STUDIES IN THE FOSSORIAL WASPS. 725 



males being placed in Anthohosca in the Thynnidaj, and the 

 females in Cosila among the Scoliidse. Guerin used both 

 names in the same work, but although he placed Cosila 

 chilensis cori'ectly, with the Scoliicla^, he failed to see the 

 relationship to Anthobosca australasice which he had described 

 in an earlier portion of the same work and classed with the 

 Thynnidae. Smith in 1868 placed a single male in his 

 genus Dwior2}ho2)tera, which is a synonym of Cosila, but both 

 before and after placed other males in Anthohosca among the 

 Thynnida3. He also described both sexes of A. albomaculaia Sm., 

 which were taken coupled by Bates, as Myzine, but this work was 

 published after his death and without his i-e vision. Burmeister 

 also in 1876 correctly associated the sexes, but placed his species 

 in Myzine. Although three or four males had been coi'rectly 

 associated with the females, they do not appear to have beexi 

 connected in any way with the males described in the genua 

 Anthobosca till my revision of the Australian species of the 

 genus appeared in 1907. The best work on the genus was done 

 by Saussure in 1892; but he treated the males as unknown, 

 except in the case of chilensis, and did not connect Cosila with 

 Anthohosca. In Dalla Torre's great Catalogue, published in 

 1897, there is much confusion in regard to this genus, species 

 appearing under the genera Thynnus, Myzine, and Cosila. There 

 is much confusion over the genus in the recent papers of Cameron 

 on South African Hymenoptera. Ashmead treats the group as a 

 family, Cosilidse, not as a subfamily of the Scoliidag, which I 

 consider the more natural course. But he places Anthohosca in 

 the Thynnida? and Dimorpho])tera in his family Myzinidse. He 

 includes in his family Cosilidse several genera of doubtful 

 affinities, of which, in my opinion, Nursea should be treated as 

 an aberrant genus of the Sphecoidea, whilst Maurillus belongs to 

 the Pompilida3. The position of Sierolomorpha and Dicrogenium 

 seems to me very doubtful, but I have not seen specimens. I am 

 compelled to look on Anthobosca as the only genus which can 

 be placed in the subfamily Anthoboscinte with any certainty, 

 Ashmead states that tlie intermediate coxse in his Cosilidfe are 

 contigvious or nearly so ; but this is quite incorrect as to the 

 females, and even in the males the separation is quite distinct. 

 In my key to the species I have included several species which I 

 have not seen, one or two of which may possibly not belong to 

 the genus. 



The females are distinguished from other Scoliidee by the 

 absence of a deep groove between the two basal ventral segments 

 of the abdomen. The hind coxee are separated as in Tiphia, 

 not contiguous as in Elis ; but the intermediate coxae are less 

 widely separated than in either of those groups, though very little 

 less so than in Elis. The males are distinguished from all other 

 Scoliidfe by the unarmed rounded hypopygium — in this character 

 approaching most nearly to the Thynnidfe of the genus Eirone. 



