THE CONSTITUTIOISr OF THE COPPER-TIN SERIES OF ALLOYS. 31 
except by the different grains reflecting light differently. The a is not, however, 
homogeneous, for after a time a pattern develops on the etched surface consisting of 
dark copper-rich combs. These are similarly oriented thioiighout a giaiii and aie not 
separated from the a outside them by a sharp margin, iliey are evidently cores 
produced by the imperfect assimilation of the last drops of licpiid. The appearance of 
this chill leads us to think that Sn 4 is completely solid at all temperatures below 
830°. Thus the microscopic examination of chills of Sn 2 and 4 gives us, at all events 
approximately, two points on the solidus Ah. 
V.s.c. chill at 800° (fig. 7). 
This contains no tin-rich mother-substance (an estimate made the amount less than 
one-tenth per cent.), but no ingot was obtained free from cores, although in the most 
perfectly slowly cooled ingot, which we reproduce in fig. 7, these cores can haiely he 
detected. In order to show the character of the coining in the a combs, we give m 
fig. 5a a photograph of an ingot, also chilled at 800°, after a cooling that was slow, hut 
not quite so slow as that of fig. 7. The combs have grown so as to form three grains 
reflecting light differently, but in two of the grains there are well-marked copper-rich 
cores. The difference between the soft indefinite boundaries of the cores and the 
sharp margins of the combs is characteristic. The very dark angular spaces between 
the combs were cavities from which the mother-substance had receded during the 
cooling. 
Sn 4. Chill at 775° (fig. 8). 
In the best slowly cooled chills there is no mother-substance, the patches that 
represent separate a crystals dovetailing into each other with no interstitial matter. 
But even in the most slowly cooled ingots there are cores in the a, and in ingots 
rather more rapidly cooled there is a complete pattern of such cores. Even in the 
presence of cores, if the cooling before the chill has been at all slow, tlieie is 
extremely little of the tin-rich mother-substance. The photograph we give closely 
corresponds to the above description, 
Sn. 4. Unchilled (fig. 9). 
The ingot fig. 9, cooled slowly down to the melting-point of tin, is important, as it 
is quite free from the brilliant white tin-rich patches of the C' complex that form such 
a well marked feature in the Sn 6 similarly treated. This slow-cooled Sn 4 shows no 
cores, and consists of several patches fitting closely together and only distinguishahle 
by the different angles at which they reflect light. The small black spots, a 
millimetre or two across, seen in the photograph are holes in the alloy. Ibis photo¬ 
graph closely resembles photographs of pure platinum and pure gold that have been 
