13 
Certain elms of the species U. campestris and other species which were 
overpoisoned, and shed most of their leaves in consequence in the last 
of June, 1883, sent out a profuse new growth of leaves and twigs. The 
foliage fell gradually for three weeks, and this was somewhat promoted 
by the succeeding rains. 
The larvee move from place to place so seldom that if the leaves are 
imperfectly poisoned from the mixture being weakly diluted, or from its 
application only in large scattered drops, which are much avoided by 
the larve, they are not killed off thoroughly for several days, and in all 
vases it requires considerable time to attain the full effect of the poison. 
This result appears on the plant and on the insect. After each rain 
the poison takes a new effect upon the plant and the pest, which indi- 
sates that the poison is absorbed more, or is more active when wet, and 
thatit acts by dehydrating thereafter. Where the tree is too strongly 
poisoned, each rain causes a new lot of leaves to become discolored by 
the poison or to fall. On some of the trees the discoloration appears in 
brown, dead blotches on the foliage, chiefly about the gnawed places 
and margins, while in other instances many of the leaves turn yellow, 
and others fall without change of color. The latter may not all drop 
from the effects of poison, but the coloration referred to is without doubt 
generally from the caustic action. The poison not only produces the 
local effects from contact action on the parts touched by it, but follow- 
ing this there appears a more general effect, manifest in that all the 
foliage appears to lose, to some extent, its freshness and vitality. This 
secondary influence is probably from poisoning of the sap in a moderate 
degree. When this is once observable, no leaf-eater thrives upon the 
foliage. Slight overpoisoning seems to have a tonic or invigorating 
effect on the tree. 
Preventive Effects of the Poison.—In this grove the elms that were poi- 
soned in 1882 were attacked in the spring of 1885 less severely than were 
those which were not poisoned the previous year. This would seem to 
imply that the insects deposit mostly on the trees nearest to where they 
develop, and are only partially migratory before ovipositing. The at- 
tack afterward became increased, probably by immigration and the new 
generation, so that later in the season the trees were mostly infested to 
the usual extent. 
In the region of Washington a preventive application of poison should 
be made before the last of May or first of June, when the eggs are being 
deposited and before they hatch. This will prevent the worms from 
ever getting a start. By the preventive method the tree escapes two 
kinds of injury: first, that directly from the eating by the insect; second, 
that which follows indirectly from the deleterious effects of the poison 
on the plant, for its caustic effect is much greater where the leaves have 
been so gnawed that the poison comes in contact with the sap. 
Treatment with London Purple.—Already early in June the insect ap- 
pears plentiful, On June 7, 1882, it was at work on all the trees, and 
