30 U. S. BUREAU OF FISHERIES 



aged 49.7 mm in length at the time of marking, and the second con- 

 sisted of 25,000 fingerlings, averaging 49.1 mm in length, transferred 

 from the Green River State hatchery to the Pnyallup River immedi- 

 ately before marking. With the run virtually complete, recoveries 

 to date number 28 from the native stock and 8 from the transplanted 

 stock. No recoveries have been made at the Green River hatchery. 



A further series of experiments on coho fingerlings have been under- 

 taken during the year to determine the effect of various periods of 

 hatchery-rearing on the return at maturity. This work has been 

 carried on at the Quilcene, Wash., station of the Division of Fish 

 Culture. One lot of these fish was released in September and another 

 in December in the Quilcene River immediately below the station. 

 Additional lots of fingerlings from the same brood are being retained in 

 the hatchery ponds for marking and release during the spring of 1938 

 at the approximate time of normal seaward migration. Exact costs 

 of rearing and handling are being recorded in order that the compara- 

 tive results of moderate and long rearing periods may be determined. 



Because the numl)er of individual lots of fish which can be marked 

 by excision of a pair of fins is definitely limited by the number of 

 possible combinations, an experiment was undertaken to determine 

 the adaptability of the internal tag for this purpose. The use of these 

 rags has heretofore been confined to relatively large fish and has been 

 attended by the disadvantage that an external mark to indicate the 

 presence of a tag is lacking. Both hatchery-reared and wild coho 

 fingerlings. ranging in size from 43 to 80 mm Avere obtained and 

 tagged with internal celluloid tags measuring 2 by 12.5 mm and of 

 varying thicknesses. One fin was removed from each fish at the 

 time of tagging and the various lots were held in aquaria f^r 75 

 days. Losses in the experimental lots ranged from mortality approxi- 

 mately equal to that of the coiitrols to 75 percent greater than that of 

 the controls. 



An operation on a larger scale was then undertaken at the Samisli 

 River hatchery to test the tags under field conditions. Tags of 0.030 

 inch thickness were selected on the basis of the previous experiment 

 and the dorsal fin was removed from each fish at the time of marking 

 for external identification. Excessive losses from a fungus infection 

 were encountered with the first lot marked, which was held in troughs 

 inside the hatchery. After treatment with salt baths for some linie 

 the remaining fish were removed to an outside concrete pond, where 

 they were held until there was no further daily loss. 



Lots of these fish were released in Friday Creek on October 8, 

 October 20, and December 11. If the results from tagging compare 

 satisfactorily with those from fin marking, greater opportunity for 

 comparative studies of individual lots of fish will be afforded. 



Sampling of commercial and sport catches for data on size, sex, 

 and age was continued and stream collections of fingerings were 

 made for additional material on early life history and rate of gi'owth 

 of coho salmon. A comparison of scale samples from various points 

 in Alaska with those from the local runs indicates that an increasing 

 number of fish in northern waters are returning to spawn at the age 

 of 4 years. All scales examined have shown only 1 year of residence 

 in the ocean, the 4-year fish in Alaska having remained in fresh water 

 until their third year. 



