THE CONSTITUTION OF THE COPPEK-TIN SERIES OF ALLOYS. 
59 
give two photographs of Sn 26, one at 656° showing the characteristic light and dark 
patches of the grains of a chemically uniform body, fig. 67 (this is just above the 
transformation line), and also a chill at 612° (fig. 68), which shows the complete 
crystallisation of the CugSn. 
An ingot of Sn 26 was slowly cooled to 380°, and maintained at that temperature 
for four hours, then chilled. The lines dividing up the t] into bars are visible in 
places, though very narrow. They contain a little of the H body, and practically 
none of the I eutectic. 
The cooling curve of Sn 26 has a peculiarity which at first looks like an accidental 
defect, but which we have verified. The first long halt gradually rounding to 
steepness is due to the almost isothermal solidification, the latent heat of solidification 
ceasing to be evolved when the temperature has fallen to 700° ; this gives ns very 
roughly the position of the solidus. The singularity of the curve lies in the second 
halt. Here there is a sudden rise due to causes analogous to surfusion, the heat 
evolved being the latent heat of transformation of the solid y into plates of CugSn. 
The short flat which follows the surfusion is unlike anything else in the cooling 
curves. It is due, we think, to a third evolution of heat rapidly following the 
second. We can only attribute this to the transformation of the residual y into a 
mixture of CugSn and liquid, a reaction that becomes much more marked in the 
succeeding alloys. If this be the right explanation of the singularity, then the 
reaction occurs at a slightly higher temperature than in the case of alloys 
with more tin. This, again, may be due to the fact that when the change, 
y = CugSn + liquid, takes place in Sn 26, there are already crystals of CugSn in the 
alloy which may serve as nuclei and precipitate the reaction. In order to verify this 
cooling curve we rejjeated it with a different pyrometer and a fresh sample of alloy, 
and found the same phenomenon. 
Sn 27. 40‘8 per cent, of tin. 
In the cooling curve of this alloy the second and third halts are completely merged 
into one. In fact, the solidus and the liquidus meet very near here at the 
temperature of the G point, and the possibility of a uniform solid solution ceases. 
The meeting of these two cuiwes at the G temperature is, as Professor Hoozeboom 
has pointed out to us, a theoretical necessity. 
Sn 27. S.c. chill at 700° (not reproduced). 
This ingot is still covered externally with large rounded rosettes of y in relief 
When cut and polished, about half the surface of the section consists of similar 
rounded combs, the rest consisting of chill primary immersed in a slender, Ijright 
network of the tin-rich residue of mother-substance. The FeClg etch blackens this 
network. The surface of the large y combs is quite uniform and free from inclusions 
