ern Mexico, therefore, are clear cut and readily recognizable 

 without the application of any greatly refined methods. The 

 obvious distinction is found in the feathers of the tail and upper 

 tail coverts which in the United States bird are ])roadly tipped 

 with rich chestnut whereas in the Mexican subspecies these parts 

 are white or nearly white. Such characters, in animals of un- 

 known history, might easily be looked upon as produced by 

 mutation; but with complete gradation from one to the other 

 known to exist in nature, it is hard, at least for some of us, to 

 believe that the difference was not accomplished by gradual 

 rather than sudden change. If it could be shown that char- 

 acters of this kind behave as hereditary units without any such 

 blending as requires ''dialectic gymnastics" to explain, it would 

 be a long step forward in the correlation of natural and man- 

 made experiments. Such characters are in fact heritable, as 

 has been shown by Sumner in his breeding and transference ex- 

 periments Avith Feromyscus. This is illustrated also by an un- 

 directed experiment to call attention to which is one of the ob- 

 jects of this communication, namely, the test of subspecific char- 

 acters which has been carried out in the domestication of the 



As is widely known to sportsnnen, breeders, and many others, 

 our domestic, so-called bronze, turkey is readily distinguished 

 from the wild bird of the eastern United States by the coloration 

 of the upper tail coverts ajid tail. The reason for this, which is 

 not so generally known, is not that the domestic bird has changed 

 in these respects under man's intluence, but because it is the 

 direct descendant of the ^Mexican wild race which differs from 

 the northeastern race by these selfsame characters. Carried 

 from Mexico to Europe in the early sixteenth century and 

 thence brought to the United States, it has continued for more 

 than three hundred generations in a new environment maintain- 

 ing its old established subspecific characters. To-day it may 

 differ from the Mexican wild race in some details, hut its <:enoral 

 coloration is the same and especially does it retain its taxonoinic- 

 ally diagnastic features. These, therefore, are heritable and 

 doubtless related to germinal conditions which became lixed in 

 the wild bird. Since the characters themselves are of the kind 

 that appear to be produced by insensible gradations and of the 

 kind that frequently bear an obvious relation to environment, 

 an easy deduction would be that the germ plasm also has 



