Dec, io, 1917 
Breeding Sweet Corn Resistant to the Corn Earworm 559 
With such small numbers little confidence can be placed in differences in 
the correlations found for the different characters. It is interesting, 
however, that damage and number of larvae have coefficients as large as 
those of morphological characters. 
Table VI .—Correlations between parents and offspring in corn experiments in iqi6 
Damage. 
Number of 
larvae. 
Damage per 
larva. 
Prolonga¬ 
tion. 
Length of 
husks. 
Length of 
ear. 
o.66±o*09 
0.72±o.og 
Grades. 
o<4o±o. 14 
Cm. 
0.68±o. 09 
Cm. 
0.72±o.o9 
Cm. 
0. s6±o. 11 
Number of 
busks. 
Number of 
layers. 
Days to 
silking. 
Husk leaves. 
Number of 
rows. 
Average. 
0.46 ±0.13 1 
0.43±o. 13 
0.6i±o. 10 
Grades. 
0.53 ±0.12 
0.77±o. 12 
0.60 
MEASURES OF INJURY 
Of the characters recorded, three were measures of injury. These 
are given under the headings “Damage,” “Damage per larva,” and 
“Number of larvae.” As might be expected from the nature of the 
characters, the three measures of injury constitute a closely correlated 
group. The damage per larva was calculated by dividing the total 
amount of damage in each progeny by the total number of larvae. The 
damage per larva is thus, of course, definitely associated with both damage 
and number of larvae. There is, however, a factor in the degree of 
damage not covered by the two other characters—that is, the number 
of ears that escape without infestation. These uninfested ears reduce 
the average damage but do not affect the damage per larva. 
Since to lessen the amount of damage is the practical object sought, 
primary consideration will be given to the relations existing between this 
character and possible protective characters. The correlations with 
number of larvae and damage per larva will be considered only as they 
may help to elucidate the correlations with damage. 
CHARACTERS CORRELATED WITH INJURY . 
In Table VII are given the interprogeny correlations of the characters 
measured. 
With 31 progenies little significance may be attached to any correla¬ 
tion that is less than 0.35, since to exceed 3 times the probable error 
the correlation must be at least 0.33. , Of the plant characters measured, 
the following showed a correlation with damage of 0.35 or closer: pro¬ 
longation, length of husks, number of layers, and days to silking. In 
addition to these, maturity and husk leaves are significantly correlated 
with number of larvae. Among the plant characters the following signifi¬ 
cant correlations appear: Prolongation with length of husks; length of 
husks with length of ear; number of husks with number of layers; 
number of layers with days to silking; days to silking with maturity, 
days silking to harvest, and husk leaves; maturity with days to harvest 
and husk leaves. 
