380 Prof. H. E. Armstrong on Loiv-Temperature Research 



arc-lamp is focused upon the black surface of the inner vanes. Even 

 when the charcoal condenser is cooled in solid hydrogen to a tempera- 

 ture of about 15" absolute, the vanes of the instrument still rotate 

 when exposed to the beam. If, however, hydrogen be the residuary 

 gas, on cooling the charcoal condenser in liquid hydrogen the radio- 

 meter can no longer be excited into action. Tested in the ordinary 

 way, by means of an induction coil, the bulb charged with helium, 

 in which the radiometer vanes still rotate, appears to be " vacuous," 

 as no discharge will pass through it. 



The crucial importance of these observations lies in the fact that 

 the method of purification adopted is so complete that all gases other 

 than helium are removed from the sphere of action. 



It is thus proved that a relatively considerable amount of gas 

 may be present and yet no electric discharge will take place — in 

 other words, that the passage of an electric current of high potential 

 through a gaseous atmosphere is dependent, apparently, not on the 

 mere presence of one fcirticidar hind but on that of appropriate 

 systems of molecules, since it can only be supposed that the ordinary 

 conductivity of " helium " in a tube prepared without using the very 

 special precautions taken by Sir James Dewar is conditioned by its 

 association with a minute proportion of some other substance — 

 perhaps impure vapour of water. 



A similar observation, which has always seemed to me to afford 

 proof of the same kind, has been brought under notice on more than 

 one occasion in the Royal Institution lectures — namely, the observa- 

 tion that when one of Sir William Crookes's tubes, containing an 

 earth which phosphoresces on passing an electric discharge through 

 the tube, is cooled locally by meaus of a wad soaked in liquid air, the 

 discharge will no longer pass and the earth cannot be excited into 

 phosphorescence. The ease with which the passage of the discharge 

 is prevented is such as to leave no doubt that some quite volatile 

 substance is present and becomes condensed on cooling the tube 

 locally. This explanation will be the more easily accepted by those 

 who have witnessed the striking demonstration often given by Sir 

 James Dewar of the efficiency of liquid air as a coohng agent, which 

 consists in placing a few cubic centimetres of the refrigerant in a 

 depression in a flask charged with gaseous bromine at a low pressure. 

 As the air boils away, the gaseous molecules within the flask can be 

 visualised as rushing towards the cold surface and, as it were, falling 

 asleep in the solid state as soon as they reach it, for all colour soon 

 disanpears from the interior of the flask, the bromine forming a dark 

 solid patch on its side where it is cooled. 



No observations which have been made of late years appear to 

 me to be of more consequence than these two, in connection with 

 the general problem as to the nature of the conditions determinative 

 of the passage of an electric discharge through a gaseous atmosphere 



