202 CONTRIBUTIONS TO NORTH AMERICAN ICHTHYOLOGY III. 



outlines somewhat serrate; lateral Hue well developed, nearly straight, 

 with 35 to -12 scales, 12 to 14 in a cross-series from ventrals to dorsal ; 

 dorsal tin beginning near the middle oi' the body, somewhat in advance 

 of the veutrals, its an tenor rays elevated, their height about equal to 

 half the base of the fin, the number of rays in the dorsal fin ranging 

 from 25 to 32; caudal fin well iorked, the lobes about equal, not fal- 

 cate; anal fin comparatively long and rather low, of 8 or 9 developed 

 rays; ventrals moderate, 10 rayed ; pectorals rather short: sexual pecu- 

 liarities, if any, unknown : coloration dull dark brown, nearly plain, not 

 silvery ; fins olivaceous or more or less dusky. 



Air-bladder with two chambers. 



Size quite large. 



In general appearance, the species of Bubaliclithys bear a considerable 

 resemblance to those of Carpiodes. The form is, however, coarser than 

 that of any Carpiodes, the dorsal fin is lower, and the coloration is 

 darker and duller. The species reach a larger size than do those of 

 Carpiodes, but whether larger or not than the species of Iclitliyobus I am 

 unable to say. In external appearance, Bubaliclitliys is intermediate 

 between Carpiodes and Ichthyobus, the one species, bubalus, resembling 

 Carpiodes most, the other, urus, being most like Ichthyobus. 



Our knowledge of the species of this genus is very incomplete. Many 

 species were named and indicated by Professor Agassiz, but with such 

 fragmentary descriptions that not a single one of them is certainly 

 known by any one. I have, however, been able to identify in specimens 

 from Quincy, 111., the fishes termed by him B. bubalus and B. niger, 

 the small-mouthed and the large-mouthed Buffalo. Assuming these two 

 well-separated species as a basis, I have compared with them numerous 

 Buffalo-fishes from various localities, and in all cases I have found them 

 identical with either the one or the other. I have therefore adopted 

 the hypothesis, possible, and perhaps probable, that all of the nominal 

 species of Professor Agassiz were based on the one or the other of these 

 two forms. As to this, I may say that the sole basis of some of these 

 nominal species was the difference in locality. From what we know of 

 the range of other species of Catostomidce, there is nothing antecedently 

 improbable in the same fish being found in the Wabash and Mobile 

 Kivers, or in the Tennessee and Osage. Hyxostoma macrolepidotum, 

 Erimyzon oblongus, Minytrema melanops, Catostomns teres, and others 

 are known to occur in all four of those streams. The questions of 

 locality may, I think, be safely eliminated from the discussion. The 



