Of special significance is the fact that in the three Boston calorimeter experiments the heat output as reported is heat eliminated and not heat produced; that is, the heat measurements have not been corrected for changes in the body-temperature. From an inspection of the data giving the increments in the water-drinking periods, it will be seen that in most cases the gaseous metabolism shows a positive increase corresponding to the increase in the heat output and in some instances it is roughly proportional to the increase in the heat. This, to a certain degree, confirms the validity of the heat measurements, even when not corrected for changes in the body-temperature.

In the final summaries, given in tables 67 and 68 (see pages 135 and 136), the total increments and the percentage increments are based solely upon the heat measurements. The values in table 67 show that in all but one of the calorimeter experiments, that with A. W. W. on March 27, 1907, there was a positive increase in the heat output as a result of taking water. In only two of the calorimeter experiments, those of January 13, 1911, with J. J. C, and January 12,1911, with T. M. C, can the increments be considered as really significant, these being 15 and 16 per cent, respectively. While the results of the experiment on January 13 may be open to the general criticism that J. J. C. was, as a rule, an unsatisfactory subject in many ways, yet so far as we can see there is nothing about the experiment which can be criticized, and we believe that the increment of 15 per cent represents a true increment. Reference to table 73 shows that in the experiment with J. J. C. on January 13, the total increase in the carbon-dioxide production was 8 per cent and in the oxygen consumption it was 13 per cent of the basal value. In the experiment with T. M. C. on January 12 (see table 72), the total increase in the carbon-dioxide production was 11 per cent and in the oxygen consumption 22 per cent. Evidently in both these cases there were actual increments in the metabolism due to the drinking of water.

The two experiments with the high increments were made with the chair calorimeter in Boston and in periods approximately 2 1/4 hours in length. The series of experiments with the universal respiration apparatus, in which the periods were approximately 15 minutes each but which covered a total period between the drinking of the water and the end of the last period of 50 minutes to approximately 2f hours, shows values considerably at variance with those obtained with the calorimeter. (See table 68.) In no case was the increment over 2 per cent and in two out of the six experiments there was, as a matter of fact, a slight decrease. The amount of water taken in these respiration experiments was much smaller than that taken in the calorimeter experiments, but this can not account entirely for the small increments, as the calorimeter experiment with A. W. W. on March 27, 1907, in which the largest amount of water was taken, namely, 3,935 grams, resulted in a decrease in the metabolism of 3 per cent.1

From the results of both series of experiments it is safe to conclude that when not over 500 grams of water are taken, as in the respiration experiments, the ingestion of water with a temperature of either 22° or 55° C. produces no significant increment of the basal metabolism. Since the two calorimeter experiments on January 12 and 13, 1911, apparently showed true increments in the metabolism due to water-drinking, there may be with more than 500 grams of cold water an increase as great as 16 per cent above the basal value. Although the subjective impressions of the two men showing the large increment were not recorded with sufficient detail to indicate any peculiar sensations, it is not impossible that we may have here a nervous phenomenon not unlike those mentioned by Loewy. (See page 140).

The pulse was counted in a considerable number of instances; measurements were likewise made of the blood pressure by means of the Erlanger sphygmomanometer. A general inspection of these results shows nothing significant in the changes in either pulse or blood pressure as a result of the ingestion of water.

1Mention should here be made of the experiments carried out by Ranke (see table 2, p. 17) in which the carbon-dioxide production for 24 hours of fasting without water was 663.5 grams, and on another day with the subject fasting with 2,100 c.c. water it was 662.9 grams.

In experiments of this kind one might maintain that it would be more logical to attempt a correlation of the metabolism with the total quantity of urine passed rather than with the amount of water taken, since it is conceivable that the mechanical work of the processes involved would be shown more clearly by the volume of urine excreted. As would be expected, the volume of urine increased considerably when large quantities of water were consumed, but the amounts of urine excreted were usually not abnormal, and we were unable to discover any definite correlation between the volume of urine and the total metabolism.