32 
MESSRS. C. T. HEYCOCK AND E. H. NEVILLE ON 
published by Mr. Andrews and others. We thus see that the whole of the 7 per 
cent, of tin has been absorbed into the a. 
We tbiidv that we have demonstrated that the condition of this solid alloy, 
provided that it has been cooled so slowly that the solid and liquid phases have 
always remained in equilibrium, is finally that of a uniform solid solution. This 
uniformity is no doubt rarely attained in practice, a not specially slow-cooled chill at 
740° (not reproduced) more nearly representing the usual state of an ingot soon after 
solidification. In this there is a complete pattern of cores forming combs about the 
size of those on the outside of the ingot, and evidently coincident with them. These 
merge into another material, which is later formed, a. Finally, sharply divided from all 
the a, there is a very perceptible amount of mother-substance, which, however, was solid 
at the moment of chilling. Owing to the action of the ferric chloride etch this mother- 
substance, which is /3, has become black. It has, strictly speaking, no right to be 
liresent in an alloy of the AB group, but is the result of imperfect transformations. 
Thus we see that the soft bronzes of the AB group, unless very slowly cooled, are 
likely to be non-homogeneous in two complementary ways—the a grains may be 
cored with a material richer m copper than the mean of the alloy, and between the 
grains there may be a little of a hard, brittle tin-rich substance. 
Bepeated rolling and annealing at a sufficiently high temperature ought to bring 
about a re-action between the cores and the tin-rich substance, and so promote a real 
chemical homogeneity which would be impossible in the alloys with more tin. The 
indication of the cooling curve that these alloys solidify by one continuous process is, 
as we have shown, contirmed by the microscope. 
The BL Alloys .-—These alloys show at least two halts in their cooling curves, the 
first at the freezing-point, the second at 790°, the temperature of the line 6/C, a halt 
common to all the group. 
We now observe a different final structure, however slow the cooling may have 
been. In the slowly cooled and unchilled alloys the a. combs no longer fill the whole 
area, but are surrounded by a new type of mother-substance, the complex W. Ibis 
is well seen in the photograjjh of the slow-cooled Sn 6 (fig. 12), where the angular 
patches of pure white are derived from a tin-rich material that was left liquid at the 
termination of the a crystallisation. 
Ingots of Sn 6, chilled above the C temperature, that is, above 790°, resemble the 
corresponding chills of Sn 4 in containing large primary combs of a surrounded by 
chill primary, the whole being imbedded in a tin-rich matrix. The chill at 96G , 
fig. 10, is a fine example of this structure. Here the large a combs are remarkable 
for a great rectangular symmetry, and are quite free from cores. TTie ingot was not 
specially slow-cooled. 
Sn 6. V.s.c. chill at 805° (fig. 11). 
This has large combs of primary a and at least 10 })er cent, of mother-substance 
