520 ANNUAL KEPOET SMITHSONIAN INSTITUTION, 1908. 



the cytologist shall have given us some clew to the manner in which 

 the sejDarate qualities are represented in the chromosomes, in which 

 the mutative changes must ensue. It is self-evident that in a muta- 

 tion some characters or qualities being borne along steadily from 

 cell to cell in the divisions must be thrown into a latent condition, 

 or perhaps totally lost, while simultaneously or separately other 

 qualities may be acquired, b}^ what actual operation we do not know. 

 It seems fairly obvious, however, that these saltations arising from 

 the nonuniform action of the chromosomes must take place in re- 

 sponse to some stimulus outside of the protoplast in which it actually 

 occurs. This by no means supposes that the stimulation comes from 

 climatic or other environmental factors, but in all probability results 

 from enzymatic or other action from neighboring masses of cells. 

 If this is so, then we may hope to be able to duplicate the process 

 in our cultures and call out a proportion of mutants at our wilL 

 The results of experiments now in progress seem to lend great favor 

 to this assumption. 



This view of the case is also favored by the facts offered by 

 bud sports, which have been designated as " vegetative " mutants, 

 although as shown above, all mutations are essentially of a vegetative 

 cliaracter. In the simplest forms of these sports lateral buds arising 

 generally near the base of a shoot develop branches wdiich diverge 

 definitely from the characters of the main shoot, and which usually 

 coincide with some known form, although this is not always the 

 case. Sometimes entirely new forms arise in this manner, as in the 

 case of seed mutants. During the present season I have been so 

 fortunate as to have three notable examples of bud sports in the 

 experimental cultures. One of these was a basal branch of Oenothera 

 ammophila which sported into the characters of 0. hiennis^ and sug- 

 gesting a j)ossible hybrid ancestry with the latter species as one of 

 the parents. A second case was one in which a seed mutant of 

 O. hiennis gave a bud sport Avhich bore the characters of the ances- 

 tral type or the true 0. hiennis. A third case was one in which a 

 plant of one of the numerous types embraced in a complex hybrid 

 progeny bore a branch which sported in a branch which resembled 

 a sister type. Other anatomical relations are found. Thus a bud 

 sport may embrace not only a branch, but a portion of the main 

 stem, from which it arises, or in other cases it may include a section 

 or longitudinal strip along one side of a branch, even dividing a 

 flower or fruit, while in other cases it may be represented by single 

 flowers or fruits scattered indiscriminately through an inflorence. 



It is to be noted that in most if not all of the sectorial variations 

 by which a part of a bud bears the divergent characters the change 

 is a reversionary one, and the qualities that appear are really latent 

 in the entire j)lant, and only need some stimulus to awaken them or 

 some agency to weaken the dominancy of the prevalent characters. 



