Anomalous Endosperm Development in Maize and the Problem of Bud Sports. 257 



variegatiou to self-color cannot be due to a segregation of factors by 

 which a self-color factor is separated from a variegation factor, for, if 

 a self-color factor, being dominant, were present originalh-, the parent 

 ear would be self-colored, not variegated. If the heterozj-gous, varie- 

 gated parent ear be pollinated by a colorless race, self-colored ears pro- 

 duced from any of its self-colored seeds always behave like Fi ears of 

 a cross between a self-colored race and a colorless race and do not 

 throw variegated ears, but only self-colored and colorless ones, in later 

 generations. Even in this case, then, the bud variation must be 

 due to a change in a genetic factor and is, therefore, a somatic 

 mutation. 



I have also shown (Emerson (12)) that results which were obtained 

 by Correns (4) with variegated-leaved Mirahilis and by de Vries (17) 

 with variegated-flowered Antirrhinum, are subject to the same inter- 

 pretation as the results with variegated-eared maize. 



If bud sports in general are mutations instead of segregations, 

 the occurence of so many recessive characters as bud sports is of con- 

 siderable interest. It has been assumed somewhat commonly that a 

 mutation is much more likely to result from the loss of a factor than 

 from the origin of a factor. But this might be true and still the chances 

 be small for the appearance of recessive bud sports in material homo- 

 zygous for the allelomorphic, dominant factor. If a change affects only 

 one of the duplex factors of a somatic cell, a potential mutation might 

 occur without a visible change in the plant concerned. The only result 

 would be to change a homozygous dominant into a heterozygous domi- 

 nant. A similar change in a dominant factor in heterozygous material 

 would, obviously, result in a bud sport, provided the character concerned 

 could manifest itself in the plant part in which the factorial change 

 took place — loss of factors for chlorophyll development, for instance, 

 would not become manifest ordinarily if it occurred in somatic cells of 

 the root. 



It would be interesting to know wliether recessive bud sports 

 actually occur much more frequently in heterozygous than in homozygous 

 material. If this is found not to be the case, it will have an important 

 bearing upon the problem of whether bud sports are mutations or segre- 

 gations, for the latter could occur only in heterozygous material. More- 

 over, it would have an important bearing upon the question of whether 

 a factorial modification generally includes both or only one of the 

 duplex factors of a somatic cell, and this in turn might prove a bit of 



Induktive Abstammunga- und Vererbungslehre. XIV. 17 



