The experiments herein recorded were initiated to determine if 

 cold storage salmon eggs contain the accessory food substances for 

 salmon fry. Salmon roe was chosen because it is cheap and easily 

 available at the canneries and cold storages on the Pacific Coast, 

 where it constitutes part of the refuse. The eggs used were taken 

 from chum salmon (Oncorhynchus keta) in November, 1920, and 

 kept in cold storage during the winter. The lowest temperature to 

 which they were subjected was — 4° F. at the time of freezing and 

 10° F. during storage. In addition to these, fresh roe of chinook 

 salmon (0. tschawytscha) was used in four quantitative experiments. 

 To be sure, the use of salmon eggs as food for fry is not new, but 

 heretofore the eggs have been cooked, thus necessitating the use of 

 fresh meat, such as liver or melts, to supply the necessary vitamines. 



The eggs were obtained from the cold storage, allowed to thaw 

 out, ground in a meat grinder, and strained through a coarse cloth 

 to remove the shells. The last operation was found necessary in the 

 quantitative work to insure complete consumption. The substance 

 was then absorbed with wheat middlings, making a mixture of 

 80 per cent salmon eggs and 20 per cent middlings. In this way, the 

 ground salmon eggs were made available to the fry. 



The water used during the first part of the experiments was 

 pumped from Lake Washington, at Seattle, into a large settling tank 

 from which it was drawn for the hatchery. But occasionally the 

 water in the tank became low and some sediment passed through. 

 Fearing that this would affect the results of the experiments, city 

 water was turned on in the hatchery on May 14, 1921. No appreci- 

 able change in the quantitative work was observed. Seattle city 

 water is no doubt purer than the water used in many hatcheries. 



On March 24, 1921, two lots of 125 each of chinook salmon fry 

 (0. tschawytscha) , four months old, were placed in two troughs of 

 equal size. The fry had been starved for four days, previous to 

 which they were fed on the fry of small redfish (O. nerka). All were 

 in the same condition. The rate of water flow in both troughs was 

 the same, about 166 gallons per minute. The inside measurements of 

 the trough were 82 x 14 x 6 inches. The space to which the fry were 

 confined is 26^ x 14 x 4 inches, or a capacity of 1,484 cubic inches 

 of water. The light was regulated by means of shades to approxi- 

 mately the same intensity throughout the trough. 



One of these lots was fed exclusively on canned salmon. This 

 was considered the control, for it is known that the antiscorbutic and 

 antineurotic vitamines are easily destroyed by heat. It was there- 

 fore expected that this lot would die in time, due to the lack of these 



83 



