26 president's address. 



vary in a similar manner to those found, but they are also com- 

 patible with other conclusions. The samples analysed were too 

 large and were collected in too haphazard a manner to give 

 figures of the accuracy needed to prove the inference by the use 

 of a mathematical study of their distribution. As a matter of 

 fact, the inference is wrong. The expired air has no continually 

 changing composition. Had the fractions of the expired air been 

 collected with accuracy and analysed, the proper inference could 

 have been made, but very great exactness is requisite if this 

 method of study is adopted. These difficulties disappear when 

 small samples are taken for analysis at considerable intervals 

 during the expiration. 



In 1905, Haldane and Priestley published the account of 

 their experiments on expired air. They made use of the dis- 

 covery of Haldane that there was little difference in the percent- 

 age of carbon dioxide in the last air expired from a deep or 

 shallow rapid expiration. Haldane had accustomed himself to 

 breathe evenly and regularly for some minutes while seated in a 

 chair. He took samples at the end of a forcible expiration. On 

 analysis, these samples were found of "identical composition'" 

 whatever was the depth of the expiration. Whether 600 c.cm., 

 1200 c.cm., or 1500 c.cm. were driven from the air passages, the 

 last part of the breath showed the same percentage of carbon 

 dioxide. Haldane recognised that the results showed the com- 

 position of the expired air leaving the air-passages was not con- 

 tinuously changing, but that a considerable part of the expired 

 air had an unvarying composition. To quote an example,* 



Volume of air expired. Percentage of CO„ in sample. 



1. 492 c.cm. 5-71 



2. 1050 c.cm. 5-72 



To obtain samples, Haldane introduced the method of breath- 

 ing into a long rubber tube fitted with a side-piece of narrow 

 bore close to the mouth. After expiration, the end of the tube 

 is closed with the tongue, and the sample for analysis withdrawn 

 from the tube through the side-piece. Haldane and Priestley 



* Haldane and Priestley, Journ, Physiol., xxxii., p. 228, 1905. 



