PACIFIC SALMON 



67 



sure, that the results given were not obtained 

 by trained observers, for the authors in- 

 spected but a small fraction of the run and 

 that at selected points only. Yet further 

 inquiry is appropriate. Are the missing 

 299 marked fish in the most favorable ex- 

 periment to be considered as representing 

 the mortality which those salmon suffered 

 during the life in the sea? If so, do not 

 the extra 167 missing in the other unfavor- 

 able experiment represent the proportion of 

 that group which went to some other areas ? 

 Others have calculated a much larger per- 

 centage of returns than the ratio 1 : 300 

 established by the most successful record that 

 Rich and Holmes cite ; calculations based on 

 those reports would give an even larger num- 

 ber of wanderers than indicated by the 

 marking experiments of Rich and Holmes. 

 The conclusions reached by Rich and 

 Holmes furnish further interesting items. 

 After stating (p. 262) that the parent- 

 stream theory is ' ' now substantiated by such 

 a wealth of evidence that it seems nearly 

 superfluous to state that none of the salmon 

 marked in the Columbia have been recovered 

 in any other river system" the authors con- 

 tinue: "the tagging experiments in British 

 Columbia in 1925 (Williamson, 1927) show 

 conclusively that a large percentage of the 

 spring (chinook) salmon caught by troll in 

 these waters originated in the Columbia 

 River." [Itjalics mine.] The "Williamson 

 article has been carefully scrutinized and 

 gives the following testimony. Under joint 

 auspices extensive tagging of adult salmon 

 was carried on off the west coast of Van- 

 couver Island, and off the Queen Charlotte 

 Islands. These fish scattered widely, reach- 

 ing many different spawning grounds. But 

 at no point in the paper is any evidence 

 furnished to show that the author who was 

 in charge of the operation knew from what 

 streams these fish had come as young or by 

 what route they had reached the points at 

 which they were captured. Accordingly it is 

 difficult to see on what grounds Rich is 

 justified in the statement quoted above. So 

 far as the Williamson article is concerned 

 this comment by Rich is purely an assump- 

 tion based on the fact that these adults after 



being tagged were recaptured in the 

 Columbia. 



Unquestionably the most safely guarded 

 experiments on marked young salmon were 

 those planned and carried out with great 

 care by Foerster at Cultus Lake, British 

 Columbia. Because the outlet of the lake 

 is completely closed by traps, records could 

 be made of every outgoing fish, fry or year- 

 lings, and of every incoming adult. In a 

 series of papers entitled "An Investigation 

 of the Life History and Propagation of the 

 Sockeye Salmon {Oncorhynchus nerka) at 

 Cultus Lake, British Columbia," and in 

 other separate articles, Foerster and his co- 

 workers have recorded results of funda- 

 mental worth. Some items from these re- 

 ports have already been cited in this paper ; 

 here I wish to add further notes on the re- 

 turn of marked salmon. 



The work at Cultus Lake started with the 

 spawning season of 1925. The first experi- 

 ments on marking sockeye migrants were 

 made in 1926 by removing both pelvic fins 

 in 101,200 out of a total run of 1,400,000 ; 

 these were due to return in the fall of 1928 

 (Foerster 1929b : 26) . In 1927 both pelvics 

 and also the adipose fins were removed from 

 91,600 out of a total downstream migration 

 of 249,700, both one-year and two-year fish, 

 due to return in 1929. Returns in this class 

 have already been cited in this paper as 

 0.88 per cent of total marked and 2.49 per 

 cent of unmarked, the discrepancy indicat- 

 ing a probable inwandering of sockeye from 

 other regions. 



In reporting the life cycle of the 1926 

 year class Foerster (1936) recorded that 

 from the group of 1926 migrants both pelvics 

 and half of the dorsal were removed from 

 99,700 sockeye, or 29.7 per cent of the entire 

 outgoing young of that year. No marked 

 three-year adults returned in 1929 and no 

 marked five-year adults appeared in 1931. 

 In the fall of 1930, adults were taken to the 

 number of 1,340 marked in that manner and 

 this group formed 1.34 per cent of the num- 

 ber operated on. Foerster calculated the 

 return of unmarked adults in this class as 

 3.2 per cent. While he regarded the data 

 for unmarked adults as subject to sampling 



