PROFESSOR W. C. ROBERTS-AUSTEN ON THE DIFFUSION OF METALS. 405 
still solid iron into steel is effected by the passage of solid carbon into the interior of 
the mass of iron, and the explanations which have from time to time been given of 
the process form a voluminous literature. Le Play considered cementation, which 
is really a slow creeping action of one solid into another, to be “ an unexplained and 
mysterious operation,” and he attributed the transmission of the carbon to the centre 
of the iron solely to the action of gaseous carbonic oxide. Gay Lussac* confessed 
that a study of the process shook his faith “ in the belief generally attributed to the 
ancient chemists that corpora non agunt nisi soluta,” for it is certain, he adds, “ that 
all bodies, solid, liquid, or aeriform, act upon each other, but, of the three states of 
bodies, the solid state is the least favourable to the exercise of chemical affinity.” 
In 1881, M. A. CoLSONt communicated a paper to the Academie des Sciences, in 
which he showed that when iron is heated in carbon there is a mutual interpenetration 
of carbon and iron at so low a temperature as 250°. The 
interpenetration of solids, as distinguished from the diffu¬ 
sion of two metals in each other, has received attention 
from many experimenters, of whose work brief mention 
will only be given, as the subject of this part of the paper 
is the diffusion of solid metals. Colson pointed out that 
pure silver diffuses as chloride in dry chloride of sodium, 
and he states that calcium passes into platinum when the 
latter is heated in lime, and that silica diffuses through 
carbon and yields its silicon to platinum. The permea¬ 
tion of strongly heated porcelain by carbon has been 
demonstrated by Marsden, Violle, and other experi¬ 
menters. Spring,} in 1885, showed that solid barium 
sulphate and sodium carbonate react on each other until 
an equilibrium is established. 
Any lingering doubt as to whether gas need necessarily intervene in the cementa¬ 
tion of iron was, I may point out, removed by an experiment of my own,§ in 1889, 
which showed that pure iron may be carburized by diamond in vacuo, at a tempera¬ 
ture far below the melting point of iron and under conditions which absolutely 
preclude the presence or influence of occluded gas. I am indebted to my friend, 
M. Osmond, for a photograph (from which fig. 6 is prepared) of a section, magnified 
100 diameters, through a piece of electro-iron, which had been heated to 1500 J and 
carburized from the upper end by contact with the diamond form of carbon, and this 
section clearly shows the gradual penetration of the iron by carbon. The white grains 
at the base are unchanged iron. The beautiful work of Osmond on the transforma- 
* ‘ Ann. de Chim. et de PLys.,’ vol. 17, 1846, p. 221. 
t ‘ Comptes Rendus,’ vol. 93, 1881, p. 1074; vol. 94, 1882, p. 26. 
X ‘ Bull, de l’Acad. Roy. de Belgique,’ vol. 10, 1885, p. 204. 
§ ‘ Nature,’ vol. 41, 1889, p. 14. 
