( xxvi ) 
the proof consists in finding, in plants with many larvae of 
H. hyerana and only one or two of the fly, not a few larvae of 
H. hyerana^ dead and blackened, but only partially collapsed, 
the result of being killed by a larva of X. comtus much too 
small to do more than suck a portion of its juices. In 
captivity I have several times observed that the attacked 
larva dies very soon after it has been seized by that of the fly. 
These circumstances prove that a larva of X. coinUis could not 
possibly, immediately it hatched, attack a larva enough to 
contain a life ration for it, and yet have that ration remain 
fresh during the whole period of its growth. 
I ought, perhaps, to dwell on this point, as it is one that 
might have occurred to the professor. Doubtless he had in 
mind the habits of the fossorial wasps, whose prey remains 
fresh for a very considerable time, fresh because it remains alive, 
though paralysed. The rapid death of its prey and its subse¬ 
quent I’apid decomposition quite difPerentiate the case of X. 
comtus from that of the wasps. 
Mr. Lyle’s larva had no doubt already disposed of, at least, 
several larvae of the Scotosia. 
Then as to S. unchdata, several points deserve notice. One 
is that its full-grown larva is of much the same bulk as that 
of H. hyerana, and of H. hyerana a nearly full-grown larva of 
X. comtus will attack and demolish a full-grown specimen, 
and even more than one. 
Perhaps the most important point in connection with S. 
undidata is that the larva is gregarious in precisely the same 
sense as the larvae of the Tortrices I have observed in this 
connection are. Whether in their younger stages they are or 
are not literally so, I do not know, but think probably they are. 
At any rate, it is the case that several or many larvae are 
always to be found on the same bush; doubtless they have all 
emerged from one batch of eggs, as in the case of the Toi trices. 
The only difference, then, that exists between the cases of 
S. undulata and that of the Tortrices, on which I have seen 
A', comtus parasitic, is that they are Tortrices and it is a 
Geometer, a condition, I should imagine, of infinite indifference 
to the larva of X. comtus. A further point is not, however, 
without interest and of some direct bearing on the subject, 
