PROFESSOR W. C. ROBERTS-AUSTEN OjST THE DIFFUSION OF METALS. 409 
the ordinary diffusion of liquids, in which the diffusing substance is usually in the 
presence of a large excess of the solvent which is supposed to exert but little chemical 
action on it. This condition has been fully maintained in the experiments on the 
diffusion of liquid metals which are described in the first part of the present paper. 
It must also be remembered that Van’t Hoff* has made it highly probable that the 
osmotic pressure of substances existing in a solid solution is analogous to that in 
liquid solutions and obeys the same laws, and it is also probable that the behaviour 
of a solid mixture, like that of a liquid mixture, would be greatly simplified if the 
solid solution were very dilute. 
Nernst expresses the hope that it may be possible to measure by indirect methods 
the osmotic pressure of substances existing in solid solutions. I trust that the 
following experiments will sustain this hope by affording measurements of the 
results of osmotic pressure in masses of solid metals at the ordinary atmospheric 
pressure, and at a temperature at which it has hitherto been scarcely possible to 
detect diffusion in them. 
The following experiments constitute, so far as I am aware, the first attempt to 
actually measure the diffusivity of one solid metal in another. It must be borne in 
mind that the union of two clean surfaces of metal, and even the interpenetration of 
two metals to a slight depth below the surfaces does not necessarily depend on diffusion 
alone, as the metals become united in a great measure by viscous flow. The nature of 
welding demands investigation, but the union of metals by welding is effected most 
energetically when the metals are in the colloidal condition, in which true diffusion is 
least marked. It may be observed that discs of gold and lead, pressed together at 
the ordinary temperature for three months, were found to have welded together more 
perfectly than two similar discs kept in contact at 100° for six weeks, although 
at least ten times more metal had interdiffused in the latter case than in the 
former. 
Diffusion of Gold in Solid Lead. 
The attempt was first made to ascertain whether diffusion of gold in solid lead 
could be measured at a temperature of 250°, that is 75° below the melting point of 
lead. AVith this object in view, thin plates of gold were fused on to the end of 
cylindrical rods of lead, 14 millims. in diameter and 7 centims. long. This could 
readily be effected by the point of a blow-pipe flame, and, when the cylinder of lead was 
kept cool by immersion in water to within a few millimetres of its end, the gold 
rapidly alloyed with the metal, but, as many analyses showed, did not penetrate 
the cylinder of lead more than a millimetre. Such cylinders were maintained for 
thirty-one days in a little iron chamber lined with asbestos, the temperature of which 
only varied by a degree or two from 250° C. The cylinders were then measured and 
* ‘ Zeitschr. Phys. Chem.,’ vol. 5, 1890, p. 322. 
3 G 
MDCCCXCVI.—A. 
