64 Cold-Storage of Fish [Oct. 



flounder {Pleiironectes americanns) . It will be seen that the gov- 

 ernment figures accord neither with Williams' nor our own. The 

 government values are based lipon the analysis of but one fish in 

 each case. Ulrich did not analyze a fresh specimen, but reports the 

 water-content of smoked flounder to be 71.66 per cent. 



On comparing the results for the three series, it will be noted 

 that there was practically no change in the water-content. This 

 would be naturally expected, because of the care taken to keep the 

 fish completely encased in ice during the storage period. In the 

 case of beef, however, Emmett and Grindley^^ report a loss of 1.3 

 per cent. of moisture after a forty-three days' period of storage. 



Total solids. As the value for total solids was determined by 

 difference, it varied inversely as the water-content. Our results 

 show that there was no change in the water-content. Consequently, 

 the value for total solids remained unchanged. 



Inorganic matter. There was no reason to believe that pos- 

 sible changes in the flesh during cold-storage would affect the ash- 

 yield, yet for an adequate analysis of the fresh specimen it was 

 desirable to make this determination. By repeating the determina- 

 tion on the cold-stored products we were able to obtain a closer value 

 for the ash-yield and at the same time detect any unexpected change. 

 Ash determinations were made only upon summer flounders. The 

 government analyst in the report already quoted gives the following 

 percentage values for the ash content of three related fish : summer 

 flounder, 1.29; winter flounder, 1.20; and plaice, 1.46. As was 

 expected, the results obtained by us showed that cold-storage was 

 without effect on the yield of ash. 



Organic matter. As in the case of total solids, organic matter 

 was determined by difference. Its variations were dependent upon 

 variations in water-content and ash-yield. There being practically 

 no variations in these, the percentage of organic matter remained 

 unchanged. 



Ammonium nitrogen. Without doubt this was our most im- 

 portant determination, especially from the Standpoint of detection of 

 bacterial influences. Our method, as already described, differs from 

 that used by Pennington and Greenlee^^ in that we substituted 



15 Emmett and Grindley : 7. Ind. Eng. Chetn., 1909, i, p. 413. 



18 Pennington and Greenlee: Journ. Am. Chetn. Soc, 1911, xxxii, p. 561. 



