8 
MESSES. C. T. HEYCOCK AXD E. H. NEVILLE ON 
liorizoiital line. In all cases in which the two components form solid solutions the 
solidus is a sloping line. 
The case where the solid phase is a i^iire component or a pure comjDound may be 
regarded as a limiting case in wliich the mutual solubility in the solid condition is 
indefinitely small. The solidus is then the vertical line through the fixed composition 
of the solid phase, while from another point of view it may be regarded as the 
horizontal line through a lower eutectic point or angle in the liquidus. This will 
become more evident if we follow Roozeboom through his very important description 
of the complete solidification of a liquid which finally forms a uniform solid solution. 
Returning again to the mixture represented by the vertical line through n, let us 
suppose that, after the formation of a very little of the solid phase defined by o, the 
system is allowed to cool very slowly. The separation of the o crystals wiU cause the 
residual liquid to be richer in B, that is, to be represented by a point on the liquidus 
a little to the right of n, and as more and more solid is formed, the point defining the 
residual liquid will travel more and more to the right. But such liquids will not be 
in equilibrium with the crystals first formed, and hence if sufficient time is allowed in 
the cooling, a continuous rc-solution or transformation of the solid phase will go on, 
while at the same time it grows in amount. Finally, when the temperature q is 
reached, defined by the intersection of the vertical through n with the solidus, the 
system will, in the ideal case of perfect equilibrium adjustments, be a uniform 
crystalline solid solution, the crystals being surrounded by a vanishing!}^ small amount 
of mother-liquid of the composition p, where p is defined as the intersection with the 
liquidus of the horizontal through q. Of course, in a real experiment, for which 
infinitely slow cooling is not possible, there will be a chance that the earlier crystals 
may never l^e wholly transformed, but may remain as cores richer in A than the solid 
outside them. In this case the temj^erature of complete solidification will be lower 
than the theoretical one given by the point q. 
Tlie fact, taught us by the phase-rule, that in a binary mixture in which only 
concentration and tenq)erature are variable a particular liquid can in general onl}^ be 
in equilibrium with one solid phase makes it certain that the solidus as defined above 
is theoretically the same curve as the melting-point curve. 
We can now proceed to consider the special form taken by the solidus in the 
copper-tin alloys. As determined by us through the study of chilled alloys, and 
drawn in Plate 11, the solidus consists of the Iwoken line Af>/cmdc/EoE 3 H'H"K'. The 
angular points on this broken line, with the exception of are pretty accurately 
determined, but the exact shape of the branch A6 has been rendered uncertain by the 
impossiljility of obtaining true equilibrium transformations during the cooling of the 
corresponding alloys, and the l)ranches Ic and mdef may, from the same cause, be a 
few degrees too lo’w. Again, we have drawn E^Eg as a vertical straight line, while it 
is possible that it may be slightly sloping and curved. The point H' is fi'iirly certain, 
while we have not yet enough evidence to fix But, on the whole, we have 
