462 THE PUR SEALS OP THE PRIBILOP ISLANDS. 



contain openings of the mucous canals. Both pairs of nostrils with short wide tubes. 

 Jaws and vomer with wide bands of cardiform teeth of equal size. A round pore 

 nearly as large as nostril immediately behind last gill. Grill membranes with a short 

 free border mesially. 



D., IX, 15; A., 13; P., 17; pores of lateral line 33 or 35. Head (measured to end 

 of opercular flap) 2^ in length; depth about half head. Least interorbital width 

 three-fifths diameter of orbit, which is one-fifth length of head to tip of opercular 

 spine. Greatest width of head If in its length. Mouth large, the lower jaw included, 

 themaxillary reaching the vertical immediately behind tbeorbit; its length2f in head. 



Third, fourth, and fifth dorsal spines nearly equal, the fifth strongest, equal to 

 length of snout and half eye. A very short interspace between the two dorsals. 

 Second dorsal very high, the longest rays equaling length of snout and eye. Caudal 

 gently rounded when spread, its length half that of head to end of opercular spine. 

 The ventrals reach halfway to front of anal, equaling height of second dorsal. Pec- 

 torals scarcely to vent, the length of the longest rays equaling distance from eye 

 to tip of opercular spine. Vent midway between base of caudal and base of lower 

 pectoral ray. Skin everywhere smooth. 



Our specimen agrees well with the description of the much smaller type (185 mm.), 

 apparently differing in the shorter pectorals and more deeply concave interorbital 

 sijace. The white spots also show no tendency to run together to form streaks either 

 along back or on the bases of the fins. 

 97. Myoxocephalus jaok Cuvier and Valenciennes. 



Cottas humilis Bean, Proo. U. S. Nat. Mus. 1881, 149; Cbamisso Island, EscUseholtz Bay. 

 Cottus polyaeantltocephalus Kner, Sitzungsb. d. K. Akad. d. Wissen. LVIII, 1868, p. 21 ; taf. IV, 



11. Decastris Bay (not of Pallas). 

 Coitus tamioptenis Bean and Bean, Proo. U. S. Nat. Mus. 1896, 384 (not of Kner;. 



In a report on the ichthyological collections of the Albatross in Alaska (Report of 

 United States Commissioner of Fish and Fisheries for 1893, p. 421), Dr. Gilbert writes 

 as follows: "A. humilis closely ri^sembles the description of A. jaolc, with which it 

 may well be identical. We do not venture to make this identification as A. jaolc is 

 said to have but 7 dorsal spines, a number we have not found in A. humilis." On 

 further consideration we have decided that the two must be identical. The type of 

 jaolc was a large dried specimen, the same which had served Pallas for his account of 

 Gottus scorpius. In such a dried specimen it would be very difficult to enumerate 

 correctly the low, feeble spines, of which the first two are very closely approximated 

 and the last one often minute and hidden in the membrane. M. humilis is abundant 

 along the coast of Kamchatka and agrees with the account of jaolc in having the 

 upper parts covered with small brown spots, the back with a series of round spinous 

 plates, and the sides below the lateral lines with posteriorly directed spines; it also 

 agrees in reaching a very large size. In the description of jaolc, the fin formula, 

 except the number of dorsal spines, is that most frequently found in humilis. 



Specimens are in the present collection from Petropaulski and from stations 3640 

 and 3G48, off Eobben Island, in 18 and 20 fathoms. All of these have the supraocular 

 and occipital crests higher and sharper than in those from the eastern portion of Bering 

 Sea, and the preopercular spines are longer, usually reaching in young specimens 

 to or beyond opercular margin. These are, however, characters subject to much 

 variation within the group, in which it will always be unsafe to recognize subspecies 



