VOLCANOES

VOLCANOES. In other parts of this work, some account has been given of the principal volcanoes, either under their own heads, or those of the countries in which they occur. (See articles ÆTNA, ICELAND, JAVA, MADEIRA, POLYNESIA, &c.) The volcanic rocks have also been described in the article MINERALOGY AND GEOLOGY; and the general features of volcanoes are treated of under PHYSICAL GEOGRAPHY. It remains now only to recapitulate those points in their structure and history which form the data for reasoning upon their nature and cause.

Number of Volcanoes.—About 400 volcanic mountains are known to exist in various parts of the earth, and more than half of these have given evidence of their igneous activity in modern times. Catalogues have been compiled by Von Buch (Les Canaries, Paris ed. 1836), Scrope (Considerations on Volcanoes, 8vo, Lond. 1825), Humboldt, and Keith Johnston. The active volcanic vents are marked on Darwin's map of the Coral Reefs; and more recently (1854), a very complete map, on which the extinct volcanoes are also marked, has been published in Keith Johnston's Physical Atlas; the number of active volcanoes is there estimated at 270. The follow-

ing numbers are given by Humboldt in the last volume of Cosmos Volcanoes. (Bohn's edition, p. 121).

VOLCANIC REGIONS—Europe.—The principal seat of volcanic activity in Europe is the kingdom (formerly of the Two Sicilies) including Ætna, Vesuvius, and the islands of Stromboli and Volcano still in action; with Ischia and the volcanoes of the Palegrean fields which have been long quiescent.¹ Among the Greek islands there is a smaller tract, where Santorin has manifested some activity in historic times.

Atlantic Islands.—All the remote islands of the Atlantic are volcanic. Iceland and Jan Mayen, Pico in the Azores, and Teneriffe, have had modern eruptions. The rest of the Azores and Canaries, Madeira, Ascension, St Helena, and Tristan D'Acunha, are quiescent, or have long burned out. On the Antarctic continent, Mount Erebus, and in the S. Shetlands, Deception Island, have been seen in action.

Africa.—In continental Africa only one active volcano has been seen, in the Cameroon Mountains, on the west coast. There are others in the Red Sea and in the Comoro Islands.

Continental India has no burning mountains. Barren Island in the Bay of Bengal, Bourbon, and the remote oceanic rock of St Paul, are active volcanoes.

Continental Asia exhibits only the volcano of Demavend on the Caspian, two reputed volcanoes in the Thian-shan, and nine in the peninsula of Kamtschatka.

Asiatic Islands.—A great volcanic belt extends through the islands of southern and eastern Asia; commencing on the coast of Birmah, and passing through the Andaman Islands, Sumatra, Java, and Timor, it turns northward to Amboina, Gilolo, and the Philippines, and then extends by the Ioo-choo Islands to Japan, the Kurile Islands, and Kamtschatka. This region includes above 100 active volcanic vents, of which half are in the eastern and half in the southern portion.

South Sea.—With rare exceptions, the whole of the Pacific Islands which exhibit any rock whatever are of volcanic origin; yet in this great region there are fewer smoking volcanoes than on the single island of Java. These active vents occur in the Mariana and Sandwich Islands, north of the equator; and in the New Guinea group, New Hebrides, and New Zealand; and in the Friendly Islands.

America, west coast.—Volcanic rocks, with cones and craters, and other indications of subterranean action, exist in almost uninterrupted series, from the Aleutian Islands and coast of Russian America to the southern extremity of the continent in Tierra del Fuego. This region contains one-fourth of the known active volcanoes, including many of the loftiest mountain-summits in the world. On the north-west coast there are five, including Mount Eliza, and Mount Regnier on Puget's Sound, and St Helen's (15,000 feet,) north of the Columbia river.

In Mexico there are four active volcanoes; and eighteen in Central America. There are none in the Isthmus of Panama, which may therefore be considered in a state of stable equilibrium at the present time.

In Quito and New Grenada there are ten active cones of extraordinary dimensions, and situated 50 to 100 miles from the coast. The group of Southern Peru and Bolivia consists of fourteen great volcanic mountains, including Chimborazo, only three of which are active. Chile exhibits the larger proportion of thirteen (out of twenty-six) active volcanoes, extending over as many degrees of latitude, and amongst which are Aconcagua, 23,000 feet; Maypu, 17,662; Antuco, 9242; Otoro, 7443; and Corcovado, 7509. In Fuegia, the loftiest summit (Sarmiento, 6800 feet) is an extinct volcano. The Galapagos and Juan Fernandez, belonging to this coast, and not to the Pacific island-group, have active volcanoes.

On the eastern coast of America there are no volcanoes; but in the West Indies there are three active vents, in the chain of the Lesser Antilles.

Distribution in Latitude.—A large proportion of the volcanoes above mentioned are situated in tropical regions, and very few exist at a distance of more than 30° from the equator. Nevertheless,

Volcanoes, they are not dependent on climate, but are seen on the grandest scale in Iceland, or blazing amidst the perpetual winter of the south polar continent. In Deception Island, South Shetland, the layers of scoria are interstratified with snow. The following examples occur in high latitudes:—

Proximity to the sea.—These statements, or a glance at the map, will show that a vast majority of the volcanoes now in action stand either as islands in the sea or near its shores. The most remote are Sangay, one of the great mountains in the Quito group, 112 miles, and Fragua, 156 miles from the sea; and the two volcanoes of Thinn-shan in Central Asia, which are 1500 miles from either the Arctic or Indian seas.

Linear arrangement.—The position of volcanic vents at intervals along lines of coast and chains of islands is a remarkable circumstance, and was recognised by Plato in the myth of Phlegethon. It seems to indicate a frequently common origin upon longitudinal fissures,—an inference which is strengthened by some instances of their simultaneous eruption at great distances. On the night of the 19th January 1835, an eruption of Osorno was seen by Mr Darwin, who learned afterwards that Aconcagua in Chile, 480 miles northwards, was in action on the same night. A great eruption of Coseguina, 2700 miles to the north, occurred within six hours, accompanied by an earthquake felt over 1000 miles; but it is difficult even to conjecture whether this coincidence was accidental, or showed some subterranean connection. In the case of Juan Fernandez, situated 330 miles from the coast of Chile, there was undoubtedly a connection between the volcanic forces acting under the island and under the continent, as was shown during the earthquake of 1835.²

FORM AND STRUCTURE OF VOLCANOES—Cone of Eruption.—The characteristic form of fire-emitting mountains is a cone, truncated at the summit by a funnel-shaped cavity or crater. They are built up of ashes and scoria, erupted from the crater and deposited in layers, which are thickest near the rim, and consequently become steeper as the cone grows in height.² The angle of the slope is pictorially represented at 45°, as in Humboldt's View of Chimborazo, but is probably never quite so much. In the Peak of Teneriffe the inclination at the summit averages 33°, and in the volcanoes of Java it varies from 20° to 35°. Humboldt says that of all the volcanic cones he has seen, that of Cotopaxi is the most beautifully regular. The loftiest eruptive cones are the Sabama, in Bolivia, 22,350 feet; and Aconcagua, in Chile, 23,004 feet.

The outer surface of the cone consists usually of loose materials. The lofty peaks of the Andes present the appearance of mere heaps of trachytic fragments; but the interior must be rendered more solid by pressure and the percolation of rain-water, which it is known will convert light tuffs and loose volcanic sand into a heavy and compact rock. The interstices of the scoria are sometimes filled with ashes, or with pumice, as at Teneriffe.

Craters.—The crater of Vesuvius exhibits a series of apparently horizontal strata—the sections of lava-currents and beds of scoria, which dip outwards at an angle of 30° to 40°. Liquid lava is seldom ejected from cones of eruption; but when it does overflow at the summit, the crater is usually broken down on one side by its escape. Obsidian, which has a lower specific gravity than other forms of liquid lava, is frequently ejected from summits; while the heavy basaltic lavas have issued from fissures at low levels, and have often been spread out beneath the sea. The fragments of melted lava ejected from craters harden instantly on the surface, while the interior becomes vesicular from the expansion of the vapour they contain. The same is the case with the surfaces of lava streams, which cool rapidly and contract, breaking up like the sunburnt earth, and becoming rugged and uneven. Some scoriae, like pumice, will float in water for a time, and have been met with far at sea, covering the surface, in masses of sufficient depth to impede the progress of a ship.

Craters are frequently filled up, either by the rising of lava from within, or by the falling in of the summit of the cone. In the intervals between great eruptions, the interior of the crater presents the appearance of a hollow plain, with sometimes one or more small cones of eruption in the centre. Such was repeatedly the case with the crater of Vesuvius. Before the eruption of 1822 it was 4200 feet high; but more than 800 feet of the summit was

carried away by explosions, and a crater excavated, which was Volcanoes, three quarters of a mile in its longest diameter, and nearly 1000 feet in depth. In 1834, this great cavity had been filled up nearly to the top with lava, which had consolidated, and formed a level and unbroken plain, except that a small cone, thrown up by the ejection of scoria, rose in the midst of it like an island in a lake.⁴

Cones without Craters.—It is stated by Humboldt, that amongst the extinct, or at least quiescent volcanoes, "unopened domes" and "bell-shaped elevations without craters" are very numerous; and he cites as examples the Chimborazo, Ararat, and the Puy de Dome. It is, however, more than probable that these mountains have had craters which are now obliterated. The original eruptive orifices of the extinct volcanic island St Helena appear to have been filled up by enormous columns of phonolite, which now stand out as pinnacles (called Lot's Wife, Little Stony-top, &c.), more or less denuded of their former matrix.

Dikes.—The steepness of volcanic cones is liable to be further increased by the pressure of the incandescent lava previous to eruptions. Great vertical fissures are formed, radiating from the volcanic focus, and extending to the surface, or even to the summit of the mountain. Through these fissures the principal lava-streams escape, and they ultimately become filled with sheets of compact lava, cooled slowly under great pressure, and are called dikes. The crater of Vesuvius exhibits seven or more great vertical dikes, some of them at least 400 or 500 feet in height, and thinning out before they reach the uppermost part of the cone. In the Val del Bové, or lateral valley of Etna, the mass of this intrusive matter sometimes appears to equal in volume the rock it penetrates. Mr Darwin describes one dike in St Helena as being 1200 feet high, and only decreasing from 9 feet wide at the bottom to 8 feet 8 inches at the top. In some of the valleys of this island, they are numerous to a degree unequalled anywhere else: they extend in less regular lines, covering the ground with a network like a spider's web, and with some parts of the surface even appearing to consist wholly of dikes, interlaced by other dikes. The total amount of upheaval produced at successive eruptions, and rendered permanent by this injection of the fissures, must have been very considerable.

Laminated structure.—In the lava bordering the Aetnean dikes, Mr Darwin has observed a laminated structure, generally vertical, and extending in the direction in which the mass has flowed. "The most probable explanation of this structure in the felspathic rocks, appears to be that they have been stretched whilst flowing onwards in a parting condition."⁵ It is mentioned here because of its resemblance to the "ribbon-structure" of glacier ice produced by lateral pressure.

Craters without Cones.—It has been observed that the great mass of the lava streams produced in eruptions are not discharged from the summit of cones, but from fissures on their flanks and base, and these sometimes take the form of craters. The Chajorra, on the side of the Peak of Teneriffe, is 3000 feet lower than the terminal crater, which it far surpasses in size and in the magnitude of its lava-streams. The most striking example of a volcanic vent unaccompanied by a cinder-cone, is the great cauldron-like crater of Kila-Sandwich-uea on the flank of Mauna Loa in the island of Hawaii. This orifice is 4000 feet above the sea-level, while the summit is 13,760 feet. It is 3 or 4 miles in diameter, and 1000 feet deep when in its ordinary state, and resembles a huge quarry surrounded by vertical walls. Scattered over its bottom are lakes and pools of lava always in a state of ebullition, but instead of overflowing the rim of the cauldron, it finds a passage by subterranean fissures for 10 or 20 miles before it breaks ground, and flows over the surface towards the sea. The subsidence of the liquid plain of lava in the cauldron after eruptions has given origin to a ledge of consolidated lava 342 feet above the lake which it encircles. The usual slope of the lava sheets of Mauna Loa is between 5° and 10°. (Dana.)⁶

Old Crater Walls.—A very common feature in the configuration of volcanic mountains consists in the ruins of an older and larger crater surrounding the active cone of eruption. It is frequently broken down on one side, and in some very ancient volcanoes the erosive power of running water, or (in certain cases) of the sea, have greatly modified the form of the circumvallation, and enlarged the breach in its wall. One of the most perfect examples of this kind is the Peak of Teneriffe, which "stands like a tower surrounded Peak of by its fosse and bastion." The bastion, or old crater-wall, extends Teneriffe, nearly all round, but is broken down on the W. and N.E.; the highest, southern side, is 9000 feet above the sea. Its outer slope is less than 12°, but on the inner side it presents to the mountain a precipice of 1000 to 1800 feet. The cavity which it surrounds is 8 miles in diameter, and 7000 feet above the sea. The Peak, or cone of eruption, is more than 5000 feet in height, and the summit (or

Volcanoes, piton 500 or 600 feet more. The crater is 300 feet across, and 70 feet deep, but much broken down and easy of access. The sulphureous exhalations are chiefly noticeable on first reaching the brink, and do not prevent multitudes of insects and some small birds from taking shelter in it. The highest point is 12,158 feet above the sea.¹

Etna. The high platform of Etna exceeds 9000 feet, but the cone of eruption is only 1100 feet. There is evidence of a great encircling crater 2 miles in diameter, having once surrounded the base of the cone, but scarcely any portion of its wall remains. The most remarkable feature of Etna is the great chasm called the Val del Bové, which excavates the eastern flank of the mountain. It is described as a vast amphitheatre, 3 miles across and 6 in length, surrounded by steep precipices which rise to nearly 2000 feet in height at the upper end, and exhibit sections of innumerable lava-streams and beds of volcanic scoria, traversed by highly inclined dikes.² This valley is not itself a crater, although there is evidence of an old axis of eruption near its upper end, but is comparable with the "Valley of Jacob in Mount Ararat," and the ravines which break the circumvallation of Teneriffe and the Caldera of Palma. It was probably formed, at least in part, by aqueous erosion; for though Etna ordinarily feeds only a few small rivulets, yet, sometimes, as in the flood of 1755, the body of water discharged from it has been so great, that it was thought that if all the snow on the summit was melted in an instant, it would not have supplied such a volume of water. The Val del Bové was the scene of the last eruption, which commenced on the 20th of August 1852.³ In many of the loftiest volcanoes of Java, the active cone and crater are of small size, and surrounded by a plain of ashes and sand, and encircled by an "old crater wall," which is often 1000 feet and more in vertical height. Many of these are no less than 4 miles in diameter, and they are attributed by Junghuhn to the truncation by explosion and subsidence of ancient cones of eruption.

Java. The present cone of Vesuvius is supposed to have been formed by the great eruption of 79, in which the elder Pliny lost his life, and the cities of Herculaneum and Pompeii were destroyed. The semicircular ridge of Somma, which girdles the northern flank of the present mountain, is believed to be a remnant of the ancient Vesuvius. It was described as a truncated cone with a level outline as seen from a distance; and the crater, which must have been 3 miles in diameter, was surrounded by precipitous walls, except on one side, where there was a single narrow breach. These walls must have been since destroyed on the side towards the sea.

Palma. Origin of Encircling Craters.—In many of the old and extinct volcanoes which have been wasted by atmospheric influences or the action of the sea, little or no trace is left of a cone of eruption within the broken rampart of the old crater-wall. This is the case in the island of Palma, the type of Von Buch's craters of elevation, which resembles a vast cauldron 3 or 4 miles across, with walls rising to the height of 7000 or 8000 feet above the sea. The interior hollow or caldera, is only 2000 feet in elevation, and above its sloping sides rise precipices of 1500 to 2500 feet, traversed by countless dikes. There is only one great breach in this barrier, prolonged downwards in a ravine (baranco), by which torrents escape to the sea. From the centre of the cauldron rises a small conical eminence of a few hundred feet, consisting of lava resembling the more modern of those on the rim of the great crater.

It was supposed by Von Buch that this caldera, like the dome-shaped bases of Etna and Teneriffe, had been formed of lava-streams poured out, probably beneath the sea, on a much lower slope than they now exhibit; and that their present position and form were due to sudden and violent explosions. This hypothesis has lost much of its popularity since it has been ascertained by Scrope and others (especially Sir C. Lyell), that lava streams of small extent are actually consolidated on greater inclines than was formerly supposed.

Nevertheless there is difficulty in believing that such great lines of circumvallation as that of Teneriffe have ever formed the base of proportionately gigantic cones of eruption. In cones of eruption the sides are concave, but the swelling dome-like form of the bases of volcanoes with encircling craters points to the influence of expansion and upheaval rendered permanent by successive injections. They are true craters of slow elevation, and not of eruption or explosion. Mr Darwin admits this explanation in the case of the great broken craters of St. Jago, Mauritius, and St. Helena, which he thinks may have been elevated slowly en masse, while their central portion became separated from the exterior by curved faults. The inclination of their beds may have been slowly acquired by that mode of elevation of which, as Elie Beaumont says, the dikes are the evidence and the measure.⁴

Rapid formation of new Cones.—The ordinary growth of volcanic mountains may be slow, or fitful, and subject to frequent interruptions; but instances are not wanting of rapid or even sudden formation. Such is the Monte Nuovo in the Phlegrean fields, formed M. Nuovo in September 1638 on the site of the old Lucrine lake, once famous for its oysters. The eruption lasted two nights and two days without intermission, and on the third day people went up to the top of the new hill, 440 feet in height, and looked down into the crater (now 421 feet deep), and saw stones boiling up, as a cauldron of water boils on the fire. It has remained quiescent ever since.

Scarcely less remarkable is the Monte Rossi, at the foot of Etna, a double cone of 450 feet high, and 2 miles in circumference at the base, which was entirely formed of ejected sand and scoriae in the course of three or four months, in the year 1669.

The volcano of Izaleo, in Central America (now 1600 feet high), also rose up suddenly on the 23d February 1770, and has since remained uninterrupted active, often serving as a beacon for mariners in the Bay of Acajutla. But the most famous modern volcano is Jorullo in Mexico, formed in the latter half of the year 1759, which attained at once an elevation of 1600 feet, and has never resumed its activity. Six smaller cones, the least of them 300 feet in height, were thrown up along the line of the same fissure, and so enormous was the quantity of ejected lava and scoriae, that the whole plain, for the space of four square miles around, seemed to swell up like a blister, till it attained the elevation of 550 feet at the base of the mountains. Those who reason habitually from the slow progress of ordinary events, are apt to omit from their calculations the influence of convulsions like these, by which (as in social revolutions) the usual work of centuries is accomplished (or destroyed) in an hour.

Magnitude of Lava-streams.—It will be sufficient to mention two illustrations of the mass of liquid stone poured out in the course of single eruptions. In 1783 a lava-torrent burst from the Skapta-Jokul in Iceland, and continued flowing for two years. It filled up Iceland, the rocky beds of rivers, in many places 400 to 600 feet deep, and near 200 wide, flowing up their channels as well as down, filling deep lakes and an abyss of 600 feet below a waterfall, and spreading out over wide alluvial plains, in broad, burning lakes from 12 to 15 miles wide and 100 feet in depth. The two principal streams were respectively 40 and 50 miles in length, and 7 and 12 to 15 miles wide, forming a mass which surpasses Mount Blanc in magnitude.⁵

In the island of Hawaii, in 1840, a burning deluge of lava broke out 10 miles below the crater of Kilauea. It spread from 1 to 4 miles wide, and reached the sea, at a distance of 30 miles, in three days, and for fourteen days plunged, in a vast fiery cataract a mile wide, over a precipice of 50 feet. In 1843 a similar stream flowed from the summit of Mauna Loa; and, in August 1855, the lava broke out at a spot 2000 feet below the summit, on the opposite side to Kilauea, and continued for ten months overflowing an area of 200,000 acres. The main stream was 65 miles long, from 1 to 10 miles wide, and from 10 to 300 feet in depth.⁶

Frequency of Eruptions.—The activity of volcanic mountains is very unequal, and their fire often becomes extinct for long periods. The little mountain Stromboli (2318 feet) has been incessantly active since the Homeric age. In the time of Strabo and Pliny it was remarkable, as at the present day, for the number of its incandescent chasms, resembling the pit of a blast-furnace, in which fluid lava ascends hourly and overflows.

Izaleo, near San Salvador, exhibits four fiery eruptions in an hour, with variable violence, but astonishing regularity of occurrence; and the volcano of Rancagua in Chile (34.15. S. Lat.) is said to be always throwing out ashes and vapours like Stromboli.

The great mountain Sangay (17,000 feet), S.E. of Quito, has been in eruption ever since 1728, and is the most active of all known volcanoes, "exhibiting every quarter of an hour the greatest quantity of fiery, widely luminous eruptions of scoriae." 267 explosions have been counted in an hour, causing a continuous roar, which has been heard at places distant 252 and 348 geographical miles. Humboldt states that, during his residence near Quito, he could distinguish, by their voices, the eruptions of four great volcanoes on the same night.

Other mountains have burst forth into paroxysmal activity, and then relapsed into repose for many years. The eruptions of Copasapo have occurred twice after intervals of twenty-three years; and Coseguina has shown renewed activity after pauses of a hundred and of twenty-six years. In Ischia there intervened between two successive eruptions a pause of seventeen centuries; and Vesuvius has repeatedly slumbered for a century, or two at a time.

Volcanoes. An attempt has been made to connect the periodicity of small earthquake shocks with the moon's action on the supposed fluid interior of the earth;¹ and if there were any foundation for the report, it might be expected to influence other volcanic phenomena. But no such connection has been traced. Stromboli and Ætna, however, are said to be most active in November and the winter months.

Antiquity of Volcanic Vents.—The extreme age of many active volcanoes is shown by the fact, that in volcanic archipelagos seldom more than one island is active at a time, and the greater eruptions usually occur only after long intervals; also by the amount of degradation, by slow action of the sea, which their coasts have suffered, wearing back their sloping borders into lofty precipices. This is still more forcibly conveyed by the study of extinct volcanic islands, like Kerguelen's Land² and St Helena, whose internal structure is more completely exposed to view.

But most of the active volcanoes are of recent geological date, and some are quite modern. Vesuvius and Ætna rest upon marine strata of newer Pliocene age, and Hecla is newer than the Miocene tertiary. The Cordilleras have been the seat of volcanic activity since the earlier cretaceous period; but half their foci have been long extinct, and the rest are of many different ages. In islands where no marine strata co-exist with volcanic rocks, no limit can be assigned to their antiquity.

Extinct Volcanoes.—The largest and most perfect assemblage of extinct craters is that of Clermont in central France, a map of which resembles a portion of the lunar surface.³ Like the volcanoes of Madeira and the Azores, they are of post-miocene date.⁴ The Peak of Teneriffe itself (probably the Atlas of the ancients) has never erupted from its highest cone since the time of its first discovery, and appears to have been cooling down for ages. According to Sir C. Lyell, the volcanic district of Olot, in Catalonia, belongs to the older Pliocene period. The Eifel, in Prussia, also exhibits volcanic phenomena in several very interesting forms, and belongs to the middle tertiary age. The Ghauts of India are a prodigious mass of tertiary basalt, and the columnar basaltic rocks which overlie the chalk of Austria, and are repeated in Staffa and Mull, are probably of miocene age. In this country there are no ancient volcanic craters: the old trappian hills, which present the greatest resemblance to volcanic peaks, are, perhaps, the Rivals in Caernarvon, and the Titterstone in Shropshire—especially as seen from the grounds of Downton Castle.

Ancient Volcanic Rocks.—Scarcely any part of the globe is without signs of the former action of volcanic fires, and although the loosely constructed cones of eruption have been swept away, and the solid crater-foundations broken and wasted, enough remains to attest their existence and activity in every age of the earth's history. It has been previously pointed out (in the article MINERALOGY), that only an arbitrary distinction can be made between ancient "trappian" and modern volcanic rocks. Counterparts of the older rocks may be found in the interior of modern volcanoes, both as regards composition, hardness, and specific gravity. Columnar basalt appears to have been formed in the interior of lava-streams poured out at low levels and beneath the sea. Greenstone is represented by the diorite of the Ætnean dikes. Some of the trachytic lavas are full of minute quartz crystals; others contain hornblende and mica, forming sometimes a complete passage to the structure of the granitic rocks. Generally the old trap rocks consist of orthoclase and hornblende, while the newer lavas contain augite, and oligoclase, and glassy feldspar, and are more easily fusible. The dolerites of Iceland, Ætna, and Stromboli, consist of augite and Labradorite.

Causes of the distribution of Volcanoes along Coast-lines.—It is admitted by all who have speculated on the causes of volcanic action, that proximity to the sea is one of the necessary conditions for its manifestation. That the sea-water finds access to the foci of volcanoes is rendered at least probable by the enormous quantities of vapour always discharged, and the extraordinary torrents which sometimes accompany eruptions.

Ancient Volcanoes remote from the Sea.—The map of volcanoes and geological map of the world, in R. Johnston's Atlas, exhibits remains of ancient volcanoes deep in the interior of continents and far removed from present seashores. Such are Fisher's Peak, in Arkansas, and Ararat in the old world; the Ural, Altai, and Himalaya are accompanied by volcanic ranges, and there are many scattered centres of igneous action in Australia. Wherever their forms are distinct, they seem to beacon the old sea margins of the former world at various stages of its history; and the signs observed in the Thian-schan may, perhaps, be the last sparks (main-

tained by salt-lakes of the Tartarian steppes) of volcanic fires which once lighted up the shores of a great inland sea.

Extinction of Volcanic Foci.—The final extinction of volcanic fires does not seem to depend on the exhaustion of materials, but on the excessive amount of subterranean movement with which they have been associated. In one class of instances removing them far inland, and, in the other, drowning them beneath the sea. "On several encircling islands of the Pacific there are old craters and streams of lava, which show the effect of past and ancient eruptions. In these cases it would appear as if the volcanoes had come into action, or become extinguished on the same spots, according as the elevating or subsiding movement prevailed." (Darwin.)

Source of Volcanic Heat.—It was suggested by Sir Humphry Davy, that if the interior of the earth contained large quantities of the unoxidized metalloids, all the phenomena of volcanoes might be occasioned by the penetration of sea-water through deep fissures; and although abandoned by its distinguished author, this hypothesis (with some modifications) has been entertained by Dr Daubeny⁵ and other chemists. There is no question that water plays a most important part in volcanic processes, or that elastic vapours supply the principal motive force of upheavals and eruptions also.

Geological Theory.—The most popular opinion has always referred the phenomena of volcanoes (and earthquakes) to "the reaction of the interior of our planet upon its uppermost strata." "A volcano, properly so called, exists only when a permanent connection is established between the interior of the earth and the atmosphere" (Humboldt). From observations made in mines, it is believed that the interior temperature of the earth increases at the rate of 1° for every 50 or 100 feet, and that a heat capable of melting granite must prevail at the depth of 25 miles,—a distance which agrees with M. Beaumont's estimate of the thickness of the solid crust. According to Mr Hunt's observations, the increase of temperature below 200 fathoms is only at the rate of 1° for every 85 feet. But then it has been shown that pressure facilitates the fusion of substances which contract in melting, so that at the depth of 50,000 feet (or 9½ miles) a dull red heat of 680° Fahr. would be sufficient to melt granite, or at least to retain it in a state of fusion. Now, as the Peak of Teneriffe is 12,000 feet high, and rises from a depth of 12,000 feet in the sea, it is probable that its focus is at least 24,000 feet below the summit, and possibly twice as distant. And it has been already mentioned, that the substances ejected from the highest orifices are those of least specific gravity and greatest fusibility. The trachyte of Ponza would probably not be solidified at a temperature less than 680°, even at the depth of 2500 feet. Granite, according to Mr Sorby,⁶ is unerupted lava, out of which the lighter and more fusible elements have been melted,—a view which agrees with some remarks of Mr Darwin, for it seems impossible that many distinct rocks should exist separately in the depths of the earth, if all are alike fluid from the intense heat.

Mr Darwin's Theory.—The geologist who, next to Humboldt, has seen most of volcanic phenomena in all parts of the world, has advanced a step further in the theory of volcanoes, by endeavouring to connect their distribution with the nature of the subterranean movements formerly or still in progress. Having marked the sites of all the active volcanoes on his map of coral-reefs,—in which he has distinguished (by structure of the reefs) the areas of subsidence from others regarded as areas of elevation,—he goes on to say, "It may be considered as almost established that volcanoes are often (not necessarily always) present in those areas where the subterranean motive power has lately forced or is now forcing outwards the crust of the earth, but that they are invariably absent in those where the surface has lately subsided, or is still subsiding."⁷ Sir C. Lyell had previously argued, that "aqueous and igneous agents may be regarded as antagonist forces,—the aqueous labouring incessantly to reduce the inequalities of the surface to a level, while the igneous are equally active in renewing its unevenness." Sir Charles, therefore, appears to regard volcanoes as active agents in raising the surface of the earth, while Mr Darwin considers them only as the index of terrestrial movements.

Statements like these, coming from the highest geological authorities, are not likely to be questioned by geologists; but to naturalists, who study the distribution and origin of the existing "nations" of plants and animals, they will appear opposed to the fact, that while active volcanoes abound in those regions which he is accustomed to look upon as the most ancient dry land on the face of the globe, they are still more absolutely and entirely wanting in

Volcanoes. the provinces which he supposes to have been last rescued from the sea, judging by the character of their living population.¹

It is quite certain that the faunas of Bourbon, Java, New Zealand, &c., are not only very ancient, but that they have once been more widely distributed, and possessed a higher importance, and are now on the decline. They are not new faunas, recently produced, or lately imported, but the remains of ancient populations, having a peculiar and marked physiognomy (facies) which we cannot find on the continents, unless dimly represented in the remains of the former world; and it is equally certain that (with local and temporary exceptions) these areas are diminishing in extent. Mr Darwin himself remarks, that the region of the Asiatic islands is "the most broken land on the globe, and that the rising parts are surrounded and penetrated by areas of subsidence."

No naturalist will doubt that the volcanic islands of the Atlantic are the summits of submerged regions. The Faroe Islands and Iceland are fragments of a once continuous barrier in the North Atlantic; Madeira and P. Santo, the Azores and Canaries, are all more or less intimately connected with the Lusitanian province.² Even the remote oceanic specks—Tristan d'Acunha and Kerguelen's Land—are regarded by Dr Hooker as particles of a great southern continent, once extending as far as the Falklands and Fuegia, and clothed with a uniform vegetation.³ Sicily must have been a promontory of Northern Africa since the establishment of the present fauna; although the temporary appearance of Graham's Island (in 1831) sought to indicate an upward tendency in the bed of the Mediterranean, near the Skirki.

Non-volcanic Coast-lines.—On the other hand (as we can see by a glance at the maps) the non-volcanic coast-lines include all the northern arctic regions which are believed to be rising (with the exception of Greenland, which is volcanic and subsiding); all the low, shelving eastern coast of America, bounding tertiary plains, which have risen above the sea in the last age; the rising borders of the Baltic, and, in short, all the shelving coasts and low islands of shallow seas. These great surfaces, in Northern Asia and America, prairies and pampas, Baltic plains, and plain of Lombardy, have been upheaved in mass, without the aid or countenance of volcanic eruptions. Or, if it should be asserted that the supposed upward movement of volcanic regions is a mere transitory operation, liable to be reversed before it has united island to island, or produced any influence on the zoological and botanical provinces, it may be replied that such minor oscillations of level are common to every coast in the world, and not peculiar to volcanic regions. The eastern shores of England exhibit churches perched on the brink of cliffs, or buried in dunes of sand, alternating with new villages, rising up beneath the old cliffs on land gained from the sea; while, on the western coast, frequent raised beaches are associated with submarine forests, and other evidence of an opposite kind, speaking as plainly of small changes of level as do the columns of Serapis at Pozzuoli. It appears, therefore, that active volcanoes do not indicate regions undergoing continental elevation; and that they only prove the existence of conditions favourable to "the penetration of the interior of the earth by external influences," as Humboldt has stated it, "because the elevation of the coast, produced by internal elastic forces, is accompanied by a neighbouring depression in the basin of the sea, so that an area of elevation borders on an area of depression, and at this bordering-line large and deeply penetrating fissures and rents are produced."⁴

Organic remains in Volcanic rocks.—It is usually said that organic remains are wanting in all igneous rocks, and such is necessarily the case with lavas poured out in a molten condition; but volcanic deposits are often aqueous, and these do contain fossils. The eruption that destroyed Pompeii consisted of showers of ashes and streams of mud, which sealed up and preserved every enduring relic. Many of the volcanic eruptions in the Andes of Quito have been accompanied by streams of liquid tufa and mud, containing the small Siluroid fishes (Psilodorus cyclopus) which inhabit the subterranean waters of the mountains. Fragments of unaltered limestone, with fossils, as well as pieces of granite and other primary rocks, have sometimes been ejected from volcanoes. The siliceous shields of microscopic Diatomaceae (m'called "Infusoria") are of common occurrence in volcanic mud, whether of marine or fresh water origin; these minute plants flourish in thermal waters, and form a large proportion of the substance termed suetoric paper. Submarine lava-streams, at a moderate depth, become overgrown with sponges and corals, and tenanted by marine animals, to which their uneven surfaces afford abundant shelter; here they may become entombed with volcanic mud, or calcareous tufa, to form deposits like those of Balxo, and S. Vincente, Madeira. With

regard to the silicified sponges ("moss agates") and wood obtained from volcanic rocks, Dr Hooker has observed that both recent and fossil (i.e., already silicified) wood are found entangled in the lavas of Kerguelen. It is known that silicious deposits are of constant occurrence in the vesicular cavities of decomposing trachytic lavas, whose further disintegration sets them free like the agates of Oberstein, found in beds of gravel.⁵ The same process may take place in deposits formed by the consolidation of volcanic mud, with cavities caused by the partial decomposition of wood, coral, and sponges. Extensive deposits of lignite are often associated with volcanic formations, as at the Drachenfels and in Iceland; they are sometimes brand interstratified with lavas and tuffs, telling of long intervals of repose, and the development of a luxuriant vegetation over the ashes of former fires. (S. P. W.)