172 
DE. A. IMATTHIESSEX OX THE ELECTEIC 
for instance, the gold-tin and the silver-tin curves, and connect the two tu rning -points, 
and we shall find that the curve so made is exactly similar to that of the gold-silver. 
Again, look at that of the silver-bismuth, and compare it in the same manner with the 
tin-silver and tin-bismuth curves, and we obtain similar results. 
The gold-bismuth and silver-bismuth cmwes are exactly similar ; the alloys of gold and 
bismuth conducting less than those of silver and bismuth, as would be expected, as silver 
is a better conductor than gold. 
From the similarity of the curves of alloys, where we may assume, from their chemical 
behaviour, that we have only a solution of one metal in another, we may always draw 
approximatively the curve of the alloys of any two metals, if we know to which class 
they belong. Thus before a single copper-gold alloy had been determined, the cune 
was almost correctly drawn, and agreed with that which was aftei’wards found by expe- 
riment. 
That some alloys are chemical • combinations may be deduced from the following 
facts : — 
1. At the turning-points of the curve we generally find the alloys contract or expand. 
2. There is no regular form of curve (see those of gold- tin, gold-lead, and silver- 
copper), so that it cannot be a 'priori even approximatively represented. 
3. At the turning-points the alloys contain large per-centages of each metal. 
4. At the turning-points of the curves the alloys are different from each other in 
appearance (crystalline form, &c.). 
Now let us for a moment examine the gold-tin curve, it being the only nearly com- 
plete one of this class, and starting from the tin side of it, we find a slow decrement in the 
conducting power to the alloy Snj Au ; then a slow increment to Sm An, and from this 
point a slow decrement to Sn Aua. For reasons before given, no alloys could be deter- 
mined between SnAug and that containing 2’7 per cent, of tin, fr’om which point the 
curve goes in a straight line to pure gold. Starting again fr’om the tin side of the ciuve, 
we may regard the alloys down to Siij Au as a solution of a chemical combination in a 
metal ; from Sng Au to Sua Au, and from Snj Au to Sn Aug, as a solution of two chemical 
combinations in each other, as all of these points are connected with each other by 
straight lines; and from Sn Aua to the alloy containing 2 ’7 per cent, of tin, as a solution 
of a chemical combination in an alloy, and from this point to pure gold as a solution of 
a small quantity of tin in gold. 
Now these turning-points contain, first, Sn^Au, 60 per cent; second, Sm Au, 37 per 
cent.; and third, SnAua, 13 per cent, of tin, which may very well be considered to be 
chemical combinations. On looking at their specific gravities, we find that Snj Au has 
almost its calculated one, whereas Sn,^ Au expands, and Sn Aug contracts more than any 
other of the gold-tin alloys experimented with. 
In their general appearance we see that Sn Aua and Sn^ Au are not at all ciy^stalline. 
and have a very glassy fracture ; but Suj Au is exceedingly crystalline, and always shows, 
on being broken, the cleavage plane of a crystal from one end of the piece to the other. 
