104 
Research Bid let in No. 2 
families. Nor was this difference in size the only feature differ- 
entiating- the several P 3 families. The variability of some F 3 lots 
was no greater than that of the parents or of F x , while other F 3 
lots equaled in variability the F 2 generation. 
Inheritance of diameter of ears was investigated in two 
crosses. The ears of F 1 were intermediate in diameter between 
those of the parents, but somewhat nearer the large-eared than 
the small-eared variety. The F 2 ranges of variation little more 
than tilled the gap between the parent races. The parent types 
were not recovered in F 3 , but the several F 3 lots were very dis- 
tinct. The smallest ears of the largest eared F 3 families were 
larger than the largest ears of the smallest-eared family. 
Weight of seeds was determined only for the cross Tom 
Thumb X Black Mexican. The seeds taken to represent the 
latter variety were probably too large, owing to previous mixture 
of strains. If allowance is made for this, the F 1 seed weight was 
only a little below the mid-parental weight and the ranges of 
variation of the F 2 fraternities well overlapped the ranges of the 
parents. Tho distinct types of seed weight were isolated in F 3 
neither parent type was recovered. One F 3 family, however, had 
a seed weight nearly equal to that of Black Mexican and the 
great variability of several F 3 lots indicated heterozygosity suffi- 
cient to enable the isolation of the parent types on further 
selection. 
Breadth of seeds was noted in the crosses of Missouri dent 
with California pop and with Tom Thumb. In both cases ¥ 1 
was distinctly intermediate between the parents in breadth of 
seeds. The variability of the F a lots was not noticeably greater 
than that of the parents. The F 2 generation, on the contrary, 
showed a markedly greater variability than F, or the parents. 
In one cross the combined range of variation of all the F 2 families 
together was not sufficient to overlap even the inner extremes of 
the parents, but in the other cross, while the F 2 ranges were not 
greater, the parents differed less in seed size, so that practically 
all the classes from below the mean of the small-seeded parent to 
above the mean of the large-seeded parent were occupied by F 2 
individuals. Two of the F 3 lots of the latter cross, grown from 
small-seeded F 2 individuals, had seeds practically as small as 
those of the small-seeded parent. Since the total number of F 2 
plants was slightly over 300 in each cross, it is thought that 
Missouri dent and Tom Thumb pop probably differed by not over 
five factors and Missouri dent and California pop by perhaps as 
many as six factors influencing breadth of seeds. 
Four different crosses were employed in the studies of the 
inheritance of height of plants. A peculiarity of these crosses 
