14 PROFESSOR FORBES ON THE VOLCANIC GEOLOGY OF THE VIVARAIS. 
to the plane of the stratum which they occupy, which is inclined about 9° to the 
horizon.* Some of the pillars are a little twisted in their upper parts, although 
quite straight below, a phenomenon which gives them an appearance of having 
yielded under the pressure of the load above them; this occurs a little to the east 
of Fingal’s Cave, and is correctly represented in one of Dr MacCuttocn’s views of 
the island. If we compare the lava-cliff of Jaujac with Staffa as a mere columnar 
display, we must give the former a decided preference. The columns are more 
numerous, more extended, higher, slenderer, better pointed, and, in every respect, 
more perfect. It is probable that this, as well as their incomparable superiority 
to any known product of altogether recent volcanoes, is due to the more perfect 
composition and fusion of the material of which they are formed. 
But this grand phenomenon might have been lost for ever to human sight 
had not the excavating action of the stream made the section which we have been 
contemplating ; and, is it not interesting to inquire in what manner the degrada- 
tion or destruction of so large a mass of intensely hard rock was effected, and 
whether it is going on still? Now, as to the present action of the elements, there 
are abundant proofs that it is going on with great energy. Just below Jaujac, a 
small stream, named Rioclat, joins the Alignon, and, opposite to their junction, a 
great mass of the principal lava-cliff fell about three weeks before my visit in 
1841. The scale of this operation, which is the removal of about 2000 cubic 
yards (40 yards in front, 3 from back to front, and 16 high), makes it very in- 
teresting, and its recent occurrence, as well as that of another great éboulement 
opposite Souillols, throws light upon the natural mode of proceeding. In both these 
cases the atmosphere seems to have acted alone. At Jaujac (the case represented 
in Plate VL., fig. 3), the river did not touch the base of the lava at all; we must, 
therefore, distinguish two methods of disintegration. 1st, The atmospheric water 
penetrating freely amongst the countless fissures which the imperfect columnar 
structure of the upper bed presents, detaches it gradually by the usual effect of 
liquid pressure, more rarely perhaps, by congelation. The principal tendency to 
fracture being exactly in a vertical direction, the cliff has a continual tendency to 
instability. A few hundred yards from the spot last described, there was, at the 
time of my visit, another huge mass in the process of separation from the cliff by 
a gap already two feet wide at the top. I passed below it during a tremendous 
thunder-storm, when the rain-water was gushing in torrents from the joints of 
the pillars, threatening instant precipitation. 2d, The lower part of the lava 
being always regularly columnar, and having, therefore, very little lateral cohe- 
sion, and probably, like a table with many legs, all the pillars not bearing equal 
shares of the load, the erosive action of the water must necessarily detach them ; 
and, we almost always, if not invariably, find that the upper part of the cliff pro- 
* A description of the Western Islands of Scotland, vol. ii. 
