out of any 
given number born. This proportion generally lies 
between the other two proportions, but always neareft. 
to thefirft* j and, in theprefent cafe, it is fufficiently 
evident that it cannot be much lefs than two^ thirds. 
In a country where there is no increafe or decreafe of the 
inhabitants, and where alfo life, in it? firft periods, is fo liable,. 
In. 
