66 Progress of Popvlation and Wealth 



whites of both sexes under twenty, and decennial for all above that 

 age and under 100, it had afforded the means of estimating, with 

 great accuracy, the probability of life of each sex at different pe- 

 riods by comparing the numbers of the several classes in the pre- 

 ceding census, with those of the classes ten years older in the suc- 

 ceeding census, if it were not for the interference of two causes, 

 whose quantities we have no means of precisely ascertaining. 

 These are, the diminution of males from boyhood to middle age, by 

 roaming and going to sea, and the increase of both males and fe- 

 males, but in unequal quantities, by immigration ; of which disturb- 

 ing influences the census affords us the most satisfactory evidence. 

 Thus, the class of females between fifteen and twenty, in the cen- 

 sus of 1840, which corresponds to the class between five and ten, 

 in the census of 1830, instead of exhibiting a decrease, by reason 

 of the deaths in the intervening period of ten years, shows an in- 

 crease of 41,427, equivalent to 5^ per cent; which effect must ne- 

 cessarily have been produced by accessions from abroad, suppos- 

 ing the ages of the females to be accurately noted.* Thus, too, 

 whilst the females of this class show an increase of 51 per cent, a 

 similar comparison of the males between five and ten, in 1830, with 

 those between fifteen and twenty, in 1840, exhibits a decrease of 

 3£ per cent ; which seems to indicate that, although immigration 

 has considerably swelled their numbers in ten years, it has done so 

 to a less extent than with females, principally by the number of 

 boys who have gone abroad, and in some degree by the greater 

 mortality of males, which is manifested by the general tenor of the 

 census. 



It is proper to add that the same sources of error which have 

 been mentioned, must affect any estimates that can be made of the 

 probabilities of life in the United States, and that, therefore, the 

 tables that have been given must be regarded as only approxima- 

 ting to the truth. 



Let us now advert to the coloured race in reference to this sub- 

 ject. 



The following tables compare the decrease of life between the 



* As it seems scarcely credible that the number, at any period of life, should have 

 gained by immigration in any given time equal to the loss sustained in the same time by 

 death, it is rational to suppose that some error has crept into this part of the census. 

 Can it be that many of this class of females, who work from home, are counted twice ? 

 or must we suppose that many, who have passed twenty, have reduced their age within 

 more desirable limits ? 



