on Basalt, &c. 
3°9 
Where there are two ranges of columns, with an intervening 
amorphous stratum, it is probable that the upper is the result of 
a second inundation of fluid basalt. It is well known that 
basaltic columns are most solid at the bottom ; and their convex 
articulations have been repeatedly observed. Since these consi- 
derations occurred to me, I have had no opportunity of exa- 
mining, whether the divisions approach nearer to the plane 
surfaces as they recede from the centre from which the prism 
Was generated, nor whether below that centre the convex sur- 
face of the articulations is inverted; but I think it by no means 
improbable, that subsequent observations may establish this to 
be the case, and thus confer on this hypothesis nearly all the 
demonstration of which it is susceptible. I may however add, 
that the phenomena of prismatic division in basaltic veins, per- 
fectly coincide with what might be inferred from the data upon 
which my reasoning has proceeded. In veins, it is obvious 
that the refrigerating or absorbing cause must operate with 
nearly equal force on each side of the vein ; and it follows, that 
two sets of prisms would be generated, which would be hori- 
zontal instead of perpendicular, and that, unless a mass of 
amorphous basalt was interposed between them, they must form 
a division in the middle of the vein, as, from the mutual impe- 
netrability of their fibres, they could not incorporate. The 
coincidence of the existing phenomena with these conclusions, is 
sufficiently remarkable; for, in numerous observations I have 
made on the basaltic veins which affect the prismatic configu- 
ration, I found the prisms were always horizontal, and often, 
that there were two ranges of them. One of their ends applied 
to the wall of the veins, the other frequently united to an 
amorphous mass which separated them ; and, when no such 
