266 ME. J. C. MAXWELL GARNETT 



same colour, and crystals of each salt acquire iinder cathode rays a beautiful violet 

 tint.* Experiment has also shown that exposure to the emanation from radium gives 

 to gold glass a ruby colour, to silver glass a yellow colour, and to potash glass a brown 

 colour. 



Now we have seent that a molecularly subdivided metal possesses the same colour 

 by transmitted light whatever be the nature of the surrounding transparent medium, 

 supposed non-dispersive and isotropic. This colour may be called the vapour- colour 

 of the metal. It has further appeared that although the transmitted colour of a metal 

 subdivided into small spheres, many to a wave-length of light, does depend on the 

 refractive index v of the medium in which the small spheres are " embedded," yet 

 this colour approaches to the vapour-colour as v approximates to unity. As is shown 

 by the dotted curve in fig. 1, the vapour-colour of gold must be red.| The colour of 

 glass containing molecularly distributed gold is thus red,;}; although when the gold is 

 collected into spheres the glass is pink. Similarly, reference to the relative values of 

 /3'/X in Table II. shows that the vapour-colour of silver is yellow. Glass coloured 

 by small spheres of silver is also yellow. Again, Professor R. W. WOOD showed to 

 the British Association^ in Cambridge that the vapour-colour of sodium is violet, this 

 colour being due to the absorption at the D lines. This violet colour is also produced 

 at the cathode in the electrolysis of sodium chloride, || the molecules of sodium formed 

 at the cathode being distributed throughout the water in its neighbourhood and 

 giving rise to the vapour-colour. 'I Analogy with the cases of gold and of silver 

 indicates that small spheres of sodium would produce in glass a colour not greatly 

 different from the vapour-colour produced by the molecularly subdivided metal. 



Thus the colours developed in gold, silver, or soda glass by the radiation from the 

 emanation from radium are approximately the same as the colours which would be 

 given to the glass by the presence of the reduced metal, either molecularly divided or 

 in small spheres (nascent crystals). 



It is therefore very probable that the metal in the glass is reduced by the action of 

 the radiation. This view finds considerable support in the discovery of VILLARU,** 

 that cathode rays exert a reducing action, as well as from the fact, already cited, tf 

 that ELSTER and GEITKL found the salts of the alkali metals, which had been coloured 

 by exposure to cathode rays, to exhibit photo-electric effects as if they contained 

 traces of the free metal. 



* GOLDSTEIN, ' WIED. Ann.,' liv., p. 371, 1898. 



t Fide ante, p. 243. 



I Or yellow ; see the second footnote, p. 243. 



August, 1904. 



|| Cf. J. J. THOMSON, 'Conduction of Electricity through Gases,' pp. 495, 496. 



IT BUNSEN found that common salt, after heating to about 900 C., exhibited a violet colour, due 

 apparently to the reduced metal, although BUNSEN suggested a, subchloride, 

 ** 'Journal de Phys.,' 3' Series, VIII., p. 140, 1899. 

 tt See 'Phil. Trans.,' A, 1904, p. 400. 



