4 BULLETIN 649, U. S. DEPARTMENT OF AGRICULTURE. 
were allowed to follow their customary dietary routine preceding and 
following the experimental period. For the purpose of identifying 
the feces of the experimental period, three or four gelatin capsules 
containing about 0.3 gram of pulverized charcoal were taken with 
the first meal of the experimental period and with the first meal fol- 
lowing it ; the separation of the feces due to the diet under investiga- 
tion was easily made at the lines of demarcation made by the por- 
tions dully colored by the charcoal. 
Inasmuch as this study is concerned principally with the coefficient 
of digestibility of the protein and fat of the fish, no attempt was made 
to maintain a nitrogen equilibrium or uniform body weight of the sub- 
jects. The urine resulting from the experimental periods was not 
collected, for it was considered that any constituents of the foods 
which had been sufficiently broken down to appear in the urine 
had undergone the process of digestion; furthermore, the results ob- 
tained by collecting" and analyzing the urine of a short test period 
are not entirely conclusive since the urine can not be separated as 
satisfactorily as the feces. 
PREPARATION OF FISH. 
In this study of the digestibility of different types of fish, fresh 
butterfish and Boston mackerel were used, and canned grayfish and 
canned salmon. A fish loaf seemed to be the best form in which to 
prepare the fish for eating, since sufficient quantities for the entire 
experimental period could be prepared at one time. Furthermore, 
it was easy to prepare a fish loaf having a uniform- composition and 
one which would not change materially on standing by the settling 
out of fat or evaporation of water. 
The butterfish and mackerel received a preliminary cooking before 
being incorporated in the fish loaf. The fish, after being cleaned, were 
thoroughly washed and placed as close as possible to each other in a 
covered cooker, water was added, and they were cooked for one-half 
hour. They were not boiled, but steamed in a very small quantity 
of water which prevented browning or sticking to the pan. To 
prevent any loss due to extracted fat and protein, the water in 
which the fish were steamed was retained and mixed with the fish 
meat, Beythien 1 having reported that the water in which fish 
were boiled contained 8.8 to 11.3 per cent of the total fish protein. 
After this preliminary cooking of the butterfish and mackerel, the 
bones, any pieces of fins, etc., were removed and the fish meat was 
cut in an ordinary household meat cutter. The bones and bits of ♦ 
skin were removed from the canned grayfish and salmon, and the 
solid meat was minced in a meat cutter. From this point the prepa- 
i Pharm. Centralhalle, 47 (1906), p. 140. 
