228 COMPARISONS OF RESPIRATORY EXCHANGE. 



A second advantage of this apparatus is the accurate picture of the 

 respiration which may be obtained from the graphic record of the 

 movements of the spirometer bell, by which any irregularity or abnor- 

 mality is very accurately shown. For instance, these records indicate 

 when the subject is drowsy; this is of special importance in comparing 

 the results of respiration experiments, as the metabolism is greatly influ- 

 enced by sleep. Information regarding any such irregularities is neces- 

 sary in interpreting the respiratory quotients, as their value depends 

 upon the normality of the breathing. 



Still another advantage of the Benedict respiration apparatus is the 

 fact that it dispenses with the use of gas-analysis apparatus and with 

 the analysis of a large number of samples of expired air. Those who 

 are accustomed to making these analyses know that such work is not 

 only tedious but somewhat difficult, requiring special training to obtain 

 accurate results. 



Roily 1 has considered it necessary, in his experiments with a modified 

 Benedict respiration apparatus, to make an analysis of the air in the 

 apparatus at the beginning and end of the experimental period. He 

 states that it is not possible to get exact values for the oxygen consump- 

 tion without such gas analysis, as the composition of the air alters and 

 the volume of the air also alters, owing to changes in barometric pres- 

 sure and temperature. As was pointed out in the description 2 of the 

 original apparatus, theoretically corrections should be made for changes 

 in barometric pressure and temperature and in the composition of the 

 air of the apparatus, but practically it is not necessary. Grafe, 3 in his 

 discussion of the advantages and disadvantages of the Benedict appa- 

 ratus and of Holly's modification, points out that theoretically the 

 results with Holly's modification are more exact than with the original 

 apparatus of Benedict, i. e., the tension-equalizer unit, but that the 

 control experiments made with the latter are proof of its accuracy. 



The slight difference between the two methods may be shown by com- 

 paring the respiratory quotients given by Roily 4 with those obtained 

 by computing them from Rolly's protocols. Rolly's respiratory 

 quotients are 0.7991, 0.7432, and 0.773. Those calculated by the 

 Benedict method from Rolly's own figures for the carbon-dioxide 

 production, oxygen consumption, and the nitrogen admitted with the 

 oxygen are 0.7968, 0.7398, and 0.7845 respectively. In the last experi- 

 ment, there was a change of 1 mm. in the barometric pressure, 

 which is unusually large. Even with this large variation in barometric 

 pressure it is seen that the values for the respiratory quotient obtained 

 by the two methods agree within 0.01. Furthermore, an examination 



and Rosiewicz, Deutsch. Archiv f. klin. Med., 1911, 103, p. 58. 

 Benedict, Am. Journ. Physiol., 1909, 24, p. 345. 



3 Grafe, Abderhalden's Handbuch der biochemischen Methoden, Berlin and Vienna, 1913, 

 7, p. 472. 



4 Rolly and Rosiewicz, Deutsch. Archiv f. klin. Med., 1911, 103, p. 58, and Roily, ibid., p. 117. 



