THE PISHES OF BEEING SEA. 481 



Greenland. We can appreciate no dififerences between the two. The size of the 

 mouth and the length of the head are the same in specimens of equal length, and no 

 difference exists in the development of the ventrals. The agreement seems to be 

 perfect in the fin rays, relative proportions, and coloration. 



Pallas's short account of Blennius twnia contains nothing distinctive except the 

 number of fin rays and the statement that the body is banded. As both of these 

 items agree with the present species, we may safely follow Bean and Bean in making 

 the identification. 



This species has been recorded by Gilbert from Bristol Bay, and by Dr. Bean 

 from Petropaulski and St. Paul. 



In a specimen from St. Paul, 29 cm. long, the length of the maxillary is contained 

 2f times in distance from tip of snout to origin of dorsal; the mandible equals the 

 length of the pectoral. In a younger specimen, 15 cm, long, from Bristol Bay, the 

 maxillary is contained 3^ in predorsal length; the mandible approximately equals 

 length of pectoral. 



171. Pholis ornatus (Girard). 



Abundant at Captains Harbor and Ilialiuk, TJnalaska. Two specimens, agreeing 

 entirely with the above, taken by Mr. Barrett-Hamilton at Bering Island, D.; LXXVII 

 and LXXIX; A., II, 35 and II, 37. 



Dr. Gilbert records the species from Herendeen Bay, and Dr. Bean from Kadiak, 

 Shumagins, Belkofski, Unalaska, Atka, Adakh, Amchitka, Attn, and Port Moller. 

 Stejneger and Grebnitzki found it on Bering and Medni islands. 



172. Pholis piotus (Kner). Plate LXXXa.) 



Uroeentrus pietiis Kner, Sitzungsb. d. K. Akad. d. Wisseusch., LVIII, 1868, ^. 51, taf. VII, 



fig. 21. 

 Centronotus pictns Steindachner, Ichth. Beitrage, IX, p. 25. 



Numerous specimens from Shana Bay, Iturup Island. As already shown by 

 Steindachner, this is a typical Pholis, Kner having been in error in ascribing to it an 

 isolated and channeled first anal spine. The ventral spines are bound down by the 

 integument more closely than usual, but they are in other respects not peculiar. 

 Each is accompanied by two short spinous rays, concealed in the membrane and 

 difficult to detect. The latter are stiff and pungent, and seem to be not articulated. 

 The ventrals of P. ornatus show the same strncture, Kner gives the anal formula as 

 II, 40. This must be a misprint for II, 49, as the artist figures 51 rays in the fin, not 

 differentiating the two anterior ones. 



D., 9^ to lOJ in length to base of caudal. Depth, 8 to 10; D., XCIII or XOIV; 

 A., II, 46 to 48. 



173. Pholis dolichogaster (Pallas). 



Gunnellas ruberrimus Cuvier and Valenciennes, Hist. Nat. Foiss., XI, 440. 



One specimen, 18 cm. long, taken by Mr. Barrett-Hamilton at Bering Island. 

 Also, taken by Stejneger and Grebnitzki on Bering and Medni islands and at Volcano 

 Bay, Yeso. 



The color is cherry-red on the body and fins; lighter on belly, lower half of cheeks, 



and under side of head. Lips blackish anteriorly, a narrow black streak running from 



them along snout to eye and from eye across cheeks and opercles toward upper edge 



of pectoral base. This line separates the deep red upper part of the head from the 



5947— PT 3 31 



