ON DIFFUSION IN SOLIDS. 3851 
The Electrolysis of Glass and Porcelain. 
Very interesting results have been obtained in the course of studies 
of the electrolysis of glass. The electrical conductivity of heated glass, 
first observed by H. Cavendish,'® was measured and shown to be 
electrolytic in character by H. Buff,?? who was followed by many 
other workers. It was even found possible *1 to construct a copper- 
zine cell with an intervening plate of glass as the electrolyte, a measur- 
able E.M.F. being obtained at temperatures of from 50° to 120°. If two 
concentric glass bulbs are used, of which the inner contains a solution 
of copper sulphate and the outer one of zinc sulphate, with platinum 
wires as electrodes, a Daniell cell of high resistance is obtained.22 A 
more detailed study revealed some remarkable facts.2* A stout plate 
or bulb of soda glass is used as a partition, separating mercury on the 
one side from sodium amalgam on the other, and a current is passed 
from the amalgam through the glass to the mercury, a temperature of 
from 200° to 350° being employed. After about thirty hours, an 
appreciable quantity of sodium is found to have been removed from the 
amalgam and transferred to the mercury, the glass meanwhile remain- 
ing clear. When lithium amalgam is substituted for sodium amalgam, 
the current passes, and sodium appears in the mercury cathode, fol- 
lowed after a time by lithium. At the same time an opaque layer of 
silica appears at the anode surface and gradually spreads through the 
glass. A glass originally containing 2°4 per cent. of potassium and 
13°4 per cent. of sodium was found after an experiment to contain the 
same proportion of potassium, but only 5°3 per cent. of sodium and 
4°3 per cent, of lithium. The transfer of sodium follows Ohm’s law. 
*- The atomic volume of lithium is less than that of sodium. When a 
metal of larger atomic volume than sodium, such as potassium, is 
used in the form of an amalgam, it is found to be impossible to cause 
it to enter the glass. Gold and copper enter the glass, but do not 
penetrate completely, and after a time coloured deposits of these 
metals appear below the surface. In a similar manner silver pene- 
trates into lead glass, producing a characteristic structure which is 
inade visible by the silver stain. In all these cases the metal entering 
the glass has a smaller atomic volume than that which is expelled 
from it, a fact which led Roberts-Austen to regard the process as one 
in which the atoms of the metal entering pass along tracks or galleries 
in the glass left by the atoms of the metal expelled. The transfer may 
undoubtedly be described as due to diffusion. Warburg’s experiments 
have been recently repeated with certain additional precautions.24 
Diffusion of Gases through Metals. 
The permeability of certain heated metals to gases, especially of 
platinum and iron to hydrogen, was observed by H. St. Claire Deville 
19 * Electrical Researches ’ (London, 1879), p. 181. 
20 Lieb. Annalen, 1854, 90, 257. 
|! W. Thomson, Proc. Roy. Soc., 1875, 28, 463. 
| 2 H. Helmholtz, Faraday Lecture, Trans. Chem. Soc., 1881, 39, i., 277. 
*8 E. Warburg, Ann. Physik., 1884 [iii.], 21, 622 ; W. C. Roberts-Austen, ‘Third 
Report to Alloys Research Com.,’ Proc. Inst. Mech. Eng., 1895, 238. 
74 M, Le Blane and F, Kerschbaum, Zettsch. Physikal. Chem., 1910, 72, 468. 
