58 BEITISH LEPIDOPTERA. 



abdominal segments of the pups of Dasychira pudibunda, Leucoma 

 salicis, Notolophus antiqua, &c. In Lymantria monacha, the hair-groups 

 on the abdominal segments are suggestive of tubercles i, ii, iii, iv, 

 and v of the larva, whilst in Portlietria dispar the hairs fringe the 

 slightly-raised bosses that appear to represent the larval tubercles. In 

 Dasychira pudibunda, Bacot notes the dorsal surface as being slightly 

 raised and roughened, apparently scars of the dorsal tufts of the larva, 

 whilst the hairs form bands across the segments, the pupa also presents 

 traces of the supra-, post-, and subspiracular groups of hairs. In 

 Notolojrfius antiqua and N. gonostigma he observes that the pupal hairs, 

 which occupy the position of the larval tufts, have the appearance of 

 whitish circular scales, that appear to be attached at the centre and 

 occasionally overlap one another. They have a semi-transparent, opal- 

 escent appearance, and are wrinkled in a more or less regular spiral 

 pattern, the flutes or crenulations, starting from a central button, curve 

 outwards to the circumference. Poulton also notes the presence of large 

 tufts on the first three abdominal segments of the pupa of Notolophus 

 antiqua, those on the 4th abdominal segment of the larva, he says, are 

 not represented. He adds that the three tufts in the pupa form 

 squarish light-coloured patches, which are very distinct against the dark 

 pupal cuticle, and are especially prominent and well-defined in male 

 pupa?, which are much blacker than those of the females. The absence 

 of any trace of the fourth tuft is very remarkable. . . . There is a 

 long pencil of hairs on the 8th abdominal segment of the larva of 

 Dasychira pudibunda of which he could detect no trace in the pupa. 

 . These appearances on pupa?, he considers, belong to a very 

 different category from the merely mechanical scars, such as those pro- 

 duced by the caudal horn in Sphingidae, &c, for the pupa does not bear 

 a scar of the larval tuft, but possesses a true hairy tuft itself. On the 

 other hand, hairy prominences may be represented by mere scars, as in 

 Triaena psi. Bacot considers that, since the pupa is older phylogeneti- 

 cally than the larva, it by no means follows that if the tubercles and 

 hairs on the larva undergo alteration of structure or position, those of 

 the pupa must be altered also. In the Psychids, the arrangement of 

 the tubercles (and their seta?) appears to be identical in the larva 

 and pupa. In the Liparids (much more specialised Lepidop- 

 tera), the pupal hairs follow the larval tufts in position, but the 

 pupal hairs differ in structure from those of the larva. Bacot 

 says (Ent. Record, xi., p. 173): "In the Liparid larva? that I have 

 examined all the larval hairs (so far as my knowledge extends) are 

 thorny, branched, or plumose, and circular in section, while the pupal 

 hairs are, with three exceptions, without thorns or branches and are 

 frequently flat or ribbon-like. In Notolophus antiqua and N. gonostigma, 

 the situation of the larval dorsal tufts is occupied in the pupa by 

 structures which are not hairs at all in the usual sense of the word, 

 although they may be homologous in a physiological sense. The three 

 exceptions to be noted are as follows — Lymantria monacha and 

 Porthetria dispar (in both of which the pupal hairs are thorny), whilst 

 Leucoma salicis also has a few of its hairs of this character. These 

 pupa? are, however, greatly exposed, in the case of L. monacha and 

 P. dispar they are suspended in a slight silken hammock, while the 

 cocoon of L. salicis is often very slight. The pupa? of Porthesia similis 

 (aurijiua) and P. chrysorrhoca are well protected by the irritating larval 



