NOVITATES ZOOLOGICAE XXIII. 1916. 355 



the edges. The upper (central) lobe projects more forward than the lateral lobes. 

 Occasionally the process is distinctly asymmetrical. Figs. 1 and 6-9 give an 

 idea of the individual variation of the process. Differences such as shown in 

 figs. S and 9 might easily be mistaken for specific, if only a few specimens were 

 examined. The intermediates and the general instability of the size and outline of 

 the process, however, dispose of that opinion. 



In the female the processes are broader than in the male, the posterior ridge 

 is usually higher, and the margins of the antennal grooves are more strongly 

 elevated. 



The second character mentioned above in the diagnosis of Group A relates 

 to the covering of the body. There are no scales on the body, only hairs and 

 bristles. In the S the hair is long, soft and somewhat silky, and beneath it the 

 upperside of the abdomen is densely studded with numerous spine-like bristles, 

 stiff, short aud sharp, lying more or less flat on the segments. In the ? the 

 ■covering of hair is much sparser, and the hairs are shorter and stiffer, resembling 

 slender bristles also on the underside of the body ; the short bristles of the 

 abdominal tergites are rather more numerous than in the cf. In Groups B and C 

 the hairs and bristles are similar, but the bristles are fewer in number, the difference 

 being very obvious if specimens are compared side by side. The development of 

 such bristles in Somabrachys is explained by the habit of pupating in the ground. 

 We find them in numerous moths which have the same habit. They represent an 

 instructive case of convergent development, due to adaptation to similar circum- 

 stances of life. The peculiar structure of the foretibia which Somabrachys has in 

 common with Lemonia is another instance of this kind of resemblance. 



The legs of Somabrachys are almost alike in the sexes, with these exceptions : 

 (1) that the legs of the male, especially the tarsi, have a covering of scales besides 

 bristles, whereas in the female there are only bristles and hairs ; (2) that the claw 

 at the end of the foretibia and the corresponding apical tooth of the mid- and hind- 

 tibiae are larger ; and (3) that the tarsi are rather stouter. The differences from 

 Groups B and C are found in the foreleg. The claw of the foretibia is usually con- 

 vex on the upperside in Group A (= Somabrachys aegrota), rarely being longitu- 

 dinally flattened or grooved. The forefemur is considerably slenderer than in 

 Groups B and C (PI. VIII, fig. 21), and beneath less flattened, the subconvex under- 

 surface being bounded on the innerside only apically by a rudiment of a ridge, and 

 the bristles placed on this side of the femur being the same colour as, or only 

 slightly darker than, the hairs on the outer surface of the foreleg. 



The most easily perceived distinction between the groups, however, is the 

 difference in the position of vein R 1 (= 6) of the forewing. This vein arises in 

 group A invariably above the angle of the discocellulars (PL VIII, figs. 16 and 18), 

 the distance of its point of origin from that angle being unstable, as it also is in 

 Group B. It is singular that nobody has taken any notice of the neuration of the 

 forewing. Differences in the veins of the hindwiug have been mentioned in the 

 original descriptions of poicelli and chretieni, which, however, do not hold good. 



The individual variability is very considerable. Studying a long series of 

 specimens, one meets with remarkable deviations from the normal Somabrachys 

 neuration, the deviations occurring either symmetrically on both the right and left 

 wings, or only on one wing. There are four subcostal branches in Somabrachys, 

 SC 2 being absent ; in two of our specimens of Group A vein SO 4 is forked on the 

 right wing, and in one specimen on the left wing. In the hindwing the first 



