CHAPTER IV.

GENERAL ACCOUNT OF THE VOLCANIC FORMATIONS OCCURRING UPON THE ELEVATED GRANITIC PLATFORM OF CENTRAL FRANCE.

§ 1. It has been already mentioned that the volcanic formations of Central France attain an elevation much superior to that of the highest parts of the granitic platform. By many of their earliest observers, especially by M. de Montlosier, and subsequently by Dr. Daubeny, they were described as of two classes, ancient and modern, according as they appear to have been produced before or after some supposed epoch of a diluvial character, to which the excavation of the existing valleys of the district was attributed. My observations of the general features of the country in 1821 led me not merely to doubt, but to deny altogether, that there is any reason for referring the denudatory action to which its valleys are due to any single cataclysm or diluvial phenomenon. It appeared to me clear that this process has been going on from the first appearance of the land above the sea-that it is still in action, being chiefly occasioned by the decomposition and erosion of rocks by rain, frost, and other meteoric agents, but especially by the direct fall of rain from the sky, and the wash of the superficial waters, which everywhere and on every scale of force, as rills, streams, and rivers, are ceaselessly engaged in sapping and mining their banks, and carrying off the detritus to the plains, which they cover with alluvium. This every flood drives further, and grinds still finer, until it is ultimately carried out in the shape of sand or mud into the sea,

where it settles as a sedimentary deposit. I thought I saw ample proofs in the relative position of the plateaux of basalt and trachyte which are seen capping so many of the hills in Auvergne at various elevations, some more than a thousand feet above the level of the plain of the Limagne, others but slightly raised above its surface, or the alluvial bottoms of its tributary valleys, that the excavation of these valleys, as well as of the plain into which they descend, has been gradual from the earliest to the latest times, and accompanied throughout by occasional volcanic eruptions, chiefly from the neighbouring granitic heights, but sometimes from within the area of the lake-basin, that is, of the existing plain of the Limagne; and consequently that no clear chronological line of separation can be drawn between the ancient and modern volcanic products; although, no doubt, some are of a very remote antiquity as compared with others.

It seemed to me, viewing as a whole the entire district of Auvergne, the Velay, and the Vivarais, that its volcanic rocks divided themselves geographically into six distinct groups, viz.— first, the three mountain masses of the Mont Dore, the Cantal, and the Mezen, each of which raises its colossal bulk from the granitic platform to a height of about six thousand feet above the sea, and appears to have been a centre of repeated eruptions on the largest scale, giving birth to a volcanic mountain like Etna, Teneriffe, and other sites of habitually recurring eruptions; 2ndly, the products of more isolated vents of eruption which broke out at various periods, but by far the greater number since the quiescence of the habitual volcanos above mentioned, upon a zone running nearly north-west and south-east from the north-west of Riom to the neighbourhood of Aubenas on the Ardèche. Some rather wider breaks than usual between the points of eruption on this line induced me to divide this group also into three sections: 1, the chain of puys of the Monts Dôme; 2, that

of the Haute Loire; and 3, the cluster of volcanic vents of the Vivarais which have broken out in some tributary gorges of the Ardèche. To these indeed should be added a fourth and independent group, which I have not myself examined, and of which I have not succeeded in finding any detailed description. It occurs south of the Cantal, near La Guiole, and on a line parallel to the eastern zone already mentioned.

The geographical convenience of this division seems to have led to its adoption by later writers, both French and English: I shall therefore continue to employ it in the following pages. But for many reasons-especially that it forms the first in order of approach to visitors from Paris and the north, and the best introduction moreover to the other volcanic phenomena of the district-I shall begin with the description of the chain of puys (as they are called) of the Limagne and the Monts Dôme, which rise to the west of Clermont-Ferrand, the chief town of the department of the Puy de Dôme, and an excellent centre from which to explore the remarkable country around.

CHAPTER V.

FIRST VOLCANIC REGION.-MONTS DOME AND THE LIMAGNE

D'AUVERGNE.

THE Limagne d'Auvergne, as has been already noticed on the subject of its freshwater formation, is an extensive valley-plain, about twenty miles in breadth and forty in length; its soil, with the exception of the calcareous hills already described, being an alluvium consisting chiefly of boulders of granitic rocks, trachyte, and basalt, through which the Allier for the most part still wears its channel in a course from south to north. The inclination of the surface of the plain towards the river on either side, where not interrupted by hills, averages perhaps twenty feet in a mile. This at least is its slope from the base of the low hill on which Clermont stands, to the low-water mark at Pont du Château, the first point being 1204, the last 1027 English feet above the sea, and the distance about nine miles.

The western limit of the plain is formed by the abrupt escarpment of the granitic platform already described, which is fringed by some lower hills that branch off from it into the plain, and furrowed by steep and short ravines. These on being explored are found to penetrate to no great distance, terminating at the base of the range of volcanic hills, or "puys" as they are locally called, which rise from the otherwise nearly level plateau in a line nearly due north and south. On the western side of this chain of puys the platform slopes more gradually towards the valley of the Sioule, which runs nearly in a parallel direction. The width of this granitic table-land is about twelve

miles; its average elevation 2800 feet, being about 1600 above the plain of the Limagne; but on some points where it has been preserved from denudation by a capping of basalt it attains 3300 feet. On its western side it is composed chiefly of gneiss, but on the east of veined granite, in which transitions from a coarse to an extremely fine grain are very frequent. Much of this rock readily decomposes, and every storm washes away heaps of crystalline sand from its exposed surfaces. Upon this platform rises the "chain of puys," comprehending about seventy volcanic hills of various sizes, several of them being grouped together in immediate contact, in other cases a considerable interval intervening; the whole, together with the scoria and volcanic ashes which cover the plain around and between them, forming a notched and irregular ridge directed north and south, about twenty miles in length by two in breadth.*

With the exception of five (among which is the Puy de Dôme itself, the loftiest and most prominent of these hills), all of them are volcanic cones of eruption,† apparently of very recent production. Their height is from 500 to 1000 feet above their base. They are generally clothed with a coarse herbage or heather; some few with thick forests of beech, once much more abundant. This covering, however, does not hinder an examina

See Map of the Monts Dôme, and the profile sketches, in Plates I. and II.

† A volcanic "cone of eruption" in its normal form, with a "crater" or cup-shaped hollow at its summit, is the result of the accumulation round the volcanic orifice or vent of the scoriæ

and other fragmentary matters projected into the air by the series of explosive discharges of elastic vapour and gases which usually characterises an eruption. The fragments which fall back

into the vent are, of course, thrown up again and again, and triturated into gravelly sand or fine ashes by the friction attendant on this violent process. Those which fall on the outside of the vent are heaped up there in a circular bank, the sides of which, both within and without, slope at an angle rarely exceeding 33. And this bank, viewed externally, has of course the shape of a truncated cone, the crater being a hollow inverted cone contained within it.