260 



Professor Sir James Dewar 



[Jan. 23, 



treated by fan-shaped Bunsen removed traces of oxygen and hydrogen. 

 Following this was a phosphoric anhydride bulb, to dry the gas 

 roughly, to which Avas sealed a stopcock with "barometer" side tube 

 dipping into a small bottle of mercury to measure the pressure. The 

 remaining part of the apparatus was a bulbed U-tube C loosely 

 packed with asbestos wool, cooled in liquid air, to dry and purify the 

 gas and the condensation bulb D on side tube with a stopcock E 

 connecting the McLeod gauge and liquid hydrogen cooling bulb. 

 D was immersed in a vacuum vessel of liquid air arranged with 

 vacuum tight coned metal fitting for exhausting liquid air to produce 

 the requisite low temperature for condensing and fractionating off 

 the nitrogen into the gauge. Only the least volatile part was examined. 

 On treating air with various reducing agents to remove oxygen, 

 there is proof that measurable quantities of hydrogen appear which 

 must be derived from the simultaneous decomposition of water. 

 It will be noticed that the total volume of un condensable gases in 

 pure air as shown in the following table amounts to about one in 

 forty thousand of the volume originally taken, while the hydrogen 

 considered alone is of the order of one in two million. The uncon- 

 densable part in chemical nitrogen or oxygen shows how difficult it 

 is to get rid of traces of helium and neon, and old liquid air, which 

 ought to be very pure oxygen for this purpose, yet contains 15 times 

 more uncondensable at 20° Absolute than the chemical product. Again 

 this must come chiefly from adhering neon. 



