RELATIONSHIP OF ENDROMIS VERSICOLOR TO THE SPHINGIDES. 229 



of the larvfe, and become mere specks on the dorsal area ; each still 

 gives rise to a short, stilt" hair, but I could not see whether the hairs 

 were still notched, as it required a high power (finch), and the larva 

 would not keep still. Horn more developed and not so thick. Dorsal 

 line still distinct ; it can be traced from head to anal flap ; on most of 

 the segments it is of a dark transparent green, but where it crosses the 

 scutellum and horn it gives place to black, and there are traces of the 

 same change of colour where the line crosses the thoracic segments. The 

 yellow lateral lines are very plain, the sub -dorsal ones, which only crossed 

 the scutellum in the last skin, are now continued backwards to segt. 4. 

 Oblique stripes are much stronger, broadest on the sub-dorsal area, 

 and, though faint and narrow on the sides, they are continued down- 

 wards to the claspers. The 7th is much stronger than the rest, the 

 short one down the sides of the horn being also very plain. Anal 

 flap outlined with yellow. There is a dark border to the upper por- 

 tion of the oblique stripes, apparently caused by the encroachment 

 of the green from the sides upon the dorsal area. 



My notes on the larvae finish abruptly here ; one larva underwent 

 a third moult, but I Avas busy preparing for a Whitsun excursion to 

 the New Forest, and had not time to examine it, and ere my return 

 it had joined the majority. Nothing very definite was suggested by 

 the newly-hatched larva as regards relationship to other species. In 

 general appearance it reminded me someAvhat of a freshly-emerged 

 larva of one of the larger Bombycides, while the short horn suggested 

 that of a young larva of Lasiocaiiipa quercifoUa,ii.u([ I could detect nothing 

 which Avas specially characteristic of either Smerinthus or Notodonta. 

 A startling change, hoAveA-er, Avas produced in its appearance at the 

 first moult ; for a dark-coloured and rather hairy larva, Avhich might 

 reasonably have been expected to develop a coat like that of BouiIii/a' 

 rnhi, changed to a vi\dd green, and, to the naked eye, smooth lai'A'a. The 

 large tubercles, each bearing several hairs, Avere replaced by minute 

 black tubercles, each surmounted by one short hair slightly clavate 

 at the tip. These and other characters irresistibly reminded one of 

 Smeiinthus, although the resemblance Avas of a somcAvhat Chinese or 

 topsy-turA'y kind. Thus the oblique stripes slope in the opposite 

 direction, and are formed, not by the massing together of the 

 tubercles, as in Smerinthus, but by their being absent on the stripes, 

 and plentiful everyAvhere else. That these characters only develop 

 after the first moult Avould seem to intimate a relationship to Sphinx 

 rather than to Smerinthus. What is probably an important point of 

 resemblance is the strength of the seventh oblique stripe, as com- 

 pared AAdth the others, this feature being present in all the British 

 species of Smei-inthus, Avhile the horn on segt. 12, though short and 

 not A'ery similar in appearance, is not altogether devoid of significance. 

 Other characters common to l)oth larviB Avill be seen from the tables I 

 haA'e draAvn up. 



With regard to habits, I did not notice any that seemed to bear on 

 its relationship. When young the larv;B are gregarious, but this 

 habit is common to many Avidely-separated species (such as Clisio- 

 campa castrensis, Vanessa urticae, I'halera hucephala, etc.). They 

 rest with their heads turned back, and if disturbed they all jerk in 

 concert. The first of these habits, if my memory does not play me 

 false, is common to other gregarious larvte, and the jerking is, I 

 believe, almost universal Avith social larvre. 



