FUMEIDI. 299 



direction of the long axis of the case. Although much has been 

 written of these insects little is known about them, and no author of 

 repute, except Bruand, has attempted to work out the life-histories of 

 the closely allied species, and some of his descriptions are far from 

 satisfactory. It would appear that the known species really to be 

 included in the subfamily are as follows : Reticulatella, raiblensis, comi- 

 tella, norvegica, saxicolella, subflavella, edwardsella, crassiorella, affinis, 

 mitfordella, hibemicella, casta, scotica, germanica, and a number of 

 casto-like forms that may or may not be entitled to specific rank. 



Of these there appear to be two groups separable on account of 

 their neuration, one group following the Epiehnopterygids and having 

 a distinct " cellula intrusa," whilst the other has the median nervure 

 running through the discoidal cell without bifurcation. These are : 



(1) Reticulatella, ? raiblensis, comitella, norvegica (teste Heylaerts). 



(2) ? Saxicolella, subflavella, edwardsella, crassiorella, affinis, mitfordella, hiber- 

 nicella, casta, scotica, germanica. 



The true Fumeids differ from their near allies, the Proutiids, by 

 having long spurs on the anterior tibiae, and, in this respect, agree 

 with the higher Macro-Psychids, excluding the Epiehnopterygids which 

 have short (or no) anterior tibial spurs and also scaleless pectinations 

 to the antennte, the Proutiids agreeing with the Epiehnopterygids in 

 the latter character. The distinction in the antenna? is probably much 

 older than that of the tibial spurs, since no transitional stages of the 

 former are known, whilst, as regards the spurs, we have, on the Epich- 

 nopterygid side, the genera Bijugis and Psychidea, which are transi- 

 tional, leading from the Micro-Psychids to Epichnopteryx (without 

 spurs), the tendency to long spurs, however, failing on this side and 

 not going beyond the transitional forms. On the Fumeid side leading 

 to the Macro-Psychids, the Fumeas present transitional forms, some 

 species with fairly short and others with long spurs. Those with short 

 spurs are reticulatella and comitella, and it is interesting to note that 

 both these species retain the tessellated wing-marking so common in 

 the Micro-Psychids. The spur, broadly speaking, reaches to the end of 

 the tibia in all cases, occasionally it distinctly projects beyond it. But 

 we may assume that the end of the tibia is its limit, without serious 

 error, and may express the length of the spur by the distance of its 

 point of origin from the end of the tibia by a decimal. The range of 

 variation in the length of the spur in the tribe, is from -56 in B. 

 reticulatella to # 88 in F. germanica. This affords a specific character of 

 some value. Based on the length of their anterior tibial spurs, the 

 Fumeids work out as follows : 



(1) Short spurs (under -64 of tibia) — reticulatella, comitella, norvegica. 



(2) Intermediate spurs (from •65- - 71 of tibia) — subflavella, edioardsella, 

 crassiorella, affinis, mitfordella, hibemicella. 



(3) Long spurs (from - 77- - 81) — scotica, casta with vars. minor, intermediella, 

 bowerella ; (spur -88) germanica. 



Another character (probably never more than specific) is afforded 

 by the number of antennal joints. In this character, Chapman says, 

 there appears to be some amount of variation in each species, and 

 although one might properly separate the crassiorella with 21 and 24 

 joints respectively, as distinct species, yet one hesitates to do so when 

 one finds examples with 22 and 23 joints, amongst specimens that it is 

 impossible to doubt are all the same species, and the same considera- 

 tions apply strongly to the forms of casta. The origin of this 



