the antmncV of IIcpiaHd,v — Lcpidoptcra Jvrjatpe. 505 



remarkably uniform. We have here a form of Hepialid 

 antenna answering the Jordan requirements of an antenna 

 covered with hairs and devoid of scales ; this form, however, 

 from our standpoint is reached by specialization, and 

 therefore is not primitive. Indeed it seems abundantly 

 evident that the tendency in the Hcpialidfe has been to 

 completely get rid of the clavola scales. This seems at 

 first sight somewhat at variance with the tendency shown 

 by the majority of the Lepidoptera, which generally preserve 

 the dorsal scales; but ffc^nalidcV are a very distinct and 

 somewhat isolated group, moreover the antennae are 

 admitted to be "sense" organs, and it seems only reason- 

 able to conclude that on the antennae and on the wings the 

 function of scales is identical, and it can hardly be as 

 aiding any special sense, such as we believe to be centred 

 in the antennae, for similar sense can hardly be supposed 

 to exist in the wings. It is not illogical to believe that 

 when the primitive Lepidopteron developed scales — how- 

 ever far back we may have to go, the development com- 

 menced somewhere, and has since extended and become a 

 permanent and characteristic feature of the Lepidoptera — 

 the scales spread over tlie several organs, proving to be an 

 encumbrance on the antennae, although of some utility to 

 many groups of Lepidoptera, since not only are scales 

 retained on the dorsum of the shaft, but they are spread 

 over the dorsa of pectinations; but the effects of elimination 

 can be traced everywhere in the localization of the scales 

 instead of their general distribution over the clavola 

 segments, which is to be observed among the very primitive 

 groups, as for instance the eriocranid already discussed. 

 Complete loss of scales by elimination has been the 

 ultimate result on the antennae of many Hcpicdidx. 



Hedomancs have bipectinate antennae generally scaled 

 from base to tip on the dorsum of the clavola, the scales 

 not dentate, with " primary " hairs, ventral " sense " hairs, 

 and numerous ventral pits. The pectinations of Hedo- 

 mancs are essentially lateral extensions, short and curved ; 

 one can hardly say whether highly specialized or no, judg- 

 ing from the scales they are not. The bipectinate, unsealed 

 antenna of Gorgopis (lihania) is a very highly-specialized 

 form. The pectinations are long slender appendages, 

 provided with "primary" hairs, "sense" hairs on the 

 inner side, and bristles on the outer side ; on the dorsum 

 of the shaft the " primary " hairs are modified or replaced 



