25(3 LNovembor, 



The antennae are not clavate, but are of anomalous form, very 

 broad, and increasing in width from the third to the eleventh joint . 

 they are inserted at the sides of the head, the point of insertion being- 

 concealed by the projecting margin. The head is short, with very 

 large eyes^ which encroach greatly on the under-surface of the head ; 

 the gense project as a sort of toothed process on each side of the 

 mentum. The parts of the mouth can be scarce!}^ seen, but they do 

 not apparently possess any great peculiarity ; the palpi are, however, 

 probably unusually short. 



The prothorax exhibits a most remarkable character, surpassing 

 perhaps in interest even the genu's Nematidium ; this consists in the 

 fact that the lower and hind part of the prosternum is formed by the 

 junction behind the coxae of the side pieces which meet in the mesial 

 line ; the base is further augmented by a remarkable development of 

 the membrane, which borders the hind margin. The front coxae look 

 as if they were quite contiguous, but they are really separated by a 

 lamina immersed between them, and which, behind the coxae, expands 

 to form a compressed prosternal process reposing on the conjoined 

 sides. The mesosternum is unusually elongate, so that a much 

 greater longitudinal space than is usual intervenes between the front 

 and middle coxae. The metasternum is very elongate, and the middle 

 and hind coxae are but slightly separated. There are no inflexed 

 epipleurae, but at the base the wing-cases are edged with a white 

 membrane, which, at the end of the first ventral segment, gives place 

 to a crenate margin. 



I know of only two other genera that can be considered as at 

 all "related to " Mecedanum, one of them is Gempylodes, Pascoe, which 

 is a rather close ally, and presents a similar anomalous sternal 

 structure, but in which the front coxae are really contiguous. The 

 other is the genus Endestes, Pascoe, but here the relation is a distant 

 one, the front coxae being only a little exserted, but distinctly separa- 

 ted, while the prosternal process reaches to the posterior margin of 

 the epimera, which, so far as can be seen, are not conjoined, though 

 their membranous borders meet and overlap behind the process as 

 they do in Gempylodes and Mecedanum. This genus, however, also 

 differs in the shorter tarsi, the basal joint of whieii is but little longer 

 than the second. 



These insects should, in my opinion, form a separate subfamily 

 in ColydiidcB, which may be called Gempylodini, and characterized as 

 follows : — 



AntenncB absque clava, extrorsum crassiores, !<uh frontis margine laterale 



