76 FORMIC1DJE. 



long, filiform, the scape extended beyond the top of the head. 

 Thorax above rounded, slightly compressed posteriorly, pronotum 

 strongly convex and as broad as the head ; pro-rnesonotal suture 

 always distinct ; legs long and stout ; femora and tibiae cylin- 

 drical, claws simple. Pedicel one-jointed, the node on a level 

 with or higher than the rnetanotum, arched and generally very 

 convex in front, flat or a little concave posteriorly, above ahvays 

 armed with two spines ; abdomen cylindrical, the constriction 

 between the basal two segments well-marked. Head, thorax and 

 pedicel ahvays, and a portion of the abdomen also in the majority of 

 species, very beautifully and regularly striate. 



d . Similar to the $ but more elongate. Head globose, wider 

 than the thorax ; mandibles short, cylindrical, acutely pointed at 

 apex ; antennae filiform, scape verv short, about half the length of 

 2nd joint of the flagellum ; eyes and ocelli large and very prominent, 

 the space between the latter equal to the space between the orbit 

 of the eye and the lateral ocelli ; thorax very gibbous, the meso- 

 notum with a Y-shaped furrow ; the scutellum large, convex, 

 separated from the mesonoturn by a crenulate transverse impressed 

 line ; postscutellum and metanotum depressed, oblique, rounded 

 above ; legs very slight, elongate, femoi'a slightly compressed, claws 

 simple. Node of the pedicel more or less conical, a little flat 

 above, without spines ; abdomen elongate, rather massive for males. 

 The beautiful regular striae on the head, thorax and pedicel, con- 

 spicuous in the $ , feeble and not distinct. 



$. Unknown*. 



Mr. Eothney (Tr. Ent. Soc. 1889, p. 361) writes very fully on 

 the habits of D. vagans, Smith. I agree with him that the species 

 of this genus are by far the most intelligent of all Oriental ants. 



Key to the Species. 



A. First abdominal segment not striate. 



. Node of pedicel very strongly laterally 



compressed, much longer than broad . . I). scalpraium, p. 77. 

 b. Node of pedicel not laterally compressed, 



quite as broad as long D. cyantientre, p. 78. 



B. First abdominal segment striate. 



a. Pronotum longitudinally striate D. assamense, p. 79. 



b. Pronotum with transversely arched striae, 



which are more or less concentric an- 

 teriorly. 

 a'. First abdominal segment with the striaa 



anteriorly in concentric arches, in the 



middle divergent D. ceylonense, p. 79. 



* In vol. xii, p. 756, of the Jour. Bomb. N. H. Soc. I gave an account of the 

 finding of an ant I thought was the $ of D. vagans, Smith. I regret to say 

 that 1 was mistaken ; further examination of the one specimen I kept has 

 shown it to be only a large <$ ; the $ of Diacamma remains, therefore, still 

 undiscovered. 



