[85] CHEMICAL COMPOSITION AND VALUE OF FISH FOR FOOD. 265 
Second. The figures of some of the kinds of food, indeed all except 
the fresh meats, milk and its products, and eggs, are based upon few 
analyses. More are needed to show the actual ranges of variation 
and the actual averages. Sometimes different samples of the same kind 
of flesh will show widely varying percentages of constituents. This is 
especially true of the fats, and to a less extent of the water. But the 
. figures in the table are probably not very far from true representations 
of the average composition of the several kinds of foods. 
Third. The nutritive valuations are made with only an imperfect 
knowledge of the digestibility of the foods and the influence of palati- 
bility and other factors upon the nutritive value. They are, therefore, 
of necessity somewhat crude, and to be relied upon rather as approxi- 
mations than as accurate quantitative statements. Much more chemical 
and physiological investigation is needed to make our knowledge of these 
as complete and satisfactory as it should be. 
This table is, I think, worth studying. As above explained, after 
taking out the refuse, bones, skin, entrails, &c., of the sample we have 
left the edible portion, the flesh (or, in case of oysters, clams, &c., the 
flesh and liquids together). Multiplying the percentage of flesh by that 
of actual nutrients in each sample gives the actual nutrients in the sam- 
ple, as is done in the first column. The composition of the edible portion 
is given in the four columns under “In flesh,” the figures being taken from 
Tables IV and VII. As shown above, the proportions of flesh and of 
water and dry substance in the flesh differ greatly in the differentsamples. 
PROPORTIONS OF NUTRIENTS. 
The proportions of albuminoids, fats, &c., are, if anything, still more 
varied. The cod, haddock, and bass, have scarcely any fat, and the 
oysters, scallops, and clams, very little. But the eels, shad, mackerel, 
salmon, herring, and turbot are very fat. On the whole, fish average 
about the same percentages of albuminoids as the meats, but generally 
have less of fats. a 
It would be interesting to note in more detail the proportions of the 
constituents of the flesh, especially as illustrated in Tables IT, IT, and 
IV. The constituents soluble in cold and hot water, for instance, which 
so far as is known are analogous to those of meats, are of considerable 
interest, but demand much more investigation. Meat extract has be- 
come an important article of commerce. There is a fortune for some- 
body, I mistrust, in the extract from menhaden. 
“FOUL” OR “SPENT” FISH vs. THE SAME IN GOOD CONDITION. 
Some very interesting results are found in comparing the composition 
of the foul or spent fish witb the same in good condition, as shown in Table 
II. As the fish becomes lean, it loses nutritive value in three ways; first, 
in total loss of weight; second, in relative increase of refuse and decrease 
