8 PROFESSOR FORBES ON THE VOLCANIC GEOLOGY OF THE VIVARAIS. 
many miles to the south of Montpezat to form one considerable river,—the Ar- 
déche. which gives its name to the department,—it will be most convenient to 
suppose the traveller ascending the Ardéche from the town of Aubenas, one of 
the most considerable in the province, and arriving at a point called Pont de la 
Beaume (see the map), where the river Fontaulier, whose rise we have taken 
notice of, joins the Ardéche, having previously passed close to the village of 
Montpezat, and immediately under the volcano of the same name, subsequently 
receiving two minor tributaries, which we shall presently have to describe. 
From the neighbourhood of the Pont de la Beaume, the greater part of the 
valleys of which we are to speak diverge almost like the rays of a fan. As we 
look wp the course of the Ardéche, Montpezat occupies nearly the centre of the 
fan. Now, in all these valleys there is a remarkable uniformity of constitution, 
and, in some respects, of general appearance. The substratum of the whole isa 
primitive rock, granite, or in some places gneiss. The distinction is not very im- 
portant in connection with the phenomena which we are to describe; and I have 
not attempted to determine the limits of the more crystalline granites, as dis- 
tinguished from those whose slaty structure may allow them to be considered as 
having a regular cleavage and direction of beds, subordinate to which hornblende 
slate also occurs. When I speak, therefore, of granite forming the predominant 
rock (wncoloured on the map), I would not be understood to do so always with 
precise mineralogical accuracy. So far as I know, there is no peculiarity in the 
volcanic action in the granitic districts, compared with that in gneiss. A small 
patch of the coal-formation appears near Jaujac on the Alignon, but it is sur- 
rounded by granite or gneiss, which again is succeeded by the lias or oolite for- 
mations near Aubenas. The coal-formation occurs in patches nearly all round 
the great primitive plateau of central France. It is extensively worked on the 
east side at St Ettienne; and, in some places at least, its strata lie horizontally 
against the granite, shewing the anterior date of the elevation of the latter.* 
The valleys we have to describe farther agree in this extraordinary particular,— 
that, as surely as they contain water they contain a stream of lava or basalt, or 
the remains of one, which stream has accommodated itself perfectly to the 
sinuosities of the channel of primitive rock in which it has run, the possession 
of which it contests yard by yard with the water; these lava streams are 
sometimes attenuated to a surprising degree, leaving but small relics for the 
space of miles; in other places they accumulate to an astonishing thickness and 
breadth, altering the configuration of the valley, the stagnant pool of lava hay- 
ing, in the first instance, created a lake of water, and compelling the river to 
alter its course and to excavate a new channel. ‘The tributary of each valley 
commonly unites with others, accumulating near the points of junction; but the 
* Burat, p. 4. 
