It was hoped that in this study both the nitrogen intake and the nitrogen output could be sufficiently controlled throughout the entire time so that a complete nitrogen balance could be obtained and at the end of the experimental period the exact nitrogen loss to the body could be determined in addition to the changes in body-weight. It was psychologically undesirable to restrict the men at every meal, and free days were thus occasionally permitted, with, however, certain suggestions as to restrictions which it was believed the men would more or less heed. As shown on page 270 (table 34), on the uncontrolled Sundays the food taken by the men contained much larger amounts of energy than they ordinarily consumed, often more than twice the amount, but the nitrogen intake was by no means so large proportionately. There were, of course, occasional exceptions, such as the 28.47 grams of nitrogen taken on January 13 by Kon and Gul, and 32.7 grams by Moy. But on the average the squad consumed on these uncontrolled days 16.62 grams of nitrogen, a value 50 per cent higher than that indicated for the average intake per man per day during the whole experimental period (10 grams). In addition to these uncontrolled Sundays we have likewise four days at the Thanksgiving recess of uncontrolled period and also a somewhat lengthy period of 18 days at Christmas time. These breaks were simply unavoidable.

Aside from these uncontrolled periods, the nitrogen in the food was determined every day. The urine was also collected with remarkable fidelity on the part of the men and the nitrogen in the urine determined. Thus, we have sufficient data with Squad A to indicate the general picture of the nitrogen balance throughout the entire 4 months. With Squad B the conditions were much more satisfactory from the experimental standpoint. The men were on the diet for a period of only 20 days, there were no days of uncontrolled diet, and the balance between the nitrogen intake and the nitrogen output may be definitely determined.

In any final summation of the nitrogen balance for the period of 4 months during which the experiment continued, the nitrogen intake during these unrestricted Sundays and the holiday periods must be taken into account. On 5 uncontrolled Sundays the men made reasonably close estimates of the food taken. These records are given with the computed energy and nitrogen content in table 34. It might be assumed that the amounts of nitrogen ingested on the uncontrolled Sundays could be taken as an index of the nitrogen intake on the Thanksgiving and Christmas holidays, but it is by no means certain that this would be a legitimate assumption. From the reports of the men regarding their dietetic habits during these vacations, it is clear that the high rate of nitrogen ingestion on the uncontrolled Sundays was by no means continued through the entire vacation periods. The fact that the men were morally obligated to return to college at or near their last recorded weight certainly acted as a deterring influence upon excessive consumption. It does not, therefore, seem justifiable to us to assume an average intake of 16.62 grams of nitrogen for each day not specifically noted in our tables between September 27 and February 3.

It may be seen, however, that any nitrogen balance which may be made without taking into consideration the nitrogen intake of these days will undoubtedly be defective in that the apparent intake of nitrogen will be perceptibly lower than the real intake. On the other hand, the record of the output of nitrogen when measured by the nitrogen of the urine and feces is also deficient in that the urinary and fecal analyses take no note of the loss through the skin, perspiration, epithelial debris, growth of hair and nails, etc. Evidence with regard to the loss through the skin with normal individuals has been somewhat extensively discussed elsewhere.1 In this earlier study it was found that on the average with men during complete muscular rest, living a restricted life inside the respiration calorimeter, about 0.1 gram of nitrogen per 24 hours was excreted through the skin. During severe muscular work with free perspiration the experimental data showed that there may be as much as 0.22 gram of nitrogen excreted per hour. As may be seen from the record of the physical activity of the men in the low-diet research, they were by no means at rest. A large number of them indulged in severe physical exercise, such as gymnasium work and running. Although Professor Johnson reports that in the 5-minute bicycle riding period the men in Squad A perspired very rarely and very slightly as compared with the men in Squad B, who were on normal diet, nevertheless it is unquestionably true that the gymnastic work these men engaged in, and their occasional trips to the swimming-pool, would cause a considerable loss of cutaneous nitrogenous material.

1 Benedict, Journ. Biol. Chan.. 1906. 1. p. 263.

Attempts have been made by various writers to estimate the loss through the skin and through the growth of hair and nails. Thus, Taylor1 summarizes that such loss of nitrogen can not be less than 0.3 gram per day on the average. With the greater average activity of our subjects, as compared with the activity of the ordinary individual, 0.4 gram per day would be a closer estimate of the loss of nitrogen in this way. Hence, it can be seen that, in each case, 0.4 gram per day should be added to the nitrogen outgo to obtain the true nitrogen balance. Since the whole research extended over 103 days, we have an excretion of probably not less than 40 grams of cutaneous nitrogen which is not considered in either the urinary or the fecal loss. This will, in large part, at least, compensate for the excess nitrogen taken on the uncontrolled days. Accordingly, since there is uncertainty both as to the actual amount of nitrogen taken on the uncontrolled days and the cutaneous excretion, it seems best for purposes of discussion to omit both factors and to make the assumption, which may fairly be challenged, that these two more or less offset each other. With this preliminary announcement, therefore, we may consider the nitrogen balance as computed from the determination of the nitrogen in food, feces, and urine.

A suggestion of great nitrogen losses appears in a previous section in a general comparison of the food intake and the nitrogen in urine, (see page 300) but no consideration was given to the nitrogen in feces. In the comparison given in this discussion, a more exact balance is made by allowing for the fecal nitrogen. These nitrogen balances are computed for each member of Squad A and combined in tabular form with the comparisons of the energy intake and output. The energy balances will be discussed later but are tabulated here for convenience. The great significance of the nitrogen balance makes it desirable, furthermore, to present all of the data in detail, and this is done in tables 46 to 58 for Squad A and in tables 59 to 70 for Squad B.

1 Taylor, Digestion and metabolism, Philadelphia, 1912, p. 485.