This felspar is derived immediately from trachyte, that being the rock which directly surrounds the focus of the volcanic action; for if we examine the strata that successively present themselves on the sides of a crater, we are sure to find that the lowest in the series is trachyte, from which is derived by fusion the obsidian, as is the case at Teneriffe.

Agreeably with this view, I have myself remarked, that the oldest lavas of Mount Etna approached most nearly to the characters of trachyte, and that there even seemed a gradation, dependent upon the relative antiquity of the beds, down to the lavas of the present period, which possess the usual cellular and semivitreous aspect of such products.

But together with these, which may be viewed in the light of essential ingredients, lava often contains augite, leucite, hornblende, mica, olivine, specular iron, and other minerals, which appear not to have been present in the matrix, but to have resulted from the play of affinities induced amongst its ingredients, in consequence of the fusion which they had undergone.

Whether therefore we look to the original nucleus of a volcanic mountain, or to the products that have at successive periods been derived from it, we shall be led to the same conclusion, namely, that the substance on which the fire had acted was a felspathic rock allied to trachyte, and that the bodies ejected during an eruption may be regarded either as modifications of it, or as extraneous substances to which the action had accidently penetrated.

To the former class perhaps belong lavas in general, as well as pumice, obsidian, and the like; to the latter the marbles and other substances not volcanic, that lie scattered round the sides of many volcanos.

But considering the peculiar characters of trachyte, and the circumstance of its being limited to countries that appear to have undergone the action of heat, we can hardly regard it as a substance which makes a part of the original constitution of the globe, and shall be disposed to set it down as

being itself a product, although a primary one, of the fusion of other descriptions of rock.

In order to determine this question, I propose to consider, 1st, amongst what formations volcanos commonly break out, and 2dly, what is the nature of those loose masses, which have been ejected without losing altogether their original characters, and therefore afford evidence of the kind of rock amongst which the volcanic action resides.

To begin with the Rhine, it appears that the formation on which the trachyte of the Siebengebirge rests, and among which the volcanos of the Eyfel have arisen, is a clay slate belonging to the transition series; in Auvergne, the rocks of Mont Dor and of Clermont rest immediately on granite, or are separated from it only by a tertiary deposit, whilst those in Cantal are incumbent on mica slate. In Hungary and Transylvania, the rock underneath is a porphyry, associated with syenite, clay-slate, &c. and referred by Beudant to the transition series; whilst in Styria, the rock most immediately surrounding the little trachytic formation of the Gleichenburg is gneiss.

In Italy the case is somewhat different yet though the trachyte of the Euganean Hills rises from beneath chalk, we have reason to believe that primitive and transition rocks lie at no great depth beneath, as they are found near Schio, and support the alternations of Volcanic and Neptunian deposits in the Braganza.

To Vesuvius, the nearest rock formation is the limestone of the Appennines, but gneiss and clay slate seem to be the substratum through which the trachyte of Mount Etna has been protruded.

Humboldt has shewn, that the rock which supports the volcanos of the New World is generally a transition porphyry, and sometimes gránite or syenite, and Von Buch reports, that the last mentioned rocks appeared as the lowest of those uplifted strata, which surrounded the crater of the Isle of Palma, and other of the Canaries.

'Now, although the preceding enumeration indicates such a variety with regard to the position of volcanic formations, as may seem at first sight to baffle all general conclusions, yet when we consider, that in the majority of instances, the rocks have been referred either to the primitive or transition series, and that in the remaining ones, the latter were at a depth far less considerable, than that at which we shall afterwards find reason to conclude the volcanic force itself to reside, I think it may not unfairly be presumed, that volcanos have universally broken out amongst the older formations, or those most near to the nucleus, whatever it may be, of the globe.

It is obvious indeed, that in those cases in which volcanos have appeared in the midst of primitive rocks, we cannot presume the seat of action to reside amongst those of a later date, but that the reverse does not hold good; so that if we only admit that any certain position is to be assigned to these products, a single case of their occurrence in the midst of older formations would overturn every inference, to be derived from their being observed to emanate from strata of a more recent date.

This presumption is farther strengthened, by considering the nature of the substances, found in the midst of lavas, which preserve any traces of their original characters, or the loose masses of unaltered rocks, that are occasionally thrown out.

Amongst the former, I have never seen or heard described any substance that bore the slightest resemblance to the constituents of secondary strata, but have often observed imbedded portions which present the appearance of altered granitic rocks.

Thus at the Puy Chopine in Auvergne, granite is found intermingled with the trachyte and greenstone, thrown together in such confusion, that we imagine the whole to have been elevated at one time, before the rock had been entirely acted upon by the heat.

In the lavas of the Vivarais, as well as in those of the

Rhine, I have met with masses imbedded that have much the appearance of an altered gneiss or granite, and the same thing is also seen in the tuff of Gleichenburg in Styria.

Humboldt mentions having found, in the midst of the new volcano of Jorullo in Mexico, white angular fragments of syenite composed of a small portion of hornblende and much lamellar felspar; and in the collection of Dr. Thomson, now in the Museum of Edinburgh, there is said to be a fragment of lava enclosing a real granite, which is composed of reddish felspar with a pearly lustre like adularia, of quartz, mica, hornblende, and lazulite.

I have likewise seen among the specimens from the Ponza Islands, presented to the Geological Society by Mr. Scrope, a piece of granite, or perhaps rather of a syenitic rock, which he states, in the annexed catalogue, to have been found in the midst of the trachyte from this locality.

But the most interesting fact perhaps of this description, is one related on the authority of a zealous cultivator of natural history, Signor Gemellaro of Catania,—I mean the presence of a mass of granite containing tin-stone, enveloped in the midst of a stream of lava from Mount Etna. The specimen I have myself seen, and though I cannot pretend to have ascertained the presence of tin-stone, am able to vouch for the general accuracy of the account he has published of the substance.

T

It may be remarked, that these specimens of granitic rocks have, in general, a degree of brittleness, which accords very well with the notion of their exposure to fire.

The general character of the ejected masses, which are not imbedded in lava, is such, that it is difficult to refer them to any rock with which we are acquainted; it is true that M. Poli of Naples, has in his possession a fragment of rock thrown out by Vesuvius, which Humboldt pronounces to be a true mica slate, but in most cases they are mixtures of nepheline, mica, augite, leucite, and other minerals, which are rarely found associated together in the same manner in any of the original rock formations,

I have already alluded leucite crystals are cotemporaneous with the rocks in which they occur, and may add in addition to the fact before stated in support of this opinion, that Von Buch observed at Borghetto, in the midst of the crystals, a black substance analogous to the rock in which the whole was imbedded, and that in some cases, where the leucite did not completely envelope it, a connection was observable between the black substance and the rock. Sometimes in the place of the former a crystal of augite occupied the centre of the leucitic mass.

to the probability, that the

Admitting then these crystals to be cotemporaneous with the rock, and the whole to have been formed by the volcano out of the materials submitted to its action, it seems most probable that the latter, if derived from any known rocks, should have belonged to the primitive or at least the transition, rather than to any more recent order of formations.

Independently of these considerations, the general characters of trachyte favour the idea of its derivation from granite, or some analogous substance; it is from this resemblance that it has been called by Dolomieu granitoid lava,* and, although the two rocks may be distinguished by the presence of quartz in the one, and its absence in the other, yet the predominance of felspar in both seems to place them under the same genus, and to distinguish them from the constituents of secondary strata, where that mineral hardly can be said to occur, except where we have reason to suspect the agency of fire.

It would be interesting to possess a comparative analysis of a series of specimens of trachyte and granite, taken as nearly as possible from the same localities, since notwithstanding the mineralogical differences above pointed out, there is reason to believe that their chemical composition may vary but little.

* See page 124.

+ Hence Strange mistook the trachyte of the Euganean Hills for granite. See his Paper in the Phil. Trans.