BED CALORIMETER AND TENSION-EQUALIZER UNIT. 103 



Errors in the determination of the residual oxygen. The principal 

 error in experiments with the bed calorimeter is incidental to the meas- 

 urement of the oxygen consumption. The oxygen consumption of 

 the subject is determined from the amount admitted to the chamber 

 during the experimental period and from the changes in the oxygen 

 content of the air in the chamber. The chief difficulty met with is in 

 the measurement of the residual oxygen, especially as this measurement 

 is affected by the combined errors in the determination of the residual 

 water- vapor and the residual carbon dioxide. 



The measurement of the residual oxygen is also affected by errors 

 in the determination of the average temperature of the air in the appa- 

 ratus. The average temperature changes of the apparatus are obtained 

 from readings of the changes in resistance of a set of copper-wire 

 resistance thermometers. These are placed at five different points 

 adjacent to the wall of the chamber and consequently give only the 

 changes in the temperature of the layer of air next to the walls. When 

 the subject is inside the chamber there is a thermal gradient of approx- 

 imately 37 C. (the temperature of the man's body) to about 12 C. 

 for the air next to the cooling pipes. If the man moves he may pro- 

 duce a slight heating of the air about him, but this heating effect will not 

 be shown by a variation in the thermometer readings. It will, how- 

 ever, affect the volume of the air in the chamber, causing it to expand 

 slightly, with a consequent rise in the rubber diaphragm of the tension- 

 equalizer or the spirometer bell. If this increase in volume occurs 

 just prior to the end of a period, and this record is used in calculating 

 the volume to C. and 760 mm., the average temperature reading will 

 be too low, so that the results supposedly correct will indicate an 

 oxygen content of the chamber which is actually too large. This 

 difficulty in the accurate measurement of the average temperature of 

 the air in the chamber has always been recognized in observations made 

 with the respiration calorimeter and has been thoroughly discussed in 

 another publication. 1 It has, however, become even more apparent 

 as experience with the apparatus has accumulated. 



To prevent such errors in the measurement of the residual oxygen, 

 the subjects have been cautioned to remain perfectly quiet during the 

 last 15 minutes of an experimental period. Various interpretations of 

 this request have been made by the different subjects and undoubtedly 

 some have made major movements which resulted in erroneous deter- 

 minations of the oxygen consumption. The importance of this body- 

 movement control at the end of an experimental period became so 

 marked that in a series of night experiments with a subject who fasted 

 31 days a procedure was adopted to register the change in volume dur- 

 ing the last 5 minutes. 2 A recording device was attached to the spirom- 



^enedict and Carpenter, Carnegie Inst. Wash. Pub. 123, 1910, p. 89. 

 "Benedict, Carnegie Inst. Wash. Pub. 203, 1915, p. 310. 



