Normal, healthy man, subsisting upon modern well-prepared and well-cooked food materials, exhibits a uniformity of digestive processes that is, in a sense, rather remarkable. The so-called "digestibility' ' of our modern food materials can be predicted from standard figures with great accuracy. Hence a digestion experiment as such, particularly when ordinary food materials are used, is hardly justifiable. On the other hand, with a great restriction in diet, the evidence is not sufficiently extensive to show whether or not there would be a disturbance in the digestive processes. One criticism of the classic experiment of Professor Chittenden1 with soldiers was that the amount of nitrogen excreted in the feces of a group of men, presumably with low diet, varied within very wide limits from that which would be expected, or was regularly found with normal individuals. This suggested the possibility, at least, that a restriction in protein had resulted in an actual disturbance of the digestion processes. It therefore became necessary with our subjects to make periodic, so-called "digestion experiments".

These digestion experiments were made with two purposes in view: (1) to note abnormalities if they existed, and (2) to give positive information as to the amount of unoxidized material leaving the body from the alimentary tract. Theoretically, it would have been best to have had a collection of feces throughout the entire time, but this presented technical difficulties which were so great as to make such collection impracticable. We were able, however, to obtain six or seven digestion periods with most of the men in Squad A, these ranging from 3 to 16 days each. The first was October 1 to 4, when the subjects were on normal diet; subsequent periods on reduced diet were from October 8 to 12, October 17 to 21, October 31 to November 4, November 12 to 18, December 10 to 15, and finally, a period of a little over two weeks, January 14 to 30. This schedule was followed by practically all of the subjects except Pec, whose habits of defecation were so abnormal as to make it impossible for him to carry out this program satisfactorily. With Squad B one digestion experiment was made, January 15 to 23, when the men were on a reduced diet.

1 Chittenden. Physiological economy in nutrition, New York, 1907, p. 131.

The common interpretation of digestion experiments is based upon a fundamentally erroneous conception that the feces represent primarily undigested food. The presence of visible portions of undigested material in feces naturally leads to this belief, but chemical and microscopic analysis shows this is far from the case, and that the feces consist in large part of bacteria, intestinal debris, and residues of digestive juices. Consequently it is clearly erroneous to determine the digestibility of any given nutrient by the comparison, for example, of the amount of protein in the food and the amount of protein in feces. In our experiments no attempt was made to determine the so-called bacterial nitrogen or the so-called metabolic nitrogen; but since our work was primarily a matter of relative comparisons, we have adhered to the archaic form and present our data for these digestion experiments in terms of the utilization of nitrogen and the availability of energy in food for digestion periods. The nitrogen in feces is subtracted from the nitrogen in food and the remainder considered as utilized nitrogen, and the percentage of the total ingested is recorded as the percentage of nitrogen utilized. Similar treatment is given the data for energy, except that not only the calories in feces are deducted from the energy in the food, but likewise the calories in the urine. The calories in the urine are computed by multiplying the nitrogen in the urine by the factor 8.0. The total net calories are thus the total calories in food less those in feces and urine. The percentage of the total net calories compared to the total energy in the food ingested is expressed as availability of energy.

It is a matter of considerable regret that experimental evidence on the character of the feces has been for so many years neglected. Even the present available data regarding this are very fragmentary and no rational method for the comparison of food intake with the several ingredients of the feces is as yet universally employed. One frequently sees the statement that feces are formed during fasting. Whatever may be the case with animals, it is certain that during the 31-day fasting experiment at the Nutrition Laboratory with man no evidence was obtained of fasting feces. On the other hand, evidence was obtained to show that fasting does not completely kill the bacterial action, for Professor Kendall1 found bacteria in the colon. With the confusion existing at the present time, not only as to a classification of the exact facts known, but more especially with regard to the absence of any logical method of procedure, we must for the present adhere simply to the original plan of considering the nitrogen in the feces as being derived from food, and express our results according to the commonly accepted method.

In our digestion experiments we do not consider the digestibility of fat. It should be pointed out here that the common methods for analyzing the feces are wholly unsuited for the proper determination of the digestibility of fats, for the very large proportion of soaps in feces which are insoluble in ether makes a crude ether extract of feces wholly unsuitable for an estimate of the total fat content. In this research we are primarily interested in the question of whether or not there is a profound disturbance in the proportion of nutrients digested, as commonly expressed, when a diet is used which is very low in calories and moderately low in nitrogen.