290 
PROFESSOR J. A. EWING AND MR. W. ROSENHAIN 
parallel boundaries are visible, and their “ twin ” character is evident when they are 
examined in the manner described above. 
In our previous pa] )er we have already given the observation that twin crystals 
are rarely, if ever, found in a cast metal, but that they are frequently developed by 
strain, and that they are very commonly found in metal that has been annealed after 
severe straining. Our present observations on lead entirely agree with these state¬ 
ments ; but it may now be possible to suggest an explanation for the frequent 
occurrence of twins in annealed lead. When the metal solidifies from the liquid state, 
it does so by the formation of skeleton crystals, starting from a great number of 
centres, and the arms of these skeletons continue to grow until arrested by meeting 
with other growths. 
From these arms other arms again shoot out, and so on until the entire metal is 
solidified ; but each crystalline element as it settles into place on any of these arms 
must assume the proper orientation to enable it to fit in, and in the process of filling 
space by means of such a system of many meeting and interlacing arms, the formation 
of a twin would be almost impossible. But when the metal re-crystallises after severe 
strain, it does so by the growth of skeleton arms that must often start from a cleavage 
plane of an actual solid crystal, and probably the new elements deposited upon such 
a plane would find it as easy to assume the twin orientation as the normal. The 
idea that twin crystals are formed in annealed metal by growth starting from cleavage 
planes which have been sheared across in the process of compression is suggested by 
the very straight boundaries observed as a characteristic feature of twin lamellae. 
Having observed the comparatively rapid growth of crystals in strained lead at 
200° C., we extended our experiments to both higher and lower temperatures. Our 
observations were recorded photographically as before, but the general character of 
the results is so similar to those already illustrated that it is unnecessary to reproduce 
these photographs. Experiments were made at temperatures of 100° C., 150° C., 
250° C., and 300° C. 
In the absence of a reliable method of measuring the rate of growth of crystals, 
the following statements must be taken as based upon a rough general estimate. 
With this reservation, our observations over this range of temperatures may be 
summed up as follows :— 
(1) In lead which has been severely strained, re-crystallisation goes on at all 
temperatures from that of an ordinary room up to the melting-point. 
(2) T1 \e higher the temperature the more rapid are the changes in the crystalline 
structure. 
(3) The rate of change varies with different specimens, probably depending upon 
the nature and quantity of impurities present, and upon the severity of the strain to 
which the metal has been subjected. 
(4) The size of the crystals ultimately produced does not vary appreciably between 
100° C. and 300° C. In lead annealed at the temperature of the air, vqry large crystals 
have not yet been obtained, but this is probably only a question of time. 
