THE SALTING OF FISH. 29 



A number of experimenters have found that the amount of amino 

 acid formed is an index of the total protein decomposition in meat 

 and fish. Hoagland, McBryde, and Powick (1917), working on the 

 decomposition of beef during* cokl storage, found that the amino- 

 acid content of beef increased relatively more than any other con- 

 stituent. They found that the amount of amino acid formed varied 

 directly with the length of the storage period. They say : 



The increases in amino nitrogen represent an accumulation of the end prod- 

 ucts of proteolysis and furnish an excellent index of the extent of protein 

 autolysis. They are produced by the combined action of various proteolytic 

 enzymes — protease and erepsin in particular — upon muscle proteins and their 

 cleavage products. * * * 



Amino nitrogen showed greater actual and relative changes than any other 

 nitrogenous constituent. This result was to have been expected, since this 

 constituent represents, in a large degree, an accumulation of the end products 

 of proteolysis. 



Bradley and Taylor (1917) used the estimation of amino acids by 

 the formol-titration method to measure the rate of digestion of pro- 

 teins. 



Ottolenghi (1913) found that the amino-acid nitrogen content of 

 meat was the best index of decomposition of meats. He used the 

 amino-acid content to trace the ripening of meat to ascertain when 

 active decomposition sets in. He found that only S-orensen's formol- 

 titration method for determining amino-acid nitrogen and the micro- 

 scopic examination for bacteria Avere of practical use for his purpose. 

 Other methods were found to require too much time and elaborate 

 manipulation. 



METHODS. 



DeteTmiTiation of ammo-acid nitrogen. — It was necessary to use a 

 method of estimating amino-acid nitrogen which did not require a 

 complicated apparatus. The procedure had to be short, so that a 

 large number of analyses coidd be made in a day. The formol-titra- 

 tion method is the only one that fulfills these requirements. There- 

 fore, the amino-acid nitrogen was estimated in all cases by Sorensen's 

 formol-titration method as modified by Benedict and Murlin. 



The writer is well aware that this method is open to many objec- 

 tions. There are many slight errors due to the loss of diamino acids 

 by precipitation, the high results given by some amino acids, and 

 the low results given by others, etc. But for use as a measure of de- 

 composition of protein a method does not need necessarily to give the 

 exact amount of amino-acid nitrogen present in the fish or brine. If 

 it gives results which are relative in all cases, it is sufficiently ac- 

 curate for use where the results are used as an index of decomposition. 



The procedure in the determination of amino-acid nitrogen in 

 pickle was as follows : Twenty-five c, c. of brine were measured into 

 a beaker. A quantity of 10 per cent phosphotungstic acid in 2 per 

 cent hydrochloric acid, sufficient to precipitate all the protein, was 

 added. This was allowed to stand overnight. The solution was then 

 filtered and the precipitate washed. One c. c. of a 0.5 per cent solu- 

 tion of phenolphthalein was added and barium hydrate in substance 

 until the solution turned a permanent red. After standing at least 

 half an hour the solution was filtered and the precipitate washed. 

 The solution was then neutralized with half-normal hydrochloric 



