CHEMICAL COMPOSITION OF FOOD-FISHES. 
807 
muscular tissue (flesli) of lislies or mammals, not only because of lack of 
information regarding the nitrogenous compounds in the muscle of the 
oyster, but also because the other tissues and fluids, including the prod- 
ucts of secretion and excretion, and with them the liquids, are analyzed. 
The liquids consist of the fluids contained within the shells but outside 
the body, together with whatever blood may have flowed out of the 
wounds made in separating the body from the shell, or rather of so 
much of these as drained off from the body and shell in the preparation 
for analysis. In all of these latter substances we have to dojwith mate- 
rials about the composition of which extremely iittle is known. All that 
can be said for the figures for protein, therefore, is that they are based 
upon the amounts of nitrogen as found in the analyses ; that the nitrogen 
determinations were made with soda-lime, but by methods of manipula 
tion which have been tested in this laboratory by numerous comparisons 
with the absolute and Kjeldahl methods ; and that as a measure of the 
comparative amountsof true protein compounds (albuminoids and gelat- 
inoids) in the body they are, presumably, not very far out of the way. 
The determinations of fat represent the amounts of material extracted 
by warm ether from the residue left after evaporating the flesh and 
liquids, at a temperature of about 96°, nearly to dryness (so far that 
the mass was friable and easily pulverized). The flesh was dried in hydro- 
gen, and the liquids in air. I can not vouch for the correctness of the 
assumption that this ether extract consists entirely of fatty matters, but 
until the opportunity comes for investigating the subject more closely 
I see no better way than to simply state the results as we found them. 
The determinations of ash and of water I judge to be reasonably 
accurate. Our experience with the methods employed has given us 
confidence in their reliability within tolerably narrow limits. I should 
explain also that all the determinations except water were first computed 
upon the water-free substance and then calculated over upon fresh 
substance. 
The extractives were estimated by subtracting the sum of the protein, 
fats, and ether extract from the total water-free substance. It is plain 
that any errors in the estimates of the other ingredients will, unless 
they balance each other, affect the estimate of the extractives. The 
actual substances which the figures for extractives thus approximately 
represent in the flesh I presume to consist chiefly of the carbohydrates 
(glycogen) of the liver, which makes up a large portion of the body of 
the animal. 
CHANGES IN COMPOSITION PRODUCED BY FLOATING. 
As was to be expected, the sojourn in fresher water resulted iu an 
increase in the percentage of total shell contents. The principal gain 
was in the flesh, which naturally lost some mineral salts but gained 
much more water, so that its actual percentage increased. The increase 
was in one case from 9.49 to 11.41, and in the other from 8.35 to 10.18 
