LAW OF POPULATION ILLUSTRATED 191 



former, which witnessed so many more marriages. 

 In the first decade the children born to 100 

 marriages were but 425, whereas in the second 

 485 children were born to 100 marriages. This 

 was not owing to the greater fecundity of the 

 marriages made in that decade, but to a large propor- 

 tion of the marriages made in the first continuing 

 their productiveness in the second. In any decade we 

 may estimate that one-half of the births are due to 

 marriages made in the preceding decade. Thus it 

 often happens that though a decade has a higher 

 marriage-rate, the following decade with a greatly 

 lower marriage-rate may have the higher birth-rate. 

 This fact makes it evident that it is utterly futile 

 to hope to ascertain the average fecundity of any 

 people by using for that purpose only the marriages 

 and births of a single decade, or even of two successive 

 decades. By taking in the half -century from 1854 

 to 1903, I find the average fecundity of Dutch 

 marriages to be 456 births to 100, a proportion 

 which is not shown in any one of the five decades. 

 I do not doubt that, if it had been possible for me 

 to obtain exact returns of the marriage-rates and 

 birth-rates for a century, I should have attained to a 

 still closer approximation to the exact ratios. It 

 must not, however, be lost sight of that in the above 

 computations I have treated all the registered births 

 as legitimate ; which doubtless renders my calcula- 

 tion as to the average fecundity of Dutch marriages 

 somewhat too high. 



