FISHES. 127 



21. Cottus quadricornis Linu. 



Coitus quadricornieHicu. , Last Arc. Voy., 1855. pp. 348-9 : Guktil.P. Z. S., 1877, 

 p. 293. 



Richardson records tlie species from the Copperiniue Eiver region and 

 Coronation Gulf (08° 30' >s"., 11 Qo \7.).—Giinther. 



22. Cottus polaris Sabiuo. 



Cottus polaris ElCH., Last Arc. Voy., 1855, p. 351. 



Eichardson refers to this species as occurring at North Georgia, lat. 

 75° K, and at the Peninsula of Boothia. The radial formula appears 

 to me more like that of a Centridermichtliys than anything else, but the 

 genus to which it properly belongs is uncertain. 



23. Gyranacanthus pistilliger (Pall.) Gill, MS. 



Cottus ventraVis Cuv. & Val., Hist. Nat. Poiss., iv, p. 194: Collett, Christi- 

 ania Vid. Selsk. Forh. 1878, No. 14, (p. 15). 



21732. (1373.) Godtliaab, Disko Id., Greenlaud, A. L. Kumlien. D. XII, I, 16. 

 A. 18. y. 3. Ventrals nearly reach vent. 



21733. (1374.) Godthaab, A. L. Kunilicn. D. XII, 15. A. 19. V. 3. Ventrals 

 reach fifth ray of anal. 



21734. Godthaab. A. L. Knmlicn. D. XI, 16. A. 18. V. 3. Ventrals extend 

 little more than half way to vent. 



21735. Godthaab. A. L. Kiimlion. D. XI, 17. A. 18. V. 3. Ventrals reach third 

 ray of anal. 



21736. Godthaab. A. L. Kumlien, D. XII, 17. A. 19. V. 3. Ventrals nearly 

 reach vent. 



21737. Godthaab. A. L. Kumlien. D. XII, I, 15. A. 18. V. 3. Veutrals reach 

 fourth ray of anal. 



21741. (151.) Niantilic Harbor, Cumberland Gulf, A. L. Kumlien. 

 21743. (180.) Niantilic Harbor, A. L. Kumlien. 

 22332. Niantilic Harbor, Aug. 1876, Lieut. W. A. Mintzer. 



17431. Christiania, Norway, M. G. Hetting, inspector of fisheries. D. XII, 14. A. 

 16. V. 3. Veutrals reach fourth ray of anal. 



Dr. Liitken rejects the name GymnocanthuSy Swainson, because the 

 genus was badly defined. There can be no difference of opinion as to 

 the fact that the genus was poorly characterized; but there is an attempt 

 at definition and a reference to a figure of the type-species, so that one 

 need not hesitate as to what is intended. If we begin to reject names 

 of genera because they are not accompanied by comidote descriptions, 

 we may find it difficult to draw the line between what we shall accept 

 and what we shall reject. In retaining Swainson's name it may not 



