[ 12 ] 
Thus, for inftaiice, I have often feeii a range of horizon- 
tal columns placed againft another, in which all have 
appeared perpendicular ; while a third mafs, adherent to 
one or other of the former, and perhaps to both, has pre- 
fented itfelf, with its columns obliquely difpofed. It ap- 
pears then, that a perpendicular pofition, with relpedt to 
the horizon, is by no means a chara(5teriific peculiar to 
bafaltine cryftallizations, as hath been commonly pre- 
tended. On the contrary, whole groups frequently occur, 
that exclufively affedt particular degrees of obliquity ; as 
is evident in thofe of Monte Roffo, San Giovanni, and 
many others. Nor is even the horizontal pofition, though 
iefs common, to be excluded, as I have juft obferved; 
though I have never yet feen an entire group of columns 
fo difpofed. Such a group may, however, poffibly be 
found, among the great variety of limilar phaenomena, 
which thofe curious and moft interefting provinces of 
Auvergne and Velay prefent to our notice. The co- 
lumns of San Giovanni feem bedded in a kind of vul- 
canic faiid, which, in many parts of the hill, intirely 
covers them; nor do I recolledl whether any other fo- 
lid maftes appear, befides the columns : thefe, however,, 
probably reft at bottom upon a bafe of bafaltine rock of 
the fame nature. Nothing is more common, in the pro- 
vinces of France juft mentioned, than to fee ifolated ba- 
faltine hills almoft exclufively compofed of different lay- 
ers of columns, which prefent themfelves in ftages, one : 
above the other, often without any other Jlratum be- '■ 
tween them, refembJing, in Tome meafure,^ magna licet \ 
co7nponere ] 
