It is further to be noticed, that many of the craters of ancient volcanoes, are of far greater size than the present ones. Vesuvius is a comparatively small cone, raised within the crater of a larger volcano. The cone of the Peak of Teneriffe, according to the descrip tion of travellers, stands within a volcanic plain, containing twelve square leagues of surface, surrounded by perpendicular precipices and mountains, which were the border of the ancient crater. If the opinion of M. Humboldt be correct, all these craters are diminutive apertures, compared with the immense chasms through which, in remote ages, subterranean fire has forced a passage through the crust of the globe.

"The whole of the mountainous parts of Quito," he says, " may be considered as one immense volcano, occupying more than seven hundred square leagues of surface, and throwing out flames by different cones, known by the denominations of Cotopaxi, Tungurahua, and Pichincha. In like manner," he adds, "the whole group of the Canary Islands is placed as it were on one submarine volcano. The fire forces a passage sometimes through one, and sometimes through another of these islands. Teneriffe alone, contains in its centre an immense pyramid terminated by a crater, throwing out, from one century to another, lava by its flanks. In the other Canary Islands, the different eruptions take place in various parts, and we no where find those isolated mountains, to which volcanic effects are restrained. The basaltic crust formed by ancient volcanoes seems every where undermined; and the currents of lava seen at Lanzerote and Palma remind us," he adds, " by every geological affinity, of the eruption which took place in 1301 at the Isle of Ischia, amid the tufas of Epimeo."

In the preceding part of the present chapter, I have endeavoured to give a succinct account of the most important volcanic phenomena. The only formations of hard crystalline rocks in the present day are volcanic; and if we trace the connection that exists between modern and ancient volcanic rocks, and between the latter and the rocks of trap and porphyry, among the ancient rock-formations, we shall extend the dominion of Pluto over a large portion of the globe.

Many of the ancient volcanic rocks, have not flowed in currents from limited apertures, like modern lavas. "The volcanic porphyries on the back of the Cordilleras," says M. Humboldt, “are undoubtedly of igneous origin; but the mode of their formation is not like that of modern lavas, which have been erupted since the excavation of valleys. The action of volcanic fire by an isolated cone or crater of a modern volcano, differs necessarily from the action of this fire, through the fractured crust of the globe." It has been observed by the same geologist, that the further back we can trace volcanic eruptions, the greater is the similarity between their products, and the rocks which are regarded as the most ancient ;-hence, the countries that have been the seats of ancient volcanoes, are particu

Jarly interesting to the geologist. In Auvergne, and the more southern parts of France, there are èxtinct volcanoes of different ages, covering with their products several thousand square miles. The most recent of these volcanoes has been extinct or dormant since the records of authentic history, and probably for a longer period. Julius Cæsar, who was encamped on this volcanic soil, and has described the country, makes no allusion to its having been the seat of active volcanoes.*

West of the town of Clermont, there is an extensive granitic plain, rising about sixteen hundred feet above the level of the river Allier. On this plain there are numerous cones, and dome-shaped hills, varying in height from twelve hundred to two thousand feet; some of these cones have well-preserved craters, and the cones themselves are formed chiefly of scoriaceous lava. These are the most recent volcanoes of that country; their products differ in no respect from those of modern volcanoes, except that the lava may often be observed passing to the state of compact basalt, exactly similar to many of the basaltic rocks in Great Britain. That these volcanoes are the most recent, is proved by the lava flowing down from them into the present valleys; and hence we are certain, that the eruptions must have taken place subsequently to the excavation of the valleys. There are other currents of lava from more ancient volcanoes, that have flowed before the valleys were excavated, and form isolated caps on the hills that enclose the present valleys. These currents of lava are composed chiefly of compact basalt: the position of these isolated caps of basalt is similar to that on the hill b, (Plate III. fig. 2.) but they are not always columnar. The openings from which these beds of basalt have flowed cannot be always traced; but as we can observe the change from scoriaceous lava to basalt in the currents of undoubted lava, we cannot hesitate to admit, that the basalt which forms these caps, must have had a similar origin. Under the caps of basalt, there are in many situations thick beds of volcanic tufa, containing bitumen, which will be subsequently noticed. Beside the volcanoes with craters, that have ejected currents of scoriaceous lava and basalt, and poured them into the valleys; and beside the more ancient volcanoes, that have formed beds of basalt before the excavation of the valleys, there are other volcanic mountains, which have rounded summits or domes, without any perforation or crater, and these are composed chiefly of whitish or grey earthy felspar, containing imbedded crystals of felspar: to this rock the name of trachyte has been given, on account of its rough fracture. It may be properly called a volcanic porphyry.

* I visited the extinct volcanoes of France in the spring of 1822, and published an account of them in the 2d volume of my Travels, accompanied with cuts, and a section and outline of the country round Clermont, which is, I believe, the first attempt to render in this manner the structure of this volcanic district intelligible to the general reader. Without the aid of sections and diagrams, it is difficult to obtain a distinct notion of the relative position of the different volcanic formations.

The first, or more recent volcanoes, resemble in every particular, the existing volcanoes in various parts of the world; and the currents of lava may be traced from their sides along the granitic plain on which the volcanoes stand, and thence into the adjacent valleys for many miles. The lava appears as fresh as the recent lavas from Vesuvius, though it has been exposed to the action of the atmosphere for some thousand years. The Puy de Pariou is the most perfect of these volcanic cones. The following description of it is taken from the 2d volume of my Travels:-"We were one hour in going from La Barraque, a mountain village, to the foot of the Puy de Pariou, where we left our char, and another hour in ascending to the summit, as we halted several times to rest. As nearly as I could estimate, the summit of this mountain rises about one thousand feet above the plain, and is, therefore, about three thousand eight hundred feet above the level of the sea. The crater, which is the best preserved of any in Auvergne, is nearly circular. I walked round it, and its circumference is about eight hundred yards. Its shape is that of an inverted cone or funnel quite perfect. The edge or rim of the crater is narrow, from which the descent or slope is very rapid on each side the depth of the crater from the highest part of the edge (which is on the southern side) to the small plain at the bottom, may be about three hundred and twenty feet; and from the western side, about two hundred and sixty English feet. The lava which flowed from Pariou to La Barraque, and thence towards the plain of Clermont, is generally supposed to have issued from the crater; but had this been the case, the crater would not have been so entire as it is; and I am fully convinced, that the eruption of such a mass of lava must have broken down one of the sides, as at Nugere, which we afterwards visited, and the Puy de Vache. There appears, I think, decisive marks of the lava having flowed from an opening on the north-east side of the mountain, to which it may be traced. Indeed on this side there are the indications of a much larger crater, which has its escarpments turned towards the Puy de Pariou like those of Mount Somma, which are turned towards Vesuvius. The Puy de Pariou was, in all probability, a volcanic cone, formed within the larger crater by its last eruption of scoriæ.

"The annexed cut, from a drawing I made near the foot of the mountain, represents the external shape of the Puy de Pariou, and the dotted lines show the form and the relative depth of the crater, the bottom of which, a a, is about three hundred and twenty feet below the highest part of the rim c. The current of lava, b b, is on the north-east side of the present mountain. The internal shape of Pariou approaches to quadrilateral, or is that of a cone compressed on each side, and somewhat elongated from north to south. The bottom of the crater is nearly flat; there was a little water, from the recent melting of the snow, remaining in some of the hollows: indeed we were told at Clermont that we should find the crater filled

with snow. It was early in May: but the snow was gone, and grass was growing in some parts; others were covered with loose masses of scoriæ. Owing to the great porosity of the soil, the crater of Pariou seems doomed to perpetual sterility, there is no tree or shrub within it; while that of Vesuvius, after a cessation of eruptions for only four centuries, was covered with large chestnut trees."-Vol. ii. page 307.

In the Puy de Pariou, and many other volcanic mountains of this district, there is nothing particularly remarkable, except, that the lavas which have flowed from them at a remote period, should preserve all the freshness of recent lavas, and that volcanoes so well characterised, both by their forms and mineral products, should have remained unnoticed until the middle of the last century. The round topped or dome shaped hill on the left of the Puy de Pariou is called Sarcoui; it belongs to that class of volcanoes that have no craters, which will subsequently be noticed. The more ancient volcanoes, that have poured out the thick beds of basalt that cap many of the valleys round Clermont, cannot always be traced, as the openings from whence it issued may be covered by the lava of more recent eruptions. In order to obtain a more distinct idea of the position of these caps of basalt, it will be necessary to remark, that the granitic plain above Clermont, and the hollows or valleys in its sides, received their present form, prior to the most ancient volcanic eruptions; these hollows or ancient valleys, were probably basins or lakes, in which were deposited a vast thickness of calcareous strata, containing fresh water shells, and the bones of land quadrupeds. Into these lakes, there has flowed a vast mass of volcanic tufa, covering the limestone, and sometimes intermixed with it. The volcanic tufa, and the fresh water strata, appear to have filled up the ancient valleys or lakes; and on this tufa, the basalt was deposited by a subsequent eruption. At a later period, diluvial currents have furrowed excavations or new valleys in the basalt, in the subjacent tufa, and in the fresh water limestone, leaving detached portions or hills composed of basalt, tufa and limestone, which once were parts of continuous beds. Into these new valleys, the lava of the most recent volcanoes has flowed. The most remarkable circumstance attending these more ancient eruptions, is the bituminous nature of the tufa, which forms the lowest bed, and covers the fresh water limestone of Gergovia, Canturges and the neighboring hills. This tufa is, in some parts, more than three hundred feet thick; it consists of earthy basalt or wacke, intermixed with lumps of scoriæ and basalt, and in some places, with limestone: it is every where impregnated with bitumen. The tufa of Auvergne bears evident marks of being the product of an aqueous or muddy eruption, intermixed with lava and scoriæ, which increase in quantity in the upper part of the mass, and at length cover it with compact lava or basalt. That the tufa was ejected in an aqueous or muddy state is proved, by the quantity