lOi:)] on Problems of Hydrogen and the Rare Gases 555 



manometer, given by any part of the flame, records the partial 

 pressure of the diffusing hydrogen for that region; this is easily 

 shown in the middle of the flame. In the same way a pure hydrogen 

 flame can be employed. The temperature of the flame is now not 

 so high, but can be increased by the introduction of a little oxygen, 

 using a simple form of quartz blowpipe. The oxygen introduced 

 passes through an annular tube round the hydrogen jet, whereby 

 the temperature of the tube is sufficiently raised by conduction 

 from the outside intense part of the flame, while leaving plenty of 

 available free hydrogen to diffuse from the central part. The 

 manometer rises then to almost any amount we please. An actual 

 rate of this diffusion has been measured by observing the growth 

 of pressure in an exhausted McLeod gauge of known volume to 

 which the platinum tube and attached U-tube was sealed. With 

 the platinum near a white heat the rate was found to be of the order 

 of 3 c.c. an hour per square centimetre through a mean thickness 

 of half a mm. When thin platinum foil was employed in a different 

 arrangement about twenty times this rate was observed at a rather 

 lower temperature. 



If instead of the flame impinging directly on the platinum tube 

 we introduce the quartz tube cover shown in Plate 6, allowing dry 

 air or oxygen to pass through the annular space, the manometer 

 remains steady, whereas, however, on passing a current of hydrogen 

 through the annular space the pressure instantly increases and 

 quickly reaches a limit. 



On replacing the platinum tube by a palladium one, similar effects 

 result, but at a greater rate. Now, however, the pressure disappears 

 when the flame is removed, as the palladium on cooling absorbs the 

 hydrogen ; but not so fast as when the tip of the flame is used for 

 heating to show the diffusion of the hydrogen out through the hot tube. 

 An iron tube similarly employed shows much smaller effects ; but 

 on the other hand this is the only tube of the three to give any 

 indications of gas diffusion when heated by a flame of carbonic oxide 

 raised in temperature by the addition of oxygen. 



An even more effective and ready way to demonstrate a succession 

 of these effects is by using an electric discharge tube to replace 

 the manometer ; the arrangement of the apparatus is seen in the 

 next figure (Plate 8). 



The platinum tube in this case is also provided when necessary 

 with an external quartz tube for leading different gases over the 

 heated platinum. For examining the electrical discharge and 

 spectrum of the diffusing gases, the discharge tube B is provided with 

 external electrodes, formed by silvering o cm. length near each end, 

 protected by thin copper foil bound on with wire. Exhaust is pro- 

 vided as usual by a charcoal bulb C with a stopcock F for putting 

 the exhaust off and on as required. B is protected at each end with 

 U -tubes D and E, as liquid air traps for volatile materials, either 



