PROFESSOR W. C. ROBERTS-AUSTEN ON THE DIFFUSION OF METALS. 385 
comparatively rapid.”* As regards mercury and its fluid amalgams, the history is 
more extended, for in 1713 HombergI may be said to have at least foreshadowed 
the diffusion of metals, both solid and liquid, in his paper “On Substances which 
Penetrate and which pass through Metals without their being Melted.” He inci¬ 
dentally showed, by experiment, the extreme rapidity with which mercury will 
penetrate a bar of tin. 
In November, 1883, Guthrie | published a remarkable paper “ On Certain Mole¬ 
cular Constants,” in which the diffusion of zinc, lead, tin, sodium, and potassium in 
mercury was studied, and he stated that these metals, which, of course, are much 
lighter than mercury, “appear after a month’s interval in appreciable quantity at a 
depth of a foot beneath the surface, when the temperature is about 16° or 17°.” He 
concludes his paper by offering “ a general curve of amalgamation,” which he thought 
would represent the rate at which the metals examined by him alloy with mercury, 
and this curve may also, he says, “ represent the relative rates of elementary atomic 
and molecular diffusion generally.” 
Guthrie held that as the mercury he employed was a good conductor of heat, 
there was not much fear of the disturbing influence of convection currents. The 
existence of such currents, nevertheless, gave me much anxiety in the earlier experi¬ 
ments with molten metals which were begun in the year 1881, and will now be described. 
From the outset of this research both molten lead and bismuth were chosen as 
suitable fluids in which the diffusion of other metals could be studied. Advantage 
was also taken of the fact that at temperatures well above the melting point, neither 
of these metals unite with iron. The precious metals, also, when alloyed with lead or 
bismuth, do not show any tendency to unite with iron unless it is very clean and 
bright. Tubes of wrought iron, therefore, proved to be most useful in conducting 
the inquiry. In the first instance single tubes filled with lead were arranged verti¬ 
cally in a bath of lead which was kept well above its melting point. Weighed 
quantities of heated, but still solid, gold or platinum were then rapidly lowered 
through the lead in little covered receptacles of iron to the bottom of the tube, and 
when by the aid of a rigid steel wire the removal of the covers was gently effected, 
the gold was exposed to the lead ; it became rapidly dissolved, and diffusion began. 
The tubes filled with molten lead, in which diffusion took place, were about 
200 millims. long, and many such tubes were arranged in a single bath, which was 
carefully kept hotter at the top than at the bottom so as to avoid as much as 
possible the carrying of the precious metal upwards by any streams of lead which 
might rise as convection currents from the bottom of the tube. The main result of 
* Roberts-Austen, ‘British Association Report,’ 1883, p. 402. 
t Homberg, ‘ Mem. de l’Acad. Royale des Sciences,’ 1713 (vol. published 1739), p. 306. 
t Guthrie, ‘ Phil. Mag.,’ vol. 16, 1883, p. 321. W. S. Humphreys has recently made experiments on 
the diffusion of zinc, lead, tin, bismuth, silver and copper in mercury. ‘ Trans. Chem. Soc.,’ vol. 69, 
1896, p. 243. 
MDCCOXCVI. —A. 3 D 
