strong web, within three or four inches of the entrance, wherein 

 they change to a brown chrysalis. The case is composed of Httle 

 bits of wood, which he bites off with his forceps : this he interweaves 

 with his web. At the expiration of two months the moth appears. 

 It must here be observed that when the moth leaves the chrysalis, 

 the chrysalis appears half out of the tree ; being armed with strong 

 spinules round every division of the tail part, it forces itself forward 

 through the case, to the entrance to the hole, and is drawn further 

 out before the moth can disengage itself." 



This, I think, was the generally accepted history of the pupal 

 methods of the species at that day, and so far as I can gather, no 

 shadow of doubt appears to have been thrown upon this being its 

 invariable habit until some sixty years later. 



Then, however, records began to be made in the entomological 

 journals that appeared to throw some doubt upon the question. 

 Larv^ were reported to have been found during the winter under 

 bark ("Zool.," 1858), in some cases of trees on which they did not 

 appear to have fed ("Zool.," 1861, p. 7322), and there are many 

 notes of them being dug up out of the earth (" Entom.," xxxii, pp. 

 17, 71 ; xxix, p. 194; XXV, p. 46; "Buckler's Larvae," ii, p. 60), or 

 detected in the act of burrowing into it (" Entom.," xxxii, p. 279). 

 There are also numerous records, generally accompanied by expres- 

 sions of surprise, of the pupa being found in the earth (" Entom.," 

 vi, p. 487 ; XX, p. 274 ; xxi, p. 56), the empty shell protruding from a 

 tennis court, the moth having just emerged ("Entom.," xx, p. 231), 

 and in "sand-cops" ("Entom.," xxi, p. no), which it is explained 

 are low embankments used as a fence or boundary ; they are made 

 of earth or sand, banked at the sides and top with grass or sods, and 

 are often planted with dwarf willows (" Entom.," xxi, p. 156), and 

 the method of finding the pupae was to feel along the top of the 

 "cop" for a soft place, caused by the cocoon being just under the 

 surface ("Entom." xxi, pp. no, 155) ; and an instance is related of 

 a dozen or more larvae being put into a small brick house in a 

 garden, with willow chips for food, and the following June the empty 

 pupa skins were found protruding from the mortar on. the outside 

 ("Entom.,"v, p. 456). 



Westwood ("The British Moths aad their Transformations," 

 vol. i, p. 48) gives the following somewhat detailed account : — " It 

 forms a rough cocoon of the chips of wood which it has bitten to 

 pieces, fastening them together with a glutinous secretion, and 

 lining them with silk. The pupa has the head-case acute, and each 

 of the abdominal segments is furnished with several rows of reflex 

 spiny hooks, which are of great service in enabling the pupa, shortly 

 before arriving at the perfect state, to push itself through its cocoon 

 and to the surface of the tree." 



Coming to the more recent authors, Newman (" British Moths," 

 p. 18), also speaking of the larva, tells us that "before changing to 

 a chrysalis it spins a very large tough cocoon, composed of silk mixed 



