94 BRITISH LEPIDOMEBA. 



butterfly pups, and that, in each family, the loss of movement 

 occurred in frequent separate instances. There are no facts to suggest 

 that movement was ever regained when lost, and as such an occurrence 

 would be contrary to the ordinary course of evolution, it is tolerably 

 certain that it never took place. The great dominating influence in 

 the butterfly pupa was unquestionably protection against enemies by 

 assimilation in form and coloration to its surroundings, so as to escape 

 observation. 



It is curious that in two of the examples of the loss of cocoon in 

 the Lepidoptera-Phala?na? — Elachistids and the Ephyras (Zonosoma) — 

 the pupa should be disposed much in the same way as in typical butter- 

 fly pupa?, and in Zonosoma the pupa itself should have many points of 

 resemblance to a Papilla pupa ; yet it is quite certain that, its spines 

 and ridges, the girth, and the cremastral pad, have all developed in 

 Zonosoma from a pupa that had no trace of any butterfly affinity, and, 

 in fact, had no such affinity. The pupal suspension in Zonosoma is 

 probably reached from some slight cocoon structure as in Pnnomos, 

 just as the slight cocoon of some Hesperiids leads to the butterfly 

 suspension in others. 



The pupa of Hepialus clearly does not belong to the Pala?o- 

 Lepidoptera. It is a somewhat advanced form of pupa-incompleta, 

 and has acquired many characters that bring it very close to Cossus. 

 It has sundry characters that shew it to be a terminal form, 

 and that Cossus is not directly derived from it, however close a common 

 ancestor they may have had. The most important of these is the 

 entirely exceptional one amongst the pupae-incompletae of having the 

 7th segment fixed in the male. The great vertical flange in this 

 segment, for use as an implement of progression, is no doubt 

 associated with this fixity of the segment, and is an equally unusual 

 character. As a pupa, it is as advanced as that of Psyche or Anthrocera, 

 and much more so than that of Coclilidion or Nepticula, and, whatever 

 primitive characters it may have retained in the imaginal state 

 (chiefly neurational), it is as far removed from Eriocrania as many 

 other Neo-Lepidopterous families. 



The progress of pupal evolution throughout the Lepidoptera, then, 

 has always been in the direction of greater soldering together of parts, 

 and of greater solidity of the enveloping surfaces ; so that at one end 

 we have the Eriocraniids with all parts movable and separable and a 

 very flexible delicate surface, on the other we have the hard solid case 

 of Spilosoma, in which no separation of parts can be effected, none are 

 movable, and only surface lines evident, the points where such 

 separation and mobility once occurred. A description of the inter- 

 mediate stages and of the many variations in the parts that may be 

 detected in the lower pupa?, and are lost in the higher, would be 

 lengthy ; but it may be noted that there is an oval head-plate existing 

 in nearly all the lower forms and persisting in some of the higher, 

 that it carries, on the dehiscence of the pupa, in the lower forms, the 

 eye-covers that remain attached to it by a chitinous sheet passing under 

 the antenna. In the higher forms, this connection becomes too 

 flimsy and evanescent to perform any such function ; all others 

 beneath the surface equally become mere films in the higher pupa?, so 

 that, in dehiscence, the separation of organs is quite irregular, and the 

 leg- and antenna-cases may even be irregularly fractured. 



