Fishes of the Western North Atlantic 575 



Pallas I 8 14 (7J: 389). These early names for the Pacific capelin seem then to have 

 been overlooked; at least all the subsequent references to them seem to have been under 

 the name villosus down to 1937 (Soldatov and Lindberg, 10^: 57; Hart, jp: 417-420); 

 and this same course was followed by Saemundsson as recently as 1949 (pj: 93). In 

 1937, however, Schultz (^8: 13—20) reported that the capelin of the Bering Sea region 

 differ statistically from those of the Murman coast and of Newfoundland in the number 

 of scales and pectoral fin rays, in the distance from the tip of the snout to the insertion 

 of the pectoral fins, and in what he termed the "character index," i. e. the number 

 of dorsal rays, plus the number of anal rays, plus the number of pectoral rays, plus the 

 number of scales below the lateral line, minus the total number of gill rakers on the first 

 arch (see Key, below). He also reported that 



Pacific capelin average smaller at maturity than Table III. Relative Frequency in the Num- 

 the Atlantic fish. ^^^ °^ Fused Rays in the Anal Fin (in per 



A J- 1 c 1. li • J ^ • cent) for Pacific and Atlantic Capelin. 



Accordmgly, bchultz revived catervartus as a ' ^ 



distinct species, in which he has been followed ^""^''^ °^ ^^^'^^ Frequency 



^ N ,1 u ^"^^'^ ^^y^ Paf/Af Atlantic 



by Clemens and Wuby {13: 23, 97) as well as by „ 



Hubbs and Chapman (^44: 296); and Vladykov 5 66 54 



{114: 201) has contributed the additional infor- 6 13.5 5 



mation that the number of fused rays in the anal 



fin tend to be greater in Pacific capelin than in Atlantic specimens (Table iii). On 



the other hand, Vladykov, while adding support to the impression generally held that 



Pacific capelin average smaller than Atlantic fish, has pointed out that the greatest 



length credited to a capelin anywhere was 10 inches or less for the Arctic coast of 



Alaska {26: 105), where catervarius rather than villosus is to be expected on geographic 



grounds, if the two are separable. 



All in all, our present state of knowledge seems best expressed by rating catervarius 



as a subspecies of villosus^ as Rumjanev has done (pj), at least until someone has an 



opportunity to study adequate series of capelin from the Arctic coast of Canada; for 



example, from Bathurst Inlet, or from Coronation Gulf, where they were reported 



many years ago (p. 574), or even from Hudson Bay — localities that are intermediate 



between those of record for villosus on the one hand and for catervarius on the other. 



Key to Subspecies of Mallotus villosus 



I a. Character index (see above) 30 or more (30—38 recorded). 



villosus Miiller 1777, p. 576. 



I b. Character index (see above) not more than 29 (22—28 recorded). 



catervarius Pennant 1784. 

 Both sides of the North Pacific, 

 southward to northern Japan and 

 Korea in the west; to the Strait 

 of Juan de Fuca in the east. 



