274 
MESSRS. C. T. HEYCOCK AND F. H. NEVILLE 
The recalesced ingots retain the new pattern when annealed at any temperature 
below 520° C., the only changes produced by the anneal being to segregate any fine 
pattern into larger particles and, if there is a redundant phase, to remove it by the 
reaction alpha+ D = Y. 
It would now be possible to determine the complete equilibrium diagram of Y, 
by giving a prolonged anneal, at successive temperatures, to a number of recalesced 
alloys. 
Experimental Details. 
The alloys used in determining the liquidus, and also for many of the cooling 
curves, were in large quantities of several hundred grammes and were carefully analysed 
at the end of each experiment; but such a large number of alloys had to be made 
that, as a rule, we dispensed with analysis, each alloy being prepared by weighing 
out appropriate quantities of pure gold and pure aluminium and melting them 
together in a vacuous silica tube. This method of making the alloy seems to exclude 
the probability of any error except perhaps that of assuming the percentage of 
aluminium to be slightly higher than the true value. 
Etching. 
The etching reagents were bromine water or aqua regia ; the two give practically 
the same pattern, aqua regia being more rapid and clean in removing the smear of 
metal produced in polishing. Alpha always etches bright whatever other phase is 
present. D etches bright in the presence of beta or of Y, but generally darkens if 
alpha is present. Beta and Y both etch dark, and are rapidly eaten away, in the 
presence of either of the other two phases. The result is a curious one in the case of 
alloys containing primaries of D ; if the ingot is chilled above the temperature of the 
L eutectoid the D is bright and the ground is dark, but if the eutectoid is present, 
then the alpha in it generally causes the D to darken ; if the alpha is removed, by 
recalescence and annealing, the D again etches bright, so that even without the 
microscope one could pretty safely infer what phases were present. Perhaps the only 
case where it is difficult to distinguish between alpha and D is in such an alloy as 
A1 20, after spontaneous recalescence and a long anneal (fig. 17). In this case there 
is so little of the bright substance, and the particles are so minute, that the 
microscope, alone, would leave us uncertain. 
We have to thank Mr. F. E. E. Lamplough for very valuable help in the 
thermometric work and for devising an excellent method for the automatic regulation 
of high temperatures. 
We have also to thank Prof. Pope, F.R.S., for his kindness in placing the resources 
of the University Laboratory at our disposal, and, in especial, for lending us valuable 
apparatus and lenses for microscopic photography. 
