The average nitrogen excretion of Squad A from December 3 to the end of the experiment was 10.5 grams. (See table 73.) This probably represents on the whole the most constant period of lower weight during the experiment, although the second part of this period is complicated by the abnormal rises in weight incident to the Christmas vacation. We may take this nitrogen value, therefore, as an approximate indication of the level of nitrogen upon which these men were capable of living with their lowered intake.

A comparison of the nitrogen excretions in the urine at these different weight-levels is of great significance as indicating the possibility of a material reduction of protein in the diet, protein being one of the most costly food constituents. After the extensive experience of Professor Chittenden, and more particularly the recent compulsory experience of the Central Powers, any real danger from a reduction of protein in the intake seems to be lacking. It was emphasized at the outset that we did not intend to complicate our problem by control of the protein intake. We were not advocating either a high or a low protein diet or a vegetarian or mixed diet. We were confessedly somewhat surprised at the conclusion of the research to find that the nitrogen excretion, especially at the lower weight-level, and with a material reduction in the caloric intake still remained so high, for we had expected it to be somewhat lower. The fact that this did not come to our attention until some time after the experiment ceased shows the complete objectivity of our method of weight reduction and food control. The criterion as to food allotment was, in every instance, the actual weight of the subject. Since the exact knowledge of the intake of nitrogen and energy depends upon chemical analysis and determination of the heat of combustion, the actual ingestion of energy and nitrogen per day could only be computed several weeks later. This in part accounts for the noticeable variations in the intake of energy and nitrogen frequently appearing in the tables.

We have no reason to believe that a somewhat lower protein level might not have readily been obtained without a correspondingly great increase in calories. This, however, remains a disputed point. The fact that nitrogen equilibrium, or at least an indication of nitrogen equilibrium in the frequent appearance of plus values at the lower weight-level, was obtained with the surprisingly low caloric intake of 1,950 calories is, we believe, a new feature. It has been supposed from earlier feeding experiments that no material reduction in the nitrogen excretion can be maintained without very considerably increasing rather than decreasing the calories. Professor Chittenden, without absolute caloric dietetic control, showed that his soldiers were able to subsist with a low nitrogen output with no great increment in calories. Indeed, his evidence strongly suggested lowered calories. This research fully confirms his inferences with regard to this point.