212 WALTER F, WILCOX 



age question, although the influence of this upon the number 

 of children under 10 is uncertain. The next largest decrease 

 was in the decade from 1860 to 1870, when the direct and in- 

 direct effects of the Civil war reduced the proportion of children. 

 But this decrease was accentuated by the serious omissions of 

 that census, especially in the southern states and among the 

 negroes, for whom the proportion of children is very high. 



Statistics as a whole suggest that there has been an almost 

 uninterrupted but irregular decrease in the birth rate from near 

 the beginning of the nineteenth century. They do not prove 

 this, for the decrease might be explained by the increasing 

 vitality of the population, leading to a longer average duration 

 of life and consequently the survival of a larger number of 

 adults. 



The method of estimating the proportion of children, by 

 comparing them with the number of women of child bearing 

 age, may be applied for the period 1850 to 1900. This method 

 has two advantages over the preceding. It makes it possible 

 to limit the children to those under 5 years of age and to ex- 

 clude from the second term of the comparison all males and 

 the females not of child bearing age. Under these conditions 

 any decrease in the proportion of children which the figures 

 may show could not be explained as due to the increased vital- 

 ity and longevity of the adult population. The limits of child- 

 bearing age are usually assumed as 15 and 44, but for the 

 censuses the limits must be taken as 15 and 49. 



The proportion of children increased from 1850 to 1860, 

 and then decreased without a break but by very unequal 

 amounts. The slight decrease from 1870 to 1880 was probably 

 due in part to serious omissions in 1870 among the population 

 having a large proportion of children. The slight decrease 

 from 1890 to 1900 was probably due in part to the great pros- 

 perity of the country between 1890 and 1900, especially in the 

 last years of the decade, in part to the many children bom to 

 the millions of immigrants of the preceding decade, and in part 

 also to the change in the form of the age question. 



In 1900 there were only three fourths as many living chil- 

 dren to each 1,000 potential mothers as in 1860. The assump- 

 tion that there has been a progressive increase in the inaccu- 



