Thus though quartz is wanting in trachyte, and abundant in granite, yet the siliceous earth contained in that mineral may have united with the alumina present, in such proportions as would form felspar, and in this manner perhaps the latter has become more abundant, at the expense of the other two ingredients of granite.

The fusion effected by the volcanic operations would be favourable to the play of affinities, and enable the particles of the silica freely to combine with the other earths in the requisite proportions.

In some cases on the contrary, where the material operated upon consisted chiefly of quartz, the result may have been that variety called millstone trachyte, which, though chiefly siliceous, betrays its igneous origin by the cells and cavities it so abundantly contains.

We have thus arrived at the conclusion, that the characters of volcanic products in general are such as lead to a fair presumption, that they are derived from some of the older rock formations, a fact fully confirmed by a consideration of the phænomena attendant on an eruption, the general tenor of which plainly denotes, that the focus of the action is situated at a depth at least as great as that to which granite extends.

I do not lay any stress on the remarks of Stukeley, who calculates from the compass of country over which earthquakes have been felt, that the force must in some instances be 200 miles beneath the surface,* because we have reason to believe, that the vibrations may be propagated laterally far beyond the immediate influence of the impelling force; but I would argue from the immense mass of materials ejected by Vesuvius or Etna, without exhausting itself, or causing any diminution in its own dimensions; from the prodigious height to which the trachytic nucleus of a volcano

As in the one that occurred in Asia Minor A. D. 17, which destroyed thirteen cities, and extended over a diameter of 300 miles. See Stukeley on the Cause of Earthquakes. Phil. Tranact. for 1750.

is raised, as at Teneriffe, and in Equinoxial America; and lastly from the immense violence of the eruptions, which would shiver to atoms any superficial covering of rock, that the elastic vapours must be disengaged at a depth at least as great as that, to which the crust of the earth can be supposed to extend.

These considerations will be viewed as more favourable to the hypothesis suggested by Sir H. Davy's discoveries, than to any other perhaps that has been proposed. Thus it has been shewn, that volcanos usually take place in situations, in which the element calculated to excite the combustion was largely present; that the aeriform fluids given out are such as would be generated by the chemical action superinduced by its presence; that the nucleus, as well as the products of a volcano, are of a nature likely to result from the action of heat upon the constituents of the nearest rocks we know of to the seat of the action; that the character of the unaltered masses ejected favour such an opinion; and finally, that the phænomena themselves indicate a cause at once deeply seated, and of wide extent.

All these circumstances on the contrary are opposed to the theory, which attributes volcanos to the combustion of beds of coal or sulphur, for though these substances may be often present in the neighbourhood of burning mountains, yet the rock, in which they are imbedded, belongs, as I have shewn, to a comparatively recent epoch in the history of our planet, and the phænomena, which they exhibit when in a state of inflammation, are such as denote a local and superficial origin.

Conceiving therefore that the former hypothesis affords the more plausible explanation of the facts detailed, it may be worth while to recapitulate the substance of the foregoing remarks, in such a manner as may enable me to point out the connection of the several phænomena one with another, and their dependence upon the cause assigned.

Let us suppose, that the nucleus of the earth at a depth of three or four miles, either consists of, or contains as a

constituent part, combinations of the alkaline and earthy metalloids, as well as of iron and the more common metals, with sulphur and possibly with carbon. These sulphurets are gradually undergoing decomposition, wherever they come into contact with air and water, but, defended by the crust of the globe, just as a mass of potassium is by a coat of its own oxide when preserved in a dry place, the action goes on too slowly to produce any striking effect, unless the latter of these agents be present in sufficient quantity. Hence under our continents, the elastic fluids generated by this process are compressed by the superincumbent mass of rock, until they enter probably into new combinations, or diffuse themselves through the solid strata.*

But under the sea, where the pressure of an enormous column of water assists in forcing that fluid through the minutest crevices in the rock, the action must go on more rapidly, and the effects consequently be of a more striking

nature.

These effects however will take place in the middle of the sea less generally than on the coast, because the pressure of the ocean itself opposes an impediment; and it will in general not be constant, but intermittent, because the heat generated by the process itself will have a tendency to close the aperture by which the water entered, first, by injecting the fluid lava into the fissure, and secondly, by causing a general expansion of the rock; nor will the water again find admission, until, owing to the cessation of the process, the rock becomes cool, and consequently again contracts to its original dimensions.

Now the first effect of the action of water upon the alkaline and earthy metalloids will be the production of a large

* Carbonic acid is known to be very commonly present in the waters of springs, and, as we are not aware that they attract it from the atmosphere, it seems most probable that it is the result of some process going on in the interior of the earth. (Bischoff, Vulkanischen Mineralquellen, p. 271.) The same remark perhaps may apply to the exhalations of nitrogen gas, which Dr. Davy detected in the warm springs of Ceylon, (see his Travels, p. 45.) and which have been also found in some parts of North America.

volume of hydrogen gas, which, if air be present, will combine with oxygen and return to the state of water, if it be absent, will probably combine with the sulphur, both being at the high temperature favourable to their union. In the former case nitrogen gas will be given off, in the latter sulphuretted hydrogen.

But in case of the presence of oxygen, the sulphur will also become inflamed, and give rise to the production of sulphurous acid, which will predominate among the gaseous exhalations emitted from the mouth of the volcano, provided sufficient quantity of air be present to combine with the hydrogen and re-convert it into water. So soon however as the oxygen is consumed, the hydrogen, no longer entering into combustion, unites with the heated sulphur, and escapes in the form of sulphuretted hydrogen, which, towards the latter period of the eruption, will predominate over the sulphurous acid, because it continues to be formed long after the want of oxygen has put a stop to the production of sulphurous acid. Now it is well known, that these two gases mutually decompose each other, and therefore cannot exist at the same time, so that the appearance of sulphuretted hydrogen from the mouth of the volcano may indicate, if not the entire absence of sulphurous acid at the place at which the process takes place, at least that its formation is stopped by the consumption of oxygen, or is going on with less energy than heretofore.

The very circumstance of the reproduction of water by the mutual decomposition of these two gases, might be the means of keeping up the action in a languid manner for an indefinite period. The slowness with which lava cools would cause it to give out for a considerable time sufficient heat to the adjoining strata, to place the sulphur at the temperature necessary to cause its combination with oxygen; hence a certain portion of sulphurous acid would be continually emitted, which however would be soon decomposed by the hepatic gas present. The water resulting from this process would percolate into the recesses of the

rock, act upon any portions of the alkaline and earthy metalloids that might have escaped the original action, and give birth to a fresh volume of hydrogen gas, ready in its turn to dissolve a new portion of sulphur, and thereby to contribute to the repetition of the same phænomena.

The separation of muriatic acid from the common salt present in sea water is explained, on the common principles of chemistry, by the superior affinity exerted by the base for the siliceous or aluminous earth than for the acid, and the sublimation of iron in the state of fer oligiste, rather than of peroxide, may result from the deoxydizing property of the sulphuretted hydrogen at the same time disengaged. The carbonic acid given off may be derived either from the carbonaceous matters that have entered into combustion, a view of the subject which is perhaps favoured by the phænomena of the pietra mala, or from the action of the high temperature upon the carbonates of lime and magnesia, existing in the strata above the seat of the volcanic action. I have already remarked, that this latter gas is chiefly found in volcanos that have become extinct, or have been long in activity, where time appears to have been given for the heat to extend itself beyond the immediate sphere of the volcanic action.

In short, on the supposition of salt water and air being brought in contact with the sulphurets of the metals and earthy metalloids, all the known phænomena of volcanos may be deduced in the order in which they appear to occur: in the first place, so long as air was present, an evolution of large volumes of muriatic, sulphurous, and nitrogen gases, together with aqueous vapour, would take place; at a later period, when the oxygen was expended, sulphuretted hydrogen and carbonic acid, with a smaller quantity of muriatic acid, would appear; lastly, when all the other effects had subsided, aqueous vapour and carbonic acid might continue to be evolved.

If it be asked, how we can account for the presence of atmospheric air in the interior of a volcano, I answer, that