METHODS OF MEASURING RESPIRATORY EXCHANGE 27 



power (3) to produce the oscillations of the CO 2 absorber is not, of 

 course, essential, but has been found to be very satisfactory because 

 of its absolute uniformity. The CO 2 absorbing efficiency of the ar- 

 rangement described is very limited, and that is why the apparatus 

 can be used only for small animals such as frogs or small reptiles and 

 possibly mice. 



In all larger closed-space respiration apparatus the ventilating 

 and CO. 2 absorbing appliances are separate, and there is moreover a 

 separate, usually more or less automatic, device for adding oxygen. 

 As a compensating vessel cannot be used to advantage on large ap- 

 paratus the internal pressure and also the composition of the air will 

 vary, and it becomes necessary to make analyses of the air in the 

 chamber at the beginning and end of each experimental period. We 

 have therefore in a complete closed-space apparatus of a large type the 

 following distinct parts : 



1 . The animal chamber. 



2. The air-circulating pump. 



3. The carbon-dioxide absorbing system. 



4. The device for adding oxygen. 



5. The air-sampling arrangement. 



6. The arrangements for measuring temperature, humidity, and 

 pressure of the enclosed air. 



A considerable number of closed-space apparatus have been de- 

 scribed since the prototype of them all was published by Regnault and 

 Reiset in 1849, notably by Hoppe-Seyler, Zuntz, Oppenheimer, At- 

 water and Benedict, Benedict and his collaborators and Tangl, but 

 none of them is satisfactory with regard to all details. It will be most 

 convenient therefore to describe the modifications of each of the essen- 

 tial parts separately, in order to bring out the principles, which should 

 guide the construction, instead of describing the instruments one by 

 one. A complete instrument of recent type is shown diagrammati- 

 cally in fig. 6. 



The essential point with regard to the animal chamber of a large 

 closed-space apparatus is absolute tightness. There is abundant evi- 

 dence in the literature of the trouble which it has cost to attain this end, 

 though most of the difficulties encountered are never mentioned in the 

 publications. Zuntz and Oppenheimer [1908] have adopted the plan to 

 immerse the whole of their animal chamber (for dogs) into a water-bath. 

 This no doubt facilitates the temperature control and the detection of 

 leaks, but the whole becomes very cumbrous and the construction must 



