( 20fi ) 



II 



joint is so short, have for tlic most part mpigecl together, anil are verv higli 

 a transverse section, or in a front view, of a middle joint, the diameter of the stem 

 (if the joint is much shoi-ter than that of the ventral dilatation, the outline lioiiig 

 somewhat like this Q. The hairs njion the ridge arc uuuli longer than in any 

 other Agauaid, being about three times the length of a joint; they stand in two 

 rows, as in heliconia, but these row^s join each other near the subdorsal bristle, as in 

 XeocJiei'a marmared Wlk. In a side view, the ventral dilatation appears as a 

 short square tooth of which the vertical diameter is shorter than the vertical diameter 

 of the stem of the joint ; while in a ventral view the processus is transverse and 

 somewhat rhombiform. The subdorsal bristle is very strong and long ; the ventnd 

 pair is present on the apical elongate joints, but is obliterated on the other joints. 



The strong subdorsal bristles of A. contoHa (Auriv.), e<ien8 (Wlk.), and other 

 species of Asota remind one of tiie lateral proce.ssus found in At/nwiis Boisd., if one 

 examines the antennae quite su[iprKcially under a very weak lens, but are liy no 

 means homologous with that processus. (.)n p. (il 1 tried to poinl mit tlic ililVcrcncr 

 between the antennae of Aganfiis Boisd. and Asofa III). ; on I'l. 1\'. tigs. 1, 2, 'A, 4, I 

 now give front and dorsal views of an antennal joint of 6 and ? of Arjandis ficnx 

 (F.), which will at once show to the reader, when comparing figures 6 to 11, the 

 remarkable differences described on p. (il. 



The antennae of all Agan/iidae a.re similar to those types described in lln'se 

 lines. The heliconin-iyiie we I'ather frequently meet with amongst Geomelridae. 



The palpi of Asota do not exhibit much variation. The second joint is generally 

 slender, at the base curved, and somewhat twisted; the third is very thin, slightly 

 thickened towards the tip, and is in A. caricae (F.) as long as, in er/eiis (Wlk.) 

 shorter than, and in hdlcouin, (L.) longer than, the second joint ; the difference 

 in length between the second and third joint is always slight, except in ..4. coutorla 

 (Auriv.) and tartuosa (Moore), which have the second joint shorter and thicker than 

 it usually is in Asota. and the third joint more prolonged. 



The thorax of Asota and all other .A/yr /((«'(/'/«, inclusive of .!//«//« Snell.. bears 

 just in front of the hinder edge of the mesoscutnm in the middle line a rounded 

 iinpres.sion often filled up with a brownish matter. This impres.sion is most easily 

 seen in Neochera, on account of the thorax of this genus being less hairy than that 

 of Asota or Peridro'ine, and looks here like a small black spot. This impression is 

 the mouth of a gland which I have not yet found in other moths; it is certainly not 

 present in A/etis Hb., Kyctemera Mb., and I'eluchyta Hh., which some authors have 

 associated with the A</aiuiidae. 



The abdomen of .^I. caricae (F.) is in the female remarkalsly different from that 

 of all other Ai/anaiilae, and resembles that of Lymantriidae. For conipari.son I give 

 two figures of the tip of the abdomen of A. plaginota (Butl.) and two of caricae 

 (F.). The .seventh segment is in the ? of jilaginola (Hutl.) and all other Ai/anaidae 

 conical; the dorsal plate is longer than tlie ventral one, and is covered, like the rest 

 of the abdomen, with narrow elongate scales, which are at the ajiical edge of the 

 segment somewhat prolonged (PI. IV. fig. 46). Fig. 47 shows the tiji of the abdomen 

 denuded. 



In A. caricae (F.) the seventh segment (fig. 48) is .shorter and thicker than in 

 other Af/anaidae, not conically narrowid towards the ajiex, and is covered, like the 

 eighth .segment, with long curly hairs, whicli form a conspicuous anal tuft. In the 



