290 F. B. SUMNER AND H. H. COLLINS 



The situation as regards one of the color varieties — the 'yel- 

 lows' — has proved to be less simple than was at first supposed. 

 This variety has appeared in several independent descent lines, 

 and two of these lines seem to differ from one another character- 

 istically in respect to the exact shade of 'yellow' which is mani- 

 fested. Furthermore, neither of these last-named races appears 

 to behave strictly like a monohybrid recessive. It seems likely 

 either that there are independent 'modifying' factors concerned, 

 which segregate according to principles not yet ascertained 

 by us or that the 'yellow' factors themselves are unstable and 

 undergo changes of some sort. 



We are quite aware that the observations here presented are 

 decidedly fragmentary in comparison with the more exhaustive 

 investigations of many recent mendelian students. This has 

 been due to several circumstances. In the first place, our studies 

 of color mutations have throughout been regarded as incidental 

 to our main program of work, which has concerned itself with 

 the characters of 'natural' subspecies. In the second place, 

 Peromyscus is not well adapted to experiments in which rapid 

 breeding is an important consideration. Not only is it commonly 

 impossible to obtain more than two or three generations in a 

 year, but the cage-born mice show a considerable percentage 

 of sterility, w^hich may at any time bring some valuable descent 

 line to a close. Questions which one could promptly settle by 

 appropriate matings with more favorable material must await 

 many months for an answer or must even remain unanswered. 

 Finally, we must mention that these studies were interrupted 

 when far from complete, and that a large part of the stock was 

 killed at that time, although certain lines of experimentation were 

 resumed later. This interruption was due to the absence of the 

 senior author from La Jolla during the greater part of one year 

 and to the junior author's permanently severing his connections 

 with the Scripps Institution in the fall of 1919. 



These explanations, w^e trust, will temper the criticisms of 

 those who may be disposed to wonder why we have failed to 

 settle certain obvious problems or even to attempt certain obvious 

 experiments. The ruthors may fairly ask to be credited with 



