The considerable amount of experimental attention given to the technique for determining the gaseous metabolism led, as would be expected, to an unusual extension of the gaseous metabolism measurements. When the experiment was first proposed, it was intended to obtain occasional records of the respiratory exchange for the individual men. As the program developed, it appeared desirable to get as frequent records for each man as possible. Therefore all of the facilities of the Laboratory were drawn upon to meet the emergency of having to determine the gaseous metabolism of 12 men as nearly as possible once each day. It was believed that in some of the observations, methods could be used in which a high degree of accuracy might be sacrificed to expedition, but it was also desirable to make determinations of the respiratory exchange by means of thoroughly tested standard methods, if only rarely. Finally, it was planned to make observations of the squad as a whole in a newly perfected respiration chamber in the Nutrition Laboratory in Boston, which had been thoroughly tested in a series of observations with a group of approximately 25 individuals several months before.

Respiratory-Valve Method

For the most reliable measurements of the basal gaseous metabolism we took advantage of the newest technique developed in the Laboratory. This included the use of a special form of mask, Thiry-Tissot valves, a spirometer, and the Haldane gas-analysis apparatus. By means of two spirometers it was possible to obtain the gaseous metabolism in at least two periods each on two men every morning, thus securing observations on each member of the squad at least once a week. In the last few weeks of the research, the routine was so altered that measurements were obtained with this procedure on three men each morning. While the various sections of the apparatus have been extensively described by Dr. T. M. Carpenter of the Nutrition Laboratory staff,1 the exact combination of appliances used (due, as a matter of fact, to the technical skill of Dr. Carpenter) has not been described in extenso and a few words are here desirable.

Believing that for the most accurate measurement of the respiratory exchange and especially the respiratory quotient, untrammeled respiration is desirable, Dr. Carpenter has used a face mask provided with inlet and outlet valves which communicate, in turn, with a pipe leading to the outdoor air and another pipe leading to a Tissot or similar type of spirometer. By means of a small canvas cap and suitable straps, this mask can be so firmly attached to the face as to make it air-tight. With reasonable care in application, the mask will be sufficiently comfortable for the subject to wear it for a half hour or more; occasionally the subjects fall asleep - a proof of absence of discomfort.

As used in the daily observations with Squad A, the mask was first attached and the valves examined to find if they were working smoothly. The expired air was then allowed to pass into the room for several moments, but at a given time connection was made with the spirometer and the air expired by the subject was collected in the spirometer for a period of approximately 10 minutes. To provide exact information as to the degree of repose of the subject, as well as graphic records of the respiration, two pneumographs were placed one about the chest and the other about the thighs; these were connected with suitable tambours which recorded upon the smoked surface of a kymograph drum. By this method the activity was instantly recorded, also each respiration. By means of the respiration record, the character as well as the rapidity of the respiration could be seen. Finally, the importance of knowing whether the subject was asleep or awake during the observations led us to obtain by means of a signal magnet writing upon a kymograph drum records of the responses to a stimulus given the subject every minute automatically by a clock. The clock was in a closed electric circuit, which caused a buzzer to vibrate for a fraction of a second. If the subject was sufficiently awake to respond to this signal, he pressed a small push button held in his hand; this movement was recorded on the drum.

1 Carpenter, Carnegie Inst. Wash. Pub. No. 216, 1915, p. 61.

At the end of the 10-minute period of collection of expired air, the time was accurately noted and the tube leading to the spirometer disconnected. Connection was then made with the second spirometer and the expired air collected during a second 10-minute period. Throughout both periods the pulse-rate of the subject was frequently counted by another observer.

While the second spirometer was being used, the height of the first spirometer, the temperature, and the barometer were recorded. A sample of the air in the spirometer was also collected in small Hal-dane samplers over mercury, properly numbered, and placed in suitable racks for subsequent analysis. Usually three samples were taken from each spirometer. The total volume of air expired in each of the two 10-minute periods could be computed from the height of the two spirometer bells. These volumes could be corrected to 0° C. and 760 mm. by suitable calculations. Using these volumes and the number of respirations per minute as obtained from the kymograph records, the total volume of air expired per minute and the total volume per respiration could be accurately determined.

The gas analyses were all made with the small Haldane portable apparatus. Two of these were kept in perfect condition for the purpose and analyses frequently made on both for control. Furthermore, the gas analyses each day were preceded by an analysis of outdoor air; until satisfactory values were found for both carbon dioxide and oxygen in these air samples, the analyses of the gas samples were not made. The gas-analysis technique was carried out by Miss Mary F. Hendry, of the Nutrition Laboratory, whose long training, particularly under the skilful guidance of Miss Alice Johnson, has made her one of the best gas analysts that the Laboratory has ever had on its staff.

With two and in the latter part of the research three subjects, each observed for two 10-minute periods, the usual duplicate analyses for each spirometer would require a total of 8 analyses with 2 subjects and 12 analyses with 3 subjects. It was early found, if the sample from one spirometer was analyzed and the complete oxygen intake and carbon-dioxide production computed from this one analysis, and a sample from the second spirometer analyzed and the complete oxygen intake and carbon-dioxide output computed from the results of the analysis, that in the large majority of cases, one could rely upon a single analysis of the air sample from each spirometer and assume that the accuracy of the whole physiological and chemical process was established by the agreement of the two samples. This procedure was not followed until several days' verification of the method had proved its uniformity. It is of interest to note that subsequently the cooperation of the subjects was so perfect, the technical skill of the operators so satisfactory, that duplicate analyses were rarely called for. Thus, the daily gas analyses (aside from the daily outdoor air control) could usually be completed after 4 analyses with 2 subjects and 6 analyses with 3 subjects.

The special care given to these measurements of the gaseous metabolism was fully justified by the unusual significance of the results. This procedure supplied us with the most exact picture of the transition in basal metabolism of the individual subjects from day to day as the reduced diet continued. It is the clearest picture of variations in the individual basal metabolism and likewise of the absolute values for the metabolism that we have throughout the whole study.

Portable Respiration Apparatus

The complexity, nicety of manipulation, and care required for metabolism measurements by means of a mask, valves, spirometer, and gas analysis, precluded the use of this method for more than two or at best three of the 12 subjects each day. On the other hand, it was agreed that changes of considerable importance might take place inside of a week which should be noted, if possible. Hence we employed a new type of respiration apparatus, which had been primarily developed for clinical work, but not extensively used in the Laboratory prior to this research.1 By means of seven of these apparatus, specially constructed for the purpose, the measurements were extended each morning so that two observations were made with the respiratory-valve type of apparatus and seven with the new portable type. We thus obtained basal measurements for 9 of the 12 men daily.