BED RESPIRATION CALORIMETER. 19 



around the chest of the subject to record his respiration and activity, 

 but more recently the latter has been recorded by means of a spring-bed 

 arrangement. 1 The subject lies upon an air-mattress placed upon a 

 metal framework which can be readily slid into the chamber. When 

 everything is in readiness, the subject is put into the chamber, the 

 front opening is closed by the glass panel and sealed with wax; the 

 heat-adjusting arrangements are put in order and the preliminary 

 period is begun. During this preliminary period the assistant in 

 charge of the calorimetric measurements brings the apparatus into 

 equilibrium, so that there is no radiation through the walls and the 

 absorption of heat by the water-current flowing through the apparatus 

 is constant. 2 When the equilibrium has been obtained, a determin- 

 ation is made of the residual content of the water-vapor and the carbon 

 dioxide. After the determination of these two gases, the experiment 

 is begun, the air-current being deflected from one side of the absorp- 

 tion system to the other and continued for a fixed period. At the end 

 of the period the temperature is obtained by readings from a series of 

 electrical resistance thermometers inside the respiration chamber, dis- 

 tributed at various points. The barometer is also read at the exact 

 end of each period and the height of the spirometer taken in order 

 to find the apparent volume of air inside the chamber. The oxygen 

 cylinder and sulphuric -acid and soda-lime containers are then weighed. 

 The experiment may be stopped at this point or another period begun. 

 The usual length of periods is 45 minutes or an hour, and an experiment 

 usually continues at least 1| hours. 



The accuracy of the measurement of the carbon-dioxide elimination 

 and oxygen consumption has been carefully controlled theoretically by 

 burning alcohol. 3 The alcohol was introduced into the chamber through 

 a copper tube, at the end of which a small enlargement was made in 

 which was placed an asbestos wick. By means of this arrangement, 

 small amounts of alcohol were burned in successive periods, these 

 periods being each an hour or more in length. The alcohol was usually 

 'burned at the rate of about 14 gm. per hour, and as the amount burned 

 could be determined to 0.01 gm., the error in weighing the alcohol was 

 about 0.1 per cent. Considerable difficulty was experienced in the 

 actual measurements of the alcohol on account of the changes in level 

 of the alcohol in the lamp. This was finally overcome by means of a 

 small manometer outside of the calorimeter; this manometer was arbi- 

 trarily filled to the same height at the end of each period. The results 

 of two typical alcohol check experiments are given in tables 4 and 5. 



Benedict, Carnegie Inst. Wash. Pub. 203, 1915, p. 311. 



"As this publication does not deal with the calorimetric features of the apparatus, these are 

 not described here. A full description is given by Benedict and Carpenter in Carnegie Inst. 

 Wash. Pub. 123, 1910, pp. 10-53. 



'Benedict, Riche, and Emmes, Am. Journ. Physiol., 1910, 26, p. 1. 



