DIGESTIBILITY OF RAW STARCHES AND CARBOHYDRATES. 3 
days in which the ordinary diet was followed. The foods were 
attractive in appearance, and the table was neatly set with white 
covers and dishes. The subjects were given weighed portions of the 
frozen pudding along with a weighed ration of oranges and sugar, 
and tea or coffee as desired. In a preliminary experiment it was 
- found that subjects were able to eat 300 grams of frozen pudding 
at a meal, and accordingly this quantity was usually weighed out for 
each person. If any of the weighed portions was not eaten, the 
residue was weighed and careful account was kept of the amount 
eaten by each subject. 
_ The quantity of raw starch usually eaten per woman per day was 
_ about 170 grams, or a little over a pound in three days. This quan- 
_ tity of raw starch quite closely approximates that eaten by the 
subjects in the experiments with men. The quantity of oranges 
- consumed was limited to that consumed by the men. The quantity 
of sugar used in beverages was within limits left to the discretion 
of each subject and varied from none at all to 100 grams per woman 
daily. In determining the coefficients of digestibility of starch 
alone, corrections were made for undigested sugar. The quantity 
of sugar consumed did not seem to affect the digestibility of the 
_ starch as so determined. | 
The feces belonging to an experimental diet were marked by car- 
- mine taken with the first meal and were separated from the feces of 
the subsequent regular diet by lampblack taken with the first regular 
meal. It was noticed that the carmine colored the feces for two or 
three days after it was taken. The drying and analysis of foods 
and feces and the correction for metabolic products were all made 
according to methods used in the experiments with men. 
It was noticed in mixing the materials to make the frozen pudding 
that the raw cornstarch did not blend well with the oil and milk, 
and this made it difficult to obtain uniform samples for moisture and 
fat determinations of the cornstarch pudding. The other pure 
starches mixed better with the oil and milk, and the flours, farina, 
and corn meal mixed well with the oil, so that in these puddings the 
samples obtained were uniform and the moisture and fat determina- 
tions agreed closely in the five samples taken of each pudding. The 
ability of the flours and starches to mix well with the oil and milk 
as contrasted with the poor mixture obtained with the cornstarch 
can probably be explained by difference in surface attraction of the 
raw materials for the oil. 
In order to determine the coefficient of digestibility of the carbo- 
hydrate of patent flour, graham flour, farina, and corn meal, an 
_ analysis was made of these substances and the amount of carbo- 
Eparate to be attributed to each in a given diet was calculated from 
the amount of the substance known to have been eaten in the frozen 
pudding. 
Microscopic examinations of the feces were made by one of the 
authors in the microchemical laboratory of the Bureau of Chemistry 
~ to eT whether or not they contained undigested grains of raw 
starch. 
The nitrogen determinations were made in the nitrogen laboratory 
and the fibre determinations in the food-control laboratory of the 
Bureau of Chemistry. Determinations not otherwise specified were 
made in the laboratory of the Bureau of Home Economics. 
q 
oak 
