ON DIFFUSION IN SOLIDS. 355 
without rise of temperature. Alloys were also formed by compressing 
their components.** Of these, the most interesting case is the forma- 
tion of brass by compressing a mixture of copper and zinc filings. The 
union is very incomplete, and it is necessary to reduce the conglomerate 
to filings and to repeat the compression five or six times. The same 
memoir describes the formation of fusible alloys by pressure, but the 
study of the constitution of these alloys was unknown at the time, and 
in‘the light of more recent knowledge it is obvious that the production 
of a readily fusible mass is no proof of chemical union, as the eutectic 
is merely a conglomerate. It was in fact soon shown *‘ that fusible 
alloys could be produced by merely mixing filings of their component 
metals. The temperature at.which such a mixture melts is above the 
true eutectic temperature by an amount which depends on the coarse- 
ness of the powders used.** Other experiments,°® in which the forma- 
tion of fusible alloys by pressure was assumed from the compact 
appearance and low melting-point of the product, are inconclusive for 
the same reason. 
These objections do not apply to the later experiments of Spring, 
in which true interpenetration was shown to occur.®7 Two cylinders 
of the same metal, accurately faced and freed from grease, were found 
to adhere firmly at ordinary temperatures.and to become fully united 
on heating. Union occurred with all the metals examined with the 
exception of anumony, and was usually. so intimate that when the 
two cylinders were subsequently separated by force the fracture passed 
through the metal on one side or the other of the plane of junction. 
Quite similar results were obtained when the two cylinders were of 
different metals. Thus, when zine and copper were placed in contact 
and heated to 400° for six to eight hours, the two cylinders became 
united by a layer of brass 0°8 mm. in thickness, whilst copper and 
antimony similarly formed a layer of the purple alloy. Vapour 
undoubtedly played a part in some of the reactions recorded, as it was 
found that a thin laver of brass was formed at 400° even when the 
zine and copper were separated by a distance of 0°8 mm., and cadmium 
also volatilised to a small extent. The intervention of vapour had 
been previously suggested in explanation of some of Spring’s earlier 
welding experiments,*® and similar conclusions have been reached by 
a study of reactions between solid salts in closed vessels.*® 
When mixtures of copper and silver filings are compressed into 
cylinders by the application of a pressure of 8,000 atmospheres, the 
pressure. being maintained for a month at the ordinary temperature, 
apparently homogeneous masses are obtained which, however, prove 
on microscopical examination to consist merely of unaltered filings 
united by adhesion, no interpenetration having take place.®° It is 
53 Ber., 1882, 15, 595. 
‘4 W. Hallock, Zeitsch. Physikal. Chem., 1888, 2, 378. 
°° C. Benedicks and R. Arpi, Metallurgie, 1907, 4, 416. 
56 C. Drewitz, Verh. Ver. Befird. Gewerbefl., 1902, 81, 325. 
57 Zeitsch. Physikal. Chem., 1894, 15, 73; Bull. Acad. Roy. Belg., 1804 (iii. < 28, 23, 
58 W. Hallock, Amer. Jour. Sci., 1889 {iii ], 87, 402. 
oe Daa 2 Perman, Proc. Roy. Soc., 1907, 799A, 310.- - 
60 G, Spezia, Atti R. Accad. Sci. Torino, 1910, 45, ii., 526; 
