THE ESTABLISHMENT OF VARIETIES IN COLEUS BY THE 

 SELECTION OF SOMATIC VARIATIONS. 



BY A. B. STOUT. 



INTRODUCTION AND HISTORICAL REVIEW OF LITERATURE. 



Judgment as to the genetic constitution of a plant is based on the 

 expression of characters in a plant itself, in members of a selfed progeny 

 and in a hybrid progeny. It is largely through the study of the last 

 named that there has developed the conception that characters are 

 represented in cells by unit factors. In considering the evidence as to 

 whether these assumed factors are in any sense units, it is highly 

 essential that the individual be studied as to the variability of char- 

 acters and the range of expressions exhibited in homologous parts. 

 For this purpose so-called bud variations are especially significant, 

 since they represent, perhaps, the extreme of spontaneous somatic 

 variability and suggest that quite as marked variations as exist in 

 hybrid progeny may develop in the parts of a single individual. 



In an excellent summary of all cases of bud variation known to him, 

 Darwin (1868) shows that the phenomenon is widely distributed in 

 the plant kingdom and that it may affect various parts of a plant. He 

 drew the highly interesting conclusions that they include: (1) rever- 

 sions to characters not acquired by crossing, (2) reversions in hybrids 

 to parental characters, and (3) cases of spontaneous variability. 

 The latter he considered as of the same sort as appear in seed progeny. 

 According to Darwin's views, " long-continued and high cultivation" 

 are conditions that induce bud variations, but he recognized that cer- 

 tain cases, especially those when only single buds or parts of buds are 

 changed, do not seem to be due to external conditions. Darwin con- 

 sidered in general that bud variations are evidences of the almost 

 unlimited variability that plants manifest, due to the nature of living 

 structures and the exciting causes of environment. 



Darwin did not believe in fixed hereditary units. He held that 

 characters may respond directly to the effects of environment, and also 

 exhibit spontaneous variability both in seed and in vegetative propa- 

 gation. Furthermore, sexual hybridization was considered to be of 

 influence in modifying and changing inherited characteristics. 



De Vries (1901, vol. i, p. 39) considers that bud variations are spon- 

 taneous changes most common in varieties with incompletely fixed 

 characters. He includes (1889, p. 13) these phenomena under the head 

 of "dichogeny," a general conception proposed for cases in which the 

 nature of the organ may be determined, as he assumes, by external 



3 



