ON THE CHEMICAL NATURE OF ALLOYS, 43. 
solvent for the particles of metal next to it, and so promotes the combination 
of the metals where this can take place. When the alloy Sn Pb,, for instance, 
solidifies, it must not be strictly regarded as a homogeneous diffusion of tin 
and lead in one another, if Pillichody’s statement be correct (although it will 
not be far from it, as all the tin-lead alloys have the theoretical conducting- 
power), but rather as a homogeneous diffusion of tin and lead in one another 
from the formation of the alloy Sn, Pb, with the excess of lead mechanically 
diffused through the mass, as this, according to him, separates out before the 
alloy Sn, Pb solidifies. 
With respect to the chemical nature of these and other alloys in a liquid 
state little is known; it is, however, very necessary to draw a line between 
the solid and the liquid alloy, when speaking of their chemical constitution. 
No doubt the determination of the electric conducting-power of the liquid will 
throw much light on their chemical nature. The investigation of those 
where the so-called stationary points have been observed, and where the 
melting and solidifying points do not coincide, will be especially interesting. 
As yet, I believe, no such observations haye been made. 
Passing on now to the alloys of the second group, the question for consider- 
ation will naturally be, what is the cause of the rapid decrement in the 
conducting-power of Class B metals when alloyed with traces of one of the 
Class A metals (in the solid alloy) ? 
That it is not due to the formation of chemical combinations, is proved by 
the following :— 
I. At the turning-points of the curves representing the conducting-power 
of this group of alloys they contain only very small percentages of the Class A 
metal. Thus, at those of the alloys 
Percentage of 
IBisma th =ftTs cop ce RTRs tin 0:6 
Bismuth-lead!< So Masse lead 2:0 
Silver-tin. -'});. orbenateets - tin 2°6 
II. The great similarity of the curves of this group speaks greatly against 
the existence of chemical combinations in the solid alloy. The curves of the 
following series show this in a very marked degree :—the bismuth-lead, bis- 
muth-tin, copper-tin, copper-zinc, silver-lead, silver-tin. 
III, On examination of that part of the curve where the rapid decrement 
takes place, we find that it requires about twice as much lead as of tin to 
reduce the conducting-power of the Class B metal to the same extent: thus, 
to reduce that of silver to 67, it would require 0-9 vol. per cent. of lead, or 0-7 
per cent. of tin; and to reduce it to 47-6, 1-4 vol. per cent. lead, or 0-7 per 
cent. tin. Again, to reduce bismuth to 0-261, it would require 0-4 vol. per 
cent. lead, or 0-62 per cent. tin; and to reduce it to 0-255 with lead, or to 
0-245 with tin, it requires 1-76 vol. per cent. lead, or 0°85 per cent. tin *, 
From these facts we can hardlyassume that the rapid decrement in the curve of 
the Class B metal is due to the existence of chemical combination. The reason 
of this great decrement in the conducting-power might be looked for in the pro- 
cess of solidification, for it is well known what a great effect traces of foreign 
metals have on the crystalline structure of some metals, Cooke has, I think, 
stated that absolutely pure antimony crystallizes with great difficulty,—an ob- 
servation which I can corroborate ; for, when trying to obtain crystals of that 
metal for thermo-electric experiments, I found that the purer the antimony 
the smaller the crystals, in fact I could not obtain any for my purpose. 
* Phil. Trans. 1860, p. 171. 
