ON DIFFUSION IN SOLIDS, 353 
would also pass through other metals, including lead, tin, zinc, and 
gold. The experiment was devised *® of causing mercury to flow from 
one vessel to another at a lower level, using as a syphon a rod of cast 
lead, bent so as to form two arms of unequal length. Tin may be 
used in the same way.*’ Copper and brass were subsequently added 
to the metals exhibiting this behaviour.*® 
It is unlikely that true diffusion in the solid state plays any 
unportant part in the process of penetration of mercury in such cases. 
Mercury dissolves lead or tin, forming a liquid amalgam at the ordinary 
temperature, and this amalgam probably makes its way through the 
intercrystalline spaces. It was observed by Henry that the transfer of 
mercury takes place much more readily through cast lead than through 
hammered or rolled metal, the intercrystalline spaces of the latter 
having been closed by the mechanical treatment. The metal also 
becomes exceedingly brittle during the process, owing to the loosening 
of the crystalline texture by the action of the liquid amalgam. 
Solid Diffusion in Metals, 
It is in metals and metallic alloys that the most complete and 
satisfactory experiments in solid diffusion have been made. The 
existence of diffusion in solid metals has been fully established, and 
although doubts may be raised as to the crystalline character of the 
material through which diffusion takes place, the range of temperature 
covered by the experiments is so great that it is difficult to conceive of 
the persistence of amorphous metal throughout. 
Leaving the cementation of iron for separate consideration, the 
earliest scientific record of the production of an alloy by union in the 
solid state appears to be the observation of M. Faraday and Stodart *° 
that iron and platinum could be alloyed by welding at a temperature 
below that at which fusion could take place. A-solid solution is formed 
in this case. 
Several early metallurgical processes have been quoted by W. C. 
Roberts-Austen *° as depending on solid diffusion. One of these is the 
ancient process of extraction of silver from gold by cementation, 
described by Pliny and others,*+ in which the alloy is packed in a 
mixture of salts and heated for a long time at a temperature well below 
the melting-point. The silver is gradually extracted, leaving a residue 
of solid gold. Another process is that of ‘kernel roasting,’ in which 
cupriferous iron pyrites are slowly roasted in air, the final product 
being a central ‘ kernel’ of cuprous sulphide, enclosed in a shell of 
ferric oxide, whilst the silver originally present wanders outwards, and 
appears in the free state as a lace-like coating on the surface.*? 
86 J. Henry, Proc. Amer. Phil. Soc., 1841, 4, 56, 84. 
87 K. N. Horsford, Amer. Jour. Sci., 1852 [ii.], 18, 305. 
88 J. Nicklés, Compt. rend., 1853, 86, 154. 
89 Quart. Jour. Sci., 1820, 9, 319; ‘ Exp. Researches in Chemistry,’ 1859, p. 57 
40 Phil. Trans., 1896, 1874, 383. 
4 A full account is given by J. Percy, ‘ Metallurgy of Silver and Gold’ (London, 
1880), i., 379. 
42 Ww. C. Roberts-Austen, loc cit, 
1912, 
AA 
