2o8 WALTER F. WILCOX 



But for most purposes a comparison with the number of women 

 of child bearing; age seems the best. The number of married 

 women of child bearing age is known only for 1890 and 1900. 

 Partly for this reason, partly because many of the influences 

 tending to decrease the birth rate tend also to decrease mar- 

 riages, and partly because limiting the comparison to mar- 

 ried women excludes the influence of illegitimacy, the com- 

 parison between children and married women should be used 

 only in a subsidiary way. 



The number of children under 5 years of age and also the 

 number of women of child bearing age in the total population 

 have been reported by the censuses only since 1850; the num- 

 ber of children under 10 years of age has been reported by 

 censuses since 1830. Accordingly, an accurate statement of 

 the proportion of children under 10 years of age to the total 

 population can be made for the last seventy years of the nine- 

 teenth century. For 1800, 1810, and 1820 the number of 

 free white children under 10 years of age was given, and for 

 1820 the number of free colored and slaves under 14 years of 

 age. From this information an effort has been made to esti- 

 mate approximately the total number of children under 10 

 years of age at each of these earlier censuses by aid of the 

 assumption that as the negro population under 14 years of 

 age in 1900 is to the negro population under 10 years of age in 

 1900, so is the negro population under 14 years of age in 1820 

 to the negro population under 10 years of age in 1820. For 

 1800 and 1810 the free colored and slave population under 14 

 years of age has first been estimated from the total free colored 

 and slaves of all ages by assuming that the proportions of 1820 

 applied, and then from these estimates the free colored and 

 slave population under 10 years of age has been estimated as 

 in 1820. It is admitted that the results are only approximate, 

 but it must be remembered that these estimates applied to 

 only one sixth of the entire population under 10, five sixths of 

 it being given by direct enumeration. 



No census can furnish all the information needed to com- 

 pute the birth rate or number of births in a year to each 

 thousand persons, nor has this information been obtained for 

 the United States, or any considerable part of it, by any other 



