On the contrary side of Murat is a segment apparently of the same bed, in which the columns are subdivided by frequent joints; the separate articulations fitting into each other by means of alternately concave and convex bases, having occasionally small wedge-shaped processes arising from the odd angles of one prism, and closing over corresponding oblique truncations in the other.

*

Both above and below Murat, in the valley of the Alagnon, patches of the lacustrine limestone are quarried for use beneath the volcanic rocks, proving the freshwater formation to have extended to the east of the central vent of the volcano. But its chief bulk is seen on the opposite or western flank, where in the vicinity of Aurillac a thick bed of breccia, alternately supported, covered, and penetrated by currents of basalt and trachyte, rests upon the siliceous marls and limestones, the strata of which are in many spots twisted and dislocated, as if they were still unconsolidated at the moment of their invasion by the volcanic torrents. Occasionally there are appearances of a subsequent deposition of limestone in confused strata moulded upon the rude surfaces of the volcanic beds. This may be observed between Polminhac and Yolet in the valley of the Cèr. In general, however, the line of junction of the lacustrine formation and superimposed volcanic products is better defined; it may be followed almost uninterruptedly from Vic en Carladèz to Aurillac. The currents both of basalt and trachyte occasionally send out thick vertical veins or dykes into the limestone below; but neither along the walls of these, where the contact of the two substances

*It is said that a sufficient quantity of gold-dust was less than a century ago contained in the sand of the river Jourdanne to repay the labour of sifting and washing; and tradition asserts that the town of Aurillac situated on this

river derives from hence its name: qu. Auri lacus? The trachytes of Mexico are auriferous; there is nothing therefore improbable in the idea that those of France may be so likewise.-See Brieude, Top. Médicale de l'Auvergne, 1782-83.

is immediate, nor elsewhere, could I trace any material alteration in the texture of the limestone. In frequent instances calcareous masses, as well as smaller fragments and even some freshwater shells, may be seen enveloped by the lava, but they still effervesce briskly with acids, and, except in a somewhat dusky tinge, seldom appear at all affected.

The lacustrine beds are found at a higher level, by several hundred feet, on the east side of the mountain, than on the west, which led Mr. Raulin* to suppose the existence of a fault traversing the central heights of the mountain, and dating its formation from the first outbreak of the volcano-a supposition by no means improbable.

Without the limits of the tertiary formation, the primary base emerges from beneath the volcanic products on every side of the Cantal. The most elevated point at which I have observed it is near the village of Thièzac in the valley of the Cèr, between five and six miles from the supposed crater, where a rock of gneiss, certainly in situ, pierces through the calcareous strata as well as the overlying trachytic breccia, and disappears again immediately.

We have no data by which to determine the relative ages of the volcanic remains of the Mont Dore and Cantal. Appearances would lead to the conjecture that these volcanos were occasionally in action at the same period.

§ 2. CANTON D'AUBRAC.

The small mountainous district of Aubrac, which lies between the three towns of La Guiolle, St. Geneis, and St. Arcize, in the department of Aveyron, is for the most part covered by massive beds of basalt, which belong, I believe, to a centre of eruption

*Bulletin xiv., p. 174.

independent of that of the Cantal; but having had no opportunity of observing this group otherwise than from the summit of the latter mountain, I am unacquainted with the true limits and disposition of its volcanic productions. I am, however, assured by M. Le Coq, who has visited it, that the basalt has been evidently erupted from the gneiss and mica-schist on the spot, and that no traces of any tertiary or other calcareous strata are to be seen there.

CHAPTER VIII.

REGION IV. DEPARTMENTS OF THE HAUTE LOIRE AND ARDÈCHE.

§ 1.

THE basaltic ramifications of the Cantal seem to have been impeded in their progress to the east and south-east by the high primary range of La Margeride, although some detached portions are found upon its surface. To the north of this range they extend nearly to the banks of the Allier in the direction of Lempde and Brioude. On crossing the river and penetrating further into the department of the Haute Loire by the high road to its capital, Le Puy en Velay, we first traverse a barren tract of granitic soil, appearing to have been much devastated by torrents descending from the heights of La Chaise Dieu, and

[* However foreign to my subject, I must be allowed to mention, as an object of curiosity but little known, from this road being unfrequented by curious travellers, the bridge over the Allier at Vielle Brioude, one of the noblest and most beautiful works of the sort in Europe. It consists of a single stone arch, with a span of 192 feet, and rising 90 from the river. It is perfectly level at top, and, though narrow, admits of the largest carriages. The piers that sustain this magnificent arch abut on either side against rocks of mica-schist which here encase the Allier for some distance. It was built as early as the 15th century; but it is only within a few years and under Napoleon that it has been widened to admit vehicles of all sizes. The airy lightness and elegance

of the arch as seen from below are not perhaps to be paralleled by any other similar piece of masonry.] In the year after this note was penned (1824), this bridge fell into the river, being probably weakened at the time the addition above referred to was made. Another equally handsome, and it is to be hoped more durable one, has since been built between the original piers, and of the same dimensions. The Allier, taking its rise in the highest range of the Lozère, is liable to floods of extreme violence. One of the most terrific ever recorded has just occurred (October, 1857), by which the works of the railway bridge, nearly completed at this spot, have been swept away, and a vast amount of other damage done.

immediately find ourselves surrounded by volcanic remains of a different character to those we left behind on the other side of the river, and evidently foreign to that focus of eruption. The Allier may, therefore, in a general way, be established as the natural boundary of the volcanic districts of the Cantal and Haute Loire.

The volcanic remains of the departments of Haute Loire and Ardèche, or, in other words, of the ci-devant provinces the Velay and Vivarais, belong to both the classes which have been distinguished above, and accordingly arrange themselves under these separate heads:

1. The Mont Mezen and its dependencies.

2. The products of more recent eruptions, which have burst out on numerous points irregularly scattered over a broad band of the primary plateau, from Paulhaguet and Alègre to Pradelle and Aubenas, seeming to be a prolongation in a S.E. direction of the chain of recent volcanos already described in the department of the Puy de Dôme.

§ 2. THE MONT MEZEN AND ITS DEPENDENCIES.

The Mont Mezen is the most elevated of an extensive system. of volcanic rocks, resting partly on granite or gneiss, and in part on the Jurassic formation, which by their position and constitution prove themselves to be the remains of a single and powerful volcano, of the same character as those which have been already described in the Mont Dore and Cantal. Its products, however, are disposed in a somewhat different manner, being spread over an almost equally extensive surface without accumulating into such mountainous masses around their centre of eruption. Two causes seem to have contributed to occasion this diversity of aspect, namely: first, that the eruptions of this volcano appear to have been less frequent