2 DR. R. ANGUS SMITH ON THE 



would sufficiently serve the purpose^ and prevent any one 

 from asserting that I was endeavouring to make out a 

 case against the air underground. This number had 

 already been taken by high authorities as indicating the 

 amount of oxygen. 



For my own satisfaction^ however^ it was needful to 

 add greatly to the data from which to judge. Finding, 

 too, that the numbers given by chemists varied exceed- 

 ingly, I felt it necessary to inquire into the cause of the 

 discrepancy. 



I believe I have shown that the air of various places gives, 

 from local causes, a difference in the amount of oxygen, 

 and that this difference, although apparently very small, 

 does in reality indicate important changes of quahty, or, 

 in other words, that the amount of oxygen is in reality an 

 important guide in considering the purity of an atmosphere, 

 although we must deal, not with percentages, but with 

 parts in ten thousand. 



Amount of Oxygen in Pure Air. 



When Priestley discovered oxygen, he examined the air 

 of various places, and found the amounts of this gas to 

 differ to the extent of 6 per cent. He used as a test nitric 

 oxide, which combines with the oxygen, after which both 

 gases, viz. the combined oxygen and nitric oxide, are 

 absorbed by water. At one time he obtained one-fifth of 

 oxygen, which is nearly the exact amount ; but he did not 

 seize the idea forcibly. 



Scheele found from 20 to 30 per cent., and others found 

 still greater variations, until Cavendish showed, by 500 

 examinations of atmospheric air, that it had a nearly con- 

 stant composition, and arrived at a mean of oxygen equal 

 to 20*833 per cent.* 



* See Dr. George Wilson's calculations in his Life of Cavendish. 



