AND PLANTS 91 



between successive weighings of the same volume of the 

 same gas seemed altogether outside the possible range of 

 errors in weighing. Lord Rayleigh therefore considered 

 the possibility that the thing he had called nitrogen, and 

 spent all this care in weighing, might not always be the 

 same thing. Now the numbers in the first column were 

 obtained by weighing nitrogen liberated from certain of 

 its compounds : those in the second column, which are, 

 about one-half per cent, higher than those in the first, 

 were obtained by weighing the gas left after a portion of 

 atmospheric air had been deprived of its oxygen, its 

 traces of carbonic acid and other known impurities, and 

 its water. When this work was done, all chemists 

 believed that the residue, left after treating air so as to 

 deprive it of the constituents I have mentioned, was pure 

 nitrogen : and the reason why they believed this affords 

 us an instructive lesson in the care with which experi- 

 mental discrepancies should be treated, before we neglect 

 them. In the year 1785, Cavendish published an account 

 of his analysis of air. He found that in every hundred 

 volumes of air there were nearly twenty-one volumes of 

 oxygen, with a small quantity of carbonic acid, and he 

 tried to find out whether the rest was pure nitrogen. 

 His way of doing this was to confine the nitrogen in a 

 glass tube like the tube of a barometer, the open end of 

 the tube dipping into mercury, and to convert it into an 

 oxide which would readily dissolve in a solution of potash 

 floating on the mercury. If the whole volume of gas left 

 after the removal of the oxygen and carbonic acid had 

 been made up of pure nitrogen, and if Cavendish had 

 been able to oxidize it all, the gas in his tube should have 

 been completely absorbed. The process by which the 

 nitrogen was oxidized in this experiment took a very 

 long time, and at the end there was a small bubble of 



