126 BRITISH LEPIDOPTERA. 



It may be well here to mention that, in our opinion, no more un- 

 satisfactory suggestion for the origin of the Bombycid (Lasiocampid 

 and Saturniid) stirps has been brought forward, than that propounded 

 by Packard, who attempts to derive it from the Lithosiids. The 

 latter form a highly specialised Arctiid group, differing from what 

 Packard terms the Bombycids in egg, larval, pupal and imaginal 

 characters as widely as can be. We have shown that the tendency 

 of this stirps is for the larval tubercle i to be enlarged, and tubercle ii to 

 become atrophied. In the Liparids, Lithosiids and Arctiids, on the 

 other hand, tubercle ii tends to be enlarged, whilst i becomes 

 atrophied, e.g., the larvae of Lithosia complana, Oallimorpha liera, etc. 



Before leaving our consideration of the larvae of this stirps, it may 

 be well to recall attention to the fact that, in this stirps, there are traces 

 of extra prolegs on the abdominal segments. We have already referred 

 to those in Lagoa crispata, Chrysopyga unduluta , and other Megalopygids, 

 also to those of the PterOphorid, Pselnuphorus brachydactylus. It is 

 remarkable that Kowalewski found ten pairs of abdominal prolegs in 

 the embryo of Sphinx, one half of which disappeared before hatching, 

 thus leaving the normal number. Tichomiroff found prolegs developed 

 in the embryo of Bombyx mori on the abdominal segments 2-10, but 

 those on segments 2, 7, 8 and 9 were absorbed again in a later 

 embryonic stage, and Graber notices that on all the abdominal 

 segments, except the 9th and 10th, of the early embryos of the same 

 species, faintly marked knob-like elevations are to be seen, which may 

 be considered as the first indications of rudimentary appendages. The 

 same author considers that in B. mori " the stage of pantopody has 

 only a very ephemeral duration." What value these ephemeral 

 structures have in Bombycid and Sphingid embryos, and what meaning 

 is to be attached to them is not quite clear. We only draw attention 

 to the fact that they appear in the embryos of two specialised super- 

 families, where probably homologous structures still occur normally in 

 the larvae of one of the most generalised superfamilies of the same 

 stirps. It may be noticed here that the mining larva? of the Nepticulids 

 have nine pairs of abdominal prolegs that bear no terminal crochets or 

 hooks. 



The sluggish habits of the larvae of many of the superfamilies in- 

 cluded in this stirps, are probably due to their large size, to the 

 highly developed condition of the prolegs, and to the complex mecha- 

 nism by means of which walking is accomplished. Anyone who has 

 watched a Saturniid or Sphingid larva walk knows that it cannot 

 hurry. It is probably on account of this slow method of movement 

 that the various larvae are so remarkably protected by spines or hairs, 

 some of which are of a most complex character. We are inclined to 

 connect these sluggish habits with an arboreal {i.e., as opposed to a 

 ground-feeding) habitat, and have already shown how the larva? of 

 the large Geometrid superfamily have met the difficulty by protective 

 coloration and a decrease in the number of prolegs, which gives 

 them greater speed and lessens the necessity of other special defensive 

 structures, and yet have maintained their arboreal environment. 

 Packard also connects these sluggish habits with an arboreal condi- 

 tion of life, and asserts that such larva? are surrounded by a purer ah?, 



* Bombyciue Motlis of America, pp. 32 and 83, 



