NOVITATES ZOOLOGICAE XXIII. 1916. 357 



bidentate as in S. aegrota, but ranch narrower than in that species. The two prongs 

 or teeth vary in length, and very rarely are obliterated. The frons, when denuded,, 

 usually shows a central depression surrounded by an irregular rim, which is exceed- 

 ingly variable, but always low. The rim and the groove are occasionally barely 

 traceable. As a rule the rim is anteriorly open, ending here at each side in a small 

 tooth or tubercle. The groove encircled is often divided up by irregular transverse 

 folds, and frequently extends backwards as a shallow, irregular channel or scar. 

 There is no ridge at the suture between the antennae. The homology of this central 

 structure with the process of S. aegrota is evident from figs. 2 and 4, and 3 and 6. 

 The scar is more or less different in every specimen (cf. figs. 10-12). 



The branches of the antennae, especially those of the distal segments, are- 

 shorter than in S. aegrota (= Group A), and the spiniform bristles on the upperside 

 of the abdomen less numerous. 



The anterior femur is considerably thicker (PL VIII, fig. 22) than in S. aegrota, 

 and flatter beneath, an obtuse edge being formed where the under and inner lateral 

 surfaces meet. The covering of stiff hairs on the innerside is usually blackish 

 brown. The thorn at the apex of the foretibia is flattened above, or more or less- 

 impressed longitudinally. 



Although vein R 1 of the forewing (PI. VIII, figs. 17 and 19) always arises 

 from below the angle of the discocellulars, it sometimes, but rarely, approaches this 

 angle so closely that a specimen with R 1 in that position might easily be confused 

 with Group C, if the wings alone were examined. The individual variability in 

 venation is no less great than in the case of S. aegrota. I mention the following 

 examples : (a) R 1 and R 3 of both forewings stalked ; (b) R 1 of right forewing 

 normal, distally obsolete in left forewing, with a short additional vein arising from 

 the cell in front of R 1 of left forewing ; (c) R 1 of left forewing entirely coincident 

 with R 2 , while in right forewing these two veins are approximated apically ; (d) R 1 

 completely coincident with R 2 in left forewing, in right forewing coincident only 

 apically; (e) R 1 and R 2 in both forewings, and R 2 and R 3 in both hindwings 

 coincident. 



The lower branch of the cell-fold of the forewing is variable in position in all 

 three groups. 



The tenth abdominal tergite of the <$ (we have no ¥ ? ) gradually tapers to a 

 sharp point, which is curved downwards. The tenth sternite is more strongly 

 granulöse than in S. aegrota, and the side-claspers are much broader, touching one 

 another dorsally when closed and thus concealing the tenth tergite (PL VIII, 

 fig. 27). 



GROUP C 



This is evidently less common than the previous groups. We have 68 males 

 from Ai'n-Draham in Northern Tunis, Batna and Guelt-es-Stel in Algeria. 

 Oberthür records chretieni from near Sebdou, khenchelae from Khenchela, albinervis 

 from Sebdou and Geryville, and the doubtful fumosa (i.e. doubtful to me) from 

 Gäryville. Our specimens represent one single species, the name of which is 

 S. chretieni, if my identification of Oberthür's figures is correct. 



S. chretieni varies much less in size than S. aegrota and S. infuscata, and is 

 always large, the forewing measuring 12 to 14 mm. in our series. The wings are 

 more uniformly brown than in S. infuscata, the veins being hardly darker than the 

 ground. Oberthür's albinervis is presumably based on specimens in which the 



