BED CALORIMETER AND TENSION-EQUALIZER UNIT. 101 



made. Before discussing this apparent discrepancy, the general ques- 

 tion of conditions under which the experiments were carried out and 

 the various sources of error may be considered. 



The chamber of the bed calorimeter is not large enough to permit 

 much movement by the subject, but is sufficiently large for a man of 

 medium size to turn from one side to the other. The subjects were 

 requested to remain quiet throughout the experimental period, and 

 although they usually made an attempt to do this, it was very difficult 

 for them to remain in one position for two or three hours. With the 

 opportunity for freedom of movement afforded by the bed calorimeter 

 there was more or less movement by the subject, and as he could be 

 seen only through the glass window at the end of the calorimeter and 

 a small glass porthole on the side, it was not possible to observe accu- 

 rately the degree of muscular repose. Some of the subjects found 

 themselves so comfortable in the calorimeter that it was impossible 

 for them to keep awake. This was especially true of J. J. C. and V. G. 

 The other subjects were awake for the greater part of the time. 



During the experiments with the tension-equalizer unit, the subject 

 lay flat upon his back, with nosepieces inserted in the nose, or a mouth- 

 piece in the mouth. The connections with the ventilating apparatus 

 were so short that movements of the head were not possible, and any 

 other major movements, such as movements of the arms or legs, could 

 easily be recorded by the observer. Accordingly, the subject lay either 

 in a comfortable relaxed position, which would be conducive to a low 

 metabolism, or in a condition of tenseness, which would tend to increase 

 the metabolism. 



It will be seen, therefore, that it was somewhat difficult to compare 

 the metabolism as measured by these two apparatus, for to determine 

 the accuracy of measurement it is necessary to assume that the metab- 

 olism to be measured is the same. With the same degree of muscular 

 repose in both cases, other conditions being alike, the results of the 

 measurement would agree, provided the two apparatus were equally 

 accurate. If, on the other hand, the two forms of apparatus gave 

 theoretically the same results, any differences in the metabolism would 

 be due to differences in the degree of muscular repose. 



In addition to this possible difference in the metabolism due to 

 difference in the muscular repose, there are various sources of error 

 with both the bed calorimeter and the tension-equalizer unit, which 

 may affect the measurement of the metabolism. Those for the bed 

 calorimeter will first be considered. 



Sources of Error in Experiments with the Bed Calorimeter. 



Errors in measuring the carbori-dioxide elimination. — The errors in the 

 measurement of the carbon-dioxide elimination have to do mainly with 

 the weighing of the absorption vessels. The carbon-dioxide absorbers 



