40 BRITISH LEPIDOPTEBA. 



development of the dorsum of the pro-thorax in the larva of Cemra ap* 

 pears to have no phylogenetic significance, nor any close connection 

 with the chitinous pro-thoracic shields of boring larvae. It is certainly 

 smooth and shining, but appears to have been modified independently, 

 for protective purposes, in this particular genus. Still, its probable use 

 for the moulding of its hard cocoon must not be altogether overlooked. 

 The anal segment has caused much discussion as to its structure 

 and homologies, especially with regard to the suranal plate, the 

 infra-anal lobe, the paranal lobes and the paranal tubercles. The 

 supra-anal, or "suranal," plate of Packard, is the "podex" of 

 Kirby and Spence, and both in its shape and ornamentation would 

 appear, especially in Bombycid and Geometrid larvre, to afford specific 

 characters. It varies much, also, among the Notodonts and Satur- 

 niids, and is especially well-developed in those larvae which constantly 

 use the anal legs for grasping, while the front part of the body is more 

 or less raised. It appears to be correlated with enlarged anal pro- 

 legs. According to Packard, this plate, morphologically, appears to 

 " represent the dorsal arch of the tenth or last abdominal segment of 

 the body, and is the ' anal operculum ' or ' lamina supra-analis ' of 

 different authors. This suranal plate is, in the Platyptericidae re- 

 markably elongated, forming an approach to a flagellum-like terrify- 

 ing appendage, and, in the larva of Aylia tau, forms a long, promi- 

 nent, sharp spine. Its shape, also, in Centra caterpillars, is rather 

 unusual, being long and narrow. In the Ceratocampidae, especially 

 in Anisota, Dryocampa, Fades and Citherunia, this plate is very large, 

 the surface and edges being rough and tuberculated, while it seems to 

 attain its maximum in Sphingicampa, being triangular, and ending in 

 a bifid point " (Bombycine Moths, p. 25). 



The " paranal lobes " are the " homologues of the two anal valves 

 observed in the cockroach, and occur in all, or nearly all, insects," 

 according to the same author. They are the " valvulfe " of Bur- 

 meister, and the " podical plates " of Huxley. They are fleshy and 

 papilliform in Geometrid larvae, and appear as if projecting backward 

 from the base of the anal legs. In the larvae of the Dicranurids they 

 are similar, and each ends in a seta. 



The "paranal forks" or "paranal tubercles" are two bristles 

 arising from the end of .a papilla, directed backward. They are found 

 in the larvae of most arboreal caterpillars, being especially well- 

 developed in those of Notodonts and many Geometrids, whilst they 

 are wanting in the larvae of Noctuids, Sphingids, Bhopalocera (?), and 

 some Geometrids- and INCOMPLETE (Micro-lepidoptera). In the 

 American Choerodes, they are very large ; so also are they in the larva 

 of our common Uropteryx sambucaria, where they become papilliform 

 and setiferous. Their use was discovered by Hellins. In his description 

 of the larva of Cerura bifida, he writes of them : " At the tip of the 

 anal flap are two sharp points, and another pair underneath, which 

 are used to throw the pellets of frass to a distance." Packard has 

 seen the frass pellets held by the two spines of the paranal tubercles 

 in Centra borealis, whilst Dyar says that he has seen the caterpillars 

 throw their pellets, with the aid of these spines, away from them, so as 

 to strike against the side of a tumbler in which they were confined. 



The "infra-anal lobe" is described by Packard as a "thick conical 

 fleshy lobe or flap, ending often in a hard chitinous point, and situated 



