50 
MESSES. C. T. HEYCOCK AND F, H. NEVILLE ON 
or a little below, 500°, contains the same two phases, a and 8, in all the alloys from 
Sn 6 to Sn 20. Beyond Sn 20, even with a mere trace more tin, the eutectic 
vanishes, and there is so little of it in Sn 20 that we are disposed to regard it as an 
accident in that alloy, or as due to the fact that the alloys that we call Sn 20 may 
have a slight deficit of tin in them. If so, there is considerable reason to regard the 
8 substance of Sn 20 and of all the previous C' eutectics as being the compound 
Cu^Sn. In this case the compound is not present, at all events in crystalline form, 
above the transformation curve, and is finally produced by crystallisation out of a solid 
solution. On this view the transition from the condition of the alloy just above the 
point D' to that just below it must be compared to the conversion of a pure substance 
from the liquid to the solid state, only that in the case of the alloy we are considering 
hoth states are solid. The change, like that from a liquid to a solid, is an exothermic 
one. 
It is, of course, possible that the 8 material found in all the alloys from Sn 6 to 
Sn 20 may be a series of solid solutions varying in composition throughout this range 
of alloys, but the great similarity in the appearance of the 8, from its first aj)pearance 
in the scanty eutectic of Sn 6 to the moment when it fills the whole alloy in Sn 20, 
makes us think it far more probable that it is a pure compound of comstant 
composition. Up to the present we have not been successful, by means of a chemical 
solvent, in removing the a from such an alloy as Sn 16 in order to analyse the 8. 
The character of the pattern of riband, bar and rosette, or fern-leaf, seen in the CD 
alloys when cold and unchilled, is partly, we think, dependent on the greater or less 
perfection of the transformations that take place during solidification. Too rapid 
cooling during solidification will leave the grains somewhat copper-rich with some tin- 
rich mother-substance round them; in fact, the D transformation will not be com¬ 
plete. This will lead, when the 8 crystallises, to its doing so in broad ribands, 
enclosing areas in which it is more sparingly distributed as rosette or fern-leaf. In 
the chill at 558°, and the same is seen in the unchilled alloy of fig. 52, the 8 forms 
many straight pointed bars. This peculiarity, which is first seen in Sn 20, suggests 
the bars of Sn 21 and succeeding alloys. 
The DE Alloys .—All slowly cooled ingots of the DE alloys, but especially the 
higher chills, are more or less covered, particularly on the top, wnth a raised primary 
crystallisation. This suggests bunches of grapes packed closely together, but some¬ 
times, as in the upper chills of Sn 24 and Sn 25, there are suggestions of combs in 
the form of long bars with rounded serrations. In the lower chills the grape pattern 
alternates with patches of linear ruling, showing iridescence. 
We have examined an'unchilled ingot of Sn 20'5. This alloy seems to consist of 
masses of 8 in close juxtaposition, the trace of the C' eutectic visible in the Sn 20 
having disappeared. The bar pattern noticeable in some of the ingots of Sn 20 is 
now more marked, the bars having pointed ends, so that they resemble elongated 
willow leaves. But the bars now seem to differ a little in composition from the rest 
