Although, with normal young men having presumably normal digestion, standard factors for digestibility of our common food materials may be used to compute the probable fecal output of nitrogen and energy per day, it seemed necessary, when there were to be material reductions in the diet, not to assume that there would necessarily be the same proportion of energy and nitrogen lost through the feces as would commonly occur with people subsisting upon full diet. Consequently arrangements were made to obtain representative samples of feces from time to time. It was impracticable, owing to the long period of the investigation, to collect the feces for the entire time. Such collections were made at the beginning of the experiment and approximately every other week throughout its continuance. The length of each period for the collection of feces ranged from 3 to 16 days.

The feces were collected in numbered sheet-iron pans, one of which was assigned to each member of the squad; care was taken to avoid loss of urine at the time of defecation. The pans were covered with tin covers and the subjects carefully instructed, especially at the beginning and end of the period of collection, to indicate the first portion of feces passed.

The separation of feces was made by administering a marker of some kind. At first powdered charcoal or lamp black was used, but later, in accordance with the extended experience at Battle Creek Sanitarium, carmine was employed. Markers were put in gelatine capsules and 3 capsules were given simultaneously according to the following schedule: On the day of the first collection of feces, the marker was given before breakfast; the feces colored with the carmine or charcoal were then considered a part of the feces belonging to the collection period. At the end of the collection period, a marker was given with the following breakfast and feces preceding the colored portion included in the feces for this period.

After the separation of the colored portion of the feces, the portions to be analyzed were placed in glass jars in exactly the same manner as the food samples were handled, 5 c.c. of concentrated hydrochloric acid added to each jar and shipments made in special cases to guard against breakage.

Constipation

As a result of the marked reduction in diet, constipation was of frequent occurrence, especially in the earlier stages of the research. To offset this and profiting again by the extended experience at Battle Creek Sanitarium, bran was administered in amounts desired by the subjects. While practically all of the men were entirely unused to bran, nearly all of the subjects asked for it and in rather large amounts. Indeed, so much bran was taken that our attention was called to the fact that this substance, instead of being mostly indigestible fiber, contained a really large amount of available energy. Consequently we found, much to our chagrin, that no small part of the energy in the diet was supplied by the large amounts of bran used which should therefore be taken into account. It thus became necessary either to curtail the use of bran or to substitute it for other food materials.

In rare instances the subjects resorted to some simple physic, as salts or oil. In no case was oil used during the periods when feces were collected; the energy of the feces was therefore not contaminated by the energy of unabsorbed petroleum oil. Patent preparations containing bran in various forms, such as biscuits, were liked by the subjects and these, combined with an extensive use of bran, readily served to hold constipation in check. Chef Hall also made some excellent bran muffins which were greatly relished not only by the squad but by the entire student body. The diet was further modified in the latter part of the period by the use of somewhat bulky food materials, such as spinach, which supplied coarse ballast. By these simple dietetic means constipation was controlled.