E. D. Clark and L. H. Almy A85 
In addition to Atwater’s classical work on American fishes, the litera- 
ture contains numerous references bearing on the chemical composition 
of food fish. The earlier investigations have been carried on mainly 
by foreign workers, among whom are Schlossberger (13), Morin (14), 
Weidenbusch (15), von Baumhauer (16), Payen (7), Kénig (6), Buckland 
(5), Almen (4), Kostytscheff (8), Popoff (17), Sempolowski (18), and 
Kianizyn (19). The results obtained by most of these authors have been 
discussed at length by Atwater (1). Their analyses consisted, generally, 
of determinations of solids, fat, ash, and nitrogen. There has been no 
concerted analytical study of the composition of American food fish since 
1888. European food fishes have been studied more recently, and among 
the investigations those of Balland (20) and Hollande (21) are especially 
noteworthy. A comparison of the food value of beef, veal, etc., with 
fish has been made by Milone (22), Beythien (23), and Kénig and Splitt- 
gerber (24). 
Methods. 
History of Samples. 
The fish were all obtained from Philadelphia fish dealers. No 
fish were purchased of which an accurate history could not be 
obtained. This history indicated good handling according to 
present commercial standards and a minimum lapse of time 
after catching. The analytical work was begun immediately upon 
the receipt of the sample. | 
Preparation of Sample for Analysis. 
The fish were cleaned by separating as completely as. possible 
from the edible portion the head, tail, skin, entrails, and 
bones. The edible portion was weighed and the difference be- 
tween this weight. and that of the whole fish represented the 
refuse, or inedible portion. The flesh was ground in a meat 
chopper, and the individual samples were mixed thoroughly by 
hand. As the subsequent analysis requires 2 or 3 days of one 
analyst’s time for completion, it was not convenient to undertake 
more than one complete analysis of a single species of fish at a 
time. In the case of large fish but one specimen was used, but of 
the smaller fish from 2 to 8 were analyzed, the edible portions 
being united and well mixed before sampling. 
