THE CONSTITUTIOX OF THE COPPER-TIN SERIES OF ALLOYS. 
47 
faint mottle and the polygons of slightly different tint, so common in the region of 
chemically uniform solid solutions. This observation, and others like it, do not, 
perhaps, exclude the possibility that some of the alloys below the solidus may be a minute 
complex of two materials. In accordance with the information supplied jointly by 
the microscope and the pyrometer as to the temperature of complete solidification, we 
have, therefore, ventured to draw the solidus Icmd with a flat between c and m. The 
position of c is, we think, correct, but that of m is very uncertain. The break is, 
however, useful as a reminder that Ic and md are the solidi of two different substances. 
If the gap cm really exists, the solid alloys below it ought to be complexes of 
two bodies. 
Although we have not been able to detect the fact, it is quite possible, from the 
way in which the /3 disintegrates, that there may be such finely grained complexes. 
For, unlike tlie C transformation, which is an obvious solution of a in a liquid 
followed by the crystallisation of the very different the D change is one of 
disintegration throughout each /3 crystal, a change that might conceivably be a 
separation into two phrases. An alternative supposition is that the points c and m 
are coincident, and that the change at D is a case of dimorphism, the /3 and y having, 
at their transformation temperature, the same composition. But the evidence we 
have for the position of the solidus makes it difficult to bring m up to c. We fear 
we must leave the question undecided. It is certain, hoivever, that there is a 
transformation at the D temperature, that the cooling curves of the Sn 18 and Sn 19 
show heat evolution there, and that during this transformation the ji combs disappear 
without having filled the alloy. 
The Transformation of the Curve C'XDt 
Whether the solid alloys below Icmd are all uniform or not, it is certain that tliey 
undergo a profound change after crossing the line C'XD', this change being 
accompanied, as the cooling curves show, by an evolution of heat. In the case of the 
alloys to the left of CO', the change consists in the crystallisation of copper-rich a 
crystals out of the solid, as we have already proved. In the case of the alloys to the 
right of C', it consists in the crystallisation of white tin-rich 8, which appears in the 
form of rosettes, bars and fern-leaf. The unchilled Sn 16, of which we give a 
photograph (fig. 40), shows these new crystals, but with more tin the new material 
grows in amount until at Sn 20 it practically fills the whole alloy. The changes are 
well shown in Sn 17. Chills at 668° and 616° were uniform. The chill at 547 is 
uniform so far as much more than 99 per cent, of the surface is concerned, but dilute 
FeCL brings out a few large polygons bounded by extremely thin lines of 8, and in 
one place the lines have thickened a little by a moss-like growth of a similar white 
tin-rich body. There are also a few short lines of minute rosettes of white 8. This 
is the beginning of the pattern that develops at a slightly lower temperature, but the 
