CRITICAL DISCUSSION OF RESPIRATION APPARATUS. 239 



not find that either one is better than the other. The manipulation 

 of the gas-analysis apparatus is also difficult and, according to our 

 experience, requires considerable training for its successful use. 



To keep the apparatus in good condition does not require a great 

 deal of attention, the parts needing most care being the valves and the 

 gas-analysis apparatus. The valves should be so cared for that they 

 will have the least resistance when used and be without leak. During 

 experiments they should always be moist, as otherwise they do not 

 functionate properly. The rubber connections in the gas-analysis 

 apparatus are also liable to leak and to deteriorate. Setting up the 

 apparatus requires skill in order to avoid breaking the different parts of 

 the capillary tubing; the gas-analysis apparatus is especially large and 

 cumbersome, with many parts to be connected. The caustic-potash 

 solution in the carbon-dioxide pipette occasionally needs renewing. 

 The phosphorus also needs attention occasionally, particularly if the 

 apparatus has been used in a warm room, as in this case there is a 

 tendency for the phosphorus to fuse together and to cause errors due 

 to the occlusion of small bubbles of gas when the air is drawn from the 

 phosphorus pipette. When in use, the apparatus should be frequently 

 controlled by analyses of outdoor air, which is uniform in composition 

 all over the world and in all kinds of weather. 1 



The apparatus used by Zuntz and his co-workers does not give a 

 graphic record of the respiration, either in volume or in rate, and in 

 order to have an accurate knowledge of the respiration-rate it is neces- 

 sary to have some accessory apparatus. 



The results obtained with the Zuntz-Geppert apparatus are reliable, 

 provided the greatest care is taken in carrying out the experiments. 

 When the breathing is normal this is particularly true as to the respira- 

 tory quotient, which represents the relation between the carbon dioxide 

 and the oxygen. Even if there is a leak in some part of the measuring 

 apparatus, the relationship expressed by the respiratory quotient would 

 not be affected, but the total quantity of expired air would be less than 

 the actual amount and consequently the total metabolism as measured 

 would be too low. This would not be true if a certain portion of the 

 expiration were lost, as, for example, the last portion or the beginning, 

 as the ratio between the carbon ctioxide and the oxygen is not the same 

 in all portions of the expired air. In the manipulation of the appa- 

 ratus there is practically no noise which would disturb the subject, and 

 this quietness is conducive to good results. 



When samples are being taken and stored continuously, as they may 

 be over mercury, a series of experiments may be carried out with but 

 very short intermissions, thus making the measurements practically 

 without interruption; this is also true with the Benedict apparatus. 

 The gas analyses necessitated by this method are, however, tedious and 



Benedict, Carnegie Inst. Wash. Pub. 166. 1912. 



