ERIC MJOBERG, PTINID.E. 3 



line. The joints are reduced to eleven, of which the nine 

 distal ones are strongly compressed laterally (note the simi- 

 larity to the Paussidce); the seeond joint is rudimentary, being 

 visible only from underneath, quite concealed from above by 

 the surrounding joints. 



Låter on (1912) Lea described another species, P. doli- 

 chognathus, but with only ten joints. It has the seeond joint 

 rudimentary and, as in P. laticornis Lea, only visible from 

 underneath. The last joint is, aecording to Lea's observa- 

 vations, »about as long as the eighth and ninth combined». 

 Apparently the antennae of this form are totally different; 

 that is the reason why I have assigned it to a new genus, 

 which I name Leaptinus in honour on the discoverer. 



Among the present material there is also a very curious 

 form with the antennae extremely compressed laterally. It 

 represents a new good genus, which I have called Paussoceros 

 on account of its shape of the antennae. 



In this form the seeond joint seems to be entirely reduced, 

 the third (the real third, but the seeond in number) is much 

 broader than the first, the fourth to the ninth are very 

 broad and leaf-like, bisinuate at the proximal margin. The 

 ninth (in number the eighth) is at the apex strongly sinuate, 

 where it receives the very small terminal joint. 



The differentiation in the way of reduction has proceeded 

 still further in the old genus Edrephes Westw. Three species 

 were hitherto known, the present material containing a new 

 form, bringing the number up to four. 



By careful examination of the genus we soon find that 

 it by no means proves to be a homogenous or natural one. 

 As Lea (Proc. Roy. Soc. Victoria 23 (N:o 8), Pl. I. 1910, p. 

 222) — very properly, it seems to me — remarks, the three 

 hitherto described species »should in fact be regarded as 

 belonging to three genera». That is what I have done here, 

 keeping the first described species E. formicamm Pascoe as 

 the type for the old genus Ectrephes, and rassigning the two 

 other ones E. Icingii Westw. and E, pascoei Westw. to a 

 new genus each, Mesecirephes and Monectrephes respectively. 

 The new form described below is apparently allied to E. 

 Pascoei Westw. and a seeond species of that genus. 



The most primitive or the least differentiated is undoub- 

 tedly the genus Edrephes Westw. A tendency of forming a 



