( •""9 ) 



IVum a sevuu-bristled, auteuna. It wUl be seen from the figures that the bristles 

 staud also in longitudinal rows. 



The dorsal bristles have mostly an irregular position, but here and there we find 

 them arranged in an irregular transverse row. 



2. LYCAEXIDAE. 



The joints of the stalk are always cylindrical; those of the club, also cylindrical 

 in most species, are sometimes somewhat flattened, especially in the species with 

 abruptly and strongly clubbed antennae. In rare cases (f. 7, Liptena) the joints of 

 the club are contracted at base, so that the club has a serrate appearance, especially 

 in a side view. In these latter L>/caenidae as well as in those forms which have the 

 antenna abruptly clubbed, the club is often folded longitudinally in dry specimens, 

 the ventral wall of the joints having sunk in when drying: this phenomenon docs not 

 seem to occur amongst skippers, but is very often met with in thin antennae of other 

 families. The shrinking, as here observed, must not be confounded with the de- 

 velopment of grooves of other families ; Lycaenid antennae are never grooved. The 

 club is ventrally not more expanded than dorsally, or the difference is very slight. 



The scales are on stalk and club either arranged in regular transverse rows, or 

 the serial arrangement is much disturbed. In the extent of the scaling there is a 

 much greater variety than in Hesyeriidae. According to the space left unsealed, we 

 can arrange the antennae into three groups: (1) Distal or all joints of club ventrall}', 

 last one to five joints dorsally, without scaling. (2) The not-scaled ventral area 

 runs down the stalk, sometimes to the base of the antenna; the greater part of the 

 club not scaled dorsally. (3) Like (1) and (2), but the more proximal joints that 

 participate of the ventral bare area have scaling at the apex ventrally, which 

 separates the jiroximal portion of the not-scaled area into patches; in Ht/pochri/sops 

 there is generally one row of scales, while in Lucia (f. 6, Lucia auri/era) the apical 

 scaling is much more extended. 



The not-scaled area is covered all over with fine sense-hairs ; it is of great 

 significance that the fine sense-hairs are distributed, not only over the ventral 

 surface, but also over the entire not-scaled dorsal surface (f. D). In a few of those 

 cases where only the last one or two joints are not scaled dorsally, there are no fine 

 seuse-hairs in the mesial line of the dorsal surface {Liphyra brassolis). In the 

 species which have the joints of the club narrowed at the base {TJptcna, f. 7), the 

 hairs are less dense at the apex than at the base of the joints. 



The sense-bristles on the not-scaled portion of the antenna are arranged (f. 8) in a 

 transverse series as in Ilesperiidae, but on the stalk the series is often very irregular. 

 In Eumaeus the bristles have also on the club an irregular position. In the forms 

 which have the distal four or five joints entirely without scaling {Tkeclu, f. 0; 

 Arhopala), the series of bristles — generally up to eleven in a series — extends all 

 round the joint, so that these joints are of the same form and bear the same organs 

 dorsally and ventrally. The middle joints of the club which are scaled above have 

 a ventral series of seven bristles, as is the case in many Ilesperiidae; this number 

 seems to be constant, inasmuch as it occurs at least on one joint in all species; to my 

 knowledge there is only one species in which the number is reduced on all joints, 

 that is the aberrant Liphi/ra brassolis. Sometimes the externo-lateral bristle is 

 double, as in many Hcspcrids. In the species with broad club {[jijcaena) the lateral 

 bristles are generally longer and thicker than the median ones. The series is mostly 

 median (figs. 8 and \)), seldom apical (f. 7, Liptena). 



