40 Annals of the South African Museum. 



segments is feebly indicated dorsally. Sting very long, three- 

 quarters the length of the abdomen. 



Coxae, especially of the 1st pair of legs, very large. Anterior 

 femora and tibiae flattened and wide. The intermediate tibiae 

 short, only half as long as the femora ; the posterior tibiae also 

 short, but a little longer than the intermediate. The 2nd, 3rd, and 

 4th joints of all the tarsi wider than long ; the apical angles of all 

 the tarsal joints produced into stout, blunt spines ; the posterior 

 metatarsi with three or four stout spines, the intermediate tibiae 

 with two rows of spines on their external margins. 



? . 7 mm. Similar to the $ but much darker, the 1st abdominal 

 segment dark reddish brown, the apical margin lighter ; the 2nd-4th 

 segments piceous, the 5th ferruginous in the middle, piceous at the 

 sides. The margin of the pronotum is much more prominent than 

 in the £ , the mesonotum is flatter, raised more above the pronotum 

 and is vertically truncate in front. The scutellum is distinct, black, 

 shining, and impunctate, and separated from the mesonotum by a 

 shallow sulcus. The metanotum is very short ; epinotum not 

 forming a hump as in the g , the dorsum short, the declivity steep, 

 rounded above and at the sides. 



Eyes and ocelli present ; the former moderately large, their 

 distance from the base of the mandibles being about two-thirds of 

 their own length ; the ocelli are set far back, close to the occipital 

 margin. The legs are proportionately more slender than in the £ . 

 "Winged (de-alated in the single specimen in my possession). 



Bulawayo, Bembesi, under large stones. The genus is termi- 

 tophilous, but this species does not appear to be so, since it was not 

 found in or near the nests of termites. The species cannot be 

 mistaken for any other Ponerine ant of our region, on account of 

 the striking fossorial structure of the legs, and the marginate 

 pronotum. This species is apparently closely related to C. sellaris, 

 Mayr, from the Cameroons, although not having had an example of 

 the latter for inspection, it is not easy to compare the two species, 

 since Mayr's description of sellaris is based mainly on a comparison 

 between it and C. Feae, Emery, an Indian species. Judging, how- 

 ever, by the descriptions of these two species, Constanciae has the 

 following characters to distinguish it from sellaris. The colour is 

 much darker ; the striae below the antennal sockets do not extend 

 back to the posterior angles of the head, hardly passing beyond the 

 middle line of the head ; the limits of the declivity above (seen from 

 behind) are distinctly marked off from the humped shaped portion 

 of the dorsum of the epinotum. The declivity is also distinctly 



