98 Selection Does Not Lead to Origin of Species. 



sowed alternating rows of Galland wheat and beardless 

 wheat; and it was discovered that the former suffered 

 more from the cold of winter than the latter and that it 

 ripened from 8-14 days later; these two differentiating 

 characters sufficed to enable the small initial admixtures 

 of the beardless wheat to get the upper hand within three 

 years. 



RiMPAU^s observation concerns the rough wheat (Ri- 

 vett's Bearded, Fig. 21) of which it is often asserted 

 that it easily degenerates and that after a few years it 

 always contains a larger or smaller proportion of beard- 

 less examples. But when kept absolutely free from ad- 

 mixture, as in RiMPAu's experiments in Saxony, this 

 form remains perfectly true; as indeed it has done in 

 Scotland for over a hundred years. "But it was more 

 liable than any of the varieties of wheat we grow to suffer 

 from the winter, and as it develops later in the spring 

 than all other sorts it is easy to understand that all chance 

 admixtures, which are unavoidable in cultivation on a 

 large scale — and are apt to be introduced by the use of 

 farmyard manure — will multiply much more rapidly and 

 soon obtain the upper hand."-^ 



The results of experiments in acclimatization, there- 

 fore, can only be used as scientific evidence when the 

 danger of crosses, or of degeneration by the chance ad- 

 mixture of native races is absolutely excluded. 



The best examples have been collected by Schubeler 

 who is himself responsible for some of the results in his 

 collection.- They relate chiefly to the acclimatization of 



* It is well known that in bad years Avena fatua multiplies rap- 

 idly among oats ; it contributes nothing to the yield. Godron, De 

 I'Espece, I, p. 163. 



^ Schubeler, Die Pilan::enwelt Norzvegens, 1875. Also, Die 

 Culturpiianzen Norwegens. 



i 



