ON DIFFUSION IN SOLIDS. 357 
As an illustration of the extent of the diffusion, it may be mentioned 
that at 251° the gold had risen against gravity to a height of 7 cm. in 
less than a month. Further experiments, continued for four years at 
18°, showed that at that temperature the gold had risen 7°5 mm., the 
quantity diffusing in a thousand years being calculated as being nearly 
equal to that which would diffuse in molten lead in one day.® 
The conditions of these experiments may be examined a little more 
fully. The equilibrium of alloys of lead and gold has been investi- 
gated.°* Gold forms two compounds with lead, Au,Pb and AuPb,. 
Only the second of these could play any part in the diffusion, owing 
to the large excess of lead present. There is a eutectic point at 215°, 
so that in the experiment conducted at 251° the transport may have 
taken place by means of capillary ascent of the liquid eutectic, but the 
other experiments recorded are at temperatures well below the eutectic 
point. Another experiment recorded in the Bakerian lecture showed 
that the rate of diffusion of gold in solid silver at 800° is of the same 
order as that of gold in lead at 200°. In this case a compound is not 
formed, and diffusion can only take place by the formation of a solid 
solution. 
Further experiments on the same lines as those of Spring have been 
carried out by G. Masing,*? who prepared conglomerates of various 
pairs of metals by compressing mixtures of the filings in a steel cylinder 
under 1,000 to 5,000 atmospheres. The extent to which union had 
‘taken place after heating to different temperatures and for different 
times was estimated by taking heating curves and comparing them 
with the curves obtained under similar conditions from alloys prepared 
by fusion. In certain cases the results were confirmed by microscopi- 
cal examination and by determinations of the electrical conductivity. 
The results show that pressure alone is without influence on diffusion 
or combination, and serves merely to bring the fragments of the two 
metals into intimate contact. Intimate mixtures of very finely ground 
zine and cadmium, or copper and silver, melt at the eutectic point 
of the corresponding series, but the mass is at once distinguishable 
under the microscope from one obtained by fusion, by the absence of 
the eutectic structure. Only traces of a compound are formed when 
the two metals combine without an appreciable formation of solid 
solutions, as in the pairs Mg-Pb, Mg-Sn, Mg-Zn, and Mg-Sb. (These 
mixtures were not examined microscopically.) Diffusion is perceptible 
in the case of cadmium and magnesium, which form a continuous 
series of solid solutions containing a compound MgCd, and of 
aluminium and magnesium, in which series the compound Al,Mg, 
forms solid solutions with an excess of aluminium. 
The alloys most fully studied by Masing were those of bismuth and 
thallium. Heating the mixture of filings, previously rendered compact 
by pressure, to 120° for 53 hours brings about the formation of a 
coloured layer surrounding the particles of thallium. This layer con- 
6 Proc. Roy. Soc., 1900, 67, 101. 
6 R. Vogel, Zeitsch. Anorg. Chem., 1905, 45, 11. 
Si Zeitsch. Anorg. Chem., 1909, 62, 265; see also G. Tammann, Zeitsch. El<ktro- 
chem., 1909, 15, 447. 
