322 THE rsVCHK'oLl-TID.l.;. 



The family and genus thus indicated were established for a single 

 small fish, named Psychrolutes paradoxus, twenty-one lines in total 

 length, obtained in the Gulf of Georgia during the voyage of II. M. 

 S. Plumper. 



In 1876Captain Button, in the "Transactions and Proceedings of the 

 New Zealand Institute"' ( Vol. VIII, p. 214), described a tisli then recently 

 found at New Zealand as Psychrolutes hit hs, considering it to be geuer- 

 ically related to the North American fish. 



In 187G Dr. Albert Giinther, in "The Annals and Magazine of Natu- 

 ral History" (!), Vol. XVII, p. 395, established for the New Zealand 

 fisb a pecuiar genus, to which he gave the name Neophrynichthys, and 

 assigned the following characteristics: 



"Head broad and depressed ; skin naked. Canine teeth none: palate 

 smooth. Gill-covers without spines. Two dorsals, the first formed by 

 nine flexible spines. Ventrals close together, thoracic, rudimentary. 

 Three gills and a hall'; pseudo-branchiae. Gill opening extending to the 

 lower angle of the pectoral." 



On this form he commented as follows: 



"One specimen 6£ inches long, from Dunedin, obtained from the 

 Otago Museum. This fish has been named by Captain Hutton Psychro- 

 lutes lata*; and from a careful comparison with Psychrolutes paradoxus, 

 I can confirm the correctness of his view as regards the affinity of these 

 two fishes; but the presence of a well developed first dorsal appears to 

 me to demand the separation of the New-Zealand fish into a distinct 

 genus. The discovery of this fish led me to reconsider the position 

 which the family Psychrolutidrc ought to take in the system. As the 

 absence of the first dorsal can not be retained as one of the characters 

 of the family (which would conuect it with the Gobiesooidae) I think 

 those fishes ought to be removed from the division of Gobiesociformes to 

 that of the Cotto-scombriformes, where it would folknv the Batrachidae.'' 



He corrected the reference of the Psychrolutidie to the Gobiesoci- 

 formes in a foot-note (p. 390), as follows: 



"In my systematic synopsis of the families of Acanthopterygian fishes 

 a misleading error has crept in (p. i.\), the family Psychrolutidie being 

 characterized by -Ventrals none' instead of 'No adhesive ventral appa- 

 ratus.' Also the diagnosis of the fourteenth division should be corrected 

 by striking out the words 'or entirely absent.'" 



Dr. Giinther, in 1881, in the "Proceedings of the Scientific Meetings 

 of the Zoological Society of London" for the year 1881 (p. 20, pi. i), 

 noticed and published a plate of a fish obtained in Swallow Bay, Ma- 

 gellan's Strait, which he referred to Neophrynichthys latus in the follow- 

 ing terms: 



"Of this veiy interesting fish, which was discovered only a few years 

 ago by Mr. Button in New-Zealand, a specimen lt> inches long is in the 

 collection. Fortunately, by the kindness of Mr. Hutton, I am in a posi- 

 tion to compare the American specimen with the one obtained on the 



