BIOLOGY OF THE LOTUS BORER. ah 
aquatic environment, is well taken, for many leaf and tree-feeding 
caterpillars, exposed to the same or to greater risks, do not construct 
protective webs and, on the other hand, many larve living in 
sheltered and well-protected situations do build burrows or webs— 
within which to work. Be that as it may, there is probably little 
danger of wind or waves washing these caterpillars off the leaves. 
DIPTEROUS PARASITES. 
From living enemies, however, they do not escape so easily. <Ac- 
cording to the literature four species of tachinid flies have been 
reared from the larve. Townsend (5) lists Lxorista hirsuta. O. S. 
(now £&. vulgaris Fall.) and Pherocera comstocku Will. as having 
been reared by Forbes in Illinois. Coquillett (7, p. 77, 19) adds 
Hypostena variabilis Coq. and Panzeria penitalis Coq. to this list. 
Because of the confusion between Pyrausta penitalis and Pyrausta 
ainsliei and the impossibility of finding from the hterature the ex- 
act source of the material from which the parasites were reared, it 
is possible,that not all of these species attack the true Pyrausta 
penitalis. -Panzeria penitalis, for one. is known to be a parasite of 
Pyrausta ainsliet, and there is no definite record of its ever having 
been reared from lotus-feeding larve. All of these records should 
be verified in the light of our more exact knowledge of their hosts. 
HYMENOPTEROUS PARASITES. 
Among the hymenopterous parasites, Bracon xanthostigma Cress. 
is listed by Riley and Howard (4, ». 439) as a parasite of this borer on 
lotus at St. Louis, Mo. Viereck (77, p. 223) lists Meteorus communis 
Cress. with the simple statement that it parasitized P. penitalis. Hart 
(6, p. 180) mentions one secondary and two primary parasites, but 
without determinations. We recognize at least one of his parasites, 
the braconid, making white cocoons singly on the leaf surface. (PI. 
III A.) This one, determined as Apanteles harti Vier. by A. B. 
- Gahan, was found to be the most common at Kimberlin Heights. Its 
small white cocoons (1.5 by 4 millimeters) are firmly fastened to the 
leaf disk either under the webbing or exposed outside. In July they 
were only occasionally seen, but by early September were much more 
common and were killing from 10 to 25 per cent of the larve. This 
species evidently attacks the smaller larvee and completes its hfe as a 
parasite when its host is scarcely more than half grown. None of 
them developed from larve larger than this. From 7 to 10 days 
elapsed between the spinning of the cocoon by the parasite and the 
emergence of the adult. 
Another parasite, very likely the other mentioned by Hart (6, p. 
181), which the writers found only a little less common than the fore- 
_ going, is an undescribed, yellowish brown species of Microbracon (de- 
