[13] CHEMICAL COMPOSITION AND VALUE OF FISH FOR FOOD. 243 
Drying.—The drying was conducted in an ordinary bath at a temper- 
ature of nominally 100°, but actually about 969, as is usual in drying: 
baths which consist of an air chamber with double walls inclosing boil- 
ing water. In each case two portions were dried, one, “A,” in a current 
of hydrogen, and one, “ B,” in air. 
Drying in hydrogen.—F rom 50 to 100 grams, in most cases preferably, 
100 grams, of the freshly-chopped substance was weighed on a watch 
glass or small sauce-plate, dried in a current of hydrogen at 969° for 24 
to 48 hours, cooled, allowed to stand in the open air for some 24 hours, 
weighed, ground, sifted through a sieve with circular holes of 1™™ di- 
ameter, bottled, and labeled “A.” A few of the fattest samples, how- 
ever, could not well be worked through so fine a sieve; for these, either a 
coarse sieve was used or the substance was crushed as finely as practi- 
cable, and bottled without sifting. For the complete drying, from 1 to 
2 grams of A” were weighed in small drying flasks, and dried in hydro- 
gen three or four hours. It is extremely difficult to get an absolutely 
constant weight, though we find that the object is in most cases approxi- 
mately attained in three hours. The total moisture and dry substance 
are computed from the two dryings. 
Drying in air.—As the drying of large quantities in hydrogen was less 
convenient, and drying in air suffices perfectly well for certain determi. 
nations, particularly sulphur, phosphorus, and chlorine, a sufficient 
quantity of material for the work was insured by drying a portion in 
air. This was effected by weighing 200 grams or more of the freshly- 
chopped substance at the same time that the portion was taken for dry- 
ing in hydrogen, drying in air, exposing to air of room, weighing, grind- 
ing, and bottling as above, this portion being labeled “B.” The amount 
of water-tree substance in “B” was calculated from the data obtained 
font A? 
Prowvimate ingredients by direct determination.—In a number of the 
samples, determinations were made of the ingredients soluble in cold 
and in hot water, and of the portion not dissolved by water, alcohol, or 
ether. 
The objects of the determinations were to obtain data comparable with 
those of other investigations*, and to test the methods, as well as to learn 
the amounts of the ingredients. The methods have proved unsatisfac- 
tory in many respects, and we have felt it advisable to make no more de- 
terminations by them than are indicated in the tables, until the subject 
is worked up more thoroughly. For that matter, an at all satisfactory 
examination of the proximate constituents will naturally involve deter- 
minations of both the total amounts and the ultimate composition of the 
ingredients of the juices, as well as solids, of the flesh. Considering the 
complicated character of these compounds, the vagueness of our present 
knowledge regarding them, and the amount of preliminary work that is 
* EB. g., Almen, Analyse des Fleisches einiger Iische (Nova Acta Reg. Soc. Sc. Ups., Ser. 
III), Upsala, 1877. 
