PHYSICAL EVOLUTION 219 



compel us to decide that over 48 per cent, of the population are 

 not to be reckoned in, when we are considering the question of 

 physical evolution. The physical future of the race has nothing 

 to do with them. 1 



But as yet we have not seen by any means the whole extent Repro- 

 of the elimination that goes on. There is constantly at work selection 

 the principle of reproductive selection, as Professor Karl Pearson 

 has called it, and this is, in reality, only a form of Natural Selec- 

 tion. 2 Twenty-five per cent, of the mated population in England 

 and Denmark produce one-half, or slightly over one-half, of the 

 next generation. Here is a very remarkable fact. In estimating 

 its importance we must bear in mind that fertility may be re- 

 garded as a proof of good physique. This can hardly be dis- 

 puted, even if we consider, not the net but, the gross fertility, i.e. 

 if we reckon all the children born, and not only those who grow 

 up to continue the race. In many vigorous families there is not 

 much difference between the two the net and the gross since 

 of a number of children born most survive till the age of 

 marriage. In fact, through the working of Natural Selection 

 fertility must inevitably come to be correlated with fitness to 

 survive. This being so, reproductive selection must be counted 

 among the causes that go to make the race strong. It is, in fact, 

 a most effective, though less obvious, form of Natural Selection. 3 

 Families which are highly unprolific are likely to become extinct; 

 the race is kept up by, and gets its character from, the more 

 fertile stocks. 



1 Illegitimate births have to be left out of the calculation as the ages of the 

 parents are not given in the returns. But they numbered in 1896 only 4-2 per cent, 

 of the total of births, and are, therefore, an unimportant factor. See the Registrar 

 General's Return, p. 17. 



The ages were unrecorded in less than 2 per cent, of the marriages. In some 

 cases there is, no doubt, an understatement of age. This is shown by the fact that 

 in each successive census the number of women returning themselves as between 20 

 and 25 is larger than the number of girls returned in the census of ten years earlier 



as between 10 and 15. See the " Census Report " for 1891, vol iv. p. 28, where it 

 is shown that the error due to this amounts to less than 5 per cent. If we assume 

 the same percentage of mis-statement in the case of brides, it will not materially 

 affect our conclusions. 



2 See Natural Scitnct, May 1896. 3 See p. 240. 



