754 Professor Sir James Deivar [June 7, 



bulb in liquid hydrogen for only half a minute, the rarefaction became 

 so great that, when the arc light was thrown on, the vanes remained 

 perfectly still. From a comparison of these two experiments we may 

 again derive an indication of the boiling-point of helium. For, in 

 the case of the hydrogen radiometer, a fall of 75 per cent, in the tem- 

 perature of the charcoal bulb, from the boiling-point of air to the 

 boiling-point of hydrogen, reduced the vanes to rest ; similarly, we 

 may infer that a fall of like amount from the boiling-point of hydro- 

 gen would reduce the vanes of the helium radiometer to rest, a result 

 which would make the boiling-point of helium about 5° or 6° absolute, 

 as before. 



A variation of the hydrogen radiometer experiment may be made 

 as follows. Into the bulb of the radiometer a branch tube was sealed 

 containing a little metallic sodium. On immersing the charcoal bulb in 

 liquid hydrogen, and throwing on the electric beam, the vanes remained 

 at rest as before ; but on gently heating the sodium, a minute quantity 

 of hydrogen was liberated which was sufficient to re-start the radiometer 

 for a short period of time. In a few minutes, however, the absorp- 

 tion of hydrogen by the charcoal became so efficient that the radiometer 

 stopped. 



Ordinaey Air Radiometer. 



A radiometer, with attached charcoal bulb, was repeatedly washed 

 out with the oxygen and nitrogen vapour coming from old liquid air, 

 and sealed off at a pressure of a fraction of about a millimetre of mer- 

 cury. On immersion of the charcoal tube in liquid air, the motion 

 of the vanes did not cease, but, on immersion in liquid hydrogen, a 

 minute or two sufficed to bring the motion completely to a stop. 

 The explanation of the experiment is that old liquid air is not a 

 mixture of oxygen with a small proportion of nitrogen and argon, 

 but it always contains traces of neon and helium. 



Oxygen and Nitrogen Radiometer. 



The effect of the residual gases contained in air can be shown as 

 follows : The bulb of a Crookes radiometer was thoroughly washed 

 out with a mixture of pure chemical oxygen, care being taken that 

 no hydrogen, neon or helium was present. This bulb was not 

 provided with a side charcoal tube, but had a long quill tube 

 attached, a few inches of which could be cooled in liquid hydrogen. 

 The radiometer, thus charged with pure oxygen, had a portion of the 

 quill tube cooled in liquid hydrogen. The motion entirely ceased 

 through the condensation and solidification of the gas at 20° Abs. 

 When air was used instead of the chemically prepared oxygen, the 

 same radiometer was not stopped owing to the presence of helium 

 and neon in the atmosphere. 



