562 
Journal of Agricultural Research 
Vol. XI, No. ii 
small larvae are frequently found among the silks inside the husks with 
no indication of having eaten their way there, it is inferred that 
these small worms can crawl down the silks to the tip of the ear. 
Larvae hatched on other parts of the plant would on the average be 
larger when they reach the silks and must eat their way to the ear. 
Against these larger larvae a greater prolongation should prove an 
adequate protection, but they are a comparatively small factor in 
the total damage, as their depredations are largely confined to the tip 
of the ear. 
In both 1915 and 1916 an attempt was made to record for each ear 
whether the larvae entered the ear by traveling down the silks or by 
eating through the husks. Of the 1,449 earworms found in the dam¬ 
aged ears 1,384 were recorded as having entered from the end and only* 
65, or 4.5 per cent, by penetrating the husks. The holes so frequently 
,observed in the husks, the presence of which led to the belief that thick, 
harsh husks would afford protection, are explained as being made by 
the larvae when emerging. The few larvae which gain access to the ear 
by eating through the husks do not leave a continuous track, for the 
husks of a young ear are elongating at different rates and the continuity 
of the hole left by the larva is soon broken. 
From the fact that in the compacted silks the larvae would be close 
together and the debris left by one larva might conceal other small larvae, 
it was thought that there would perhaps be a greater tendency to overlook 
larvae in the ears with great prolongation with the result that the number 
of larvae in such ears would be underestimated. If errors of this kind 
were of sufficient magnitude, they might account for the correlation 
between prolongation and number of larvae. 
As a check against errors in counting the number of larvae, the per¬ 
centage of ears that had no larvae was correlated with prolongation. 
Whatever may be the difficulties in counting the larvae, whether an ear 
contains larvae or not is a fact easily observed, and in dividing the ears 
into those with and those without larvae there would be no tendency to 
overlook the presence of larvae in ears with the greater prolongation. 
The correlation between prolongation and the percentage of ears with 
larvae was found to be —0.59, a very close agreement with the —0.60 
correlation between number of larvae and prolongation. 
If the relation between prolongation and damage as measured by the 
interprogeny regression is one of cause and effect—that is, if each increase 
of a centimeter actually reduces the damage by 1 per cent, the same rela¬ 
tion should hold among the individual plants of a progeny. In other 
words, the intraprogeny regression should be the same as the interprogeny 
regression. The average intraprogeny correlation of damage and pro¬ 
longation, which seems the best expression we have for the relations 
