Practically all of the men in Squad A show much greater differences between the predicted and found values at the period of minimum weight than at the beginning of the experiment when they were on normal diet. Thus we find that under normal conditions of diet 7 out of the 12 men considered show values greater than predicted. The average difference for the entire squad was 1.1 per cent above the predicted, the striking exception being Mon, with a plus difference of 12.5 per cent. At the period of minimum weight-level, namely, the latter part of November, all of the men, with the single exception of Mon, had values considerably below those predicted. Mon's metabolism was in excess of the predicted values by 1.5 per cent. The average computed value for the 12 men was 11.7 per cent below the predicted value. At the end of the reduced-diet period every man shows a computed heat production lower than the predicted value, although in the case of Mon this difference is insignificant, being only 2 calories, or 0.1 per cent. The average for the 11 men was 14.6 per cent below the predicted values. It is a matter of considerable importance that throughout the entire study Mon retained a consistent relationship in his total metabolism not only with the rest of the squad, but with the predicted values. When it is considered that we are comparing the total basal heat production as actually measured by the gaseous metabolism with that predicted for individuals of similar age, height, weight, and sex, as derived from a series of prediction equations, it is obvious that the basis of comparison is much more fundamentally and firmly established than when the comparison is made on the basis of body-weight or body-surface.

Table 133. - Heat Production At Different Diet Levels As Computed From Respiration Experiments And As Predicted By Harris And Benedict - Squad A; Subjects Post-Absorptive And In Lying Position

1 The normal diet period includes experiments during Sept. 27 to Oct. 4, 1917, inclusive.

2 See tables 115 to 127.

3 The normal values for Kon were obtained on Oct. 27, 29, and 30.

4 The period of minimum weight-level represents 5 experiments on or about Nov. 24.

1 The period at the end of the reduced diet represents 5 experiments on or about Jan. 26, 1918. 2 See tables 115 to 127.

It is seen, therefore, that not only on the basis of body-weight, as shown in earlier tables, as well as the questionable basis of body-surface, but also from these biometric prediction values, the basal metabolism was profoundly lowered with every member of Squad A. It is furthermore of significance that of the four men with pronounced negative values in column g on the normal diet, namely, Bro, Can, Tom, and Vea, three of them, Bro, Tom, and Vea, show absolutely greater negative values at the end of the experiment than all of the other subjects except Pea, thus suggesting a tendency to retain a relative position in the group so prominently indicated in the positive values noted with Mon.

Group Measurement Of Basal Metabolism With Low Diet, Squad B

Immediately following the night of January 6 inside the respiration chamber, Squad B was put upon a much restricted diet which averaged 1,375 net calories. The men came to Boston for experiments in the group respiration chamber on January 13, 19, and 27, respectively. The minimum carbon-dioxide excretion per hour for the low-diet period, as found with the group chamber, also the computed heat per kilogram per hour and per square meter per hour are given in table 134. The normal values, previously recorded in table 114, are also included in this table. With the change to a much-restricted diet, there was a pronounced decrease in the carbon-dioxide production, a decrease in the heat per kilogram per hour amounting to but 1 per cent, and an appreciable decrease in the heat per square meter per hour. On January 20 the heat on both bases of computation was lower; on January 28 the minimum values were found.

Table 134. - Minimum Metabolism During Sleep As Measured In Group Respiration Chamber - Squad B

In examining this table one particularly regrets that normal values immediately subsequent to January 6 were not obtained before the restriction in diet was begun. The pronounced fall in the metabolism from December 16 to January 6, which, as previously stated, we are unable to explain, should have been further studied.1 Although the fact that the heat per kilogram fell only slightly on January 14 might indicate that a further fall without dietetic restriction was hardly probable, nevertheless it represents a faulty point in the experimental plan. To begin a study of the effect of a superimposed factor upon a sliding base-line is experimentally bad. Furthermore, when so pronounced a fall in metabolism was shown on January 28, it is much to be regretted that the experiment could not have continued further. It will be remembered that Squad B was placed upon a restricted diet for the main purpose of obtaining data for comparison with some of the important findings noted in the earlier phases of the experiment with Squad A. Squad B, with a praiseworthy self-sacrificing spirit, volunteered to live upon this greatly restricted diet for a short period, so that the desired data might be obtained.

1 See note on page 500.

In analyzing the figures obtained in these experiments, one is confronted by the difficulty of selecting the proper normal value for comparison. As previously stated, it is obvious that the only one that can be logically selected is that for January 6, since it was obtained immediately before the restriction in diet. Using this value, we find that the heat production per kilogram of body-weight per hour fell until on January 28 it was 13.3 per cent lower than it was on January 6. The heat production per square meter of body-surface per hour also fell until on January 28 it was 17.1 per cent lower than on January 6. With Squad A the computed heat values for the final experiment of February 2-3 gave values which indicated that there had been a decrease of 19 per cent in the heat production per kilogram of body-weight over the initial value found on September 29-30. For the heat production per square meter of body-surface it was 23 per cent. These findings with Squad B of 13.3 and 17.1 per cent, respectively, for the same factors are not quite so large, but the loss in body-weight and what is, we believe, of still more importance, the loss of body-nitrogen was by no means so great as it was with Squad A.

Using the normal values obtained with this squad in October, we may compute the total decrease in the heat output as shown by the value obtained on January 28. On this basis we find that there has been a decrease in the heat production from October 7 to January 28 (including both the seasonal variation and the variation due to the restricted diet) of 22.7 per cent, and on the basis of the heat per square meter of body-surface 24.7 per cent. These values actually exceed somewhat those found with Squad A. Had it not been for the series of observations on Squad B with uncontrolled diet on November 3-4, November 17-18, December 15-16, and January 5-6, these values would have been extremely difficult to interpret. The difficulty of applying an arbitrary base-line or an arbitrary so-called "standard" value to a series of experiments is apparent, even when group measurements are made.

The general picture of reduction, then, is approximately what would be expected from the loss in body-weight and the loss in nitrogen with Squad B as compared with Squad A. Both show pronounced alterations of the basal metabolism as the result of the restricted diet. The decrease is not only absolute but relative, being from 13 and 23 per cent lower per kilogram of body-weight and from 17 and 25 per cent per square meter of body-surface.