28 
MESSRS. C. T. PIEYCOCK AND E. II. NEVILLE ON 
The growth of the large combs, which now so nearly fill the field, shows that the 
process of solidification has made great progress since the previous chill, but the 
minute cond)s of chill primary are an evidence that the crystallisation of the a was 
not quite complete. 
Sn 2. Chilled at 957° (fig. 3a and fig. 3). 
The surface, when examined under the usual normal illumination, was full of 
uniformly dark a combs in a slender network of pure white. It is evident that before 
the moment of clnlling tlie solid a had grown until it almost filled the ingot. No 
cores can 1)e detected (fig. 3a). Tlie photograph we reproduce, in fig. 3, was taken 
with oblique illumination in order to sliow the differences in orientation that exist in 
the different grains, or filled-out skeletons, that make u}) the ingot. A change in the 
direction of the incident light would have completely altered the disposition of light 
and shade. By this method of illumination the mother-substance, which does not 
scatter much light, is less evident. Tlie three figures, 2, 3, and 3a, taken together, 
illustrate excellently the growth of a granular solid from skeletons. Figs. 3 and 3a 
are photographs of the same portion of the surface of the ingot. 
Sn 2. Chilled at 932° (not reproduced). 
As usual in this group of alloys, the white substance between tlie combs becomes 
visil^le by polishing alone. H(Jl turns the combs a rich copper colour and leaves the 
mother-substance white. It is now present in smaller amount, not in a continuous 
network, but as rows of isolated dots and short lines, and in the three-branched 
jiatches to be found wfiiere three lobes of a meet. The amount ot mother-substance 
was estimated by counting tlie squares of a ruled grating. Several independent 
determinations gave values between 2 and 3 per cent, for the fraction of the whole 
area occupied by this white mother-sub.stance. A higher power still shows the 
merest trace of chill primary. The a combs must be very uniform, for a prolonged 
polish leaves tliem a uniform purple, and by no method of etching can cores be 
developed in them. It may be mentioned here that a slowly cooled chill at 940° was 
heat-oxidised to the second series of tints, but at no tint did any trace of cores appear. 
This is the most severe test of homogeneity in the a that we know of 
Sn 2. Chill at 907° (not reproduced). 
The whole section of the ingot is similarly oriented, and it evidently consists of one 
crystal. We can now see cores in tlie a. 
Sn 2. Chill at 882° (fig. 4). 
In this ingot the specks of tin-rich mother-substance are reduced to much less than 
half a per cent, of the whole area. Our photograph shows a very uniform surface 
