34 RESPIRATORY EXCHANGE OF ANIMALS AND MAN 



(4), and a third vessel (5) with sulphuric acid on pumice stone to absorb 

 the moisture given off from the soda lime. The gain in weight during 

 an experimental period of the two last bottles will correspond to the 

 carbon dioxide given off from the animal and the gain in weight of the 

 whole system : animal chamber, first H 2 O absorber, CO 2 absorber, and 

 second H 2 O absorber corresponds to the oxygen which has been re- 

 tained, since the air enters and 'leaves this system dry and CO 2 free. 

 A very important point in Haldane's apparatus is the avoidance of re- 

 sistance to the passage of air through the absorbers. When the tubing 

 employed is of sufficient bore the internal pressure is practically equal 

 to the atmospheric and the danger of leakage is minimized. The 

 rubber connections should be made as short as possible as rubber is 

 not impervious to water vapour or CO 2 . 



y8. The prototype of the larger apparatus with measurement 

 of the air current is the famous Pettenkofer respiration apparatus 

 [1862], in which atmospheric air taken from outside was allowed to 

 enter the animal chamber directly while a constant current of air was 

 drawn out by a large pump and measured by a gas meter. Two small 

 pumps were at the same time continuously taking samples of the 

 atmospheric and of the outgoing air. These samples were measured in 

 gas meters and pressed through baryta water contained in Pettenkofer 

 tubes, in which the carbon dioxide was absorbed and afterwards titrated. 

 When the quantity of CO 2 found was divided by the volume of the 

 sample and multiplied by the total volume of outgoing air, the carbon 

 dioxide elimination could be deduced. 



Voit [1875] introduced the important improvement that a gas 

 meter actuated by a small motor was used as a pump for the main 

 air current (fig. 10). An extremely accurate measurement is hereby 

 secured as Voit's calibrations show, and in all modern air-current ap- 

 paratus gas meters have superseded the older and complicated types 

 of pumps. It is perhaps a little strange that they have not so far 

 been used in closed-circuit apparatus. 



The Pettenkofer- Voit apparatus could only be used for CO 2 deter- 

 minations. 1 The same is the case with Tigerstedt-Sonden's apparatus 

 [1895] which is ventilated in the same way as Voit's, but in which only 

 a small sample of the air (60 to 100 c.c.) is analysed in the extremely 

 accurate Petterson-Sonden gas-analysis apparatus for CO 2 . It is 



1 Though the determination of oxygen is theoretically possible by weighing the subject 

 before and after the experiment and determining the excretions and the output of water 

 vapour it has never become accurate enough for practical use. (See Benedict, 1910.) 



