DARWINISM DEFENDED. 151 



tainly have come from the wild ones. It is also true that 

 man has made use of only natural factors, and whoever will 

 compare the extraordinary creatures of the deep sea with 

 even the most bizarre of our cultivated races, will see that 

 the fluxing life-conditions of free nature can modify or 

 reform the animal world in no less degree than the intelli- 

 gence of man can do it. ... 



"Recently de Vries, 14 in his book on 'Mutations/ has tried 

 to deny the worth of the selection principle, and although 

 I fully recognise the high worth of his contribution to science 

 based on such extensive series of experiments, yet I must 

 oppose him in this position. In various places in the book 

 he writes that nothing fixed can be produced by selection, 

 and that therefore it can have no importance as a working 

 factor in descent. For example, in the introduction (p. 6) 

 he says : 'Artificial selection never, as far as experience 

 reaches, leads to the origin of ne\v complete types.' The 

 reversion of modified domesticated races is indubitable, and 

 de Vries himself has brought forward new illustrations of 

 this fact which has been so long known. But convincing 

 proof that natural selection cannot lead to constant forms 

 cannot be deduced from these observations, because they 

 refer in all cases to forms which have been highly modified 

 in the course of a few years or decades, so that the pre- 

 sumption lies close at hand that there has not been sufficient 

 time really and lastingly to modify the original heredity 

 established bv centuries. Many facts indicate that the in- 



* J 



tensity of heredity depends upon the number of generations, 

 that is, upon time. Long-inherited characters are difficult 

 to eradicate ; recent ones easy. We can, therefore, not 

 expect to meet such a constancy in the products of a century 

 as we find in Nature. Many gradually selected races of 

 doves are now almost entirely constant, that is, no longer 

 revert to the primitive race when they are inter-bred. The 

 way in which the reversions appear shows that the duration 



