MancJiester Memoirs, Vol. li. (1907), No. 15. 5 



not enable us to see individual brickbats; we are looking, 

 so to speak, at fragments of masonry which the brickbats 

 build up. 



There are other lines of evidence which bring out 

 clearly the crystalline character of the separate grains of 

 metal. If you cast metal on to a smooth surface of glass, 

 under some conditions of casting which are not very easy 

 to reproduce, but which you sometimes have the luck to 

 light upon, you find when you examine the surface that 

 there are curious geometrical markings upon it which 

 appear to be due to the presence of very minute bubbles 

 of air or gas which are caught on the plate or perhaps have 

 been occluded in the metal, and have come out in the process 

 of solidification. These bubbles instead of taking a casual 

 round formation, have a more or less distinct geometrical 

 figure corresponding to the particular type of brickbat of 

 which the crystal is built up. They are, as it were, 

 hollow brickbats. They are holes representing the 

 absence of certain brickbats in the surface of the metal. 



Examples of this action are shown in Figs. 5 and 6, Plate 

 III, which are photographs of a cadmium surface cast on 

 smooth glass. Fig. 6 is part of a single grain only, with a 

 very high magnification (over 4,000 diameters). These 

 show in an unmistakeable way the geometrical character 

 of the holes formed by tiny air bubbles, caught between 

 the metal and the glass at the moment of solidification. 



The crystalline markings which are a familiar sight 

 on the free surface of a metal solidifying after being melted 

 in an open dish, serve to show that in the growth of crystal 

 grains the process goes on in a manner that is described 

 as dendritic. That is to say the grain grows, as a tree 

 grows, by spreading branches. These branches rush out 

 in various directions, and then the intermediate spaces are 

 filled up at leisure. Again, in the electrolytic deposit of 



