RELATION OF DIMORPHISM TO MUTATION. 33 



The most obvious (liU'eroiU'c between such a variation and the 

 dimorphic ])ranclies and leaves of coffee, cotton, or cacao hes in the 

 fact tliat the bud mutations are of rare and irreo:uhir occurrence, 

 while the chan*>:es of characters shown in dimorphism are rejjjularly 

 repeated durinoj the development of each individual plant or tree. 

 The production of fertile branches in the cotton plant involves a 

 mutative clian<^e of characters away from those that are expressed 

 in the main stalk and the veo;etative branches. But instead of pro- 

 ducino; normality only the one kind of branches with rare mutations 

 to other kinds, the regular course of development for the cotton plant 

 involves the production of two forms of branches, the vegetative 

 form near the base and the fertile form farther up the stalk. In the 

 Triumph variety of Upland cotton there may be said to be a double 

 dimorphism, resulting in three fcirms of branches. The internodes 

 of the lower fruiting branches are very short, like those of the branches 

 of "cluster" varieties, though branches with internodes of normal 

 length are ])rotluced farther up. 



De Vries has proposed to associate bud variations with the form 

 of alternative expression of characters shown in accommodations to 

 external conditions (dichogeny). Accommodations and mutations 

 are alike in the general sense that both may be considered as phenom- 

 ena of alternative expression, but the changes of expression are evi- 

 dently determined in different ways in the two cases. Bud varia- 

 tions represent definitely determined changes in the expression of the 

 characters, like seminal mutations, but in the condition of dichogeny 

 there is no such definite determination of expression on the part of 

 the plant. Changes of expression continue to be dependent on the 

 external conditions and are readily reversible if the conditions are 

 changed. 



The analogy between bud variations . and dimorpliic branches is 

 much stronger, for both of these changes of expression are determined 

 by the plant instead of depending on changes of environment and 

 after the changes are made they are not readily reversible. It is 

 also to be noted that all the three kinds of changes of expression 

 shown in accommodations, dimorphism, and bud mutations take place 

 during the processes of vegetative growth without any apparent rela- 

 tion to the special organs of the germ cells that have been supposed 

 to control the process of heredity. 



For purposes of the study of heredity a very definite distinction 

 is to be made between changes of expression of characters that arise 

 by mutation and those that appear in response to differences of 

 external conditions. The increase in the proportion of simple leaves 

 on cotton plants grown under greenliouse conditions is not the same, 



221 



