92 BIOLOGY AND SOCIAL PROBLEMS 



have the coat-color of the wild parent, but 

 of their progeny one fourth show the mark- 

 ings of the hooded stock. This hooded char- 

 acter, then, is what may be called a unit char- 

 acter. 



Castle has attempted to ascertain whether 

 this unit character can be modified by a proc- 

 ess of selection, the object being to produce 

 from the hooded stock by selection two ex- 

 treme conditions : a completely light animal, 

 and a completely dark one. In the beginning 

 there were selected from the common hooded 

 forms, two sets of individuals : one with as 

 little pigmentation as possible, the other with 

 as much. These two sets ajfforded the mate- 

 rial with which the test began. Their descend- 

 ants were selected in the desired directions 

 generation by generation till, after somewhat 

 more than six years of work on over twenty- 

 five thousand rats, two clearly divergent stocks 

 were produced. One of these was composed 

 of members entirely light except for a small 

 amount of dark on the head, and the other 

 was made up of individuals completely dark 

 except for some light on the belly. Thus two 

 stocks were obtained that bred true and ap- 

 jroximated closely to the two extremes aimed 



