342 Professor Liveing, On Differences 



kathode rays, for it makes the glass fluoresce exactly as if it 

 were exposed to kathode rays. The spectrum of the thread of 

 light also is that of the kathode glow superposed on the con- 

 tinuous spectrum of the anode. The fluorescence of the tube 

 falls away when the pressure is reduced, and so does the brilliance 

 of the light emitted, especially in the wide part of the tube, while 

 the glow at the. kathode becomes plainly developed ; nevertheless 

 I think the thread of discharge continues to emit kathode rays 

 to some extent, and that the lines observed at the anode are 

 really due to such rays. It is in accord with this supposition 

 that bromine and iodine, which produce the strongest fluorescence, 

 are also those which shew the lines at the anode. 



It is difficult to manipulate the halogens because the mercury 

 pump cannot be used directly, so that I had to reduce the 

 pressure by connecting the sparking tube with a globe previously 

 exhausted, and this involved the use of stopcocks which had to 

 be lubricated ; and further the electrodes are rather rapidly 

 attacked by halogens. I did not see striations in any one of the 

 three substances, or well defined dark space. 



Every metallic vapour, so far as I have yet observed, shews 

 but one spectrum at both electrodes and throughout the tube. 



Mercury, when a globule is made the kathode, shews a 

 brilliant green negative glow at the surface of the metal, and 

 a less bright patch of light where the kathode rays strike the 

 surface of the metal when this is the anode. In both the 

 spectrum is the same, with the well known green line, pair of 

 yellow lines and indigo blue line, of mercury. When the tube 

 is warmed to increase the pressure of the mercury vapour the 

 glow which fills the tube shews the same spectrum, and there 

 is no other spectrum at the anode nor is any definite positive 

 column developed, or any striation. 



Other metallic vapours examined were those of sodium, 

 cadmium, and thallium. Plain tubes, without capillary, about 

 15 cm. long, enclosing fragments of the several metals, were 

 pumped out, and, while still connected with the pump, were 

 heated by gas jets, in a small furnace, built up of tiles from a 

 combustion furnace, so that the whole tube was tolerably equably 

 heated. Small openings left between the tiles allowed the 

 spectrum to be observed. After the carbonic acid gas emitted 

 by the glass had been pumped out as completely as possible, 

 the tube and its contents were gradually heated, and though 

 the pressure of the vapour could not, be measured, it was varied 

 by raising the temperature more or less. 



The sodium vapour at very low pressure shewed its presence 

 on passing the discharge by a long positive column with a well 

 marked dark space about the kathode. The kathode glow was 



