250 COMPARISONS OF RESPIRATORY EXCHANGE. 



able, and that when the respiration is quiet and free there is a liability 

 toward back-leak. The more forcible part of the expiration passes 

 through the expiration valve, but the end of the expiration, which is 

 slower, may go back through the inspiration valve; consequently, if 

 the portion lost has not the same ratio of carbon dioxide to oxygen as 

 the portion collected, true respiratory quotients may not be obtained. 

 A test 1 of one of the valves showed a recovery of only 78 per cent of the 

 air drawn through it. These valves may be safeguarded by attaching a 

 long tube to them, so that the air which passes out through the expira- 

 tion valve may be drawn in again with the next inhalation. 



The Douglas method has recently been used by Carter 2 on tubercular 

 patients in preference to the Zuntz-Geppert method. Henderson and 

 Prince 3 have also employed it in some observations on "oxygen pulse 

 and systolic discharge" and state that it is much simpler and easier to 

 use, more accurate, and affords more nearly normal conditions as to the 

 air breathed by the subject than any other device with which they are 

 familiar. 



In general, it is apparently more difficult to obtain reliable results 

 with this method than with the other open-circuit methods. The 

 bags used must be tested for diffusion and always handled in the same 

 manner when emptying them before and after the experiment. Care 

 must be taken not to have the periods long enough to cause the subject 

 to exhale against pressure. The valves used should be of a reliable 

 type or carefully safeguarded by a long tube on the ingoing valve. The 

 apparatus is of advantage because of its portability. 



VALVES. 



In all methods for determining the respiratory exchange in which the 

 inspired and expired air are separated, it is necessary to use some kind 

 of valve for the separation. In this investigation several types of 

 valves have been employed and their individual merits have been dis- 

 cussed in connection with the apparatus with which they were used. 

 Those most easily and cheaply constructed are the Mueller valves, 

 which can be made of materials found in almost any laboratory. 

 The principal requirements are that they should have a wide opening 

 through which the air passes; that the water seal should be so thin that 

 it offers no resistance and yet at the same time sufficiently deep to 

 prevent air from returning through the ingoing valve; and that they 

 should be suspended or set in such a manner that they are perfectly 

 level, so as to give an effective closure with a minimum amount of water. 



The Zuntz valves, which are actually of the type devised by Speck, 

 are effective in operation; the chief objections to them are their size 



See p. 252. 



*Carter, Journ. Expt. Med., 1914, 20, p. 87. 



'Henderson and Prince, Am. Journ. Physiol., 1914, 35, p. 109. 



