METHODS OF MEASURING RESPIRATORY EXCHANGE 43 



it was afterwards delivered and measured through a meter, but his 

 bags were probably not tight against diffusion and his technique very 

 faulty. The principle, however, is excellent for certain types of ex- 

 periments, and it has recently been revived by Douglas who has worked 

 out a method which is specially adapted for the study of the respira- 

 tory exchange during open-air exercise in circumstances where all 

 other devices would fail, but which will also prove extremely useful in 

 a number of other cases, e.g. on bed-ridden patients (fig. 17). The 



FIG. 17. Douglas's respiration apparatus. From " Journal of Physiology" 

 (Cambridge University Press). 



subject breathes during an introductory period through the mouth- 

 piece and valves. When it is desired to make an experiment the three- 

 way tap is turned so as to connect with the bag and the expired air 

 collected over a certain period. With violent exercise a bag taking 

 60 litres will not hold the air expired during one minute, but it has 

 been shown (Krogh [1913]) that experiments of even much shorter 

 duration are sufficient to give perfectly reliable results. The air col- 

 lected in the bag is afterwards analysed and measured by connecting 

 with a gas meter of suitable size and pressing the air slowly out of the 

 bag. When a gas analysis is considered a thing to be avoided, the 



