ARMY ANTS IN BRITISH GUIANA. 319 



tion of the cheeks and the distinctly advanced clypeus in the smaller 

 workers remind one of Dorylus, and the simple pygidium, the toothed 

 claAvs and the antennpe are Ecitine, whereas the structure of the 

 mandibles of the soldier and the scale-like transverse petiolar node 

 are peculiar to the genus. In the male the venation (Fig. bh) and 

 toothed claws conform to the Ecilon type and the flattened femora, 

 especially the hind pair, and the shape of the antennal funiculi are as 

 strongly reminiscent of Dorylus, while the structure of the mandibles, 

 though sickle-shaped and toothless, and more like those of certain 

 Ecitons, ate nevertheless very singular. The genitalia (Fig. 10) con- 

 form in the main to the Eciton type (Fig. 6) in having the stipites 

 articulated and not fused with the basal annulus, but their shape is 

 more like those of certain Dorylus. The structure of the subgenital 

 plate (Fig. lOc/), however, with its four terminal teeth is peculiar. 

 Emery (1910, p. 16 nota) believes that Chcliomyrmcx is "very close 

 to Eciton, subgen. Labidus and especially to E. coecum," and there- 

 fore includes it in the tribe Ecitini. I am unable to follow him in 

 this procedure since I believe that the resemblance to caecum is super- 

 ficial and illusory and due to similarity of habits (convergence) and 

 that the genus Chcliomyrmcx should constitute an indepentlent tribe,, 

 the Cheliomyrmicini. It is, in fact, the most archaic and generalized 

 of all the tribes of the Doryline subfamily, just as the Leptanillii 

 constitute the most specialized, or degenerate tribe. Undoubtedly 

 Chcliomyrmcx is a very ancient group, very near the ancestral stem 

 from which both the Dorylini and the Ecitini sprang and diverged, 

 possibly during the late Cretaceous. The discovery of the female 

 will probably yield additional arguments in favor of this contention. 

 C. mrgalonyx was first found at Kartabo July 20, on the edge of the 

 jungle and just back of the laboratory, where it was foraging in 

 columns under prostrate logs. At first sight this species may be 

 mistaken for E. coccum, but the columns are denser and more popu- 

 lous and the soldiers and workers exhibit much less variation in stature. 

 Cheliomyrmex seems also to be even more photophobic than coccum, 

 disappearing at once into the soil when the light is let into its galleries. 

 It stings much more painfully and will attack fiercely when inter- 

 cepted in its movements. On several successive days I saw detach- 

 ments of the same army under logs in the same locality. There were 

 no larvae nor pupse and very probably w hat I saw were merely hunting 

 columns of a huge colony which I failed to locate till a month later. 

 August 21, on \'isiting the taxidermist's hut, behind the lal)oratory 

 and less than a hundred yards from the spot in which I first found the 



