198 ROLLINS 



(1891) and the more recent work of Clausen, Keck, and Hiesey (1940) have 

 forcefully shown the fallacy of these attempts to demonstrate that the en- 

 vironment can provide the causal factors required to produce the far-reaching 

 changes that would make one well-recognized species become identical with 

 another. However, the modifications of plants by the environment can be 

 rather great. Environmental modifications are as numerous and varied as 

 the combinations of factors that produce them. The taxonomist early learns 

 to be wary of placing any reliance upon such modifications for purposes of 

 classification. Indeed, one of his first tasks in any detailed study of species 

 is to find out which part of the phenotypic variation he is dealing with is 

 correlated with the environment and which part is under direct genetical 

 control. The complicated interplay of genetical and environmental factors in 

 producing the ultimate phenotype makes the distinguishing of these forces 

 a task of large proportions. It calls for experimental procedures that are so 

 expensive in time, effort, and money that for many plant groups the kinds of 

 information needed will be exceedingly slow in coming, if at all. 



Fortunately, men like Hall (1932), Turesson (1922), Clausen, Keck, and 

 Hiesey (1940), and Marsden- Jones and Turrill (1931) have pointed the 

 way by developing methods of differentiating between environmental modi- 

 fications and hereditary differences. Each such study adds immensely to our 

 understanding of the plant group that is so studied. But one point in this 

 connection should be reemphasized. It would take a millennium of hard work 

 by many more botanists than are now active to produce the kind of study for 

 all plants we now have as models from the work of Clausen, Keck, and 

 Hiesey. Fortunately, too, a wide general experience with growing plants will 

 often permit an investigator to short-cut to a sound position with respect to 

 the variation in a given group, without resorting to exhaustive experimenta- 

 tion. For example, an experienced taxonomist quickly learns the nature of a 

 whole constellation of phenotypic effects that are produced by bringing a 

 wild plant into cultivation, where abundant fertilizer and water are supplied. 

 He knows what modifications to expect, when shade plants are found in the 

 sun, and vice versa. In fact, he is always wary of quantitative characters asso- 

 ciated with vegetative growth up to a point, and unless the differences are 

 considerable, he will not use them for the purpose of trying to distinguish 

 between taxa. Even after he feels confident of quantitative information he 

 must remain wary of chimeras, sucker shoots, and the like. His aim is to 

 discern the characters, the variations of which are genetically controlled, for 

 it is those characters that reflect phylogenetic and evolutionary relationships, 

 which in turn make up the framework of a sound taxonomical system. 



