Vol. 7, 1921 PHYSIOLOGY: LANGWORTHY AND HOLMES 
121 
was divided into groups of five subjects each, and one group was served 
the 54%, 70%, 85%, and 100% flours in the order given while the second 
group was served the flours in the reverse order. The results of the 139 
digestion experiments reported below do not indicate that the first group 
of subjects acquired any tolerance for the coarser flours not possessed by 
the second group which was served the coarser flours first. 
The experimental diet was so chosen that its preparation should involve 
a minimum of labor, that the bread should supply the larger portion of 
the total protein, and that it should be representative of a simple mixed 
diet. The diet consisted of bread, fruit (orange), butter, sugar, with tea 
or coffee as a beverage if desired. 
The fruit, butter, and Sugar were served as purchased from a nearby 
market. A quantity of bread sufficient to supply all the subjects for one 
day was prepared daily, the ingredients being used in the following pro- 
portion : 
3 cups sifted flour 
1 teaspoon salt 
1 tablespoon sugar 
1/8 to 1/2 cake yeast 
1 tablespoon hydrogenated vegetable oil 
1 cup water 
While data relative to the digestibility of the diet as a whole are of value, 
interest was primarily centered in the digestibility of the bread or rather 
the flour from which the bread was made. The digestibility of the pro- 
tein and carbohydrate supplied by the different breads has been estimated 
by making a correction for the undigested protein and carbohydrate re- 
sulting from the accessory foods by a method outlined in detail in earlier 
publications. 4 It has been assumed from the results of the early investi- 
gations of this Office 5 that the digestibility of the protein of butter is 97% 
and protein of fruit, 85%; of the carbohydrate of fruit, 90%; and of sugar, 
98%. 
In planning the investigation, it was decided to make a sufficient number 
of digestion experiments so that the average results obtained should be of 
general application. The experimental periods were to be long enough 
so that the effect of any irregularity in the rate of passage of food residues 
through the alimentary tract, and any error in the separation or collec- 
tion of feces would be practically negligible. The experimental periods 
which were from 15 to 25 days in length were subdivided and considered 
as separate three-day experiments, following one directly after the other. 
The subjects for the experiments were students in local educational 
institutions, and were selected with care in order that the conclusions might 
be applicable to the average normal adult. An attempt was made to 
include in the squad some accustomed to strenuous exercise, some ac- 
customed to take light exercise, and some who took little recreation, in 
