MACRO- PS YCHINA. 273 



duced, unless, perchance, within a very limited period by atavism. 

 The Solenobiids and Taleporiids have short anterior tibial spurs, as also 

 have the Lumids. On the other hand most of the Fumeids have long- 

 ones, as also have the Oiketicids and Acanthopsychids, whilst Proutia 

 (betulina) is in an intermediate position, having anterior tibial spurs of 

 about f the length of the tibia, and, so far as the material at hand goes, 

 this appears to be almost the only genus of the Psychids in this posi- 

 tion. The Epichnopterygids (? including Bijuijis), however, have short 

 anterior tibial spurs, which clearly show that these forms are derived 

 in common with Proutia aud Fumea from Luffiid forms but not 

 through these (or similar genera) as intermediate forms. Fumea (casta, 

 &c), with its long anterior tibial spurs is, as far as this character goes, 

 on the direct line to the higher Macro-Psychids, and we have the Psyche- 

 oidinae (Diabasis) as intermediates, in so far as the first step in the loss 

 of the posterior tibial spurs is concerned. In the higher Macro- 

 Psychids the posterior spurs are evanescent or lost ; the anterior spurs 

 are also often lost, but when they are preserved they are long, showing 

 the derivation from a form of which Fumea with the transitional Proutia 

 is the lowest preserved. 



When we examine the antenna?, we find they divide the Macro- 

 Psychids into two groups, just as is determined by the tibial spurs, and 

 the division is precisely the same in both cases". In Luffia, the 

 pectinations of the antennae are devoid of scaling, and the long 

 sense-hairs occur more or less all round them, the scaling being con- 

 fined to the dorsal aspects of the bodies of the antennal joints. Pre- 

 cisely the saine structure occurs in Proutia, Bijwjis, and Epichnopteryx. 

 In Fumea the scaling runs right down the dorsal aspect of the pectina- 

 tions to their tips, and this structure obtains throughout the (Oiketicids 

 and) Psychids, except that certain Acanthopsychids appear to have lost 

 the scaling on the pectinations ; the smoothness of the dorsal aspects 

 of the pectinations makes this tolerably certain. If the want of scales 

 was due to their being rubbed off or lost, their points of attachment would 

 be very evident, and if scales had never been there the surface would 

 present some sense-hairs, which it certainly does not. Of those examined, 

 Canephora unicolor preserves a few scales on the antennal pectinations, 

 the other Acanthopsychids seem to have lost them. The following 

 tabulation is an attempt to summarise the chief imaginal characters of 

 the two groups, viz. — (1) The Macro-Psychids derived from a Luffiid 

 base. (2) The Macro-Psychids derived from a Fumeid base : 



A. Primary Macro-Psychid (still with some Micro characters, 

 as those of s pupa and possession of secondary cell in 

 apex of discoidal cell) ; short anterior tibial spurs (£) ; 

 four posterior tibial spurs ; antennas with unsealed 

 pectinations : ? araneiform . . . . . . . . Luffiidae. 



I. Anterior tibial spurs short (£) ; antennal pectina- 

 tions unsealed, posterior tibiae with 4 spurs — 



1. Semi-araneiform ¥ . . . . . . . . Bijugis. 



* Chapman, to whom we are indebted for the facts in this part of our work, 

 says that he " was very pleased that the antennal structure and anterior tibial 

 spurs gave exactly the same indications," although he adds that he " was unpre- 

 pared for their absolutely separating the Fumeids and Epichnopterygids, and 

 throwing the former over to the higher Psychid branch." He concludes that " when 

 two such apparently unconnected characters agree so closely, it is impossible to 

 avoid giving great weight to their indications." 



R 



