TtiE CONSTITUTION OF THE COPPEK-TIN SEPJES OF ALLOYS. 
41 
Sn 15. 
The higher chills resemble those of Sn 14. 
o 
Sn 15. Chill at 775°, s.c. (not reproduced). 
The ingot contained no a, but consisted of large ^ combs in a white tin-rich mother- 
substance. A glance at the diagram shows that this is what we should expect. The 
yS was at first not striated, hut an ignition to a temperature of about 300° developed 
striation when the surface was again etched. 
Sn 15. Chill at 600°, v.s.c. chill (fig. 33). 
This is important. It consists of several large polygons differing, on account of 
orientation, very much in brightness. There are also fainter polygonal lioundaiies, liut 
these which do not separate regions of different brightness, may be cracks. There is 
no a, and the y8 is not striated. The ingot is undeniably a uniform solid solution. 
We have no other chills of Sn 15. 
The fact that in this group of alloys all the primary a is absorbed at 790°, and that 
only at lower temperatures does a. again crystallise, accounts foi fhe singnlaily 
disjointed appearance of the combs seen in the unchilled alloys. The photogiaph of 
Sn 12, published in our paper in the ‘ Proceedings,’ vol. 68, shows this peculiarity, 
but in slowly cooled ingots of Sn 14 it is much more marked. For in Sn 12 a little 
of the primary a has persisted, and so decided the orientation of the new ct, while in 
Sn 14 there were no such nuclei, and the character of the primary a combs is not m 
the least reproduced in the later crystallisation. The chills of Sn 14 at 530° and 
470° illustrate this when compared with the chill at 800° (fig. 28). 
Sn 15'5. 25*5 'pcr cent, hy weight. 
The alloy is important, because at this composition occurs the angle C m the 
liqiiidus, and also, apparently, the eutectic angle C of the curve ^C'XD'E', which is 
to the solid solutions yS and y what the liquidus is to a liquid. The cooling cuives of 
Sn 15 and Sn 16 enable us to see what the cooling curve of Sn 15-5 would be like. 
The short upper flat of 15 would be absent and also the long flat at C, because there 
is neither crystallisation of a at the freezing-point nor an isothermal transfoi mation 
of a into /S. Instead, the commencement of solidification is marked, as in Sn 16, by 
a very gradual slope due to the heat evolved in the continuous reaction between the 
solid yS and the remaining liquid, a reaction determined by the solidus Ic and the 
liquidus CD, the process of solidification being quite similar to that of the AB group. 
The solidification having completed itself at about 740°, the cooling curve becomes 
steeper and corresponds to the cooling of a solid with no evolution of latent heat, 
VOL. ccii.—A. G 
