Stout: The Origin of Dwarf Plants 447 



nodes which later appeared more intensified in the dwarf plants. 

 This, coupled with the variability of the progeny, might be consid- 

 ered by Mendelians to indicate that the parent plant was a half- 

 mutant. While it is no doubt true that such cases do occur (a 

 most notable case is that of Oe. semigigas), it is evident that the 

 conception has been applied to many cases of irregular inheritance 

 and sporadic variation and even to instances of cumulative 

 variation. 



A most important type of discontinuous variation is that of 

 simultaneous variation in a group of characters, well illustrated by 

 the dwarf Hibiscus. Not only is a number of characters modified 

 but the habit of profuse branching from the base appears as a char- 

 acter quite new to the parent species. Such phenomena are not 

 well explained on any conception of continuous unit characters. 



Bateson does not admit that such simultaneous variation can 

 occur as a sporadic variation. He does not believe in the mutation 

 and inheritance of group-characters as described by De Vries and 

 already mentioned with respect to characters of Oe. nanella. He 

 prefers to think of the phenomena as due to a recombination of 

 factors. His strongest evidence for this view is the apparent 

 marked reversions t'lat appear in certain crosses. If, however, 

 recombinations can give new groups or develop characteristics 

 new to the particular race and species, his analysis is of doubtful 

 validity. 



To assume that a variation such as the dwarf Hibiscus is due 

 to a single c lange in a single hereditary unit is to assign to the 

 units different values and to admit of interaction between units. 

 On the other hand to assume that there are hereditary factors that 

 are themselves stable but that can interact upon each other in 

 various ways is to assign different values to the supplementary or 

 coordinating units and to the various interactions between them. 

 This makes the comprehensive description of the processes of 

 heredity in terms of units other than characters of doubtful valid- 

 ity. 



It is quite clear that hereditary variations giving dwarf forms 

 of various degrees of intensity and extensity do arise. If in all 

 cases single characters, factors or other hereditary units are con- 



