6 ESTABLISHMENT OF VARIETIES IN COLEUS 



attempting to explain them, as they do all heritable variation, by the 

 presence or absence of a unit factor. While this is the general attitude 

 of those who have investigated the inheritance of variegation and the 

 nature of bud variations involving variegation, the results of their studies, 

 as a whole, show great diversity and many necessary modifications of 

 the general Mendelian doctrine of the integrity of unit characters or 

 unit factors and the purity of the segregations of such assumed factors. 



East (1908, 1910 a) considers that a large majority of the known 

 cases of bud variation are due to the loss of a dominant character and 

 that 70 per cent of all known cases are color variations. His discussion 

 and suggestions do not claim to be comprehensive or critical, and he 

 excludes from his treatment all cases of bud variations in variegated 

 plants, because some types of these are known to be pathological. It 

 is interesting to note that he states that no important potato has arisen 

 as a bud sport. He reports four cases of bud variation in potatoes 

 giving white from red or pink which appear to be constant; also several 

 cases giving colored or purple blotched from pure white, concerning 

 which no data are given except the statement that they are not con- 

 stant. East's view is that bud variations are due to loss or latency of 

 hereditary units that stand for characters. 



Bateson (1909, p. 273), especially, has advocated, on theoretical 

 grounds largely, that bud variations are due to qualitative cell-divisions 

 in somatic tissues, giving somatic segregation of unit factors. The 

 idea is quite identical in its main features with Weismann's concep- 

 tion of qualitative divisions, giving tissue differentiation in ontogeny. 



In considering the nature of the albomarginate types of variegation, 

 with special reference to the origin and the development of the green 

 and white areas, Baur (1909) has made criticial anatomical studies of 

 a white-margined Pelargonium zonale which indicate very clearly the 

 relationship of white and green tissues as periclinal chimeras and 

 explain the appearance on them of branches wholly green or white. 

 In testing, by crossing experiments, his assumption that the green and 

 white tissues are each pure, he is forced to the further assumptions that 

 the loss of green in this case is due solely to the condition of the chro- 

 matophores and that male sex-cells carry chromatophores. He does 

 not consider that white cells can arise spontaneously from green, or 

 vice versa. Yet this probably did occur when the chimera was first 

 produced. In fact, the numerous varieties that are peripheral chimeras 

 not only in Pelargonium but also in other genera indicate that such 

 spontaneous loss of power to produce green is frequent. We may say 

 that Baur's interesting results, however, do show that if spontaneous 

 loss occurs in young leaves after they are formed, mottled or striped 

 variegation is produced, but if the loss occurs in the growing-point itself, 

 then chimeras will result, their constancy depending largely on the 

 relative permanency of the change. 



