( 380 ) 



3. ERYCIISUDAE. 



The joints ol' rlu- stalk are cylindrical, seldom somewhat conical {Uelicopis); 

 distill joints often flattened veutrally, as are iu most species the joints of the clnb. 

 The not-scaled ventral areas of the proximal joints of the stalk are always convex: 

 those of the distal joints of the stalk are convex, or, as is mostly the case, either 

 flattened in middle or distinctly impressed. The middle of the joints of the club is 

 always flattened or impressed. The impression is often rather deep and well defined, 

 at least proximally: in many cases it has a more or less ovate sl'.aj)e (f. 10, Xecyria 

 suuiulersi), while in other forms it is a wide furrow extending from the base to the 

 apex of the joint (f. 11, Nemeobius lucina). There is a gradual transition from the 

 not-grooved to the grooved antenna. As in Lycaenidae, the clnb is very often folded 

 longitudinally in cabinet specimens, in whicli case tlie grooves of the club are not 

 well visible. As one fluds specimens in which the grooves are deeper than iu other 

 individuals of the same species, it is most probable that the grooves are more 

 shallow in live specimens than in dried ones. 



The scaling, which is rather rough in IIelicojji», is in most species of great 

 extent. The scales are irregularly arranged on the stalk, but on the clnb they stand 

 sometimes iu regular transverse rows. According to the extent of the not-scaled 

 area, we distinguish the following principal t)'pes : (1) The seven apical joints 

 ventrally and the five apical ones also dorsally not scaled {Taxila only). (2) Four 

 distal joints not scaled dorsally, a greater number of distal joints not scaled vent- 

 rally, the i)roximal joints of the club and the distal joints of the stalk with small 

 not-scaled middle areas {Diculloneura'). (3) Only two apical joints not scaled 

 dorsally, and either only five not scaled ventrally {Saribia tepahi [1835]), or there is 

 a ventral middle stripe of variable width extending sometimes down to the base of 

 the antenna, on the stalk very frci(uently divided up into often small areas in con- 

 sequence of the scaling encroaching upon the distal portion of the ventral surface 

 of the joints (f. 10, Neciji-ia saundersi). (4) Like (3), but the small ventral bare 

 spaces stand at the apex of the joint and extend a little upon the base of the 

 following joint (f. 12, Uelicopis). 



1 JKive not met with a species in which the ajiical joint is scaled dorsally. 

 The sense-hairs are distributed over the ventral not-scaled area; in the species 

 with grooves the hairs are not restricted to these grooves, though they are some- 

 times denser in the groove than without. The length of the hairs varies consider- 

 ably; the proxiiniil joints of grooved antennae have the hairs often very long. On 

 the not-scaled dorsal surface of the distal joints (one to six) the tine sense-hairs are 

 modified, being very short and placed in punctures: the dorsal and ventral surfaces 

 are, therefore, in opjwsitiou to what we found iu I.i/ciicniddc, different here — the 

 ventral side has a covering of sense-hairs as in that family, while the dorsal side is 

 covered with setiferous punctures. 



The sense-bristles resemble in arningement sometimes those uf Li/caenidae. 

 There is a transverse series as in that family, extending nearly all round the joint 

 in the few forms where several distal joints are not scaled. The number of bristles 

 in the ventral serii's on the joints of the club wliich are scaled above is not constant; 

 the number is mostly reduced, the complete series beiug seldom met with {Er;/ciiia; 

 f. 13, A/w'/luris melibofus). On the club there is at least one bristle at each side 

 left, while on the joints of the stalk often only the externo-lateral one is preserved 

 (i. 14, yymphidium). In all (?) Erycinidae the ventral bristles of tlic club are 



