3o8 
Journal of Agricultural Research 
Vol. XXV, No. 7 
In each of the three groups of stature, tall plants are found associated 
with long leaves, a relation to be expected from the standpoint both of 
physiology and genetics. It is of interest, however, that the relation¬ 
ship is no closer, especially in the group of normal stature. 
With leaf width, height is found to be negatively correlated in the 
brachytic group, while the coefficients of correlation in the other two 
groups are essentially the same as for leaf length. The negative correla- 
Fig. 7.—Frequency distribution for total number of leaves. Shaded portion, dwarf plants; solid lines, 
brachytic plants; broken lines, normal plants. 
tion in the brachytic group indicates that some sort of segregation of the 
dwarf stature and wide leaves is taking place in this group, since a posi¬ 
tive correlation would be expected if the association were due to physi¬ 
ological causes. The correlation with leaf index indicates that the 
leaves were broad, not only absolutely but relatively, this fact being in 
the nature of a substantiation of the hypothesis that some of the plants 
classed as brachytic were potentially dwarfs with respect to these char¬ 
acters. 
