76 



THE AMERICAN NATURALIST [Vol. XLIV 



greater than ± 2 rows. If conditions are more favorable 

 at the time when the upper ear is laid down it will have 

 two more rows than the second ear; if conditions are 

 favorable all through the season, the ears generally have 

 the same number of rows ; while if conditions are unfavor- 

 able when the upper ear is laid down, the lower ear may 

 have two more rows than the upper ear. Furthermore, 

 seeds from the same ear have several times been grown 

 on different soils and in different seasons, and in each 

 case the frequency distribution has been the same. Hence 

 it may be concluded that in the great majority of cases 

 fluctuation is not greater than in ± 2 rows, although fluc- 

 tuations of ± 4 rows have been found. 



A second question worthy of consideration is: Do 

 somatic variations due to varying conditions during de- 

 velopment take place with equal frequency in individuals 

 with a large number of rows and in individuals with a 

 small number of rows ! From the fact that several of my 

 inbred strains that have been selected for three genera- 

 tions for a constant number of rows, increase directly in 

 variability as the number of rows increases, the question 

 should probably be answered in the negative. This 

 answer is reasonable upon other grounds. The eight- 

 rowed ear may vary in any one of four spikes, the sixteen- 

 rowed ear may vary in any one of eight spikes ; therefore 

 the sixteen-rowed ear may vary twice as often as the 

 eight-rowed ear. By the same reasoning, the sixteen- 

 rowed ear may sometimes throw fluctuations twice as 

 wide as the eight-rowed ear. 



A third consideration is the possibility of increased 

 fluctuation due to hybridization. Shull 8 and East have 



when maize biotypes are crossed — a phenomenon apart 

 from inheritance. There is no evidence, however, that 



