422 Dr. J. H. Vincent. 



influence of the electrical stress, likewise flew off, and were deposited 

 in the metallic state near the pole. 



An experiment was tried to discover if the ions of metal, deposited 

 in this manner on a metal plate connected with an idle pole, in the 

 full stream of + ions and - electrons, showed any special + or - 

 electrification. In all cases the electrification was positive. This 

 lends support to Mr. Strutt's view that + ions as well as - electrons 

 will fly off from a radio-active body. Even these results, however, 

 must not be taken as conclusive, for in a paper published in 1891,* I 

 showed that at a high vacuum nearly the whole of the interior of a 

 tube through which an induction spark was passing was electrified 

 positively, negative electrification being detected only in the immediate 

 neighbourhood of the negative pole. 



During the course of my experiments a curious circumstance was 

 observed, which deserves record as it may elucidate some of these 

 obscure phenomena. While the volatilisation of the silver pole is 

 rapidly proceeding, the metal glows as if red-hot. This " red heat" is 

 superficial only. The metal instantly assumes, or loses, the appearance 

 of red heat the moment the current is turned on or off, showing that 

 the high temperature does not penetrate below the surface. The 

 volatilisation of the positive ions is confined to the surface, and the 

 surface glow is connected with that action. If instead of silver, a good 

 conductor of heat, I take diamond, a bad conductor, the surface 

 layers are changed sufficiently to convert them into a form of graphite,, 

 which from its great resistance to oxidising agents, cannot be formed 

 at a lower temperature than 3600 C. 



' The Density and Coefficient of Cubical Expansion of Ice." By 

 J. H. VINCENT, D.Sc., B.A., St. John's College, Cambridge. 

 ComYnunicated by Professor J. J. THOMSON, F.R.S. Eeceived 

 January 22, Eead February 6, 1902. 



(Abstract.) 



After an account of the methods employed by previous experi- 

 menters in the subject, reference is made to the views of Nichols, 

 according to which two distinct kinds of ice have been subjected to 

 experiments. The density of artificial ice is about O916 gramme 

 per cubic centimetre, while that of natural ice is more than one part in 

 a thousand greater. 



* "Electricity in Transitu: from Plenum to Yacuum," 'Journ. lust. Elect. 

 Engin.,' TO!. 20, p. 10. 



