386 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.xxvii. 



dislocated toward the median line and directed straight backward. 

 Some of our species {Idahoensis, canadensis^ petlolatus^ arizonensis^ 

 hrunei'i, arctomys, taherculatiis, and hirsutus) are t3'pical CeratophyUus 

 as defined by Wagner. Some were found which, \\k.&j[)roxirm(s, had 

 the first pair of spines only slightly dislocated and bent inward, and it 

 was puzzling indeed to find that lucidus and charlottensls were t^^pical 

 Ctenophthalm'us on the fore legs, while the former was a genuine 

 CeratophyUus on the hind legs and the latter a Pulex! Then occurred 

 the peculiar aberrant perpinnatus with a middle pair of the spines 

 dislocated -toward the center. 



After all this I was prepared for the group of certainly Ceratophyl- 

 lus species, which ha^ the first pair of spines dislocated, as in lahiatus^ 

 keeni, pseudarctomys, calif o7'nicus, ciliatus, wagneri, ignotus, divisiis, 

 coloradensis, oculatus, wicJihaml, and sexdentatns^ all of which are cer- 

 tainh^ to be regarded as more clearly congeneric with the type of Cera- 

 tophyUus rather than with that of CtenopJithxxlin us. For the time being 

 there was but the one recourse of falling back upon the artificial (?) 

 character of the presence or absence of a genal ctenidium, and so far as 

 my studies have progressed it is the sole means by which I can sepa- 

 rate these numerous species into two more or less homogeneous groups 

 about the original t} pes. The onh' alternative would seem to be the 

 establishment of numerous genera, which will eventuall}" have to be 

 done, but which would seem to be unwise in the present very frag- 

 mentary condition of our knowledge of the existing species of the 

 world. 



All of the species of CeratopliyUus, so far as I have examined them, 

 have on the inferior disk of the fifth tarsal joint numerous very min- 

 ute hairs, and between the hind coxa and its epiphysis distally there is 

 always formed a more or less deeply excavated emargination. In 

 some other genera there are wide departures from these conditions 

 though their uniformity has yet to be tested for all the species. 



SYNOPSIS OF AMERICAN SPECIES. 



a. Hind coxte with one or more rows of minute teeth on inside distally. 

 h. Eyes well developed; teeth on inside of hind coxae in several rows. 



c. Pronotal ctenidium of 40 spines vndtispinosus (p. 389) . 



cc. Pronotal ctenidium of 24 spines dentatiis (p. 390) . 



bb. Eyes rudimentary; teeth on inside of hind cox« in one row; pronotal ctenidium 



of 14 spines charlottensis ( p. 390) . 



««. Hind coxae without minute teeth on inside. 



b. Fifth tarsal joint with the middle pair of lateral spines dislocated toward median 

 line and replaced by two supernumerary spines; pronotal ctenidium of 36 



spines }>erpinnntug (p. 391 ). 



bb. Fifth tarsal joint never with middle pair of spines dislocated; pronotal cteni- 

 dium of 26 spines or less, 

 c. Males without a strongly developed style projecting over pygidium; size 

 medium to small; eyes usually present. 



