ON THE CRYSTALLINE STRUCTURE OF METALS. 
283 
the constituents. We therefore abandoned the attempt to observe this process in 
iron, but our attention was directed to the study of similar processes of annealing 
or re-crystallisation as seen in more fusible metals, particularly lead. 
The question, What is the temperature required to produced re-crystallisation in 
lead ? was raised by the appearance we observed in specimens of plumbers’ sheet- 
lead that had been etched with dilute nitric acid. When thus treated, ordinary 
sheet-lead exhibits a remarkably brilliant crystalline structure on such a large scale 
that no previous polishing is required. The etched surface shows all the appearances 
due to the selective effect of oblique light on etched crystalline surfaces; when the 
specimen is turned the light catches the various crystalline grains in turn, the colour 
and brightness being uniform over each grain, but different on different grains. 
Fig. 2, Plate 3, is a photograph of such a surface magnified two diameters : these 
surfaces tarnish immediately on drying, and must therefore be observed and pho¬ 
tographed while wet with the acid solution. 
An examination of this etched surface reveals a peculiarity in the configuration of 
the crystalline grains ; these are seen to have many remarkably straight boundaries 
meeting at sharp angles, several sets of parallel boundaries being frequently observed. 
These features, which strongly resemble what we had previously observed in wrought 
copper, are to be ascribed to the frequent occurrence of twin crystals. In our earlier 
observations on twin crystals their presence had always been readily detected by the 
configuration of slip-bands produced in them by slightly straining the specimen after 
polishing. An instance of twinning observed in sheet lead by this method has been 
given by us in a previous paper (‘ Phil. Trans.,’ A, vol. 193, 1900, Plate 26, fig. 40). 
In the present instance this method of detecting twins is not available, as the rough¬ 
ness of the surface and the great depth of etching employed make it impossible to 
study the slip-bands. But the presence of twin lamellae nevertheless becomes evident 
under slightly higher magnification with oblique light. Fig. 3, Plate 3, is a photo¬ 
graph of crystals with twin lamellae magnified 40 diameters. The figure illustrates 
the selective effect of oblique illumination, which has picked out a few isolated 
crystals, lighting these brilliantly while neighbouring ones remain almost dark. 
Within the area of the brightly-illuminated grains, a number of dark patches are 
seen, and these show the straight boundaries occurring in parallel sets which are 
characteristic of twinning. In this instance there are three distinct parallel sets 
of dark bands, and the fact that they are twin lamellae becomes apparent when the 
stage carrying the specimen is rotated, thus altering the incidence of the light. As 
the specimen is turned, the grains that were bright become dark, but presently 
some of the patches that were previously dark shine out brilliantly, all the bands 
which are parallel to one another flashing out simultaneously. Fig. 4 is a photograph 
of the same field as fig. 3 after a rotation of about 30°, and illustrates this appear¬ 
ance. But those parallel bands which catch the light simultaneously are evidently 
portions of the crystal in all of which the orientation of the elements has been 
changed by the same amount ; in other words, they are twin lamellae. 
