10 BULLETIN 1093, U. S. DEPARTMENT OF AGRICULTURE. 
destroyed will have expanded into new shoots, and feeding is then 
confined almost entirely to the stems of these shoots, not to the leaves 
and buds. (PI. I, Fig. 1.) In this way a large percentage of the new 
growth may be entirely destroyed, provided the infestation is suffi- 
ciently heavy. The same habit of feeding is continued through the 
later stages, until the supply of new growth is exhausted, then the 
older leaves may be attacked, and the vines stripped, provided, again, 
that the caterpillars are numerous enough to be forced to seek this 
less attractive food. 
It is obvious that a few caterpillars can do a great amount of dam- 
age, first to the terminal bud from which is produced the season's 
fruit, and second by causing a branching growth of the plant. This 
damage to the terminal bud may occur .before the owner realizes that 
his bog is infested. 
FEEDING HABITS ON WHITE OAK FOLIAGE. 
In the observations made in 1916-19 concerning the dispersion by 
wind of first-stage gipsy moth larvae, it became evident, from the 
count of larvae taken from the traps, that there are two periods of 
the day during which maximum dispersion occurs. From this fact 
it was inferred that there are, correspondingly, two daily periods 
of activity, as the larvae would be more likely to be carried by the 
wind when crawling over leaf surfaces, limbs, or trunks of trees 
than when at rest. To verify this inference it was decided to make 
a study of the habits of first-stage larvae as they are found in tree 
tops, where the earliest feeding usually appears, and where also 
the relation between the feeding habits and wind dispersion could 
be more satisf actorily studied. 
A white oak about 30 feet high was selected near the border of 
Muddy Pond Bog. Around this tree was erected a tower 16 feet 
square at the base, 8 feet square at the top, and 25 feet high. A 
flooring was placed around the branches on top of this tower, thus 
bringing the upper 5 feet of the terminal branches in position for 
close examination and observation of the movements of voung 
larvae. (Pi. IV.) 
On May 19, 1919, when observations were started, the leaves on 
this tree had begun to unfold, some of which would measure an inch 
in length, giving considerable protection to the young larvae, as 
well as supplying food. It was found that the first feeding by first- 
stage gipsy moth larvae after reaching the foliage was confined to 
the leaf hairs, principally on the underside of the leaves. After 
feeding in this manner for a day or two they began to feed on the 
tissue of the leaf, later eating through the tissue. The information 
relating to the feeding periods of first-stage gipsy moth larvae, col- 
lated for the whole period of observation, furnishes what can be 
