215 



swammerdami, Forel, of Madagascar, as its type. This form 

 and several other paleotropical and neotropical species have 

 the head in the worker and female constricted behind to form 

 a narrow neck, whereas another series, also with a single 

 cubital cell in the wings (cockerelli and albisetosa of North 

 America, sagei of the 'Himalayas, and longiceps), do not have 

 the head narrowly constricted behind. Forel preferred to 

 retain these species in the subgenus Deromyrma, because he 

 attributed more value to the wing venation than to the shape 

 of the head (Rev. Zool. Afr., 2, 1913, p*. 350). Viehmeyer 

 continued the subdivision of the genus .1 phae nogaster by 

 separating out the Papuan species dromedarius, loriae, and 

 quadrispina, as a distinct subgenus, Planimyrma (Zool. 

 Jahrb. Abth. f. Syst., 37, 1914). More recently Emery has 

 carried the subdivision considerably further. He regards the 

 Mediterranean species of the testaceopilosa group as constitut- 

 ing the subgenus A phae nogaster sens, str. (with sardoa, Mayr, 

 as the type). These forms have small females, with narrow 

 thorax, which seem never to bear wings. The large series of 

 circumpolar species allied to subterranea, Latreille, with 

 large, winged females and two cubital cells, Emery places in 

 a new subgenus, Attomyrma. The North American species, 

 cocherelli and albisetosa , he assigns to another subgenus, 

 Novornessor. Although the males and females of the Aus- 

 tralian species and the Himalayan sagei have only one cubital 

 cell, he prefers to place them in Attomyrma rather than in 

 Deromyrma. While I agree that the Australian species 

 should be removed from the subgenus Deromyrma, I do not 

 believe that they should be assigned to Attomyrma. The 

 three species form a very compact group, characterized not 

 only by the presence of only one cubital cell, but also by 

 their geographical isolation and the size of the female, which 

 is much larger in proportion to the worker than in any of the 

 other species of Apha enogaster . They may, therefore, pro- 

 perly constitute a distinct subgenus, for which I propose the 

 name Nystalomyrma, subgen.. nov., with Myrmica longiceps, 

 Smith, as the type. This leaves only the Himalayan sagei 

 without a definite position. It may be placed in Attomyrma, 

 as Emery suggests, especially as the female is only slightly 

 larger than the largest worker forms. 



It seems necessary to dwell at length on the taxonomic 

 matters, because the three species of Aphaenogaster, and 

 particularly longiceps, are among the commonest or most 

 conspicuous ants in the Eucalyptus forests, and therefore 

 familiar to all Australian entomologists. The species of 

 Nystalomyrma are insectivorous, like most other species of 

 Aphaenogaster, and live by preference in sand or sandy soil, 



