1873.] <"" [Cope. 



and the S. curilus, especially in the rounded or oval outlines of the oper- 

 cular bones. From the former it differs in the larger head, much longer 

 pectoral fin, larger scales, &c. ; from the latter in the absence of spots 

 below the lateral line, absence of "two tubercles on the upper jaw," 

 stouter form, ic. 



Spratelloides bryoporus. Cope, sp. nov. 



Form compressed and moderately elongate. Belly rounded, except 

 between ventral fins and vent, where it is angular. Head five times in 

 length to end of scales at basis of caudal fin; depth 4.5 times in the 

 same. Eye 4.75 times in head, measuring without regard to vertical 

 palprebse. Maxillary bone broad, flat, extending to beneath the middle 

 of the pupil. Pectoral fin 2.5 times in the space between the base and 

 that of the ventral fin. Latter originating opposite a point a little in 

 front of the middle of the dorsal. Radii, D. I. 18; A. 18; scales 1. tr. 

 12; 1. 1.51. Anal radii short; ventrals rather short. Length, fourteen 

 inches. 



General color golden; the dorsal region dark, with blue reflections. 



The vertex of the head is occupied by a bifurcate depression, which 

 is occupied by a transparent thickening of the epidermis, which is filled 

 with delicate branching mucous tubules in great numbers. Opercular 

 and post-frontal regions filled with delicate, elongate mucous tubes. 

 Coast of Alaska. The first species of the genus of herring from the 

 Northern Pacific Ocean. 



HEMIBRANCHII. * 



The position of the Pegasidce has been for a long time a matter of 

 discussion, and without important result. Gunther, the latest writer, 

 places them in the Physontomi near the Lopliobranchii; the older authors 

 had regarded them as belon ;ing to the latter order. Steenstrup as 

 quoted by Kner (Sitzungsber. Akad, Wiss, Wien, xli., p.p. 821-22), thinks 

 them near to Aspidophorus, which is one of the Scyphobraneliii. I am 

 sure that the reference to the Hemibranchii is a step in advance of 

 previous conclusions, and for the following reasons : 



The characters which distinguish Pegasus most widely from Aspido- 

 phorus, are the following: The ventral fins are abdominal, the inter- 

 clavi.clcs are present ; the first and fourth superior branchihyals are 

 wanting ; the branchiostegal rays are nearly wanting, the centra, neural 

 arches and spines of the dorsal vertebrse are suturally united. It agrees 

 with the Hemibranchii in all of these characters except in the presence 

 of but one branchiostegal; but these are reduced in number in both 



* In the Transactions of the American Philosophical Society 1S71. 457 an account of the osteo- 

 logical characters of the two orders of eels, Enehelycephali and Colocephali, is given. 

 Under the head of the former, the Congridce and Anguillida are included, while a third is 

 added, the " Gymnot horacidCB. " This is now omitted, for it is synonymous witli the 

 Hfuraenidce of the succeeding order, as the characters given show. Its presence out of place, 

 is probably the result of a clerical mistake in not eliminating it from a previous MS. , written 

 before the distinction between the orders Enchelycephali and Colocephali was recognized. 

 As it was inserted under the latter head, its omission from the former was to be understood. 



In that essay it is also stated, that the suboperculum is wanting in the OstraciidCE. In sunn 

 of my specimens this is the case; in others it is present, but is a very thin lamina, easily lost. 



