16 
leaving the calorimeter. As a rule there was a change of diet in 
respect to one class of ingredients; for example, the replacement of 
sugar by an isodynamic amount of fat. 
In all cases a preliminary period, usually of four days' duration, 
preceded each experiment or series of experiments with the subject in 
the chamber of the calorimeter. During this preliminary period the 
subject received the same diet as was to be served in the experiment 
which would immediately follow. In addition, the excretory products 
were collected and analyzed, thus making this period both a digestion 
and a nitrogen metabolism experiment. On the evening of the last 
day of the preliminary period J:he subject entered the chamber of the 
respiration calorimeter and usually retired at about 11 o'clock. At 
about 1 o'clock in the morning the heat measurements were begun in 
order to get the calorimeter into normal condition for the commence- 
ment of the second period, or the metabolism experiment proper, at 
7 o'clock the following morning. 
The programme followed during the experiment or series of experi- 
ments was arranged beforehand. A copy of it was furnished to the 
subject, and copies were also posted in convenient places for the guid- 
ance of those outside. 
PREPARATION, SAMPLING, AND ANALYSIS OF FOODS. 
The method of preparing, sampling, and analyzing the food mate- 
rials was the same as that employed in experiments Nos. 9 and 10, 
already described. a 
With the exception of milk, the proper quantity of each kind of 
food, either for each meal or for the day, was placed in glass jars pre- 
vious to the beginning of the experiment, and materials which might 
spoil during the course of the experiment, such as meat and bread, 
were thoroughly sterilized. Butter and sugar, in quantity sufficient for 
each day, were passed in with the breakfast, and the subject used, as near 
as he could judge, the proportions called for by the menu for each meal. 
Sufficient bread for the day was usually passed in at breakfast time 
in two jars, and as nearly as possible the scheduled amounts used at 
each meal. Two or more duplicate jars of each food material, as thus 
prepared, were analyzed, the usual determinations being made accord- 
ing to the methods already described. The milk was obtained fresh 
each day; an aliquot portion was taken for the preparation of a com- 
posite sample for analysis, and the proper amounts for each of the three 
meals were placed in bottles. 
In the following table is shown the percentage composition of the 
different food materials used in the thirteen experiments reported 
herewith. The values are for, the fresh, edible material as served to 
the subject. 
a U. S. Dept. Agr., Office of Experiment Stations Bui. 69, pp. 20-29,88-89. 
