In order to study the intensity of oxidative processes in the organism, determinations of the oxygen consumption are, as a rule, performed; one is accustomed to consider the values obtained for the consumption of oxygen in a resting subject, early in the morning, before the ingestion of any food, as an index of the intensity of the oxidative processes in this individual. For, thanks to the investigations of Zuntz (38) and Magnus Levy (39), we know that the consumption of oxygen under these conditions is practically constant in the same individual. In our studies, therefore, the early morning values for the oxygen consumption were determined in a number of subjects - this was a fore-period during which the use of water was unrestricted; the same determinations were then performed in a main-period during which liquids were restricted, and again in an afterperiod during which liquids ad libitum were again permitted.

In order to exercise some control over these subjects, who were confined in different rooms, and in order to be certain that they did not exceed the quantity of liquid allowed them, the quantity, the specific gravity and the appearance of the urine, also the specific gravity of the blood and of the blood serum, were determined. Without entering into a detailed dis cussion and interpretation of the significance of the latter figures, I have simply entered them into the tables, because I believe that they present certain points of interest that are independent of the questions under discussion.

During the main- and middle-period of the experiment the amount of liquids was restricted. Sometimes the withdrawal of liquids was reinforced by the administration of electric-light baths that produced abundant loss of water by profuse sweating.

The water contained in the solids of the diet was not included in the calculation, because otherwise the analytic part of the investigation would have been unnecessarily increased out of all proportion to any possible value it might have possessed. Instead, great care was exercised to maintain uniformity of diet, both quantitatively and qualitatively throughout the whole course of the experiment; particular directions were also given to have the food prepared in the same way, so that the water content of the various dishes remained approximately constant.

In several cases, e.g., in cases 1, 2, 15, the patients did not take the whole quantity of weighed solid food prescribed that, in preliminary studies, had been found to correspond to their individual tastes and requirements. Consequently, the total quantity of water was still more limited in these cases than was shown by the determination of the actual amount of liquids consumed. It is interesting to note in this connection that thirsting led to an involuntary and distinct re striction of the total amount of food ingested, an observation that is of the greatest importance in interpreting the reduction of obesity that follows the restriction of liquids.

In nearly all of the cases the loss of weight, the quantity, the specific gravity and the appearance of the urine, as well as the concentration of the blood, showed that a condition of thirst really existed. The withdrawal of liquids was in no case carried to excess, if for no other reason than the one that such extreme reductions in the fluid-intake are not employed in practice in reduction cures for obesity.

The determination of the interchange of gases was performed with the Zuntz-Geppert apparatus, an apparatus that has already aided us in discovering so many truths in regard to the physiology and the pathology of the gas-metabolism of the organism. The short duration of each experiment (20-45 minutes), it is true, rendered it necessary to exercise considerable care in interpreting the results obtained; the great facility, however, with which a large number of single determinations that controlled and supplemented each other could be performed in the same individual fully compensated for this defect.

All the gas-figures were reduced to O°, 760 mm. barometric pressure, and dryness.

The determinations of the specific gravity of the blood were performed with the pyknometer and were always made at 11 a. m.

In order not to extend and complicate the tables too much, the quantity of the urine and its specific gravity are only given on those days on which determinations of the respiratory interchange of gases were made.

Particular care was taken to accustom the patients to the technique of the experiments before the proper investigation, i.e., the "fore-period" was begun. We did not begin the "fore-period" until the patients had learned to breathe quietly and regularly, without performing any voluntary muscle-movements, as long as they were in the apparatus.

In the text of this dissertation we are giving only the average values obtained in the different periods. The numerical details will all be found in the exhaustive tables appended to this article.

All the patients gave their explicit consent to these experiments.