400 BULLETIN UNITED STATES GEOLOGICAL SURVEY. 



horizontal in' position tban is usual; operpuluin striated; the head in 

 the typical specimen entirely scaleless. 



Gill-openings wide, the gill-membraues not much connected below. 



Scales quite small, not closely imbricated, seemingly partly imbed- 

 ded in the skin; lateral line obscure, nearly straight, the number of 

 scales included in it not ascertainable. Dorsal fin single, beginning 

 just behind the head, the length of its base just half the length of the 

 fish from snout to base of caudal. It consists of about eight flexible 

 spines, gradually increasing in length backward. There are about 

 fifteen soft rays, which are much higher than the spines, also increasing 

 in height backward. The condition of the specimen renders it impos- 

 sible to exactly count either soft rays or spines without danger of 

 breaking them. The above count, as well as that of the aifal and ven- 

 tral fins is only an approximation, made by counting the bases of the 

 rays. The last and longest ray of the dorsal is about two fifths of the 

 length of the fish from the snout to the base of the caudal. The anal 

 fin is a little shorter than the dorsal, and consists of about two spines 

 and some eighteen or twenty soft rays. The form of the fin is similar 

 to that of the dorsal, and the hinder part of the fin is similarly ele- 

 vated, the longest rays being about two-fifths as long as the fish. 



Caudal fin greatly elongated, rounded or lanceolate in outline, pro- 

 duced behind, its length nearly half that of the. rest of the fish. 



Ventral fins thoracic, apparently 1, 5, and apparently united into one. 

 I regret that this important matter cannot be certainly decided. When 

 first examined, the two were united into one lanceolate fin, but in hand- 

 ling they were split apart, apparently not naturally, but the possibility 

 remains that they were merely stuck together by some adhesive sub- 

 stance. At all events, the two were very ©lose together, and the bases 

 still seem to be connected. Pectorals rather narrow, apparently short; 

 the ends, however, broken ; their position as in percoid fishes. 



Colors vanished. In spirits, at present, uniform pale. 



Length of typical specimen two inches. It is in good condition, every 

 part being present, but like most museum specimens it has become ex- 

 tremely soft. I have not attempted to dissect the fish as I do not wish 

 to destroy or injure it, and I therefore leave for future investigation the 

 determination of its affinities, referring it provisionally to the Gohiidce, 

 on the ground of the apparent cohesion of the ventrals. 



Family CYPRINODONTID^. 

 Genus HYDKARGYRA Lacepede. 



4. — Hydrargyra siMiLis Baird & Girard. 



1853 — Hydrarfjyra similis Baird & Giuard, Proc. Ac. Nat. Sc. Phila. 389. 



Uydrargyra similis Girard (1859), U. S. audMex. Bound. Ichtb. 68, pi. 35, f. 1-8. 

 ? Fundulus similis Gunther (1866), Cat. Fishes Brit. Mu,s. vi, 323. 



Several small female specimens, stout and full-bodied. They agree 

 closely with Girard's description and figure, but not very well with 



