498 Annals of the South African Museum. 



front of the constriction (i. e. the posterior articulation of the node). 

 There is a very distinct transverse impression in the middle of the 

 mesonotum which does not occur in the type. In addition, the abdomen 

 has a fairly long pubescence, whereas it is very short and more 

 scattered in the type. The colour is yellower and duller, but similarly 

 distributed. Otherwise identical." 



The transverse impression on the mesonotum is absent in the 

 smaller $ £ , and even in some of the larger ones. The epinotum is 

 armed with very short, triangular teeth, somewhat variable in length, 

 but not longer than their basal width. The scapes do not reach the 

 occipital margin, falling short of the same by as much as the apical 

 width of the scape. The dorsal face of the 1st node is distinctly con- 

 cave from side to side, more so than in castanea, tricolor and ferruginea. 

 The median tubercle of the mesonotum is very prominent, especially 

 when seen from the side, and the dorsum of the epinotum is convex, 

 short, hardly more than half the length of the declivity. 



Fig. 33. — C. Neuvillei, race Cooperi. 



9 . 9 mm. (hitherto undescribed). Castaneous brown, the petiole 

 a little brighter, the apical margins of the abdominal segments dirty 

 ochreous and translucent. Sculpture, pilosity and pubescence as in 

 tricolor, but the epinotum above and at the sides is duller and more 

 densely rugulose. The epinotal teeth as in tricolor, the declivity of 

 the epinotum shorter. The scapes are shorter and thicker, not 

 extending back beyond the posterior ocelli, also more strongly curved 

 near the base. The 1st node is twice as wide as long, more deeply 

 excavated in front than in tricolor ; 2nd node feebly grooved 

 posteriorly. 



Durban. (S.A.M., E.M., G-.A. colls. ; type of $ in my collection.) 



Eace Cooperi, var. ingravis, Forel. 



Loc. cit., p. 341, ^ • 



$ . 3*1-4 "8 mm. Much darker than the race, of a warm reddish- 

 brown. The sides of the head are a little less convex. This form 

 hardly merits separation as a variety ; I am unable to confirm the 

 other differences noticed by Forel. 



Durban (L. Bevis and C. B. Cooper). (S.A.M., E.M., a. A. colls.) 



