14 
BL'LLETIX 204, U. S. DEPARTMENT OF AGRICULTURE. 
in order to establish the essential facts concerning the identity of 
the organism and the conditions most favorable for its increase and 
development. 
FOOD-PLANT INVESTIGATIONS. 
As has already been pointed out. the species of tree growth have 
an important relation to the ability of the gipsy moth to increase 
and cause serious damage in the field. Nearly 20 years ago food- 
plant experiments were conducted in this country to determine the 
species upon which the gipsy moth would subsist, and a long list of 
food plants was published by Forbush and Fernald in then excellent 
book on the gipsy moth. Most of these experiments, however, were 
carried on by using large caterpillars and feeding them in jars or cages 
hi the laboratory. At that time it did not seem important to deter- 
mine whether there was variation in the feeding habits of the cater- 
pillars in different stages. As early as 1908 it was observed and 
proven by an extensive field experiment that the first-stage cater- 
pillars of the gipsy moth could not develop on white pine in the 
absence of other food. This naturally led to the question of unfavor- 
ability of other species to gipsy moth attack. In 1912 a careful 
series of food-plant experiments was begun. Mr. F. H. Mosher took 
charge of this work and has been furnished with a number of assist- 
ants during the feeding season. The feeding was carried on in indi- 
vidual trays, which were specially constructed for the purpose. 
One hundred first-stage caterpillars were placed upon a branch of 
foliage in each tray. In this way the feeding habits could be observed, 
the foliage renewed daily, and it was possible to determine which 
food plants were least subject to attack under laboratory conditions. 
Similar experiments were carried on with caterpillars in the succeed- 
ing stages. This work was continued in the summers of 1913 and 
1914, so that up to the present time about 250 species of trees and 
shrubs have been tested. 
As a result of these experiments some of the more common species 
are rated as follow- : 
I. — Species favored by the gipsy-motn larvae in all stages: 
Alder, speckled. 1 
Ash, mountain. 1 
Aspen. 1 
Balm of Gilead. 1 
Basswood. 
Beech. 
Birch, gray. 1 
Birch, paper. 
Birch, red. 1 
Boxelder. 1 
Larch. 
Oak, black. 
Oak, chestnut. 
Oak, post. 1 
Oak, red. 
Oak, scarlet. 
Oak, swamp white. 
Oak, white. 
Poplar, big toothed. 
Shadbush. 1 
Willow. 1 
AVitch hazel. 1 
Species of low commercial value. 
