88 



THE AMERICAN NATURALIST [Vol. LV 



changed gradually and, at least as a working hypothesis, one 

 might suppose that the germ plasm had been affected directly or 

 indirectly by the environment. Such an explanation is, of 

 course, far too simple and old-fashioned for present-day stu- 

 dents of evolution. It does not explain a multitude of undeni- 

 ably important and fascinating results of experimental work. 

 But neither do theories of mutation and the maze of modern 

 genetics explain the intergrading subspecies and perhaps there is 

 room for at least a little experimental work which does not deny 

 such an hypothesis at the outset. 



In the case of the turkey, while the relatively inconspicuous 

 subspecific character has proved itself stable, violent saltatory 

 changes have been established easily. These are of the sort 

 common among close bred domesticated animals but so rare 

 among wild vertebrates that no one has yet found a case in 

 which they can be shown to have been perpetuated by natural 

 process. Thus we now have self-colored breeds of turkeys re- 

 spectively black, white, buff, and blue gray as well as the breed 

 called Narragansett in which the feathers are tipped with steel 

 gray. Hence it seems that the turkey may offer an opportunity 

 for comparative study of the hereditary behavior of characters 

 which have developed naturally by what seems to be continuous 

 variation and those which have appeared diseontinuously and 

 been perpetuated artificially. Sumner (1. c, 1918) has found 

 with Peromyscus that hybridization of different subspecies pro- 

 1 s tl 3 and generations a blending of the subspecific 

 characters comparable to the gradations found in nature, 

 whereas mutant characters (partial albinism, etc.) act as simple 

 Mendelian units. In other words, natural subspecific characters 

 act in hybridization experiments as they would be expected to 

 do on the assumption that they were produced by continuous 

 variation. Whether or not the same results would follow with 

 the turkey and other forms would seem to he well worth deter- 

 mining. In a general way, it is known to breeders tliat hybrids 

 between wild and domestic turkeys are of intermediate type, 

 but so far as I know careful well-controlled work has not been 



Wilfred H. Osgood 



Field Museum of Natural History 



