EROSION OF LAVA—CRATER OF JAUJAC. 15 
jects, as in Plate II., fig. 6, and gravity thus assists the atmospheric causes of 
disintegration. 
To these I would add one consideration as to the commencement of erosion. 
When the form of a lava-bed confines the running water to the centre of a stream, 
as in fig. 7, we find that it acts extremely slowly ;* but if the river takes one side 
of a lava-bed, as in fig. 8, there being a crevice between the granite and the lava, 
the water must penetrate, and, by its pressure, tend to separate the columns, and 
to wash them out, so that I know of no existing case of water running under the 
condition last described. But portions of the lava often adhere to the granite 
even on the side of the valley most eroded. This is the case on the bank of the 
Alignon. 
The often excessive fragility of the lava also assists its division and removal. 
At the éboulement of Jaujac the basalt is singularly brittle, almost the whole 
fallen mass is shattered into bits of a few pounds weight. In texture it 
slightly resembles pitch-stone, and the lustre is that of animal glue. The frag- 
ments include many pieces of granite. At the junction of the Alignon and Rio- 
clat is a considerable mass of pure feldspathic granite, not like that of the country, 
surrounded by, and cased in basaltic columns, which have formed almost as re- 
gularly as if it had not been there. This, though very interesting, is conformable 
to the illustration which we gave of the law of direction of the columns. [If it 
depended upon the contact of the lava with the granite otherwise than as the 
cooling is thereby affected, the columns might be expected to radiate from the 
enclosed mass; but as any mass, not enormous, enveloped in such a stream, 
might acquire the temperature of the melted matter, the whole would cool with 
reference only to the conditions extraneous to it. The lava extends but little 
way up the Rioclat, which, however, presents a curious deposit of volcanic ashes, 
which no doubt must have fallen from the crater of Jaujac. 
The Coupe de Jaujac (being an exact translation of the word crater, by 
which the ancients denoted a volcanic orifice) is distant only half-a-mile or a 
mile from the village, in a natural opening or cavity between two primitive 
mountains, filled with coal-formation sandstone, whose character is well marked 
by abundant impressions of fossil plants. This formation (in which coal was at 
one time worked)} is interesting on several accounts, although the occurrence of 
the volcano in the midst of it cannot but be regarded as entirely accidental. Its 
extent is small, being nearly confined to the valley of Prades, adjoining that of 
the Alignon, at least if the Geological Map of France be correct. It is, as already 
observed, part of a widely-spread but often-interrupted ring of the coal-formation, 
surrounding the primitive plateau of Central France. Iexamined, very carefully, 
* Ewamples:—The tributary on the left of Ardéche near Thuez; lava of Burzet at the cascade 
near the village. 
} On the authority of Fausas Sr Fonp. 
