226 Dr. Wallace on 
in such a case if the autumn temperature be warm, probably many 
of the early cocoons will produce moths in September and October, 
whose produce will be valueless, as the eggs cannot be kept alive 
through the winter. 
The cocoon (see Pl. XV. ) is of an elongated flask-shaped 
form, of a colour more or less pale grey, verging into light grey- 
brown, but occasionally of a rusty brown, especially in the earlier 
and later cocoons; of very close texture, from 13 inch to 2% inches 
long and about 3? of an inch broad, in girth 24 inches, varying in 
size and weight according to the health and size of the larva; 
hanging pendent from the leaf-stalk they frequently measure 4 
inches from the stalk to the lower extremity of the envelope. 
There are two classes of cocoons which furnish silk: one the 
closed cocoon, of which the type is that of the B. Mori, spun 
equally all round, requiring to be unwound before the moth 
emerges; otherwise the moth emits a fluid from its mouth for the 
purpose of dissolving the gum which binds the silken threads to- 
gether, and makes an exit between the diverging threads, and the 
cocoon, being left with a hole in it, becomes useless, as, in the 
common method of reeling, the hot water enters the cocoon at the 
hole, renders it too weighty to revolve, and the thread breaks 
during the process. 
But the second class of cocoons, which is a very extensive one, 
and to which B. Cynthia belongs, has hitherto been much neglected 
by the reeler, because of his ignorance how to manipulate success- 
fully cocoons with a hole in them. These cocoons are so constructed 
that an aperture is left at the apex for the exit of the moth, and 
are therefore bottle- or flask-shaped cocoons, ‘“ naturellement 
ouverts ;” the threads at the apex, which form the exit hole, are not 
broken ‘off by the larva in the process of spinning, but are doubled 
back in a continuous thread ; the larva may be observed during 
the process stretching upwards the head, fastening a thread and 
retreating with it as far inside as he can stretch, thus doubling and 
redoubling the thread by a series of loops. These loops are set 
in layers, of which there are several formed at different periods of 
the work, corresponding to the different skins; they may be dis- 
sected out of the cocoon, when it will be found that being in a con- 
tinuous thread, there is no real impossibility, but only a mechanical 
difficulty presented to the reeler who wishes to wind these cocoons. 
Further it will be noticed that these layers of looped threads are 
so grouped one inside another as to present a very great obstacle 
to the entrance of an enemy, e. g., an earwig or a tomtit’s beak ; 
they are not only very strong, but being folded inwards and con- 
