No. 600] PIEBALD RATS AND MULTIPLE FACTORS 



723 



different ones are combined with the first ones, the results 

 may be reversed. Now to stndy one variable (germ 

 plasm) through a measure (soma) influenced by a second 

 variable (environment) will seldom give correct results 

 if the effect of the second variable is not clearly recog- 

 nized and discount made for its influence. In the present 

 case, it appears that the curve of the averages can only 

 show the degree to which the variations due to environ- 

 ment and the germinal variations tend to go in the same 

 direction. That there is a rise in the curve shows that, 

 on the average, they are a little more likely to agree in 

 direction than to contradict each other. On the other 

 hand, since the environmental variations can not be ac- 

 counted for and eliminated, the curve gives no informa- 

 tion as to the actual or relative potencies of either set of 

 variables. That there are no fluctuations in the curves 

 may have been assumed to prove that environment is con- 

 stant and therefore does not demand consideration. But 

 this conclusion can not be safely drawn from the facts. 

 The curves probably do mean that, when generation is 

 compared with generation, the variations of the environ- 

 ment are cancelled out; they mean that these environ- 

 mental, or extra-germinal, variations occur within a gen- 

 eration, and probably within a family or within the gonads 

 of the parents. Environment might well be ignored were 

 the ultimate question to be answered, "How much can 

 selection change the average grade of hooded rats?" 

 But this is not the main question. The question to be an- 

 swered is, "What is the nature of the changes in the germ 

 plasm V 



In view of all this, one can find slight justification for 

 assuming that the germinal variations were small and 

 constantly occurring. It seems entirely possible that the 

 environmental, or extra-germinal influences were strong, 

 perhaps even more effective than the germinal constitu- 

 tion. In this case, there would be no need to assume a 

 very large number of factors to find a multiple factor ex- 

 planation for the slow advance wrought by selection. 



