THE GENUS ACRONYCTA AND ITS ALLIES. 
131 
seizing strongly the loose silk of the bottom of the cocoon, and 
this is always done so effectually, that it is only with great care 
and pains that the silk can be unravelled, without damage to 
the pupal spines. The spines are arranged on the tridens 
formula, i.e., with 2 dorsal and 6 (3 on each side) ventral 
spines. They are curved so as to be hooks rather than spines, 
the dorsal curved downwards and the ventral upwards. The 
latter are, however, not so terminal as in tridens, but are spread 
round a semicircle, so that the outer ones are lateral rather 
than terminal. The, roughly speaking, round boss, on which 
the spines are situated, is separated at its base from the rest of 
the pupa by a shallow groove ventrally, but dorsally by a deep 
channel or incision, so cut into it that the basal portion of the 
boss has a thin round margin over it, with a smooth surface, 
passing down to the dorsal spines, which is somewhat shield- 
shaped. The pupal margin of the incision presents a series of 
longitudinal ribs or buttresses, four on either side, with indica- 
tion of a minute one in the intervals between them. 
When the moth emerges, it leaves very little trace of an 
opening in the outer diaphragm, but there is always a little 
wool rubbed off the moth to be observed in the opening. If 
the pupae are kept too dry, the moth is unable to force the 
outer diaphragm, either because this becomes too hard and 
dry, or because the moth has no spare fluid to soften it, the 
moth then perishes in the outer chamber. 
The larva appears to eat almost anything arboreal, is perhaps 
even more omnivorous than psi\ but I have little doubt it is 
correctly named, in so far that in the wild state it is fonder of 
alder than anything. It is very widely distributed and supposed 
to have its headquarters in the New Forest ; I fancy it is as 
abundant here as anywhere. My friend. Dr. Wood, finds a larva 
or two most years, and I have twice taken it here on birch. 
On only one occasion have I searched for it, and then Dr. 
Wood and myself (really looking ostensibly for Cerura bicuspis, 
which we did not get) each took two larvae off alder. 
Alni feeds up the most rapidly of any species I have bred. 
In each skin it occupies four days in feeding, one or two days in 
moulting, with a few extra days in the last skin. One month 
is all the time spent as a larva when the season is reasonably 
warm. 
{To be contmued) 
