272 
MESSES. C. T. HEYCOCK AND F. H. NEVILLE 
change caused by recalescence followed by annealing. The white pattern of the 
440° C. chill is of D, and is characteristic of the D that crystallises out of uniform 
beta when the temperature of the ingot reaches that of the cL line. In the second 
ingot, at one time like the first, the L complex has been formed and destroyed, and 
the reaction between the alpha, remaining from the complex, and the D bars has 
removed the latter completely. A comparison of these two ingots only might lead 
one to think that the residue was D, but a study of ingots intermediate in temperature 
of chilling proves that the white residue is alpha. 
Forced Recalescence. 
The fact that recalesced ingots lose all trace of the recalescence pattern when 
heated above 520° C., at which temperature the beta is solid, makes us doubt if it 
would be possible to trace the liquidus of Y by seeding liquid alloys with fragments 
of a recalesced alloy. But we have been able to show that beta, which, spontaneously, 
does not often recalesce above 410° C., can be stimulated to do so at all temperatures 
up to 515° C. 
The experiment was conducted as follows :—An ingot was melted, a thermocouple 
placed in it and the ingot allowed to cool slowly while its cooling curve was traced on 
the recorder. At a selected temperature the ingot was touched with a cold steel 
wire. The contact generally caused a slight and momentary drop in temperature 
followed by a sudden jump which sometimes rose as high as, but never above, 524° C. 
We have in this way forced the recalescence of alloys varying in composition from 
A1 15 to A1 23. 
The recalescence of A1 15 could not be effected above 408° C. The change of 
structure is well seen by comparing the pattern of a recalesced ingot with that of 
a control ingot that, was not recalesced but otherwise had precisely the same thermal 
treatment. In the case of A1 15 the control proved to be uniform solid solution with 
hair-line boundaries between the polygons and a few small triangular pools of matrix. 
In the recalesced ingot there are broad jagged edges of a new material separating 
the polygons, as well as a certain amount of this material inside the polygons. 
The change in the case of A1 15 took place in the alpha and is not altogether 
comparable with the cases of forced recalescence of the beta that we are about to 
describe. In all the cases in which a uniform beta was forced to recalesce two new 
substances appear, a network or diaper of a somewhat golden material that resists 
etching and has all the character of alpha, and which we shall call alpha, and another 
material the crystallisation of which determines the pattern; this second substance 
etches dark and is eaten away by the etching reagent, it fills the cells of the alpha 
network ; we have already called this substance Y. These two substances preserve 
their character in all the cases of forced recalescence from A1 18 to A1 23. 
A1 18 was forced to recalesce at 500° C. and the temperature rose to 504° C, The 
