762 



THE AMERICAN NATURALIST 



[Vol. XLIV 



. . . Since the beginning of the 

 to any new races or types. He 



But even those of us who do not believe that all evolutionary 

 change is saltatory, as it seems to be in the (Enothera mutants, 

 can agree with de Vries that the difference between fluctuation 

 and mutation lies in the fact that fluctuation is due to environ- 

 ment and is not hereditary, while, when a step has actually 

 been accomplished in permanent evolutionary change, the norm 

 about which fluctuation occurs is different from the old One. 

 "We can accept this doctrine even if we deny that the difference 

 between the new and the old is not a "unit." We can not, how- 

 ever, accept the idea, repeatedly brought forward in this book, 

 that "There is no question that improvement takes place in the 

 experimental garden'' (p. 110) when selecting for improvement 

 in pure lines, or when he says "In the case of no single char- 

 acter can selection be relaxed" (p. 106) ; or when he quotes 

 Halley (p. Ill), with approval, to the effect that in improving 

 wheat by selection, "the rate of improvement gradually falls off 

 year by year until at the end of many years the race reaches 

 a maximum and becomes constant. But, of course, it will not 

 remain so if it is not subjected to continuous selection." 



W. J. Spillman. 



(To be continued.) 



