352 REPORTS ON THE STATE OF SCIENCE.—1912. 
and L. Troost,*° and further studied by Thomas Graham in his cele- 
brated memoir on diffusion.** The question has been frequently 
investigated, the metals most commonly employed being platinum and 
palladium. The permeability of heated silver to oxygen was also deter- 
mined by L. Troost.?7 In all cases the rate of diffusion increases 
rapidly with the temperature. The metal exercises a selective action 
and behaves as a semi-permeable membrane, palladium and platinum 
allowing hydrogen to pass, but not most other gases, whilst silver trans- 
mits oxygen, but not hydrogen or nitrogen. The diffusion is not pro- 
portional to the pressure of the gas, but the figures obtained may be 
explained on the assumption that the molecules of hydrogen diffusing 
are dissociated into single atoms, and that the diffusion is proportional 
to the pressure of the dissociated gas.?8 This view has, however, been 
disputed,* adsorption being invoked to explain the anomalies. 
The diffusion of nascent hydrogen through metals has also been 
studied. Thus, when iron is made the cathode in an electrolytic cell, 
hydvogen rapidly diffuses through the metal, and pressures of many 
atmospheres are obtained.*° Platinum, palladium, and nickel show the 
same effect in a much smaller degree. The diffusion increases very 
rapidly with the temperature, the increase being of the order of T°.*! 
There is a close relation between the diffusivity of a gas through a 
metal and the absorptive power of the metal for that gas. The semi- 
permeable character is extremely well marked, gases other than hydro- 
gen not diffusing through platinum at a white heat, even in sufficient 
quantity to be detected by means of the spectroscope.*? The diffusion 
of gases through colloidal membranes, such as caoutchouc, is of a 
simpler character, as is shown by the simple proportionality of the 
diffusion to the pressure and to the temperature.** Its discussion does 
not fall within the range of this Report. 
The Passage of Liquids through Meials. 
The only liquid which has been observed to pass through metals 
without destroying their structure is mercury. It was observed in 
1713 ** that a complex mixture, consisting essentially of an amalgam 
of silver and mercury, was able to pass through a plate of silver without 
rendering it brittle. Ata much later date it was shown ** that mercury 
2% Compt. rend., 1863, 57, 965; H. St. Claire Deville, zbid., 1864, 59, 102. 
26 Phil. Trans., 1866, 156, 399. 27 Compt. rend., 1884, 98, 1427. 
28 W. Ramsay, Phil. Mag., 1894 [v.], 88, 206; A. Winkelmann, Ann. Physik., 
1901 [iv.], 6,104; O. W. Richardson, J. Nicol, and T. Parnell, Phil. Mag., 1904 [vi.], 
8,1; O. W. Richardson, ibid., 7, 266; Proc. Camb. Phil. Soc., 1905, 18, 27. 
2 G. N. St. Schmidt, Ann. Physik., 1904 [iv.], 18, 747. See reply by A. Winkel- 
mann, ibid., 1905 [iv.], 16, 773. 
30 M. Bellati and S. Lussana, Atti R. Ist. Ven., 1890 [vii.], 1, 1173; 1891 [vii.], 
2, 987. 
31 A, Winkelmann, Ann. Physik., 1905 [iv.], 17, 590. 
82 W. W. Randall, Amer. Chem. Jour., 1897, 19, 682. 
33 See, for instance, Graham, loc. cit., and 8. von Wroblewski, Ann. Physik., 1876, 
158, 539. 
’4 Homberg, Mém. Acad. Roy. Sct., 1713, 306 (not actually published until 1739). 
%> J, F. Daniell, Jour. Roy. Inst., 1831 [ii.J, 1, 1. 
