The InJi< ritance of Quantitative Chmmcters in Maize 83 
ably a mere expression of the effect of the seasonal differences 
in weather, accelerating the growth of the early plants and re- 
tarding that of the late ones. If, however, this correlation were 
genetic, or even merely physiological, so that, without respect 
to seasonal weather changes, the early plants tended to have 
longer internodes than the late plants, then the weather actually 
experienced in 1910 would have tended to make the long inter 
nodes longer and the short ones shorter, thus tending to extend 
the range of variation of F 2 while lessening that of the parents. 
But it is difficult to see how there can be any physiological cor- 
relation between earliness and long internodes. And it is even 
more difficult to believe in genetic correlation betAveen earliness 
and long internodes when it is remembered that in case of the 
parents the early one has the shorter internodes. On the whole, 
therefore, it seems very probable that the production of inter 
node lengths in F 2 both above the upper and below the lower 
extremes of the parents was due in part at least to new combina 
tions of inter node-length factors. 
If the above is the correct explanation of the very long and 
the very short inter node types that appeared in F 2 family No. 
510, it should be possible to isolate strains from it that have 
longer or shorter internodes than do the parent varieties. The 
selection of F 2 plants to be tested by F 3 progenies was based 
upon height of stalk and upon earliness rather than upon inter- 
node length, and it happened that no F 2 plant with extremely 
short or extremely long internodes was chosen. The F 3 fam- 
ilies grown in 1911 are arranged in Table 33 in order of the 
internode length of the F 2 plants of which they are the progenies 
as indicated by black-faced figures in the F 3 arrays. The exact 
reverse in seasonal weather conditions between 1910, when in 
Nebraska the early summer was favorable for growth and the 
late summer unfavorable, and 1911, when in Massachusetts the 
early summer was unfavorable and the late summer very favor- 
able, made it impossible to determine from a single season's 
study of the F 3 families whether types more extreme than the 
parents could be isolated. By reference to Table 33 and to 
Table 39, which latter contains the data for earliness of these same 
families, it will be noted that, with the marked exception of No. 
1 1.37, there was a strong tendency for the early F 3 families to 
have much shorter internodes and the late families much longer 
internodes than the F 2 plants of which they were the progenies. 
In short the early F 2 plants had longer and their F 3 progenies 
shorter internodes while the late F 2 plants had shorter and their 
F 3 progenies longer internodes than would have been the case 
had the weather been more nearly uniform thruout the two 
seasons. 
