288 
PROFESSOR J. A. EWING AND MR. W. ROSENHAIN 
precaution to keep the incidence constant, a small change will sometimes cause the 
entire disappearance of a crystal whose facets happen to be near the critical angle. 
But, provided any portion of a crystal is shown bright in the photograph, any changes 
in its outline or surface are certainly genuine, for light that catches the facets on 
one part of the crystal must necessarily catch all similarly oriented facets also. 
Where, therefore, the same crystal appears in successive members of a series, the 
photographs may be taken to represent its life-history accurately. 
Another feature of these photographs requires explanation. In figs. 18 and 19, 
and more or less in most of the others, an irregular network of dark lines or channels 
is seen to cover the surface, and these lines obviously have little or no relation to 
the crystals. As a matter of fact, these lines are only found in specimens that have 
been etched, and then annealed and re-etched. In such a case as fio\s. 3 and 4, 
where only one etching has been applied, similar lines are seen to follow the inter¬ 
crystalline boundaries. In that case they are simply “ corrosion channels ” caused by 
the action of the acid at the crystalline junctions. These channels are of compara¬ 
tively great depth, and when such a specimen is allowed to re-crystallise, and is then 
re-etched to develop the new crystalline structure, this second etching is rarely 
carried deep enough to remove the old etched corrosion channels, and these remain as 
a network over the surface, simply indicating the positions in which crystalline 
boundaries formerly existed before the structure became changed. This view is borne 
out by the observation that deeper etching will eventually remove all trace of this 
network. In many instances these traces of the former crystalline boundaries are 
very useful as indicating the nature and extent of the growth that has taken place. 
On the other hand, the deep etching that would be required to remove these lines was 
considered undesirable, as very deep etching might of itself alter the apparent shape 
of the crystals. By observing the effect of successive deep etchings on the pattern of 
a stable specimen we satisfied ourselves that it would require exceedingly vigorous and 
prolonged etching to cause any visible change of pattern, and that, in tire specimens 
illustrated in this paper, the changes cannot possibly be ascribed to that cause. 
Our next experiment was to expose a freshly-crushed specimen of lead to a tem¬ 
perature of 200° C. for a long time, and to watch the changes produced in its 
crystalline structure. The annealing ovens* used for this purpose consisted of wooden 
boxes, lined with asbestos cloth, and heated by means of one or more incandescent 
electric lamps. The temperature could be regulated by regulating the current in 
the lamps, and, more conveniently, by adjusting the lids of the boxes so as to admit 
more or less air. Variations of 10° C. in the temperature were rare. 
Figs. 20 to 27 inclusive, Plates 8, 9, and 10, illustrate our observations on one 
specimen treated in this way. Fig. 20 represents a marked area on the etched surface 
* These annealing ovens were devised and used by Mr. S. R. Roget in his researches on effects of pro¬ 
longed heating on the magnetic qualities of iron. ‘ Proe. Roy. Soc.,’ May 12 and Dee. 8, 1898. 
