480 THE FUR SEALS OF THE PRIBILOF ISLANDS. 



168. Pholidapus dybo'wskii (Steindachner). 



Centronotus dyhoivskii Steindachner, Ichthy. Beitriige, IX, 1880, 22, northern Japan. 

 f I'liolidapus grebnitskii Bean and Bean, Proc. U. S. Nat. Mns., 1896, 390, Yeso, Japan. 



Five specimens, the largest 25 cm. long, from Shana Bay, Iturup Island. Stein- 

 dacbner's excellent and detailed description leaves nothing to be desired, and corre- 

 sponds perfectly with our material except in the character of the scales. A careful 

 examination of these under high magnification fails to show that they are " am hinteren 

 Rande mit kurzen Zahnchen bewafinet." The posterior border is entire and the 

 scales strongly marked with concentric stria'. Dorsal spines number (52, 03, 03, 04, 04. 

 Dorsal ocelli are present in all our specimens, two of them being faintly visible even 

 in the youngest, 55 mm. long. 



rholidapuH grebnitslcii may differ in the shorter dorsal fin (57 spines) and in the 

 absence of teeth on the vomer, if these details are correctly reported. 



The genus Pholidapun is closely related to Opisthocentrus, diftering in the naked 

 cheeks, the dentition, and in the less ditterentiation of the posterior dorsal spines. 



169. Opisthocentrus ocellatus (Tilesius). (Plate LXXIX.) 



Upislhocentrus qiiinqitemaculatiis Kner. 

 BlenniojMditim pelropauli Boulenger. 

 f Opisthocentrus tenuis Beau and Bean. 



Numerous specimens: Tareinsky Bay, Kamchatka; Petropaulski Harbor; Shana 

 Bay, Iturup Island. Recorded by Dr. Bean from Petropaulski. 



The number of dorsal ocelli varies from 5 to 9 in our specimens, being the pre- 

 vailing number. Of 24 s])ecimens whose tins we have examined, 4 have 58 dorsal 

 spines, 10 have 59, 5 have 00, and 5 have 01. In addition, 1 si)ecimen has but 55 

 spines. The latter is the only male in the collection, and is conspicuous by the absence 

 of distinct dorsal ocelli and the great lieight of the vertical tins, the longest dorsal 

 spine exceeding the length of the pectoral and contained 1^ times in head. In females 

 the longest spine is 2A in head. 



The anal contains 30 to 39 rays in all our specimens. The dorsal tin is composed 

 exclusively of spines, the anterior flexible ones passing into the strong iiungent ones 

 near the i)osterior end. The stronger si>ines vary trom 7 to 12 in number in our 

 specimens. 



Our material answers the description of the type, which had 57 dorsal spines and 

 30 anal rays. It also agrees with si)ecimens from Petropaulski, reported on by Bean 

 and Bean (Proc. U. S. Nat. Mas., 1896, 391), with dorsal spines 58 in number. Blcnni- 

 ophidium jjeiropatdi Boulenger (Proc. Zool. Soc. Lond., 1892, 583) has but 52 dorsal 

 spines, but it is otherwise not to be distinguished from 0. quinquemaculatus. Still 

 more aberrant are 4 specimens from northern Japan reported on by Steindachner 

 (Ichthy. Beitriige, IX, 25), with but 50 to 53 spines and 32 to 34 anal rays. 



Opisthocentrus tenuis Bean and Bean, from Yeso, D., 54; A., 38, is probably not 

 a distinct species, although differing from any si)ecimen examined by us. Our first 

 impression on examining the type was that it was a valid species. 



170. Pholis fasciatus (Bloch and Schneider). 



Blennius tnnia, Pallas, Zoogr. Kosso-Asiat., Ill, 1811, 178. Kuril Islands. 

 Muraenoides viaxiUaris Bean, Proc. U. S. Nat. Mns., 1881, 147. (St. Paul.) 

 Pholis tainia Bean and Bean, Proc. U. S. Nat. Mus., 1896, 388. 



Three large specimens from St. Paul Island, the type locality of P. maxillaris, 

 have been compared with a number of individuals of Pholis fasciatus from Upernavik, 



