CRITICAL DISCUSSION OF RESPIRATION APPARATUS. 263 



Durig 1 has pointed out that differences of small amounts in the 

 oxygen consumption and the carbon-dioxide elimination per minute 

 may result in large variations in the respiratory quotient if the differ- 

 ences are in opposite directions. There is, then, a double effect upon 

 the respiratory quotient, and in that case the quotients are very variable. 

 For example, in gas analysis, with a difference of 0.1 per cent, differ- 

 ences may be obtained of 0.04 to 0.05 in the respiratory quotient if the 

 errors in the carbon-dioxide determination are in the opposite direction 

 to those in the oxygen determination. Such variations, however, would 

 be very large for gas analyses in which differences of not more than 0.02 

 to 0.04 per cent should be expected. 



With many methods of gas analysis the errors tend to compensate 

 one another, particularly if the gas analysis is made by means of a 

 Haldane apparatus, when the low carbon-dioxide absorption will be 

 compensated by a greater absorption in the potassium pyrogallate. 

 The result in this case would be that the carbon-dioxide increase would 

 be too small, while the oxygen percentage would be too high; the oxygen 

 loss would then be too small, but unless the error due to incomplete 

 absorption of carbon-dioxide by the potassium hydroxide was large, 

 the ratio between the carbon dioxide increase and the oxygen deficit 

 would not be markedly different from the actual ratio obtained by a 

 correct analysis. With the Regnault-Reiset or closed-circuit method, 

 on the contrary, the two determinations are made independently and 

 there may be an error in one but no compensating error in the other. 

 Consequently, wider variations maybe found in the respiratory quotient 

 by this method than with the open-circuit method. The determi- 

 nation of the respiratory quotient by the analysis of expired air is, 

 therefore, the more logical method. 



Respiratory quotients below 0.7 or above 1.00, which are obtained 

 with individuals without food and in a resting condition, must be looked 

 upon with considerable suspicion. Thus far the accumulation of 

 reliable evidence has not been sufficient to show that respiratory 

 quotients much below 0.7 may be obtained, even with abnormal or 

 pathological conditions. On the other hand, respiratory quotients 

 over 1.00 can not be expected to occur unless there is some trans- 

 formation of sugar into fat, but this is not likely to occur with a man 

 who has not had food for 12 hours or more. Abnormal quotients such 

 as these should be controlled by repeated observations in successive 

 experiments in order to make certain of their accuracy. It must be 

 pointed out that a very sharp distinction should be made between the 

 probable accuracy of respiratory quotients obtained with an apparatus 

 and the probable accuracy of the values obtained for the carbon- 

 dioxide elimination and oxygen absorption. Accurate respiratory 



^urig, Denkschriften der mathematisch-naturwissenschaftlichen Klasse der kaiserlichen Akad- 

 emie der Wissenschaften, Vienna, 1909, 86, p. 118. 



