188ti. 1 PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 323 



New-Zealand coast. Structurally they are identical; only some tenta- 

 cles are developed in the American specimen above the eye and ou 

 some parts of the body. Tlie coloration is a blackish -brown, marbled 

 with lighter browu and gray. These differences are not sufficient to 

 indicate specific distinctness. The specimen was obtained in Swallow 

 Bay, Magellan's Straits." 



Finally Dr. Giinther, in 1830, in his " Introduction to the Study of 

 Fishes," (p. 469), interpolated the family Psychrolutidm between the 

 families Batrachidae and Pediculati. 



in 1832, Professors Jordan and Gilbert, in their "Synopsis of the Fishes 

 of North America" (pp. 083, 686), identified a fish from Kodiak, Alaska, 

 with the Psychrolutes paradoxus, and gave their views as to the rela 

 tionships of the genus, differing entirely from Dr. Gunther as to such 

 relationships. They considered the genus Psychrolutes to be composed 

 of "small fishes, closely resembling Liparididce, from which group they 

 are distinguished by no character of much importance," and referred 

 the genus to the family Cottidae, associating it with the genus Got- 

 tunculus in a subfamily Psychrol utilise, which they proposed for the 

 Cottidae with '• ventral fins present, spinous dorsal little developed, 

 continuous with the soft dorsal, the spines slender, concealed in the 

 loose, naked skin, gill-membranes loosely joined to the isthmus, no slit 

 behind last gill" (p. 683). 



Professors Jordan and Gilbert have specifically attributed to Psy 

 chrolutes a "suborbital stay narrow, not reaching preopercle." 



If the identification of Professors Jordan and Gilbert had been cor- 

 rect, the material would now be at hand for a comparison of the genera 

 Psychrolutes and Neophrynichthys, but if the description of Dr. Gunther 

 is reliable, the identification lias been erroneous. (1) The spinous dor- 

 sal was externally invisible in the type so that it was denied, in the di- 

 agnostic phrase, "a single dorsal fin on the tail, without spinous por- 

 tion," whereas, in the specimen of Jordan and Gilbert, the spinous 

 dorsal is conspicuous, although covered by skin; (2) the eye in the 

 type has a diameter only "about one-seventh of the length of the head, 

 one-half of that of the snout," while the eye in the Kodiak fish is about 

 one fourth the length of the head and about equal to that of the snout; 

 (3) the ventrals in the type are "composed of two rays, the inner of 

 which is bifid," while the Kodiak fish has, apparently, a spine and five 

 rays. The Alakan fish is in poor condition, and it does not seem ad- 

 visable to allocate it until the typical Psychrolutes is reexamined. But, 

 although the fishes described under the same name by Gunther and 

 Jordan and Gilbert appear to be distinct,* they may lie allied, and the 

 specimen described by the latter may thus be made serviceable in the 

 inquiry as to the relations of the type. 



*Dr. Bean informs rue that lie has himself distrusted the identification of the speci- 

 men described by Jordan and Gilbert, now in his custody, wiili tin' Psychrolutes par~ 



adoxus. 



