THE CONSTITUTION OF THE COPPER-TIN SERIES OF ALLOYS. 
57 
bold to assert that this ingot was, immediately before the chill, a uniform solid, but 
we are convinced that such was the case. 
The other chill (fig. 64) of Sn 24 was also a slow-cooled chill at 635°, a little way 
below the transformation curve. It is at the same magnification as the preceding 
chill. We now see that the true crystallisation of -q has taken place, and that the p 
nearly fills the ingot. Perhaps we may remind the reader that, on account of the 
ferric chloride etch, the tin-rich is dark and the copper-rich mother-substance is 
white. Had the jDattein been broug'ht out by heat oxidation, it would h^-ve lieen 
reversed. 
Sn 25. 38'4 per cent, of tin. 
This alloy is especially interesting because it has the formula CugSn, and because 
of the many physical properties wdiich reach their maximum intensity in the 
unchilled alloy of this composition. There has on such grounds been a consensus of 
opinion that it is a chemical compound. So far as the slowly cooled and unchilled 
alloy is concerned, this view is almost certainly correct. But we find that when a 
liquid of this composition begins to solidify, the crystals first formed are considerably 
richer in copper than the mother-substance, so that the crystals cannot be CugSn. 
At a lowmr temperature, howmver, the compound itself crystallises out of a solid 
solution. This occurs at E , the summit of the transformation curve, a point where 
the cooling curve shows a large evolution of heat and a phenomenon analogous to 
surfusion. 
Sn 25. F..9.C. chill at 720° (fig. 65). 
The ingot was not granulated by the process of chilling, though the figure makes 
it evident that there was a considerable amount of matter liquid at the moment of 
chilling. Polishing alone shows the combs in a more brilliantly white network of 
mother-substance. The usual etching reagents do not give a contrast betwmen the y 
combs and the mother-substance sufficiently well marked for photography, \ve 
therefore adopted the method of heat-oxidation, the ingot being heated over a 
Bunsen until the first tint of orange appieared. Under this treatment the y combs 
oxidised more rapidly than the mother-substance, becoming a dark orange while the 
lattei lemained almost white. This selective oxidation of tlie y, which is reproduced 
in the photograph, proves that the y combs are richer in coppier than the liquid out 
of which they are crystallising, and therefore that they cannot be the compound 
CuoSn. The heating has, however, to some extent decomposed the y, leaving a rim 
of whiter material round each lobe of the combs. This and other features are better 
seen in fig. 65a, which is a more highly magnified photograph of the same ingot. 
Here we can see that not only the borders of the combs but also the whole of some of 
the smaller masses of y are paler than the rest; this probably indicates that they are 
VOL. CCII.—A. J 
