Roedel did not find that these conclusions 

 were compatible with evidence from tagging 

 experiments ( Fry and Roedel 1949) . Tagged 

 fish from Soledad and Viscaino regions were 

 later recovered in both southern and central 

 California . (No facilities were available for 

 recovery in Baja California.) Returns from 

 southern California -tagged fish were far more 

 numerous than returns from Viscaino fish for 

 the first two years after release, but in the third 

 and fourth year after tagging returns from both 

 regions were of similar magnitude. The separa- 

 tion of California and Viscaino stocks was not 

 as complete as seemed indicated by the morpho- 

 metric study. No British Columbia fish were 

 tagged, but of 11 tagged near the mouth of the 

 Columbia River, one was recaptured off south - 

 ern California . No fish were tagged south of 

 the Viscaino region. 



Royce (this collection of papers) uses 

 Roedel' s data on morphometric comparisons of 

 head length in illustrating the overlap of 

 measured characters when regression has been 

 used. He then compares the results with the 

 tag return data, and does not find that the two 

 kinds of data are incompatible . The reader is 

 referred to Royce' s paper. 



Northern anchovy, Engraulis mordax 



Until recently, the northern anchovy has 

 been but little investigated. The systematic 

 sampling of the commercial catch and age an- 

 alyses were begun only in 1952 . (Miller 1955, 

 Miller ^t jil 1955.) Anchovies have not been 

 tagged, so nothing is known of the extent of 

 their migrations or mixing over their distribu- 

 tional range, which extends from the Queen 

 Charlotte Islands, British Columbia, to Cape 

 San Lucas, Baja California. 



Two studies have been made of meristic 

 characters. Hubbs (1925) reported on variation 

 in vertebral counts off central and southern 

 California. He separated the population inhabit- 

 ing the brackish waters of San Francisco Bay as 

 a distinct subspecies, Engraulis mordax nanus 

 Girard. The San Francisco Bay anchovy differs 

 from the ocean form in having a lower vertebral 

 count (av. of 43.80 as compared to 45.73 for 

 the ocean form in the San Francisco- region), 



smaller size, longer head and deeper body . The 

 lower vertebral count found in the bay form par- 

 allels the situation found in the nehu in Ala Wai 

 Canal, Oahu (Tester and Hiatt 1952), and in the 

 Australian anchovy, Engraulis austral is, in 

 Gippsland Lakes rivers (Blackburn 1950) . 



McHugh (1951) reported on a study of 

 Engraulis mordax mordax "designed to deter- 

 mine the extent of variation in space and time 

 in morphological characters of the northern 

 anchovy, and on the basis of these findings to 

 interpret the population structure of the sub- 

 species." (ibid. : 124.) He studied the distribu- 

 tion in mean values of five meristic characters: 

 vertebrae, anal fin rays, dorsal fin rays, 

 pectoral fin rays and gill rakers. 



The vertebral data grouped by stage of 

 development and area are presented in table 7. 



There is little difference in the mean num- 

 ber of vertebrae in the samples from British 

 Columbia, Washington, Oregon and California 

 (range 45.69 to 45.82 vertebrae), but there is 

 a falling off in mean number in samples from 

 Baja California, amounting to approximately 0.4 

 vertebrae off southern Baja California . (Turtle 

 Bay to San Juanico Bay) . McHugh concluded 

 "there is thus no evidence from vertebral counts 

 that racial differences occur along the entire 

 Pacific coast of Canada and the United States . 

 It appears, however, that intermingling between 

 these areas and Baja California is either incom- 

 plete or lacking." (ibid. : 131) 



McHugh found differences between samples 

 from the Pacific Northwest and southern California 

 in mean fin ray counts. The differences were 

 most marked in the mean number of anal fin rays . 

 Individuals from southern California averaged 

 over 0.5 ray higher than individuals from the 

 Pacific Northwest . However, there was con- 

 siderable variation in the means of samples from 

 each area, as is apparent from the following 

 tabulation: 



Locality Anal fin rays 



Range in mean values in samples 

 British Columbia 21.89-22.38 



Washington 21.89-22.25 



Oregon 21.89-22.43 



Central California 22 . 22 - 22 . 60 



Southern California 22 57-23. 07 



Baja California 22 . 08 - 23 . 00 



65 



