11^ [May, 



19. Thorax not constricted before base ; humeri evident. 



16. Rabocertjs Muls, 



Thorax constricted before base ; humeri not developed. 



17. PSEUDORABOCEIIUS Pic, 



20. Rostrum about as long as wide, scarcely wider towards apex. 



19. ViNCENZKLLUS Reitt, 



Rostrum longer than wide, widened towards apex 21. 



21. Antennae with last 6 joints enlarged ; rostrum feebly widened towards 



apex 20. Cabiderus Muls. 



Antennae with last 4 or 5 joints enlarged ; apex of rostrum much widened. 



21. Rhinosimus Latr. 



1. IsiSTEiA Lewis. 

 A monotypic genus from Japan, I. rufobrunnea Lew. 



2. SosTHENES Champ. 



Type, 8. dyschirioides Champ.; Guatemala. 



Salpingus niponicits Lewis and Lissodema myrmido Mars., both 

 from Japan, have an abrupt 3-jointed club to the antennae, and the sides 

 of the thorax rounded, not denticulate ; I am unable to separate them 

 generically from Sosthenes. 



3. Neosalpingus Blackb. 



Blackburn says of this genus ; " from Lissodema it is at once sepa- 

 rated by its very different tarsi, which are shorter and stouter, with the 

 apical two joints conspicuously more slender than the rest, which are 

 densely clothed beneath with long hairs, the basal joint of the hind tarsi 

 being scarcely longer than the second." This tarsal structure I believe 

 to be merely a sexual character of the male and without generic signifi- 

 cance. In the iy^Q of Blackburn's lirst species, N. corfical/s, it is 

 scarcely evident, and a second specimen in the B.M. collection with 

 similar tarsi is evidently a female with the ovipositor extruded. In the 

 tj'pe of iV. dentaticolUs it is much more evident, though even here 

 the tarsi are not so strongly expanded as in ^S*. aeratus Muls., J- 

 Again, the British Museum possesses a specimen, also from Adelaide, 

 indistinguishable from the type except that it has simple tarsi ; this I 

 take to be the $ N. dentaticolUs. JSfeosalinngits Blackb. therefore 

 merges into Lissodema, except that in N. corticalis the sides of the 

 thorax are not denticulate, as they are in N. dentaticolUs. If we retain 

 N. corticalis as the type, the genus may be maintained on this latter 

 character, Salpingus hyhridus Er. being included in it, while N. denta- 

 ticolUs Blackb. is referred to Lissodema. The other species placed by 

 Lea (Trans. Roy. Soc. S. Austr. xh, 1917, pp. 159-161) in Neosalpingus, 



