504 Mr. A. Quail on 



is typical in this respect. The majority of species of 

 Charagia have only a few dorsal scales on not more than 

 half-a-dozen basal segments, the scales of the antenna 

 like those of the wing being green in colour, striated, not 

 dentate, and the segments have numerous pits bearing 

 superficial resemblance to sockets of long-lost scales. C. 

 daiyhnandrx preserves some basal segments scaled dorsally 

 and laterally, with " primary " and " sense " hairs and 

 bristles located on a ventral unsealed area ; the segments 

 beyond are devoid of scales, "sense" hairs are not numer- 

 ous on the dorsum, they are so, however, both laterally 

 and ventrally, with bristles on all aspects at the distal end 

 of the segments. C. lignivora has gone a long way 

 ahead, the scales having almost completely vanished, and 

 " sense " hairs are numerous on all aspects. 



European Hcpialus have likewise gone a long way 

 towards complete loss of clavola scales. H. lupulimis 

 retains single scales per segment curiously enough on the 

 apical half of the clavola, the scales are not dentate and 

 are longer than ordinary clavolar scales ; one finds here 

 "sense" hairs on the dorsum as on otlier aspects. The 

 antennal scales of If. lupulinus suggest that Hcjnalus, 

 which now so generally have only a few clavolar scales, 

 somewhere in the past had antennoB scaled from base to 

 tip. The subgenus Oihyra preserves exactly this condition, 

 and must have separated from Hcpialiis when the antenna: 

 were scaled to tip ; it is not in the least likely that Cihyra 

 has since developed scales after Hcpialus once lost them. 

 Oihyra then retains this character, but has developed latero- 

 ventral enlargement — the beginnings of pectination — there- 

 by increasing the " sensory" area, which indeed is wholly 

 covered with "primary" and "sense" hairs, there being 

 none of the latter on the scaled dorsum. 



The latero-ventral enlargement of Cihyra is not properly 

 unipectination although in profile it appears to be so, it 

 is a widening and ventral lengthening of the segment as a 

 whole. Definite ventral unipectination does exist else- 

 where among Lepidoptera {Incurvaria muscahlla), and 

 may be observed among Diptera (Tijndid/e). 



Among the latero-ventrally enlarged forms known to me 

 the clavola segments of hacotii {Gorgo'pis ?) are most special- 

 ized ; " sense " hairs completely cover the segments on all 

 aspects together with " primary " hairs, even the bristles 

 are reduced in size, so that the clothing of the segments is 



