president's address. 25 



There were, however, other scientists who noticed that the air 

 collected after its expulsion from the respiratory passages into a 

 jar might have a very different composition in the body. They 

 perceived that the air of the mouth and nostrils might be differ- 

 ently constituted than the air in the trachea, that the air in the 

 trachea might have another composition from the air in the small 

 bronchial tubes, and that the air from the deep recesses of the 

 lungs might differ from the remainder. They tested their 

 hypothesis by collecting the expired air in successive portions, 

 usually in two parts, but sometimes in three or more fractions. 

 They put into practice the noble maxim of John Hunter, "do not 

 think, but try." They found that each sample showed a different 

 composition. The quantity of oxygen diminished from the first 

 to the last fraction, while the amount of carbon dioxide increased. 

 They inferred that the air in the nostrils and windpipe contained 

 the most oxygen and the least carbonic acid gas, and that the air 

 from the deep recesses contained most carbon dioxide and least 

 oxygen. They were of opinion that the air in the upper respir- 

 atory passages closely resembled in composition the atmosphere, 

 while that from the deeper parts was most altered. They con- 

 cluded finally, that the expired, air had a continually changing 

 composition, that first breathed out being richest in oxygen, and 

 that last expelled most laden with carbonic acid gas. 



More than a century has elapsed since these conclusions were 

 drawn from the experiment of analysing fractions of the expired 

 air. Scores of scientists and hundreds of students repeated the 

 experiment with many modifications. For eighty-five years no 

 one perceived that the conclusions extended beyond what the 

 experiment proved. The inferences would have been correct if 

 the fractions had been collected in sufficient numbers to show 

 that the air was changing in composition in some regular manner. 

 As the volume of expired air is rarely less than 500 c.cm., and 

 often is more than 1000 c.cm., it is obvious that analysis of the 

 whole in three or four fractions does not furnish figures for this 

 purpose. The inference as to the continually changing composi- 

 tion may or may not be true. If it were true, the figures would 



