150 
Journal of Agricultural Research 
Vol. XXIX, No. 3 
time. In the fourth stage, which lasts 
from 2 to 6 days, they are usually 7 to 
9 mm. in length. The fifth stage also 
requires from 2 to 6 days and is con¬ 
cluded by the fifth molt just prior to 
pupation; the larvae in this stage are 
from 14 to 18 mm. long, but when 
full-grown they are about 18 mm. long 
(PI. 2, D, E). Observations based on 
171 larvae representing five genera¬ 
tions indicate that it requires from 15 
to 30 days, with an average of about 21 
days (see Table VI), to complete the 
larval development. 
Table VI .—Duration of larval period 
of Phlyctaenia rubigalis 
Generation 
Number 
of larvae 
Duration of larval 
period 
Minimum 
Maximum 
Second.. _ 
25 
Days 
22 
Days 
30 
Third_ 
82 
17 
28 
Fourth_ 
2 
17 
21 
Eighth.. _ 
36 
18 
23 
Ninth_ 
26 
15 
20 
Larval coloration varies with the 
kind of plant on which the larva is 
feeding. On cinerarias the larvae are 
more yellowish-green than on marguer¬ 
ites, and on primulas they acquire a 
bright-green color; but the relation is 
most pronounced in the case of coleus 
which gives them a dull reddish-brown 
appearance. 
Feeding Habits 
The larvae upon hatching show more 
or less activity and move about on the 
surface of the foliage, apparently in 
search of an abrasion or place to begin 
feeding; they remain, however, grouped 
on the leaf upon which eggs were de¬ 
posited and from which they hatched. 
They usually feed on the lower side of 
the leaf, or in the rolled leaf, or in leaves 
webbed together for their protection. 
When feeding their heads are bent at 
almost right angles to the body and 
small pieces of the leaf are gnawed out 
from circular holes which are enlarged 
,as the larvae continue to feed along the 
edges. In the later feeding the small 
areas coalesce, resulting in the destruc¬ 
tion of the entire leaf surface upon 
which they are feeding. There is a 
marked change in the size of the larvae 
between the time of hatching and a 
few hours after feeding. In some in¬ 
stances they become at least four times 
their original size. As newly hatched 
larvae they apparently have some 
difficulty in beginning a feeding area, 
and it has been noticed that they tug, 
tear, pull, or dig at the leaf surface in 
order to break through the tissues, but 
whenever an abrasion occurs it is quite 
promptly made their point of attack. 
The exact type of injury varies some¬ 
what in the case of different host 
plants. On cinerarias and marguerites- 
they skeletonize the leaves, usually 
feeding on the lower surface, and on 
the latter they also draw the leaves 
together, while on the rose they appear 
to chew small pieces of the under 
surface of the leaves. They ate small 
round holes through violet leaves or 
else they fed on the under surface, 
leaving only the epidermis of the upper 
surface intact, giving it a blotched ap¬ 
pearance. Even though the rose is 
reported as one of its favorite hosts in 
some sections, several attempts to rear 
larvae on it in the laboratory failed 
since they would invariably spin silken 
strands to lower themselves from the 
plant, and crawl away. 
During periods of rest the trans¬ 
parency of the body tissues permitted 
observations on the complete process 
of digestion through the alimentary 
canal. The young larvae when dis¬ 
turbed suspend themselves from the leaf 
by spinning a silken strand on which they 
sway backward and forward, clinging 
to any object which they may en¬ 
counter. They are able to cross gaps 
from one leaf surface to another by 
stretching or reaching across, resting 
on the prolegs during the operation. 
Feeding has been observed on the sides 
of the veins and tips of the leaves, and 
at times they may eat a narrow strip 
along the under side of the leaf and 
afterward draw it together with web¬ 
bing. 
Web Making 
As a general rule web making or 
spinning does not take place until after 
the larvae begin feeding. In the case 
of a larva hatched four hours before, 
the following procedure was observed: 
The anterior portion of the body swayed 
backward and forward, evidently in 
search of a resting place for the legs 
while hanging on with the prolegs, the 
head and first anterior segment moving 
together, with only slight motion of the 
second and third segments. Since 
strands of silk were noticed emanating 
from the mouth, these actions were 
apparently associated with web mak¬ 
ing. The first pair of legs was used 
in attaching webbing to the tips of the 
leaf hairs. A single strand was first 
made, then several others were joined 
to it at a central point to form a web 
surface, similar to a spider web. As 
soon as a small area of the leaf surface 
had been covered with webbing to 
afford protection and support, the 
larva devoured a portion of the leaf 
tissue equal to the size of its head. It 
