A Monograph of Formicidae of South Africa. 617 



Sub-genus MYRMOTURBA, Forel. 

 For characters see table of subgenera. 



Qj.. In this sub-genus, the head of the Qj. is usually more or less 

 triangular and with the angles rounded, or subcordate ; the posterior 

 margin always more or less concave. The scapes are rarely long, 

 and in few of them do they extend beyond the hind margin by as 

 much as one-fourth of their length. In taking the measurement, it 

 should be noted that the length of the head is taken from the 

 anterior margin of the clypeus to the level of the posterior angles; 

 in measurements of the scape, that limb should be turned back 

 parallel to the long axis of the head, not directed obliquely to the 

 posterior angles. The pronotum in usually more or less convex 

 transversely and slopes downwards in front. In many species the 

 pronotum has two shallow impressed lines, one on each side of the 

 middle, oblique and convergent anteriorly. In the §§ these lines are 

 more or less parallel. They are inconspicuous as a rule, and not 

 easily seen unless the insect is held at a certain angle. 



The mesonotum is narrowed gradually backwards and the epinotum 

 is frequently widened towards the brow of the declivity. The thorax 

 is therefore always somewhat constricted at the meso-epinotal level. 

 Usually a trace of the metanotum is visible, but as this segment is 

 very small and inconspicuous, the suture between it and the epinotum 

 is referred to in the following pages as the meso-epinotal. The dorsum 

 of the epinotum is narrow, the sides converging above to form a 

 narrow and rounded angle (i. e. the actual dorsum) the dorsum 

 nearly always merging gradually into the declivity, very rarely 

 subangularly. The degree of convexity of the dorsal profile of the 

 thorax varies, but in many cases (maculatus and its races and 

 varieties), to so slight a degree that it is dj/ficuit to appreciate* The 

 scale of the petiole also exhibits slight variations, but on the whole 

 it cannot be relied on as a diagnostic character of much value. 

 Some authors have placed a great deal of emphasis on the absence 

 or otherwise of short spines or setae on the lower surface of the 

 tibiae and also on the shape of that joint. In cross-section it is 

 either round, quadrate-oblong or prismatic. The normal fundamental 

 sculpture of the sub-genus, as exemplified by C. maculatus, consists 

 of a very fine and even rugulosity on the thorax and abdomen, 

 somewhat reticulate in parts, and on the head of a reticulation which 

 merges into a rugulose sculpture here and there. The sculpture in 

 most of the species is barely appreciable under a magnification of 

 less than 1(3 diameters. The pubescence is conspicuously scanty on 



