1885.] PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 71 



14. Etheostoma variatum, Kirtlaud. 



Skull narrow, tlw paiietal region strongly convex transversely, supra- 

 occipital and parietals with liue radiating striae. Supraoccipital large, 

 its crest obsolete. Sutures obscure. Scapula shorter and broader than 

 in Percina; shoulder girdle otherwise similar. Profile much as in Per 

 cina, highest at occiput. Lower pharyngeals narrowly triangular. Py- 

 loric cceca 3. Vertebrce 15 + 21 = 36. 



15. Etheostoma lepiduni, Baird & Girard. 



Skull essentially as in PJ. variaUim, the sutures more distinct. Pyloric 

 cceca 2. A^ertebra? 16 + 19 = 37. 



16. Etheostoma maculatum, Kirtland. 



Frontal region narrow. Parietal region narrow, very convex trans- 

 versely. Profile highest at occiput, essentially as in E. variaUim^ the 

 skull a little narrower. Pyloric cceca 4. Vertebrae 15 -^- 24 = 39. 



17. Etheostoma -whipplei, Girard. 



Skull essentially as in E. variatum, the ])rofile lower above the eyes. 

 Pyloric coeca 3. Vertebrai 15 + 21 = 36. 



18. Etheostoma flabellare, Ralinesqne. 



Frontal region longer and narrower than in any other species. Pari- 

 etal region very strongly convex transversely, surface of skull smooth ; 

 a trace of supraoccipital crest. Profile low, highest at occiput. Dif- 

 fers from Etheostoma variatum chiefly in the greater length and com- 

 l)ression of the head. Pyloric cceca 3. Vertebrit', 13 -f 20 = 33. Lower 

 pharyngeals very narrow. 



19. Etheostoma fusiforme, Girard. 



Skull essentially as in E. variatum. Pyloric coeca 3, VertebrsB 

 16 + 20 = 36. 



20. Microperca puuctulata, Putnam. 



Frontal region not very narrow. Parietal region transversely convex, 

 its bones smooth. N^o supraoccipital crest; sutures very distinct; skull 

 highest above posterior part of eye. Vertebr,® 14 + 16 = 30. 



So far as the skeletons are concerned, we seem to be justified in the 

 following inferences : 



1. The Etheostomatince are near allies of the Percince, and .should not 

 form a separate family. 



2. Thej" are among themselves closely related, and the extreme forms 

 are so connected by intermediate forms that they might with no great 

 violence to nature be regarded as forming a single genus. 



3. The species nearest allied to the typical Percina? is Percina caprodes. 

 This is the largest in size, and of the others in general those smallest 

 in size are most aberrant in structure. 



4. Those species which have usually been grouped together on ex- 

 ternal charaters agree in general in regard to the skeleton. 



