56 
i\lESSES. C. T. HEYCOCK AND F. H. NEVILLE ON 
Sn 23. Slotv-cooled chills at 690°, 685°, 680°, 650° (not reproduced). 
These are all very much alike : when ignited they are uniform, at all events to the 
eye or a lens, but FeClg brings out a number of rather small angular and interlacing 
patches, which show the effect of varying brightness on tilting or rotation. A 
moderate magnification makes it evident that each joatch is broken up into narrow 
bands with still narrower lines of a mother-substance between them. In fact, the sepa¬ 
rations proper to the region below the transformation curve have taken place although 
on a very small scale. This detail is not perceptibly larger in the chill at 650° than 
in that at 690°, although a long period of slow cooling was allowed between the two 
chills ; hence we suspect that it occurred at lower temperatures during the chilling. 
Fig. 63, a chill at 685° of Sn 24, gives a fair idea of these ingots. 
Sn 23. S.c. chill at 620° (fig. 62). 
We are now below the transformation curve; the iridescence has increased and 
the patches are much larger, but, in addition, tliere has grown a large scale pattern 
of broad -q bands and sprays of what may be likened to foliage. This is the true 
breaking-up of the solid solution into two phases, and we see that the rj phase has 
very much increased as compared with Sn 22 and Sn 21. The chills at 590° and 
570°, and indeed all lower chills, have the same j^attern, so that the D' transformation 
has now vanished or become unimportant. Oblique illumination shows, by the 
position of the shadows, that the darker iiigredient has been etched out below the 
level of the lighter ground. Thus the darkening of the tin-rich phase is due to a 
real attack by the etching reagent. The pattern seen in these lower chills reminds 
one of the alloys from C to D when chilled below the transformation curve. In both 
cases there are ribands of a tin-rich material, often enclosing polygonal areas, and, 
branching from these ribands, a fantastic crystallisation of the tin-rich phase 
partly fills the enclosed areas. But in the CD alloys the ribands and fern-leaf 
crystallisation is left white and not attacked l:)y the etching, while in the DE alloys 
it is attacked and darkened; in the CD alloys it is S, in Sn 23 and Sn 24 it is r]. 
Sn 24 is similar throughout to the above, with the exception that in the chills in 
which the u phase appears it is more abundant; in fact, nearly filling the alloy. 
Primaries and chill primaries are obtained by chilling above 700°, and between the 
solidus and the transformation curve the chills are uniform when ignited; but when 
strongly etched show the iridescence and patchiness due to minute separations, while 
as soon as the chilling temperature falls Ijelow the transformation curve, the solid 
solution breaks up and the bars, or really plates, of rj, nearly fill the field. We give 
two photographs of Sn 24. The first (fig. 63) was a slow-cooled chill at 685°, that is, 
a little below the solidus. It shows no primaiy, but is full of iridescence due to 
minute linear separations; there are also some long slender bars of t]. It may seem 
