246 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



Table 11. — Count of anal fin rays in North American Salmonidae 



[Counts adjusted to a complete count (see text); x indicates rays present in frequency column but no number given) 



1 Chamberlain (1907); southeastern Alaska: complete count made. 



' Milne (1948); Skeena River; data adjusted by adding 3 rays (see table 10). 



' Foerster and Pritchard (1935a); southern British Columbia and Puget 

 Somid; data adjusted by adding 3 rays (see table 10). 



* Milne (1948) ; Skeena River; data adjusted by adding 2 rays (McCrimmon 

 (1949) says 1 rudimentary and 1 unbranched in S. aalar and S. trutta). 



' Shapovalov (1947); California; 2 rays added. 



* Mottley (1936); Kootenay Lake, British Columbia; 2 rays added; standard 

 deviation 0.5. 



' Kendall (1935, p. 137); Penobscot River; 2 rays added; McCrimmon 

 (1949). 

 ' McCrimmon (1949); count includes rudimentary rays. 



• Vladykov (1954); complete count. 



i« Wilder (1952); Nova Scotia; complete count. 



" DeLacy and Morton (1943); Karluk, Alaska; count may be incomplete. 



Table 12. — Count of dorsal fin rays in O. nerka 



' Chamberlain, 1907. Because his published data are in percentages, a few of the reconstructed samples differ slightly in sample number. 

 ' Foerster and Pritchard, (1935a) ; counts do not include all rays. 



The meager data on dorsal-ray counts for all 

 species are summarized in table 13, in which I 

 have attempted to adjust all data to a complete 

 count. This shows that the overlap in the fre- 

 quency distributions of the dorsal-ray count is 



sufficiently large that many individuals of Oncor- 

 hynchus can not be distinguished from the charrs 

 on the basis of dorsal-ray count. 



It is worthy of note that 0. kisufch is lower than 

 the remaining Oncorhynchus in both anal- and 



