Dec. io t 1917 Breeding Sweet Corn Resistant to the Corn Earworm 565 
The difference between the intra- and inter-progeny regression re¬ 
mains unexplained, and in the light of this disparity it should be kept 
in mind, no asurance can be given that an increase in the prolongation 
in other stocks will be followed by the same rapid increase in immunity 
found in the course of these experiments. 
LENGTH OF HUSKS 
Length of husks and prolongation are measurements of nearly the 
same thing. Prolongation, however, may increase in either of two ways. 
The husks may be longer or the ear may be shorter. The first selec¬ 
tions were made for prolongation without special regard to the length 
of the ear, and it was feared that in so doing there might have been a 
loss in the length of the ear. Fortunately prolongation is more closely 
correlated with length of husk than with length of ear. The loss in 
length of ear has not been material, and the worm-resistant strains 
have a satisfactory ear length. The average for the different progenies 
ranges from 15.2 to 26.7 cm. Since little is to be gained by reducing 
the damage at the expense of the length of ear, it would probably be 
safer in future work to use length of husks as a basis of selection than to 
rely on the prolongation. 
HUSK LEAVES % 
The correlation between damage and husk leaves is 0.31. Since husk 
leaves afford additional surface on which moths can deposit eggs the 
larvae of which may gain access to the ear, a positive correlation would be 
expected. On the other hand, eggs so deposited are to some extent at 
the expense of eggs which in the absence of husk leaves would be deposited 
on the silks. Larvae hatching on the husk leaves would be somewhat 
delayed in reaching the ear. These larvae might be expected to do less 
damage than those hatching on the silks, and for this reason the 
damage per larva should be negatively correlated with husk leaves. 
That the husk leaves do attract the moths or at least afford a location 
for the eggs is indicated by the positive correlation of husk leaves with 
number of larvae, 0.52. The second assumption of an opposite relation 
with damage per larva does not appear in the direct correlation, which is 
also positive though only 0.12. It will be shown, however, that in gen¬ 
eral as the number of larvae increase the damage per larva also increases, 
and the partial correlation of husk leaves with damage per larva for con¬ 
stant number of larvae is, in fact, negative, — 0.32. This makes it appear 
that there is also support for the view that husk leaves tend on the average 
to reduce the damage done by each larva. 
The final results are therefore in accord with the supposition that the 
manner in which husk leaves increase the damage is through providing 
additional opportunities for the moth to deposit eggs near the tip of the 
ear. 
