228 
MK. J. C. HUMFliEY OX THE EFFECTS OF STEAIX 
metal, but that in the centre there rvas simply the uj^per crust, about inch to 
inch in thickness, and this could be easily cut away from the surrounding portions. 
The crystals forming the crust fell away from one another to a certain extent when 
unsujjported by tlie metal beneath. This was due to the well-known fact that the 
metallic impurities form a eutectic with part of the lead and collect in the crystalline 
boundaries. This eutectic has a lower freezing-point than the pure lead, and hence 
remains molten after the main mass has solidified. The boundaries are, therefore, 
lines of weakness, and the crystals tend to fall aj^art when the siij^porting metal 
below is poured away. 
The u})per surface of the crust which had been in contact with the air was, apart 
from the skeletal markings, fairly plane, but the lower surface was covered with 
numerous small spiky projections, these being regular in shape and position over each 
individual crystal, but varying from one crystal to another. Fig. 4 (Plate 2) shows such 
a surface natural size, and the various crystalline grains can be easily distinguished by 
their different textures. It will l)e seen that in some cases certain spikes are raised 
rather aljove the others, forming ridges which run in different directions across the 
crystals. These correspond to those upon the upper surface vrhich have been 
mentioned above. A portion of the same surface at the junction of three crystals is 
shown magnified to 10 diameters in fig. 5, and the different shape and orientation of 
the spikes upon each crystal can be clearly seen. These spikes have more or less 
curved surfaces, but generally roughly assume the form of octahedra. Some of them 
exhibit a somewliat wavy surface, as if the metal had solidified in successive layers. 
Such markings are probably due to a certain unsteadiness when the molten metal is 
poured away, the latter flowing two or three times over the solidified surface before 
finally leaving it. 
The various crystals were now separated one from another with a fretsaw. The 
rough sides were smoothed by cutting away the spiky projections with a sharp knife, 
and the crystals were then etched with dilute nitric acid. 
The etching consisted of two o})erations; the crystals were first placed in a 20 per 
cent, solution of nitric acid for about half an hour. This removed any roughness 
from the surface and produced a fairly smooth plane specimen. They were next 
immersed in a 5 per cent, solution, in order to produce the geometrical pits by means 
of which most of the following phenomena have been studied. In order to obtain a 
good development of sucli pits it was found necessary to prolong the etching for from 
one to two days; if particularly large pits vrere recpuired the |)rocess ^vas even longer. 
It was found in general that the stronger the acid solution the smaller were the pits, 
and vice versd. With a 5 2 )er cent, solution a beautiful system of contiguous pits 
covering the entire surface of the specimen could be produced which was easily visible 
at a magnification of 20 diameters or less. 
The shape of these etched pits has been found to be that of a portion of a negative 
cubo“Octahedron. Such a figure is shown in fig. G (a) and (d ); in (a) one of the 
