196 



[September, 



having, very generously, presented me with a mounted preparation of his 

 species, I find that they are amply disi'mct, form/ ceticohi being without 

 the conspicuous submarginal series of pores that characterizes lonr/icornis. 



The tarsi of the anterior limbs, as shown in the accompanying 

 figure, are seen to be more or less completely bisected by a deep fold, 

 starting from the dorsal edge and usually extending to a depth of from 

 half to three-quarters of the limb {a, h) ; in extreme cases, however, the 

 fold is continued, very faintly, to the ventral edge of the limb (c, d) ; 

 but there is nothing that can be described justly as an articulation. It 

 is noticeable that the division occurs at varying parts of the member ; it 

 may be at or near the middle (as in a and h), nearer the distal extremity 



a b c d 



Fig. 8. — Exaeretopus longicornis. a-d, various forms of anterior tarsi, X 220. 



(as in c), or nearer the base (as in d). Moreover, there are often other 

 incipient folds, marked b}^ small indentations on the edge of the limb, 

 which in some cases (as in b) may extend inwards to a depth almost 

 equalling that of the princijjal fold. Similar incipient folds are to be 

 seen on the tarsi of the other limbs, more especially upon those of the third 

 pair ; and, in one example, a nearly complete division — almost compai-able 

 with that shown at (d) — occurs on one of the tarsi of the hinder limb. 

 In another example I have observed a fold on the hind tarsus exactly 

 comparable with that shown at (a). All the tarsi in both formiceticola 

 and longicornis have an unusually crumpled character, which confirms 

 me in my view that the apparently dimerous condition of the anterior 

 tarsi is merely the result of an exaggerated fold. 



The name lonqicornis^ appn)priate enough so long as the species 



