( 573 ) 



a set of three or four, of which one is somewhat prolonged, there being also some 

 additional minute bristles at the ventral edge farther p^oximall}^ 



The frons is a trifle longer than in D. h/piisus. The meso- and metanota bear 

 five rows of bristles, the former having a few additional bristles dorsally near the 

 base. The mesopleura are studded with eighteen bristles, the metepimeram with 

 twenty-one to twenty-four, and the first abdominal sternite with three to nine. The 

 abdominal combs consist of the following spines in the three mounted specimens, 



Fig. 38. — Diimpsyllus gnjpuriis. 



the figures in brackets referring to the combs of the other side of the same 

 specimen : 3, 2, 4 (2, fl, 5)- 4, 4, 4 (4, 4, 4)— 3, 4, 6 (5, 3, 4). 



Length (mounted specimens) : S 3-0-3'5 mm. 



4t?c? from Mntaragwa, Aberdare Mts., British East Africa, March 24, 1910, 

 off Dendromijs nigrifrons ; — type. 



1 (? from Mt. Kinangop, Aberdare Mts., March 1, 1910, off Crocidurafumosa. 



36. Dinopsyllus eremus spec. nov. (text-fig. 39). 



¥ . Two females obtained by Mr. Kemp in Belgian Congo and one in British 

 East Africa differ so much in the receptaculuiu seminis from all other females that 

 we have no doubt of their specific distinctness. In all the females, inclusive of 

 D. echinus and hirsutus, the receptaculum has a prominent hump on the upper 

 side of the " head," and apparently does not present any differences in the various 

 species. On the other hand, the receptaculum seminis of D. eremus has no hump 

 whatever (text-fig. 39). 



The sj)ecimen, moreover, is distinguished from D. li/piiituK and apistiii by the 

 slightly different position of the vestigial frontal tubercle. In the two species 

 mentioned the tubercle is situated in the ? , a little nearer to the frontal oral 

 corner than to the hind edge of the frons, wiiile in D. eremus the reverse is the case. 

 The occiput measured along the dorsal line is twice as long as the distance of 

 the antennal groove (on the vertex) to the frontal tubercle in l>. eri'iniM, and 

 only about half as long again as that distance in D. lypiiHUx and ci./>i.t>uii. The 



