HETEROGENEA CRUCIATA. 381 



series of specimens showing the different stages iu the process of extru- 

 sion, it is equally clear, in fact quite certain, that the spines before 

 hatching are invaginated into the interior of the larva, precisely as in C. 

 avellana, and are extruded and assume their exterior position shortly after 

 the larva leaves the eggshell, in the manner thus described in the case of 

 the latter species. There is, however, a very important difference in 

 the structure of the dorsal spines, viz., each one has two branches 

 instead of being simple. As it evaginates at first a simple straight 

 portion protrudes, but is seen to enclose not one but two terminal 

 portions, the portions that from the first appear to be stiff and har- 

 dened ; then the soft evaginating portion divides into an anterior 

 and posterior horn, separating from each other at an angle of 80°-90°, 

 and the hard terminal portions, at first parallel to each other, cross 

 one another at an angle, and finally form the extremities of these two 

 branches. There is in G. avellana a short process about half way up 

 the soft (invaginated) portion of the spine, that seems to have no use 

 or meaning, but is probably the representative of the second spine in 

 H. cruciata (asella). The principal one of these, that is the longer and 

 rather thicker one, inclines slightly backwards, and is altogether of a 

 length about equal to f of the diameter of the larva, or about 0*13 mm. 

 The shorter and rather more slender one, which might be regarded as 

 a branch of the other, but which is more nearly equal than that 

 description would imply, points decidedly forwards. The larger branch 

 terminates in a three-spined point, and has several very minute points 

 on its stem ; the smaller one terminates usually in two points. The 

 lateral spines appear to be simple (as regards branching), and expand 

 terminally into a three-spiked coronet (Chapman). 



Variation of larva. — The larva varies very much. Its ground colour 

 is sometimes yellowish instead of green, and the cross on the back light 

 red, flesh-coloured, or yellow tinged with red on the edges. Occasionally 

 it is interrupted by the ground colour (Borkhausen). 



Cocoon.— In confinement the cocoon is sometimes placed on a dried 

 beech leaf, but is more frequently spun very tightly into the forks of 

 beech twigs, and Hewett states that neither he nor Tate could ever 

 find them in the wilds. It is about a quarter of an inch long, two lines 

 broad, and of a very short elliptical form. It is firm in texture, and 

 bears a remarkable resemblance to a gall excrescence. A few fine 

 threads form a kind of network around its base, and attach it to the 

 leaf. It is of a dark, dull brown colour, with blotches of a pale grey 

 or dirty white, spreading irregularly over the upper surface. Buckler 

 says it looks as if it bore a delicate lichenous growth. Fletcher 

 observes that this mottling accurately matches the beech bark. The 

 cocoon is lined inside with pale yellow silk. The larva remains 

 unchanged in the cocoon from October until the following May or 

 June, the pupal stage lasting only about a fortnight. 



Pupa. — Structurally, the pupa is very like that of Gochlidion 

 avellana (testudo), but the eye-collar is remarkably well-developed, 

 stretching right across from the antenna? to the mouth-parts, and on 

 dehiscence it remains attached to the head coverings. Buckler de- 

 scribes it as a trifle less than a quarter of an inch long, and thick in 

 proportion, the abdomen bent under, giving the pupa a rather rounded 

 form ; the wings and appendage cases, not attached to the abdominal 

 segments beyond the second ; the abdominal segments distinct, as also 



