48 AMERICAN FISHERIES SOCIETY. 



of protein in the flesh remained unchanged during the floating. 

 It seems probable however that the flesh may have lost a small 

 amount of nitrogenous material. If this was the case the actual 

 gain of flesh and of water must have been less and the loss of 

 fats, carbohydrates and mineral salts, greater than the estimates 

 make them. But there appears to be every reason to believe 

 that the error must be very small, and since it would affect ail 

 the ingredients in the same ratio, the main result, namely, that 

 there was a large gain of water and a considerable loss not only 

 of mineral salts, but of fats and carbohydrates as well, can not 

 be questioned. 



CHANGES IN THE COMPOSITION OF THE LIQUID PORTION. (lIQUOR). 



3. The liquids might be expected to receive material from 

 the flesh, and to yield material to the surrounding water. The 

 materials coming from the flesh would be such as the latter 

 parted with by either osmose or secretion. Those yielded to the 

 water would either escape by diffusion or be washed away when 

 the shells v^ere open wide enough to allow. What share each 

 of these agencies had in effecting the changes that actually oc- 

 curred in the liquids, the experiments do not and, in the nature 

 of the case, cannot, tell. Comparing the percentage composi- 

 tion of the liquids before and aftei floating, as shown by the 

 averages of the analyses in the two experiments, it appears 

 that : — 



Tho percentages of j^^g^^ j^n^fi^ 



Water rose from 94.9 to 95.5 



Water-free substance fell from 5.1 " 4.5 



Total loo.o " loo.o 



Protein rose from. 1.9 " 2.1 



Carbohydrates, &c., rose from 0.7 " i.i 



Mineral salts fell from 2. 5 " 1.3 



The increase in the percentage of water, and the decrease in 

 that of mineral salts are very marked. The quantities of fats 

 ("ether extract") are too small to be taken into account. The in- 

 crease of nitrogen and that of carbohydrates, though absolutely 



