no. 2357. SOME WESTERN FLUVIAL FISHES— SNYDER. 27 



tification. The type of T. lineata is lost and the locality in which it 

 was found is unknown. One pharyngeal arch is preserved (Cat. No. 

 2783, U.S.N.M.). The teeth are in two rows, 2-4, and they closely 

 resemble those of R. hydroplilox. Girard writes of T. lineata: "The 

 general aspect is elongated, the body being subfusiform, anteriorly 

 thickish, and quite tapering posteriorly." Also the same author 

 remarks of his T. egregia: "By its general aspect this species resem- 

 bles T. lineata." It is quite clear, therefore, that T. lineata was a 

 fish of slender, graceful form like R. egregius or R. hydroplilox, while 

 S. atraria is a deep-bodied, thick-tailed form, with a comparatively 

 arge head. The name Richardsonius atrarius should, therefore, 

 stand for the common chub of the Bonneville basin, where it is very 

 generally distributed, inhabiting both lakes and streams, and where 

 it is the species which was apparently able to hold out longest in 

 bodies of water which have dwindled during the slow desiccation 

 which followed the Quaternary period. 



Acomus generosus was said to have been taken in the Bonneville 

 basin, and the writer of the present paper unwittingly accepted the 

 statement in a recent brief review 6 of the genus to which it is sup- 

 posed to belong. In the description of the species, Girard 6 gives the 

 locality "Cottonwood Creek, an affluent of the Great Salt Lake of 

 Utah." In the table presented on a previous page it will be noted 

 that other species from the same locality have been identified with 

 Mississippi basin forms, and no one would now presume to assign 

 species of Notropis or Lepomis to the Great Basin. Clearly, then, 

 in the case of at least four or possibly five nominal species a mistake 

 was made in the locality. There is evidence that at four places in 

 Beckwith's travels a Cottonwood Creek was approached. Two of 

 these creeks are named and located in his maps of the route and two 

 were apparently unknown by that name. The banks of the first one 

 encountered furnished a camp site for two nights and a day early in 

 July. This creek is a tributary of the Arkansas, and at the point 

 crossed by the old Sante Fe trail is no doubt the type locality of A. 

 dulcis, C. gunnisoni, C. lugubris, and C. ludibunda. At this place it 

 would also be possible to collect specimens of B. Tiumilis and A. dul- 

 cis. The second Cottonwood Creek noted by Beckwith was passed 

 on the 8th of November, 1S53, after camping there on the previous 

 night. The party was then without a naturalist, and it was quite 

 probable that no collecting was done. This stream is Girard's afflu- 

 ent of the Great Salt Lake of Utah, and, as is now evident, an im- 

 possible locality for the species which he assigned to it. Returning 

 to Acomus generosus, the t} 7 pes are two specimens measuring about 

 3* inches (Cat. No. 256, U.S.N.M.). The head is short, the body 

 robust, the lips small and not very pendent. The skull is thick; the 

 fontanelle completely obliterated. There are 79-81 scales in the 



• Proc. U. S. Nat. Mus., vol. 49, 1915, p. 575. 



