as illustrative of Geology. 79 



masses is the globular. In favourable circumstances this form 

 appears very distinct in glassy slags. It displays itself on the 

 free surface of blast-furnace slags, and is found likewise in 

 every other kind of slags, for instance, in the melting processes 

 of lead and copper. In like manner obsidian takes a globular 

 shape, whether it be confined in small quantities like mareka- 

 nite in peailstone, or form, as in Iceland, whole streams of lava. 



The most regular form assumed by amorphous masses is that 

 of hexagonal prisms. We conceive their formation to resemble 

 the mutual flattening of equally large globes, placed in juxta- 

 position, occupying mutually such a station, that lines drawn 

 through their centre form equilateral triangles. This structure, 

 which every glass can assume, but which, however, is often 

 found in such substances not glassy, as have passed from the 

 fused into the solid state, is sometimes found in perfection in 

 slags. We may here and there indeed trace on their actual 

 surface the original transition from the globular formation to 

 the perfect hexagonal prism, and at the same time understand 

 how irregular prisms arise from a difference of magnitude in 

 the adjoining spheres of attraction ; the same phenomena, which 

 in certain circumstances are displayed on a grand scale in tor- 

 rents of lava in common basalt, and in various other ignigenous 

 rocks. 



The formation of vesicular cavities in the slags merits parti- 

 cular consideration, because, in regard to this, the most perfect 

 analogy is displayed with what is perceived in many volcanic 

 rocks, and especially in streams of lava. The vesicular struc- 

 ture has in both cases a twofold origin ; either resultino- from 

 vapours and gases which, during the formation of the fused 

 mass, were received by it ; or produced by a formation of va- 

 pour which is to be ascribed to an accidental approximation of 

 moisture. The formation of cavities of the former kind has 

 been already treated of, in the description of the appearance of 

 iron and graphite in the vesicular spaces of blast-furnace slags. 

 Where the latter cause exists, both slags and torrents of lava 

 display the greatest number of cavities in places where they 

 come in contact with a moist surface, and on the actual sur- 

 face on which moisture was operating in one way or another. 

 In the case of glassy slags, if water be poured upon the sur- 



