A Treatise on Geology/Chapter 9

by John Phillips (1800-1874)

CHAP IX.

MODERN EFFECTS OF HEAT IN THE GLOBE.

To know the temperature of the interior parts of the globe at the present period, and the effects depending on its condition in this respect, is important, as furnishing one, and that, perhaps, the most instructive, of the elements for computing the changes which have, in earlier times, affected its structure and configuration, and varied its adaptations for organic life. By combining such knowledge of the subterranean parts of the earth as they now are, with inferences concerning more ancient periods, we are to seek the laws of action and variation of terrestrial heat, and, with the help of chemical and mechanical philosophy, to arrive at a general contemplation or "theory" of this part of geological science. Once well established, such a "theory" will be fertile of deductions bearing on all the known phenomena of organic and inorganic action: the recorded facts of geology form, on the other hand, a parallel series of terms, which involve the same elements: by comparison of these two scales, the progress made in the interpretation of nature will readily appear, and the lines of further research will be clearly indicated.

The phenomena indicative of the presence and degree of heat below the surface of the earth, are either such as mark its ordinary and regular state, as HOT SPRINGS, which, with a few exceptions, are not known to vary in their temperature, and VOLCANOS, which mark, in their epochs of critical action and their periods of repose, the measure of the intermitting agencies connected with their origin, growth, and decay. The conclusions which arise from these cognate phenomena may be further tested by experimental inquiries into the statical temperature at small depths below the surface of the earth.

Volcanic Action.

Volcanic action, considered in its full meaning, includes, perhaps, the largest class of phenomena, attributable to one predominant agent, which falls within the province of geology. These phenomena are the more interesting and instructive, because they extend through an immensity of past duration, with many variations distinctly related to geological and historical time. The facts known by history and tradition respecting particular vents of subterranean fire, go back to the origin of history and civilisation, and other phenomena of the same volcanoes are undoubtedly to be referred to a part of the scale of geological succession, corresponding to the forms of plants and animals which lived and died before the present races occupied the surface. Each volcanic mountain has its own peculiar history, its accident of origin, its law of progressive increase, its period of inevitable decay; it is a monument more venerable than the pyramids; recalling, by its mysterious agitation of the fertile plains around, the remembrance of movements affecting other lands and seas than those on whose boundaries volcanic fires are now excited.

What augments the interest naturally attached to problems regarding the long duration and varying energy of volcanic fires, is the completeness of the series of phenomena which, taken collectively, they present. New vents are opened in every few years to show us the origin of volcanic accumulations on the land or in the sea; an hundred ignivomous mountains bring up to the surface abundant examples of substances most instructive on points which otherwise could only be sources of vain conjecture; and the last stage of these frightful disorders of nature is seen in many districts where, only at particular points, mephitic vapours rise to darken the smiling picture of general fertility.

Origin of Volcanos.

A mountain which has long been silent, and on whose slopes the cultivation has spread for ages, is yet the centre of great subterranean disturbance, shaken by earthquakes, and surrounded by hot springs and sulphureous exhalations. It cannot be known, from such phenomena alone, whether the volcanic energy of this particular region is sinking slowly to the entire decay, which the perishing craters of the Eifel indicate, or reawakening to violent efforts, like those which Vesuvius made in the year 79 of our era, after many centuries of entire repose, while the older crater of Monte Somma was falling in decay.

The renewal of action in an old volcano, after a long period of repose, may be looked upon as exhibiting, in a considerable degree, the phenomena which accompany the first origin of a volcanic vent. Earthquakes, subterranean noises, the bursting forth of new springs, and the suppression of old sources, are symptoms of a particular kind of subterraneous disturbance, of which they record the violence, and in some degree moderate the effects. Volcanic forces are in action wherever such phenomena appear; and, unless the imprisoned powers acquire an extraordinary intensity, these are their only effects; volcanic eruptions are the consequence of forces which have accumulated beyond the relief afforded by displacements of the crust of the earth.

The terrific aspect of a burning mountain, and the immense volumes of melted rocks and scattered ashes which remain as measures of its fury, affect the imagination too strongly; and in this scene of temporary violence we forget the less marked, but really important, changes occasioned by the disturbance of interior temperature, which in sudden earthquakes, or more gradual and extensive changes of position among the masses of matter, is slowly modifying the aspect of the globe.

But, independent of the information to be gathered from the renewed activity of particular volcanos, like Etna and Vesuvius, whose changes of condition are matter of history, the prolific energy of heat has raised up islands in the sea, and mountains on the land, within our own days; and though these new volcanos are always near to the situation of old ones, and are really only new chimneys to the same subterranean fires which those conducted to the surface, the circumstances of their origin are very instructive.

In what form does the ground open for the formation of a new volcanic vent? This question has been answered by Von Buch's hypothesis of "craters of elevation," which, taken as the origin of a volcanic mountain, are described as being formed by the uplifting of the ground in a dome-shaped or conical elevation, with a central aperture. The correctness of this opinion has been disputed by Mr. Lyell, and both observation and calculation have been employed to determine the truth. What is now seen of volcanic mountains in general, proves them to be accumulations of ashes and lava currents, heaped in a conical shape round a central aperture. Supposing the aperture made, it is obvious that lava streams from its edges would flow only to limited distances, and scoria and dust would fall in showers round the opening: and thus every volcanic cone would show, in a vertical section, as fig. 95., layers (l) more or less irregular, sloping

each way from the crater (e). In a horizontal section, the layers of ashes and streams of lava would be distinguished, as in fig. 96.

[The dotted parts correspond to the depositions of ashes falling all round the crater, and enveloping the lava currents, which ran down different sides of the mountains at different times. In one part the lava is seen filling a cross rent in the mountain, like a dyke of

older rocks.]

Mountains thus constituted have been, doubtless, formed by successive eruptions; they may be called "craters of eruption;" but still the question recurs,— What was the origin of the opening through which those ejections began, which in their continuance have formed craters of eruption?

In several cases which have occurred within the reach of authentic history, eruptions on Etna and Vesuvius have commenced in the opening of a fissure through the previously aggregated masses of volcanic substances. This happened in 1538, when the Monte Nuovo rose (the greater portion in a day and a night) on the shore near Puzzuoli, which had been previously (for two years) disturbed by earthquakes. Fissures appeared en Etna in 1669, when the Monte Rossi, which is a double cone of 450 feet in height, was formed by explosion, and lava currents ran down the mountain.

The year 1759 witnessed the formation of a new volcanic vent, and the accumulation of the new mountain of Jorullo (1695 feet high), west of the city of Mexico. According to Humboldt's relation, "a tract of ground from 3 to 4 square miles in extent rose up in the shape of a bladder;" and the bounds of this convulsion are still distinguishable from the fractured strata. This important statement has been controverted by the opponents of Von Buch's hypothesis of "Erhebungs Cratere," and defended by its favourers; and if Humboldt's account remains the only authority for such a mode of origination of a crater on land, we must also remember that it is the only authority for any mode of origin of the opening of a new volcanic region.

But it may be asked,—Are there no characteristic arrangements of the volcanic rocks, which may be employed to determine whether they were accumulated in a level, or in an inclined conical position? are there no characters of form or fissures by which a mountain of elevation can be distinguished from a crater of eruption? It is maintained by De Beaumont and Dufrenoy that there are. If we attend to the forms necessarily assumed by lava flowing from the crater of a volcano, we shall see the almost impossibility, that the melted matter should flow equally on all sides, so as every way to invest the cone with a concentric strata of rock. Wherever the crater is lower, or the slope of the cone depressed, there the liquid would be directed, and long streams (or coulées), not zones of rock, be solidified. If then, in any case, the structure of volcanic masses is such that the distribution of once melted rock is concentric to the conical surface, and not in narrow streams parallel to the slope, such a mass of rocks may be thought to have been raised by expansion, by elevation from an originally nearly horizontal strata. If, indeed, we suppose the lapse of immense time, many streams of lava may successively flow down, and cover the whole conical slope; but not regularly, nor with that uniformity and mutual union here meant by the term concentric sheet of rock.

The cases are few in which this arrangement of the volcanic layers appears. The insulated hills of trachyte ("domite") near Clermont, in Auvergne, are supposed by Dr. Daubeny to be of this nature; the Mont d'Or and Plomb du Cantal have been specially quoted and illustrated in proof of Von Buch's speculation, by MM. De Beaumont and Dufrenoy. Distinguishing clearly, in their prefatory remarks, between the enveloping of a mountain slope by many streams of lava, and the elevation, with fractures, of broad floors of rock, into a conical mass, they attempt, by a consideration of the structure, form, and fissures of these mountains, to determine rigorously to which of the two cases they belong. In this argument the fissures yet existing in a volcanic mountain are an important part of the data;—it requires no great exercise of calculation to see plainly that, on the supposition of a conical elevation, the fissures will grow wider and wider, till they meet in a large subcentral hollow; and the sum of their breadth, will vary as the inclination of the cone; and it depends upon a careful examination of the district whether these conditions be fulfilled. In the opinion of the able geologists quoted, the state and appearance of the sheets of rock which concentrically form the Plomb du Cantal, is such as to agree with the hypothesis, which, besides, is supported by an examination of the nature of the rocks. The Plomb du Cantal, they observe, is in no manner assimilated to a denuded cone of eruption: this supposition of its origin is, on several accounts, inadmissible; it is, on the contrary, by all its characters, the result of elevation operated on a great basaltic plateau, resting on trachyte. The group of Mont Dor requires, on this hypothesis, several centres of elevation; on Mr. Lyell's view, as many points of eruption.

The conclusion of Dufrenoy and De Beaumont has been objected to by Mr. Lyell on various grounds, principally the unequal thickness of the presumed plateaux of volcanic rock now found sloping from the Plomb du Cantal; for these, according to Prevost, are thickest toward the centre. It is satisfactory to refer to an independent inquirer, very competent to deliver a just decision on all the bearings of this subject. Professor Forbes, visiting Auvergne in 1835, directed his attention to Von Buch's hypothesis, and has recorded the result (Edinb. New Phil. Journal, July, 1836). He noticed the radiation of valleys from the Cantal, their narrowing from the centre of the elevation outwards, and their wanting lateral valleys. These radiating valleys, so numerous, from a single mountain, appear to have originated in fissures of disruption. The alternation of "stratified" tufa, with trachyte, under a capping of basalt, in the slopes of the mountain, is an argument of weight with professor Forbes, and leading to the same conclusion. "Upon the whole," says this careful observer, "it seems to me that the evidence of earthquakes subsequent to the deposition (in whatever way) of the Cantal and Mont Dor, is a fact so indisputable as to render the argument about craters of elevation to a great extent merely verbal."—"There seems, therefore, so much of probability, and so little of extravagance, in Von Buch's theory, that we wonder how it could possibly have given rise to such animated opposition."

Let us turn from volcanic districts to others in which stratified rocks have been subjected to vertical displacement, in order to see in what forms the dislocated rocks are combined. Are there in such rocks "hollows of elevation" such as may be compared with the erhebungs cratere of Von Buch? It appears that there are such elevations, unless, with regard to the lake of Laach, we reject the obvious inference from its general figure, and are prepared to doubt the exactitude of the description of the "valley of elevation" of Woolhope. Such cases are, however, rare; they seldom occupy an exactly, or even approximately, circular area: the Woolhope valley is elongated, the Laach crater imperfect; the valley of elevation of Kingsclere is very little allied to a conical mountain; Greenhow Hill, in Yorkshire, though a double or transversely divided elliptical elevation, is, perhaps, as good a case in point as any that can be mentioned in England, to show the analogy which really obtains between the elevation of ancient strata and that of some modern volcanic tracts. To what extent the admission of this analogy bears on the origin of particular groups of mountains remains to be seen, but it seems probable that most of the volcanic mountains are, like Vesuvius, Etna, and Stromboli, craters of eruption, while a few may be better explained by a general or partial elevation, at the origin or during the continuance of their action.

It must not be thought that the discussion regarding the first opening of volcanos is unimportant: the history of ancient elevations of the strata is closely connected with that of modern earthquakes; and the occurrence of volcanic fires along mountain lines is a circumstance very intelligible, upon the supposition that they were caused by the opening of the ground along a great fissure, and perhaps hardly to be explained otherwise. If volcanic regions, arranged in line, owe their origin to the rupture of the ground along that line, its length, and the degree of displacement of the rocks on its sides, are measures of the repressed force which at length found vent. "Volcanos in line," as Von Buch calls them, are thus connected with the traces of the grandest movements which the crust of the earth has experienced; and those who contend against the origin, by elevation, of single volcanic hills, oppose the doctrine of mountain elevation by one or a few violent struggles of nature, anterior to volcanic eruptions along them, and attribute the elevation of ranges like the Andes to many successive efforts of the volcanic action seated below them. The further discussion of this subject is part of a general inquiry, comprehending alike the modern and ancient movements of the land, which will be found in the next chapter.

Volcanos in Action.

Earthquakes, and the other premonitory symptoms of a volcanic crisis, are succeeded by eruptions from conical hills which have previously yielded passage to the fiery floods pressed upwards to the surface, from new orifices on the flanks or at the base of ancient cones, or from situations where volcanic action is a novelty. The effects vary according to the diversity of conditions. The materials issue in the form of melted rocks (lava), or are driven up in the state of ashes and dust (scoria, &c.), or burst forth as gas or steam. The lava, lifted by great mechanical pressure from some depth in the earth, rises in the tubular passage of the mountain toward its summit; and if the sides of the cone are strong enough to resist the accumulating pressure, it may even overflow the top, as has happened in the Peak of Teneriffe, to whose very summit Humboldt traced a stream of vitreous lava. But, generally, the slowness with which an eruption proceeds, is such as to allow of the lava making for itself lateral passages to the surface, on the flanks of the mountain, through fissures which yield to the pressure of the column above, or are opened by earthquakes. Such lateral eruptions have raised many minor cones on the slopes of Etna, and round the base of Vesuvius. Portions of the lava which enter fissures in the sides of the mountain, and are consolidated therein, may be compared to the dykes of the older pyrogenous rocks.

Lava, whatever be its chemical composition, puts on very different appearances, according to the circumstances which accompany its consolidation. The main circumstances which thus modify its aspect, are the volume of melted rock, the exposure of its surface to air or water, the nature and position of the surface on which it rests. Prismatic structures seldom appear in the rocks, except where the mass of the lava was great; cooled in sea-water, the lava of Torre del Greco became more dense than that which was cooled in air, and assumed rudely prismatic structures. On sloping surfaces it is found that the cellular cavities, common to lava which is cooled in the air, are elongated in a direction parallel to the slopes,—an effect clearly intelligible by considering the viscidity of the moving mass, and easily imitable by art.

The minerals which enter into the composition of lava are, as already stated (p. 83.), chiefly felspar, augite, and titaniferous iron. But besides these, many varieties of substances are produced in a crystallised state during the cooling of the fused mass; and, as is commonly observed among the old rocks, such as granite and basalt, these occur most plentifully, and in the finest crystallisations, in cellular cavities and small fissures of the lava. Eighty-two species of minerals are enumerated in a catalogue of the products of Vesuvius by Monticelli and Covelli, and others have been added to the already large list of this unusually rich locality.

"Lava, when observed as near as possible to the point from whence it issues, is, for the most part, a semifluid mass of the consistence of honey, but sometimes so liquid as to penetrate the fibre of wood. It soon cools externally, and therefore exhibits a rough unequal surface; but, as it is a bad conductor of heat, the internal mass remains liquid long after the portion exposed to the air has become solidified. That of 1822, some days after it had been emitted, raised the thermometer from 59° to 95° at a distance of 12 feet; 3 feet off, the heat greatly exceeded that of boiling water. The temperature at which it continues fluid is considerable enough to melt glass and silver, and has been found to render a mass of lead fluid in 4 minutes, when the same mass, placed on red-hot iron, required double that time to enter into fusion."—" Even stones are said to have been melted when thrown into the lava of Vesuvius and Etna. On the other hand, the temperature in some cases does not appear to have been sufficient to fuse copper; for, when bell-metal was submitted to the action of the lava of 1794, the zinc was separated, but the copper remained unaffected." (Daubeny, On Volcanos, p. 381.) These experiments on the heat of lava at the surface are not at all discordant with what is known of the easy fusibility of basaltic and trachytic compounds. In lava at such temperatures it might happen that fragments of granite, mica schist, &c. should escape fusion; and such are stated, on good authority, to have been found in the midst of the lava of Vesuvius, Etna, and the Ponza Isles; while limestone in a similar situation is found of that crystalline texture often observed in calcareous rocks which have undergone fusion.

The volume of melted rocks poured forth in a single short eruption of Vesuvius is considerable; far greater during some of the long-continued periods of activity of the Icelandic volcanos; enormous, if we contemplate the united effect of a whole chain of volcanos like those of South America. In 1737, the current of lava from Vesuvius which destroyed Torre del Greco, and ran into the sea, is supposed to have accumulated no less than 33,587,058 cubic feet (equal to a cube of above 322 feet by the side, or a cone of the same height and above 630 feet diameter at the base). In 1794, another current, which flowed also through the same ill-fated town, was calculated by Breislac, who saw the eruption, to equal 46,098,766 cubic feet.

Etna, which rises above 10,000 feet in height, and embraces a circumference of 180 miles, Dr. Daubeny assures us, is composed entirely of lavas, which appear to have been emitted above the surface of water, and not under pressure. "In the structure of this mountain, every thing wears alike the character of vastness. The products of the eruptions of Vesuvius may be said almost to sink into insignificance, when compared with these 'coulées,' some of which are 4 or 5 miles in breadth, 15 in length, and from 50 to 100 feet in thickness; and the change made on the coast by them is so considerable, that the natural boundaries between the sea and land seem almost to depend upon the movements of the volcano." (On Volcanos, p. 203.) The great current of 1669, which destroyed Catania, is estimated by Borelli to contain 93,838,950 cubic feet.

But it is in the great eruptions of Iceland, as that of Skáptaa Jokul (in 1783), that the effect of the continued energy of the subterranean fires, in ejecting matter to the surface, becomes most astonishing. The fearful eruption alluded to did not entirely cease till the end of two years: in its course the lava filled valleys 600 feet in depth; dried and took the place of lakes; accumulated in rocky gorges; spread in wide plains till they became broad burning lakes, sometimes from 12 to 15 miles wide, and 100 feet deep. The lava may be said to have taken two principal and nearly opposite directions; flowing in one 50, and in the other 40 miles, with a breadth in the former case of 15 miles, in the latter of 7. The ordinary depth of the accumulated mass was about 100 feet, but in narrow defiles it sometimes amounted to 600 feet. Mr. Lyell, from whose admirable summary of this destructive eruption the above abstract is taken, makes an ingenious comparison of this prodigious mass of modern pyrogenous rock with older effects of the interior heat of the globe, and illustrates its effect on the geology of England, if spread like the basaltic plateau of Antrim. Spread upon the stratified rocks of England, before their elevation from the sea bed, the lava would have occupied a vast continuous surface; and, after the rising of the rocks and their waste by watery action, the original extent might be traced. The Skápta branch of the lava might rest on the high oolitic escarpment which commands the vale of Gloucester, 100 feet in thickness, and from 10 to 15 miles broad, exceeding any which can be found in Central France. Great tabular masses might occur at intervals, capping the summit of the Cotswold hills, between Gloucester and Oxford, by Northleach, Burford, and other towns. The same rocks might recur on the summit of Cumnor and Shotover hills, and all the other oolitic eminences of that district. Plateaus 6 or 7 miles wide might have crowned the chalk of Berkshire, and masses 500 or 600 feet thick might have raised the hills of Highgate and Hampstead to rival or surpass Salisbury Craigs and Arthur's Seat. (Principles of Geol. book ii. ch. xii.) To this prodigious fiery flood, there are certainly few phenomena of superior grandeur among the "wonders" of geology.

Dispersion of Ashes.

The currents of lava, though they may appear to flow with a certain regularity, are really urged by forces which continually rise to explosive energy, and dissipate parts of the liquid columns within the crater into scoriæ and ashes. This effect appears in no small degree due to a circumstance almost universally observed in volcanic excitement,—the extrication of vast volumes of aqueous vapour. To the mechanical energies which steam exerts at the base of the fiery funnel, and in the substance of the mass of lava, we may, perhaps, refer most of the phenomena attesting great expansive power. The ashes, scoriæ, and stones which are shot upwards from the mouth of the volcano, and fall in showers around, are of the same mineral composition as the solidified parts of the lava: they mostly rest on the slopes, and augment by external layers of growth the diameter of the volcanic mound. The white lapilli, and black ashes, remind us, in this pulverulent state, of the felspathic and augitic rocks whence they are derived; and it is probable that in this way much larger accumulations happen on and around Vesuvius, Etna, and some other volcanos, than those which are produced from flowing lava. Pompeii, Stabiæ, and Herculaneum were buried in ashes and sediments derived from ashes, to depths of 60, 80, and 100 feet; and it has been calculated that the masses ejected from Vesuvius vastly exceed the whole bulk of the mountain. (Daubeny on Volcanos, p. 155.)

The ashes, instead of falling round the volcanic cone, are sometimes carried for great distances by the winds. Owing to the commotion of the atmosphere during these paroxysms of the earth, rains often descend, and sweep away the falling ashes in rivers of mud ("lava d'acqua"), which flow according to the slopes of the ground, and cover up cities, and fill lakes and valleys. To this cause a part of the accumulation covering Herculaneum has been ascribed, while Pompeii was overwhelmed in dry ashes. It is easy to perceive that alluvial accumulations will from this cause spread over a large extent of country round the base of an ignivomous mountain, the arrangement of which is purely the effect of water, though the materials are exclusively the products of heat. Such volcanic sediments will be arranged in a consistent geological classification as aqueous deposits; they may contain as well as cover many organic productions, wood, shells, bones, &c., and be thus, in some cases, referred to their true geological age.

"Volcanic sandstones" as Mr. Murchison calls the marine deposits of ashes and disintegrated trap rocks, which are inter laminated among the rocks of the silurian system, may have had, in some instances, a similar origin.

Another mode of aggregation of similar ingredients is exemplified by some part of the "trass" deposit, as it is called, in the country near Andernach, where it abounds on the borders of the Eifel volcanos. Showers of ashes falling in lakes would be arranged therein exactly as other sediments from a different source, except that the areas and depths of the distributed substances must vary according to the circumstances of their admission to the water. Much of the trass in the Valley of Brohl is, however, in too irregular a state of arrangement to admit of this view. It probably was deposited rather as a mass of liquid mud, bursting from some old crater, and bearing the spoils of the surface (wood and rock fragments) with it. The wood in this trass is carbonised. The puzzolana of Naples is of similar nature to the trass, and contains shells and bones, with fragments of pumice, obsidian, and trachyte. It forms considerable hills round Naples, some of which have regular craters.

When, as in the case of Graham Island, a new volcano bursts up in the sea, and scatters ashes and scoriæ, these, falling in the sea, are variously disposed of, and may be borne by currents far from their origin. If, like the same island, the volcanic heap, after subsisting for a time, is wasted away by the waves, we can easily predict the effect on the sea bed near;—sloping strata of volcanic sediments, which may cover or envelop abundance of mollusca, and even fishes poisoned by mephitic gases, which frequently break forth in points not far from the centre of the eruption. Among the singularities of the eruptions of Vesuvius is the pouring forth of boiling water from the sides of the mountain (Daubeny, 156.). Eruptions of this nature are less rare in the New World. Humboldt mentions the singular fact, that with these aqueous eruptions pass multitudes of small fishes along with abundance of mud.

"When (on the 19th of June, 1698) the Peak of Carguairazo sunk down, more than four square leagues around were covered with clayey mud, called in the country "lodozales." Small fish known by the name of "prenadillas" (Pymelodes Cyclopum),—a species which inhabits the streams of the province of Quito,—were enveloped in the liquid ejections of Carguaiiazo.

These are the fish said to be thrown out by the volcano, because they live by thousands in subterranean lakes, and, at the moment of great eruptions, issue through crevices, and are carried down by the impulsion of the muddy water that descends on the declivity of the mountains. The almost extinguished volcano of Imbaburu ejected, in 1691, so great a quantity of prenadillas, that the putrid fever, which prevailed at that period, was attributed to miasmata exhaled by the fish." (Humboldt on Rocks, p. 455.)

Fetid mud, called "moya," burst, in enormous quantity, from the foot of the volcano of Tunguragua, in Quito, in 1797, and filled valleys and dammed the course of rivers. Sulphuric acid is mixed with the waters which flow from Purace, in Quito, and some other extinct volcanos.

Besides ashes, scoriæ and stones even of size are thrown out by the volcanic forces, and sometimes take their course with the drifts of mud, so as to form part of the re-aggregated mass of trass, or constitute a volcanic conglomerate.

The last class of volcanic products which come to the surface are the gaseous and vaporous substances, to which much of the grandeur of the exhibition, as well as much of its general power and momentary energy, is owing. The most abundant of these is steam, which rises in white clouds over the craters of active, and from rents in extinct, volcanos. The most abundant of the gases are muriatic acid, sulphuretted hydrogen, sulphurous acid, carbonic acid, and nitrogen. (Daubeny.) Sublimations of particular solids occur, as boracic acid in the crater of Volcano, muriate of ammonia, muriate of soda, specular iron ore. The boracic acid cannot be sublimed by the heat of our furnaces; but Dr. Daubeny has shown by experiment, that, when heated and traversed by steam, a portion is taken up and carried with the steam.

Extinction of Volcanos.

The suppression of volcanic excitement lasts so long in some cases, that the long and quiet sleep is not to be distinguished from a real extinction of the local energy of heat. Between two eruptions in Ischia seven-teen centuries elapsed. In this respect the history of Vesuvius is very instructive, especially when compared with the aspect of the long decayed volcanic mounds of the Eifel and Auvergne, whose fires were, perhaps, never beheld by man.

The cone of Vesuvius is of comparatively modern date, formed within the larger and more ancient crater of Monte Somma. The descriptions given by Latin writers seem applicable to this latter mountain, up to the great eruption of A. D. 79, which Pliny's narrative has rendered famous. Previous to that event the mountain was cultivated; its crater, perhaps, served as the encampment of Spartacus; and only obscure tradition or uncertain inference raised the conjecture that this smiling tract was based on subterranean fire. If the passage of Lucretius (vi. 748.) has any reference to Vesuvius, the only symptoms of activity of heat were sulphureous exhalations, such as might rise many centuries after the volcano had sunk to rest, such as now rise in the Solfatara, and have risen, with little difference, for 1600 years! (See Dr. Daubeny on Volcanos, p. 166. first edition.)

The great eruption of A.D. 79 was followed by six others, at long intervals, averaging 164 years, till 1036, when, for the first time, the flowing of lava is mentioned, the previous eruptions being of ashes and lapilli. Three eruptions are on record between 1036 and 1306. Vesuvius has never, since the first outbreak on record, been at rest for so long a period as between 1306 and 1631, between which epochs only one slight revival of action happened in 1500. Throughout this period Etna was in a state of unusual activity, as if the rival craters of Sicily and Campania were connected to the same subterranean channels. Before the eruption of 1631, the crater of Vesuvius was a pasture for cattle, its sides were covered with brushwood, in which wild boars sheltered. The old surface was all blown into the air, and seven streams of lava poured at once from the crater, committing enormous destruction. Since that time the mountain can hardly be said to have been ever tranquil, and the frequency of eruptions appears to have progressively augmented to the present time. In the seventeenth century, the intervals between the outbreaks of Vesuvius are, on an average, twenty years; in the eighteenth, five years; and since 1800, two years.

Etna has experienced, within the reach of history, sufficient variations of volcanic energy to justify the use made of its changes in the Pythagorean philosophy.

                      Nee quæ sulphureis ardet fornacibus Ætna,
                      Ignea semper erit; neque enim fuit ignea semper.

Ovid. Metam. XV.

The speculations with which this opinion is accompanied,

show the activity of inquiry which was excited among the people adjoining the Mediterranean volcanos.

The early eruptions of Etna are lost in the obscurity of history, and the great mass of the mountain was probably accumulated during the later tertiary periods of geology. The first recorded eruption, in 480 b. c., was followed by others, 427 and 396 b. c.; the intervals averaging 42 years. After 256 years, in which no eruptions are recorded, four more are noticed between 140 and 122 b. c.; average interval 6 years. After 66 years of rest, three other eruptions appear between 56 and 38 b. c.; average interval, 9 years. No eruption is mentioned till 40 a. d.; interval, 78 years; a pause till 251 a. d.; another, still longer, till 812 a. d.; a third to 1169 a. d.; and then, after twelve centuries of rarely interrupted quiet, the mountain became, agitated, and has since continued to manifest its violence, more and more frequently, to the present century. In the twelfth and thirteenth centuries, 3 eruptions; in the fourteenth century, 2; in the fifteenth, 4; in the sixteenth, 3 of unexampled duration; in the seventeenth, 8; in the eighteenth, 14; in the nineteenth, to 1832, 6 eruptions. (See Dr. Daubeny on Volcanos, and Mr. Lyell's Principles of Geology, for details, which are here unnecessary.)

The Lipari Isles present us with yet another variation in the phases of volcanic action and rest. Stromboli is always active, but almost never violent; no cessation having ever been noticed in its operations, which are described by writers antecedent to the Christian era in terms which would be well adapted to its present appearances; while in Lipari, the only indications of volcanic action now existing are the hot springs; and the island of Volcano, in an intermediate state, still emits gaseous exhalations.

Since the first colonisation of Iceland by the Norwegians, the eruptions of the volcanos in that country have been frequent, and almost regularly distributed, through the ten centuries during which it has been known to us. Dr. Daubeny notices as the first eruption recorded, that at the end of the ninth century (894 a. d.). One also occurred in 900. The subsequent dates of eruptions are, 1000, 1004, 1029; 1104, 1113, 1157, 1158; 1245, 1262, 1294; 1300, 1311, 1332, 1340, 1359, 1374, 1390; 1416, 1436, 1475; 1510, 1554,1580, 1587; 1619, 1622, 1625, 1636, 1660, 1693; 1717, 1720, 1724, 1728, a series of eruptions, 1748 to 1752, 1753, 1772, 1783. In 1724 occurred the first eruption of Krabla. Eruptions have subsequently occurred in 1821, 1823.^[1]

During this period submarine eruptions happened from 1224 to 1240; in 1422; in 1563; 1783; and new islands were thrown up in 1563, 1783.

Here, therefore, we have the recorded history of four volcanic systems, which appear very unequal in their progress toward decay, as if their energy depended upon conditions differently apportioned to the several regions. Without repeating all the hypotheses in Ovid, which commence with the notion of the earth being an animal that breathes flame through many variable spiracles, we may inquire whether the fluctuation of volcanic energy in particular districts depends upon local and temporary stoppage of the channels to the surface, or upon the failure in some of the essential conditions of igneous excitement? To answer this question we must survey volcanic phenomena in a variety of other aspects.

Extinct Volcanos.

The Solfatara, near Puzzuoli, is in shape like other volcanic cones, with craters at the summit; it is formed of a trachytic rock, naturally hard and dark-coloured, but in proportion as it is exposed to the vapours given off from the "fumaroles" in the crater (the steam contains sulphuretted hydrogen and a minute proportion of muriatic acid), its texture and colour undergo a remarkable alteration. It passes through various stages of decomposition, and finally appears a white siliceous powder. Saline compounds effloresce on the surface of the rock (muriates of ammonia, &c., sulphates of alumine, lime, soda, magnesia, iron, &c.), and sulphur (not sublimed alone, but derived from sulphuretted hydrogen) lines the walls of its cavities. The ground is hollow (probably in fissures) below, and a stream of trachyte has formerly flowed from it, and formed the promontory called the Monte Olibano.

Craters in which the volcanic fires are utterly extinct are sometimes filled by water, as the Lago Agnano, and the more celebrated Lake Avernus, where no longer rise the sulphureous fumes which once procured it the formidable character of a gate of hell. (Æneid, vi.) They may be compared with some of the craters of the extinct Rhenish volcanos, as the Laacher See, near Andernach, which is about 2 miles in circumference, the Meerfeld, and other circular lakes or "maars" of the Eifel district. Sulphureous exhalations, which resemble those of the Solfatara, and lakes in craters like those of the Eifel, occur in Hungary and Transylvania; and the central districts of France show us, in addition, a variety of facts, to complete the view of the condition of countries where, though volcanic action, as commonly understood, is entirely extinct, the effects of subterranean heat and chemical decompositions are manifested by evolutions of particular gases, and the issue of hot springs.

Geographical Distribution of Volcanos.

Though volcanic accumulations abound in all quarters of the globe, the area which they occupy on the land is not to be compared to that of any one of the systems of stratified rocks, and is inferior to that of most of the individual formations. On a first view, volcanic mountains seem to be so many insulated points of ignition, productive of distinct mineral compounds, and subject entirely to independent local conditions. The history of eruptions, though very incomplete, is, however, sufficient to destroy this notion, by showing on the line of the Andes corresponding movements of the land, and ejections of ashes into the air, at points very far removed from each other.

Thus, a few days after the earthquake which destroyed Concepçion, on the 20th of February, 1835 several volcanos within the Cordilleras, to the north of Concepçion, though previously quiescent, were in great activity, and the island of Juan Fernandez, 360 miles to the north-east of the city, was violently shaken. The volcanos of Osorno, Aconcagua, and Coseguina (the first and last being 2700 miles apart), burst into sudden activity early on the same morning (June 20. 1835).

This connexion and sympathy of the phenomena of volcanos and earthquakes at considerable distances, is an important element for the determination of the true condition of the subterranean spaces where these phenomena are excited. The town of Riobamba, near Tunguragua, was destroyed by a tremendous earthquake on the 4th of February, 1797; and at this moment, the smoke which had been seen to issue in a thick column from the volcano of Pasto, 65 leagues north of Riobamba, suddenly ceased. Volcanic mountains appear to act as safety-valves to a boiler, and some of them relieve completely and continually the subterranean pressure, so as to free from earthquakes a considerable region round their bases. Thus, while earthquakes agitate the neighbouring islands of the Canary group, the Peak of Teyde appears to be the cause of the comparative immunity from these disasters which the island of Teneriffe enjoys.

By combining observations of this nature with considerations of the grouping of volcanic craters, the direction and extent of earthquakes, the ebullition of hot springs, and analogous phenomena, we arrive at the notion of volcanic regions, and may by this means class the active and extinct volcanos, which are scattered over the globe into a modern number of systems, convenient for description, even if the association rests on an uncertain basis.

In Europe, the purely volcanic phenomena which are from time to time manifested, appear at points referrible to one of seven centres of action. Iceland, with its Geysers and six or more active volcanos, Hecla, Skaptaa Jokul, Skaptaa Syssel, Eyafialla Jokul, and Kattlagiaa, and its formidable eruptions and coast elevations, stands almost alone; Jan Mayen, visited by Scoresby, being its only volcanic neighbour. The Azores form another region of considerable importance, where the rising of islands has happened within the reach of history. Sicily, like Iceland, has, besides the great chimney of Etna, lateral escapes of the imprisoned forces labouring below; and Sciacca, the island which rose and disappeared in 1 831, is a remarkable proof of their energy. The Lipari Isles form another group, where Stromboli and Vulcano are still feebly active; while Vesuvius stands and burns amidst many older vents, long since extinguished, and Ischia and the Ponza isles bear to it the same relation that the Lipari group does to Etna.

The last centre of activity in European volcanic systems is in the Greek Archipelago, where Santorini has undergone many violent displacements, and some igneous exhibitions since the Christian æra.

Similarly the extinct volcanos of Europe may be grouped very conveniently in a few systems of connected groups. The basaltic mountains and cliffs of the Farö Isles are stated to be on many points allied to those of Staffa and Antrim, and perhaps the whole region in which igneous rocks are scattered, from the Farö Isles to Antrim, Arran, the Vales of Clyde and Forth, if not to Teesdale and Derbyshire, should be viewed as the theatre of one great and long enduring system of submarine volcanic forces.

Another great system, of more recent date, is the tripartite volcanic tract of Central France, included in the districts of Auvergne, Cantal, Velai, and Vivarais, to which we may attach some points scattered about the Cevennes mountains near Rodez, at Agde near Montpellier, and Beaulieu, near Aix en Provence.

Just before the Rhine enters the Low Countries, which conduct it to the sea, it divides hilly districts, principally of transition rocks, among which, on the left bank, is a large exhibition of ancient volcanic energy, in the numerous cones and "maars" of the Eifel country, lying east of the Ardennes. On the right bank, lower down, are the celebrated trachytic and basaltic mountains, called the Siebengebirge, which by some detached rocks of like nature, lying to the east, appear to be connected with the great basaltic masses of the Westerwald.

The Westerwald, Vogelsgebirge, and Rhongebirge, may be taken as the principal volcanic group of Western Germany. Many insulated cones and masses of basaltic rocks about Limberg and Wetzlar, the Habichtwald near Cassel, some basaltic hills near Eisenach, Fulda, Hanau, and Frankfort, may perhaps be contemplated as parts of this ancient system.

The Kaiserstuhl mountain, near Freyburg in the Brisgau, the cone of Hohentwiel, and some small points in Würtemberg may be grouped together, though it is not known that they have really any particular geological relations.

A few detached basaltic cones appear near Egra, amidst the Fichtelgebirge. A much larger volcanic tract begins in the Mittelgebirge, and extends parallel to the Erzgebirge, and across the Elbe to near Zittau. The line of this system is continued by many detached cones across the range of the Riesengebirge into Silesia.

Volcanic appearances are mentioned near Ho north of Olmutz.

In Hungary, as described by Beudant, the effects of extinct volcanic action are extensive and remarkable. Five distinct groups of mountains, composed wholly of trachyte, are enumerated by Beudant, who attributes to each group a separate origin. One of these groups, larger than that of the vicinity of Clermont, being 20 leagues long, and 15 broad, is situated in the porphyritic mining district of Schemnitz and Kremnitz; another, smaller, but similarly circumstanced as to the porphyry, crosses the Danube, near Grau; another, extending east and west elliptically, forms the mountains of Matra, near Eger; the fourth, a large mass, ranging north and south, for 25 or 30 leagues, from Tokai to Eperies; the fifth is the group of Vihorlet, east of the last, apparently related to the trachytic mountains of Marmarosch, in Transylvania. Most of the Hungarian volcanic rocks can be classed as varieties of trachyte, according to a method of M. Beudant; opal, opalised wood, pearl-stone, and pumice, and scarified masses, abound; and from the porphyries of earlier date, there appears to be an easy mineral gradation to some of the trachytes. The latest volcanic action is placed by Mr. Lyell in the Meiocene tertiary period. The subjacent strata are mostly of the transition æra. Another group of Hungarian volcanos adjoins the Flatten See, on the north-west side.

In the eastern part of Transylvania volcanic rocks of tertiary æra occur in a range of hills covered with thick wood, extending from the hills of Kelemany, north of Remebyel, to the hill of Budoshegy, 10 or 12 miles north of Vascharhely. The principal mass of the range is trachytic conglomerate, from beneath which, at intervals, trachyte of different kinds emerges, and encloses craters at the southern end of the range. Some of these craters are, like those in the Eifel, converted to lakes. Exhalations of hot sulphureous vapours are poured out from rents in the hill of Budoshegy; sulphureous, chalybeate, and carbonated waters rise at the foot of this mountain in many places.

In Styria, the Gleichenberg, a trachytic mountain enveloped in strata of ashes, perhaps accumulated in water, indicates considerable volcanic energy during the tertiary æra. At several other points in Styria volcanic masses appear.

The Euganean hills south of Padua constitute a very remarkable mass of volcanic deposits, consisting principally of trachytic rocks, associated with semi-vitreous masses, and at Monte Venda with basalt. The subjacent calcareous strata of "scaglia" contain many fossils of the European chalk. North of Vicenza, the variety of volcanic products is considerable, and it is thought their differences are partly related to the place which they occupy in the series of strata there occurring, between the primary slates and the scaglia. On the volcanic rocks rest calcareous and tufaceous deposits; and at particular places, especially Monte Bolca, fishes occur abundantly in slaty bituminous marls, which alternate with volcanic sediments, often containing shells, like the trass and puzzolana.

Near Viterbo (Monte Cimini), trachytic rocks abounding in leucite, associated with basalt, and beds of pumiceous tufa, covering bones of quadrupeds, are connected with tertiary marls and shells. Near Radicofani the same trachyte occurs in the Monte Amiata.

A few miles south-west of Volterra, near Monte Rotondo, and near Monte Cerboli, sulphuretted hydrogen rises abundantly from little crater-shaped openings (lagunes), and boracic acid is sublimed therewith, as well as in the crater of the island of Volcano. Mount Albano, 12 miles from Rome, from which a current of lava is traced nearly to the city, as well as the volcanic tuff which alternates with other sediments below the soil of Rome, sufficiently prove the former activity of volcanic forces in this vicinity. Near Albano are four lakes, once probably craters. The Rocca Monfina, a mountain of great antiquity, on the road from Rome to Naples, surrounded with igneous volcanic deposits, carries on the line, of connection to the Phlegrean fields and Vesuvius.

In the Ponza Islands, Mount Vultur, the Lake Amsanctus, volcanic action, though long extinct, has left proofs of its former force and extent at points more or less connected with Vesuvius; while in the Val di Note, the early energies of Etna are manifested among tertiary strata.

Some of the Grecian islands and shores have exhibited volcanic fire, and great elevations of land in modern times, as Santorini; and extinct volcanic action is manifested in the Solfatara of Milo, and the convulsions of Methene and Trœzena, mentioned by Strabo and Ovid.

If we compare this brief notice of the situations where active and extinct volcanos have poured eruptions on the land and in the sea, with the extent of country included by Mr. Lyell in his "Volcanic Region from the Caspian to the Azores," it will immediately appear, that, with the exception of Iceland and Jan Mayen, all the points in Europe which have produced eruptions during the reach of history, are included in that region. The whole space between the Caspian and the Azores, a distance of 1000 miles, within the parallels of 35° and 45° north latitude, has been from time immemorial agitated by earthquakes; which also extend their effects farther to the north, so as perhaps to unite the Mediterranean band of volcanic energy with the distant fires of the Icelandic group. Near to and beyond the latitude of 45° are situated many of the most conspicuous of the older volcanic systems, but the most modern lie farther to the south. The western continuation strikes the Azores. As a general conclusion, it appears that earthquakes extend the evidence of subterranean disturbance much beyond the area covered by volcanic ejections.

If, taking another view of the subject, we inquire the relation of this distribution of volcanic vents to the features of European physical geography, it immediately appears that all the active volcanos are situated in islands or peninsulas, or, in general, very near to the sea-side. Further, it is evident that the same law of proximity to water applies to the ancient volcanos of Auvergne, the Rhine valley, the Hungarian and Transylvanian volcanos, and the Euganean hills, &c.; for these points, now far removed from wide sheets of water, were bathed by fresh waters (Auvergne, Transylvania), or the sea (Euganean hills, &c.), at the time when they were theatres of igneous violence.

Asiatic Volcanos.

Proximity to the sea, or to large surfaces of inland waters, characterises, in like manner, the points where volcanic action is now, and has formerly been, manifested on the continent and islands of Asia. On either side of the sea of Marmora, from the Dardanelles to Constantinople, volcanic accumulations appear. Syria and Palestine, often desolated by earthquakes in early periods, abound in volcanic appearances. Near Smyrna these are extensive^[2], and the vicinity of the Dead Sea is volcanic. The Caucasian chain of mountains is full of volcanic accumulations; Ararat is of this character. At Bakur, on the western side of the Caspian, is the celebrated "field of fire," where excavations in the soil yield naphtha, and gas rises, which is easily inflamed. The Elburz range of mountains, on the southern side of the Caspian, presents one important volcano in action, the Peak of Demavend, which is 1 4,000 feet above the sea.

On the Arabian side of the Red Sea, volcanic phenomena appear at Aden, Medina, Mount Sinai, and other points; the island of Zibbel Teir is said to contain an active volcano. Volcanic phenomena are mentioned at the mouth of the Persian Gulf, in the island of Ormus, and at some distance inland north of Kerman.

Beyond the limits of the Mediterranean and Caspian volcanic regions just described, Humboldt has added to the previous reasons for admitting the existence of some volcanic action in the midst of the Altai Mountains (lat. 42° to 46° N., long. E. 80° to 87°). These volcanos, which are 400 leagues from the Caspian Sea, are nevertheless situated among some considerable lakes so as to invalidate in no degree the generality of the inference drawn from the consideration of European volcanic districts.

On the eastern border of Asia is an immense sigmoidal band of intense volcanic activity, which constitutes one of the most remarkable physical features of the globe. Commencing with Barren Island, in the Bay of Bengal, the line passes south-eastward through Sumatra, where Marsden describes four existing volcanos, one of which is 12,000 feet high. Through Java the line passes nearly east and west, amidst thirty-eight large volcanic mountains, conical in figure, and rising separately from a plain to 5000, 11,000 and even 12,000 feet above the sea. In 1772 one of the largest fell in, so that an extent of ground 15 miles long and 6 broad, with 40 villages, and 2957 persons, were destroyed.

From Java the volcanic hue continues eastward through Sumbawa, known from the formidable eruption recorded by sir Stamford Raffles, and Flores, and Timor, where the burning peak sunk in 1637, and is changed to a lake. Between Timor and Ceram, also, in one of the Banda Isles, in the northern part of Celebes, the volcanic action is manifested among the Molucca Isles. Ternate, Tidore, and Sangir, continue the line in a northerly direction to the Philippine Islands, Mindanao, Fugo, and Luçon. From Formosa, by the Loo Choo Isles to Japan, the line runs north-eastward; a course which it continued through the ten volcanos of Japan, and the nine active vents of the Kurilian islands to the burning mountains of the peninsula of Kamschatka.

The Aleutian Islands continue the line of volcanic activity (an island having been thrown up in 1795 3000 feet in height, according to Langsdorff) to the point of Russian America called Alaschka, which is believed to be also volcanic.

American Volcanos.

Traces of powerful volcanic action, now extinct, appear about the head waters of the Columbia and Missouri rivers; and probably along more southern parts of the lofty ranges of the Rocky Mountains, yet but imperfectly known to Europe. The peninsula of California possesses, besides the lofty Mount St. Elia (17,875 feet above the sea), two other active volcanos. The line of igneous action is continued through Mexico, but not in the general direction of the high mountain range. This goes to the south-east, and it is in a line crossing it obliquely, nearly east and west, that the fire active vents of Mexico, Tuxtla, Orizaba, Popocatepetl, Jorullo, and Colima, are situated. The distance of Jorullo from the sea is 36 leagues, and that of Popocatepetl somewhat greater; and this circumstance may be thought to invalidate the seeming necessity of proximity to water as an element of volcanic excitement. But it appears not unreasonable to admit the existence here of a great transverse fissure, on whose prolongation westward are situated the volcanic (extinct) group of the Revillagigedo. Several intermediate points of extinct volcanic action connect the five active vents above noticed in Mexico.

Between Mexico and the Isthmus of Darien, in the provinces of Guatimala and Nicaragua, are no less than twenty-one active volcanos, running in the line of the great mountain chain. On the southern side of the isthmus three volcanos occur in the province of Pasto, as many in Popayan, and six of surpassing height and grandeur in Quito, viz. Cayambe, Cotopaxi, Pichincha, Antisana, l'Altar, and Tunguragua. The fire comes out from one or other of these giant cones, but, according to Humboldt, they all are parts of a single swollen mass, or immense volcanic wall, covering a surface of 600 square leagues. In Peru one active volcano is known, and there is no other between Quito and Chili, but the whole country is so remarkably subject to earthquakes, that it must be presumed the subterranean connection is continued from Quito to Chili.

In Chili, at least nineteen points of eruption are ranged in the general mountain line of the Andes, here passing southwards: Villarica, one of these, burns continually, but is seldom subject to violent excitement. One point of eruption appears to have been ascertained by captain Hall, in Tierra del Fuego. This extraordinary range of volcanos, which appears to indicate a continuous area of excitement as much as 6000 miles in length, is equally remarkable for the narrowness of its area, and its proximity and uniform parallelism to the boundary of the Pacific. According to Humboldt, all the volcanos of America have burst through older igneous products, such as basalts, trachytes, and porphyries. Granite is the basis of the trachytic masses of Mexico.

If the line of the great Mexican volcanos be prolonged to the eastward, it enters the volcanic portion of the West Indian Islands, on the west it cuts the Revillagigedos. The Gallapagos Islands are volcanic, and the same may be the case with Juan Fernandez.

In a large proportion of the West Indian Islands, volcanic appearances have been recognised; and in several the igneous action is still important. The large islands exhibit least of this. In Trinidad is a great expanse of asphaltum; in Jamaica the Black Hill is volcanic: but all the smaller islands are either of volcanic or coralligenous growth. Grenada, St. Vincent, St. Lucia, Dominica, Montserrat, Nevis, St. Christopher, St. Eustachia, are entirely volcanic; Martinique, Guadaloupe, Antigua, St. Bartholomew, St. Martin, St. Thomas, are partially volcanic, and partially calcareous. The line of these volcanic islands forms an arch convex to the eastward.

African Volcanos.

On the continent of Africa, the notices of volcanic districts are slight and incomplete. Perhaps between the Nile and the Red Sea, as Ruppell and Jomard state, volcanic action is not extinct. In Mount Atlas hasaltic eruptions appear. The African islands, on the contrary, are nearly all, almost exclusively, volcanic. From the Azores, which are usually reckoned as European, the Madeira Isles continue the Atlantic system of volcanic action to the group of the Canaries. Further south, the Cape de Verde Isles, Ascension, Fernando Po, Prince's Island, St. Helena, Tristan d'Acunha, Gough's Island, are so many points of active or extinct volcanic fire. Madagascar, Bourbon, and Mauritius contain abundantly the effects of the same cause.

The circumstances observed in these various groups differ extremely. In Madeira and Porto Santo, Ascension, St. Helena, Tristan d'Acunha, the volcanic fires are extinct, and their effect has generally been to up-heave stratified rocks covered by volcanic accumulations. The Canary group has, in Lanzerote, a low volcanic tract liable to burst suddenly after long intervals (from 1736 to 1834.), and a vent immensely elevated for the escape from gaseous emanations and explosions, in the Peak of Teneriffe, which rises to between 11,000 and 12,000 feet, out of a concentric base of basaltic jocks, between 3000 and 4000 feet high. Von Buch believes this fact to be in favour of his general doctrine of craters of elevation, which is also supported by him upon the evidence of the form of Palma, another of these islands, which have all (according to this view) been raised from the sea, by the up heaving of submarine lava and sediments.

Australia exhibits traces of former volcanic action, in the "Dividing Range," New South Wales; and when the interior of this vast region shall have been fully explored, such may, perhaps, be found extensively.^[3]

Indian Ocean.—A submarine volcano is noticed on the maps in lat. 7° S., long. 87° E. St. Paul's Isles, lat. 38° S., long. 77° E., are also volcanic.

Pacific Ocean.—This great expanse of water appears to be really one vast theatre of volcanic action; for almost every island raised above the water to a considerable height is more or less full of volcanic rocks, some having high and powerfully active craters; while the low islands are of coralligenous growth, and appear in forms that suggest (though Mr. Darwin offers another explanation) the belief of their being founded on volcanic mountain tops. The volcanic systems of the Pacific appear connected with the great littoral band of Asia, already described, by the line of the Banda Isles, New Guinea, New Britain, New Ireland, and other neighbouring islands. The New Hebrides, Norfolk Island, the Friendly Isles, the Society Islands, and the Sandwich Islands, and many of the detached islets which adorn the Tropical part of the Pacific, are principally of volcanic origin. The Ladrone Islands constitute a mountain chain of active volcanos. In Tahiti (Otaheite) there is an extinct crater at a height of many thousand feet; in Hawaii (Owhyhee), the enormous crater of Kirauea has been described by Mr. Ellis in his Missionary Tour. After traversing a vast surface of consolidated lava, the crater was seen at a distance, in a vast plain, sunk below a high precipice, which encircled the plain with a rugged border 15 or 16 miles round. The crater was of a crescent shape, 2 miles long (N. E. and S.W.), 1 mile broad, and 800 feet deep. Lava melted, and in violent agitation, filled this singular furnace, round which fifty -one conical islands rose, twenty two of them emitting smoke and flame, and many ejecting great streams of lava, which rolled down to the fiery gulf below. This had evidently, at some time previous, been full to its edge, and was now partly emptied by lateral subterranean discharge. All Hawaii is of volcanic origin.

Geological Age of Volcanos.

Volcanos, properly so called, may perhaps have existed during all the periods of geology marked by the succession of stratified rocks; but volcanic eruptions on the surface of the land or bed of the sea are rarely known by their effects previous to the commencement of the tertiary eras. Perhaps the earliest certain exceptions to this generalisation have been found by Mr. Murchison, and Mr. De la Beche, in the silurian and Devonian rocks. In each of these instances, the evidence of volcanic eruptions such as are here meant, is found in the occurrence of layers of volcanic sediments, analogous to trass and puzzolana, in alternating succession with the ordinary deposits of water. Such facts abound on the borders of the Malvern Hills, the range of the Caradoc Hills, the Corndon Hills, and among the trappean rocks which border the granite of Dartmoor.

If the great masses of basalt, which, under the name of the whin sill, are interstratified with carboniferous limestone in the mining dales of Northumberland, were (as Mr. W. Hutton believes) spread out like lava on the bed of the sea, the occurrence of volcanic eruptions is proved from the early carboniferous eras. The mixture of porphyritic pebbles and red sandstone in Germany, on the east side of the Harz, and near Exeter, seems to render plausible the conjecture that volcanic eruptions were not infrequent during the pœcilitic era. (See De la Beche s Manual, 2d edit. p. 365.)

The trap rocks of Skye and other islands on the west of Scotland, which are in contact with lias and other rocks of the oolitic system, can scarcely, upon good grounds, be admitted as originating in volcanic eruptions: they are mostly unerupted lavas.

In tertiary periods of geology, traces of eruptions became frequent. According to Lyell, the oldest volcanic rocks of the Limagne d'Auvergne belong to the eocene tertiary period, being associated with freshwater strata at Pont du Chateau near Clermont, and in the Puy de Marmont near Veyres. None of the volcanic eruptions of Central France had, however, commenced when the older subdivisions of the freshwater groups originated.

The newer portions of the Mont Dor and Plomb du Cantal are stated by this author to be of meiocene date, as well as some cones which stretch from Auvergne, through Velay, into the Vivarais, where they are seen in the basin of the Ardeche. Finally, Etna, which commenced its operations during the newer pleiocene era, has continued them down to recent times with undiminished energy.

Had we included in this review the cases of unerupted lavas (basaltic and porphyritic dykes and interposed masses), there would have been an unbroken series of igneous products, cooled in subterranean, submarine, or subaerial situations, from the earliest primary eras down to the present day; and from the whole we should clearly see how very probable, or rather certain, it is, that granitic and other plutonic, as well as volcanic, rocks are not so much the products of the particular times as of the particular circumstances in which igneous action has been manifested.

Volcanic Eruption Forces. Earthquakes.

The degree of mechanical pressure under which lava is effused, and ashes are ejected, from volcanic vents, is of importance in the theory of their action; and when combined with indications of the same kind in earthquakes, enters among the data requisite for comparing the agencies of subterranean movements in ancient and modern periods.

Ejections of Ashes and Stones.—The distance to which these are transported after leaving the volcano, is a useful indication of the quantity ejected, and thereby of the general power of an eruption, but not a measure of its momentary violence. During the eruption of Vesuvius in 472-473, the ashes thrown out were transported by the winds even to Africa, Syria, and Egypt, and fell in Constantinople. In 1631, ships were covered with ashes 20 leagues from Vesuvius. In 1812, the eruption of the Souffrier Mountain, in St. Vincent's, gave forth ashes which were carried by the winds to Barbadoes. During the terrific eruption of Tomboro, in Sumbawa (1815), clouds of ashes obscured the sun, and fell, inches deep, on the streets and houses in Java, at a distance of 300 miles.

The intensity of the volcanic force can be better appreciated by the magnitude of the stones ejected from the crater, and the height and distances to which they are thrown, than by any other criterion. It appears that stones 8 lb. in weight were thrown from Vesuvius to Pompeii, a distance of 6 miles; and stones were observed by sir W. Hamilton to be thrown so high above the mountain top, that they occupied 11″ in falling, which gives a height of 2000 feet, and an initial velocity of above 350 feet in a second. In 1798, during a violent eruption in Teneriffe, the mountain Chahorra threw out stones to such a height that from 12 to 15 seconds were reckoned during their descent. The height was consequently from 2500 to 3600 feet, and the initial velocity from 380 to 480 feet per second. The pressure of a whole column of lava, which should overflow the crater of Tenerifie, would, according to Daubuisson, be equal to 1000 atmospheres, and might eject lava, at the base, with a velocity of nearly 850 feet per second. These forces are much inferior to those with which cannon balls are projected. The intermitting character of these "fits" of volcanic violence is favourable to the notion of their principally depending on the sudden evolution of the force of steam, with whose operation in this way we have been familiarised by the steam gun of Mr. Perkins.

The formation of New Mountains is another phenomenon which strongly indicates the importance of volcanic operations in changing the aspect of the globe. The cases are numerous. In 1538, in or near the site of the ancient Lucrine Lake, in the Bay of Baiæ, the Monte Nuovo was thrown up, in 48 hours, to a height of 440 feet, with a circumference of 8000 feet, from a crater of eruption, which has been measured to the depth of 418 feet in the middle. In 1669, the Monte Hossi was thrown up on the slope of Etna, 450 feet in height, and 2 miles in circumference; this was accomplished in three or four months. The formation of Jorullo, in 1759, to a height of 1695 feet, is one of the most remarkable effects of this kind. (See page 204.)

The New Islands which have been raised from the sea by volcanic explosion or movement of the sea bed, furnish additional facts; and probably a large proportion of these striking phenomena is unrecorded, and many more must pass away without notice, notwithstanding the increased facilities which extended commerce and general scientific education have afforded for recording them in future.

The changes which have occurred in and about the Island of Santorini, from an epoch 237 years before Christ, to almost the present year, are remarkable, the general effect being an augmentation of the land. The new island of Sciacca, which appeared in July, 1831, and disappeared in the course of the following winter, is one of the most interesting events of this kind known in modern times. It appears that a line of earthquakes may be traced from Corfu, by Calabria, to Etna, which, in its extension westward, nearly strikes the volcanic island of Pantellaria. Between Pantellaria and Sicily, on this line, submarine movements were noticed in June, 1831: soon afterwards the signs of an eruption were seen by Neapolitan fishermen; and on the 18th of July, a British man of war passing near the spot, white columns were seen in the sea, rising from a dark hillock, which threw up stones and ashes. It was then judged by captain Swinburne to be 70 or 80 yards in diameter, and about 20 feet high. In August it had grown to a circumference of 3240 feet, its height being 107 feet; and in the middle was a crater 780 feet in circumference; the columns of ashes rose to a height of 3000 or 4000 feet. The evolution of gases was inconsiderable. When examined, the mass of the island was found to be a dark vesicular lava, with a few fragments of limestone, and other non-volcanic rocks. On the 28th of September, according to Prevost, the circumference of the island was 2300 feet, and the height from 100 to 230 feet. In the winter of 1831-2, its loose and perishable fabric had yielded to the action of the waves, and disappeared from the surface. It is now a dangerous shoal, shelving gradually to the deep sea bed (100 fathoms), out of which it originally sprung; on the neighbouring parts of the sea bed, probably, a considerable deposit of volcanic sediment is spread. Such is the history of the vanished island of Sciacca.

In the Azores, in 1628, an island rose from 160 fathoms water, in 15 days, to a height of 360 feet above the sea; Mr. De la Beche has found in the MS. of the Royal Society, a notice of another island, which had been thrown up in 1690, but soon afterwards, like Sciacca, was dissolved and sunk again in the sea.

In 1811, off St. Michael's, in the same group of islands, a volcano was observed to be active in the sea, on the 13th of June. On the 17th it shot up black columns of cinders to the height of 700 or 800 feet above the sea, and at other times clouds of vapour; the eruptions being accompanied by great noises and vivid lightnings. On the 4th of July, the island which was formed was 1 mile in circumference, almost circular, and about 300 feet high; the crater discharged hot water. This island, to which the name of Sabrina was given, disappeared like Sciacca.

In 1783, a new island rose in the sea near Reykiavich, in connection with the Icelandic volcanic system: it was 1 mile in circumference, but soon disappeared like so many of these already mentioned.

The ejections from the summits and sides of volcanos go to enlarge the mean diameter of the globe, whether they be heaped on the land or laid on the bed of the sea. The amount of this augmentation of diameter has never been estimated (we believe), nor would the estimate, perhaps, be worth the slight trouble of the calculation, were it not useful to moderate the false impressions which a contemplation of the violence of ignivomous mountains occasions. If we suppose the volcanic lines and groups known on the globe to be collected in one line, it would not equal a great circle of the sphere. If we take as the breadth of this volcanic band a surface of 10 miles, we shall much exceed the average. To assume that half the mass of active or extinct volcanic mountains above the sea is the product of subaerial or submarine eruptions is an ample allowance. Finally, if the figure of the mixed volcanic and rocky mass be taken as a series of cones, 2 miles in height, which is far above the truth, the mean volume of igneous products resulting from the calculation is $\textstyle {\frac {24000\times 10\times 2}{3\times 2}}$ miles = 80,000 cubic miles; which, if spread over all the globe equally, would augment its diameter about 2¾ feet. Now, as all the conditions have been taken in a sense the most favourable for the magnitude of the result, we see how feeble, after all, is the change of the general conditions of the globe, produced by the agents of violence put in action during volcanic excitement.

The cavities left within the globe, by the ejection of this mass of matter, are probably so circumstanced by the overarching of their roofs, that they may resist for a long time the tendency of the superincumbent weights to fall in; but there is a limit to this resistance. When the superficial accumulations are of vast height and great lateral extent, as in some of the mighty volcanos of America, the internal heat rises upward, in the substance of the mountain, so as to re-absorb the base of the cone, and weaken its strength. From this cause, perhaps, it happens that sometimes volcanic mountains fall into the cavity below them, and are swallowed up. Thus the great mountain mass of Papandayang, in Java, fell into the greater cavity out of which it had been raised; and l'Altar, in Quito, lost its commanding summit.

The subterranean connection of even distant volcanic mountains, and the reciprocity of action between what appear on the surface to be distinct volcanic groups, justify the belief that the sources from whence the eruptions are supplied with mineral matter spread widely around the volcanic vents; an inference still further strengthened by the extension of earthquakes beyond the regions of burning mountains. It follows that movements of subsidence, which are occasionally witnessed in really volcanic districts, may, and indeed must, happen sometimes in other situations, where lines or surfaces of weakness exist, in the earth's crust, Such depressions may be either gradual or sudden, according to the circumstances which determine the points and degrees of relative weakness in the earth's crust, and from all the considerations it is easy to perceive that the real change of the earth's diameter, by the explosive action of volcanos, is very small, and much counterbalanced, in all periods, by the contrary effects of subsidence; and that in the progress of volcanic operations a limit must at last be reached, when the two opposite effects of the same cause must be exactly balanced, though not necessarily in the same physical regions. The general result, then, is an augmentation of the heights by volcanic energy, and a deepening of the depths by the consequent subsidence.

Far from the centres of volcanic excitement, the compensating depressions of the earth's crust would probably be gradual and almost insensible: in such situations there may also occur equally gradual and almost imperceptible elevations of particular tracts of land, because if there be a real sinking over lines and surfaces of weakness, there will be relatively a rising over points having the contrary properties. There may also be a real rising of such parts, with a relative sinking of others, if the arrangement of the rocks is such as to give maxima of strength in opposite directions. The ordinary and well-known forms of anticlinal and synclinal axes are examples of such figures; for an upward general pressure, such as accompanies volcanic violence, may more easily extend and raise an anticlinal mass, and a subsequent general collapse would act with more force on the synclinal surfaces of stratification. Other causes concur to augment these effects, which are certainly exemplified in observed phenomena of the relative levels of land and sea.

In conformity with this reasoning, we find, on the testimony of all writers who have examined the history of earthquakes, that they are by far most abundant and most violent, in countries which surround or lie between volcanic districts. Before and during volcanic excitements, earthquakes abound, so as evidently to make part of the same phenomenon; and, even under countries where volcanic fires are dormant or extinct, these convulsions of the solid framework of the earth are more powerful than in remoter districts. It is in volcanic countries that proofs have been found of the real displacement and positive elevation of land, on particular days, and during particular earthquakes; while at points far remote from Vesuvius and Hecla, the land is slowly rising in Scandinavia, perhaps slowly sinking in Greenland, perhaps alternately elevated and depressed on some parts of the shores of Britain.

Examples of permanent displacements of land, arising from convulsive movements near the seats of igneous activity, are furnished by the Calabrian earthquakes of 1783, the Lisbon earthquake of 1755, the Chilian earthquakes of 1822 and 1835. In 1822, according to Mrs. Graham, the Chilian coast was agitated by a movement which extended in length 900, 1000, or perhaps 1200 miles (including Copiapo and Valdivia), and raised the whole line of coast for a distance of 100 miles; at Valparaiso 3 feet; at Quintero 4 feet; the greatest movement being about 1 5 miles N. E. of Valparaiso: the beds of oysters and other shells were raised clear to the surface. The whole region between the Andes and a line far out in the sea is supposed to have been permanently raised, 2, 3, or more feet (in the interior the elevation is said to have reached even 7 feet). The area under which, ashore, the earthquake extended, is estimated at 100,000 square miles.^[4] If, as Mr. Lyell supposes, the whole of this vast area was raised, and the elevation be taken at 1 foot on the average, the whole augmentation of the earth's diameter caused by it will be $\scriptstyle {\frac {1}{4000}}$th part of that which we attribute to the whole mass of visible volcanic accumulations on the surface. It is unnecessary to re-open the discussion of the accuracy of the data above assumed, because in 1835 similar phenomena happened on another part of the same coast.

This second great disaster on the Chilian coast has been described by Mr. Caldcleugh, from his own and other observations, with much care. It was heralded by the landward flight of immense flocks of sea-birds (the same thing occurred previous to the shock of 1822), and by the remarkable activity of the volcanos of the Andes, An enormous wave, rising 28 feet in height, destroyed Talcahuano, and was followed by a greater. Columns of smoke rose in the sea, followed by whirlpools. In the Bay of Conception the strata of clay slate were elevated about 3 or 4 feet. At San Vicente, a port a little south of Talcahuano, the land rose about 1 foot and a half. In the small island of Santa Maria, the rise was estimated by Captain Fitzroy at 8, 9, and 1 feet! At Nuevo Bilbao, 70 leagues north of Concepçion, the earthquake was violent, but there is no permanent elevation of the land. Throughout the entire provinces of Canquenes and Concepçion, the crust of the earth has been rent and shattered in every direction. An hundred miles from the coast vessels experienced the shock. The island of Juan Fernandez was included in the area of the submarine disturbance, which below the land reached northward as far as Coquimbo.

It is remarkable that Acosta speaks of very similar effects of waves and violent movements in the same range of coast, in the 16th century; and other instances have been collected by Mr. Woodbine Parish.

Though, for reasons before stated, we cannot expect to find cases of sudden depression in volcanic regions so frequent or extensive as those of elevation, enough is known to assure us that in and beyond these regions, earthquakes have very often caused subsidence of land. We read, that in the year 541 Pompeiopolis was half swallowed up in an earthquake; that in 867 Mount Acraus fell into the sea; that in 1112, the city of Liege was flooded by the Meuse, and that of Rotemburg on the Neckar was ruined. In 1186, a city on the Adriatic shore is described as sinking into the sea; in 1596 the sea covered many towns in Japan; in 1638 St, Euphemia became a lake; and in 1602 Port Royal is commonly believed to have sunk. In 1755, the great earthquake caused a new quay at Lisbon to subside, and its place was occupied by water 100 fathoms deep, and other similar cases of engulphment occurred on the Portuguese and African shores. In 1819, extensive subsidence occurred with the submersion of a town and large tracts of country, at the mouth of the Indus, and in the same vicinity rose a compensating elevation, called "the Ullah Bund."

That earthquakes are experienced over regions far from volcanic mountains is easily ascertained by consulting the imperfect records which have been preserved of these phenomena in Europe. For it thus appears that in Norway, Scotland, England, Belgium, and many parts of Germany and France, considerable earthquakes have occurred, not only at a distance from European volcanos, but also without any definite relation of time to the eruptions of the Icelandic or Mediterranean volcanos. In a long catalogue which we have drawn up for the purpose of comparing the dates of earthquakes in Great Britain with the recorded eruptions of those volcanos, &c., the last 1000 years, we have found scarcely any accordance.

The movement of the ground during an earthquake is described variously,—as a vibration, a rolling, an undulation, a shock; but it is to be regretted that these terms do not always convey a definite and exact notion of the kind of disturbance which really takes place. Some observers speak only of vertical movements, such as were experienced during the Lisbon earthquake by a vessel far west in the Atlantic; others mention horizontal movements, as during the Chilian earthquake of 1835. In general, there is an impression that the movement of the ground travels in one certain direction, like a wave upon water; this direction was remarked to be different in the northern and southern portions of country shaken in Chili in 1822. There is sometimes one shock, in other cases several, seldom many successive impulses from below. The most violent movements appear to have been experienced on the sea-side, and in the sea itself, which, retiring and returning with mighty waves, 10, 20, or even 60 feet high (in the Lisbon earthquakes of 1755), produce incalculable mischief and destruction of life and property.

Were the globe a solid mass at great distances from the seat of the original disturbance, these effects could not happen, unless, as Mr. Mallet has shown, a wave of elastic compression were generated, which should travel like a great wave of translation in water, with velocities corresponding to the elasticity of the rocks, so as to reach Lisbon, Loch Lomond, Italy, and the West Indies. Rocks, we know, are elastic in their parts, but very imperfectly so in their mass, owing to the numerous divisions which intersect them. Earthquakes cannot be compared to the vibrations of a string, or the pulsations of sound, gradually falling to rest; the motion observed is more similar to the undulation of a flexible lamina over an agitated liquid;—as when a long cloth is shaken in a particular manner, so that a wave of air travels below its parts successively to the end.

Mitchell, to whom physical geology is largely indebted, was the first to explain earthquakes by wave motion, and he employs for the purpose the mechanism of a fluid thrown into undulation, or vapour operating by expansion beneath or between the strata.^[5] He assigns 1750 feet per second for the velocity of the Lisbon earthquake. Professors H. D. and W. B. Rogers, following in the same track, make the phenomena of earthquakes depend on undulations propagated in molten rock below the solid crust, trace the path of some of these phenomena, and give measures of the rate of progress of the wave: viz., 27 to 30 miles per minute, or about twice as fast as the wave of sound in air. They find for the velocity of sea waves generated by the earthquake shock, 3½ and 5 miles an hour.^[6] They find the area agitated by earthquakes at any one epoch to be very long and narrow, corresponding to the great wave of translation, and trace the synchronous lines of movement for several hundred miles in length.

Mr. Mallet, in a paper communicated to the Royal Irish Academy^[7], 1 followed by a Report to the British Association, has entered fully on the dynamics of earthquakes, and on the history of these phenomena; and has performed some capital experiments on the rate of movement of earth waves in incoherent sand, and in granite of perhaps the average degree of consolidation.^[8]

From the point where the earthquake originates, two sets of waves proceed in the solid crust of the earth, viz., the wave of elastic compression, propagated in every direction with a velocity proportioned to the elasticity and density of the parts of the earth-crust in its path. In different sorts of rock the velocity will not be the same: it will be greatest in the solid, and least in the loose aggregations of matter. Another set of waves is that of sound. And, if the origin of the earthquake be under the sea, a water wave of translation will be generated in the sea, of much less velocity than that in the earth. Sound waves will be communicated to the water and to the air; but of these we need not say much. If the earthquake originate under the land, and be propagated under the sea, it will reach the extreme border of the sea, and raise the shore so as to force the water to appear to retire, and afterwards to return and flow higher than before, a phenomenon distinctly observed. Supposing the first shock to have happened under the sea, and all the waves to be noticed on the extreme edge of the water, we should have,

1 . The earth waves of shock and sound together, or nearly so.

2. The forced sea wave lost upon the beach.

3. The sound wave through the sea.

4. Sound waves (possibly) through the air.

5. The great water-wave, which has been found so destructive.^[9]

According to Mr. Mallet, the velocities to be expected in the sound-wave would be 4700 feet per second in water, 1140 feet in air; and, judging from the elasticity, in lias 3640; in coal measure sandstones 5248; in oolite 5723; in primary limestone 6696; in carboniferous limestone 7075; and in hard slate 1 2,757; and in granite and igneous rocks still higher rates. Perhaps the speed of the great earth-wave may be nearly the same; but the masses of rock in the earth are so much interrupted by joints, by unequal condensation, varying inclination, and other circumstances, that as appears in the case of the Lishon earthquake, in Mitchell's, Rogers's, and Humboldt's estimates, (⅓ to ½ a mile in a second,) the real velocity is much less. Mr. Mallet has ascertained it in the case of sand and granite to be even less than the above instances, and his experiments were so arranged in the sand at Killiney and the granite of Dalkey, as to give very accurate results. For his beautiful process the reader must be referred to the Brit. Assoc. vol. for 1852.

It appears very desirable, for the completion of this theory of earthquakes, to carry out the seismometrical observations recommended by the British Association, especially at the great public observatories.

Mr. Hopkins in treating this subject mathematically, has shown how, by proper observations of this kind, the local origin of the earthquake can be determined in depth, as well as in geographical position.^[10]

The force of an earthquake shock diminishing at points removed from its origin as the square of the distance increases, we see how great must have been the shock in the case of the Guadaloupe earthquake (1843), when, as Rogers has shown, an area not less than 2300 geographical miles in length by 770 in breadth was agitated. According to the observations made on this occasion, the shock was simultaneous in lines nearly north and south, and felt moving in opposite directions from a curved central axis, at the rate of 27 miles in a minute. This seems to indicate a linear subterranean fracture of great length—a fault geologically speaking—such as might occur over a cavity left by the withdrawal of a fluid support to the earth's crust.

By a mathematical investigation of this subject, founded on the phenomena of precession and nutation which arise from the action of the sun and moon on the unspherical mass of the earth, Mr. Hopkins has shown that whether the earth be partially fluid or wholly solid within, there would be no material difference in the precession and nutation, provided the ellipticities of the interior and exterior surface of the supposed solid crust were equal, and the density of the crust and fluid equal and uniform. But if these limitations were not observed; if the solid shell and interior fluid were heterogeneous, and the ellipticities of the interior and exterior surface of the crust different, then the amount of the precession and nutation would depend on the difference between the ellipticities of the interior and exterior surface of the crust, and on its thickness: or on this latter quantity alone, if the solidity of the shell resulted from refrigeration. And the result of the whole inquiry appears to be, that the thickness of the solid crust cannot be less than $\scriptstyle {\frac {1}{4}}$th or $\scriptstyle {\frac {1}{5}}$th of the radius of its external surface.^[11] This conclusion is probably decisive against any universal ocean of molten rock below us, at depths accessible to the disturbing agents which generate earthquakes and volcanos: but it seems not to preclude the admission of limited fluid masses, at various and far smaller depths than 1000 or 800 miles. In harmony with this view is the opinion of Mr. C. Darwin, who, from considering the circumstances which accompany volcanos and earthquakes in the Cordilleras of the Andes, proposes, as a fundamental point of reasoning, the recognition of the existence of a vast internal sea of melted rock below a large part of South America.^[12]

This conclusion appears liable to so little objection; it is, besides, so perfectly in harmony with the fact historically proved of the perpetual readiness of volcanos for action, and with the geological inference of the perhaps unlimited extent below our feet of rocks once fused; that we shall venture to adopt it as a datum sufficiently established, and applicable to the whole series of volcanic phenomena, in every country, and during all past periods of time.

But this ocean of melted rock may sleep, and does remain at rest, beneath enormous areas, for centuries, or much longer periods, till some particular causes concur to "change (as Mr. Darwin expresses it) the form of the fluid surface," and develop extraordinary chemical energy and fearful mechanical violence. What are these causes? and what is the condition of the subjacent fluid masses whose repose they disturb?

Hypotheses of Volcanic Action.

To answer the questions just proposed, is the object of a just theory of volcanic action. The conditions already established, of the great extent of the phenomena, the appearance of volcanic fires in every kind of rock, and the continuity of such operations not only through historical but through earlier geological periods, negative completely the trifling notion of any particular combustible substances, or decomposable chemical compounds, being sufficient to maintain such long-enduring and powerful operations of heat. We must adopt larger and yet more definite views on the subject. No supposition will be of the smallest value, which provides an agency inferior to the area, unequal to the mechanical violence, or inconsistent with the chemical characters of volcanic excitement.

Accordingly, only two hypotheses have been deemed worthy of attention in the modern consideration of this subject. Humboldt, Cordier, and other eminent geologists, reviving the opinion of Leibnitz, look upon volcanic action as the necessary result of the influence exerted by the heated interior upon the cooled exterior masses of the globe. If the earth be now generally hot within, it must formerly have been hotter; in the process of cooling, the exterior solidified part and the interior fluid parts contract unequally, a general pressure and tension result, and the crust breaks locally to restore the equilibrium. Hence earthquakes, and fissures, on some of which volcanic vents are established, which serve more or less to relieve the subterranean pressure, as earthquakes also do. If, in addition to this general view, we suppose the admission of water through fissures to particular parts of the "ocean of molten rock, "it is easy to see that the observed mechanical phenomena of volcanos and earthquakes will result as the effect of a local excitement superadded to a general operation. Such is an outline of the explanation offered by the hypothesis of a general heat pervading the interior of the globe.

Mr. Darwin, in his summary of the phenomena attending earthquakes on the coast of Chili, in 1835, regards, very justly, the submarine outbursts, the renewed volcanic activity, and the permanent elevation of the land, as forming parts of one great action, and being effects of one great cause, modified only by local circumstances; and that, therefore, "no theory of the cause of volcanos, which is not applicable to continental elevations, can be considered as well established." This appears a just inference. He is further of opinion that the following conclusions may be drawn from the phenomena of earthquakes.

1st. That the primary shock of an eathquake is caused by a violent rending of the strat, which, on the coast of Chili and Peru, seems generally to occur at the bottom of the neighbouring sea.

2d. That this is followed by many minor fractures which, though extending upwards, do not, except in submarine volcanos, actually reach the surface.

3d. That the area thus fissured extends parallel, or approximately parallel, to the neighbouring coast mountains.

Lastly. That the earthquake relieves the subterranean force precisely in the same manner as an eruption through an ordinary volcano.

Now every thing here said may be adopted, without hesitation, into the general speculation of Humboldt, of which, in fact, these inferences from observation are strongly illustrative.

Another view, which is strongly supported, is usually considered by its defenders as "the chemical hypothesis" of volcanic action. It presented itself both to Davy and Gay-Lussac, as a natural consequence of the discovery of the metallic and metalloid bases of the earths and alkalis; and though the former eminent philosopher abandoned his speculation, it has found able support in Dr. Daubeny. The account given by Daubuisson will clearly exhibit the opinion of Gay-Lussac. "If we admit, what is in fact almost certain, that water enters in great quantity to the foci of volcanos, and there comes in contact with the metalloid bases of the earths and alkalis, and some chlorides (especially the chloride of sodium), the following effects will happen: One part of the liquid will be quickly decomposed; the metals and the chlorides will seize oxygen, and be thereby converted to silica, alumina, lime, magnesia, soda, &c.—substances which predominate in lavas; the hydrogen will be liberated in the state of gas, or in combination with chlorine will form hydrochloric acid, which is known to be very often present in the vaporous exhalations of volcanos."^[13] The heat generated by the primary chemical action (oxygenation) and the energetic action of steam, to which part of the water is converted, are thought sufficient to account for the mechanical phenomena of volcanos.

Dr. Daubeny has given to this speculation a character of greater completeness, by an examination of the actual products of volcanos, for comparison with a regular deduction of chemical phenomena from the fundamental postulates of Gay-Lussac and Davy.

If, at a depth of 3 or 4 miles, the nucleus of the earth consists of the metalloid bases of the earths and alkalis, with iron and other metals, partially combined with sulphur,—the new oxygenation to which, under ordinary conditions, they would be subject, may be productive of no other phenomena than a moderate rise of temperature in the neighbouring rocks or in thermal springs.

But with access of water, and especially sea water, the effects of the heat generated will become more formidable. Oxygenation on an extensive scale; evolution of a large volume of hydrogen, again to combine with oxygen (supposing atmospheric air present), or with sulphur, at a high temperature. In the former case, nitrogen gas will be liberated, which may rise uncombined, or may unite with hydrogen to form ammonia; and this will be neutralised by free muriatic acid, and produce sal ammoniac.

The hydrogen not thus disposed of may combine with sulphur to form sulphuretted hydrogen gas; but this may be again decomposed by rising and meeting with oxygen; as long, therefore, as oxygen abounds, there will be evolution of water and sulphurous acid; afterwards sulphuretted hydrogen will prevail toward the end of the eruption. As long as heat remains in the lava, the combustion of sulphur, and the decomposition of the sulphurous acid by sulphuretted hydrogen, would regenerate water, to maintain, by combination with metals and metalloids, a continuance of similar though feebler actions.

There is not, we believe, any attempt on record to deduce all the chemical phenomena of volcanos from the hypothesis of general heat below the surface of the earth: we must therefore, at present, suppose this is difficult, except upon the admission of that powerful absorption of oxygen from water, which the "chemical" hypothesis provides. Granting, then, the truth of these opinions as to the origin of the substances ejected from volcanos, do they involve the rejection of the hypothesis of a pervading high temperature below the surface of our planet? Surely not.

For what account does the peculiar series of gaseous and earthy ejections from a volcano give of the origin of the volcanic action? What opens the fissure and gives passage for the water to the base of volcanic mountains? The whole crust of the globe, stratified and unstratified, is a mass of metallic oxidation; how can there yet remain, at so many points, access for water through this oxidated crust, to the unseen primitive nucleus? How happens it, that really volcanic effusions are so limited and so few among the older strata, which were formed when the stratified crust of the globe was thinner, and (by this hypothesis) the unsaturated metalloid bases were more plentiful near the surface?

It appears to us very clear, that the union of the two speculations here brought into comparison is not only practicable, but reasonable, and even necessary. A general cause of change of form of the earth's surface and interior parts is supplied by the doctrine of a change of interior heat; abundant admission for water is afforded by the fractures necessary (upon this view) to adjust the balance of pressures; and the chemical products can only be properly understood by a suitable hypothesis of chemical action. The interior mass of the globe may yet retain the uncombined bases of earths and alkalis; but the chemical products resulting from admission of oxygen to these are not at all the less intelligible, if we suppose the whole mass of the interior to have those general conditions of heat which appear to suit the mechanical disturbances of the land and sea. On this point, however, further researches on collateral phenomena may be prosecuted with advantage, and to these we now proceed.

Thermal Springs.

In general, the springs which issue from the earth derive their origin from rain which has descended through fissures of the rocks (especially calcareous rocks), and, in consequence of meeting with natural impediments to further descent,—as beds of clay, dykes, mineral veins,—faults, collects in the rocky reservoirs, rises to the surface, and issues at the point to which access is easiest, whether it be the lowest point of the vicinity or not. The rains which supply such springs descend irregularly; yet, if the subterranean reservoirs be considerable, the discharge is nearly constant in all parts of one year, and in many succeeding years. To each of such springs usually one particular chemical quality is imparted by the rocks through which the water passes; and one particular average temperature belongs to each, generally identical with that of the ground through which it passes.

This temperature seldom differs much from the mean annual heat of the locality, and, unless the stream be subject to variation of quantity, hardly varies with seasons or years.

It has, however, been found that the small differences which appear between the mean temperature of the air and of springs at particular localities, are of a somewhat regular character, and bear a general if not a precise relation to latitude. It was found, for instance, by Dalton (1793), that the springs at Kendal gave a somewhat higher range of temperature than the air: the same observation has been made at Berlin, Paris, and other places in the North Temperate Zone; but in the equatorial region the contrary appears to be the fact. The tables of Kupffer (which may be consulted in De la Beche's Manual of Geology), constructed from observations of Humboldt, Von Buch, Cordier, Wahlenberg, Kupffer, &c., appear to give as much as from 1° to 5° superiority of air temperature above that of the ground; while in latitude 54° to 60°, in Russia, the springs were warmer than the air by 5° or 6°.—This fact appears to show clearly that the temperature of the earth and of springs is influenced by some general cause independent of solar heat.

Besides the class of ordinary springs, which may thus differ by small amounts from the temperature of the air, there are "thermal springs" which often deserve the name commonly assigned to them of "hot springs," and sometimes approach even the boiling point. These are usually found to be almost, or even absolutely, constant in their discharge, uniform in their temperature, and unvarying in their chemical composition. Some of the sources frequented by the luxurious nations of antiquity still retain their efficacy,—in Greece, in Belgium, and at Bath; and the inquiry into the cause of this continued heat becomes the more important when we consider the great geographical area over which hot springs are scattered, the singularity of their association with cold and mineral waters, which is often noticed, the variety of their contents, and the geological circumstances which accompany their efflux.

It is unnecessary to dwell at any length on the question, how far any peculiar chemical quality is characteristic of hot waters, so as to offer a satisfactory explanation of their warmth from chemical action. There is no such peculiarity. Thermal waters are found to be, on the average, neither more nor less pure than springs of common temperature; they exhibit, in fact, the same scale and variations of chemical constitution as common waters. The chemical quality of hot waters, offers no explanation of their heat, though, combined with other considerations, it may help to guide to a right view of the manner in which that heat has been acquired.

There is no one product of thermal springs, constantly found in them, which never occurs in cold waters; but it appears from Dr. Daubeny's important researches, that nitrogen gas is very common in hot springs, and perhaps very rare in cold waters. This circumstance appears to him of great importance in the argument whereby he connects the origin of hot springs with volcanic action. In Dr. Daubeny's admirable Essay on Mineral and Thermal Waters^[14], the catalogue of thermal waters exhibits the prevalence of nitrogen, among the gases evolved, in a striking degree; carbonic acid is also plentiful, and, in particular districts (Nassau), predominant. As examples, we may select the notices of the warm springs of the British islands, and of those which adjoin the Ardennes and Nassau mountains,—in both instances only obscurely dependent on volcanic formations; the Pyrenean and other springs may also be noticed.

Warm Springs of the British Islands, Yielding Nitrogen, &c.

1. Bath.—The King's Bath spring rises through lias^[15], at a temperature of 66° above that of the neighbourhood; contains saline ingredients, 15 grains in a pint (muriate of lime and magnesia); evolves 96.5 per cent, nitrogen, 3.5 oxygen, and some carbonic acid.

2. Bristol.—The Hot Well rises in carboniferous limestone, at a temperature of 25° above that of the place; contains saline ingredients, 6 grains in a pint (sulphate of soda and muriate of lime); evolves 92 per cent, nitrogen, and 8 oxygen.

3. Buxton, Derbyshire.—St. Anne's Well rises in carboniferous limestone, at a temperature of 33° above the vicinity; contains saline ingredients, only 1.8 grains in a pint (muriates of magnesia and soda); evolves nitrogen only.

4. Baltewell, Derbyshire.—The Bath spring rises in carboniferous limestone, at a temperature of 13° above the vicinity; contains saline ingredients, 3½ grains in a pint (sulphate of lime and muriate of soda); evolves nitrogen only.

5. Stony Middleton, Derbyshire.—The spring rises in carboniferous limestone, at a temperature of 14° above that of the vicinity; contains saline ingredients, 2 grains in a pint (sulphate of soda and magnesia, and muriate of lime); evolves nitrogen only.

6. Taafe's Well, near Cardiff.—Rises from coal strata, at a temperature of 21° above that of the vicinity; contains saline ingredients, only 1.2 grain in a pint (sulphate of magnesia); evolves 96i per cent, of nitrogen, and 3½ per cent, of oxygen.

?. Mallow, Co. Cork.—The Spa well rises in carboniferous limestone, at a temperature of 23° above that of the vicinity; contains saline ingredients, only 0.3 grain in a pint (carbonate of lime); evolves nitrogen 93½ per cent., and oxygen 65.

It is very surprising that the only hot springs of Great Britain should all rise through strata of the carboniferous system (mostly below the coal), or through others which rest unconformable upon them.

Warm Springs of a part of Germany, &c., yielding Carbonic Acid, &c.

Aix-la-Chapelle.—The Kaiserquelle rises at the junction of clay slate and carboniferous limestone, with a temperature 85° above that of the vicinity; contains of saline ingredients 32 grains in a pint (muriate, carbonate, and sulphate of soda, &c.); evolves nitrogen 69.5, and carbonic acid 30.

Borset.—The Mühlenbend rises with the same geological relations as the last, with a temperature 121.5° above that of the place; contains of saline ingredients 34 grains in a pint (muriate, carbonate, and sulphate of soda, &c.); evolves nitrogen 80 per cent., oxygen 2, and carbonic acid 18.

Ems.—The Rondul rises in argillaceous slate, with a temperature of 81° above that of the place; contains of saline ingredients 28.9 grains in a pint (carbonate, muriate, and sulphate of soda); evolves carbonic acid gas only.

Wiesbaden.—The Kochbrunnen rises in chloride slate, with a temperature of 108° above that of the vicinity; contains of saline ingredients 57 '6 grains in a pint (muriate of soda, lime, and potash); evolves nitrogen 27 per cent., and carbonic acid 73.

(The above springs all rise in or adjoining the slaty rocks.)

Warm Springs of the Pyrenees.

Those of Arles, Preste, Fernet, and Molitz, in the Dep. des Pyrenees Orientales, having temperatures above the vicinity of 85.3°, 71.0°, 72.2°, and 40°; contain of saline ingredients 2, 1, 1.3, 1.3 grains respectively (sulphuret of sodium, &c.); and evolve nitrogen gas only. They rise from granite.

The following are in the same department:—

That of Sorède, having a temperature above the vicinity of 9°; contains of saline ingredients 6.8 grains in a pint (carbonate, sulphate, and muriate of iron); and evolves carbonic acid gas only.

Those of Reynez, Enn, and Thuez, having temperatures above that of the vicinity of 23.7°, 62.0°, and 71° have almost nosaline contents; and evolve no gases. The two former rise in mica slate, the latter at the junction of granite and limestone.

Those of Enaldes, Dorros, and Los rise at the boundary of granite, with temperatures 47.1°, 44.4°, and 24.2° above the vicinity; contain very little saline admixture (1 grain hydro sulphuret of soda, &c.); and yield nitrogen gas only.

The waters of Barège and Cauteretz, in the Pyrenees (51.9° and 70.1° above the temperature of the place), rise in primary rocks, and yield nitrogen only.

The baths of Loueche (74.1° above the temperature of the place) yield nitrogen only.

To complete this view of the chemical characters of hot springs, we may notice some of those which rise in volcanic countries.

At Mont Dor, Caesar's Bath rises in trachyte, with a temperature 52° above that of the country; contains of saline ingredients 11.4 grains in a pint (carbonate, muriate, and sulphate of soda); and evolves 9.85 nitrogen, 0.85 oxygen, and 90 carbonic acid.

The springs of Chaudes Aigues, near Aurillac, rise in gneiss, with a temperature 118° above that of the place; contain 14.5 grains of saline ingredients In a pint (carbonate and muriate of soda, magnesia, lime, and oxide of iron); evolve from 12 to 30 nitrogen, 1 to 15 oxygen, 57 to 87 carbonic acid.

None of the facts disclosed by chemical analysis of these springs, justify the belief that it is to any peculiar chemical action in their channels that their heat above the atmosphere is owing. On the contrary, their heat is derived by communication from the heated rocks through which they pass, whatever may be the cause of their chemical differences. (See professor Forbes's remarks, Phil. Trans. 1836, p. 576.) That the heat of the rocks, and therefore that of the springs, is derived from volcanic action, appears to Dr. Daubeny probable, because nitrogen gas, so commonly evolved from hot springs, is also a product of volcanos, both subaerial and submarine, and because "the majority of thermal waters arise, either from rocks of a volcanic nature, from the vicinity of some uplifted chain of mountains, or, lastly, from clefts and fissures caused by disruption."

These arguments, when taken in connection, appear to us to prove that the heat of the springs is derived from the depths of the channels in which they flow below the surface. The presence of nitrogen may establish the existence of substances, at considerable depths, capable of decomposing atmospheric air; but when we find that in volcanic Ischia a whole group of springs yields no nitrogen, and that it is not in volcanic regions, but on the borders of granitic elevations, and fractures of ancient strata, that nitrogen is most uniformly the predominant gaseous product, it seems unnecessary to appeal to local volcanic excitement for an effect which spreads both in time and area far beyond the traces of purely volcanic phenomena.

That hot springs are numerous in volcanic regions is a certain and even necessary truth; but they appear quite as abundant on the ancient lines of uplifted rocks, like the Pyrenees, where professor Forbes has traced so many to their origin at the junction of stratified and unstratified rocks, that it seems in that region almost an invariable concomitant circumstance.^[16] "The general connection of the hot springs with the granite is so remarkable in that country, as to strike the observer at once; but there are several other peculiarities worthy of note. The abundance of hot springs increases in a very remarkable manner as we advance eastward in the range; nor can any one have a just idea of the prodigal abundance of these thermal waters, who has not visited the departments of the Arriège and the Pyrénées Orientales. Their temperatures are also the highest. In this part of the chain, granitic formations preponderate; yet in almost every case which I have examined, if springs rise in granite, it is just at the boundary of that formation with a stratified rock.

More striking instances of the immediate connection between thermal waters and disturbed strata, than the Pyrenees afford, cannot be desired. The same thing, however, is very generally true; even in England, under the Bath springs, at the Buxton spring, at the Bristol spring, the dislocations of the strata are very remarkable. In connection with professor Forbes's result, Mr. Henwood's curious observation, already stated, that the temperature of the waters issuing from the granite of Cornwall is always lower than that of such as flow from slate rocks at the same depth, deserves to be remembered. This is found to be the case at the surface, and to the depth of more than 200 fathoms.

Thermal springs are thus found to have, as their most general characteristic of origin, a peculiar geological position;—they burst forth (more remarkably than other springs) at points of extreme displacement of the strata, anticlinal elevations, &c., or, in general terms, at points where it is conceivable that a communication exists downward to the regions of interior heat. For this important generalisation we are indebted to Dr. Daubeny.

Further, it appears that these springs are scarcely less abundant or less heated in countries far removed from the regions of powerful volcanic excitement, than amidst active or extinct volcanos. Dr. Daubeny supplies an excellent catalogue of European springs, in his Report to the British Association, 1836; and Mr. De la Beche has collected examples of hot springs in all quarters of the globe.^[17] The following brief summary will suffice for the purposes of reasoning on their geographical relations to existing volcanos.

In the British Islands, the average of 7 springs connected with carboniferous limestone gives an excess of temperature above that of the atmosphere of 28°.^[18]

In Germany, the average temperature of 20 springs is 58.9° above that of the atmosphere.

In France, connected with its central volcanic chain, 12 springs average 69.2° above the air.

In France, connected with the granitic Vosges mountains, 4 springs average 80.2° above the atmosphere.

Connected with the Alps of Dauphine, rising in Jura limestone, &c., 3 springs average 46.7° above the air.

Connected with the Pyrenees, 36 springs average 48.6° above the air.

Connected with the Swiss Alps, 16 springs average 44.8° above the air.

In Croatia, 5 springs average 58.2° above the temperature of the region.

In Styria and Carinthia, 5 springs average 44.6° above the temperature of the place.

In volcanic Hungary, 14 springs average 47° above the temperature of the air.

In volcanic Iceland, springs of various temperatures occur, from the boiling Geysers to a moderate warmth; the hottest being near the site of active volcanos.

In the volcanic island of Ischia, 6 springs average 55.9° above the air.

In Sicily, 2 springs average 55° above the air.

In Italy generally, 19 springs average 52.4° above the air.

In Sardinia, 4 springs average 57.7° above the air.

In Corsica, 2 springs average 59° above the air.

In Portugal, 35 springs average excess of temperature above that of the country about 30°.

In the Caucasus, average of those mentioned by De la Beche, about 60° above the air.

In the Himalaya, on the Jumna River, the temperature of springs appears to exceed that of the air at least 80°.

In China, 3 springs, which issue from granite, probably exceed the temperature of the air from 70° to 120°.

In Japan (volcanic) is a boiling spring.

In Ceylon are springs which exceed the mean temperature about 30°.

On the American continent, the most remarkable collection of springs is on the Ozark mountains; the 70 springs, which here rise in slate rocks, have temperatures which exceed those of the vicinity by 40° to 100°. Others occur in the Rocky Mountains.

In Jamaica, the bath springs in St. Thomas in the East are about 50° above the mean temperature.

It is to be regretted that the information concerning the temperature of hot springs is, in general, insufficient to determine whether they suffer periodical variations with seasons or cycles of years. Until lately, the means of instrumental research were inadequate for delicate experiments such as those required in this branch of study, nor has much been done to furnish future observers with the power of settling these questions. It appears probable that thermal springs may vary their temperatures, because it is an established fact, that a part of the contents of some of them is withdrawn, by cutting off their connection with subterranean springs of cold water.^[19] Variation of temperature is asserted as a fact, in respect of the spring of Gargitello in Ischia, Pfeffers Baths, a spring at Cannea in Ceylon, and Bagnères de Luchon in the Pyrenees. During earthquakes and volcanic violence, thermal springs have been affected, both in their quantity and in their temperature: in 1755, the year of the Lisbon earthquake, the temperature of the Source de la Reine at Bagnères de Luchon was raised 75°. In 1660, a great earthquake desolated the country from Bordeaux to Narbonne, displacing large masses of ground, and caused one of the hottest of the Pyrenean springs to become so cool as to be no longer of any value. (Kircher, Mundus Subterraneus.) On the contrary, two springs in South America, far from any native volcano, have increased in temperature by 4 centigrade, in the interval between an observation by Humboldt and its repetition by Boussingault. (Forbes, On Pyrenean Springs, Phil. Trans. 1836.)

The general conclusions fairly derivable from a study of thermal springs are few, but important. Their heat is not the effect of local causes peculiar to each locality, but is communicated to water which has fallen on the surface, and penetrated to great depths in the earth. Returned to the surface by hydrostatic pressure, these springs bring with them the temperature of the interior, modified and slightly diminished by the comparatively cool rocks near the surface of the earth. This diminution of their heat is perhaps but slight, owing to the feebly conducting power for heat which the rocks possess; yet upon some very small streams it may have a powerful influence. Most of the very warm waters, as those of Bath, Aix-la-Chapelle, the springs of Nassau, and the Pyrenees, are very abundant. To see these rivers of hot water pouring forth for a thousand years undiminished in heat or abundance, is one of the most remarkable and even (as professor Forbes truly says) romantic circumstances which fall under the notice of geology. The conclusion to which they obviously point of the existence of a general heat below the surface of the earth, is indisputable, whether, with Dr. Daubeny, we view that heat as the result of chemical action, and call it volcanic, or, with Humboldt and Arago, regard it as the residue of the original ignition (chaleur d'origine) of our planet.

Experimental Inquiries into the Heat of the Globe.

That the earth has below its surface a source of great heat, independent of solar influence, is perfectly ascertained by volcanic phenomena; that this heat is very generally diffused, is equally certain, from the extent of country in which thermal springs are found; that it is universally spread below our feet, becomes continually more and more probable from experimental researches in countries uninfluenced by any chemical actions supposed to go on at the base of volcanos, where no hot springs burst to the surface, and where the fractures of the strata yield both pure and mineralised waters at common temperatures. Before, however, stating the important facts thus established, it is convenient to direct attention to the conditions of the experiments; for thus the truth and applicability of the inferences drawn from them will more clearly appear.

No truth is more firmly established in meteorology, than the primary dependence of the temperature of each point on the earth's surface upon the calorific influence radiated from the sun. The evidence is found in the conformity of the diurnal and monthly changes of temperature, at each place, to the changing position of the sun, and the proportionality of the annual mean temperatures at different places to the quantity of solar rays received.

Neither of these satisfactory parts of evidence can, however, be completely gathered, except by long averages of years, which neutralise the irregularities of particular years; nor properly understood, without attending to many secondary influences.

The heating influence of the sun, though continually acting, has not been found to have any cumulative effect on the globe; which, upon the whole, has perhaps undergone no perceptible change in this respect since the reach of history; but many parts of its surface have experienced real alterations of climate from drainage, in closures, destruction of forests, and other causes. There is a cooling as well as a heating power constantly at work. The earth is a warm body plunged in a relatively cold medium, for the planetary spaces are cold compared to our globe, and the incessant radiation from the surface of the earth into the vast spaces around is uncompensated by any counteracting influence, though diminished in the cold regions of the world by peculiar provisions of a beneficent Providence.

The temperature of the ethereal spaces around is supposed to be pretty well represented by the minimum of observation on the earth's surface, during the long absence of the sun. It is therefore generally taken at about 50° centigrade, below the freezing point,—a supposition confirmed by some astronomical considerations, and sanctioned by Fourier and Swanberg.

Preserving between their joint effects a variable equilibrium of temperature at the surface of the earth, the calorific power of the sun and the refrigerating influence of the planetary spaces affect every point on the terraqueous globe; and, as far as geographical position with respect to the poles and equator is concerned, the result may be nearly calculated. The mean temperature of any zone of land and sea is, in fact, nearly proportional to the cosine of its latitude.^[20]

But the globe is enveloped in an atmosphere, which produces further modifications of climate, according to the elevation of places above the level of the sea. The sun's rays traverse this atmosphere without heating it; the warmth which it possesses is derived from the earth by conduction, and dissipated by radiation. Owing to the diminution of density in the upper regions of the atmosphere, the air heated near the earth's surface expands into larger and still larger spaces as it rises, and thus the upper parts of the atmosphere have a temperature always growing lower and lower as the density grows less and less. The variations of heat in the atmosphere are greatest at and near the earth's surface; they may become insensible in the upper aerial regions, above the clouds. The cold, thus permanently fixed in the high atmospheric spaces, necessarily reacts upon the land which is raised above the general level of the sea. The temperature of the surface of such land is the resultant of the general influence of the sun, planetary spaces, atmospheric modifications, and conducting power of the ground. In general, the effect of elevation above the sea level in diminishing the heat of the surface of the ground, is nearly in proportion to the height, in all latitudes.^[21] Hence it happens, that as the mean temperature of the equator is about 81½°, the height in the air at which the mean snow line should be found = 49½ x 352 feet = 17,424. (obs. 16,829), and in any other lat. = (81½° N. cos. lat.—32) 352. In W. lat. 56°50′, nearly that of Ben Nevis, this gives (44.6—32) 352 = 4435 feet; and as Ben Nevis is 4350 feet high, and is not covered perpetually with snow (which melts in July and August, except in shaded parts), the rule appears exact enough for the longitude of Britain.

Another cause productive of differences of temperature on surfaces equally exposed to the sun's influence, is the peculiar distribution of land and water; for these dissimilar parts of the globe unequally absorb and unequally give out heat; and one of them diffuses itself so as to obliterate many original differences of climate. Thus, on different circles of longitude, places which, having the same latitude, should have the same mean annual temperature, may, and do, differ in this respect several degrees, from the dissimilitude of the ground, and from the different relations they bear to the masses of land and surfaces of sea. Under the equator the land is generally hotter than the sea; towards the poles the reverse obtains. The sea climate admits of less extreme variations from the torrid to the frigid zone, than the land, and sea-shores participate in this mildness. Thus we have oceanic, littoral, insular, and continental climates, which differ sometimes by several degrees. The formula, therefore, given above (which expresses the average mean temperature in terms of the latitude) requires modification from this cause, as sir D. Brewster has shown in the essay already quoted. From what has been said, it is plain that the principal causes which influence the earth's surface temperature are known. One of the circumstances which mask the regularity of the results, and their real dependence on the position of the sun, is the delay which occurs between the moment of exertion of the greatest heating and cooling power and its visible effect on the surface of the land and sea. In the influence of the moon on the tide, we have an instance of the same kind lately reduced to law: the highest tides take place after the moon has passed her point of power. Just so the warmest epoch of the day is after the sun has crossed the meridian, when most rays fall on the earth: the hottest and coldest epochs of the year follow by an interval of about three weeks (in northern latitudes) the summer and winter solstices. When these allowances of time are made, and the local circumstances previously adverted to allowed for, the coincidence of calculation for hourly, daily, monthly, and annual temperatures, with the result of long continued and regular observation, is surprisingly close, and fully justifies the general conclusion that the earth's surface temperature is the balance of the variable heating energy of the sun and the uniform cooling power of the ethereal spaces in which the earth's orbit is situated. (What effect on surface temperature the peculiar condition of the interior of the earth may occasion, will be seen hereafter.) This being established, we may appeal to observation for proof that it is at the surface of the earth the greatest variations of temperature take place, and from this surface they are propagated upwards with diminishing force into the air above, and into the water and earth below, till in each direction they terminate, or become insensible. The communication of solar heat into the earth constitutes the first branch of our inquiry, and it has been quite sufficiently prosecuted to authorise the following positive statements.

1. By Leslie's experiments, made in 1816, 1817, at Abbotshall, in Fife, with long thermometers, plunged in the earth at depths of 1, 2, 4, 8 feet, their stems rising above the surface, so as to be easily inspected, we find that the variations of temperature continually diminish downwards;—at 1 foot, the extreme monthly differences corresponding to summer and winter were 21° and 19.6°; at 2 feet, 16.5° and 16.3°; at 4 feet, 12.8° and 11.5°; at 8 feet, 8.0° and 8.2.°

2. The epochs of highest and lowest temperature continually differ more and more from the summer and winter solstices, according as the depth in the earth is greater; or, in other words, the time taken by the sun's rays to penetrate and warm the ground augments with the depth.

Thus, at 1 foot from the surface, January is the coldest and July the hottest month; at 8 feet from the surface, February and March are the coldest months, and September the hottest.

3. The average mean temperature of the year augments from the surface downwards; but does not reach the average of the air temperature, in the range of these experiments.

These results have been more than confirmed—they have been enlarged—by the experiments of Arago in Paris, Quetelet in Brussels, and Forbes in Edinburgh, and extended to the depth of 25 feet. M. Quetelet has founded on the experiments at Brussels a mathematical investigation of the highest interest.

Among the data for computation employed by M. Quetelet are experiments analogous to those of Leslie, made in 1762 at Zurich by M. Ott; a series made at Strasburg by Herrenschneider, in 1821, 1822, and 1823; another at Heidelberg by M. Muncke; one made at Upsal in 1832-3, by M. Rudberg; others at the observatories at Paris and Brussels descending to 25 feet. The original memoir^[22] must be consulted for the mathematical part of the subject; but we shall present the conclusions which the investigation has established.

1. In descending from the surface of the earth, to depths continually augmenting, the mean temperature of the year augments gradually; yet, immediately below the surface, and at depths of half a foot or a foot, the mean temperature is found to be a minimum.

2. The rate at which the annual variations of temperature are transmitted to the interior of the earth, may be estimated at 6 or 7 days for 1 foot thickness of earth.^[23]

3. Observation and theory agree in showing that the extreme temperatures of the year decrease in geometrical progression, while the depths below the surface are taken in arithmetical progression.

4. The annual variations of temperature may be considered as insensible at depths from 60 or 75 feet; that is to say, at the depths where the maxima and minima will occur at the same epochs (after an interval of one year!) as at the surface.

5. On descending several feet below the surface, the annual variations of temperature are as the sines of the elapsed times, in a circle whose circumference corresponds to the period of one year.

6. When different latitudes are compared, it appears that the annual variations of temperature penetrate to the least depths in the higher latitudes.

7. The rate with which diurnal variations of temperature are transmitted to the interior of the earth, may be stated at somewhat less than 3 hours for 1 decimetre in thickness (3.9 inches English).

8. The diurnal variations become insensible at a depth of 1.3 metre (51 inches), which is 19 times less than the depth reached by the annual variations, as theory also indicates.

The important conclusion of the entire disappearance of all trace of annual or diurnal variation of temperature at a depth so moderate as from 60 to 100 feet, is perfectly confirmed by the well known experiments in the caves under Paris; and is the more satisfactory, that it falls much within the limits assigned to the annual variation by Fourier in his mathematical theory of heat.

The condition of the interior of the earth below the point of invariable temperature cannot be assumed upon any ground of probability independent of geological observations, nor foretold by any mathematical theory of heat, nor determined by any experiments made at the surface; but may be easily detected by direct thermometric experiments, even at the moderate depths already reached by human enterprise. If the earth be very cold within, the influence of the interior cold will begin to be felt below the depth of 100 feet; if very hot within, the rate of increase of this heat may be inferred from exact and numerous observations.

The experimental inquiries for this object have been prosecuted with great success in Europe, and partially in America, to depths amounting in England to 1584 feet (at Monkwearmouth), and about 1800 feet in Mexico. They consist of three divisions. In the first case, the experiments are made in or very near to mineral veins, which, by their character of filling fissures on lines of disruption, remind us of the general geological conditions of appearance of hot springs; the second set of experiments takes place in collieries and other excavations of like nature, among the stratified rocks, with or without dislocations. In each of these cases, either the temperature of the rock, of air, or of a constant subterranean spring may be tried. In the third case, wells or boreholes are sunk, in a country where little or no water naturally springs to the surface, to considerable depths, and till strong streams of water are let up, bringing with them the temperature of the subterranean regions at those depths.

First Class of Experiments. Metalliferous Veins. (From Daubuissons Traité de Géognosie.) The degrees are centigrade.

In the middle of the last century, Gensanne, director of the mines of Giromagny (Vosges), concluded that

At	100	mètres depth, the temperature was 12.0°
	308		18.8
	432		23.1
	Ratio deduced, 333 metres = 11.1° centig.
	or 30 ———— = 1.0° ———

M. Dauhuisson made, in 1802, a large series of experiments on the waters in the mines of Freyberg, where the mean surface temperature is 8 cent. The general results are contained in the following table.

Depth in Metres.	Name of the Mine.
	Beschertgluck.	Himraelfahrt.	Kiihschacht.	Junghohebirke.
0	8°	8°	8°	8°
80				10
100		10		10
120	10			11
160				12½
180		12½
200				14
220	12½
240		14		15
260	14 & 15	14½	14
280				16
300	15½		15 & 16
320				17

The general result is an augmentation of 8° for 300 metres; or, all observations included, 1° for 40 metres: if the extremes alone be taken, 1° in 35 metres.

This conclusion was confirmed by new experiments, in 1805, under the direction of Mr. Trebra. The temperature of the rock was now tried, with great care, for two years, the observations being registered thrice in each day. The temperature never varied in this time.

At	the surface (as before)	8°
	180 metres	11½
	260	15

as M. Daubuisson had found in 1802.

The ratio deduced is about 1° in 37 metres in the upper part, and 1° in 22.2 metres in the lower part.

Again, under the same direction, thermometers placed in gneiss in the mine called Alte Hoffnung Gottes, gave

At	the surface	(as before)	8.00°		72	mètres		8.75
	170		12.80
	270		15.00
	382		18.75

From these experiments it is concluded that the augmentation of temperature is 1° in 38 metres.

In the mines of Poullaouen and Huelgoat, in Brittany, M. Daubuisson found results which he considered to be partly influenced by local causes. In Poullaouen, at 140 fathoms, the augmentation was 3.1° or 1° for 45 metres. In Huelgoat, at 230 metres, the augmentation was 8.7°, or 1° in 26.4 metres.

In Cornwall, Mr. Fox's observations, at various periods, yield corresponding results. In a spring Dolcoath copper mine, 439 fathoms deep, the temperature was 27.8°, and that of the surface 10°.

In the same mine, 421 fathoms deep, the temperature of the rock of a gallery for 18 months was 24.2°.

Lately (1837) Mr. Fox communicated to the British Association some further observations, made below the lowest workings, in the Levant tin and copper mine, and the consolidated copper mine. At 230 fathoms from the surface, in the Levant mine (in granite), a thermometer, sunk 3 feet below the "sump," stood at 80; another, sunk only a few inches, was at 78.5°; and the air in the mine 67°. At 190 fathoms, the corresponding indications were 78°, 72.5°, and 67°. The general ratio is 1° Fahr. for 46 feet English; or, allowing 10 fathoms to the invariable temp., 1° in 46 feet.

In the consolidated mines, at a depth of 290 fathoms, thermometers sunk in a cross course of the rock (killas) indicated at the vein 92° and 88°; 10 fathoms from it, 86.3° and 85°; 24 fathoms from it, 85.3° and 84°. Here the metallic vein is the hottest part; Mr. Fox thinks, because of its allowing hot waters to ascend. The temperature 85.3° is at least 35° above that of the climate, and the ratio is consequently 1° in 49.6° feet; or, allowing 10 fathoms to the depth of variable temperature, 1° in 48 feet.

And still more lately, experiments made by Captain Oats inTresavean copper mine gave the following results.

					Rock $\overbrace {\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ }$
Depth.				Air.	No. 1.	No. 2.
26	feet	In	granite	53.3°	57.0°	52.8°
200			lode	77.2°	76.0°	75.5°
200			again	77.7°	76.0°	75.5°
250			lode	83.2°	82.5°	82.0°
262			lode	85.5°	82.5°	82.0°
	Ratio,	1° in 48 feet from the surface; or,
		1° in 46 feet from a point 10 fathoms below surface.

Humboldt observed, in mines near Guanaxuato (Mexico),

At	502	mètres depth in	Valenciana	mine	36.8°	centig.
	193	Rayas			33.7	———
	134	Villalpand			29.4	———

The surface temperature is 16°. The volcanic character of the country is perhaps unfavourable for accurate inferences.

Second Class of Experiments. In Stratified Rocks.

Saussure, in the salt mines of Bex, found

At	108	mètres depth the temperature	14.4°
	183		15.6°
	220		17.4°
	Ratio deduced, 1° centigrade for 37 mètres.

Mr. Hodgkinson, at the request of the British Association, has made some experiments in the comparatively shallow salt mines of Cheshire which evince an augmentation of 1° in 70 feet from the surface.

But the greatest strength of observation, independent of mineral veins, has been concentrated in the coal districts. Mr. Bald, Mr. Buddie, and other observers, have long since collected much information in the collieries of the Tyne and Wear; of which, however, we can make only partial use, because the experiments were mostly made on the air, which, for many reasons besides miners' lights and chemical actions, is unlikely to yield accurate ratios, such as are now attainable.

The following are some of Mr. Bald's results, published in the Edinburgh Royal Transactions. The scale is Fahrenheit's.

Whitehaven.—Spring at surface 49°

480 feet 60 

Ratio from surface, 1 for 44 feet.

Workington.—Spring at surface 48°

504 feet 60 

Ratio from surface, 1° for 42 feet.

Percy Main Colliery, Northumberland.—Mean

temperature at surface 49°^[24]

900 feet depth 70 

Ratio from surface, 1 for 43 feet.

Jarrow colliery.—Surface assumed 49.5°

Water at 882 feet 68.0 

Ratio from surface, 1 for 48 feet.

Killingworth colliery.—Surface assumed 48°

Water at 1200 feet depth 74 

Ratio from surface, 1° for 46 feet.

The near accordance of these results is remarkable. The ratios are all in error by a small quantity, because HO allowance is made for the depth of variable heat.

It is a very usual and easy objection to these results, that the lights, the respiration of horses and men, pyritous decompositions, &c. raise the temperature. The contrary is generally true, as we have shown in narrating the particulars of an experiment (1834) at Monkwearmouth, where the coal had been reached only a few days previous, no horses had entered the mine, few miners were at work, no chemical decompositions apparent, and the air supposed to be heated was many degrees cooler than the coal and rocks, and grew hotter only in proportion to their influence.^[25] The depth of this pit was 1584 feet; mean temperature at the surface 47.6°; thermometer at the bottom, in coal, 71.5°, 72.0° and 72.6°. Ratio deduced 1° Fahr. for 20 yards English.

This ratio, lower than Mr. Bald's, derived from the water in the coal mines, may perhaps be more correct; and it is supported by experiments at Wigan, under the care of Mr. Peace, which give 60 feet for 1°. At Manchester, Mr. Hodgkinson obtained a ratio from the surface, of 1° in 69 feet; while at Bedminster, under the care of Mr. W. Sanders, the ratio was found to be as high as 1° in 30 feet, and some anomalous facts were observed. (In each case 100 feet are deducted from the depth as an allowance for the depth of variable heat.)

M. Cordier gives the following summary of observations in the coal mines of Carmeaux, Littry and Decise (1827)

Carmeaux.
Water in the well	Veriac,	at	6.2	mètres	12.9°	cent.
	Bigorre		11.5		13.1
Rock at the bottom of	Ravin mine		181.9		17.1
	Castellan		192.0		19.5
Littry.
Surface temperature			0	mètres	11.00°	cent.
Rock at the bottom of St. Charles mine			99		16.18
Decise.
Water of the well	Pelisson	at	8.8	mètres	11.40°
Puits des Pavilions			16.9		11.67
	Rock in the Jacobe mine		107.0		17.78
Ditto			171.0		22.10

The general result of a complete discussion of these observations on subterranean temperature made in mines and collieries, appears, to give a ratio of 1° cent, for about 25 metres, or 1° Fahr. for 45 feet English.

Mr. Kenwood's observations on subterranean temperatures in the rocks, made on the waters issuing from them, extend to no less than 95 in slate, and 39 in granite, and from the surface to SOO fathoms and upwards. The following is a summary.

Slate.			Granite.
Average Depth. (ft.)	No. of Observa- tions.	Temperature.	Average Depth. (ft.)	No. of Obser- vations.	Temperature.
35	21	57.0°	31	7	51.6°
73	19	61.3	79	17	55.8
127	29	68.0	133	12	65.5
170	21	78.0
221	5	85.6	237	3	81.3

Thus at all depths the slate appears to be about 3.9° warmer than the granite at the same level.

The progressive increase of temperature in descending is in a mean of

95	observations on slate	1°	for 6.5	fathoms	(39 feet).
39	granite	1		6.9	(41.4).
(Reports of British Association for 1837.)

The Third Class of experiments includes chiefly Artesian weils. One of the most important is the well of La Rochelle, described by M. Fleuriau de Bellevue. The mean temperature of the district is

Air at the surface			1.87°
Water in the well, at	316	feet depth	16.25
Ditto	369½		18.12
Ratio from surface, 1° cent, for 58.5 to 72 feet, or 20 metres.

At Southampton, a well 133 yards deep was found to have a temperature of 56½° Fahr.; the mean temperature of surface being 50°. The ratio deduced is 1° Fahr. in 46 feet English.

The importance of this branch of evidence induced M. Arago to publish a short but valuable notice of Artesian wells, which is inserted in Jameson's Journal for 1835, p. 235. The following table is extracted, and the ratios appended to each observation:—

		Mètres.		Ratio.
Paris.—Mean temp, of surface	10.6°
Well of Port St. Ouen	12.9	66	1°	in	29.0
Departement du Nord.—Mean temp.	10.3
Well of Marquette	12.5	56	1		25.5
Aire	13.3	63	1		21.0
St. Venant	14.0	100	1		27.0
Sheerness.—Mean temp, of surface	10.5
Well	15.5	110	1		22.0
Tours.—Mean temp, of surface	11.5
Well	17.5	140	1		23.3
Mean result, 1° cent, for 24.6 mètres; or,
1 Fahr. for 45 feet.

The coincidence of this with the former result is unexpected.

The conclusion from experimental observation is in harmony with that authorised by hot springs, that the earth has a general and pervading high temperature below the surface.

---

1. The total number of recorded eruptions appears to be the following:—
   From Hekla, since the year	1004	inclusive,	22
   From Kattlagiaa Jokul	900		7
   From Krabla	1724		4
   In different parts of the Guldbringd Syssel	1000		3
   At sea	1583		2
   From the lake Grimsvatn, in	1716		1
   From Eyafialla Jokul	1717		1
   From Eyrefa Jokul, in	1720		1
   From Skaptaa Jokul, in	1783		1
   			—
   			42

2. Strickland, in Geol. Proceedings, 1837.
3. See Geol. Proceedings, Dec. 1834.
4. Journal of Science, vol., xvii. p. 46. It is not here asserted that 100,000 square miles were "permanently altered In level" It is stated that the "principal force was exerted in a circle of 50 miles diameter, having its centre S. E. of Valparaiso," and again the force diminished in proportion to the distance from Valparaiso.
5. Phil. Trans. 1760.
6. Reports of the British Association, 1843.
7. Ibid. 1846.
8. Communicated to the British Association in 1851.
9. Mallet in Brit. Assoc. Reports for 1850.
10. Brit. Assoc. Reports, 1847.
11. Phil. Trans. 1839, 1840, 1842.
12. Geol. Proceedings, 1838.
13. Traité de Geologie, tom. i. p. 206.; and Ann. de Chimie, tom. xxii.
14. Reports of British Association, 1836.
15. Dr. Daubeny places the source of this spring in red sandstone, but we conjecture that it is likely the spring originates in the mountain limestone which lies unconformable below the lias and new red sandstone.
16. Phil. Trans. 1836, part ii.
17. Geological Manual, p. 17.
18. St Amand, near Valenciennes, in the same strata, has the same excess of temperature.
19. This appears to be ascertained in the case of the hot spring of Aix, in Provence; and though, in the late diminution of the Bath waters by sinking a well In Bath (1836), the new well was filled by warm water, it was believed, that during the sinking of the Batheaston Trial coal pit, the Bath waters were reduced. The water was slightly warm in the Batheaston pit, if we correctly remember the statement of Dr. Smith, who was employed on the occasion.
20. The mean annual temperature of the equator being taken at 81.5°, that of any other lat. = 81.5 x nat. cosine lat. This is in error toward the north pole, owing to the distribution of land and water, which makes two poles of maximum cold in Asia and America, nearly coincident with the magnetic poles. See a paper by sir D. Brewster (Transactions of the Royal Society of Edinburgh).
21. About 1° of Fahrenheit for every 100 yards of ascent is a common correction used with the mountain barometer. A more exact proportion is supposed to be 1° for every 352 feet, as found by comparing Geneva and Great St Bernard. Mr. Atkinson, in Memoirs of the Astronomical Society, Professor Challis (Cambridge Phil. Trans.), have treated the subject mathematically. A general view of the state of knowledge on the distribution of terrestrial heat may be found in Professor Forbes's Report on Meteorology to the British Association.
22. Sur les Variations des Températures de la Terre. Bruxelles, 1837.
23. Forbes's experiments in different sorts of rock show the effect of these, in modifying the range of subterranean temperature, in altering the rate of its progress, and changing the epochs of maximum and minimum temperature. (Edinb. Trans. 1846)
24. It is really under 48°.—Author.
25. Phil. Mag. and Annals, 1834.