322 WHEELER. 



burchelli and 39 of a form very close to E. morosus but of a paler, more 

 reddish brown color and with somewhat longer mandibles. No other 

 Dorylines appeared at the lights during my stay at Kartabo. I con- 

 clude that these 39 males had escaped from the CheUontyrmcx colony 

 just back of the laboratory for the following reasons: First, their 

 flight coincided with the sojourn of the large Cheliomyrmex colony 

 behind the laboratory. Second, the laboratory is on a point of land 

 at the confluence of two great rivers, the Mazaruni and Cuyuni, which 

 cannot be crossed by such feeble flyers as these male Dorylines. 

 Furthermore, the building is screened on the sides from the rivers 

 by huge clumps of bamboo. Hence the males must have come from 

 the imxnediate vicinity behind the laboratory. Third, my days of 

 intensive search for ants about the laboratory convinced me that the 

 only Doryline with dentate claws and unknown male in the immediate 

 \acinity was the Chciiomyrvicx, the males of the other species encoun- 

 tered {E. burchelli, hamatum and coecum) being known. Fourth, 

 both the males taken at lights and the soldiers and workers of the 

 Chdiomyrincx from the colony behind the laboratory difl'er specifically, 

 or at least subspecifically from morosus. I therefore regard all of 

 these phases as belonging to the same species, which may be dis- 

 tinguished from the Mexican and Honduran form as follows: 



Cheliomyrmex megalonyx sp. nov. 



Soldier (Fig. 9d, f). Very similar to morosus (Fig. 9a, c) but differ- 

 ing as follows: Color of body more deeply ferruginous red; head and 

 manchbles proportionally somewhat longer, the latter narrower in the 

 region of the median tooth, which is longer and narrower at its base. 

 The basal tooth is more acute and separated by a distinct diastema 

 from the median tooth, not arising from its base as in viorosus. Anten- 

 me somewhat longer, the median joints especially. Occipital border 

 of the head, seen from above more deeply excised and sharply mar- 

 ginate, the margination continuing down the inferior occipital angle 

 on each side as a sharp ridge on to the gular surface. In morosus 

 these lateral ridges are absent and the median margination of the 

 occipital border is feeble. Epinotum less swollen and convex, some- 

 what lower and more sloping than in morosus, its base and declivity 

 more distinct, subequal and nearly straight in profile. Petiole 

 slightly more compressed anteroposteriorly, the anterior and pos- 

 terior surfaces of the node more flattened, the sides and dorsal surface 



