MUTATION 175 



gained uniformly red birds. The second genera- 

 tion of the hybrids did show more red than the 

 first, but during the last two years no advance 

 has been made. Again, a cock having a high 

 single comb was crossed with a hen having a typ- 

 ical low pea comb ; the hybrid offspring had high 

 combs with papilla placed high up on each side. 

 An effort to establish by quantitative selection 

 a high pea comb has failed. Dr. Castle tells me 

 that his continued attempts to modify color types 

 of rats by quantitative selection have of late been 

 ineiScacious, since regression is very strong to- 

 ward the original types. The evolution of the 

 American trotter is often cited as a clear case of 

 the results of quantitative selection. Yet is it 

 not true that the advances in recent years have 

 been quite as much determined by the evolution 

 of the sulky and certain technical improvements 

 in handling the trotter and training him? The 

 running record, the result of a larger selection, 

 has, I understand, stood quite still for the last 

 twenty years. Thus even race horses form no 

 exception to the rule that selection, within given 

 characters, soon reaches a period, and improve- 

 ment must wait on the appearance of a new char- 

 acter by mutation. 



This conclusion, far from being opposed to 

 Darwin, would doubtless have been cordially ac- 

 cepted by him, as certain passages in his writings 

 indicate.^ After describing the early improve- 

 ment of the gooseberry, he says : — 



^ Compare the instance given at p. 48. 



