ESTABLISHMENT OF VARIETIES IN COLEUS 



conditions, but he also assumes that the internal constitution may 

 admit of development in several directions. 



De Vries discusses bud variation with special reference to variega- 

 tion. For the frequent and striking cases of the production of varie- 

 gated branches on green plants and the development of green branches 

 on variegated plants, he offers the old explanation of "latent poten- 

 tiality." They belong in general with mutations in that they appear as 

 clear-cut discontinuous variations; the former are classed as progressive 

 mutations and the latter retrogressive (1901, vol. i, p. 606). 



De Vries makes a most comprehensive analysis of the nature of 

 variegation and concludes (1901, vol. i, p. 616) that the capacity for 

 variegation is more widely distributed in the plant kingdom as a latent 

 or semilatent character than perhaps any other character. He notes 

 that true aurea varieties are few and are remarkably constant. Most 

 variegated races show rather wide fluctuations and constitute what he 

 calls intermediate races. His scheme (1901, vol i, p. 424) of represent- 

 ing the relationship is as follows : 



In a later pubhcation (1913) de Vries considers that the pangens, 

 which he assumes are directly concerned with the transmission and 

 expression of characters, may be not only active, semilatent, and latent, 

 but also labile. 



The transition from constant green varieties through variegated 

 varieties to constant aurea varieties is conceived to be dependent on 

 the degree of activity displayed by antagonistic qualities within the 

 cells. The essential changes are conceived to be intracellular and not 

 dependent on qualitative cell-divisions; all the cells are potentially 

 aUke, but different processes within the cells give differences in expres- 

 sion. Thus the conception of de Vries does not consider that the hered- 

 itary physiological units (pangens) are fixed and uniform. They are 

 subject to effects of environment and thej- exhibit spontaneous change 

 even to the degree of sudden appearance bj' progressive mutation. 

 All these phenomena may be exhibited either in vegetative or in seed 

 progeny. Sexual hybridization may entirely modify or change the 

 effect and nature of the basic physiological units of heredity. De 

 Vries's general attitude, however, places the emphasis on discontinuous 

 or mutational changes as the only really stable variations. 



