354 ALBERT C. HUNTER 



were responsible for the decomposition of the salmon and, if such 

 was found to be the case, to establish a correlation between the 

 physical appearance of the fish and the presence of the bacteria 

 after holding the salmon for known lengths of time under known 

 conditions. This necessitated the determination of total counts, 

 and later of the kinds of bacteria, present in the muscular tissue 

 of the salmon and also an investigation as to how these bacteria 

 gain access to the flesh. 



Four species of salmon were used in the investigation, namely, 

 the sockeye or red salmon (Oncorhynchus nerka), the humpback 

 or pink salmon (0. gorbuscha), the silver or coho salmon (0. 

 kisutch) and the chum or dog salmon (0. keta). All the fish 

 used were caught in fish traps near the San Juan Islands in Puget 

 Sound. The fish were handled as they are handled commercially 

 by the fishermen in that locality and accurate data were kept 

 as to the methods of handling, the length of time out of water, 

 the temperature at which they were held, etc. 



The method of procedure was usually as follows: A trip was 

 made to the trap late in the afternoon and the night was spent 

 on the fishing boat alongside the trap. Early the next morning 

 the trap was lifted and emptied of fish. One or two fish were 

 selected for immediate examination and the others were placed 

 in the hold of the boat to be transported to the fish house. On 

 arrival at the fish house the desired number of salmon were placed 

 in a large box in a corner of the fish house. A thermograph was 

 kept with the fish throughout the whole period. Each morning 

 one fish was removed from the box and examined. The general 

 appearance of the fish was noted and recorded. When the odor 

 and appearance of the salmon indicated that the fish were in an 

 advanced stage of decomposition the experiment was terminated 

 and a new experiment begun. Whenever it was found impossible 

 to visit the trap on the fishing boat the fish were received at the 

 fish house and the data in regard to the time and place of the 

 catch were obtained from the fisherman. The fish usually ar- 

 rived about eight hours after they were taken from the water. 

 In the case of sockeye and humpback salmon three separate 

 catches were allowed to decompose and were studied. The first 



