GABRIELLE L. C. HOWARD. 



77 



combine to make up the length of the decurrent portion of the 

 lamina in Type 51 is therefore probable. 



The results in cross Type 16 x Type 35 (Table XV), are 

 also interesting. Here both parents have the same average 

 value, and the range of variation in the F^ generation is exactly 

 similar to that of the parents. The length in the F^ generation 

 varies from .5 cm. to 6.5 cm., far beyond the limits of the 

 parents. No plants with non-decurrent leaves were found, but 

 this may have been due to the comparatively small number 

 measured. In the third generation there were great differences 

 among the cultures, and plants with non-decurrent leaves 

 occurred in three cultures. The numbers in all these cases 

 (cultures 231, 202, 27) point to a 63: 1 ratio and consequently to 

 the existence of three factors in these particular cultures. Higher 

 values were found in the F^ than in these cultures indicating 

 the existence of additional factors in the parents. This supposi- 

 tion is confirmed by the high values found in the .F3 culture 163. 

 The large number of factors involved would explain the absence 

 of plants with non-decurrent leaves in the F,,. The factors which 

 are responsible for the length of the decurrent portion of the 

 lamina in Types 16 and 35 respectively would seem to be entirely 

 different, even though their external expression is identical. 



