IEMICAL COMPOSITION OF FOOD-FISHES. 855 
The experiments give no reliable data for the determinations of the 
absolute increase and decrease of the liquids and their constituents, so 
that it is impossible to say with entire certainty whether there was or 
was not an actual gain of protein or fats or carbohydrates. It seems 
very probable, however, that the liquids received and retained small 
quantities of these materials from the flesh (bodies) of the animals. 
CHANGES IN THE COMPOSITION OF THE WHOLE SHELL CONTENTS, FLESH AND LIQUIDS. 
4. Comparing the average percentage composition of the total shell 
contents before and after floating in the two experiments, it appears : 
Beforo A ffer 
The percentages of— dialysis. dialysis. 
Water rose from 85.2 to 87.1 
Water- free substance fell from 14.8 to 12.9 
100. 0 100. 0 
Protein (N X 6. 25) fell from 6.8 to 6.5 
Fats (ether extract) fell from . 1.4 to 1. 2 
Carbohydrates, etc., fell from 4. 3 to 3. 7 
Mineral salts (ash) fell from 2. 3 to 1. 5 
14.8 12.9 
The changes in the total shell contents are mainly those of the flesh, 
since the amounts of the organic and mineral ingredients of the liquids 
are small. The water-free substance of the liquids makes only a small 
part of the whole water-free substance and consists very largely of 
mineral salts, a large part of which are the salts of the sea water. 
Here accordingly, as in the flesh, we have to do mainly with an increase 
of water and. decrease of water-free substance and of its ingredients, 
though there was more loss of protein, carbohydrates, and fats than the 
gain in the water should account for. 
5. The absolute gain and loss of material in the total shell contents 
can not be exactly computed from the data of these experiments ; but 
it is safe to say that it consists almost entirely in a gain of water and 
loss of salts, and that, though there seems to be a loss of fats, and espe- 
cially of carbohydrates, and perhaps a loss of protein also, the amount 
of these latter that escape during the floating must have been small. 
CHANGES DUE TO OSMOSE. 
6. From the standpoint of the physiological chemist the chief inter- 
est of the experiments is in the quantitative indication of the changes 
due to dialysis. As was to be expected, the change in composition 
during the dialysis consisted chiefly in a small loss of soluble materials, 
especially mineral salts, with some carbohydrates, and a large gain of 
water. But the flesh (body) of the animals, in which the chief change 
occurred, appeared to lose a little fat and protein also, which would 
hardly be taken out by osmose. Their removal is easily explained by 
assuming processes of secretion or excretion to accompany that of 
dialysis. 
