268 BRITSIH LEPIDOPTERA. 



which remain in the case for copulation," as are the former from the 

 Taleporiidae and Solenobiidae, which are excluded because they have 

 " les antennes du male tout autrement conformees, et dans le genre 

 Taleporia, Hb., les especes $ et 2 ont des palpes et des ocelles." 

 Surely the female Solenobia which has all the characters of the Fumeid 

 female given by Heylaerts is nearer the latter (by his own definition) 

 than is this to the female of the higher Psychids. We are quite 

 willing to accept the positive evidence offered by the want of " the 

 tongue, ocelli, palpi, the apterous females," &c, as showing a relation- 

 ship between the various families of which the Macro- Psychids are 

 composed, but we object to a character such as " the difference of male 

 antennae" being considered sufficient to separate two other groups 

 which are evidently somewhat closely related. 



We find, as might be expected from our remarks above, that 

 Heylaerts' genera are based on the same comprehensive scale. Among 

 other details he subdivides Epichnopteryx (after the withdrawal of 

 Bijugis) into groups a and b, and fails to see that if the genus be the 

 first step in grouping above a species, each of these groups must be a 

 distinct genus. Thus he gives us : 



a. Les tibias anterieurs sans epine tibiale — me7itonella,pnlla,hofmanni, ardua, 

 flavociliella, tamierella, undulella, reticella. 



b. Les tibias anterieurs avec une epine tibiale ne depassant pas la nioitie de la 

 longueur du tibia anterieur — sapJw, noctumella, nudella, vestalis, staudingeri, 

 millierei, flavescens. 



Heylaerts' system appears to us to fail in a proper appreciation of 

 detail. Had he studied the characters offered by the species, and then 

 grouped such of the species together as were evidently most closely 

 allied, to form his genera, grouped these again to form his tribes, and 

 so on to subfamilies and families, instead of (as is evidently the case) 

 fixing his higher divisions first and then discovering the species that 

 fitted into them, we might suppose that a more logical result would 

 have been reached. One recognises also that this author, basing his 

 classification almost entirely on the characters presented by the male 

 imagines, has often failed to recognise the true relationships of some 

 of the smaller groups, OAving to his not having been able to check the 

 results arrived at, by using the characters offered by the early stages. It 

 must be conceded that his main divisions of the higher Psychinae are, on 

 the whole, sound, and we ourselves have come to the conclusion that 

 the Oiketicidae and Psychidae are divisible into subfamilies agreeing in the 

 main with Heylaerts' genera, so that our family Psychidae subdivides 

 naturally into the (/ Animulinae), Acanthopsychinae, Oreopsychinae, 

 Empedopsychinae (Psychinae), and Apteroninae. It is in the subdivi- 

 sion of his Canephoridae that we find ourselves in the strongest dis- 

 agreement, since the structural peculiarities of Fumca, Diabasis, 

 Proutia, Bijugis, and Epichnopteryx suggest, for these at least, a 

 diphyletic origin, in which Epichnopteryx and Bijugis present Luffiid 

 affinities, whilst Fumea and Diabasis have more definite Psychid 

 characters. Proutia appears to be somewhat intermediate between 

 Luffiids and Fumeids, with such strong Luffio-Epichnopterygid affini- 

 ties as to suggest a not very distant alliance with the latter group. 



Had we been able to get the material requisite for a thorough 

 study of the genus Bijugis we might be less inclined to cavil with those 

 who insist on a close alliance between Epichnopteryx and Fumea. The 



