On consideration it seems improbable that the great sheets of augitic lava, such as cover the surface of the land of Bashan, are altogether the product of the volcanic mountains which appear to be confined to special districts in this wide area. Some of the craters do indeed send forth visible lava-streams, but they are insignificant as compared with the general mass of the plateau-basalts; and the crater-cones themselves appear in some cases to be posterior to the platforms of basalt from which they rise. It is very probable, therefore, that the lavas of this region have, in the main, been extruded from fissures of eruption at an early period, and spread over the surface of the country in the same manner as those of the Snake River region, and the borders of the Pacific Ocean of North America, and possibly of the Antrim Plateau in Ireland, afterwards to be described.
The volcanic hills which rise above the plateau are described in detail by Schumacher. Of these, Tell Abu Nedir is the largest in the Jaulan. It reaches an elevation of 4132 feet above the Mediterranean Sea, and 1710 feet above the plain from which it rises; the circumference of its base is three miles, and the rim of the crater itself, which is oval in form, is 1331 yards in its larger diameter. The interior is cultivated by Circassians, and is very fruitful; the walls descend at an angle of about 30 deg. on the inside, the exterior slope of the mountain being about 22 deg.. The cone seems to be formed chiefly of scoriae, and the lava-stream, which issues forth from the interior, forms a frightfully stony and lacerated district.
Another remarkable volcano is the Tell Abu en Neda (Fig. 24). This is a double crater, with a cone (probably of cinders) rising from the interior of one of them. The highest point of the rim of one of the craters reaches a level of 4042 feet above the sea. A lava-stream issues forth from Abu en Neda, and unites with another from a neighbouring volcano.
Tell el Ahmar is a ruptured crater of imposing aspect, reaching an elevation of 4060 feet, and sending forth a lava-current, which falls in regular terraces from the outlet towards the west and north.
The ruptured crater of Tell el Akkasheh, which reaches a height of 3400 feet, has a less forbidding aspect than the greater number of the extinct volcanoes of this region, owing to the fact that its sides are covered by oaks, which attain to magnificent proportions along the summit. Numerous other volcanic hills occur in this district, but the most remarkable is that called Tell el Farras (the Hill of the Horse). It is an isolated mountain, visible from afar, and reaches an elevation of 3110 feet, or nearly 800 feet above the surrounding plain. The oval crater of this volcano opens towards the north, and has a depth of 108 feet below the edge, with moderately steep sloping sides (17 deg.-32 deg.), while the slope of the exterior, at first steep, gradually lessens to 20 deg.-21 deg.. These slopes are covered with reddish or yellowish slag. The above examples will probably suffice to afford the reader a general idea of the size and form of the volcanoes in this little known region.
It has been stated above that the great lava-floods have probably been poured forth intermittently. The statement receives confirmation from the observations of Canon Tristram, made in the valley of the Yarmuk. This impetuous torrent rushes down a gorge, sometimes having limestone on one side and a wall of basalt on the other. This is due to the fact that the river channel had been eroded before the volcanic eruptions had commenced; but on the lava-stream reaching the channel, it naturally descended towards the valley of the Jordan along its bed, displacing the river, or converting it into clouds of steam. Subsequently the river again hewed out its channel, sometimes in the lava, sometimes between this rock and the chalky limestone. But, in addition to this, it has been observed that there is a bed of river gravel interposed between two sheets of basalt in the Yarmuk ravine; showing that after the first flow of that molten rock the river reoccupied its channel, which was afterwards invaded by another molten lava-stream, into which the waters have again furrowed the channel which they now occupy. The basaltic sheets descend under the waters of the Sea of Galilee on the east side, and were probably connected with those of Safed, crossing the Jordan valley north of that lake; owing to this the waters of the Lake of Merom (Huleh) were pent up, and formerly covered an extensive tract, now formed of alluvial deposits.
(d.) Land of Moab.—Proceeding southwards into the Land of Moab, the volcanic phenomena are here of great interest. Extensive sheets of basaltic lava, described as far back as 1807 by Seetzen, and more recently by Lartet and Tristram, are found at intervals between the Wadies Mojib (Arnon) and Haidan. On either side of the Mojib, cliffs of columnar basalt are seen capping the beds of white Cretaceous limestone, while a large mass has descended into the W. Haidan between cliffs of limestone and marl on either hand.
Around Jebel Attarus—a dome-shaped hill of limestone—a sheet of basaltic lava has been poured, and has descended the deep gorge of the Zerka Main, which enters the Dead Sea some 2000 feet below. This gorge had been eroded before the basaltic eruption, so that the stream of molten lava took its course down the bed of this stream to the water's edge, and grand sections have been laid bare by subsequent erosion along the banks. Pentagonal columns of black basalt form perpendicular walls, first on one side, then on the other; while considerable masses of scoriae, peperino, and breccia appear at the head of the glen, probably marking the orifice of eruption. Other eruptions of basalt occur, one at Mountar ez Zara, to the south of Zerka Main, and another at Wady Ghuweir, near the north-eastern end of the Dead Sea. There are no lava-streams on the western side of the Ghor, or of the Dead Sea.
The outburst of the celebrated thermal springs of Callirrhoe, together with nine or ten others, along the channel of the Zerka Main, is a circumstance which cannot be dissociated from the occurrence of basaltic lava at this spot. In a reach of three miles, according to Tristram, there are ten principal springs, of which the fifth in descent is the largest; but the seventh and eighth, about half a mile lower down, are the most remarkable, giving forth large supplies of sulphurous water. The tenth and last is the hottest of all, indicating a temperature of 143 deg. Fahr. Thus it would appear that the heat increases with the depth from the upper surface of the table-land; a result which might be expected, supposing the heated volcanic rocks to be themselves the source of the high temperature. To a similar cause may be attributed the hot-springs of Hammath, near Tiberias, and those of the Yarmuk near its confluence with the Jordan. Some of these and other springs break out along, or near, the line of the great Jordan-Arabah fault which ranges throughout the whole extent of this depression, from the base of Hermon to the Gulf of Akabah, generally keeping close to the eastern margin of the valley.
(e.) The Arabian Desert.—The basaltic lava-floods occupy a very large extent of the Arabian Desert, from El Hisma (lat. 27 deg. 35' N.) to the neighbourhood of Mecca on the south, a distance of about 440 miles, with occasional intervals. The lava-sheets are called "Harras" (or "Harrat"), one of which, Harrat Sfeina, terminates about ten miles north of Mecca. The lava-sheets rest sometimes on the red sandstone, at other times, on the granite and other crystalline rocks of great geological antiquity. In addition to the sheets of basalt, numerous crater-cones rise from the basaltic platform at a level of 5000 feet above the sea, and two volcanic mountains, rising far to the west of the principal range, called respectively Harrat Jeheyma and H. Rodwa, almost overlook the coast of the Red Sea.
(f.) Age of the Volcanic Eruptions.—It is very clear that the first eruptions, producing the great basaltic sheets of Moab and Arabia, occurred after the principal features of the country had been developed. The depression of the Jordan-Arabah valley, the elevation of the eastern side of this valley along the great fault line, and the channels of the principal tributary streams, such as those of the Yarmuk and Zerka Main, all these had been eroded out before they were invaded by the molten streams of lava. Now, as these physical features were developed and sculptured out during the Miocene period, as I have elsewhere shown to be the case, we may with great probability refer the volcanic eruptions to the geological epoch following—namely, the Pliocene. How far downwards towards the historic period the eruptions continued is not so certain. Dr. Daubeny, quoting several passages from the Old Testament prophets, says it might be inferred that volcanoes were in activity even so late as to admit of their being included within the limits of authentic history. The poetic language and imagery used in these passages by the prophets certainly lends a probability to this view, but nothing more. On the other hand, these regions have suffered through many centuries from the secondary effects of seismic action and subterranean forces, and earthquake shocks have laid in ruins the great temples and palaces of Palmyra, Baalbec, and other cities of antiquity. The same uncertainty regarding the time at which volcanic action died out, with reference to the appearance of man on the scene, hangs over the region of Arabia and Syria, as we have seen to be the case in reference to the extinct volcanoes of Auvergne, the Eifel, and the Lower Rhine. In all these cases the commencement and close of eruptive action appear to have been very much about the same period—namely, the Miocene period on the one hand, and that at which man entered upon the scene on the other; but in the case of Syria and Western Palestine, the close of the volcanic period may have been somewhat more than 2000 B.C.
 Lake Phiala, near the Lake of Huleh, is also situated to the west of the Jordan valley. Its origin, according to Tristram, is volcanic.
 Schumacher, "The Jaulan," Quarterly Statement of the Palestine Exploration Fund, 1886 and 1888; and Across the Jordan, London, 1886.
 Lartet, Voyage d'Exploration de la mer Morte (Geologie), Paris, 1880.
 Tristram, Land of Moab, London, 1873; and Land of Israel, 1866.
 Niebuhr, Beschreibung von Arabien, 1773.
 C. M. Doughty, Arabia Deserta, 2 vols., 1888. A generalised account of this volcanic region by the author will be found in the "Memoir on the Physical Geology of Arabia Petraea, and Palestine," Palestine Exploration Fund, 1887.
 Schumacher, loc. cit., p. 248.
 Land of Israel, p. 461.
 "Geology of Arabia Petraea, and Palestine," Memoirs of the Palestine Exploration Fund, p. 95.
 Doughty, loc. cit., vol. i., plate vi., p. 416. An excellent geological sketch map accompanies this work.
 "Memoir of the Geology of Arabia Petraea, and Palestine," chap. vi. p. 67.
 Nahum, i. 5, 6; Micah, i. 3, 4; Isaiah, lxiv. 1-3; Jeremiah, l. 25.
THE VOLCANIC REGIONS OF NORTH AMERICA.
(a.) Contrast between the Eastern and Western Regions.—In no point is there a more remarkable contrast between the physical structure of Eastern and Western America than in the absence of volcanic phenomena in the former and their prodigious development in the latter. The great valley of the Mississippi and its tributaries forms the dividing territory between the volcanic and non-volcanic areas; so that on crossing the high ridges in which the western tributaries of America's greatest river have their sources, and to which the name of the "Rocky Mountains" more properly belongs, we find ourselves in a region which, throughout the later Tertiary times down almost to the present day, has been the scene of volcanic operations on the grandest scale; where lava-floods have been poured over the country through thousands of square miles, and where volcanic cones, vying in magnitude with those of Etna, Vesuvius, or Hecla, have established themselves. This region, generally known as "The Great Basin," is bounded on the west by the "Pacific Range" of mountains, and includes portions of New Mexico, Arizona, California, Nevada, Utah, Colorado, Idaho, Oregon, Wyoming, Montana, and Washington. To the south it passes into the mountainous region of Mexico, also highly volcanic; and thence into the ridge of Panama and the Andes. It cannot be questioned but that the volcanic nature of the Great Basin is due to the same causes which have originated the volcanic outbursts of the Andes; but, from whatever cause, the volcanic forces have here entered upon their secondary or moribund stage. In the Yellowstone Valley, geysers, hot springs, and fumaroles give evidence of this condition. In other districts the lava-streams are so fresh and unweathered as to suggest that they had been erupted only a few hundred years ago; but no active vent or crater is to be found over the whole of this wide region. A few special districts only can here be selected by way of illustration of its special features in connection with its volcanic history.
(b.) The Plateau Country of Utah and Arizona.—This tract, which is drained by the Colorado River and its tributaries, is bounded on the north by the Wahsatch range, and extends eastwards to the base of the Sierra Nevada. Round its margin extensive volcanic tracts are to be found, with numerous peaks and truncated cones—the ancient craters of eruption—of which Mount San Francisco is the culminating eminence. South of the Wahsatch, and occupying the high plateaux of Utah, enormous masses of volcanic products have been spread over an area of 9000 square miles, attaining a thickness of between 3000 and 4000 feet. The earlier of these great lava-floods appear to have been trachytic, but the later basaltic; and in the opinion of Captain Dutton, who has described them, they range in point of time from the Middle Tertiary (Miocene) down to comparatively recent times.
(c.) The Grand Canyon.—To the south of the high plateaux of Utah are many minor volcanic mountains, now extinct; and as we descend towards the Grand Canyon of Colorado we find numerous cinder-cones scattered about at intervals near the cliffs. Extensive lava-fields, surmounted by cinder-cones, occupy the plateau on the western side of the Grand Canyon; and, according to Dutton, the great sheets of basaltic lava, of very recent age, which occupy many hundred square miles of desert, have had their sources in these cones of eruption. Crossing to the east of the Grand Canyon, we find other lava-floods poured over the country at intervals, surmounted by San Francisco—a volcanic mountain of the first magnitude—which reaches an elevation, according to Wheeler, of 12,562 feet above the ocean. It has long been extinct, and its summit and flanks are covered with snow-fields and glaciers. Other parts of Arizona are overspread by sheets of basaltic lava, through which old "necks" of eruption, formed of more solid lava than the sheets, rise occasionally above the surface, and are prominent features in the landscape.
Further to the eastward in New Mexico, and near the margin of the volcanic region, is another volcanic mountain little less lofty than San Francisco, called Mount Taylor, which, according to Dutton, rises to an elevation of 11,390 feet above the ocean, and 8200 feet above the general level of the surrounding plateau of lava. This mountain forms the culminating point of a wide volcanic tract, over which are distributed numberless vents of eruption. Scores of such vents—generally cinder-cones—are visible in every part of the plateau, and always in a more or less dilapidated condition. Mount Taylor is a volcano, with a central pipe terminating in a large crater, the wall of which was broken down on the east side in the later stage of its history.
(d.) California.—Proceeding westwards into California, we are again confronted with volcanic phenomena on a stupendous scale. The coast range of mountains, which branches off from the Sierra Nevada at Mount Pinos, on the south, is terminated near the northern extremity of the State by a very lofty mountain of volcanic origin, called Mount Shasta, which attains an elevation of 14,511 feet (see Fig. 25). This mountain was first ascended by Clarence King in 1870, and although forming, as it were, a portion of the Pacific Coast Range, it really rises from the plain in solitary grandeur, its summit covered by snow, and originating several fine glaciers.
The summit of Mount Shasta is a nearly perfect cone, but from its north-west side there juts out a large crater-cone just below the snow-line, between which and the main mass of the mountain there exists a deep depression filled with glacier ice. This secondary crater-cone has been named Mount Shastina, and round its inner side the stream of glacier ice winds itself, sometimes surmounting the rim of the crater, and shooting down masses of ice into the great caldron. The length of this glacier is about three miles, and its breadth about 4000 feet. Another very lofty volcanic mountain is Mount Rainier, in the Washington territory, consisting of three peaks of which the eastern possesses a crater very perfect throughout its entire circumference. This mountain appears to be formed mainly of trachytic matter. Proceeding further north into British territory, several volcanic mountains near the Pacific Coast are said to exhibit evidence of activity. Of these may be mentioned Mount Edgecombe, in lat. 57 deg..3; Mount Fairweather, lat. 57 deg..20 which rises to a height of 14,932 feet; and Mount St. Elias, lat. 60 deg..5, just within the divisional line between British and Russian territory, and reaching an altitude of 16,860 feet. This, the loftiest of all the volcanoes of the North American continent, except those of Mexico, may be considered as the connecting link in the volcanic chain between the continent and the Aleutian Islands.
(e.) Lake Bonneville.—Returning to Utah we are brought into contact with phenomena of special interest, owing to the inter-relations of volcanic and lacustrine conditions which once prevailed over large tracts of that territory. The present Great Salt Lake, and the smaller neighbouring lakes, those called Utah and Sevier, are but remnants of an originally far greater expanse of inland water, the boundaries of which have been traced out by Mr. C. K. Gilbert, and described under the name of Lake Bonneville. The waters of this lake appear to have reached their highest level at the period of maximum cold of the Post-Pliocene period, when the glaciers descended to its margin, and large streams of glacier water were poured into it. Eruptions of basaltic lava from successive craters appear to have gone on before, during, and after the lacustrine epochs; and the drying up of the waters over the greater extent of their original area, now converted into the Sevier Desert, and their concentration into their present comparatively narrow basins, appears to have proceeded pari passu with the gradual extinction of the volcanic outbursts. Two successive epochs of eruption of basalt appear to have been clearly established—an earlier one of the "Provo Age," when the lava was extruded from the Tabernacle craters, and a later epoch, when the eruptions took place from the Ice Spring craters. The oldest volcanic rock appears to be rhyolite, which peers up in two small hills almost smothered beneath the lake deposits. Its eruption was long anterior to the lake period. On the other hand, the cessation of the eruptions of the later basaltic sheets is evidently an event of such recent date that Mr. Gilbert is led to look forward to their resumption at some future, but not distant, epoch. As he truly observes, we are not to infer that, because the outward manifestations of volcanic action have ceased, the internal causes of those manifestations have passed away. These are still in operation, and must make themselves felt when the internal forces have recovered their exhausted energies; but perhaps not to the same extent as before.
(f.) Region of the Snake River.—The tract of country bordering the Snake River in Idaho and Washington is remarkable for the vast sheets of plateau-basalt with which it is overspread, extending sometimes in one great flood farther than the eye can reach, and what is still more remarkable, they are often unaccompanied by any visible craters or vents of eruption. In Oregon the plateau-basalt is at least 2,000 feet in thickness, and where traversed by the Columbia River it reaches a thickness of about 3,000 feet. The Snake and Columbia rivers are lined by walls of volcanic rock, basaltic above, trachytic below, for a distance of, in the former, one hundred, in the latter, two hundred, miles. Captain Dutton, in describing the High Plateau of Utah, observes that the lavas appear to have welled up in mighty floods without any of that explosive violence generally characteristic of volcanic action. This extravasated matter has spread over wide fields, deluging the surrounding country like a tide in a bay, and overflowing all inequalities. Here also we have evidence of older volcanic cones buried beneath seas of lava subsequently extruded.
(g.) Fissures of Eruption.—The absence, or rarity, of volcanic craters or cones of eruption in the neighbourhood of these great sheets has led American geologists to the conclusion that the lavas were in many cases extruded from fissures in the earth's crust rather than from ordinary craters. This view is also urged by Sir A. Geikie, who visited the Utah region of the Snake River in 1880, and has vividly described the impression produced by the sight of these vast fields of basaltic lava. He says, "We found that the older trachytic lavas of the hills had been deeply trenched by the lateral valleys, and that all these valleys had a floor of black basalt that had been poured out as the last of the molten materials from the now extinct volcanoes. There were no visible cones or vents from which these floods of basalt could have proceeded. We rode for hours by the margin of a vast plain of basalt stretching southward and westward as far as the eye could reach.... I realised the truth of an assertion made first by Richthofen, that our modern volcanoes, such as Vesuvius and Etna, present us with by no means the grandest type of volcanic action, but rather belong to a time of failing activity. There have been periods of tremendous volcanic energy, when instead of escaping from a local vent, like a Vesuvian cone, the lava has found its way to the surface by innumerable fissures opened for it in the solid crust of the globe over thousands of square miles."
(h.) Volcanic History of Western America.—The general succession of volcanic events throughout the region of Western America appears to have been somewhat as follows:—
The earliest volcanic eruptions occurred in the later Eocene epoch and were continued into the succeeding Miocene stage. These consisted of rocks moderately rich in silica, and are grouped under the heads of propylite and andesite. To these succeeded during the Pliocene epoch still more highly silicated rocks of trachytic type, consisting of sanidine and oligoclase trachytes. Then came eruptions of rhyolite during the later Pliocene and Pleistocene epochs; and lastly, after a period of cessation, during which the rocks just described were greatly eroded, came the great eruptions of basaltic lava, deluging the plains, winding round the cones or plateaux of the older lavas, descending into the river valleys and flooding the lake beds, issuing forth from both vents and fissures, and continuing intermittently down almost into the present day—certainly into the period of man's appearance on the scene. Thus the volcanic history of Western America corresponds remarkably to that of the European regions with which we have previously dealt, both as regards the succession of the various lavas and the epochs of their eruption.
(i.) The Yellowstone Park.—The geysers and hot springs of the Yellowstone Park, like those in Iceland and New Zealand, are special manifestations of volcanic action, generally in its secondary or moribund stage. The geysers of the Yellowstone occur on a grand scale; the eruptions are frequent, and the water is projected into the air to a height of over 200 feet. Most of these are intermittent, like the remarkable one known as Old Faithful, the Castle Geyser, and the Giantess Geyser described by Dr. Hayden, which ejects the water to a height of 250 feet. The geyser-waters hold large quantities of silica and sulphur in solution, owing to their high temperature under great pressure, and these minerals are precipitated upon the cooling of the waters in the air, and form circular basins, often gorgeously tinted with red and yellow colours.
 J. W. Powell, Exploration of the Canyons of the Colorado, pp. 114, 196. Major Powell describes a fault or fissure through which floods of lava have been forced up from beneath and have been poured over the surface. Many cinder-cones are planted along the line of this fissure.
 Capt. C. E. Dutton. Sixth Ann. Rep. U.S. Geol. Survey, 1884-85.
 Dutton, loc. cit., chap. iv. p. 165.
 Amer. Jour. Science, vol. 3., ser. (1871). A beautiful map of this mountain is given in the Fifth Annual Report, U.S. Geol. Survey, 1883-84. Plate 44.
 Daubeny, loc. cit., p. 474.
 Gilbert, Monograph U.S. Geol. Survey, vol. i. (1890).
 Powell, Exploration of the Colorado River, p. 177, etc. (1875). Hayden, Rep. U.S. Geol. Survey of the Colorado, etc. (1871-80).
 Richthofen, Natural System of Volcanic Rocks, Mem. California Acad. Sciences, vol. i. (1868).
 Geikie, Geological Sketches at Home and Abroad, p. 271 (1882).
 Prestwich, Geology, vol. i. p. 370, quoting from Richthofen.
 The origin of geysers is variously explained; see Prestwich, Geology, vol. i. p. 170. They are probably due to heated waters suddenly converted into steam by contact with rock at a high temperature.
VOLCANOES OF NEW ZEALAND.
One other region of volcanic action remains to be noticed before passing on to the consideration of those of less recent age. New Zealand is an island wherein seem to be concentrated all the phenomena of volcanic action of past and present time. Though it is doubtful if the term "active," in its full sense, can be applied to any of the existing craters (with two or three exceptions, such as Tongariro and Whakari Island), we find craters and cones in great numbers in perfectly fresh condition, extensive sheets of trachytic and basaltic lavas, ashes, and agglomerates; lava-floods descending from the ruptured craters of ashes and scoriae; old crater-basins converted into lakes; geysers, hot springs and fumaroles which may be counted by hundreds, and cataracts breaking over barriers of siliceous sinter; and, lastly, lofty volcanic mountains vying in magnitude with Vesuvius and Etna. All these wonderful exhibitions of moribund volcanic action seem to be concentrated in the northern island of Auckland. The southern island, which is the larger, also has its natural attractions, but they are of a different kind; chief of all is the grand range of mountains called, not inappropriately, the "Southern Alps," vying with its European representative in the loftiness of its peaks and the splendour of its snowfields and glaciers, but formed of more ancient and solid rocks than those of the northern island.
(a.) Auckland District.—We are indebted to several naturalists for our knowledge of the volcanic regions of New Zealand, but chiefly to Ferdinand von Hochstetter, whose beautiful maps and graphic descriptions leave nothing to be desired. In this work Hochstetter was assisted by Julius Haast and Sir J. Hector. From their account we learn that the Isthmus of Auckland is one of the most remarkable volcanic districts in the world. It is characterised by a large number of extinct cinder-cones, in a greater or less perfect state of preservation, and giving origin to lava-streams which have poured down the sides of the hills on to the plains. Besides these are others formed of stratified tuff, with interior craters, surrounding in mural cliffs eruptive cones of scoriae, ashes, and lapilli; these cones are scattered over the isthmus and shores of Waitemata and Manukau. The tuff cones and craters rise from a floor of Tertiary sandstone and shale, the horizontal strata of which are laid open in the precipitous bluffs of Waitemata and Manukau harbours; they sometimes contain fossil shells of the genera Pecten, Nucula, Cardium, Turbo, and Neritae. As the volcanic tuff-beds are intermingled with the Upper Tertiary strata, it is inferred that the first outbursts of volcanic forces occurred when the region was still beneath the waters of the ocean. Cross-sections show that the different layers slope both outwards (parallel to the sides) and inwards towards the bottom of the craters. Sometimes these craters have been converted into lakes, as in the case of those of the Eifel; but generally they are dry or have a floor of morass. Of the crater-lakes, those of Kohuora, five in number, are perhaps the most remarkable; and in the case of two of these the central cones of slag appear as islets rising from the surface of the waters. The fresh-water lake Pupuka has a depth of twenty-eight fathoms. To the north of Auckland Harbour rises out of the waters of the Hauraki Gulf the cone of Rangitoto, 920 feet high, the flanks formed of rugged streams of basalt, and the summit crowned by a circular crater of slag and ash, out of the centre of which rises a second cone with the vent of eruption. This is the largest and newest of the Auckland volcanoes, and appears to have been built up by successive outpourings of basaltic lava from the central orifice, after the general elevation of the island.
Before leaving the description of the tuff-cones, which are a peculiar feature in the volcanic phenomena of New Zealand, and are of many forms and varieties, we must refer to that of Mount Wellington (Maunga Rei). This is a compound volcano, in which the oldest and smallest of the group is a tuff-crater-cone, exhibiting very beautifully the outward slope of its beds. Within this crater arise two cones of cinders, each with small craters. It would appear that after a long interval the larger of the two principal cones, formed of cinders and known as Mount Wellington, burst forth from the southern margin of the older tuff-cone, and, being built up to a height of 850 feet, gradually overspread the sides of its older neighbour. Mount Wellington itself has three craters, and from these large streams of basaltic lava have issued forth in a westerly direction, while a branch entered and partially filled the old tuff-crater to the northwards.
Southwards from Manukau Harbour, and extending a short distance from the coast-line to Taranaki Point, there occurs a plateau of basalt-conglomerate (Basaltkonglomerat), with sheets of basaltic lava overspreading the Tertiary strata. These plateau-basalts are intersected by eruptive masses in the form of dykes, but still there are no craters or cones of eruption to be seen; so that we may infer that the sheets, at least, were extruded from fissures in the manner of those of the Colorado or Idaho regions of America. Proceeding still further south into the interior of the island, we here find a lofty plateau of an average elevation of 2,000 feet, interposed between the Tertiary beds of the Upper and Middle Waikato, and formed of trachytic and pitch-stone tuff, amongst which arise old extinct volcanic cones, such as those of Karioi, Pirongia, Kakepuku, Maunga Tautari, Aroha, and many others. These trachytic lavas would seem to be more ancient than the basaltic, previously described.
(b.) Taupo Lake, and surrounding district.—But of all these volcanic districts, none is more remarkable than that surrounding the Taupo Lake, which lies amidst the Tertiary strata of the Upper Waikato Basin. The surface of this lake is 1,250 feet above that of the ocean, and its margin is enclosed within a border of rhyolite and pitchstone—rising into a mass of the same material 1,800 feet high on the eastern side. The form of the lake does not suggest that it is itself the crater of a volcano, but rather that it was originated by subsidence. On all sides, however, trachytic cones arise, of which the most remarkable group lies to the south of the lake, just in front of the two giant trachytic cones, the loftiest in New Zealand, one called Tongariro, rising about 6,500 feet, and the other Ruapahu, which attains an elevation of over 9,000 feet, with the summit capped by snow. These two lofty cones, standing side by side, are supposed by the Maoris to be the husband and wife to whom were born the group of smaller cones above referred to as occupying the southern shore of Taupo Lake. The volcano of Tongariro may still be considered as in a state of activity, as its two craters (Ngauruhoe and Ketetahi) constantly emit steam, and several solfataras break out on its flanks.
(c.) Roto Mahana.—In a northerly direction from Tongariro, and distant from the coast by a few miles, lies in the Bay of Plenty the second of the active volcanoes of New Zealand, the volcanic island of Whakari (White Island), from the crater of which are constantly erupted vast masses of steam clouds. The distance between these two active craters is 120 nautical miles; and along the tract joining them steam-jets and geysers issue forth from the deep fissures through which the lava sheets have formerly been extruded. Numerous lakes also occupy the larger cavities in the ground; and hot-springs, steam-fumaroles and solfataras burst out in great numbers along the banks of the Roto Mahana Lake and the Kaiwaka River by which it is drained. Amongst such eruptions of hot-water and steam we might expect the formation of siliceous sinter, and the deposition of sulphur and other minerals; nor will our expectations be disappointed. For here we have the wonderful terraces of siliceous sinter deposited by the waters entering Roto Mahana as they descend from the numerous hot-springs or pools near its margin. All travellers concur in describing these terraces as the most wonderful of all the wonders of the Lake district of New Zealand—so great is their extent, and so rich and varied is their colouring.
The beautiful map of Roto Mahana on an enlarged scale by Hochstetter shows no fewer than ten large sinter terraces descending towards the margin of this lake, besides several mud-springs, fumaroles, and solfataras. But the largest and most celebrated of all the sinter terraces has within the last few years been buried from view beneath a flood of volcanic trass, or mud, an event which was as unexpected as it was unwelcome. In May, 1887, the mountain of Tarawera, which rises to the north-east of Roto Mahana, and on the line of eruption above described, suddenly burst forth into violent activity, covering the country for miles around with clouds of ashes, and, pouring down torrents of mud, completely enveloped the beautiful terrace of sinter which had previously been one of the wonders of New Zealand. By the same eruption several human beings were entombed, and their residences destroyed.
The waters of Roto Mahana, together with the hot-springs and fountains are fed from rain, and from the waters of Taupo Lake, which, sinking through fissures in the ground, come in contact with the interior heated matter, and thus steam at high temperature and pressure is generated.
(d.) Moribund condition of New Zealand Volcanoes.—From what has been said, it will be inferred that in the case of New Zealand, as in those of Auvergne, the Eifel and Lower Rhine, Arabia, and Western America, we have an example of a region wherein the volcanic forces are well-nigh spent, but in which they were in a state of extraordinary activity throughout the later Tertiary, down to the commencement of the present epoch. In most of these cases the secondary phenomena of vulcanicity are abundantly manifest; but the great exhibitions of igneous action, when the plains were devastated by sheets of lava, and cones and craters were piled up through hundreds and thousands of feet, have for the present, at least, passed away.
 Geol.-topographischer Atlas von Neu-Seeland, von Dr. Ferd. von Hochstetter und Dr. A. Petermann. Gotha: Justus Perthes (1863). Also New Zealand, trans. by E. Sauter, Stuttgart (1867).
 Tongariro was visited in 1851 by Mr. H. Dyson, who describes the eruption of steam.
 Mr. Froude figures and describes the two terraces, the "White" and "Pink," in Oceana, 2nd edition, pp. 285-291.
TERTIARY VOLCANIC DISTRICTS OF THE BRITISH ISLES.
It is an easy transition to pass from the consideration of European and other dormant, or extinct, volcanic regions to those of the British Isles, though the volcanic forces may have become in this latter instance quiescent for a somewhat longer period. In all the cases we have been considering, whether those of Central Italy, of the Rhine and Moselle, of Auvergne, or of Syria and Arabia, the cones and craters of eruption are generally present entire, or but slightly modified in form and size by the effects of time. But in the case of the Tertiary volcanic districts of the British Isles this is not so. On the contrary, these more prominent features of vulcanicity over the surface of the ground have been removed by the agents of denudation, and our observations are confined to the phenomena presented by extensive sheets of lava and beds of ash, or the stumps and necks of former vents of eruption, together with dykes of trap by which the plateau-lavas are everywhere traversed or intersected.
The volcanic region of the British Isles extends at intervals from the North-east of Ireland through the Island of Mull and adjoining districts on the mainland of Morvern and Ardnamurchan into the Isle of Skye, and comprises several smaller islets; the whole being included in the general name of the Inner Hebrides. It is doubtful if the volcanic lavas of Co. Antrim were ever physically connected with those of the west of Scotland, though they may be considered as contemporary with them; and in all cases the existing tracts of volcanic rock are mere fragments of those originally formed by the extrusion of lavas from vents of eruption. In addition to these, there are large areas of volcanic rock overspread by the waters of the ocean.
(a.) Geological Age.—The British volcanic eruptions now under consideration are all later than the Cretaceous period. Throughout Antrim, and in parts of Mull, the lavas are found resting on highly eroded faces either of the Upper Chalk (Fig. 27), or, where it has been altogether denuded away, on still older Mesozoic strata. From the relations of the basaltic sheets of Antrim to the Upper Chalk, it is clear that the latter formation, after its deposition beneath the waters of the Cretaceous seas, was elevated into dry land and exposed to a long period of subaerial erosion before the first sheets of lava invaded the surface of the ground. We are, therefore, tolerably safe in considering the first eruptions to belong to the Tertiary period; but the evidence, derived as it is exclusively from plant remains, is somewhat conflicting as to the precise epoch to which the lavas and beds of tuff containing the plant-remains are to be referred. The probabilities appear to be that they are of Miocene age; and if so, the trachytic lavas, which in Antrim are older than those containing plants, may be referred to a still earlier epoch—namely, that of the Eocene. As plant remains are not very distinctive, the question regarding the exact time of the first volcanic eruptions will probably remain for ever undecided; but we are not likely to be much in error if we consider the entire volcanic period to range from the close of the Eocene to that of the Miocene; by far the greater mass of the volcanic rocks being referable to the latter epoch.
In describing the British volcanic districts it will be most convenient to deal with them in three divisions—viz., those of Antrim, Mull, and Skye, commencing with Antrim.
(b.) Volcanic Area.—The great sheets of basalt and other volcanic products of the North-east of Ireland overspread almost the whole of the County Antrim, and adjoining districts of Londonderry and Tyrone, breaking off in a fine mural escarpment along the northern shore of Belfast Lough and the sea coast throughout the whole of its range from Larne Harbour to Lough Foyle; the only direction in which these features subside into the general level of the country being around the shores of Lough Neagh. Several outliers of the volcanic sheets are to be found at intervals around the great central plateau; such as those of Rathlin Island, Island Magee, and Scrabo Hill in Co. Down. The area of the basaltic plateau may be roughly estimated at 2,000 square miles.
The truncated edges of this marginal escarpment rising to levels of 1,000 to 1,260 feet, as in the case of Benevenagh in Co. Derry, and 1,825 feet at Mullaghmore, attest an originally greatly more extended range of the basaltic sheets; and it is not improbable that at the close of the Miocene epoch they extended right across the present estuary of Lough Foyle to the flanks of the mountains of Inishowen in Donegal in one direction, and to those of Slieve Croob in the other. In the direction of Scotland the promontories of Kintyre and Islay doubtless formed a part of the original margin. Throughout this vast area the volcanic lavas rest on an exceedingly varied rocky floor, both as regards composition and geological age. (See Fig. 28.) Throughout the central, southern, eastern, and northern parts of their extent, the Chalk formation may be considered to form this floor; but in the direction of Armagh and Tyrone, towards the southwestern margin, the basaltic sheets are found resting indiscriminately on Silurian, Carboniferous, and Triassic strata. The general relations of the plateau-basalts to the underlying formations show, that at the close of the Cretaceous period there had been considerable terrestrial disturbances and great subaerial denudation, resulting in some cases in the complete destruction of the whole of the Cretaceous strata, before the lava floods were poured out; owing to which, these latter are found resting on formations of older date than the Cretaceous.
 Mr. J. Starkie Gardner, from a recent comparison of the plant-remains of Antrim and Mull, concludes that "that they might belong to any age between the beginning and the end of the warmer Eocene period; and that they cannot be of earlier, and are unlikely to be of later, date."—Trans. Palaeont. Soc., vol. xxxvii. (1883).
 Having dealt with this district rather fully in The Physical Geology and Geography of Ireland (Edit. 1891, p. 81), and also in my Presidential Address (Section C.) at the meeting of the British Association, 1874, a brief review of the subject will be sufficient here, the reader being referred to the former treatises for fuller details. The following should also be consulted: Gen. Portlock, Geology of Londonderry and Tyrone (1843); Sir A. Geikie, "History of Volcanic Action during the Tertiary Period in the British Isles," Trans. Roy. Soc. Edinburgh, 1888; and the Descriptive Memoirs of the Geological Survey relating to this tract of country.
 Owing to the superposition of the basaltic masses on beds of chalk throughout a long line of coast, we are presented with the curious spectacle of the whitest rocks in nature overlain by the blackest, as may be seen in the cliffs at Larne, Glenarm, Kinbane and Portrush. (See Fig. 27.)
SUCCESSION OF VOLCANIC ERUPTIONS.
(c.) First Stage.—The earliest eruptions of lava in the North-east of Ireland belonged to the highly acid varieties, consisting of quartz-trachyte with tridymite. This rock rises to the surface at Tardree and Brown Dod hills and Templepatrick. It consists of a light-greyish felsitic paste enclosing grains of smoke-quartz, crystals of sanidine, plagioclase and biotite, with a little magnetite and apatite. It is a rock of peculiar interest from the fact that it is almost unique in the British Islands, and has its petrological counterpart rather amongst the volcanic hills of the Siebengebirge than elsewhere. It is generally consolidated with the columnar structure.
The trachyte appears to have been extruded from one or more vents in a viscous condition, the principal vent being probably situated under Tardree mountain, where the rock occurs in greatest mass, and it probably arose as a dome-shaped mass, with a somewhat extended margin, above the floor of Chalk which formed the surface of the ground. (Fig. 27.) At Templepatrick the columnar trachyte may be observed resting on the Chalk, or upon a layer of flint gravel interposed between the two rocks, and which has been thrust out of position by a later intrusion of basalt coming in from the side. It is to be observed, however, that the trachytic lavas nowhere appear cropping out along with the sheets of basalt around the escarpments overlooking the sea, or inland; showing that they did not spread very far from their vents of eruption; a fact illustrating the lower viscosity, or fluidity, of the acid lavas as compared with those of the basic type.
(d.) Second Stage.—After an interval, probably of long duration, a second eruption of volcanic matter took place over the entire area; but now the acid lavas of the first stage are replaced by basic lavas. Now, for the first time, vast masses of basalt and dolerite are extruded both from vents of eruption and fissures; and, owing to their extreme viscosity, spread themselves far and wide until they reach the margin of some uprising ground of old Palaeozoic or Metamorphic rocks by which the volcanic plain is almost surrounded. The great lava sheets thus produced are generally more or less amorphous, vesicular and amygdaloidal, often exhibiting the globular concentric structure, and weathering rapidly to a kind of ferruginous sand or clay under the influence of the atmosphere. Successive extrusions of these lavas produce successive beds, which are piled one over the other in some places to a depth of 600 feet; and at the close of the stage, when the volcanic forces had for the time exhausted themselves, the whole of the North-east of Ireland must have presented an aspect not unlike that of one of those great tracts of similar lava in the region of Idaho and the Snake River in Western America, described in a previous chapter.
(e.) Third Stage (Inter-volcanic).—The third stage may be described as inter-volcanic. Owing to the formation of a basin, probably not deep, and with gently sloping sides, a large lake was formed over the centre of the area above described. Its floor was basalt, and the streams from the surrounding uplands carried down leaves and stems of trees, strewing them over its bed. Occasionally eruptions of ash took place from small vents, forming the ash-beds with plants found at Ballypallidy, Glenarm, and along the coast as at Carrick-a-raide. The streams also brought down sand and gravel from the uprising domes of trachyte, and deposited them over the lake-bed along with the erupted ashes. The epoch we are now referring to was one of economic importance; as, towards its close, there was an extensive deposition of pisolitic iron-ore over the floor of the lake, sometimes to the depth of two or three feet. This ore has been extensively worked in recent years.
(f.) Fourth Stage (Volcanic).—The last stage described was brought to a termination by a second outburst of basic lavas on a scale probably even grander than the preceding. These lavas consisting of basalt and dolerite, with their varieties, and extruded from vents and fissures, spread themselves in all directions over the pre-existing lake deposits or the older sheets of augitic lava, and probably entirely buried the trachytic hills. These later sheets solidified into more solid masses than those of the second stage. They form successive terraces with columnar structure, each terrace differing from that above and below it in the size and length of the columns, and separated by thin bands of "bole" (decomposed lava), often reddish in colour, clearly defining the limits of the successive lava-flows. Nowhere throughout the entire volcanic area are these successive terraces so finely laid open to view as along the north coast of Antrim, where the lofty mural cliffs, worn back into successive bays with intervening headlands by the irresistible force of the Atlantic waves, present to the spectator a vertical section from 300 to 400 feet in height, in which the successive tiers of columnar basalt, separated by thin bands of bole, are seen to rise one above the other from the water's edge to the summit of the cliff, as shown in Fig. 30. Here, also, at the western extremity of the line of cliffs we find that remarkable group of vertical basaltic columns, stretching from the base of the cliff into the Atlantic, and known far and wide by the name of "The Giant's Causeway," the upper ends of the columns forming a tolerably level surface, gently sloping seawards, and having very much the aspect of an artificial tesselated pavement on a huge scale. A portion of the Causeway, with the cliff in the background, is shown in the figure (Fig. 31). The columns are remarkable for their symmetry, being generally hexagonal, though occasionally they are pentagons, and each column is horizontally traversed by joints of the ball-and-socket form, thus dividing them into distinct courses of natural masonry. These are very well shown in the accompanying view of the remarkable basaltic pillars known as "The Chimneys," which stand up from the margin of the headland adjoining the Causeway, monuments of past denudation, as they originally formed individuals amongst the group belonging to one of the terraces in the adjoining coast. (Fig. 32).
(g.) Original Thickness of the Antrim Lavas.—It is impossible to determine with certainty what may have been the original thickness of the accumulated sheets of basic lavas with their associated beds of ash and bole. The greatest known thickness of the lower zone of lavas is, as I have already stated, about 600 feet. The intermediate beds of ash and bole sometimes attain a thickness of 40 feet, and the upper group of basalt about 400 feet; these together would constitute a series of over 1,000 feet in thickness. But this amount, great as it is, is undoubtedly below the original maximum, as the uppermost sheets have been removed by denuding agencies, we know not to what extent. Nor is it of any great importance. Sufficient remains to enable us to form a just conception of the magnitude both as regards thickness and extent of the erupted matter of the Miocene period over the North-east of Ireland and adjoining submerged tracts, and of the magnitude of the volcanic operations necessary for the production of such masses.
(h.) Volcanic Necks.—As already remarked, no craters of eruption survive throughout the volcanic region of the North-east of Ireland, owing to the enormous extent of the denudation which this region has undergone since the Miocene Epoch; but the old "necks" of such craters—in other words, the pipes filled with either solid basalt, or basalt and ashes—are still to be found at intervals over the whole area. Owing to the greater solidity of the lava which filled up these "necks" over the plateau-basaltic sheets which surround them, they appear as bosses or hills rising above the general level of the ground. One of these bosses of highly columnar basalt occurs between Portrush and Bushmills, not far from Dunluce Castle, another at Scawt Hill, near Glenarm, and a third at Carmoney Hill above Belfast Lough. But by far the most prominent of these old solidified vents of eruption is that of Sleamish, a conspicuous mountain which rises above the general level of the plateau near Ballymena, and attains an elevation of 1,437 feet above the sea. Seen from the west, the mountain has the appearance of a round-topped cone; but on examination it is found to be in reality a huge dyke, breaking off abruptly towards the north-west, in which direction it reaches its greatest height, then sloping downwards towards the east. This form suggests that Sleamish is in reality one of the fissure-vents of eruption rather than the neck of an old volcano. The rock of which it is formed consists of exceedingly massive, coarsely-crystalline dolerite, rich in olivine, and divided into large quadrangular blocks by parallel joint planes. Its junction with the plateau-basalt from which it rises can nowhere be seen; but at the nearest point where the two rocks are traceable the plateau-basalt appears to be somewhat indurated; breaking with a splintery fracture and a sharp ring under the hammer, suggesting that the lava of Sleamish had been extruded through the horizontal sheets, and had considerably indurated the portions in contact with, or in proximity to, it. Amongst the vents filled with ash and agglomerate, the most remarkable is that of Carrick-a-raide, near Ballycastle. It forms this rocky island and a portion of the adjoining coast, where the beds of ash are finely displayed; consisting of fragments and bombs of basalt, with pieces of chalk, flint, and peperino, which is irregularly bedded. These ash-beds attain a thickness of about 120 feet just below the road to Ballycastle, but rapidly tail out in both directions from the locality of the vent. Just below the ash-beds, the white chalk with flints may be seen extending down into the sea-bed. Nowhere in Antrim is there such a display of volcanic ash and agglomerate as at this spot.
(i.) Dykes: Conditions under which they were Erupted.—No one can visit the geological sections in Co. Antrim and the adjoining districts of Down, Armagh, Derry, and Tyrone, without being struck by the great number and variety of the igneous dykes by which the rocks are traversed. The great majority of these dykes are basaltic, and they are found traversing all the formations, including the Cretaceous and Tertiary basaltic sheets. The Carlingford and Mourne Mountains are seamed with such dykes, and they are splendidly laid open to view along the coast south of Newcastle in Co. Down, as also along the Antrim coast from Belfast to Larne. The fine old castle of Carrickfergus has its foundations on one of those dyke-like intrusions, but one of greater size than ordinary. All the dykes here referred to are not, however, of the same age, as is conclusively proved by sections amongst the Mourne Mountains where cliffs of Lower Silurian strata, superimposed on the intrusive granite of the district, exhibit two sets of basaltic dykes—one (the older) abruptly terminated at the granite margin, the other and newer penetrating the granite and Silurian rocks alike. It is not improbable that the older dykes belong to the Carboniferous or Permian age, while the newer are with equal probability of Tertiary age. Sir A. Geikie has shown that the Tertiary dykes of the North of Ireland are representatives of others occurring at intervals over the North of England, and Central and Western Scotland, all pointing towards the central region of volcanic activity; or in a parallel direction thereto, approximating to the N.W. in Ireland, the Island of Islay, and East Argyleshire, but in the centre of Scotland generally ranging from east to west. The area affected by the dykes of undoubted Tertiary age Geikie estimates at no less than 40,000 square miles—a territory greater than either Scotland or Ireland, and equal to more than a third of the total land-surface of the British Isles; and he regards them as posterior "to the rest of the geological structures of the regions which they traverse." It is clear that the dykes referred to belong to one great system of eruption or intrusion; and they may be regarded as the manifestation of the final effort of internal forces over this region of the British Isles. They testify to the existence of a continuous magma (or shell) of augitic lava beneath the crust; and as the aggregate horizontal extent of all these dykes, or of the fissures which they fill, must be very considerable, it is clear that the crust through which they have been extruded has received an accession of horizontal space, and has been fissured by forces acting from beneath, as the late Mr. Hopkins, of Cambridge, had explained on mechanical grounds in his elaborate essay many years ago. This view occurred to myself when examining the region of the North-east of Ireland, but I was not then aware that it had been dealt with on mathematical principles by so eminent a mathematician. The bulging of the crust is a necessary consequence of the absence of plication of the strata due to the extrusion of this enormous quantity of molten lava; and the intrusion of thousands of dykes over the North-east of Ireland, unaccompanied by foldings of the strata, must have added a horizontal space of several thousand feet to that region.
 A peculiar form of crystalline quartz first recognized in this rock by a distinguished German petrologist, the late Prof. A. von Lasaulx, who visited the district in 1876.
 Sir A. Geikie has disputed the correctness of the view, which I advocated as far back as 1874, that the trachytic lavas of Antrim are the earliest products of volcanic action; but at the time he wrote his paper on the volcanic history of these islands, it was not known that pebbles of this trachyte are largely distributed amongst the ash-beds which occur in the very midst of the overlying basaltic sheets, as I shall have to explain later on. This discovery puts the question at rest as regards the relations of the two sets of rocks.
 This remarkable section at the chalk quarries of Templepatrick the author has figured and described in the Physical Geology and Geography of Ireland, p. 99, 2nd edit. (1891), where the reader will find the subject discussed more fully than can be done here.
 These pebbles were first noticed by Mr. McHenry, of the Irish Geological Survey, in 1890.
 The vertical position of the columns of the Giant's Causeway is rather enigmatical. The Causeway cannot be a dyke, as has often been supposed, otherwise the columns would have been horizontal, i.e., at right angles to the sides of the dyke. Mr. R. G. Symes, of the Geological Survey, has suggested that the Causeway columns have been vertically lowered between two lines of fault, and that originally they formed a portion of the tier of beautiful columns seen in the cliff above, and known as "The Organ."
 Sleamish and several other of the Antrim vents are described by Sir A. Geikie in the monograph already referred to, loc. cit., p. 101, et seq. Also in the Expl. Memoirs of the Geological Survey of Ireland.
 A diagrammatised section of the Carrick-a-raide volcanic neck is given by Sir A. Geikie, loc. cit., p. 105.
 Geikie, loc. cit., p. 29, et seq.
 P. 32. The view that the crust of the earth has been horizontally extended by the intrusion of dykes is noticed by McCulloch in reference to the dykes of Skye.
 Hopkins, Cambridge Phil. Trans., vol. vi. p. 1 (1836).
 As suggested in my Presidential Address to Section C. of the British Association at Belfast, 1874.
ISLAND OF MULL AND ADJOINING COAST.
The Island of Mull, with the adjoining districts of Morvern and Ardnamurchan, forms the more southern of the two chief centres of Tertiary volcanic eruptions in the West of Scotland, that of Skye being the more northern. These districts have been the subject of critical and detailed study by several geologists, from McCulloch down to the present day; and amongst the more recent, Sir Archibald Geikie and Professor Judd hold the chief place. Unfortunately, the interpretation of the volcanic phenomena by these two accomplished observers has led them to very different conclusions as regards several important points in the volcanic history of these groups of islands; as, for example, regarding the relative ages of the plateau-basalts and the acid rocks, such as the trachytes and granophyres; again as regards the presence of distinct centres of eruption; and also as regards the relations of the gabbros of Skye to the basaltic sheets. Such being the case, it would appear the height of rashness on the part of the writer, especially in the absence of a detailed examination of the sections over the whole region, to venture on a statement of opinion regarding the points at issue; and he must, therefore, content himself with a brief account of the phenomena as gathered from a perusal of the writings of these and other observers, guided also to some extent by the analogous phenomena presented by the volcanic region of the North-east of Ireland.
(a.) General Features.—As in the case of the Antrim district, the Island of Mull and adjoining tracts present us with the spectacle of a vast accumulation of basaltic lava-flows, piled layer upon layer, with intervening beds of bole and tuff, up to a thickness, according to Geikie, of about 3,500 feet. At the grand headland of Gribon, on the west coast, the basaltic sheets are seen to rise in one sheer sweep to a height of 1,600 feet, and then to stretch away with a slight easterly dip under Ben More at a distance of some eight miles. This mountain, the upper part of which is formed of beds of ashes, reaches an elevation of 3,169 feet, so that the accumulated thickness of the beds of basalt under the higher part of the mountain must be at least equal to the amount stated above—that is, twice as great as the representative masses of Antrim. The base of the volcanic series is seen at Carsaig and Gribon to rest on Cretaceous and Jurassic rocks, like those of Antrim; hence the Tertiary age is fully established by the evidence of superposition. This was further confirmed by the discovery by the Duke of Argyll, some years ago (1850), of bands of flint-gravel and tuff, with dicotyledonous leaves amongst the basalts of Ardtun Head. The basement beds of tuff and gravel contain, besides pebbles of flint and chalk, others of sanidine trachyte, showing that highly acid lavas had been extruded and consolidated before the first eruption of the plateau-basalts; another point of analogy between the volcanic phenomenon of Antrim and the Inner Hebrides. These great sheets of augitic lava extend over the whole of the northern tract of Mull, the Isles of Ulva and Staffa, and for a distance of several miles inwards from the northern shore of the Sound of Mull, covering the wild moorlands of Morvern and Ardnamurchan, where they terminate in escarpments and outlying masses, indicating an originally much more extended range than at the present day. The summits of Ben More and its neighbouring height, Ben Buy, are formed of beds of ash and tuff. The volcanic plateau is, according to Judd, abruptly terminated along the southern side by a large vault, bringing the basalt in contact with Palaeozoic rocks.
(b.) Granophyres.—The greater part of the tract lying to the south of Loch na Keal, which almost divides Mull into two islands, and extending southwards and eastwards to the shores of the Firth of Lorn and the Sound of Mull, is formed of a peculiar group of acid (or highly silicated) rocks, classed under the general term of "Granophyres." These rocks approach towards true granites in one direction, and through quartz-porphyry and felsite to rhyolite in another—probably depending upon the conditions of cooling and consolidation. In their mode of weathering and general appearance on a large scale, they present a marked contrast to the basic lavas with which they are in contact from the coast of L. na Keal to that of L. Buy. The nature of this contact, whether indicating the priority of the granophyres to the plateau-basalts or otherwise, is a matter of dispute between the two observers above named; but the circumstantial account given by Sir A. Geikie, accompanied by drawings of special sections showing this contact, appears to prove that the granophyre is the newer of the two masses of volcanic rock, and that it has been intruded amongst the basaltic-lavas at a late period in the volcanic history of these islands. A copy of one of these sketches is here given (Fig. 33), according to which the felsite is shown to penetrate the basaltic sheets at Alt na Searmoin in Mull; other sections seen at Cruach Torr an Lochain, and on the south side of Beinn Fada, appear to lead to similar conclusions. These rocks are penetrated by numerous basaltic dykes.
(c.) Representative Rocks of Mourne and Carlingford, Ireland.—Assuming Sir A. Geikie's view to be correct, it is possible that we may have in the granite and quartz-porphyries of Mourne and Carlingford representatives of the granites, granophyres, and other acid rocks of the later period of Mull. The granite of Mourne is peculiar in structure, and differs from the ordinary type of that rock in which the silica forms the ground mass. In the case of the granite of the Mourne Mountains, the rock consists of a crystalline granular aggregate of orthoclase, albite, smoke-quartz, and mica; it is also full of drusy cavities, in which the various minerals crystallise out in very perfect form. As far as regards direct evidence, the age of this rock can only be stated to be post-Carboniferous, and earlier than certain Tertiary basaltic dykes by which it is traversed. The granophyres of Mull are traversed by similar dykes, which are representatives of the very latest stage of volcanic action in the British Islands. The author is therefore inclined to concur with Sir A. Geikie in assigning to the granite of the Mourne Mountains, and the representative felsitic rocks of the Carlingford Mountains, a Tertiary age—in which case the analogy between the volcanic phenomena of the Inner Hebrides and of the North-east of Ireland would seem to be complete.
 Geikie, Proc. Roy. Soc. Edinburgh (1867); Brit. Assoc. Rep. (Dundee, 1867); "Tertiary Volcanic Rocks of the British Isles," Quart. Journ. Geol. Soc., vol. xxvii. p. 279; also, "History of Volcanic Action in British Isles," Trans. Roy. Soc. Edin. (1888); Judd, "On the Ancient Volcanoes of the Highlands," etc., Quart. Journ. Geol. Soc., vol. xxx. p. 233; and Volcanoes, p. 139.
 Brit. Assoc. Rep. for 1850, p. 70.
 Judd, Quart. Jour. Geol. Soc., vol. xxx. p. 242.
 History of Volcanic Action, etc., loc. cit. p. 153, et seq. The "Granophyres" of Geikie come under the head of "Felsites," passing into "granite" in one direction and quartz-trachyte in another, according to Judd; the proportion of silica from 69 to 75 per cent.—Quart. Jour. Geol. Soc., vol. xxx. p. 235.
 This view the author has expressed in a recent edition of The Physical Geology of Ireland, p. 177 (1891).
ISLE OF SKYE.
This is the largest and most important of all the Tertiary volcanic districts, but owing to the extensive denudation to which, in common with other Tertiary volcanic regions of the British Isles, it has been subjected, its present limits are very restricted comparatively to its original extent. Not only is this evident from the manner in which the basaltic sheets terminate along the sea-coast in grand mural cliffs, as opposite "Macleod's Maidens," and at the entrance to Lough Bracadale on the western coast, but the evidence is, according to Sir A. Geikie, still more striking along the eastern coast; showing that the Jurassic, and other older rocks there visible, were originally buried deep under the basaltic sheets which have been stripped from off that part of the country. These great plateau-basalts occupy about three-fourths of the entire island along the western and northern areas, rising into terraced mountains over 2,000 feet in height, and are deeply furrowed by glens and arms of the sea, along which the general structure of the tableland is laid open, sometimes for leagues at a time.
It is towards the south-eastern part of the island that the most interesting and important phenomena are centred; for here we meet with representatives of the acid (or highly silicated) group of rocks, and of remarkable beds of gabbro, which have long attracted the attention of petrologists. These latter beds, throughout a considerable distance round the flanks of the Cuillin Hills, are interposed between the acid rocks and the plateau-basalts; but towards the north, on approaching Lough Sligahan, the acid rocks, consisting of granophyres, quartz-porphyries, and hornblendic-granitites, are in direct contact with the plateau-basalts; and, according to the very circumstantial account of Sir A. Geikie, are intrusive into them; not only sending veins into the basaltic sheets, but also producing a marked alteration in their structure where they approach the newer intrusive mass. Equally circumstantial is the same author's account of the relations of the granophyres to the gabbros, as seen at Meall Dearg and the western border of the Cuillin Hills—where the former rock may be seen to send numerous veins into the latter. Not only is this so, but the granophyre is frequently seen to truncate, and abruptly terminate some of the basaltic dykes by which the basic sheets are traversed—as in the neighbourhood of Beinn na Dubhaic. All these phenomena strongly remind us of the conditions of similar rocks amongst the mountains of Mourne and Carlingford in Ireland; where, at Barnaveve, the syenite (or hornblendic quartz-felsite) is seen to break through the masses of olivine gabbro, and send numerous veins into this latter rock.
The interpretation here briefly sketched differs widely from that arrived at by Professor Judd. The granitoid masses of the Red Mountains (Beinn Dearg) and the neighbouring heights are, in his view, the roots of the great volcano from which were erupted the various lavas; the earlier eruptions producing the acid lavas, to be followed by the gabbros, and these by the plateau-basaltic sheets, which stretch away towards the north and west into several peninsulas. Thus he holds that "the rocks of basic composition were ejected subsequently to those of the acid variety," and appeals to various sections in confirmation of this view. To reconcile these views is at present impossible; but as the controversy between these two observers is probably not yet closed, there is room for hope that the true interpretation of the relations of these rocks to each other will ere long be fully established.
 Geikie, loc. cit., p. 161, etc.
 Physical Geology of Ireland, 2nd edition, p. 174 (Fig. 21). Professor Judd has also come to the conclusion that the granite of Mourne is of Tertiary age, Quart. Jour. Geol. Soc., vol. xxx. p. 275.
 Judd, loc. cit., p. 254.
THE SCUIR OF EIGG.
Amongst the more remarkable of the smaller islets are those of Eigg, Rum, Canna, and Muck, lying between Mull on the south and Skye on the north, and undoubtedly at one time physically connected together. The Island of Eigg is especially remarkable for the fact, as stated by Geikie, that here we have the one solitary case of "a true superficial stream of acid lava—that of the Scuir of Eigg." (Fig. 34.) This forms a sinuous ridge, composed of pitchstone of several kinds, of over two miles in length, rising from the midst of a tableland of bedded basalt and tuff to a height of 1,289 feet above the ocean; the plateau-basalt is traversed by basaltic dykes, ranging in a N.W.-S.E. direction. But what is specially remarkable is the evidence afforded by an examination of the course of the Scuir, that it follows the channel of an ancient river-valley, which has been hollowed out in the surface of the plateau. The course of this channel is indicated by the presence of a deposit of river-gravel, which in some places forms a sort of cushion between the base of the Scuir and the side of the channel. Over this gravel-bed the viscous pitchstone-lava appears to have flowed, taking possession of the river-channel, and also of the beds of several small tributary streams which flowed into the channel of the Scuir. The recent date of the pitchstone forming this remarkable mural ridge, once occupying the bed of a river-channel, is shown by the fact that the basaltic dykes which traverse the plateau-basalts are truncated by the river-gravel, which is, therefore, more recent; and, as we have seen, the pitchstone stream is more recent than the river-gravel. But at the time when this last volcanic eruption took place, the physical geography of the whole region must have been very different from that of the present time. From the character and composition of the pebbles in the old river-bed, amongst which are Cambrian sandstone, quartzite, clay-slate, and white Jurassic limestone, Sir A. Geikie concludes that when the river was flowing, the island must have been connected with the mainland to the east where the parent masses of these pebbles are found.
Effects of Denudation.—The position of the Scuir of Eigg and its relations to the basaltic sheets show the enormous amount of denudation which these latter have undergone since the stream of pitchstone-lava filled the old river channel. The walls, or banks, of the channel have been denuded away, thus converting the pitchstone casting into a projecting wall of rock. That it originally extended outwards into the ocean to a far greater distance than at present is evident from the abruptly truncated face of the cliff; and yet this remarkable volcanic mass seems to have been, perhaps, the most recent exhibition of volcanic action to be found in the British Isles. It is perhaps, on this account, the most striking of the numerous examples exhibited throughout the West of Scotland and the North-east of Ireland of the enormous amount of denudation to which these districts have been subjected since the extinction of the volcanic fires; and this at a period to which we cannot assign a date more ancient than that of the Pliocene. Yet, let us consider for a moment to what physical vicissitudes these districts have been subjected since that epoch. Assuming, as we may with confidence, that the volcanic eruptions were subaerial, and that the tracts covered by the plateau-basalts were in the condition of dry land when the eruptions commenced, in this condition they continued in the main throughout the period of volcanic activity. But the eruptions had scarcely ceased, and the lava floods and dykes become consolidated, before the succeeding glacial epoch set in; when the snows and glaciers of the Scottish Highlands gradually descending from their original mountain heights, and spreading outwards in all directions, ultimately enveloped the whole of the region we are now considering until it was entirely concealed beneath a mantle of ice moving slowly, but irresistibly, outwards towards the Atlantic, crossing the deep channels, such as the Sound of Mull and the Minch, climbing up the sides of opposing rocks and islands until even the Outer Hebrides and the North-east of Ireland were covered by one vast mantle of ice and snow. The movement of such a body of ice over the land must have been attended with a large amount of abrasion of the rocky floor; nor have the evidences of that abrasion entirely disappeared even at the present day. We still detect the grooves and scorings on the rock-surfaces where they have been protected by a coating of boulder clay; and we still find the surface strewn with the blocks and debris of that mighty ice-flood.
But whatever may have been the amount of erosion caused by the great ice-sheet, it was chiefly confined to the more or less horizontal surface-planes. Erosion of another kind was to succeed, and to produce more lasting effects on the configuration of the surface. On the disappearance of the ice-sheet, an epoch characterised by milder conditions of climate set in. This was accompanied by subsidence and submersion of large tracts of the land during the Interglacial stage; so that the sea rose to heights of several hundred feet above the present level, and has left behind stratified gravels with shells at these elevations in protected places. During this period of depression and of subsequent re-emergence the wave-action of the Atlantic waters must have told severely on the coast and islands, wearing them into cliffs and escarpments, furrowing out channels and levelling obstructions. Such action has gone on down to the present day. The North-west of Scotland and of Ireland has been subjected throughout a very lengthened period to the wear and tear of the Atlantic billows. In the case of the former, the remarkable breakwater which nature has thrown athwart the North-west Highlands in the direction of the waves, forming the chain of islands constituting the Outer Hebrides, and composed of very tough Archaean gneiss and schist, has done much to retard the inroads which the waves might otherwise have made on the Isle of Skye; while Coll and Tiree, composed of similar materials, have acted with similar beneficent effect for Mull and the adjoining coasts. But such is the tremendous power of the Atlantic billows when impelled by westerly winds, that to their agency must be mainly attributed the small size of the volcanic land-surfaces as compared with their original extent, and the formation of those grand headlands which are presented by the igneous masses of Skye, Ardnamurchan, and Mull towards the west. Rain and river action, supplemented by that of glaciers, have also had a share in eroding channels and wearing down the upper surface of the ground, with the result we at present behold in the wild and broken scenery of the Inner Hebrides and adjoining coast.
 Geikie, loc. cit., p. 178; also Quart. Jour. Geol. Soc., vol. xxvii. p. 303.
ISLE OF STAFFA.
Reference has been made to this remarkable island in a former page, but some more extended notice is desirable before leaving the region of the Inner Hebrides. Along with the islands of Pladda, Treshnish, and Blackmore, Staffa is one of the outlying volcanic islands of the group, being distant about six miles from the coast of Mull, and indicates the minimum distance to which the plateau-basaltic sheets originally extended in the direction of the old marginal lands of Tiree and Coll. The island consists of successive sheets of bedded basaltic lava, with partings of tuff, one of which of considerable thickness is shown to lie at the base of the cliff on the south-west side of the island. The successive lava-sheets present great varieties of structure, like those on the north coast of Antrim; some being amorphous, others columnar, with either straight or bent columns. The lava-sheet out of which Fingal's Cave is excavated consists of vertical prisms, beautifully formed, and surmounted by an amorphous mass of the same material. At the entrance of the Boat Cave we have a somewhat similar arrangement of the columns; but at the Clam-shell Cave the prisms are curved, indicating some movement in the viscous mass before they had been fully consolidated.
Fingal's Cave is called after the celebrated prince of Morvern (or Morven), a province of ancient Caledonia. He is supposed to have been the father of Ossian, the Celtic bard rendered famous by Macpherson. The cave, one of many which pierce the coast-cliffs of Western Scotland, is 227 feet in length, 166 feet in height, and 40 feet in width. On all sides regular columns of basalt, some entire, others broken, rise out of the water and support the roof. The cave is only accessible in calm weather.
 A drawing of this cliff is given by Geikie in the Manual of Geology (Jukes and Geikie), 3rd edition, p. 277.
 Prestwich, Geology, vol. i. p. 281, where a view of this cave is given.
PRE-TERTIARY VOLCANIC ROCKS.
THE DECCAN TRAP-SERIES OF INDIA.
The great outpourings of augitic lava of Tertiary and recent times which we have been considering appear to have been anticipated in several parts of the world, more especially in Peninsular India and in Africa, and it is desirable that we should devote a few pages to the description of these remarkable volcanic formations, as they resemble, both in their mode of occurrence and general structure, some of the great lava-floods of a more recent period we have been considering. Of the districts to be described, the first which claims our notice is the Deccan.
(a.) Extent of the Volcanic Plateau.—The volcanic plateau of the Deccan stretches from the borders of the Western Ghats and the sea-coast near Bombay inland to Amarantak, at the head of the Narbudda River (long. 82 deg. E.), and from Belgaum (lat. 15 deg. 31' N.) to near Goona (lat. 24 deg. 30'). The vast area thus circumscribed is far from representing the original extent of the tract overspread by the lava-floods, as outlying fragments of these lavas are found as far east as long. 84 deg. E. in one direction, and at Kattiwar and Cutch in another. The present area, however, is estimated to be not less than 200,000 square miles.
(b.) Nature and Thickness of the Lava-flows.—This tract is overspread almost continuously by sheets of basaltic lava, with occasional bands of fresh-water strata containing numerous shells, figured and described by Hislop, and believed by him to be of Lower Eocene age. The lava-sheets vary considerably in character, ranging from finest compact basalt to coarsely crystalline dolerite, in which olivine is abundant. The columnar structure is not prevalent, the rock being either amorphous, or weathering into concentric shells. Volcanic ash, or bole, is frequently found separating the different lava-flows; and in the upper amygdaloidal sheets numerous secondary minerals are found, such as quartz, agate and jasper, stilbite and chlorite. The total thickness of the whole series, where complete, is about 6,000 feet, divided as follows:
1. Upper trap; with ash and inter-trappean beds 1,500 feet 2. Middle trap; sheets of basalt and ash 4,000 " 3. Lower trap; basalt with inter-trappean beds 500 " ———— 6,000 " ========
Throughout the region here described these great sheets of volcanic rock are everywhere approximately horizontal, and constitute a table-land of 3,000 to 4,000 feet in elevation, breaking off in terraced escarpments, and penetrated by deep river-valleys, of which the Narbudda is the most important. The foundation rock is sometimes metamorphic schist, or gneiss, at other times sandstone referred by Hislop to Jurassic age; and in no single instance has a volcanic crater or focus of eruption been observed. But outside the central trappean area volcanic foci are numerous, as in Cutch, the Rajhipla Hills and the Lower Narbudda valley. The original excessive fluidity of the Deccan trap is proved by the remarkable horizontality of the beds over large areas, and the extensive regions covered by very thin sheets of basalt or dolerite.
(c.) Geological Age.—As regards the geological age of this great volcanic series much uncertainty exists, owing to the absence of marine forms in the inter-trappean beds. One single species, Cardita variabilis, has been observed as occurring in these beds, and in the limestone below the base of the trap at Dudukur. The facies of the forms in this limestone is Tertiary; but there is a remarkable absence of characteristic genera. On the other hand, Mr. Blanford states that the bedded traps are seen to underlie the Eocene Tertiary strata with Nummulites in Guzerat and Cutch, which would appear to determine the limit of their age in one direction. On balancing the evidence, however, it is tolerably clear that the volcanic eruptions commenced towards the close of the Cretaceous period, and continued into the commencement of the Tertiary, thus bridging over the interval between the two epochs; and since the greater sheets have been exposed throughout the whole of the Tertiary and Quarternary periods, it is not surprising if they have suffered enormously from denuding agencies, and that any craters or cones of eruption that may once have existed have disappeared.
 The Deccan Traps have been described by Sykes, Geol. Trans., 2nd Series, vol. iv.; also Rev. S. Hislop, "On the Geology of the Neighbourhood of Nagpur, Central India," Quart. Journ. Geol. Soc., vol. x. p. 274; and Ibid., vol. xvi. p. 154. Also, H. B. Medlicott and W. T. Blanford, Manual of the Geology of India, vol. i. (1879).
 Blanford, Geology of Abyssinia, p. 185.
Another region in which the volcanic phenomena bear a remarkable analogy to those of Central India, just described, is that of Abyssinia. Nor are these tracts so widely separated that they may not be considered as portions of one great volcanic area extending from Abyssinia, through Southern Arabia, into Cutch and the Deccan, in the one direction, while the great volcanic cones of Kenia and Kilimanjaro, with their surrounding tracts of volcanic matter, may be the extreme prolongations in the other. Along this tract volcanic operations are still active in the Gulf of Aden; and cones quite unchanged in form, and evidently of very recent date, abound in many places along the coast both of Arabia and Africa. The volcanic formations of this tract are, however, much more recent than those which occupy the high plateaux of Central and Southern Abyssinia of which we are about to speak.
(a.) Physical Features.—Abyssinia forms a compact region of lofty plateaux intersected by deep valleys, interposed between the basin of the Nile on the west, and the low-lying tract bordering the Red Sea and the Indian Ocean on the east. The plateaux are deeply intersected by valleys and ravines, giving birth to streams which feed the head waters of the Blue Nile (Bahr el Arak) and the Atbara. Several fine lakes lie in the lap of the mountains, of which the Zana, or Dembia, is the largest, and next Ashangi, visited by the British army on its march to Magdala in 1868, and which, from its form and the volcanic nature of the surrounding hills, appears to occupy the hollow of an extinct crater. The table-land of Abyssinia reaches its highest elevation along the eastern and southern margin, where its average height may be 8,000 to 10,000 feet; but some peaks rise to a height of 12,000 to 15,000 feet in Shoa and Ankobar.
(b.) Basaltic Lava Sheets.—An enormous area of this country seems to be composed of volcanic rocks chiefly in the form of sheets of basaltic lava, which rise into high plateaux, and break off in steep—sometimes precipitous—mural escarpments along the sides of the valleys. These are divisible into the following series:—
(1) The Ashangi Volcanic Series.—The earliest forerunners of the more recent lavas seem to have been erupted in Jurassic times, in the form of sheets of contemporaneous basalt or dolerite amongst the Antola limestones which are of this period. But the great mass of the volcanic rocks are much more recent, and may be confidently referred to the late Cretaceous or early Tertiary epochs. Their resemblance to the great trappean series of Western India, even in minute particulars, is referred to by Mr. Blanford, who suggests the view that they belong to one and the same great series of lava-flows extruded over the surface of this part of the globe. This view is inherently probable. They consist of basalts and dolerites, generally amygdaloidal, with nodules of agate and zeolite, and are frequently coated with green-earth (chlorite). Beds of volcanic ash or breccia also frequently occur, and often contain augite crystals. At Senafe, hills of trachyte passing into claystone and basalt were observed by Mr. Blanford, but it is not clear what are their relations to the plateau-basaltic sheets.
(2) Magdala Volcanic Series.—This is a more recent group of volcanic lavas, chiefly distinguished from the lower, or Ashangi, group, by the occurrence of thick beds of trachyte, usually more or less crystalline, and containing beautiful crystals of sanidine. The beds of trachyte break off in precipitous scarps, and being of great thickness and perfectly horizontal, are unusually conspicuous. Mr. Blanford says, with regard to this group, that there is a remarkable resemblance in its physical aspect to the scenery of the Deccan and the higher valleys of the Western Ghats of India, but the peculiarities of the landscape are exaggerated in Abyssinia. Many of the trachytic beds are brecciated and highly columnar; sedimentary beds are also interstratified with those of volcanic origin. The Magdala group is unconformable to that of Ashangi in some places. A still more recent group of volcanic rocks appears to occur in the neighbourhood of Senafe, consisting of amorphous masses of trachyte, often so fine-grained and compact as to pass into claystone and to resemble sandstone. At Akub Teriki the rocks appear to be in the immediate vicinity of an ancient vent of eruption.
From what has been said, it will be apparent that Abyssinia offers volcanic phenomena of great interest for the observer. There is considerable variety in the rock masses, in their mode of distribution, and in the scenery which they produce. The extensive horizontal sheets of lava are suggestive of fissure-eruption rather than of eruption through volcanic craters; and although these may have once been in existence, denudation has left no vestiges of them at the present day. In all these respects the resemblance of the volcanic phenomena to those of Peninsular India is remarkably striking; it suggests the view that they are contemporaneous as regards the time of their eruption, and similar as regards their mode of formation.
 W. T. Blanford, Geology of Abyssinia, pp. 151-2.
 Blanford, loc. cit., p. 182.
Basalt of the Plateau.—The extensive sheets of plateau-basalt forming portions of the Neuweld range and the elevated table-land of Cape Colony, may be regarded as forerunners of those just described, and possibly contemporaneous with the Ashangi volcanic series of Abyssinia. The great basaltic sheets of the Cape Colony are found capping the highest elevations of the Camderboo and Stormberg ranges, as well as overspreading immense areas of less elevated land, to an extent, according to Professor A. H. Green, of at least 120,000 square miles. Amongst these sheets, innumerable dykes, and masses of solid lava which filled the old vents of eruption, are to be observed. The floor upon which the lava-floods have been poured out generally consists of the "Cave Sandstone," the uppermost of a series of deposits which had previously been laid down over the bed of an extensive lake which occupied this part of Africa during the Mesozoic period. After the deposition of this sandstone, the volcanic forces appear to have burst through the crust, and from vents and fissures great floods of augitic lava, with beds of tuff, invaded the region occupied by the waters of the lake. The lava-sheets have since undergone extensive denudation, and are intersected by valleys and depressions eroded down through them into the sandstone floor beneath; and though the precise geological period at which they were extruded must remain in doubt, it appears probable that they may be referred to that of the Trias.
 Green. "On the Geology of the Cape Colony," Quart. Jour. Geol. Soc., vol. xliv. (1888).
 The district lying along the south coast of Africa is described by Andrew G. Bain, in the Trans. Geol. Soc., vol. vii. (1845); but there is little information regarding the volcanic region here referred to.
VOLCANIC ROCKS OF PAST GEOLOGICAL PERIODS OF THE BRITISH ISLES.
It is beyond the scope of this work to describe the volcanic rocks of pre-Tertiary times over various parts of the globe. The subject is far too large to be treated otherwise than in a distinct and separate essay. I will therefore content myself with a brief enumeration of the formations of the British Isles in which contemporaneous volcanic action has been recognised.
There is little evidence of volcanic action throughout the long lapse of time extending backwards from the Cretaceous to the Triassic epochs, that is to say, throughout the Mesozoic or Secondary period, and it is not till we reach the Palaeozoic strata that evidence of volcanic action unmistakably presents itself.
Permian Period.—In Ayrshire, and in the western parts of Devonshire, beds of felspathic porphyry, felstone and ash are interstratified with strata believed to be of Permian age. In Devonshire these have only recently been recognised by Dr. Irving and the author as of Permian age, the strata consisting of beds of breccia, lying at the base of the New Red Sandstone. Those of Ayrshire have long been recognised as of the same period; as they rest unconformably on the coal measures, and consist of porphyrites, melaphyres, and tuffs of volcanic origin.
Carboniferous Period.—Volcanic rocks occur amongst the coal-measures of England and Scotland, while they are also found interbedded with the Carboniferous Limestone series in Derbyshire, Scotland, and Co. Limerick in Ireland. The rocks consist chiefly of basalt, dolerite, melaphyre and felstone.
Devonian Period.—Volcanic rocks of Devonian age occur in the South of Scotland, consisting of felstone-porphyries and melaphyres; also at Boyle, in Roscommon, and amongst the Glengariff beds near Killarney in Ireland.
Upper Silurian Period.—Volcanic rocks of this stage are only known in Ireland, on the borders of Cos. Mayo and Galway, west of Lough Mask, and at the extreme headland of the Dingle Promontory in Co. Kerry. They consist of porphyrites, felstones and tuffs, or breccias, contemporaneously erupted during the Wenlock and Ludlow stages. Around the flanks of Muilrea, beds of purple quartz-felstone with tuff are interstratified with the Upper Silurian grits and slates.
Lower Silurian Period.—Volcanic action was developed on a grand scale during the Arenig and Caradoc-Bala stages, both in Wales and the Lake district, and in the Llandeilo stage in the South of Scotland. The felspathic lavas, with their associated beds of tuff and breccia, rise into some of the grandest mountain crests of North Wales, such as those of Cader Idris, Aran Mowddwy, Arenig and Moel Wyn. A similar series is also represented in Ireland, ranging from Wicklow to Waterford, forming a double group of felstones, porphyries, breccias, and ash-beds, with dykes of basalt and dolerite. The same series again appears amidst the Lower Silurian beds of Co. Louth, near Drogheda.
Metamorphic Series presumably of Lower Silurian Age.—If, as seems highly probable, the great metamorphic series of Donegal and Derry are the representatives in time of the Lower Silurian series, some of the great sheets of felspathic and hornblendic trap which they contain are referable to this epoch. These rocks have undergone a change in structure along with the sedimentary strata of which they were originally formed, so that the sheets of (presumably) augitic lava have been converted into hornblende-rock and schist. Similar masses occur in North Mayo, south of Belderg Harbour.
Cambrian Period.—In the Pass of Llanberis, along the banks of Llyn Padarn, masses of quartz-porphyry, felsite and agglomerate, or breccia, indicate volcanic action during this stage. These rocks underlie beds of conglomerate, slate and grit of the Lower Cambrian epoch, and, as Mr. Blake has shown, are clearly of volcanic origin, and pass upwards into the sedimentary strata of the period. A similar group, first recognised by Professor Sedgwick, stretches southwards from Bangor along the southern shore of the Menai Straits. Again, we find the volcanic eruptions of this epoch at St. David's, consisting of diabasic and felsitic lava, with beds of ash; and in the centre of England, amongst the grits and slates of Charnwood Forest presumably of Cambrian age, various felstones, porphyries, and volcanic breccias are found.
Thus it will be seen that every epoch, from the earliest stage of the Cambrian to the Permian, in the British Isles, gives evidence of the existence of volcanic action; from which we may infer that the originating cause, whatever it may be, has been in operation throughout all past geological time represented by living forms. The question of the condition of our globe in Archaean times, and earlier, is one which only can be discussed on theoretic ground, and is beyond the scope of this work.
 The reader is referred to Sir A. Geikie's Presidential Address to the Geological Society (1891) for the latest view of this subject.
SPECIAL VOLCANIC AND SEISMIC PHENOMENA.
THE ERUPTION OF KRAKATOA IN 1883.
I propose to introduce here some account of one of the most terrible outbursts of volcanic action that have taken place in modern times; namely, the eruption of the volcano of Krakatoa (a corruption of Rakata) in the strait of Sunda, between the islands of Sumatra and Java, in the year 1883. The Malay Archipelago, of which this island once formed a member, is a region where volcanic action is constant, and where the outbursts are exceptionally violent. With the great island of Borneo as a solid, non-volcanic central core, a line of volcanic islands extends from Chedooba off the coast of Pegu through Sumatra, Java, Sumbawa, Flores, and, reaching the Moluccas, stretches northwards through the Philippines into Japan and Kamtschatka. This is probably the most active volcanic belt in the world, and the recent terrible earthquake and eruption in Japan (November, 1891) gives proof that the volcanic forces are as powerful and destructive as ever.