148 MODERN CLASSIFICATION OF INSECTS. 
appearance until after they have become perfect insects. Ratzeburg, 
in Nova Acta (vol. xvi. pl. 9.), has figured the larva and pupa of He- 
miteles luteator, with the view of showing that the head of the pupa 
is formed from the two first segments of the larva, representing the 
eye of the pupa as visible through the skin of the second segment of 
thelarva. In my unpublished memoir upon Trichiosoma lucorum 
and its parasites, observed ten years ago, I had not only noticed this 
circumstance, but had observed that the fifth and following segment 
of the larva, immediately before assuming the pupa state, is contracted ; 
but on carefully examining the insect in this state, it is evident that it 
is owing to the increased development of the head of the inclosed pupa, 
that an alteration has taken place in the relative position of the ante- 
rior segments, which cannot be considered as a proof that the head of 
the pupa had occupied two segments instead of one. My figure 76. 
13. represents another larva (preserved in spirits), ready to undergo its 
transformation to the pupa, showing the head of the latter disengaged 
within the skin of the larva, and occupying the second segment of the 
body, the antennze extending into the first segment. 
The larve of the majority of these insects spin a silken cocoon, in which 
they pass the pupa state; these cocoons are sometimes found in a mass 
together, and are often enveloped in a general covering of yellow or white 
glossy silk (which has been incorrectly stated by some authors to be 
spun by the dying caterpillar for the covering of its parasites, although 
it ordinarily spins but a few threads for supporting itself by a girdle round 
the body ), forming a large oval mass, whichmay often be observed upon 
railings, walls, stems of plants, &c. Such are the habits of Microgas- 
ter glomeratus /?dse/, and some allied species (Réaumur, Mém. tom. ii. 
tab. 33. and 35. fig. 1, 2.6.; and T. H. [Gen. Hardwicke], in Mag. 
Nat. Hist. No. 11. incorrectly named Ichn. ovulorum). In one spe- 
cies the cocoons are arranged in a symmetrical manner, like the cells 
in a miniature bee-hive, without any external covering (Réaumur, Mém, 
tom. ii. pl. 35. f. 7, 8., representing the cocoons of Microgaster alvea- 
rius). My figure 76.17. represents a mass of fifty-two of these cocoons 
arranged symmetrically, discovered by Mr. Ingpen, which appeared 
to have been attached on one side to a sprig, and from which the 
insects had escaped both at the top and bottom, by cutting open the 
circular lid ofeach cocoon. In aspecies of Cryptus, of which I have 
observed the transformations, the cocoon is long, cylindric, and rounded 
at each end; and I distinctly noticed that it was composed of three 
separate layers, the two interior being very shining, and of a gummy 
