30 THE SALTTITG OF FISH. 



acid. Enough tenth-normal sodium hydroxide was then added to 

 turn the sohition very slightly pink. Then 10 c. c. of neutral 40 per 

 cent formalin were added and the solution titrated with tenth-normal 

 sodium hydroxide. A blank on the formalin was run and this sub- 

 tracted or added, as the case might be. 



When the determination was made on the fish, the procedure was 

 varied as follows : Sections were cut from four fish. These were cut 

 into small pieces and mixed. Twenty-five grams of this mixture 

 were weighed out and then ground in, a mortar for 5 minutes. 

 Twenty-five grams of salt were added, and the sample was washed 

 out of the mortar. The sample was then made up to 250 c. c. volume 

 and was allowed to stand on ice overnight; then sufficient was filtered 

 off through a dry filter to obtain a 50 c. c. sample which was treated 

 as in the case of the brine. 



Before sampling the salt the residual salt was thoroughly mixed 

 after draining off the brine. A 25-gram sample was then weighed 

 out. After dissolving the salt in water sufficient phosphotungstic 

 acid was added to precipitate the protein and ammonia present. 

 From this point the procedure was the same as for the brine and 

 fish. In every case, after precipitation with phosphotungstic acid, 

 the supernatant liquid was tested to see if enough phosphotungstic 

 acid had been added. 



Salthuj of fish. — Great emphasis was placed on uniformity of 

 procedure throughout this experimental work. The experiments 

 concerning protein decomposition as affected by the impurities in 

 salt were carried out on river herring caught in the Albemarle 

 Sound. These fish were purchased from fishermen and were iced as 

 soon as they were brought ashore. This was about 1| hours after 

 capture. They were then immediately cleaned thoroughly. The 

 following pi-ocedure was followed in salting lots numbered FA-1, 

 FA-4, FA-5, and FA-6. These results are reported in Table 5, 

 page 32. 



The heads and all viscera were removed and the fish split along 

 the belly to the vent. They were then soaked in ice water for an 

 hour. The under side of the backljone was scraped free from blood, 

 and all blood was washed out. The fish were then drained free from 

 water, weighed, and salted. The river herring were rolled in salt 

 and packed tightly, belly up. in layei^ in kits. Some salt was 

 sprinkled over each layer of fish. No brine was added; the fish de- 

 veloped their own piclde. When packed, a weight of salt equal to 

 one-third the weight of the fish was added. The following day a 

 weight of salt equal to one-twelfth the weight of the fish Avas added. 



The salt used was either the commercial brand of salt known as 

 Diamond Flake, or mixtures of this salt with chemically pure salts 

 of similar degrees of fineness. Diamond Flake salt is a fine-grained 

 salt prepared in Michigan, the analysis of which is given on page 18. 



Previously, an experiment had been run to determine whether the 

 impurities present in commercial Diamond Flake salt were of suffi- 

 cient importance to cause a variation in its preservative action on 

 fish from that of chemically pure sodium chloride taken as a stand- 

 ard. The result of this work showed that the differences in the 

 amounts of amino-acid iiitiogcn formed in these two cases was 

 within the limit of the experimental error. 



