242 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



TSHAWYTSCHA 



• KETA 



• KISUTCH 

 • • NERKA 



D NAMAYCUSH 



• GORBUSCHA 



X SALAR 



X G. KAULOOPS 



O HI ALU A 



O— O ALPINUS 



O— O FONTINALIS 



O AUREOLUS 



I — 

 O UARSTONI 



O OOUASSA 



X TRUTTA 



Oe6=ExPECTE0 

 H WtTHiN SPECIES 

 VARIATION 



L_ 



_L_ 



_1_ 



BRANCHIOSTEGAL RAYS 



-Mean numbers of branehiostegal rays. 



Figure 3.- 



for British Columbia and Puget Sound the dif- 

 ference is 0.354. Considering that a difference of 

 0.506 was noted between adjacent localities in 

 southeastern Alaska, it would seem logical to add 



this geographical difference of 0.354 to the previ- 

 ous difference of 0.506, which gives a difference of 

 0.860 rays that can be expected between means of 

 samples of the same species. 



The branehiostegal ray counts for various Sal- 

 monidae are summarized in table 7 and figure 3. 

 If we apply to the other species the criterion found 

 above for nerka of an expected "within species" 

 difference of 0.86 rays between samples we find 

 that the table clearly sets apart 0. tshawytscha. 

 The next three species of Oncorhynchus, heta^ 

 Mmtch, and nerka are close together but separated 

 from gorhtischa. 



C. Tuimaycush is clearly distinct from the re- 

 maining charrs. 



Another interesting point is that S. trutta is 

 quite separate from saJar or g. kamloops. This is 

 reminiscent of the position of S. trutta. (in fig. 1) 

 between the charrs and the other Salmo. 



Table 7. — Count of branehiostegal rays on left side in North American Salmonidae 



[x in frequency column indicates rays present, but no numbers given] 



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



' Chamberlain (1907); soutlieastern Alaska. 



> Kendall (1935, p. 137). 



' McCrimmon (1949); eastern Canada. 



• Shapovalov (1947). 



Pyloric Caeca 



Since more material is available for Oncorhyn- 

 chi(.s it has been considered first (table 8). The 

 published material on caeca is usually listed by 

 categories and since different authors have used 

 different breaking points for their categories, some 



» Mottley (1936); Kootenay Lake. 

 ' Vladykov (1954). 



* Wilder (19.'J2); Nova Scotia and New Bnmswick. 



• DeLacy and Morton (1943); Karluk, Alaska. 



of their material may be listed slightly in error; 

 thus, the number of caeca if listed from 96-105 

 would be given in table 8 under the category 95- 

 104. 



The material for tshawytHcha is extremely vari- 

 able but this is caused chiefly by the great differ- 

 ence between the counts for the Sacramento River 



