Di- Daubeiiy on the Columnar Structure of Trdp-Rochs. 65 
valley which bounds its progress, I intend to trouble the Society 
with a few remarks. 
It is on the aboye fact, indeed, that M. Faujas ha& founded 
an opinion as to the cause of the prismatic arrangement which 
basaltic rocks exhibit ; for, as in the narrow and steep defiles 
of the Vivarais, the bottom of the valleys is usually occupied 
by a running stream, he conceives that the columnar structure 
developed in the lava in these situations, arises from the sudden 
congelation produced among the particles of the liquified mass, 
by falling into water, and seems disposed to extend the Same 
hypothesis to basaltic rocks in general. This theory, however, 
appears not only quite untenable as applied to basaltic rocks in 
general, but even very questionable when confined to the case 
we are at present considering; Sudden congelation, so far 
from being favourable, as M. Faujas supposes, to a regular ar- 
rangement among the particles of matter, proves, so far as we 
know, almost destructive of it. 
Without entering, however, into the merits of this theory, 
which I conceive will find but few supporters, it will be suffi- 
cient for my present purpose to remark, that the columnar struct 
ture of the lava found in the bottom of tlie valleys in the Vi- 
varais, must have been exposed by some process subsequent to 
its original formation, for the bed of the existing rivulet, which 
exhibits so many beautiful examples of columnar basalt, ap=- 
pears in every ease that I have seen, to have been scooped out 
subsequently to the flowing of the lava * ; for the latter must 
have blocked up the course of the original stream, if any such 
existed at that period, as appears from the fact of the lava ex- 
* The abtupt face which the beds of the tivulots pteseht ori eithOr side, com- 
|)ared with the more gradual slope of the valley above, seems to afford a beautiful 
illustration of the difference bet\^een the effect produced by streams now existing, 
and by the retiring of a body of waters considerable enough to cover the whoSe 
face of the country. On this account j the Vivarais shotdd be visited by those who 
feel interested in the theory that has been so much insisted upon, with respect to 
the origin of a great proportion of our preseiit valleys from the Mosaic Deluge, or 
at least from the action of a body of waters equally considerable. Possibly, too, 
an accurate examination of the depth of lava cut through in these instances by 
the existing streams, might lead to some inferences as to the age at which th©: 
lavas must themselves have flowed. 
VOL. VII. 2^0. 13. JULY E 
