PSYCHIDAE. 371 



certain directions, yet, as a whole, they are of a very generalised type. 

 As in the Megalopygidae, nearly the whole of the surface of the scape 

 and pedicel is covered with long, narrow, hair-like scales, and all of 

 the clavola excepting the ventral aspect is clothed with scales of but 

 little higher type. The greatest development is reached in the 

 pectinations. Relatively, they are the longest found among the 

 lepidoptera. In Psyche confederata some of those near the middle of 

 the clavola attain a length equal to one-half that of the whole antenna. 

 The bases of the pectinations have migrated proximad and occupy a 

 central position on the segments. Hairs of the third type are numerous 

 on the ventral surface of both shaft and pectinations. They are well 

 developed, but have no regular arrangement in their insertion. Pits 

 are rare, and are limited in the forms studied to the two or three distal 

 pairs of pectinations. Cones also are rare, and when present are 

 situated at the ends of the pectinations. The antennae of Thyridopteryx 

 ephemeraeformis present a peculiarity in the joint between the scape 

 and the pedicel. The latter segment is jointed, not at the apex of the 

 scape in the ordinary way, but is set obliquely on the caudal edge of 

 the apex." The interesting facts observed by the study of the Psychid 

 antennas give considerable clues as to the line of evolution of the 

 superfamily, and, apart from the Epichnopterygid antenna already 

 discussed (ante, p. 337), one finds that whilst the Psychinae have 

 densely scaled shaft and pectinations, those of the Acanthopsychinae, 

 tend to lose the scaling first from the pectinations and afterwards 

 from the shaft (vide, })ost. pp. 375-376). 



Kellogg notes (Taxonomic value of scales of Lepidoptera, p. 85) of 

 the scales of Psychidae (which, in North America, comprise only ten 

 known species in the five genera — Psyche, Pseudopsyche, Platocoeticus, 

 Thyridopteryx, and Oiketicus) as follows : "Of four genera examined 

 the correspondence in scale-specialisation is obvious. The wings are 

 sparsely scaled (in Thyridopteryx the scales are disappearing, the wings 

 being mostly clear and unsealed) and there is little arrangement of the 

 scales into rows. The specialised scales in the family are small, 

 narrow, strongly pectinated, usually with two short points (as in 

 Thyridopteryx and Pseudopsyclie), or with three short teeth (as in 

 Pseudopsyche), or with one point (as in Psyche and Oiketicus). The line 

 of specialisation is as follows : The hair-form shortens, widens, and 

 divides at the tip into two very short points, which persist or disappear 

 during the continued shortening and widening of the scale. The 

 points are acute and never more than one-fifth the length of the whole 

 scale. The striae average about •002mm. apart." One suspects that 

 the clear wing-membrane found in many Psychids is the result of 

 specialisation, normal scaling being characteristic of the most 

 generalised Psychids. We find in the scale-clothing of these insects 

 not only the generalised scale-hairs and well-developed scales, but 

 specialised hair-scales in varying numbers in different species, the 

 latter being more particularly abundant in those families that other 

 structures suggest as the most specialised. We are inclined, therefore, 

 to look on these hair- scales, as specialised by degradation, and leading 

 up to the total suppression of all scale-structure, a line of evolution 

 which finally culminates in the development of a perfectly transparent 

 wing-membrane practically devoid of any scale-clothing. 



The immense abundance of some species in their restricted 



x2 



