In our effort to secure basal metabolism we have selected and reported in table 114 the values representing the minimum carbon-dioxide production per hour throughout the night. Most of the periods of experimentation throughout the night were from a half hour to an hour in length, and the values for the carbon-dioxide production here given are based upon the average of two or three and frequently more periods. Since there is, however, some discrimination permitted in the selection of the minimum period and an error in selection might profoundly affect the low value found, we have computed the average carbon-dioxide production for the measurements made throughout the night. The values for the minimum carbon-dioxide production per hour given in table 114 are naturally somewhat lower than the averages found for the entire night, being usually not far from 5 to 10 grams.1 Thus, on the night of October 6-7 the total average carbon-dioxide production for Squad B was 297 as against a minimum of 292 gm.; on November 3-4, 297 as against 295 gm.; on November 17-18, 299 as against 292 gm.; on December 15-16, 296 as against 287 gm.; and what is of special interest, on the night of January 5-6, 276 as against 265 gm. Under these circumstances, it is perfectly clear that the value of 265 gm. is a probable minimum value and not an erroneous selection based upon one or two periods. On account of the construction of the group chamber and the fact that duplicate weighings are made in each period, there is almost no possibility of an error in the determination. Evidently, not only the minimum value but also the total carbon-dioxide production was at a distinctly lower level on this night than it had been on the previous occasions. Since the total body-weight of the group varied only from 805 to 818 kg., the values for the total carbon-dioxide production per hour on the different nights may properly be compared. An examination of the original protocols for this night shows no reason whatever for questioning these results.

Among other explanations for this variation in metabolism found with Squad B we have considered the change in temperature environment. As the season progressed the weather became colder, but the temperatures inside of the group respiration chamber on these nights give no suggestion of change in temperature environment during the various experiments. Furthermore, no evidence has as yet been obtained to show an influence of small changes in temperature environment upon normally-clothed men. The widest variation in average temperatures throughout the entire season's experiments was but 4° C; the average temperature for January 5-6 was 19.3°. We thus have here a perceptible lowering in the basal metabolism for which we have no satisfactory explanation.1

1 For detailed values of a typical night, see table 5, page 119.

Heretofore it has been assumed that the basal metabolism is constant,2 particularly when average values are considered, but this group of 12 men, with only minor changes in personnel, showed on the average an actual decrease in the metabolism per kilogram of body-weight amounting to about 11 per cent and a lowering of the metabolism per square meter of body-surface amounting to 9 per cent, in the period from November 4 to January 6. We have no means of knowing to what extent this fall in metabolism is a normal procedure, nor have we evidence as to whether Squad A would have shown a similar lowering, but we must consider the possibilities of having to deal with a normal seasonal fall in basal metabolism of 11 per cent, on the basis of body-weight, or 9 per cent on the basis of body-surface. In any event the value for January 5-6 must be considered as the basal value for Squad B for subsequent comparison.

To sum up these findings regarding the basal metabolism of Squads A and B, we may say, first, that in the early part of October the two squads gave identical results for heat production in the group respiration chamber both per kilogram of body-weight and per square meter of body-surface. The average basal value for Squad A determined with the respiration apparatus in Springfield (see table 113) was 39.5 calories, which was but 3 per cent lower than that found with the group respiration chamber in Boston. We thus have three indices of uniformity in the basal values for these squads. Second, the fact that Squad B gave identical values with Squad A indicates that the number of men in the squad was sufficiently large for a general average result. Third, and most difficult to interpret, is the fact that, after 4 weeks of constancy, there was a slight seasonal variation in the metabolism, with a tendency towards a decrease in metabolism from November 4 until December 16, which amounted to 3 per cent, and a profound lowering of metabolism on January 6, which amounted to 11 per cent per kilogram of body-weight and 9 per cent per square meter of body-surface over the values obtained in the early autumn.

1The experiment on the night of January 5 immediately followed the Christmas vacation, the men coming direct from their homes to the Laboratory. The train travel, and the unusual social activities engaged in during the previous 18 days, with consequent loss of sleep, produced fatigue and probably a deeper sleep during the night experiment. The possible factors of fatigue and depth of sleep thus entered into experiment as in no other one in this series. Attention must be called to the fact that psychological observations on January 5 show, in general, aberrant values which may logically be ascribed to this factor of fatigue. A further complicating feature is the fact that the pulse-rate on the morning of January 6 was, as a matter of fact. 3 beats higher than the last observation on the morning of December 16. This would lead us to expect an in-creaee rather than a decrease in metabolism. 2Benedict and Carpenter. Carnegie Inst. Wash. Pub. No. 261, 1918.