4 BULLETIX 310, U. S. DEPARTMENT OF AGRICULTURE. 
EXPERIMENTAL METHODS. 
Each test period included three days, or nine meals, a period long 
enough for the purpose, but, as shown by experience, not of sufficient 
duration to make the diet distasteful. In order that the subject 
should experience no sense of monotony, the ration in question was 
served for only one test period per week, followed by a rest period of 
four days, in which the subject lived as usual on a mixed diet. Fur- 
thermore, the tests with fat were alternated with tests of an entirely 
different sort of diet, so that the same sort of food was not eaten 
oftener than on alternate weeks. 
The subjects changed from time to time according to circumstances, 
but, since their ages, occupations, and modes of living did not vary 
materially, the results obtained for the different fats are believed to 
be directly comparable. The men were between 20 and 30 years of 
age, students in medical and dental schools, and sufficiently informed 
in physiological questions to appreciate the importance of accuracy 
in carrying out such directions as were given them. They were urged 
to observe special care in the collection of feces, for accuracy in this 
particular is essential in determining the amount of food retained and 
assimilated by the body. They were asked to supply data as to their 
physical condition, both before and after, as well as during the time 
occupied by the experimental periods, but since all remained in 
normal physical condition none of the individual reports are incor- 
porated in this paper. 
The amounts and kinds of food eaten during each three-day exper- 
imental period were recorded as well as the total quantity of feces, 
and samples of both food and feces were analyzed to determine the 
quantities of protein, fat, carbohydrate, and mineral matter retained 
and assimilated by the body. Since sufficient quantities of blanc 
mange were prepared to serve for an entire test period, only one 
analysis of the blanc mange was necessary. Only occasional analyses 
of the wheat biscuit were made, as it was found to be of very uniform 
composition. 
In order to f acilitate the identification of the feces of a test period, 
charcoal, 1 which imparts a dark color to the feces, was taken with 
the first meal of the test period and with the first meal following the 
test period. The feces first showing a dark color and all excreted 
until the color imparted by the charcoal was again noticed were re- 
tained for analytical purposes. The feces were freed from water by 
drying at 95° C, and analysis was made on the dry basis. The urine 
was not collected and no determinations of the heats of combustion 
of food or feces were made, since factors have been reported by 
means of which the available energy value can be computed with 
i U. S. Dept. Agr., Office Expt. Stas. Bui. 143 (1904), pp. 6&-77. 
