36 



Attiini (notably species of Sericomyrmex and Aptero- 

 stigmaJA 2 ) These peculiarities, and the fact that the colony 

 was completely shut off from the outside world by a layer of 

 mycelium, strongly suggest that Prodiscothyrea velutina is a 

 fungus-growing and fungus-eating ant. Should further 

 research prove this supposition to be correct, the insect would 

 be of considerable interest as the first case of a fungus- 

 gardener among the ants of the Old World, and the first 

 to exhibit the habit among the members of the subfamily 

 Ponerinae. As the Ponerinae are by common consent 

 regarded as a very ancient and primitive group, we should 

 have to infer that the habit was probably of great antiquity, 

 and had been acquired independently at least twice during 

 the phylogeny of the Formicidae. The possibility of a third 

 independent development of the habit is, indeed, suggested 

 by the extraordinary Myrmicine ant Proatta hutteli, recently 

 described by Forel from Sumatra, because it is structurally 

 very similar to certain species of the Attiine genus Mycoce- 

 purus, but von Buttal-Reepen did not find it associated with 

 fungi, but in the carton nest of a termite (Hamitermes 

 dentatus, Harv.), which does not cultivate fungi like many 

 of the Old World termites. 



Prodiscothyrea evidently belongs to Emery's tribe 

 Proceratiini, and is very closely related to Discothyrea. In 

 this genus, however, the antennae are 9-jointed* the clypeus 

 forms a semicircular disc instead of a very short, transverse 

 plate, the frontal carinae are very much smaller and shorter, 

 there are no large impressions on each side of the head for 

 the accommodation of the antennal scapes, the eyes are smaller 

 and at the middle of the sides of the head, and in the female 

 the eyes and ocelli, judging from the descriptions, are much 

 larger. Like Discothyrea and Spaniopone, Prodiscothyrea is 

 related to Bradoponera of the Baltic amber, which is of the 

 lower Oligocene age. The Australian genus is evidently an 

 ancient Mesozoic genus, like Paranomopone, which I recently 

 described, from the same locality in Northern Queensland 

 (Psyche, xxii., 1915, pp. 117-120, pi. 1). The great antiquity 

 of the tribe Proceratiini is attested by its wide geographical 

 range and the sporadic occurrence of species of Procerativm, 

 Discothyrea, and Sysphincta in widely separated regions of 

 both hemispheres. 



(2) The peculiar structure of the toothless mandibles, with the 

 dense, regular row of short setae along their apical borders, is also 

 suggestive of habits which require the manipulation of such delicate 

 bodies as fungus hyphae, though I know of no similar structures 

 in our American fungus-growing ants. 



