NORTH AMERICAN SALMONIDAE 

 Table 10. — Count of anal fin rays in O. nerka 



245 



1 From Chamberlain (1907). 



' Because published data by Chamberlain is in percentages a few of the 

 samples reconverted to actual numbers differ slightly from original sample 

 size, undoubtedly owing to rounding oH of percentages. 



' From Milne (1948). 



< From Foerster and Pritchard (1935a). 



In determining how much variation to expect 

 between anal-ray counts within a species (table 

 10) we can only compare counts made by the same 

 investigator. In Chamberlain's data, the maxi- 

 mum dilTerence between sample means is 0.85 

 (18.24-17.39). In Milne's (1948) data we can 

 compare only the 1947 data (see quotation above) 

 which leaves a difference of 1.25 (14.98 — 13.73). 

 Because of the small size of the Lakelse sample 

 this difference may be too large. 



A comparison of the means and ranges of the 

 anal-ray count in table 11 shows that counts in all 

 Oncorhynchus are definitely higher than in the 



other genera. Sahno gairdneri occupies an inter- 

 mediate position between Oncoi'hynckun and the 

 charrs. 



For dorsal rays, as for the anal, counting meth- 

 ods dift'ered between investigator. Table 10 indi- 

 cates that Foei-ster and Pritchard (1935a) were 

 counting about 3 less anal rays than Chainberlain 

 was. Tlie doi-sal-ray count appears to vary some- 

 what less than the anal-ray count ; thus, for Cham- 

 berlain's data on southeastern Alaska sockeye the 

 maximum diU'erence between sample means is 0.85 

 rays for the anal-fin count but only 0.51 for the 

 dorsal count (table 12). 



