PHYSICALGEOGRAPHY

Physical Geography. A TREATISE on physical geography should comprehend a description of the general structure of our planet, of the arrangement of the solid and liquid materials of which its surface consists, of the nature of the elastic invisible fluids with which it is surrounded, and an account of the distribution of the organized beings by which it is inhabited. A detailed description of a field so vast, of a range of subjects so multifarious, would embrace every branch of natural history; but the province of physical geography is understood to be limited to general views of the different subjects now indicated. These views, if founded on legitimate deductions from observed phenomena, should enable us to reduce to a few general principles the infinite variety of appearances presented to us by nature; and it is this generalization which in modern times has elevated the study to the dignity of a science. Physical geography, however, in this philosophical sense, is of very recent origin; much of every domain of terrestrial nature yet remains to be explored; the surface of our earth is still but partially known; and although each succeeding age adds prodigiously to the knowledge of mankind, this exhaustless field will, for ages to come, afford materials to exercise the industry and reward the investigations of the philosophical inquirer, ere physical geography approaches to its limits of perfection.

SECT. I.—Position of the Earth in the System of the Universe; its Form, Magnitude, Density.

The sublime conceptions of astronomy, founded on observation and on mathematical inductions, have taught us to consider the innumerable stars visible to the naked eye and indicated by the telescope, as so many suns, probably enlightening and gladdening unseen worlds, which revolve round each, as our earth and its kindred planets revolve round our glorious luminary. The stars nearest to our earth are grouped into constellations, for the convenience of descriptive astronomy; but the more distant appear to the eye as nebulous patches or streaks of diluted light, which the highest powers of optical instruments have barely enabled us to resolve partially into clusters of stars, which analogy would lead us to regard as myriads of other suns; while imagination, ranging through illimitable space, pictures still more remote orbs, whose light has for ages travelled the profound, without as yet reaching the abodes of men. Such lofty conceptions of infinite power and beneficence are calculated to elevate the mind which entertains them, and to afford a compensation to human vanity for the comparative insignificance of our earth in the scale of creation. The solar system, of which our planet forms only a small part, seems to be itself but a fragment of a more extended system, round the centre of which the observations of Maskelyne and Herschel have rendered it probable that the sun and the fixed stars all revolve.

The terraqueous globe, without including comets, is one of twenty-nine bodies wheeling round our sun. Of these, eleven are termed primary planets, and revolve round the sun as a common centre; the remaining eighteen are named moons, satellites, or secondary planets, because they also revolve round some one of the primaries. The earth is accompanied in her orbit by one of these satellites. The orbits of two other planets are included within her path round the sun. The orbit of the earth is elliptical, but differs so little from a circle, that, in estimating its diameter, it may be considered as such; and its semidiameter, or the earth's distance from the sun, as determined from various as-

tronomical data, especially from the transits of Venus over the sun's disc, equals ninety-five millions of miles.

That the earth is globular, is proved by ships steering in a general direction, either westward or eastward, arriving again at the point of departure; and very satisfactorily by the phenomena exhibited by vessels in receding from an island in the ocean. At a short distance the hull of the ship is sunk below the horizon; a little farther and the courses or lower sails are lost; and soon only the upper sails are described on the circular horizon. The rotundity of the earth is such, that when the eye is elevated six feet above the water, an object on its surface may be seen at a distance of three miles; and at all other altitudes of the eye, the distance of the utmost visible point varies as the square root of the altitude.

The earth, however, is not a perfect sphere, but is somewhat flattened at the poles, or forms what is termed an oblate spheroid. Newton attempted to ascertain the ellipticity of the earth by calculation; and, considering the effect of its diurnal velocity on a homogeneous fluid of the mean density of the earth, he determined its equatorial to be to its polar axis as 236 to 230; but he imagined, that if the density were greater at the centre than at the surface, the compression would be greater than in a spheroid of equal density. This error was pointed out by Huygens, who, on the supposition that the density increases regularly with the depth in the earth, estimated the relative axes as 578 to 579. But the determination of both philosophers is at variance with the results of geodesical experiments made about that time, and with the more refined measurements of meridional degrees in different parts of the earth in later times, as well as with the results of observations with the pendulum made by Maskelyne on the mountains of Perthshire, and the very recent investigations on the same subject for Forster, Sabine, Duperrey, Freycinet, Kater, Brisbane, and others.

On comparing the two methods of determining the figure of the earth, we are inclined to place more confidence in the result of measurements of meridional arcs than on the pendulum experiments, because the latter are liable to be disturbed by the geological structure of the regions in which they are made; and this is probably the cause why they accord less with each other than we should have expected from the delicacy of the apparatus, and the expertness of the observers. They give a greater ellipticity to the earth than the comparison of meridional arcs. The latter are not only more accordant with each other, but agree very nearly with the form deduced from the inequalities in the moon's motions, the equations for which were discovered by Laplace. The result of the whole is, that we may safely assume the ellipticity as equal to $\backslash frac\{1\}\{305\}$th of the earth's axis; or the polar is to the equatorial diameter as 305 to 306.

The solution of this problem is essential to the determination of the exact length of the radius of the earth; and is a most important problem in astronomy, as it is the unit by which we measure the planetary distances. But the determination of meridional arcs affords the solution of the point in question; and, aided by calculations, we obtain the following results in English miles:

The oblate figure of the earth appears to be a consequence of its rotation on its axis. Mathematicians have de-

monstrated, that if the earth had been a homogeneous body, its ellipticity would have been greater than it really is; and if it had been so constituted that its density increased so as to be infinitely great at the centre, the ellipticity must have been much less. But the ellipticity is intermediate between these extremes, and perhaps coincides with that of a spheroid of equilibrium. The effect of centrifugal motion has evidently affected not only the liquid, but the solid materials of the globe; and this has been adduced by Playfair as a striking proof of the former fluidity of the solid materials of the earth.

The revolution of the earth on its axis is performed in a natural day, or, more strictly speaking, once in $23^h\backslash \; 56\text{'}\backslash \; 4\text{'}\text{'}$; and as its mean circumference is $=\; 24871$ miles, it follows that any point in its equatorial surface has a rotatory motion of more than 1000 miles per hour. This velocity must gradually diminish to nothing at either pole. Whilst the earth is thus revolving on its axis, it has a progressive motion in its orbit. If we take the length of the earth's orbit at 600,000,000, its motion through space must exceed 68,490 miles in the hour.

Besides these two motions, the refinements of modern astronomy have determined two secular movements of extreme slowness, produced by the attractions of the other heavenly bodies of our system, namely, the motions which cause the precession of the equinoxes, and the nutation of the earth's axis. These motions, though important astronomical speculations, are little connected with geography; but it is otherwise with the diurnal and annual movements of our planet, the former of which produces the grateful vicissitudes of day and night, the latter the changes of the seasons. Both exercise remarkable influences upon the organized beings inhabiting the earth; and their consequences, therefore, deserve notice in a treatise of physical geography.

Various astronomical phenomena indicate that the density of the earth increases towards the centre. Cavendish endeavoured to estimate the mean density of the earth from the force of gravitation; and, from experiments with spheres of lead, he estimated it at 5.48. The investigations with the pendulum by Maskelyne and more recent observers have given it as equal to 4.71; and perhaps we shall not greatly err, if in round numbers we consider the mean density of the earth as equal to five times that of water. But the average density of the rocky masses which compose the crust of the earth does not exceed 2.5, or 2.8 at the utmost; and therefore the density of the interior parts must be considerably greater than that of the materials on the surface. Laplace considers it as probable that the density increases progressively with the depth, and that the globe may be composed of concentric beds symmetrically arranged round the centre of gravity. From these determinations, some philosophers have conjectured that the interior of the earth must consist of matter as ponderous as the metals, and that probably the metallic veins which we find near the surface may only be small portions of the central masses, expanded and projected through the concentric stony materials by the internal heat of the earth. If, however, we admit that the density of all matter is in the direct ratio of the pressure which it sustains, it has been shown, by a philosopher not less remarkable for the originality and boldness of his theoretical speculations, than for the profundity and sagacity of his philosophical investigations, that the interior of the earth cannot be a solid mass of any material with which we are acquainted, but "that our planet must have a very widely cavernous structure, and that we tread on a crust or shell whose thickness bears but a very small proportion to the diameter of its sphere;"¹ which, since an absolute void is inadmissible, he conjectures to be filled with elemental heat,

or light in its most concentrated state, shining with the most intense resplendence.

SECT. II.—General Distribution of the Land and Water.

On glancing at a map of the world, we are immediately struck with the great extent of its superficies occupied by the waters, which are united into one great ocean, filling up the depressions which separate the more elevated portions of the solid crust of the globe, and dividing the land into detached masses of very various extent, which we denominate continents, islands, or insulated rocks. These divisions, however, differ only in size; the continent is but a larger island; and the insulated rock is merely the summit of a mountain, the flanks and base of which are buried below the general level of the sea.

We are in the habit of speaking of the four quarters of the world, and sometimes of its four continents; but it is sufficiently obvious that there are only two great distinct divisions or principal masses of the dry land to which the name of continent should be applied; the eastern, including Europe, Asia, and Africa, which, from being the earliest abodes of civilization, may be termed the Old Continent; and the western, or American, which is usually known as the New Continent. By confining the designation of continent to these prime divisions of the land, we shall find no difficulty in classing Australia with islands. The old continent includes three of the great divisions of the earth, Europe, Asia, and Africa; the western continent, though chiefly occupied by the descendants of Europeans, is naturally divided into two great portions, which in future ages may become as distinctly marked by political and moral differences as they are by geographical peculiarities.

The vast extent of Australia has induced some geographers to consider it as a continent; but it may more justly be reckoned the first of islands in magnitude; and it might be grouped with the numerous and extensive islands lying to the south-east of the old continent, as a fifth great division of the globe, under the title of the Oceanic or Polynesian.

The terms employed in describing the various forms of the land, and the subdivisions of the liquid surface of our planet, have been sufficiently noticed under the article GEOGRAPHY, to which the reader is referred.

The slightest glance at a map of the world will show, that there exists a greater accumulation of dry land in the northern than in the southern hemisphere. Recent voyages have indeed proved that this disproportion was over-rated by the geographers of the last age. The late discoveries of the extensive tracts of land to which the names of New Shetland, New Orkney, and Sandwich Land have been given, prove that considerable deductions must be made from what was at one time considered as the domain of the Antarctic Ocean. But the voyages of Cook, Furneaux, Weddell, and others, have shown that these, although extensive groups of islands, do not belong to a great antarctic continent; and even if such existed, the general form of the ancient, as well as of the western continent, towards the north, would still leave a great preponderance of dry land in the northern hemisphere. It is not unworthy of notice, that the proximate sides of the old and new continent seem as if they had mutually influenced the forms of each other; for where any remarkable projection of the one is perceived, we find an indentation nearly opposite in the other. Thus, the principal eastern projection of South America is almost opposite to the vast African bight, extending from Cape Palmas to Cape Negro; the great western projection of Africa is opposed to the basin of the Caribbean Sea and

¹ Leslie's Elements of Natural Philosophy, note I.

Physical Geography. the Gulf of Mexico; and the north-eastern trending of the coasts of Europe is balanced by the opposite direction of the shores of North America.

Land. The general disposition of the land is different in each. In the old continent the principal extension is to the west and east; in the new it is to the north and south. In the former we may draw continuous straight lines of immense extent in different directions without encountering one sea. Thus a line from Cape de Verd in Western Africa will pass through continuous land till it reaches the promontory which separates the Yellow Sea from the Gulf of Leatong, a distance of 7800 geographical miles. One projected from Cape Palmas, on the Gulf of Guinea, to the Asiatic promontory called Tschoutskoi-nos, will extend 8700; and another, from Tangier to the Cape of Good Hope, will be nearly 5800 geographical miles. The longest straight line which we can project on the land of the new continent is from Cumana to Tierra del Fuego, a distance of 4600 geographical miles; and the next, from Icy Cape to Puerto de los Angeles, on the western coast of Mexico, is about 4100. In the old continent the lines we have described have large portions of the surface on either hand; but in the new continent the greater part of the land is on the eastern side of the imaginary lines.

The greatest extension of each considerable division of the two continents lies in the direction of its principal mountain chains. This is strikingly exemplified in America, where the Rocky Mountains and Western Cordillera of Mexico in the northern, and the gigantic Andes in the southern hemisphere, pursue the general direction of the American continent. The greatest extent of Europe is from west to east, and this is the general direction of the chains of the Pyrenees, the Alps, the mountains which form the northern boundaries of Turkey in Europe, and the Carpathians. A similar direction is pursued by Mount Taurus, and other ranges of Central Asia, the Himalaya chain, and the mountains of Thibet. In Africa the position of the mountain chains is less known; but the direction of the rivers, and the notices of travellers, would lead to the inference, that, on its eastern side, a spine of mountains, more or less interrupted, extends from Abyssinia towards the southern extremity of Africa. Even the direction of smaller portions of both worlds, as of islands and peninsulas, coincides with their principal mountain chains. This is well illustrated in Norway, in Italy, in the so-called peninsula of India, in Siam, Cambodia, Florida, and California, and on a smaller scale in the islands of Ceylon, Madagascar, Cuba, and Great Britain.

Ocean. The expanse of the waters is termed the Ocean; but, for the convenience of descriptive geography, it is divided into several, though, strictly speaking, there is but one ocean; and the different seas form but parts of a vast whole, which communicate freely with each other, and can scarcely be considered as marked by definite limits. Most geographers speak of two grand oceans, which have derived their appellations from their position relative to the old continent.

I. The Western Ocean, which includes three principal subdivisions. 1. The Northern Ocean, of which the southern limits are the British coasts, and a line drawn thence to the southern extremity of Greenland. 2. The Atlantic Ocean, which should be limited to the south by the extreme projections of Guinea and Brazil. 3. The Ethiopic Ocean, which lies between the limits of the last, and a line joining the southern extremities of Africa and America. All three are sometimes, though less correctly, denominated the Atlantic.

II. The Eastern Ocean, which is also subdivided into three principal portions. 1. The Indian Ocean, comprehending the seas between Southern Africa, Australia, and the Malayan Islands. 2. The Pacific Ocean, extending from Behring's Straits to Cape Horn, and the southern side of Australia. It is often subdivided into the Northern and Southern Pacific, of which the limits are not well defined.

3. The Austral or Southern Ocean, which in general terms may be described as extending southwards from a line passing round the globe and touching Van Diemen's Land, and the Cape of Good Hope, and Cape Horn. With these main oceans various seas communicate, either by narrow straits, as the Persian Gulf, the Red Sea, the Mediterranean, and the Baltic; or by wider channels, as the Sea of Okhotsk, the Yellow Sea, the Sea of Hudson, and of Baffin; or by numerous openings, as the Sea of China, the Caribbean Sea, and the Mexican Gulf. The Euxine, the Caspian, the Sea of Aral, and of Baikal, and the magnificent fresh-water lakes of North America, have many of the characteristics of lakes; but their gigantic scale, and their important influence on the climate of the adjacent countries, entitle them to the designation of Inland Seas.

The portions of the earth's surface which are covered by the waters bear a large proportion to the dry land. Dr. Halley first suggested the idea of ascertaining the comparative surface of the land and water, by cutting them out of a map, and weighing them separately. It was by this means that he ascertained the relative area of the English counties. The astronomy of Dr. Long, published in 1742, contains the application of this principle for ascertaining the relative surface of the land and water on the earth. He describes his having cut out each from the gores of Senex's sixteen-inch globes, and on carefully weighing them he obtained the ratio of the water to the land as 349 to 124, or very nearly as 3 to 1. Dr. Long's experiment appears to have been conducted with great nicety; for Professor Rigaud of Cambridge has very lately confirmed it by a laborious and careful repetition upon the great scale. This latter gentleman used the gores printed for Addison's three-feet globes, which contain a superficies of 4071½ square inches. The gores of this magnificent globe are twenty-four in number. The torrid zone was divided at these lines and at the equator, and thus formed into forty-eight divisions; the divisions of the two temperate zones were also forty-eight, and these together made ninety-six divisions, each of which was carefully examined separately, divided into land and water, and accurately weighed in a balance which easily indicated one tenth of a grain.

The southern frigid zone was considered as wholly consisting of water; but the northern frigid zone, from longitude 180° to 270° west, was taken as land, and, from 270° to 360°, was computed to consist, one half of land and one half of water. The whole surface of the earth was supposed to be divided into 1000 parts, and all the estimates of the different regions were expressed in proportions of this standard. Rigaud's paper contains several tables, of which the following are the general results.

This differs surprisingly little from the estimate of Dr. Long, considering that each was deduced from the works of different geographers, and at a considerable interval of time. If we express the ratios of each determination, we shall have, in

Dr. Long's estimate, the land to the water..... = 1 : 281
Professor Rigaud's..... = 1 : 276
And it is also deserving of notice that Rigaud's last esti-

mate differs but little from his former experiment on Carey's twenty-one-inch globe; but both differ considerably from the calculations of Malte-Brun. As, however, that author has not furnished us with the elements of his calculation, it ought not to be put in competition with the elaborate investigations of Long and of Rigaud.

The dry superficies of each great division of the globe was ascertained by the same method; and the tables of Rigaud enable us to give an accurate comparative view of their relative area in thousandths of the earth's surface.

265.9233

If we take the whole superficies of the globe = 196,836,658 square miles, as the land is to the water in the proportion nearly of 266 to 734 it follows that the whole land occupies a surface of 52,363,231 square miles, and the ocean an area of 144,473,427 square miles.

SECT. III.—Variations of the Surface.

Besides the inequality produced by the oblate figure of the earth, different portions of its surface are at different distances from its centre. The dry land is elevated into mountains, depressed into valleys, or stretched out into plains, varying in extent and elevation above the sea; and the ocean itself must be considered as effused into a vast basin, seemingly diversified, like the dry land, with mountains, precipices, valleys, and plateaux of an infinite variety of shapes and dimensions. In fact, the whole dry land may be considered as a series of islands, or of plains and mountains, elevated above the general surface of the ambient fluid which occupies the deeper valleys that divide them.

The whole surface of the dry land is elevated more or less above the general level of the ocean; with one remarkable exception, which, however, has only of late years been detected. Barometrical measurements have shown that a vast area of Central Asia, no less than 18,000 square leagues, is considerably below the level of the ocean. This tract includes the Caspian and the Aral, the surfaces of which have recently been shown to be 101.2 feet lower than that of the Euxine. Should any convulsion of nature, such as earthquakes are known to produce in other quarters of the earth, depress the low sandy tract which now separates the sea of Azof and the Caspian, the waters of the Euxine, and also of the Mediterranean and the Atlantic, would inundate an enormous extent of the sandy steppes of Asia, and entirely change the climate and face of that portion of the earth.

Plains of greater or less extent, with inconsiderable inequalities, are found in every region. The most considerable plain in Europe is that which, commencing on the Northern Ocean, between the extremity of Jutland and the mouths of the Scheldt, extends eastwards through Prussia and Poland into Russia, and constitutes in that empire an almost unbroken plain, from the confines of the Frozen Ocean to the shores of the Euxine and the Caspian, until it meets the Uralian chain, which separates Northern Europe from Northern Asia. In this vast expanse, the surface is scarcely broken by any hills, except the inconsiderable ridge of Valdai, between Toropetz, Smolensk, and Moscow, which nowhere rises to 1200 feet above the sea. This magnificent plateau is divided by the Uralian Mountains from a still more extensive plain, which may be denominated the great plain of Northern Asia, the boundaries of which are

the Altaian chain and the Icy Ocean. It commences at the eastern side of the Ural, and inclines towards the north-east until it almost touches the Northern Pacific Ocean. Both plains together form a remarkable feature in the surface of our earth, and being only interrupted by the Ural chain, which never exceeds 4500 feet in height, have been regarded by some geographers as but one plateau; which, thus considered, is the most extensive on our planet, with a medium breadth of 1100 miles, a length of 6000, and a surface of about 6,600,000 square miles. The surface of much of these two great plains is fertile; some of it is covered with heaths or swamps, producing dwarf birches, willows, and a stunted vegetation; a considerable part is shrouded in dark pine forests.

Far different in appearance are the enormous wastes of arid sand in Northern Africa and Central Asia, the vast savannas of the Mississippi and the Missouri, and the pampas and llanos of South America.

A prodigious zone of sand stretches like an ocean, in Northern Africa, between the ridges of Mount Atlas and the parallels of the Senegal and the Niger, and from the Atlantic to the narrow valley of the Nile. It may be said to cover a tract across that continent between the eighteenth and thirty-first degrees of north latitude, extending from long. 15° W. to 30° E. This enormous area may be considered as having a length of 2470 miles, a breadth of 900 miles, and consequently a superficies of about 2,200,000 square miles. It is, however, by no means an uniform surface of loose sand. In many parts the dreary waste is broken by low hills of naked sandstone, or by tracts of arid clay; occasionally it is enlivened in a few spots by verdant isles, or oases, of various extent, according as water finds its way to the parched surface. The principalities of Fezzan and Darfur are the most considerable of these islands of the desert; but the most celebrated of these insulated spots is the far less considerable oasis of Siwah, the palms of which still wave over the ruins of the Temple of Ammon, and indicate the most ancient seat of African civilization.

Egypt, Nubia, and Senaar, the eastern boundaries of this desert, would long ago have been overwhelmed by the moving sands of the western desert, but for the annual inundation of the majestic Nile; and thus the empire of sand would have stretched uninterrupted into Arabia. As it is, the narrow valley of that river, and the Red Sea, are all that separate the African deserts from those of Asia. Soon after quitting the Nile, the traveller by the route of Suez encounters sand, which is continued into the centre of Arabia, where it forms the Desert of Nedjed, extending to the valley of the Euphrates. The sandy zone then inclines northwards, enters Persia, and forms the saline deserts of Adjemi, Kerman, and Mekran; it is turned north-east by the valley of the Indus, passes through Caubul and Little Bukharia, till it joins the vast deserts of Cobi and Shamo, which occupy so large a portion of Central Asia between the Altaian and Mustag chains, and reach to the confines of China. The sandy zone thus traced throughout the breadth of the ancient continent, from Western Africa to the 120° of east longitude, has been computed to cover an area of 6,500,000 square miles; but the Asiatic portion of this tract includes many chains of mountains and fertile valleys. It is characterized by the occurrence of arid wastes of sand or clay, sometimes with saline incrustations on the surface, and is remarkably deficient in considerable rivers; except the Nile, the Euphrates, the Indus, and the Oxus, there are no large rivers in a region which embraces almost a fourth part of both Africa and Asia. This portion of Central Asia forms a series of elevated plains six thousand miles in length from west to east. "Some of these plains," says Humboldt, "are covered with herbage; others produce only evergreen saliferous plants, with fleshy and jointed stems; but a great

Physical Geography. number glitter from afar with a saline efflorescence, that crystallizes in the semblance of lichens, and covers the clayey soil with scattered patches like new-fallen snow."

From these elevated plains have issued the barbarous hordes which at different periods have swept like a whirlwind over the regions where the human intellect had achieved its proudest triumphs, and devastated the seats of Asiatic and European civilization. From this officina gentium have successively issued the Huns, the Avars, the Mongols, the Alans, and Uzes—the scourges of humanity; and possibly also the tribes by whom the American continent was originally peopled.

North American plains. The great valley of the Mississippi is a magnificent plain, covered with primeval forests, or spreading in vast savannas; but a great portion of it is liable to periodical inundations from the river. At 100 miles from the sea its annual inundation has a breadth of from 80 to 100 miles; and even at the mouth of the Ohio, which is 1000 miles from the sea, the inundating waters have the breadth of thirty miles. This shows the extreme flatness of the soil. The river has been extensively embanked, to rescue a portion of the rich soil from these inundations; but, during the season of the inundations, the greatest part of this extensive plain is still abandoned to the alligator and other reptiles, and to enormous herds of wild animals during the rest of the year.

The banks of the Lower Missouri, for 2500 miles, present a succession of fertile plains beyond the reach of inundations. This district, which may be considered as a continuation of the great valley of the Mississippi, exhibits an agreeable intermixture of natural prairies and forests, abounding in wild animals, which retreat before the rapidly-advancing tide of human colonization.

These immense plains are probably destined to become the abodes of a dense population, and perhaps the seat of a mighty empire, subject to few of the natural vicissitudes that impede the march of improvement, if we except the contingent risk of some mighty convulsion, by which the waters of the extensive fresh-water seas of Canada may be permitted to overwhelm this extensive realm with a far more formidable catastrophe than the deluge of Deucalion. Lake Superior has a depth of about 600 feet, and its surface is more than 300 feet above the level of the plains of the Mississippi. Should earthquakes disrupt the intervening barrier, the waters of that and the other great inland seas would sweep these plains with a devastation quite unparalleled in the history of our planet since the deluge of Noah.

South American plains. The interior of South America presents plains not less extensive, the peculiarities of which have been sketched with a masterly hand by Humboldt. The first may be termed the valley or plain of the Orinoco. It is bounded on the north by the maritime ridge of Caraccas, on the south by the forests of Guiana, and extends westward to the mountain chain of New Granada, the summits of which are lost in perennial snow. It is a solitude of more than 20,000 square leagues.

The second is divided from the first by the mountain-arm which the Andes sends eastward to separate the waters of the Orinoco from those of the Marañon, an interrupted branch, stretching from the towering ridges of Popayan, to the humbler granitic Sierra de Tumucuragua, on the confines of Guiana. The southern boundary of this second plateau consists of the chains of Santa Cruz de la Sierra, of Chiquito, and Cuyaba, which separate it from the plains of the Paraguay, the Parana, and La Plata. The western boundary is the cordillera of the Andes. This territory includes the llanos, extended solitudes, with a scanty soil, destitute of trees, but covered with a luxuriant herbage, the consequence of the periodical inundations, which convert this portion of South America into inland seas; amidst the turbid waters

of which, as has been forcibly described by Humboldt, may be seen herds of wild cattle and of horses, swimming from one gradually-diminishing island to another, in search of food, or escaping from the attacks of the jaguar, to become the prey of the alligator, or perhaps to be stunned by the strokes of the electric gymnopus. In the immense tract of country included in this second division, are some detached groups of mountains, indicated on the magnificent map of Olmedilla; but they are generally of inconsiderable elevation, and the face of the country may be considered as one enormous plain, or succession of contiguous plains, little elevated above the general level.

The third system of great South American plains forms the basins of the Paraguay and La Plata. This region includes the vast pampas of Buenos Ayres, extending from the foot of the eastern ridge of the Andes to the "sea-like Plata," and stretching southward into the cheerless deserts of Patagonia, presenting to the eye interminable plains, with a surface often as unvaried as the ocean, inhabited by vast herds of wild cattle and droves of horses; while the tall form of the Rhea Americana, a species of ostrich, frequently gives local peculiarity to the picture.

Besides the extensive plains already described, which are either not much elevated above the sea, or rise by perceptible degrees, there are other plains, with steep acclivities on every side, which have thence been termed table-lands. Some of these are very extended, and retain a general level of several thousand feet above the sea. Sometimes they are bordered with chains of mountains, much less elevated on the side towards the table-land than in the opposite direction; sometimes they have surfaces much undulated or broken. There are few table-lands in Europe of sufficient importance to be here noticed. The most considerable is that of Central Spain, embracing the two Castilles, which has a general elevation of nearly 2000 feet above the level of the sea. The barometer at Madrid has a mean height of about twenty-seven English inches, which, supposing the instrument on the shores to have a mean of thirty inches, would give to the capital of Spain an elevation of about 1900 English feet. The thermometer in winter often descends below the freezing point; and the chamersops and the agave, which flourish luxuriantly in the plains around the table-land, are unknown at that elevation, though under the fortieth parallel of north latitude. The descent from this table-land on all sides is steep; and the transition from the arid and almost treeless plains of Castille, into the milder climate at the foot of the mountain ridges which support it, is marked by the vegetable productions of a warmer latitude.

Our imperfect knowledge of Central Asia prevents any accurate notion of the extent and form of these remote regions; but there can be little doubt, that in this quarter there is the most extensive table-land on the face of our globe. The severity of the climate, the direction of the rivers flowing from it on every side, and the scanty information derived from Asiatic caravans, all indicate that there is a prodigious central elevation, supported on the north by the chain of the Altai, on the west by that of Belug Tag, on the south by the mountains of Thibet, and on the east by the chain of Sialki, which divides the country of the Mandshurs from their Tatar brethren of Mongolia. This enormous tract, however, can scarcely be described as consisting of one table-land; for it is intersected by the lofty mountain ranges of Alaki and Mus Tag, which are capped with perennial snows, and it contains the vast, treeless, sandy deserts of Cobi and Shamoo.

The central plateau of Asia, from the severity of its climate, is inferred by Humboldt to have an elevation of not less than 9000 or 10,000 feet above the sea; and this region, including Thibet and the desert of Shamoo, is believed to have a superficies of 1,500,000 square miles.

According to the observations of Olivier, Persia should likewise be considered as a table-land of great elevation. Southern India affords an excellent instance of a table-land, which includes the whole kingdom of Mysore. It stretches from the banks of the Kistna on the north, and is supported by the steep acclivities of the Western, and of the less abrupt Eastern Ghauts. The general elevation of this region is about 3000 feet; but at its southern extremity it abuts against a smaller though much loftier table-land, the district of the Neigherry Hills, which, till lately, was almost unknown to Europeans. A good account of it has recently been published at Calcutta, by Dr Robert Baikie, from which we learn that the district occupies an area of 600 or 700 square miles, between the eleventh and twelfth degrees of north latitude, and the seventy-sixth and seventy-seventh degrees of east longitude. The general level of its finely-undulated surface is between six and seven thou-

sand feet above the sea. It is connected with the Mysore country by a neck of land about fifteen miles in width; but its northern flank towers above that kingdom by 3500 feet, and it is separated from the subadjacent plains of Malabar and Coimbetore by an almost precipitous boundary 7000 feet in perpendicular altitude. The surface is fertile, the climate mild and salubrious; and there the invalid, suffering from the baneful influence of a residence in the sultry plains of India, finds a European climate. The barometer varies from 22.840 to 23.220; whilst the thermometer throughout the year ranges from 73° 51 to 42°, and has a mean of 52° 50. The quantity of rain is 42.251 inches; heavy rain falls on nineteen days, showers on eighty days, and the clear days are 179 in the year.

The great similarity of this Indian table-land, in elevation, in salubrity, and in climate, to the more extensive table-land of Mexico in North America, is remarkable.

The Mexican plateau (No. 1) has also a general elevation of 7000 feet above the sea; it rises as abruptly as the Neigherry district, especially from Vera Cruz, and in a single day the European may escape from the pestilential atmosphere of the plains, to an elevation at which the black vomit is unknown. The sea-coasts of Mexico are hot and unhealthy at certain seasons, and are known to the natives by the name of Tierras Calientes, or burning regions, producing the vegetable riches of the torrid zone. A single day, from Vera Cruz, suffices to reach the height of 4000 feet, where there reigns a perpetual spring; and this region, termed Tierras Templadas, though often enveloped in clouds and fogs, is healthy. On ascending to the height of 7000 feet, the traveller gains the Tierras Frías of the Mexicans, or great table-land, which has generally a mean temperature of 63° Fahrenheit. It produces the cerealia of Europe, and the vine and the olive flourish on its plains. But its surface is varied by hills; and along its margins, stupendous peaks, such as Orizaba, Popocateptl, and the Sierra Nevada of Mexico, have their summits far above the limits of perpetual snow. There are other plains in Mexico which are more than 1000 feet above the general level of the great plateau. Such is the plain of Toluca, where the climate is severe, and the mean temperature so low that the olive there refuses to ripen its fruit.

Several less extensive table-lands are included between the two lofty parallel chains which constitute the Southern Andes. Thus the plains of Quito and Riobamba are true table-lands; the general level of the first being 8000 feet above the sea, and the city of Riobamba having an elevation of 10,800 feet, so that it is probably the most elevated city in the world.

Mountain chains and ranges of hills are separated by valleys, which have obtained the names of ravines, dells, defiles, or passes, when narrow and difficult of access. These chiefly occur among lofty mountains, and sometimes present mural precipices of wonderful sublimity. Narrow valleys often present a striking coincidence in the position of the salient and receding angles of their opposite sides. This has been well described, as observed in the Pyrenees, by Ramond; and some geologists have considered it as a proof that such valleys had been formed by the sudden disruption of continuous strata.

into different kinds, according to their respective relation to the mountain ranges on which they occur. Principal valleys are such as run parallel to and separate great chains of mountains. Those which have a direction at a considerable angle with the principal chain, separating the branches or arms of the mountain chain, are termed lateral valleys; and the smaller ones, running into these lateral arms, are denominated subordinate valleys. The lateral open into the principal, and the subordinate into the lateral valleys.

Sometimes we find a valley, consisting of a chain or succession of basin-shaped cavities, at different levels, separated from each other by narrow gorges. Such appearances are not uncommon among all mountains; and several examples occur in our Highland glens, as in Glen Roy, Glen Gluy, and Glen Spean, where the fancied parallel roads mark the successive heights at which the waters formerly stood, until the rocky barriers which pent them up were opened.

Sometimes a valley has a circular form, or is basin-shaped, and on all hands surrounded with mountains, excepting at a narrow gorge, where the waters of the valley make their escape. Two of the most remarkable instances of this are found in Bohemia and in Cashmere. The former consists of a single circular valley, nearly 200 miles in diameter, which appears at one time to have been a vast lake, until the formation of the defile in the Erzgebirge, through which the Elbe escapes into Saxony. Cashmere is a fertile valley of difficult approach, nearly circular, and ninety miles in diameter, embraced by the arms of the Himalaya Mountains. Native tradition represents it as a lake, until one of the Hindu deities burst the rocky boundary, by the formation of the ravine through which the Vidusta now pours its waters.

Some valleys contain large rivers and lakes, from which there is no issue or outlet. Such valleys have not yet been drained by the giving way of some point of their boundaries. Central Asia affords many instances of this species of valley, as in the districts around the Koko-noi, the Loke-nor, the lakes of Zaizan, Kirgha, and Palkati, and the Sea of Aral. A very remarkable instance occurs amidst the loftiest peaks of the Andes, in the valley which contains the great lake of Titicaca, which the ancient Peruvians regarded as sacred, the residence of the founder of their mo-

narchy, and which is sufficiently remarkable, by its extent, and the sublimity of the surrounding scenery.

The most elevated portions of the earth's surface are either connected in extensive mountain chains, or they exist in aggregated groups branching from a common centre, or they form insulated heights, rising more or less abruptly from plains. We have seen that the direction of mountain chains is that of the principal extension of the regions in which they occur; and the same thing holds good in peninsulas and islands, in which we generally find the chains longitudinal. Thus the principal mountains of Europe, Asia, and America form extensive chains in the direction of the greatest extent of each of these divisions of the earth.

The direction of peninsular chains is well seen in Norway, in the Apennines, in the mountains of the western side of Arabia, in the Ghauts of India, and in the mountains of Corea, Kamtschatka, and California. The direction of insular chains, as corresponding with the principal dimension of each island, is also illustrated in Madagascar, Ceylon, Japan, Cuba, St Domingo, the British Isles, Sardinia, Sicily, and in many of the smaller islands.

If we endeavour to generalize the observations on the direction of mountain chains, one very remarkable peculiarity seems deducible, namely, that they very generally present their steepest acclivities towards the great basins to which they are contiguous, while they slope more gradually in the opposite direction; and on examining their intimate structure, we find their stratified beds dipping generally from the basins to which their escarpments are presented. Thus the ridges of the Scandinavian peninsula present their boldest escarpments to the basin of the Northern Atlantic; while the opposite ranges of Greenland and Iceland also show their steepest acclivities towards that basin. Round the shores of the Mediterranean the same arrangements are especially observable. The lofty ridge of the Atlas is very bold on the northern side, and declines more gradually towards the Sahara. The chain of Spanish mountains which skirt the Mediterranean, from Gibraltar to the Pyrenees, present their escarpments towards that sea; and the Maritime Alps of France, the mountains of Switzerland and the Tyrol, of Istria, Dalmatia, and Thrace, all present their most precipitous sides to the basin of the Mediterranean. We believe that the considerable chain which in Asia Minor extends from Mount Ida to the country around Scanderoon, has the escarpment directed towards the Levant seas; while the mountains of Armenia present their boldest declivities to the basin of the Euxine, and the Caucasus and mountains of Mazanderan towards that of the Caspian. The escarpment of the great chain of Africa, termed that of Lupata, which seems to be prolonged from the lofty mountains of Abyssinia to the south of the Mozambique Channel, would appear, from the little we have learned of its structure, to face the basin of the Indian Ocean; and we know that the steepest declivities of the Western Ghauts of India are directed towards the same basin. The mighty spine of the American continent, from the shores of the Arctic Frozen Ocean to the extremity of South America, through a course of more than 8600 British miles, presents a series of rugged precipices to the vast basin of the Pacific; and, if we might indulge in one sweeping generalization, it would seem that the chains stretching from the Persian Gulf eastward through Thibet, and thence bending to the north-east through Mongolia and Northern China to Tscheoutskoinos on the Frozen Ocean, present their fronts also to the Pacific Ocean.

As such coincidences can scarcely be considered as accidental, they afford a wide field for speculation. Can we suppose that appearances, on such an immense scale, have any relation to the operation of the force which caused the

elevation of the land, acting towards a central point, and producing the dip of the elevated strata all around, from a common axis of movement? Such speculations, aided by the position of volcanoes, and other mountains of igneous formation, might lead us to infer the direction of the great lines of subterranean disturbances which have modified the appearance of the crust of the earth.

The inequalities of the earth's surface, however considerable they may appear to the eye that contemplates them, bear but a trifling proportion to a sphere of nearly 8000 miles in diameter. The loftiest peaks are not quite five miles in height above the sea; and we may obtain some idea of their effects on its general form, by considering, that such inequalities on the surface of our earth produce no greater proportional deviation from its spherical shape than a roughness equal to $\backslash frac\{1\}\{18000\}$th of an inch would on a ball an inch in diameter; not so considerable as the inequalities produced upon the form of an orange by the papillae of its rind. The heights of mountains, however, exercise an important influence on the climate and productions of the regions where they occur; and it becomes an important point in physical geography to determine their respective altitudes. This has usually been attempted by two methods; by geodesical mensuration, and by marking the difference of the barometrical columns on their summits and at their bases. The first is the most accurate when we can obtain a level base at a moderate distance from the height to be ascertained, and when the measurements of the angles are carefully taken by expert operators, and with good trigonometrical instruments. The use of the barometer is confined to accessible heights; but it is much more easily managed, under such circumstances, and requires little more than attention to the true height of the barometrical column, corrected for the temperature of the instrument and of the ambient air. In variable climates, accuracy would require synchronous observations at the foot and on the summit of the height to be ascertained; but in tropical climates, or even in the south of Europe, the barometer scarcely varies throughout the year at the level of the sea, and therefore such measurements are more easily made. The methods of both kinds of observation are sufficiently detailed in the articles BAROMETRICAL MEASUREMENTS AND TRIGONOMETRY.

Accessible heights have also been estimated by the fall of the thermometer on ascending. It has been found that, in our latitudes, an ascent of every 300 feet may be considered as equivalent to a fall of 1° of Fahrenheit's thermometer; but this formula does not answer so well either in high or in low latitudes.

It has also been proposed to measure such elevations by marking the temperature at which water boils at the base and on the summit. A convenient instrument for such observations was invented by the Rev. F. Wollaston, but it is liable to accidents in carriage; and, on the whole, geologists trust either to barometrical or to trigonometrical observations for measuring mountain elevations.

The most lofty mountains in general form parts of extensive chains; but there are many instances of great elevations being attained by isolated peaks. Instances of the former are familiar in the Alps, the Carpathians, the Himalaya, and the Andes; of the latter, in the towering summits of Ætna, Teneriffe, and Mouna Roa in the Sandwich Islands. But such detached elevations are either active or extinct volcanoes.

In the following table we have arranged the heights of mountains, according to the regions in which they occur; and marked the method of determination by the letters B. and T. according as the determination has been made by the barometer or by trigonometrical operations, while E. signifies estimated height.

From this table it appears that the loftiest pinnacles of our earth are to be found in the Himalaya range. Some of them attain an elevation a mile higher than Chimborazo, which was long regarded as the loftiest mountain on the globe. Of late years our countryman Mr Pentland has measured several mountains in Upper Peru, especially near the Lake of Titicaca, much higher than Chimborazo; and Sorata, one of these, is only surpassed by a few peaks in the great Indian chain. Although those who first attempted to measure the summits of the Himalaya appear to have exaggerated their real altitude, succeeding observers have shown that we must consider that range as the most elevated on the globe, and the peak of Dhawala-giri as the sovereign of mountains.

Human habitations, and even considerable towns, have been erected in very elevated stations on mountain chains. Dr Gerard states, that in the valley of Sulei, in the Himalaya Mountains, there is a village at the enormous elevation of 14,700 feet above the sea, and that rye is cultivated as high as 14,900 feet. Several travellers have found

the revered temple of Kedar-Nath to be elevated not less than 12,000 feet. America presents some extraordinary examples of the same kind.

In Europe we have the following.

SECT. IV.—General Arrangement of the Solid Materials of the Globe.

A particular description of the solid materials of our earth has been given under the articles GEOLOGY and MINERALOGY; but it is our business to present here a brief statement of their general distribution.

Persons unaccustomed to examine the structure of the crust of our earth are apt to imagine that mineral masses present but a confused congeries of stony matter, without any order or particular arrangement. Nothing, however, can be more erroneous. Wherever water-courses, artificial excavations, or abrupt precipices, expose the structure of the crust of the earth, we find very striking appearances of the agency of causes which must have acted with great uniformity over vast portions of its surface, and which have produced a general resemblance between the structure of countries widely distant from each other. A large portion of the solid materials is arranged in beds, varying in extent and thickness, but everywhere indicating the operation of one common agent. These beds are sometimes slightly coherent, as when composed of clay or sand; at other times they are consolidated stony bodies, arranged in parallel layers, which are often subdivided into thinner portions by seams or joints, and preserve their parallelism for a great extent, whether their position be horizontal, or at different degrees of inclination. Such beds are termed strata, and the general fact is expressed by the term stratification. Sometimes we find a succession of strata of the same rock in juxtaposition; at other times there are strata of different substances interposed or alternating with the principal rock. The position of strata in plains is generally but little inclined towards the horizon; but as we ascend mountains, the strata are usually inclined, sometimes nearly, or even absolutely vertical. When a mountain, or a chain of mountains, is lofty, it is not unusual to find the strata inclined from the centre or axis of the chain towards either hand (No. 2);

and in such cases we generally find the strata reposing on a different kind of rock, which has no appearance of having a seamed or stratified structure. Such rocks may also occur in plains; they have frequently a granular or crystalline structure, and form an important class of stony bodies. Hence the distribution of rocks into two great divisions, the stratified and unstratified. Both divisions present instances of rocks which either appear simple or homogeneous to the eye, or consist of stony bodies, which are evidently composed of two or more materials. Instances of the one

occur in clay-slate, limestone, quartz-rock, and basalt; of the other in mica-slate, gneiss, granite, and porphyry. The visible axis of a mountain chain is very often some species of unstratified rock, and stratified materials usually form its flanks. In other instances no unstratified material is visible at the axis, as far as we can trace the structure of a mountain. But whenever the mountain is much elevated, or

steep, and its sides ploughed by torrents, it is very rare not to find some unstratified material as the nucleus upon which the inclined strata rest (No. 3, g). We are not, however, to assume, that in the order of superposition the stratified materials are invariably above the unstratified. Though this holds good in a great number of cases, we frequently observe vast masses of unstratified rock directly resting upon regular strata (No. 3, t). In some cases we are able to show such irregular masses connected with other similar materials much lower in the series of rocks, by one or more lines of communication penetrating the strata, which, according to their shape and size, have obtained the name of veins or dykes. Instances of this occur frequently in mountains into the composition of which trap-rocks enter (No. 4).

There are, however, other kinds of veins which we cannot so readily trace to larger masses of mineral substances. The most familiar instances of such occur in metallic veins (No. 5). They may often be observed penetrating both stratified and unstratified matter, in the greatest diversity of forms and directions; sometimes in very minute threads, expanding to the dimensions of many feet, or even yards, and repeatedly contracting again to the thickness of a fraction of an inch.

Sometimes mineral bodies form irregular masses entirely enclosed in the rocks in which they occur, varying in weight from a few grains to many tons, without any perceptible trace of the mode by which they were so introduced. Occasionally long fissures occur, which are not filled up by any solid material; sometimes also cavities are left in the more solid species of rocks, the sides of which are lined with crystalline bodies of various kinds. Sometimes these fissures and cavities are of enormous extent, forming vast caverns in the bowels of the earth, which are either empty or filled with water. Instances of these on the great scale oc-

Physical Geography. cur in the limestone strata of many countries, especially of Carniola and North America; and on a less scale in France, Germany, and in our own island. The cavern of Adelsberg, in Carniola, extending to three leagues in its various ramifications, and containing a subterranean lake and river, affords an example of the extent of caverns in limestone rocks; and the no less remarkable Mammoth Cave of Kentucky, which has been traced to the extent of nine or ten miles, is an American example of the same formation. Almost every extensive cavern of Europe is also in limestone, as the interesting ossiferous caves of Gailenreuth and Künloch, in Franconia; of Bize, Somières, and Racogne, in France; the Kirkdale Cave, and the caverns of the Craven and Mendip Hills, in England; but the fine stalactitic cave of the Isle of Skye is in sandstone. We do not recollect any considerable caverns found in other rocks, excepting one in the schist of the island of Thermie in the Archipelago, and the vast Icelandic cavern of Surtshellir, in a current of lava which had flowed from the Bald Yökul. This last cavern, as described by Henderson, is nearly a mile in length, and appears to have been formed in the flowing lava by the disengagement of gaseous matters; as it has all the appearance of being the consequence of an enormous air-bubble, having its arched roof ornamented with coralliform projections of black slag. Perhaps we may also give as an exception the small but remarkable columnar cave at Staffa, which is in trap-rock.

Such is the general structure of the solid crust of our earth. But when we examine the rocks more minutely, we find other striking peculiarities belonging to each, which, besides the order of their superposition, mark different eras in their formation.

The regularity of the lines dividing stratified rocks early suggested the idea of their deposition from water; and this was rendered still more apparent when observers began to remark that the remains of organic beings were often included in them. Plants, shells, and corallines were first noticed as entering into the composition of rocks; and next the remains of fishes, and of other vertebrate animals, were distinctly recognized. Some organic remains were observed to pervade limestone rocks at great depths in the earth, and upon very high mountains; and remains of plants, and even stems of trees, were found fossilized, or converted into stone, in the strata of the coal formation. Some remains of marine genera of animals were detected in slaty strata which do not belong to either of these two kinds of rocks; and, lastly, the remains of quadrupeds were recognised in the sandy marl and gypseous deposits which in some countries lie immediately below the soil, or were imbedded in sedimentary matter in the floors of caverns in various parts of the world. In most instances there is an alternation of marine and fresh-water organic remains; and the general fact appears to be, that the fresh-water shells are most numerous in the upper beds, and those of marine origin in the subordinate strata of the low country, while the former seldom attain any considerable height on mountains; but the marine animal remains are found on the summits of ridges as lofty as the Pyrenees.

Minute examination of the organic remains collected from different strata have established the fact, that comparatively but few of the fossil animals are of the same species with those of the same families now living; and we cannot doubt that whole genera and orders have disappeared from the face of the earth. It is only in the upper deposits that any species identical with those now existing have been detected. Where the organic bodies are completely fossilized, very few if any living species can be recognised; and generally the recent species most nearly allied to the fossils are now no longer to be found in the adjacent seas or regions. Thus the fossil-shells and corallines of northern countries, in both continents, have their congeners (when

such are known) generally in tropical climates; and the fossil-trees of the British coal-fields have a greater affinity to the tree-ferns and cyacadea of southern regions, than to any sort of European vegetation.

These circumstances are too remarkable not to have suggested theoretical speculations. Accordingly, geology has long been conversant with sublime, but hasty, and sometimes unwarranted generalizations; and though she is still remote from the period when a lasting theory shall safely and satisfactorily embrace the explanation of observed phenomena, yet she has begun to assume a more philosophical aspect. Facts are rapidly accumulating, and the deductions of ingenious men, applying the science of number and calculation to this interesting study, have already established a basis on which future observers may erect a philosophical theory of the earth.

The crystalline arrangements of the particles of a considerable portion of the solid materials of the earth's crust, indicate, beyond doubt, that the particles have been once in a state of such freedom of movement as to assume a determinate order, that is, in a certain degree of fluidity; and hence the geologist assumes that such minerals must have either been softened, or melted by heat, or have been suspended in a fluid medium. The phenomena of stratification, and the incorporation of organized bodies in different rocks, are not less certain indications that a large portion of stony matter has obtained its present form by subsidence from water; and this constitutes the grand basis of geological reasoning. The second link in the chain, the relative age of the different strata, is derived from the order of their succession. As the strata bear the marks of deposition from water, it is obvious that the lowest in the series must have been the first deposited, and the upper beds the most recent formations. No such chronological data are afforded by the unstratified rocks, though their relations to stratified materials afford just inferences respecting their true era; but it is from the examination of the strata of any country that we can arrive at a just conception of its geological history.

Relative geological age, then, is to be sought by a careful examination of the superposition of the strata; and as the observation of organic remains has shown a succession of living forms to be peculiar to each series of rocks, the knowledge of such remains forms an important feature in modern geology. Certain fossils are peculiar to certain formations, and we are sometimes able to identify particular rock formations by their included organized remains. It is on such a basis that the most generally received classification of strata has been established; and geologists now usually divide them into three great classes, the primary, secondary, and tertiary.

The primary strata are such as do not include any organic bodies. They are the lowest in the series of strata, and are found to repose on crystalline unstratified rocks, generally on granite, which, as far as we have penetrated the crust, has the appearance of being the general fundamental rock. The name primitive was originally given to these rocks because they were supposed to have been deposited before the creation of organized bodies; but this involves a hypothesis which we cannot prove, as the disappearance of organic remains may be owing to the cause which crystallized such rocks. The term, however, as indicative of the oldest strata, is convenient, although the theory which suggested it may be exploded.

The secondary strata are generally less crystalline in structure, and in their lower beds the organic remains are few, being chiefly marine productions; but these increase in number and variety in the upper beds, which may be said to end with the chalk formation. The organized bodies are chiefly marine productions, as zoophyta, crustacea, shells, fishes; sometimes reptiles, amphibians, and plants are also found in secondary strata. During the deposition of the

secondary strata the land appears to have been gradually rising above the level of the waters which once covered it; and during the same period there are marks of more sudden changes or disturbances among them, which are most probably to be accounted for by the forcible eruption of igneous rocks from the bowels of the earth. The land appears to have attained extensive elevations above the waters after the deposition of the cretaceous formation, the newest of the secondary rocks; and another system of deposits now begins to be developed. Stratified rocks appear now also to have been formed by deposition from fresh water, as is proved by the abundant remains of animals inhabiting that element; and this is one characteristic of the tertiary strata. It is true, that in the coal-fields of the secondary rocks, we occasionally find alternations of sea and fresh water animals; which may be explained by such formations having taken place in extensive estuaries, into which rivers have flowed. But the decided alternations of fresh and salt water deposits are peculiarly characteristic of the tertiary period; and this, with the limitation of such formations to particular basins or circumscribed hollows, enables us to recognise them. Some of the best recognised of these beds are the Paris Basin, the London Basin (No. 6), and that of the Isle

of Wight and the opposite coast of Hampshire. Tertiary beds appear to be entirely wanting in Scotland and in Ireland; but they occur in Germany, in Spain, in Italy, and in different parts of Asia and America. In the upper strata of the tertiary period we find many shells similar to living animals in the adjacent seas; and bones of extinct mammalia also occur among their beds. Sometimes the remains belong to genera of which no species now exists; at other times to genera of which we still possess living types.

M. Deshayes has attempted to ascertain the dates of tertiary deposits by the number of living species they contain, considering that which contains the greatest number of existing species the most recent. He distinguishes three tertiary epochs. The most ancient is that of the Paris, London, and Hampshire Basins, in which he finds 3.5 per cent. of living shells, corresponding to the Eocene deposits of Mr Lyell. The middle group includes the basins of the southwest of France, of Baden, Vienna, Hungary, and Moravia. Among these he found 15.2 per cent. of living shells; and this corresponds to Lyell's Miocene period. The most recent strata include the tertiary beds of Italy, Sicily, Spain, of the south-east of France, the Morea, and the crag formation of England. In it he detected 45.4 per cent. of living shells; and it corresponds to the Pliocene formation of Mr Lyell.

Besides these great formations, there are local deposits of more limited extent, which we may slightly notice. 1. The formations at present going on, or elevated within historical periods, by earthquakes and volcanic agency from the bed of the sea, and added to the dry land. 2. Lacustrine deposits, now forming the rock called travertino, or giving rise to marly sediments. This last deposit, at Big-bone-Lick, in Kentucky, contains bones of the mastodon and other mammalia, with fresh-water shells; and at Market Weighton, in Yorkshire, the bones of the elephant, with those of the lion and the wolf. 3. The incrustations of calcareous matters cementing fragments of other rocks, as in the nagelfluhe of the Germans; or including bones of living species, which have fallen or have been washed into caves and fissures. 4. Superficial deposits of gravel and sand, evidently the produce of rivers or of floods, which sometimes contain bones

of elephants, the rhinoceros, &c., and are distinguished by the name of diluvian. 5. Erratic blocks of different rocks, which have been removed from their original beds, and scattered over extensive tracts. The most remarkable instances of this transportation are the profusion of large blocks of the zircon syenite of Norway, scattered over all the north of Germany, to the very foot of the Jura chain; of the granites of the Alps on the summit of the Jura, and beyond it into France; and of the granite of Shap Fell, over the intervening mountains, into Lancashire and Yorkshire. 6. The deposits formed by volcanoes at present in activity, consisting either of ashes, mud, or lava; to which we may add the obsidian, pumice beds, trachyte, and basalts of extinct volcanoes. The best characterized and most remarkable of the former are the ashes which have buried Pompeii, the mud eruptions of Tunguragua in 1797, the enormous lava streams of the Skaptar Yökul in 1783, and of Vesuvius and of Ætna at numerous periods: of the latter, we have examples in the obsidian of Lipari, Ascension Island, and Teneriffe; in the pumice beds of Lipari and of Auvergne; in the trachytes of the latter country, of Germany, and Italy; and in the basalts of almost every region.

The general opinion now entertained by geologists is, that the stratified materials of the earth's crust have been horizontally deposited from water at different eras, but that the unstratified rocks have been produced by the action of a high temperature. The consolidation of the older strata is supposed to have been accomplished by heat, produced either by internal changes among the elements of which they consist, or derived from a source of heat deep seated in the earth. The less crystalline state of the secondary strata, with the more perfect preservation of the forms of these organic remains, are ascribed to their having been subjected to a less intense temperature; while the still more earthy texture of the beds, and less altered nature of the organic bodies among tertiary strata, are ascribed, with seeming justice, to the still less heat they have experienced. The loose beds of clay and sand appear to have been deposited when the crust of our earth was cooled down to its present state.

The unstratified rocks are now generally allowed to be of unstratified origin, and believed to have been introduced into their present situations after the deposition of the greater part of the stratified materials. These are believed to have been liquefied by heat in the bowels of the earth, and forced upwards by the expansion of elastic matter, either in a softened state, or, in some instances, after they had been melted and again consolidated. In this view, granite, which so often forms the nucleus of mountains, and the lowest basis we can discover of other rocks, is supposed to have been acted on by heat, and slowly cooled after it had been elevated by the agent which liquefied it. This elevation is considered as having produced the fractures, dislocations, and elevations of the stratified rocks; and the heat of the melted material, thus introduced, is believed by most geologists to be the cause of the crystalline state of the older strata which rest on it. In some instances the elevating force has tilted

the strata to an absolutely vertical position; at other times it has carried up detached portions of the strata, and separated them from the principal bed, or even inverted the order of their superposition in considerable masses (No. 7). The injection of the fluid matter of granite between some of the strata, has given rise to the rare appearance of beds of this

Physical material; and when the force applied has rent or fissured the different strata, some of the liquefied matter has formed granitic veins, penetrating the stratified matter in different directions.

Trap-rocks, in like manner, are now generally considered as the immediate progeny of fire. They are occasionally found as interposed beds, or sometimes alternate with other matter, in distinct but thick beds, though not exhibiting the seams of true stratification. Most geologists regard such appearances as the consequence of subaqueous volcanic eruptions, in which liquid matter had flowed out of craters, spread over the adjacent strata, and been covered by aqueous depositions; after which a repetition of the same process may have given rise to such alternations. The more ordinary appearance of trap-rocks, especially of greenstone and basalt, is either in the form of veins penetrating stratified rocks, or as forming mountain caps, or huge beds superimposed on secondary, or sometimes on tertiary strata. Familiar instances of these veins occur in all coal-fields, more especially in the vicinity of the great coal deposits around Edinburgh and Glasgow. Arthur's Seat, near the former, affords a good example of a mountain capped by trap reposing on sandstone strata; and Salisbury Craig, of a vast bed of greenstone overlying the same strata (No. 8). The chalk of Antrim is in the same manner covered by greenstone and basalt; and the tertiary strata of Auvergne are often concealed by beds of trachyte or other ignigenous rocks.

The movements which produced these changes are believed to be comparatively of considerable antiquity, that is, before the records of history; but there are some causes, in operation at the present day, which may affect the stability of those more ancient arrangements, and which therefore we must now briefly consider.

SECT. V.—Circumstances affecting the present Order of Geological Phenomena.

These may be considered as four: Atmospheric action; the effect of the waters of the ocean and of floods; disturbances occasioned by earthquakes; and the influence of volcanoes.

1. Atmospheric action is slow, but constant. The united influences of winds and rains serve to form detritus, by which the higher portions of the earth's surface are gradually lowered, and the materials thus abraded carried by streams and rivers to the sea. The consequence must be, a very gradual change in the relative level of the land and the ocean. The abrasion of mountains by torrents and impetuous rivers produces deposits of gravel, sand, and mud in plains at their feet. These effects are increased by the sudden bursting of water long pent up in the bowels of a mountain, or dammed up in an alpine lake; thus giving rise to the phenomena which are termed debacles. Instances of the fall of mountains have been frequently afforded in the Alps and other lofty mountains. The fall of the Rossberg was foretold by General Pfeiffer, several years before it happened, from the existence of enormous collections of water in its bowels; and in the recent fall of one of the Aiguilles,

which border the southern side of the valley of Chamouni, a vast torrent of liquid mud, bearing before it woods and vast masses of rock, proved the influence which dammed-up water had in the catastrophe. The fall of the mountain of Chiavenna, which in 1618 buried the town of Pleurs and all its inhabitants, appears to have been produced by water insinuating itself among the shaly strata which underlie the limestone beds in the Italian Alps. Among lofty mountains, another cause of disintegration is perpetually at work. The accumulation of snow, and the formation of glaciers on their steep acclivities, where there is a considerable difference between the summer and winter temperature of the climate, is a powerful cause of the degradation of these elevations. The fall of the thundering avalanche, and the movements of the increasing glacier, from its own weight, detach vast masses of rocks, and hurl them into the valleys below. Even the action of minute particles of sand, moved by the winds for a succession of ages, is capable of wearing the hardest stones. Thus the polished surface of the granite of Sinai has been, with great probability, attributed to the perpetual descent of quartz sand over the rock; and we have in our possession nodules of Egyptian jasper, the exterior of which has been worn into smooth and deep channels, apparently by the long-continued action of the sands of the desert of Nitria on the surface of this hard mineral.

The heavier detritus of mountains is accumulated at their bases; but land-floods often convey large blocks of stone to astonishing distances. The finer particles are swept by rivers to the ocean. Humboldt has pointed out how the Gulf of Mexico is rapidly filling up by the sedimentary deposits from the waters of the Mississippi and other rivers which fall into that inland sea. The turbid waters of the Orinoco and the Marañon discolour the ocean to the distance of 300 miles from their mouths; and there can be no doubt that the formation of the flat land of Guiana, now about 100 miles in width, from the most eastern ridge of hills, is principally the alluvial deposition of these mighty rivers. The waters of the Ganges, at the highest inundation, are stated by Rennell to contain an enormous quantity of sedimentary matter; and we can hence conceive how the flat shores of the Hoogly, and the other branches, have originally been formed, and how the Bay of Bengal may, in process of time, be materially altered in its depth. Another striking instance of the effect of rivers in silting up a confined sea is to be observed in the Gulf of Leatong, or upper part of the Yellow Sea, into which a great part of the detritus of Northern China is deposited by its numerous large rivers, and by the effects of the strong currents in that inland sea. Staunton has calculated that the Yellow River alone discharges into the Yellow Sea more than 2,000,000 of cubic feet of earthy matter every hour. The matter carried down by the Nile and the Po is well known to affect the extent and deepness of the adjacent seas; and the distance of Ravenna from the present shores of the Mediterranean (between three and four miles) proves how a comparatively small river like the Montone, pouring its sediment into a still bay, has altered the situation of a port, where, in the ages even of the lower empire, the Roman fleets used to anchor. These facts would lead to the supposition that the land was gaining upon the sea; but there are other circumstances which are at least equivalent to this change.

The natural consequence, as was well pointed out by Playfair in his Illustrations of the Huttonian Theory, and previously noticed by Manfredi, must be to elevate the surface of the ocean, by the deposits of detritus in its bed. This rise is most conspicuous in inland seas. Thus, on the western side of Italy, the ruins of the palace of Tiberius at Capreae are now wholly covered by the sea; and the far-famed temple of Serapis, situated on the Bay of Baiae, has its pavement below the present level of the Mediterranean. Man-

fredi states, that when the ancient pavement of the cathedral of Ravenna was discovered in 1731, it was found to be a foot below the present high-water mark; from which we must infer that the level of the Adriatic had risen at least as much in 1500 years. The shores of Baia may have had the level altered by the vicinity of an active volcano, and the temple alluded to shows indications of more than one upheaving and depression of the land; but the district of Ravenna is all alluvial; and a similar rising was noticed at Venice by Zendrin.

The same fact is attested by observations along our own coasts. The submarine forest on the coast of Lincoln, first described by Correa de Serra in the Philosophical Transactions, 1799, proves a rise in the level of the sea; and the tradition, that the Godwin Sands once formed part of the fertile domain of the Saxon earl of that name, has probably a fair claim to be arranged with the same phenomenon.

2. The wasting of the land by the waves of the ocean, and the overwhelming of large tracts by the sea, are nearly allied facts. The changes on our own coasts, and on the countries between the Rhine and the Elbe, afford striking instances of the encroachments of the ocean on the land. Even where the rocky boundaries are the most solid, the ceaseless action of the waves has wasted and disintegrated the cliffs around all our shores; but the destruction of the softer strata along the east coasts of England, and the chalky heights of Dover, is still going on, and affords undeniable proofs of the power of this agent. The swallowing up of whole districts along the coasts of Yorkshire is shown in the annihilation of the villages of Auburn, Hyde, and Hartburn, and in the ravages of the sea on the neck of land uniting the Spurnhead with the rest of the country. We have authentic records of the destruction of Dunwich, on the coast of Suffolk; and near Harwich, in Essex, the sea is annually encroaching. The Start islet was within the memory of man a part of the island of Sandey, in the Orkneys.

Still more remarkable changes have taken place on the coasts of Holland and Friesland within historical periods. Well-known instances of this occurred in the devastation which accompanied the formation of the Zuyder Zee in the beginning of the thirteenth century; in the successive encroachments of the ocean, which formed the islands near the Texel in 1251; and in the disappearance of the once fertile district of North Friesland, the only remnants of which, since the disastrous 11th of October 1634, are the three small islets of Noordstrand.

Such changes have occurred in various parts of the earth. Some of them, especially the overwhelming of low sea-coasts, may be owing to the gradual elevation of the level of the sea by detritus washed into it; but probably still more is due to changes in the level of the land by unequal subsidence, the consequence of earthquakes.

3. The terrible effects of earthquakes have attracted attention in every age, and records of some of such convulsions have descended to us from very remote ages. No period of our earth's history has been more noted for the violence and extent of earthquakes than that between the first, third, and the middle of the fourteenth century. China was terribly convulsed for ten years from 1333, when Kiang-si, its capital, was swallowed up; mountains were engulfed, and floods, occasioned by the obstruction of the course of rivers, destroyed vast multitudes of human beings. The succussions extended westwards, and Asia Minor and Egypt were violently shaken in 1346; while in the following year severe earthquakes were experienced in Cyprus, Greece, and Italy. In 1692, the island of Jamaica was visited by a most terrible earthquake, in which enormous masses of earth and rock were detached from the Blue Mountains, and vast quantities of timber, hurled from their flanks, covered the adjacent sea like floating islands. It was during this earth-

quake that the city of Port Royal, with a tract of adjacent land, estimated at 1000 acres, sunk, in one minute, into the deep. In the succeeding year great earthquakes in Sicily destroyed the city of Catania, and a hundred and forty other towns and villages in that island, where upwards of 100,000 persons perished. In 1746, Lower Peru suffered severely from this calamity. The ocean burst in upon the land with irresistible force, when the barrier of land sunk into the sea, Lima was overwhelmed, and the present port of Callao formed. These convulsions were accompanied by eruptions of water and mud from several volcanoes among the Andes. In 1750, the city of Concepcion, in Chili, disappeared during an earthquake, and the sea rolled over it. In 1755, Lisbon was in a great measure destroyed by one of the most terrible earthquakes that ever visited Europe. The mountain chains between the Douro and the Tagus were most dreadfully convulsed. The new mole at Lisbon, to which multitudes had fled, as to a place of safety, suddenly sunk into an hideous abyss, and not one body floated to the surface, nor were any fragments of the vessels sucked into the chasm rendered up, and on the spot there is now an hundred fathoms of water. In this awful convulsion at Lisbon sixty thousand persons perished in about six minutes. A violent shock threw down the greatest part of the city, and the sea retired, leaving the bar momentarily dry; but suddenly a mighty wave, fifty feet high, rolled in on the devoted city. The extent of the terra motus, on this occasion, is very remarkable. The violence of the shocks, which were accompanied by a fearful subterranean noise, like the loudest thunder, was chiefly felt in Portugal, Spain, and Northern Africa; but its effects were perceived over a considerable part of Europe, and were even experienced in the West Indies. Our Scottish lakes, particularly Loch Ness and Loch Lomond, rose and fell repeatedly on that dreadful day. Ships at sea were affected by the shocks, as if they had struck on rocks, and their crews were in some instances thrown down by the violence of the concussions. In 1766, the island of Trinidad, and great part of Colombia, were violently agitated by earthquakes; an islet in the Orinoco disappeared, and land in other parts of the coast was raised above the waters. In 1772, the lofty volcano of Papandayan (the loftiest mountain in Java) disappeared, and an area around, fifteen miles by six, was swallowed up. Most terrible earthquakes desolated Calabria in 1783. This calamity has been admirably described by several writers, among whom we may mention Dr Vivienzo, Grimaldi, Sir William Hamilton, Dolomieu, and the commission of the Royal Academy of Naples. The violence of the shocks was chiefly felt in the Farther Calabria, and in the neighbourhood of Messina. The principal agitation was felt over an area of five hundred square miles. Many sudden sinkings of the land were perceived, numerous fissures were formed, and partial elevations were effected in some places. The greatest depressions happened at Terra Nuova, Oppido, Sinnopoli, and Santa Christina: a fissure a mile in length, a hundred and five feet in width, and thirty feet in depth, opened at Plaisano; another about the same length, and a hundred feet in depth, was formed at Cerzulle; and at La Fortuna a chasm a quarter of a mile in length, thirty feet in width, and two hundred and twenty-five feet in depth, suddenly opened in the ground. In some districts considerable mountain-slips, or the separation of huge masses of earth and rock, took place; and along the Straits of Messina, in the neighbourhood of Scylla, the cliff of Gian-Greco, a mile in length, was precipitated on the subjacent houses and gardens. A fragment detached by the earthquake from Monte Jaci crushed multitudes who had fled to the shore for safety; and at the same moment a wave broke on the devoted shore, and swept away the aged prince of Scylla, and numbers of his people.

In the year 1797, Upper Peru was terribly convulsed. The

Physical Geography. shocks of earthquakes continued with great violence for three months, and the face of the country in the centre of the convulsion was totally changed. In 1811, violent earthquakes shook the valley of the Mississippi, by which lakes of considerable extent disappeared, and new ones were formed; but these were less terrific than the catastrophe which destroyed the city of Caraccas in 1812. On the 26th of March there were heard subterranean thunderings; the ground undulated, as if agitated by a boiling liquid, and at one shock this fine city entombed in its ruins 10,000 of its inhabitants. During the earthquake the great lake of Maracaibo had its level lowered, and the riven earth at Puerto Cabello and Valencia poured forth enormous torrents of water. It is remarkable, that the volcano of St Vincent, which had been perfectly quiescent for a century, burst out with prodigious violence on the 27th of April in the same year, and threw out clouds of ashes, which rose to an immense height into the air. Much of the island was ruined by showers of scoriae and ashes; and such was the violence of the eruption, that the decks of vessels 200 miles to windward of St Vincent were covered with an impalpable dust. On the day of this eruption subterranean thunderings were distinctly heard at Caraccas, and even on the Rio Apure, 210 leagues in a right line from St Vincent. In the eruption of Tomboro, about to be noticed, the earthquakes extended throughout an area a thousand miles in diameter, a considerable tract of land at the foot of the mountain disappeared in the waves, and the port of Bima, where ships of war could formerly anchor, was ruined by the up-raising of a shoal. On the 19th of November 1822, Chili was visited by a most destructive earthquake. The shock was strongly felt at the same time throughout a line of coast 1200 miles in extent. It is stated, on good authority, that the coast for one hundred miles sustained an elevation of from two to four feet; and about a mile inland from Valparaiso it was raised from six to seven feet. The sudden elevation of the coast was indicated by shell-fish being found adherent to the rocks considerably above high-water mark. The shocks continued until the end of September in the following year; and the area over which the permanent alteration of level extended is believed to embrace the country from the base of the Andes to the sea, a surface of not less than 100,000 square miles. In 1827 Popayan and Bogota suffered most severely from earthquakes, during which vast fissures opened in the elevated plains around the latter city.

The last earthquake in Europe occurred in Murcia in 1829, near Alicante. Several villages in an area of above four square miles were thrown down by vertical movements in the valley through which the Rio Segura flows, and many small fissures were formed in the alluvial soil; while black mud, sand, and marine shells were thrown up from small cavities formed near the sea. Such are some of the severer earthquakes on record; but less considerable shocks are of frequent occurrence in various countries, especially in South America and in Italy. Smart shocks are occasionally felt in Scotland. They have often occurred at Comrie in Perthshire. A smart shock rent the spire on the town-hall of Inverness in the year 1816; and another earthquake was felt at Lancaster in 1834, which shattered chimneys, and alarmed the inhabitants. But all the shocks experienced in our island have fortunately been insignificant compared to those which have been felt in many other countries.

4. The alterations produced on our earth's surface by volcanoes in activity within historical periods appear to be underrated by some writers. Few geologists now will be disposed to doubt that most of the phenomena of earthquakes are the consequences of volcanic fire. The coincidence of those formidable shakings of the earth with volcanic eruptions, and the distance to which such agitations have been felt, show that the causes of both are deep-seated in the earth, and that they are probably the consequences of elastic matter formed by subterranean heat. We shall say nothing of

the almost universal diffusion of trap-rocks, which, under the name of trachyte, basalt, greenstone, and trap-tuffa, are now generally admitted to be of igneous origin, nor of the true craters of extinct volcanoes which abound in France, Germany, and many other parts of the earth, but which have ceased to emit any eruption within known periods, and confine our attention to those mountains now either in active combustion, or which have been so within historical periods. Gay-Lussac enumerates 163 volcanoes active at the present time; and Professor Jameson gives the following distribution of 193 volcanoes in a state of present activity:

To these we may add the gigantic crater of Kirauea, in Hawaiah, one of the Sandwich Islands; the volcano of Tomboro, in the isle of Sumbawa; the volcano of Tofua, in the Friendly Islands; that of Barren Island, one of the Andamans, in the Bay of Bengal; to which we may perhaps add the submarine volcanoes which have suddenly risen from the bosom of the deep, but soon afterwards have again disappeared, as that of Nyoë, off the coast of Iceland; Sabrina, off St Michaels, in the Azores; and Graham's Island, not far from the east coast of Sicily.

The substances ejected by volcanoes consist either of melted stony matter termed lava, of scoria, of ashes, of sand, of muddy eruptions, or of vast quantities of hot and fetid water, always accompanied by earthquakes and subterranean thunderings, and generally by columns of flame or of dense smoke.

The general form of a volcano is a truncated cone, often isolated, and having a circular depression or spiracle on or near its summit, which is named a crater. From this crater smoke is perpetually issuing; sometimes flame appears, or volleys of ignited stones are projected into the air. From

the dark cloud of smoke emitted during eruptions, vivid lightnings are usually seen to dart; and the vapours of sulphur, ammonia, or of muriatic acid, are often emitted. Sometimes the boiling lava is elevated from the bowels of the mountain to its terminal crater, fills that cavity entirely, and descends from the cone in streams of liquid fire, which concretes as it cools on the surface, but continues to move on with resistless force till it congeals into stone, though it often retains its heat for long periods. At other times the elasticity of the pent-up matter is incapable of forcing the lava to the highest crater; but the liquefied material bursts through the flanks of the mountain, and pours into the plains. Sometimes the volcanic cone is included in an exterior one; a circumstance on which Von Buch has founded his theory of craters of elevation (No. 10). The strata or beds in both cones are inclined from the general axis. The projectile power of volcanoes is prodigious. The dust of the volcano of St Vincent was carried more than 200 miles

eastward of that island, and must have attained a vast elevation, to obviate the effect of the trade-wind in its descent. It is stated, on credible authority, that Vesuvius has projected large stones 3600 feet above its summit; and Cotopaxi was ascertained by the French academicians to have hurled a rock, calculated to weigh 200 tons, to the distance of three leagues, or more than ten English miles.

Instances of volcanic eruptions are by no means rare. There are fifty-nine or sixty eruptions of Ætna recorded by historians, and about fifty-two of Vesuvius. The first eruptions of Ætna are lost in the remoteness of antiquity. Vesuvius was not in existence as an active volcano till the year seventy-nine of the Christian era, when the ancient crater presented a slight circular hollow, the sides of which were overgrown with wild vines, and the exterior of the cone was well cultivated. In the eruption which then took place, Stabiae, Herculaneum, and Pompeii were destroyed, and the elder Pliny lost his life. Stromboli appears to have been an active volcano from 300 years before Christ, and is still in a state of incessant activity, ejecting flame and red-hot stones in quick succession. The most noted eruption of Ætna was probably that of 1699, when, from a long rent on its flank, it poured out a stream of lava, which in its course to the sea destroyed fourteen towns and villages, some of which contained three or four thousand inhabitants. This current is fifteen miles in length, about 600 yards at its broadest part, and there it is forty feet deep. Vesuvius has more frequently been active in our own times than Ætna. One of its most terrible eruptions was that of 1538, when Monte Nuovo was forced up in two days to the height of 413 feet, and with a circumference of 8000 feet. Another happened in 1631, which covered with lava most of the villages at its foot, and sent forth torrents of boiling water. In little more than a century it has had eighteen eruptions. The eruption of 1794 has been well described by Breislac, who estimates one of its streams of lava on that occasion as containing upwards of forty-six millions of cubic feet. It again burned with great fury in 1805, 1813, and in 1822. We possess lavas of this latter period so strongly impregnated with sulphuric and muriatic acids, that they destroy every species of vegetable material they rest on; and, when recent, the specimens smelled strong of the latter acid. The volcanoes of Iceland have been in activity ever since the island was discovered, in the ninth century. Hekla, though not the most considerable of these, from its position and its former activity, is the best known. It has had many formidable eruptions, twenty-two of which have been noted in about 800 years; and in the same period we have notices of twenty eruptions from five other Icelandic volcanoes. A succession of eruptions of Hekla lasted for six years; but the most severe convulsions of that country happened in 1783, when the dreadful eruption of the Skaptár Yökul burst forth, and did not cease till the following year. About a month before this terrible catastrophe a submarine volcano elevated the crater of Nyoë, seventy miles south-west of Cape Reikianes, and threw out such an immense quantity of scoriae as to cover the sea, to the distance of 150 miles, with a stream which impeded the progress of ships making the island; and portions of this eruption floated as far as the Shetland and Orkney Islands. Nyoë emitted smoke and scoriae from several apertures;

but within a year the island disappeared, and a shoal marks its former site. On the 8th of June the Skaptár Yökul threw out smoke; on the 10th an enormous current of molten lava flowed from numerous cones on the Yökul, which, dividing into two main streams, pursued its course to the sea, filling up the beds of two large rivers, and covering an immense extent of once productive country. The horrors of the scene were aggravated by the enormous torrents of boiling water produced by the liquefaction of the glaciers that covered the Yökul, and by incessant showers of ashes which darkened the sun; stream of lava succeeded stream from the 10th of June to the end of August, at short intervals; and noxious emanations destroyed numbers of those whom fire and water had spared. From this calamity Iceland has never recovered; for within the space of two years the island, in consequence of this eruption, lost 9336 persons, 11,460 head of cattle, 28,000 horses, and 190,480 sheep. The extent of the principal stream of lava is fifty miles in length; its greatest breadth is from twelve to fifteen miles; in the plains its general depth is 100 feet, but in the channel of the Skaptár River, which it dried up, it is 600 feet in perpendicular depth. The south-western side of Iceland appears to be one vast focus of subterranean fire; for the several eruptions of the Oræfi, the Skeidera, Sida, and Skaptár Yökul, seem but as occasional outbreaks from one immense volcanic fissure, which really belongs to the same chain of icy mountains.

The volcanic formations of Java have been well examined by Dr Horsfield, who describes them as extending through the centre of that noble island. They form a chain, in which thirty-eight large mountains, evidently volcanic, are conspicuous, and attain an elevation of from 5000 to 12,000 feet. Most of them are no longer in activity. Tankuban Prahun has been quiescent for a series of ages; but its crater contains a large lake, a mile and a half in circumference, and 250 feet deep, from one side of which sulphureous vapour issues with great violence. Papandayang was once the largest of the Javan volcanoes; but in August 1772 the greatest part of it was swallowed up during an eruption. A luminous cloud at first seemed to envelope the mountain, prodigious quantities of ashes were ejected, which destroyed forty villages, and an area of fifteen miles by six was entirely swallowed up.

The eruption of Tomboro, in the island of Sumbawa, in the year 1815, described by Sir Stamford Raffles, is one of the most violent on record. The earthquakes caused by this eruption were felt over a circle with a radius of 1000 miles; and all around to the distance of 300 miles the quantity of ashes excited terror and dismay. In Java, though 300 miles in a direct line from Sumbawa, the sky at mid-day was darkened by dense clouds of fine dust ejected from the volcano. In Sumbawa it was attended by a whirlwind of surprising fury, which lasted for an hour, sweeping away trees and houses in its career. The explosions commenced on the ceasing of the whirlwind. They began on the 5th of April, and continued until the 15th of July. Whole villages were swept away, and upwards of 12,000 persons perished in this eruption. Vast numbers of trees, torn up by the whirlwind, strewed the ocean to a considerable extent. Raffles states that the explosions of Tomboro were distinctly heard in a circle with a diameter of 1700 miles; and that the quantity of light scoriae and pumice it threw out within a week formed a bed on the ocean two feet in thickness, and many miles in extent, which impeded ships in approaching Sumatra.

The volcanoes of America claim notice, not only from their magnitude and number, but from the peculiarities of South their position. Several of the most lofty mountains of the South American Andes are active volcanoes. Cotopaxi has an elevation of 16,800 feet above the sea; the volcanoes of Mecas and Antisana equal 17,000 feet, and Tun-

Physical Geography. guragua 16,500 feet. Humboldt considers these enormous cones as the spiracles of one prodigious volcanic vault, stretching from south to north, which presents an area of 600 square leagues; and he regards them as merely different summits of the same volcanic mass. "The fire issues sometimes from one, sometimes from another, of these summits. The obstructed craters appear to us to be extinguished volcanoes; but we may presume, that while Cotopaxi or Tunguragua have only one or two eruptions in the course of a century, the fire is not less active under the city of Quito, under Pichincha, and Imbaburu."

The connection of the volcano of Pasto, in New Granada, with the volcanoes of Quito, was strikingly displayed in 1797. A column of black smoke had continued for several months to issue from the Volcan de Pasto; but, to the surprise of the inhabitants, this smoke suddenly disappeared on the 4th of February 1797, at the precise moment when, at sixty-five leagues from the city of Pasto, the city of Riobamba, near Tunguragua, was destroyed by a terrific earthquake.

The eruption of Tunguragua, on the 4th of February 1797, was of a very extraordinary nature. The country around, to the extent of forty leagues from south to north and twenty leagues from east to west, sustained an undulatory movement of extreme violence, which lasted four minutes. In the district around the mountain every town was levelled with the ground, and the cities of Riobamba and Quero were buried under the ruins of the impending mountains. The base of Tunguragua was riven asunder, and poured out, from numerous apertures, streams of water and mud, which accumulated in the valleys to 600 feet in depth. The mud covered everything, blocked up the channels of the rivers, forming lakes which remained for more than eighty days. Fetid, suffocating exhalations were emitted by the lake Quilotoa, and it is said that flames also issued. The covering of mud, and the levelling of elevated portions of the country, totally changed its face. The shocks continued to be violent for upwards of three months, and were felt over a district 170 leagues from south to north, and 140 from west to east. It was during this eruption that the curious fishes, Pimelodes Cyclopum, were discovered in the ejected water of the volcano. They are supposed to be generated in subterranean lakes in the bowels of the mountain, and only to be brought to light by such an explosion. There are no less than sixteen active volcanoes in Chili, the eruptions of which are confined to the valleys of the Andes. A line of volcanoes in South America may be distinctly traced from 5° to 6° of south latitude, running parallel to the ocean, and occupying the highest ridges of the Andes.

In North America, Mexico presents a chain of active volcanoes crossing the general direction of the cordillera of Anahuac from east to west. It commences on the coasts of the Mexican Gulf, with the small but energetic Volcan de Tuxtla, which lies a little to the southward of a line joining the rest; next is the vast cone of the peak of Orizaba (No. 9), which attains an elevation of 16,365 feet; and then the still loftier Popocateptl, rising to 17,060 feet. These lie to the east of the city of Mexico; and on its west are the cones of Jorullo and Colima. They are all active; but at no very distant period the snow-clad summits of Iztaccihuatl, of Toluca, and of Tancitara, lying in the same line, were also burning. This chain extends for 140 leagues; and perhaps the volcanoes should be considered as spiracles, all connected with an huge ignited fissure, passing throughout the table-land of Mexico. Five of these mountains penetrate far into the region of perpetual congelation; and from the inaccessible summits of Orizaba and Popocateptl are pinnacles of frozen snow, from which smoke continually issues.

The extent of these eruptions is attested by the prevalence of volcanic rocks over the whole table-land; and the present form of this magnificent plateau would seem to be

derived from the accumulation of igneous formations Phys in a series of valleys among primary rocks. Tuxtla had a Geogr strong eruption in 1793, which covered the roofs of the houses at Oaxaca, Vera Cruz, and Perote, with ashes; and at the latter, which is 196 miles distant, the subterranean noises resembled heavy discharges of artillery.

The formation of the volcano of Jorullo dates not farther back than the year 1759. Until that period, the present site of the volcano was a fertile plain, well cultivated, and producing indigo and the sugar-cane. In the month of June in 1759 subterranean thunderings commenced, accompanied by frequent shocks of earthquakes; and these alarming phenomena continued to terrify the inhabitants for fifty or sixty days. About the beginning of September everything seemed to announce the re-establishment of tranquility; but on the night of the 28th the horrible subterranean explosions were renewed, and an area of four square miles rose up, like an enormous bladder, to the height of above 500 feet. Those who gained the mountains around saw flames shoot up, from a surface of a square league in extent, and huge fragments of ignited rock hurled into the air through a dense cloud of ashes. The softened surface was seen to undulate like an agitated ocean. The waters of two streams precipitated themselves into the burning chasm, and this seemed to add fresh fury to the flames. Thousands of small burning cones, from six to eight feet high, issued from the softened and inflated surface; and in the midst of these hornitos or ovens, which still pour out steam, and emit the sounds of internal ebullition, six vast mountain masses arose, to a height varying from 300 to 1690 feet above the former plain. The most elevated of

these is the Volcan de Jorullo, which is perpetually burning, and has ejected a vast quantity of ashes and scoriaceous lava, often mingled with fragments of primary rocks. The ashes of this eruption were scattered to a distance of 160 miles. The eruptions of the Jorullo continued till February 1760 without intermission, but they have since become less and less frequent. On this catastrophe Humboldt justly remarks, that an event, "by which so considerable an extent of country entirely changed its appearance, is perhaps one of the most remarkable physical revolutions in the annals of our globe. Geology indicates parts of the ocean where, within the last two thousand years, several small volcanic islands have been formed; but it gives no other example of the formation, from the centre of a thousand burning cones, of a mountain of scoriae and ashes 1695 feet above the level of the adjoining plain, upwards of thirty-six leagues from the sea, and forty-two leagues from every other active volcano;" and he adds, that though within six days' journey of the Mexican capital, this wonderful event had remained unknown in Europe until the period of his journey (No. 11).

In the works of the South Sea missionary Ellis there is an interesting account of the volcano of Kiraua, in the Sand-islands (which Islands. The devastating effect of volcanic fire in the South Island of Hawaiah had been noticed by Captain Cook, but it has been particularly described by Mr Ellis. The route to the present volcano is over a vast waste of naked lava. Instead of a mountain-crater, however, Kiraua presents a chasm in a plain, perhaps occasioned by the engulfing of the

previous elevation. He describes it as a flat area of fifteen or sixteen miles in circumference, sunk from 200 to 400 feet below the level of the country by which he approached Kirauea. In the bottom of this area yawns an immense abyss, in the form of a crescent, about two miles in length from north-east to south-west, and about a mile in breadth at its widest part. One portion presents a flood of glowing matter in a state of terrific ebullition. Fifty-one conical islets, containing as many craters, rise around the edges, or from the surface of the burning lake. Some of them emit flame and smoke, and others occasionally vomit streams of lava, which roll in blazing torrents down their black sides into the boiling flood below. The sides of the gulf just described are perpendicular, and rise from a wide horizontal ledge of solid lava of irregular breadth, but extending completely round the chasm; and beneath this ledge the sides slope to the burning lake, the surface of which appears to be from three hundred to four hundred feet lower.

Besides this volcano, there are several others now extinguished in the island. Of these, one is Mouna Roa, which Captain King calculated, from the position of its perennial line of snow, to be 16,020 feet in height; and the same navigator assigns to another, Mouna Kaah, the vast elevation of 18,000. In a word, the researches of Mr Ellis would lead us to infer that the whole island of Hawaiah, an area of about 4000 square miles, is covered by lavas and other volcanic matters in various stages of decomposition.

The whole islands in the Pacific are divided by Kotzebue into two classes; the low and the elevated. The former are coral reefs, of which the formation has been so well described by Captains Flinders and Beechy. Their bases would appear to be submarine mountains, which are brought to the surface of the waters by the depositions of the coral animalcules. Captain Beechy examined thirty-two of these islands, the largest of which was thirty miles, the smallest one mile, in diameter. Twenty-nine had lagoons within their centres, which seem once to have existed in them all, until they had been filled up by the operations of the zoophytes. These low coral islands present a singular appearance. (No. 12, a section.) They often form chains, extending for hun-

dreds of miles with little interruption. Their exterior is a border of dead coral rock, and stands a few feet above the waves, on which the seeds of plants, wafted by the sea, soon fix their roots, and speedily a verdant wood encircles the island, whilst its central lagoon is still the habitation of living zoophytes. The elevated islands of the Pacific very generally abound in volcanic products, and numbers of them are still the seats of active volcanoes. Many of them have the appearance of having been raised from the deep by submarine volcanoes; and Mr Lyell has with much force of reasoning contended, that numerous coral islands ought to be considered as the cones of extinguished submarine volcanoes covered over by coralline formations. If we adopt this idea, and extend it to the enormous coral reefs described by Flinders, King, Horsburgh, and Chamisso, the surface of the Pacific must repose upon a prodigious magazine of volcanic energy, which is from time to time destined to increase the surface of the habitable globe. A very extensive line of active volcanoes exists in the Great Eastern Ocean, which has been noticed by Lyell, and forms too remarkable a feature on our globe to be omitted in a sketch of physical geography. It commences in Barren Island, off the coast of the Malayan peninsula; is continued by a chain of active volcanoes through Sumatra, Java, Sumbawa, and Flores; it then turns northward through the Moluccas, the Philip-

pines, Formosa, and Loo-Choo; it passes through Japan, Jesso, and the Kuriles; reaches the peninsula of Kamtschatka, and terminates in the Aleutian Islands, after a course of between 9000 and 10,000 miles, besides the several branches which are sent off to the eastward in the directions of Papua and the Ladrone Islands.

The cause of volcanic phenomena is involved in obscurity; but various conjectures have been proposed by geologists. The Neptunian hypothesis has ascribed them to the inflammation of beds of coal, or of other combustibles, in the burning mountain; regarding them as local, or very limited in their range. Lemery's experiment on the heat extricated by moistening a mixture of iron filings and sulphur, and burying it in the earth, was supposed to afford a satisfactory explanation of the manner in which the fire might originally be kindled; and the extinction of volcanoes was supposed to give probability to the local nature of the heating cause. Abbé Breislac modified this hypothesis, by supposing that volcanic fires were fed by collections of petroleum collected in caverns, and set on fire by the chemical action of several other substances. Both these views seem to afford too narrow a basis for the extent of volcanic energy, the apparent connection between distant volcanoes, and the great length of time that some of them are known to have remained in activity. The Huttonians consider volcanoes as spiracles connected with the great source of central heat. (Playfair, 86.)

Another hypothesis has been suggested by the discovery of the metallic bases of the alkalis and earths. Their bases have been supposed to constitute the interior of the earth; and when water finds its way to them, through fissures in its crust, it will be rapidly decomposed, with the extrication of intense heat and the evolution of elastic gases, and thus the phenomena of earthquakes and volcanoes may be produced. This view, originally suggested by Sir Humphry Davy, has been illustrated by Humboldt, and especially by Daubeny, with great ingenuity; and although it has not yet arrived at the state in which, perhaps, we are entitled to designate it a theory, it is an hypothesis which affords a satisfactory explanation of the observed phenomena.

All the enumerated causes have produced and are still producing changes upon the surface of the globe; some of which are general, though slow in their operation, whilst others act much more violently, but are more local in their effects.

Mud and air volcanoes, as they have been termed, appear to owe their existence partly to the disengagement of gases by subterranean fire, but in a great measure to water from a higher level, compelled to pursue a subterranean course, by a thick bed of tenacious clay suddenly escaping by a perpendicular fissure, with a force proportional to the height of the aqueous column. An ingenious explanation of the phenomena of the mud eruptions of Sicily and Italy, on this principle, is given by Dr Daubeny. One of the most noted eruptions of this sort is that of Maculuba in Sicily, which in 1777 threw up a vast column of mud, mixed with stones and naphtha, and the disengagement of sulphureous vapours, with much hydrogenous gases. Similar eruptions occur near Catania, and also near Modena. The gaseous exhalations of the Grotta del Cane and of the lake of Ansanto in the kingdom of Naples, are probably disengaged from mineral bodies by heat. The inflammable air of Pietra Mala near Bologna, and of the petroleum wells of the country around Baku, on the western shores of the Caspian, are evidently connected with subterranean fire. In the latter district there are strong proofs of volcanic agency still active in the bowels of the earth. The singular eruption which M. Figalgo saw at the northern extremity of Prince William's Sound in North America, belongs to the same class. "The Indians conducted him into a plain covered with snow, where he saw great masses of ice and stones

Physical Geography: thrown up to prodigious heights in the air with a dreadful noise." (Humboldt.)

SECT. VI.—Distribution of the Waters of the Globe.

Natural water never exists in a state of absolute purity. When collected as it descends from the atmosphere, it always contains air; and when derived from springs, lakes, or rivers, it invariably contains more or less of saline ingredients. Dr Henry found in spring-water between four and five per cent. of its bulk of gas, of which nearly three and a half consisted of carbonic acid, and the rest of atmospheric air. Besides the gases, natural waters contain the following salts; muriate of soda, muriate of lime, sulphate of potassa, sulphate of lime or of magnesia, carbonate of lime, and sometimes carbonate of soda. Hail and snow water, when collected at a distance from habitations, contains scarcely any trace of these salts. Rain-water is next in purity; but dew-water is generally contaminated with vegetable matter, as well as with salts. The water of springs usually abounds with these salts, and often with other ingredients. River-water differs much in its purity; but when it flows through a siliceous rock it is often very pure. Lake-water is subject to much contamination; but often, when the lakes are deep and extensive, the water is depurated by subsidence and rest; though stagnant water, when shallow, becomes contaminated with vegetable and animal matter. The water of the ocean contains about three per cent. of saline contents, or one thirty-third part of its weight. The quantity of saline matter appears to increase slightly from the highest latitudes towards the equator. The ocean, from its extent, claims our first attention.

1. OCEAN.—From data already given in section second, we shall probably not err if we estimate the surface of the whole ocean and its branches at 144,473,000 square miles. It is not so easy to arrive at a safe conclusion with regard to its mean depth. Laplace and later investigators seem to believe that the depths of the ocean may be in proportion to the elevation of the land, and the ratio of their respective areas. Mrs Sommerville states, that although it is probable that profound cavities of small extent exist in the bed of the ocean, its mean depth does not exceed the mean height of the continents and islands above its level. According to Lagrange, the velocity of the tidal wave is a function of the depth of the channel through which it moves; and recent investigations respecting the velocity of the tidal wave in these seas, would lead us to infer that the mean depth of the sea around the coasts of England is about 120 feet, on those of Scotland 360 feet, on the western coast of Ireland 2000 feet, and in the main Atlantic 50,000 feet, or a little more than nine miles and a half. The examination of the co-tidal lines on Whewell's charts would lead to the conclusion that no considerable extent of the ocean has a greater depth than the Atlantic. If the depths of the ocean are indeed proportional to the elevation of the land, in the ratio of their respective surfaces, the greatest profundity of the oceanic abyss would be about fifteen miles, and its mean depth, at a distance from the land, would vary from one to five miles. If we assume four miles as its mean depth, this would give 577,892,000 cubic miles as the volume of the whole waters of the globe; or, if we reckon it five miles in depth, the mass of waters would equal 722,365,000 cubic miles.

The waters of the ocean are in perpetual movement, from the effects of winds, currents, and the tides. The effect of winds in agitating the ocean is chiefly confined to its surface. The most powerful waves in the free sea only rise to the height of a few feet; and it is highly probable, that in the profound depths of the ocean, there reigns a perpetual stillness, unruffled by any movement, except in as far as general currents may disturb the tranquillity of the waters. The surface, acted upon by strong gales, is thrown into long

ridgy waves, which rise with an uniform swell, and maintain the same velocity for considerable distances, the top curl. Geometry in foaming surges, which are precipitated on the undulating surface below, often to the great danger of the navigator. No region of the ocean is subject to more violent oscillations of this nature than that portion of the Atlantic termed the Bay of Biscay, when agitated by a smart gale of westerly wind. Geometry has been applied to measure the velocity of waves, and Lagrange estimates, that where the depth is not very great, the velocity will be the same as a heavy body would acquire in descending from a height equal to half the depth of the water, and that the velocity of the wave is in the duplicate ratio of the depth of the channel in which the water moves. It is believed, however, that the wildest tempest in the open sea does not agitate the water at the depth of an hundred feet; but when waves meet with resistance, as when they dash upon lofty cliffs, we have known them rise apparently in one unbroken mass to the height of two hundred feet.

There are other movements of the waters of the ocean, termed currents. The most general currents are produced by the movement of the waters perpetually setting from the polar regions towards the equator, and the progression of the tropical seas towards the west. Both are chiefly caused by the centrifugal force of the earth's diurnal rotation. This produces a culmination of water towards the equator, and thus a current is generated towards the part of the earth in most rapid motion; but the water accumulated there passes off before the influence of the trade-winds; and the strength of the western current is increased by the mobility of the particles of water preventing the fluid immediately acquiring the velocity of that portion of the earth over which it arrives in its progression towards the equator, by which it will appear to have a motion from east to west, or in an opposite direction to the earth's rotation. The higher density of the sea, when it has been cooled by the polar ices, will increase its tendency to displace the lighter waters, which are forced westward by the constant pressure of the trade-winds, and the other causes already mentioned. The currents from the north have been well described by Mar- res, Scoresby, and other navigators. The set of the current from the antarctic regions has been illustrated by the voyages of Cook, La Pérouse, Flinders, and others. The equatorial current becomes very apparent in the Atlantic and in the Pacific, between the parallels of 30° on each side of the equator. According to the best observers, it has a mean velocity of nine or ten miles per day in the open sea. In the Atlantic it produces that remarkable movement of the waters termed the gulf-stream. This current passes between the West Indian islands, and sweeps along the northern coasts of South America, from Cumana to the isthmus of Darien; from thence it stretches toward Cape Catoche in Yucatan, and, after wheeling through the Mexican Gulf, it rushes through the channel between Cuba and Florida, passing between the Bahamas and the United States of North America towards Newfoundland. In the latter part of its course it can be distinctly traced, not only by its effect on the dead reckoning of ships, but by the superior temperature of its waters, which betray its tropical origin, and by its greater depth; for it appears in a long succession of ages to have hollowed out for itself a submarine channel. In the Bahama Channel it has a breadth of fifteen leagues, and its waters have, in latitude 42°, a temperature of 71° Fahrenheit, whilst that of the adjacent sea beyond its influence is only 63°. The gulf-stream flows with diminished velocity as it proceeds northward, until it is lost on the banks of Newfoundland, where it encounters the great current from the arctic regions. This vast sub-marine bank, indeed, probably owes its origin to the deposition of the solid particles borne along by both currents, at the point where their opposing forces meet; and there can be little

doubt that the eastern movement of the waters of the Northern Atlantic is the resultant of the compound motion of these mighty oceanic streams. It has been computed from the average velocity of both, that the waters of the Atlantic, between the parallels of $10^\{\backslash circ\}$ and $43^\{\backslash circ\}$ north, perform a circuit of about 8800 nautical leagues in thirty-four months. The retrograde current preserves a general easterly direction after leaving Newfoundland; but it reaches considerably farther north, for its influence frequently throws on our western shores the trees and ligneous fruits of the Antilles; but the great body of it is deflected to the southward at the Azores, partly by the influence of the polar current from the north; and then sending a branch to enter the Straits of Gibraltar, gives rise to that dangerous current which has been described by Major Rennell as the cause of the frequent shipwrecks on the African coast within the Canary Islands, when the mariner, by reckoning, considers himself as far to the westward of Teneriffe.

Another branch of the great western current of the Atlantic appears to be deflected by the peculiar form of South America, from the latitude of Pernambuco, along the eastern coasts to Cape Horn; and, doubling that promontory, it finds its way into the Pacific. There is also a general western current in the Great Pacific. Ships are carried by this with great celerity from Acapulco to the Philippine Islands; but, in returning, it is found advisable to sail northwards, for the benefit of the variable winds and of the polar currents. In the Southern Pacific, the polar currents, being little interrupted by land, proceed with less deviation from their course than those in the northern hemisphere; and hence they carry icebergs nearer to the tropical regions than we ever find to the north of the equator. The western motion of the waters of the Pacific is interrupted and broken by the vast chain of islands, shoals, and sub-marine banks which stretch from China to New Zealand. The general direction is changed or modified by the form of these lands; and the vast mass of New Holland is one cause of those dangerous currents around its shores which were noticed by Cook, La Pérouse, and Flinders. The weakened force of the general current in the Indian seas permits its interruption by the variable direction of the monsoons; but the current which always sets from Australia towards the Bay of Bengal seems to be produced by the modifying influence of the south polar current on the general equatorial current when weakened by the interruption of islands. The direction of this current, finding obstacles to its progress in a north-west direction, and being resisted by the banks and islands which stretch between Cape Comorin and Madagascar, is deflected to the south-west, forming the current in the Mozambique Channel, and thence sweeping round the Cape of Good Hope in a powerful stream, not less than forty-three leagues in breadth. The entrance into this current may be ascertained by the sudden increase in the temperature of the water to eight degrees beyond that of the adjacent ocean. It is probably to this current that the vast shoal off the Cape, known by the name of the Aguillas Bank, is to be attributed. It appears to round off upon both sides, forming a bank forty leagues in breadth, and 160 leagues from east to west.

The strong current through Bass's Straits, separating New Holland from Van Diemen's Land, is but a part of the general western current; but in its passage through the intricate channels between the islands, reefs, and shoals of the ocean north of Australia, it is divided and modified in a thousand different ways. There is another current less easily accounted for. On the western coast of Africa, about thirty leagues from the land, a strong current appears to set eastward in the Gulf of Guinea, and often draws ships far off their southern course. It seems to be a species of eddy, produced by resistance to the polar current caused by the sudden contraction of the Atlantic from the form of

the opposite coasts of America and Africa. These currents may be considered as general; but there are other currents which, though special, are remarkably constant. Of these we may notice the perpetual current which sets into the Mediterranean through the Straits of Gibraltar. It occupies the middle of the Straits. On either shore a less powerful and less extensive current alternates with the tides; but there is a constant and powerful influx from the Atlantic into the Mediterranean. A perpetual current sets out of the Baltic to the Northern Ocean; and here also smaller counter-currents are found on the coasts of Sweden and Jutland. Another considerable local current perpetually sets from the Black Sea in the Mediterranean. The causes of these peculiarities we shall afterwards explain.

Besides the motions already described, the waters of the ocean are subject to alternations in the level of its surface in any particular place several times a day. The rising of the water is termed flood, the fall is named ebb-tide. The connection between the tides and the moon were long observed before any theory was proposed to account for them. Descartes ascribed the flood to the pressure of the moon on the earth's atmosphere; but Kepler first suggested that it might arise from the attraction exerted by the moon; and this theory was confirmed and enlarged by Newton, who showed how well the general phenomena might be explained on the principle of universal gravitation. The theory and the phenomena are given under the article ASTRONOMY; to which we refer for the explanation of the two floods and two ebbs daily, how the united influence of the sun and moon causes the two spring and two neap-tides during each lunation; and why the highest tides of all take place about the equinoxes, when the sun and moon are on the equator, and in perigee.

The phenomena of the tides accord well with the theory in the open sea; but in the vicinity of land, amongst numerous islands, or in deep inland basins, their regularity is affected by various causes, as the occurrence of strong gales in particular directions, the form of points of land, of banks, and shoals, and the mechanical influence of narrow channels. Thus the tides of the German Ocean take twelve hours to reach London bridge; and consequently the time of high-water there is just as a new flood is commencing in that sea. The usual rise of the tides in mid-ocean is about three or four feet; but on the coasts of Britain, especially in the mouths of rivers, the attraction of the land raises them far higher. Thus, at Chepstow, the tide rises from forty-five to sixty feet; and it is recorded that once, from the influence of a strong gale blowing at the same time into Bristol Channel, the rise was seventy feet. In the river of Annapolis, which falls into the bay of Fundy, on the coast of America, the spring-tides have been known to reach the surprising height of 120 feet. In such cases the flood-tide often rushes in with a vast mass, like a huge wave, several feet above the general surface, which is termed a bore, and is dangerous to ships that may have been stranded on the receding of the tide. Something similar is often seen in our smaller rivers, and on such extensive flat sands as those in Morecambe Bay on the coast of Lancashire, and in the Solway Frith. In narrow straits the tide runs with great velocity, such as no ship can stem even with the most favourable breeze. In the embouchure of the Mersey, the ebb-tide has a rate of five or six miles an hour. In the Pentland Frith, between Caithness and the Orkney Islands, the spring-tides have a velocity of nine miles an hour.

Whirlpools appear to be occasioned by currents meeting with sub-marine obstacles, which throw them into gyration. When the movement is rapid, the centre is the most depressed portion of the rotating circle, and objects drawn within it are submerged in that point. Several small whirlpools, capable of whirling round a boat, are seen among the Orkney Islands. That of Coryvreechan, in the narrow channel between Scarba and Jura, in the Western Islands, is

Physical Geography. caused by a rock of a conical form rising abruptly from the bottom, where the depth is 600 feet, and reaching to within ninety feet of the surface. This obstruction, in a tortuous rocky channel, causes a succession of eddies; and when the flood-tide sets in, with a fresh breeze in the opposite direction, the eddying waters rise in short heavy waves, which are highly dangerous to boats, and even to decked vessels.

The Malström on the coast of Norway, near the island of Moscoe, is a whirlpool of a similar kind, the perils of which are probably much exaggerated. The flood-tide setting from the south-west amongst the Laffoden Isles, especially when it meets with a strong gale from the north-west, produces a great agitation of the waves, and a whirlpool is formed, the roaring of which is heard at the distance of many miles. Its agitated vortices are dangerous to vessels, and it is said that seals and whales, when caught within its eddies, are unable to extricate themselves from destruction.

It is now well ascertained that Charybdis, in the Strait of Messina, owes its terrors to the imagination of seamen in the infancy of navigation, and all its celebrity to poetic fancy.

Colour of the sea. The sea, when viewed near land, has its colour affected by the nature of the bottom. Where the bottom is sandy, and the depth not very considerable, it has a greenish tint; and this is the general hue of the German Ocean; but when viewed out of soundings, the colour of the ocean is generally of a beautiful deep blue. This may be considered as the natural hue of a deep column of water, from which only the most refrangible colours reach the eye of the spectator placed on a ship's deck; and the greenish tint is produced by the mingling of this hue with the yellowish tint of the sandy bottom. Sometimes partial colours are observed in the ocean. Captain Cook and other navigators describe a brown colour of the sea, produced by myriads of minute mollusca and crustacea. In the Greenland seas, the waters in which the whales chiefly delight are streaked with green. This colour appears in broad streams in the sea; and, when examined by Scoresby, was found to be produced by innumerable medusæ and minute crustacea, the olive-green water containing sixty-four in one cubic inch. Captain Horsburgh mentions a milky appearance of the sea, observed in a passage from China to Australia, which was occasioned by minute globular bodies linked together, and doubtless forming some species of beroe or medusa. The occasional red colour of the sea on the coast of Brazil and of China is believed to proceed from minute mollusca which float in myriads in the waters at certain seasons of the year. Ehrenberg asserts that the name of Red Sea has been given to that gulf from the prevalence of a species of oscillatoria.

Phosphorescence of the sea. The phosphorescence of the sea is a very brilliant appearance which the ocean and arms of the sea frequently assume. It is best seen when a vessel is in brisk motion through the waves, or when a boat is impelled by oars through a calm sea. We have seen this appearance so brilliant that we could distinguish letters in a book by its phosphorescence in the wake of a ship, where it will sometimes form a line of pale lambent light several hundred yards in length. We have also seen from the deck, by its light, the hour on a watch, when in other situations on the deck we could not distinguish the features of the seamen. When the hand is immersed in luminous water, we have found the shining particles adhering to the fingers; and on examining such water in a globular glass vessel by transmitted light, we scarcely ever failed to observe in it several minute animals of the genera medusa and beroe. These animals appeared to have the power of emitting light by their own efforts, as the water, when left at rest, shone by fits. On counting their number in a glass of water, and suddenly dashing it on the deck, in a dark night, it was observed that the number of luminous points generally coincided with that of the animals in the water, and never exceeded it. We have found these animals in sea-water, from 60° to 37° north latitude,

in the Atlantic, and in innumerable bays; and believe that they are very general in the ocean. Physical Geography.

The dependence of the phosphorescence of the sea on living animals was long ago asserted by Captain Cook and Sir Joseph Banks. In the voyage from the Canaries to Rio de Janeiro they drew a net through luminous water, and caught a large medusa, which shone brilliantly when taken on board; they obtained from it also three minute crustacea, about one tenth the size of the glow-worm, which also were very luminous in the dark. Banks named the first Medusa pellucens, and the most brilliant of the second Cancer fulgens. Other naturalists have enlarged the list of marine luminous animals, which includes occasional instances amongst the larger fishes.

A considerable number of experiments have been made on the temperature of the ocean by Ellis, Piron, Flinders, Parry, Ross, and others; but the results are not uniform. Ellis found in the Atlantic that the temperature remained stationary at 50° at depths a little exceeding a mile. Piron, who had sounded to greater depths, found it to decrease with the depth. The mean temperature of the ocean at the surface, and at a distance from land, is higher than that of the atmosphere with which it is in contact; but it is generally lower than the air at mid-day, and higher than the atmosphere at midnight. The ocean over a bank is colder than where it is of great depth. This is ascribed to cold currents being forced upwards by the obstruction; but whatever be the cause, the fact is important, as it may enable a seaman to recognise his approach to land when the atmosphere is obscure and his reckoning doubtful, as was first remarked by Mr Williams.

The Southern Ocean, receiving less fresh water than the Atlantic, is rather saltier, as is also the Mediterranean; but the Baltic, White and Black Seas, from the size of the rivers they receive, are fresher. According to Marcet their specific gravities are as follows:

Sea-water has a mean specific gravity of 1.0280 to distilled water as 1.0000. The late Dr John Murray published an admirable analysis of sea-water, 10,000 parts of which afforded him

Dr Murray detected no trace of muriate of lime, which Lavoisier has stated as high as 20.38, and Pfaff at 31.25. In reflecting upon these discrepancies, he was led to conclude that several of the products obtained may be the results of the mutual decomposition of the salts during the processes for separating them from the water; thus the sulphates of magnesia and of lime would be decomposed by muriate of soda, and when both are obtained from sea-water, we should regard them as results of decomposition. The real ingredients of sea-water would thus be,

This doctrine he has happily applied to explain the activity of several mineral waters, the analysis of which only affords inert ingredients.

The saline contents of the ocean do not appear to vary considerably; they, however, increase as we approach to the equator, from $\backslash frac\{1\}\{2\}$ to $\backslash frac\{1\}\{3\}$, as is proved by the increase of density. The numerous experiments of Bladh of Sweden proved that the specific gravity of the waters of the Atlantic on both sides of the equator increase as they recede from the poles. This table was republished by Kirwan in the Geological Essays, with a reduction of the specific gravity to temperature 62° Fahrenheit. The longitudes of Bladh are from the meridian of Teneriffe.

Dr Traill has published the results of some experiments on water taken up in different parts of the Atlantic, from 53° to 6° N., which also show a progressive increase in the specific gravity of the ocean towards the equator, and likewise as the depth increases.

This last was apparently affected by the waters of the Orinoco, though out of sight of land.

The phenomena of sea-ice have been admirably described by Scoresby, who divides it into two kinds, field and mountain ice. Field-ice has a thickness of from two to three or four fathoms. He mentions single fields 100 miles in length and fifty in medial breadth, with a general elevation above the water of from four to six feet. When the ice is of smaller extent it is termed a floe. Mr Scoresby found the specific

gravity of ice to sea-water to be as 8 to 8.97, or nearly as 8 to 9, so that ice floats in sea-water with $\backslash frac\{1\}\{3\}$th above water; and a field which has six feet above water will have forty-eight feet below it, if the ice be of uniform thickness. When a field or floe is exposed to the dashing of the waves, especially to what is called a ground-swell, it is speedily broken up; but the pieces are apt to collect into what is termed a pack, when from the mast-head no opening can be discovered in it. In smaller collections it forms patches. The ice is loose or open when ships can pass between the fragments of floes or packs. By the collision of pieces of ice, the fragments are often heaped on the larger portions, and form hummocks, which are thus sometimes raised thirty feet above the general level of the ice. Some ocean-ice is whitish and porous. This affords only a brackish water; for though freezing separates the salt from the ice, the salt water is retained in its pores, and the water yielded is not potable; but ice which looks blackish whilst swimming has a solid consistence, a beautiful beryl-green hue in the air, and when melted affords a potable water, a fact of great importance to navigators in high latitudes. Sea-ice assumes different appearances, according as it is formed in smooth or in rough water. The first process in freezing is the formation of minute prisms; when these form in smooth water, they speedily unite into sheets of ice, which in twenty-four hours of keen frost will attain a thickness of two or three inches, and in forty-eight hours will be able to bear the weight of a man. This sort often forms in the interstices of fields, and is termed bay-ice. It sometimes increases to three feet in thickness. When the congelation takes place in rougher water, the crystals of ice accumulate, rendering the water sluggish; and this is termed by sailors sludge. When the motion of the waves prevents the concretions into larger masses, it forms what sailors term pancakes, which generally attain the thickness of a foot. Both these sorts of ice are termed young ice, to distinguish them from the more solid field-ice. They often rapidly form around ships engaged in the whale fishery, till no open water can be seen from the mast-head; and they often disappear, from the effects of gales or currents, with an almost magical celerity. The heaviest ice is liable to disruption by the waves, especially when detached from the shores, and when the fields are carried by currents into the open sea. Through the openings thus produced, Scoresby penetrated in one voyage as far north as 81° 30' N. between Spitzbergen and Greenland; and at another time reached the long-lost eastern coast of Greenland, which he surveyed to the extent of 800 miles. The drifting of fields of ice is a proof of the existence of currents from the circumpolar regions. Scoresby mentions that he has seen a field drift 100 miles to the south in a month, although the wind had been very variable during that time. When ice-fields drift they often acquire a rotatory motion, which brings them into collision with other fields; the crushing of the weaker ice is attended with an awful noise; and the utmost presence of mind in the mariner is often required to extricate his vessel from the encounter of such irresistible forces. Mountain-ice or Iceberg is the term applied to those enormous masses of ice that are more common in Baffin's Bay than in any other part of the northern seas, and are also very frequent in the Great Southern Ocean. It is now generally believed that they are glaciers originally formed on the cliffs of lofty islands or in abrupt valleys, where their lower part being undermined by the waves, in process of time they thus become detached from the land, and float about, as impelled by currents. Some of the icebergs seen in Davis' Straits have an area of five or six square miles, rise 100 feet above the surface, and are aground in water of 100 fathoms; from which data we calculate, that such enormous masses weigh more than two thousand millions of tons. Icebergs assume most fantastic shapes, as may be seen in Cap-

tain Ross's and Captain Parry's Voyages. Large icebergs can only be produced where there is deep water close to the cliffs on which they are formed; and this is probably the reason why they abound in Davis' Straits. In the Southern Ocean icebergs have been said occasionally to be 200 feet above the surface, and Forster describes them as composed of layers of different coloured ice. It is dangerous to approach some icebergs, still more to make fast to them; for sometimes they suddenly split into a thousand pieces, with a thundering explosion, as if they were of unannealed glass; at other times they suddenly lose their equilibrium, and are overturned by the wasting of the submerged parts, whilst the upper portion has lost little of its weight. Icebergs have been wafted by currents and storms far from their native seats. Occasionally they have been encountered in the Atlantic, in latitude 40° north; but in the southern hemisphere they have been frequently seen as low as 35°. The Ice-blink is one of those singular effects of extraordinary refraction, by which the existence of open water is often recognised, though it really be beyond the limits of direct vision.¹

The polar ices appear to be subject to periodical increase and diminution, depending on causes not yet perfectly understood. One of the last and most remarkable disruptions of the arctic ice is that described by Scoresby. That able philosopher has traced the usual boundary of the firm ice in close and open seasons; and during one of those periodical disruptions, in which more than two thousand square leagues of ice had broken up, he was enabled to penetrate through the boundary, which had seemingly remained undisturbed for four centuries. Large, however, as this quantity disrupted seems, it is a very small part of the whole polar ices. Leslie computed that a total eclipse of the sun for a single hour deprives our planet of as much heat as would be capable of melting a circle of ice 500 miles in diameter and 150 feet in thickness, which he estimates to be sixty times the extent of what is annually melted in the Greenland Seas; yet the effect of a total eclipse of the sun does not cool the atmosphere more than 3° of the thermometer. We cannot, then, ascribe any important influence on our climate to the melting of this portion of the polar ices.

2. INLAND SEAS.—A few of the inland seas yet remain to be noticed. We begin with the Baltic. This sea differs from the arms of the ocean already mentioned, by its comparative freshness, by its want of tides, and by its sending a current continually into the German Sea. Its freshness arises from the number and size of the rivers it receives, in proportion to its extent. It is a narrow, irregular sea, the extreme length of which is about 1200 miles, with a mean breadth of about 140 miles. This will not give an area of 70,000 square miles, and its mean depth is only about sixty fathoms. Yet it receives forty considerable rivers, and the waters of innumerable lakes, which are the drainage of a country five times as extensive as its whole surface. The immense influx of fresh water into such a small basin, from which evaporation is but slow for a large portion of the year, sufficiently accounts both for its deficiency of salt and for the out-going current. The want of regular tides in the Baltic is a consequence of the same causes that operate in all confined collections of water having narrow communications with the ocean, viz. that the moon must act almost equally on all parts of its surface at the same time, and the oceanic tidal-wave, obstructed by the sinuities of the narrow channel, only penetrates a little way into the three passages which lead into the Baltic.

The Mediterranean is the noblest and most remarkable inland sea on the face of our earth. A strong current is perpetually setting into the Mediterranean through the

Straits of Gibraltar, especially in the centre; and although there be an eddy on each side, this reflux is far inferior to the quantity flowing in from the Atlantic. This inland sea, which is 2350 miles in length, with a breadth varying from 100 to 650 miles, covers an area of about 1,000,000 of square miles. Its depth is much more considerable near the Straits than that of the adjacent Atlantic. The fine hydrographical chart of Don Vincente Tofino de San Miguel, republished in this country with additions by Captain W. H. Smyth, R. N., shows, that from the longitude of Cape Trafalgar, where the Atlantic is from fifty to sixty fathoms, the channel deepens as we proceed upwards, until, a little within Gibraltar, it is a thousand fathoms deep, and soon after is beyond soundings. It is also very deep towards its eastern extremity. Between Sicily and Africa its depth varies from a hundred to thirty fathoms; so that there exists in that part of the sea a submarine chain or shoal, which may be considered as an extension of the rock-formations of Italy and Sicily, with deep water on either hand. Besides the perpetual current from the Atlantic, it receives the surplus waters of the Euxine, by the constant stream passing through the strait of the Dardanelles. These immense additions appear necessary to supply the enormous evaporation from its surface. It has been remarked, that the rivers which fall into it are the contributions of a smaller extent of territory than the river-domain of any other inland sea, in proportion to its area; for they are to each other rather less than one and a half to one. Some have contended that there is a current below that setting in from the Atlantic. The only fact offered in proof is stated to have happened in 1712. A Dutch merchantman, laden with wine and brandy, was sunk by a broadside from a French privateer in mid-channel, and a few days afterwards the ship and cargo were cast on shore near Cape Spartel, twelve miles below where the accident happened; but the nature of the cargo probably prevented the complete sinking of the vessel, and she may have drifted within the range of the outgoing current on the African shore. The eastward slope of the bottom of the strait would certainly favour the ingress of the cooler waters of the Atlantic into the Mediterranean, and afford an argument against the existence of this alleged under-current.

There is, however, no need of the hypothesis of a counter-current to explain the preservation of the level of this sea, notwithstanding that it receives several large rivers, and a perpetual supply from the Euxine and the ocean; since Dr. Halley has shown, that the amount of evaporation from its extensive surface, in so warm a latitude, is sufficient to preserve its level unchanged.

Tides are but little felt in the Mediterranean. They are only sensible in certain parts of that sea, and seldom rise to six inches above the mean level; but, in the Straits of Messina, the form of the land, and the peculiar set of the currents, give them a more considerable range. These tides, however, are irregular, and are liable to be affected greatly by the direction and force of the winds. The want of considerable tides in the Mediterranean kept the Romans ignorant of the extent of ebbs and floods, and caused them several disasters when they carried their arms to the shores of Western Europe. Even in the time of Tacitus they were imperfectly acquainted with the currents produced by the tides around the northern coasts of Scotland, the difficulty of stemming which the historian ascribes to a "mare pigrum, remigantibus infestissimum."

That large inland sea the Euxine, with its branch the Sea of Azof, has somewhat of the characters of the North American lakes in its comparative freshness, which is one cause of the duration of ice on waters that have the same latitude as the finest regions of Italy and France. The sur-

¹ Scoresby's Arctic Regions, and Ross's and Parry's Voyages

face of the Euxine and the Sea of Azof is estimated as equal to 170,000 square miles; but as they receive some of the largest rivers in Europe, and smaller streams from Asia, the drainage of an extent of country equal to five times their own surface, we see a strong reason why there should be a constant efflux through the Bosphorus. This current in the Thracian Bosphorus has generally a velocity of from three to five miles an hour, according to the direction and force of the wind. Before this outlet was formed, the Euxine probably united with the Caspian, and constituted an hyperborean sea on the plains of Russia, and the provinces on the Lower Danube; and most probably the floods of Deucalion and Ogyges were connected with débâcles arising from disruptions of the barrier of this sea.

According to Pallas, there are traces of the former union of the Caspian with the Euxine, in the saline marshy ground and numerous lakes which intersperse the flat countries of Ulagann-Ternik, Alabuga, and Byeloe. The present Caspian is below the level of the Euxine. From barometrical measurements, Parrot stated its surface to be 300 feet below the level of that sea; but the recent trigonometrical levelling of Fuss, Sabler, and Sawitch, which has been conducted with the utmost care, reduces this depression to 101.2 feet. Its length is about 700 miles, and its breadth about 210, which would give it an area of 147,000 square miles. It has a very unequal bottom. Hanway states, that in some places 450 fathoms of line could not reach the bottom; but in other places it abounds in shallows, which render its navigation perilous, independently of the sudden storms to which it is liable. It is without tides; but its shores are subject to irregular floods, from the size and number of its rivers, and the direction of the winds. It is extremely salt, and abounds with fish; amongst which the most important is the isinglass sturgeon, and the sterlet (Acipenser huso, and A. Ruthenus). Several species of Salmo and Cyprinus, and a species of herring, are numerous; and there are extensive fisheries of several kinds of seals. How seals found their way into this inland sea, unless at one time it communicated with the ocean, we cannot easily explain.

The Sea of Aral seems once to have communicated with the Caspian, from the similarity of their fishes, and the numerous salt lakes and marshes that still crowd the intervening space, especially towards the northern extremity. The Sea of Aral is 200 miles by 100, and therefore has an area of about 20,000 square miles. It is extremely salt, though it receives several very large rivers, particularly the Sion and Gihon, the ancient Jaxartes and Oxus.

The inland sea or lake called Baikal differs from those already mentioned in being perfectly fresh. Its length is 560 miles, and its breadth forty; so that its area is about 14,400 square miles, including the islands which it contains. This lake or sea is remarkable for its green tint, and the very great inequalities of its bottom. In some parts its depth is asserted to be more than 3000 feet, but then the water suddenly shoals to 150 feet. Considering the mass of water it receives from the rivers Bargazin, Selinga, Angara, and many less considerable tributaries, and that its only visible outlet is the Lower Angara, which is not capable of discharging above a tenth of the water poured into this sea, it is conceived by some geographers, that evaporation, in a climate where the whole surface is annually frozen for five months, could not carry off the surplus, and that therefore the Baikal must have some secret outlet, by which its superfluous waters may reach the ocean. The same difficulty occurs here as with the Caspian and the Aral. How could the seals which it contains have been introduced? They are extremely numerous in its northern part, and as many as two thousand have been killed in one year. Of the fishes, the omul, or Salmo migratorius, is the most important in an economical point of view; but it also con-

tains S. salelinus, S. oxyrhinchus, the carp, the tench, the sterlet, and the isinglass sturgeon. One fish is peculiar to this sea, Calyonymus Baikalensis, which occurs in such quantities as to corrupt the air by its putrefaction on the shores after storms, to which this sea is very liable. There are many appearances of volcanic fire in the vicinity of the Baikal. It is probable that the surface of the waters of this inland sea once stood on a higher level than at present, and that thus the salmon and the seal might have been able to surmount the rapids of the Yenisei, or have found their way into the Baikal by the channel of the Lena.

The Red Sea is the only other Asiatic sea which we shall here notice particularly. It is a long arm of the ocean, extending between Asia and Africa for nearly 1500 miles, whilst its medial breadth is not above 100 miles. Until Lord Valentia's voyage, our charts of this sea were very imperfect, and much still remains to be done; but his lordship has established the important point of the superior depth of the western side, although the general depth is inconsiderable. These peculiarities of its form give it more the character of a river than of an arm of the ocean; a circumstance which, with the position of its strait in the direction of the tidal wave from the Arabian Sea, makes it subject to very considerable tides, especially in its lower part; whilst their rise is not prevented by the influx of any river. The less powerful influence of shape, and the influx of the united streams of the Euphrates and Tigris, tend to diminish the height of the tides in the Persian Gulf, another inland sea, about half the length of the Red Sea.

The Tchad, an inland sea of fresh water, is one of the most considerable of the kind in the world; and constitutes that Bahr el Soudan, so long considered as the common receptacle of the rivers of Central Africa. The discoveries of Denham and Clapperton prove that it is one of the grandest features in the kingdom of Bornou. It seems to be not less than 200 miles in length by 150 in breadth; but in the rainy season it becomes greatly enlarged, and encroaches on its flat shores. It is principally fed by the rivers Shary and Ycou. The former flows from the south into the lake, and is a noble, rapid river, half a mile in width. The latter, once supposed to be the Niger, comes from the west, but is not considerable.

We have already noticed the general direction of the Mexican equinoctial current into the Gulf of Mexico, from which it has obtained the name of the Gulf Stream. The eddying of this current in the great inland basin, the prevalence of the trade-winds, and the enormous quantity of mud brought down by the Mississippi and the less important tributary streams descending from the Mexican Sierra Madre, are, according to Humboldt, rapidly filling up the gulf. The eastern coast of Mexico, from 18° to 26° north latitude, abounds with bars, over which large vessels cannot pass; and, far inland, that distinguished traveller found banks filled with sea-shells but little changed. Were the Isthmus of Darien annihilated, the equinoctial stream would pass round the globe; and this change would probably, ere long, place the feeble governments of Japan and China at the mercy of some European conqueror.

Hudson's Sea, improperly called a bay, is during a great part of the year encumbered with ice. This basin is almost three times larger than the Baltic, comprehending an area of about 500,000 square miles. It is inhabited by some kinds of cetacea, especially the Greenland whale and the beluga.

Baffin's Sea is now found to communicate with the polar sea by several openings. Its extent, if we include Davis's Straits to the latitude of Cape Farewell, is only inferior, as an inland sea, to the Mediterranean. It is locked up for a considerable period of the year in impenetrable ice; but in the summer months the ice is broken up, and partially carried off by a strong current, which is continually setting down

Physical Geography through the strait. The centre of that sea, however, is occupied either by a series of rocky islands, or by shoals, on which vast icebergs are perpetually grounded. Its desolate shores have been much frequented by our whale fishers, on account of the multitudes of the balæna mysticetus that resort to this inlet.

Buffon has justly termed North America the country of lakes or of fresh-water seas. Commencing in the north with Great Bear Lake, we may trace a connection by means of minor lakes to Slave Lake, a vast sea of fresh water, about 200 miles in length and 100 in breadth. This last lake communicates with the Arctic Ocean by Mackenzie's River, and with Alhabsca Lake by Slave River. A chain of smaller lakes conducts thence to Winnipeg, which is scarcely inferior to Slave Lake in extent of surface. Advancing to the south-east, we arrive at Lake Superior, which, with the Michigan and the Huron, may be considered as forming the great inland sea of Canada. Lake Superior has an extreme length of 380 miles, a breadth of 161, and a circumference of 1525 miles; the coasts are rocky, and its depth, in some points, is 1200 feet, but its mean depth may be about 900 feet; and its surface is 641 feet above the sea. The Huron communicates with the Michigan by a channel four miles in length; it is 250 miles in length, varies from sixty to 180 miles in breadth, and has a circumference of 1100 miles. The Michigan is 262 miles in length by 55 in breadth, and has a circumference of 915 miles. The mean depth of the three lakes is 900 feet, which shows that their bottom is 300 feet below the level of the ocean. Their whole surface is computed at 72,930 square miles. We have already pointed out the danger to the whole valley of the Mississippi, should earthquakes ever burst the rocky barrier at the southern extremity of the Michigan.

Lake Erie is connected with the Huron by a narrow channel, which expands into Lake St Clair, and again contracts ere it reaches the lower lake. Erie is 230 miles in length, about sixty in breadth where it is widest, and about thirty-five at the mean. It differs from the upper lakes by its comparative shallowness, the mean depth not exceeding 120 feet. From many observations, it would appear that Lake Erie is rapidly filling up. Long Point, at the Big Creek, is stated to have advanced three miles in three years, in consequence of the alluvion carried into the lake by its numerous rivers. The connection of this lake with that of Ontario is by the river Niagara. The difference in level between the lakes is 330 feet; and, after the river has passed down twenty-one miles, it suddenly precipitates itself over the celebrated cataracts of Niagara, in a fall with a depth of 160 feet. It then flows for seven miles through a deep rocky chasm to Queenstown, where the country suddenly sinks; and it flows seven miles farther, through a champaign country, to Lake Ontario, another fresh-water sea, 171 miles in length and about sixty in breadth. The lakes above the falls abound with many kinds of fish, as salmon, sturgeons, several species of cyprinus, &c. The five last-mentioned seas or lakes are computed to have an area of 70,000 or 80,000 square miles.

3. LAKES.—Several of the large collections of water already mentioned have been considered as lakes; but that designation properly belongs to those less considerable bodies of water which we are now to notice. Many lakes are salt; and of such, a considerable number exist in Asia. The most celebrated salt lake in that quarter is the

Dead Sea, or Lake Asphaltites. Independently of the account in the sacred Scriptures relating to the destruction of the cities of the plain the site of which is now occupied by this lake, the nature of the surrounding country bears evident traces of the ravages of fire in remote ages. The frowning cliffs on its eastern shore are evidently volcanic, thermal waters occur near them, the country around is often to this day shaken by earthquakes, and asphaltum is yet col-

lected on its shores. Some have supposed it the crater of a volcano. The lake is much more saline than the ocean; and although the waters are limpid, their specific gravity is 1.2110. Mr Gordon attests their great buoyancy, and an analysis of the waters by Dr Marcet shows that they contain one fourth of their weight of salts. An hundred parts of water yielded,

This saltiness of the Dead Sea is evidently owing to the abundance of salt in its vicinity, which all the streams falling into it perpetually deposit in this lake. Although the waters of the Jordan, the most considerable of these, are nearly tasteless, and do not contain above one three-hundredth of their weight of salts, the latter are of the same kinds as are obtained from the waters of the Dead Sea; and it is easy to imagine how this constant supply of saline matter to a lake of narrow dimensions, without any visible outlet, must produce an accumulation of salt, which cannot pass off by evaporation. On the western shores of this lake are the ruins of considerable edifices, and still more ancient remains are said to be covered by its waters.

Messrs G. Moore and Beek have very lately proved, by the temperature at which water boils, that the surface of the Dead Sea is 500 feet lower than that of the Mediterranean; and the correctness of this estimate they confirmed by the barometrical results of Professor Schubert of Munich, which give 500 feet as the depression of the Lake of Tiberias, and 598 feet as that of the Dead Sea.

The no less celebrated Lake of Genesareth, or Sea of Genesareth, is still smaller. According to Ali Bey, it is only seven leagues in length by two in breadth. It is embosomed in lofty mountains of a volcanic character, amongst which Mount Thabor is pre-eminent. Ali Bey considers this lake as the crater of an ancient volcano; and the nature of the rocks around, and the thermal springs of Tiberias, favour the idea of the volcanic nature of the country.

The want of any outlet, and the influx of rivers passing through a very saline soil, are the causes of the saltiness of the considerable lakes of Van in Armenia, and of Ourmia in Persia. The former has a circumference of 168 miles; and its waters, though potable, are brackish. Ourmia, or Ourumia, is not less than 300 miles in circuit, and is probably the saltiest of any known lake. Its waters are stated to contain thirty-two per cent. of salt; and the sinking of the waters, during the dry season of the year, leaves a thick crust of salt on its shores, which at some points extends a considerable distance into the lake. There are many other lakes of considerable size in Central Asia; many of them are salt; but the most singular is that in Thibet, which produces tincal or borax, a substance which is imported in flat prismatic crystals into Europe; where it undergoes purification.

Lakes of America.—If North America be the country of American lakes, South America is remarkably deficient in them. That of Titicaca, in Upper Peru, already noticed, is the only considerable one in the southern division of the new world. It is true, that in the rainy season, the mighty rivers which water the plains of central South America inundate the low country to an immense extent, converting it into seas with a prodigious area, but of small depth; and it appears that inundations of this sort gave rise to the fable of the Lake of Parima, or El Dorado.

The Mexican lakes are more remarkable for their elevation of 7450 feet above the sea than for their extent. They are essentially saline, being more salt than the Baltic. The specific gravity of the waters of Tezucuo is 1.0215, and the impregnation is derived from a soil containing muriate

of soda and carbonate of soda. These lakes have diminished since the Spanish conquest, and the capital is no longer in the lake, but connected with it by a canal. The natural decrease of these lakes has been accelerated by the felling of trees, and by a tunnel driven, with immense labour through a mountain, to carry off the superfluous water that formerly often flooded part of the city. Between the inland seas of North America and the Icy Ocean there are innumerable lakes; a circumstance in which the northern portion of the new continent resembles the boreal regions of the old. A lake amongst the defiles of the Rocky Mountains, about the fiftieth degree of latitude, we were assured by the late Mr D. Douglas (a botanist, who several times crossed that chain, and was afterwards unfortunately killed by a wild bull in the Sandwich Islands), presents a remarkable peculiarity. From each extremity it discharges a considerable stream, one of which, falling into the Columbia, finds its way to the Pacific, whilst the other forms a branch of the Saskatchewan, and therefore may be said to fall into the Atlantic, through Lake Winnipeg, the inland seas of Canada, and the St Lawrence.

European Lakes.—Amongst these some of the most celebrated are the Lakes of Constance and Geneva. The first receives the Rhine, which was fabled to pass through it without mingling with its waters. Like other alpine lakes, it uniformly increases in summer, by the melting of the mountain snows; and its lowest period is in winter, when its sources are locked up in ice. Its length is thirty-five miles, its greatest breadth is eighteen miles, and its mean depth is 600 feet, but in some places it requires a line of 2130 feet to reach the bottom. Its waters thus fill a chasm about 1000 feet below the level of the ocean.

The Lake of Geneva has an area less than that of Constance; its extreme length being fifty-four miles, and its greatest width fifteen. Its depth is very variable; in some parts it was found by Deluc to be 960 feet, which shows that its bottom is only 200 feet higher than the level of the Mediterranean. Like the Lake of Constance, it receives a river, the Rhone, at one end, and discharges it at the other extremity, near to which the river precipitates itself into a subterranean channel termed Perte de Rhone. The beauty of the banks of this lake, and the prospect of the distant glaciers, have given to it great celebrity. The other Swiss lakes, though eminently beautiful, are not of sufficient importance to be described here.

Italian Lakes.—The beautiful scenery on the Lago di Como and the Lago Maggiore have been the theme of general praise; but some of the smaller Italian lakes, which seem to be the craters of extinct volcanoes, deserve notice. The lakes of Nemi and Averno are most evidently so; and the poetical representations of the sulphureous exhalations from the latter may possibly have had a real origin in former periods; but now birds rest with impunity on its surface, and fishes inhabit its waters.

British Lakes.—Our lakes are in general of small extent. The lakes of Westmoreland and Cumberland are more remarkable for their picturesque beauty than their size, and appear but as insignificant specks in a map of Europe. A few of the Scottish lakes are on a larger scale. Some of them are remarkable for their depth, and probably on that account Loch Tay and Loch Ness are never frozen. Loch Ness in some parts is 900 feet deep; so that its bottom is not less than ninety or a hundred feet below the bottom of the deepest part of the German Sea, on the eastern coast of Great Britain. The depth of Loch Tay varies from eighty to 600 feet. Loch Lomond is the largest of the Scottish lakes, being thirty miles in length by nine in its widest part, and at its northern extremity it is 600 feet deep. In the terrible earthquake at Lisbon, Loch Lomond repeatedly rose and fell several feet. The same took place in Loch Ness. A huge wave rolled with impetuosity from east to

west, and the waters sunk and fell repeatedly in an hour. Physical Geography. Loch Tay was remarkably agitated in 1783 and 1784, during the violent earthquakes in Calabria. On 22d September of the latter year, two large waves were observed to rush from each extremity of the lake, and coalesce near its centre, leaving about 100 yards of its banks at the eastern extremity momentarily dry; but the water returned and overflowed its usual boundary. The same appearances were repeated for several successive days with less violence; and the agitations of the loch did not cease till the 15th of October.

The agitation to which some alpine lakes are subject when the air is perfectly calm, in which even whole breakers have been observed, is a very obscure phenomenon. It has been often observed on the lakes in the north-west of England, in the Highlands of Scotland, and in the Lake of Geneva. It may possibly be explained by the compression of air in caverns existing in the adjacent mountains, the roofs of which are on a higher level than the surface, and communicating with the bottom of the lakes. If water much charged with air find its way into such caverns, the air separated may be compressed in their roofs, until its elasticity becoming superior to the pressure of the column of water, forces its way into the lake, and thus gives rise to what in England is termed a bottom wind.

The Irish lakes of Killarney and Lough Erne are highly picturesque; but the petrifying qualities of the waters of the more extensive Lough Neagh, in Antrim, ought rather to be attributed to the qualities of the soil in its vicinity. We possess specimens from that locality of wood converted into siliceous petrifications.

Periodic Lakes.—The inundations of rivers may be said to form periodic lakes; but there are other instances where the cause is more occult, although the effect be no less conspicuous. The best-known instance of this is the celebrated Lake of Cirknitz, in Carolina. It is only about eight miles in length by two in breadth. In the beginning of June its waters disappear through several fissures in its basin, and the peasants immediately begin to cultivate its bottom, or to pasture their herds on the rich herbage which its oozy bed soon produces. The crop is removed, and about the end of September or middle of October the waters return, spouting up through several apertures in the earth with great force. With these waters various species of fish appear; and sometimes is seen amongst them that curious animal, so puzzling to systematic naturalists, the Anguina Proteus of Laurenti, the Syren Anguina of Shaw. This lake is placed in a valley amongst limestone hills, which are well known to be hollowed out into vast caverns. Of these the most considerable is Mount Javornick. The caves are supposed to be the receptacles of subterranean rivers, which, when augmented by the autumnal rains, overflow into the channel which communicates with the basin of the Cirknitz; and when the sources of these floods fail, the waters retreat to their summer level. Other lakes of a similar kind are said to occur in Dalmatia.

4. Rivers.—The size and course of rivers are chiefly determined by the height and direction of the mountain chains in which they originate. Thus the Rhine, the Danube, and the Rhone, the largest rivers in Europe, take their rise in the Swiss Alps. The great rivers of Asia have their origin in some of its lofty central chains. The northern rivers, the Irtysh, Ob, Yenisei, and Lena, may all be traced to the Altai; the Ho-yang-Ho and Yang-tse Kiang of China arise in the mountains forming the eastern abutment of the central table-land of Asia; whilst the southern ramparts of that table-land contain the sources of the great river of Cambodia, the Irawaddy, the Brahmaputra, the Ganges, and the Scind or Indus. In Africa, the Nile has one of its sources in the lofty mountains of Abyssinia, and the other in the more distant central chain; the Niger rises in the western chain, runs eastward, is deflected in the kingdom of Houssa

Physical Geography. towards the south and west, and finally pours its waters into the Atlantic in the Gulf of Guinea. In America, the Magdalena and Marañon spring directly from the Cordilleras of the Andes, which also contain the principal sources of the Orinoco; whilst the more sluggish streams of the Paraguay and Parana derive their waters from a lower transverse chain, stretching to the coast of Brazil. In North America, the Arkansas, Red River, and Missouri, the great feeders of the Mississippi, spring from the Northern Andes, where also are the sources of the Columbia to the west, and of the Saskatchewan, the principal stream, which, flowing into Lake Winnipeg, becomes thus a part of that vast chain of lakes or seas, the outlet of which is the St Lawrence.

The channels of rivers are partly produced by the action of their own waters; but undoubtedly also by some of the causes which we have indicated as disturbing the original arrangement of the solid materials of the earth. Thus earthquakes have often opened passages for rivers through barriers of rocks and mountain chains. It may be admitted, that the continual action of water will excavate the hardest stone; but the ravines through which rivers flow frequently bear no traces of having been formed by the action of water; and the depth of some channels, compared to the scantiness of the stream, scarcely permits us to ascribe the whole to the excavation produced by the water. Thus, it is impossible to suppose that the small stream of the Lauricocha, in South America, has cut its way through the chain of the Andes, which it traverses in a long mural chasm, only twenty-five fathoms wide, but of astonishing sublimity. In other instances, the bottom of the channel itself is even below the level of the ocean; thus, the bed of the Marañon, above its junction with the Rio Negro, and more than 1000 miles from its embouchure, was found by Condamine to exceed 620 feet; which, by the estimate of Humboldt, is more than 100 feet below the level of the Atlantic. This may be a chasm produced by an earthquake, like that which, in 1752, suddenly opened, and received the whole waters of the Rio del Norte (a river which has a course of 1000 miles), so as to lay dry its channel for a distance of fifty leagues; and it was several weeks before the river assumed its usual level. Sometimes the channel of a river near its mouth is much below that of the adjacent sea. It is so in the Marañon; and the same has also been proved with regard to the Mississippi, which, just above New Orleans, is 153 feet deep, that is, at least fifty feet below the mean depth of that part of the Mexican Gulf. In such circumstances it is obvious, that the lower part of these hollows in river channels must be filled with salt water; which, by its superior gravity, will insinuate itself below the descending current. Indeed this takes place in the mouths of all rivers communicating with the sea by considerable channels of small declivity. Thus the lowest stratum of water at London Bridge, at the highest flood, is salt, though the upper current flowing over this stratum is quite potable. The tides in the Marañon are sensibly felt 600 miles from its mouth; and probably the sea-water forces itself up at least thus far beneath the mighty current of the river.

Rivers occupy the lowest parts of valleys through which they flow, and when they are very large their declivities are generally small. The surface of the Marañon, 3000 miles from the sea, has only an elevation of 1235 English feet, which would give less than five inches per mile for its mean declivity; but, from the point to which the tides of the ocean reach, the declivity is considered as not above 0.2 inch per mile. The Ganges, from Hurdwar, where it issues from the defiles of the Himalaya, has a mean declivity of four inches per mile; that of the Volga is about five inches.

The general form of the channels of rivers, however, especially in a champaign country, indicates, that if not wholly formed, they are greatly modified, by the action of their waters. In fact, their beds are usually proportional to the force of the stream. On the banks of many large rivers we can still trace the different heights at which the waters have formerly flowed. Playfair mentions the existence of four or five successive terraces on the Rhine,¹ each of which has evidently been formed in succession by the river; and hence he concludes that the Rhine once flowed 360 feet above its present level. Similar remarks were made on the Upper Rhone by Saussure. The general effect of the action of the water must be the erosion of the channels in the lapse of ages; and even in the rocky beds of rivers this action of water is very perceptible. Thus the waters of the Niagara have apparently worn away the limestone rock of the falls, and formed a deep ravine through the stony bed, six miles in length from Queenstown, where the cliffs terminate abruptly, to the present site of the Falls. The undermining of the cliffs by this stupendous cataract has caused a recession of the Falls equal to about eighteen feet in thirty years; but we cannot thence infer the length of time that this cataract has existed, because we have no certainty of the equality of the disintegration, nor of the other causes which may have aided or facilitated the process. When rivers which have their sources in lofty mountains flow through extensive plains, they are subject to periodical inundations at the rainy season, or when the snows which feed them are melted by the summer heat. Under these circumstances, the swollen river bears along with it mud, sand, and gravel. A portion of these is deposited on the banks of the river, and thus tends to elevate them above the level of the adjacent country; whilst a portion is swept to the sea by the current, and the materials are there collected into sand-banks, or, by the repellent action of the waves, serve to augment the triangular plain often formed at the mouths of rivers, to which the name of delta has been applied. This evident and gradual increase of a delta has given rise to the idea of attempting to estimate the antiquity of existing geological arrangements by such a natural chronometer. Thus, if the increase of a delta at the mouth of any river be a certain number of feet in a given time, it has been supposed that we could thence obtain a measure of the age of the whole alluvial land at that point. But for the accuracy of such deductions several data are required. We must ascertain that the supposed delta is really the formation of the river, and that the rate of increase at distant periods is uniform. In every attempt to solve such problems there is danger of mistaking the effects of general for local causes; and there is strong reason to disbelieve the alleged uniformity of the increase of the alluvial soil at the mouths of rivers. The Po, of late years, appears to have gained upon the Adriatic, at the rate of 228 feet per annum, at its two principal embouchures; but had the annual increase always equalled this, the mouth of the Po at the commencement of the Christian era should have been seventy-eight miles from the present shore, or extended beyond Mantua. But towns fifty miles nearer the Adriatic, in the valley of the Po, have existed from before the birth of Christ. The calculations of Volney, on the advance of the delta of the Nile, are founded on very insufficient grounds; and as yet it must be conceded, that we have no determinations of this sort worthy of confidence.

The currents of large rivers may frequently be traced, by their colour, to great distances in the ocean. Thus the Ganges, when in flood, discolours the sea in the Gulf of

¹ Works, vol. i.

Bengal to the distance of sixty miles, and its mud covers the bottom at eighty miles from the land. The mud of the Orinoco is carried far into the Atlantic; and on examining water drawn from the sea in latitude 8° 20' north, that is, in the parallel of the Orinoco, and 200 miles from its embouchure, Traill ascertained that the specific gravity of the water suddenly fell below what it was to the north and to the south of this parallel. The enormous mass of water discharged by the Marañon, according to Sabine, discolours the ocean, and has even considerable rapidity in a direction inclined to that of the equinoctial current, at the distance of 300 miles from the American coast. Fresh water may be obtained from the surface of this current, at a great distance from the shore.

The periodical inundations of rivers depend on great falls of rain in mountainous regions, or on the melting of snows in the neighbourhood of their source. The period depends on the return of these seasons in different places. Within the tropics the rainy season is usually about the time when the sun passes the meridian towards the tropic; and continues until his return to the same place. The floods of the Marañon and La Plata cover a vast extent of unexplored country. The rise of the Orinoco commences in May, its inundation begins in June, and the waters return to their channel in September; from which time they decrease until April of the succeeding year. The inundations of the Lower Mississippi begin in March, are at their height in June, and the river is lowest in October. The waters, at their height, 1000 miles from the sea, attain a rise of fifty feet; at 300 miles, of twenty-five feet; and at 100 miles, of twelve feet. The western bank of the Mississippi forms a vast series of lakes, which are dried up in autumn into arid plains, interspersed with swamps; but the delta below the junction of Red River is a dismal swamp, scarcely elevated above the sea. In the Red River, the Arkansas, and the Lower Mississippi, are found those enormous rafts of drift-wood, formed during the river floods, which sometimes extend for ten or twelve miles in one mass, rise and fall with the stream, yet have a luxuriant vegetation on their summits. The inundations of the Ganges follow a nearly similar course. The waters begin to increase in April, when the rains commence in the mountains in which it has its sources; the rate of its increase is about three inches daily until the month of July, when the rains descend in torrents in the plains. The increase of the river is then more rapid, being about five inches daily; and the country for 100 miles along its banks presents in the end of July the appearance of a vast lake, interspersed with insulated villages and woods. The general height of the inundation-waters in Bengal is about twelve feet, but in some places they have a depth of more than thirty feet. The great river of Ava, the Indus, the Euphrates, and the Tigris, have also their periods of inundation, depending on the circumstances determining the setting in of the rains on the mountains in which they originate.

The inundations of the Nile have long been celebrated. Shortly after the commencement of the rains on the mountains of Abyssinia, in June, the river begins to rise, and attains its greatest height in August. At Cairo the greatest rise is twenty-eight feet; at Rosetta it is no more than four about the middle of August, when the valley of Egypt, with a mean breadth of three or four leagues, and the greatest part of the delta, are covered with one sheet of water. Its increase is irregular; and it decreases gradually until the following May. As soon as the waters are within their usual channel, the soil, moistened and enriched by the sediment deposited from the inundation, is diligently cultivated by the natives.

By such inundations the banks of rivers are usually elevated above the adjacent plains; and when the stream carries down much mud and gravel, the whole bed of the river

may become considerably raised above the general level of the country through which it flows. This natural process has been considerably accelerated by art on the Po in Italy. To obviate the evil of sudden overflows of this river, its channel, and that of the tributary streams, have for many ages been carefully embanked; but the quantity of mud and gravel carried down by its waters is so great, that the bottom of the river is gradually raised, further embankments become necessary, and the repetition of these processes have elevated the whole bed of the river below its junction with the Mincio, until it is now actually thirty feet higher than the general level of the adjacent plains; a circumstance fraught with the utmost danger to the whole country between that point and the sea, should any accident happen to the main embankment. Where the evil is of less magnitude, the Italian engineers have long followed a remedy, devised, it is said, by Torricelli. When streams bring down much mud, and their channels have been suffered to become elevated above the level of the adjacent plains, the lower ground is subdivided by embanked enclosures; the turbid waters of the river are admitted by flood-gates into the subdivisions, and are there allowed to deposit their sediment. The addition of fertilizing mud thus made, in many situations amounts to four inches in depth by each irrigation, which may be repeated several times in the year. By such means, in five or six years, the soil is raised to a level with the river, and the risk of extensive inundations greatly diminished. The process is termed la colmata. It has been very advantageously applied in Tuscany and Lombardy; and if steadily and judiciously directed, under the auspices of a just and vigorous government, might possibly still avert a catastrophe which seems to impend over the districts of the Lower Po.

When river-courses lie amongst mountains, they are subject to sudden breaks, which, according to their depth, give rise either to rapids or to cataracts. There are many picturesque waterfalls in Scotland, as the Cascade of Fyers and that at Lochleven-head in Inverness-shire, Bruar Cascade in Athole, the Devon Lynn in Clackmannanshire, and the magnificent Falls of the Clyde near Lanark. The most stupendous cascade in Great Britain is the Fall of Glomach, in the district of Applecross, in the parish of Kintail, and county of Ross. At the head of a wild and solitary glen, seven miles from the inn of Shealhouse, the river Girsac is precipitated in one unbroken fall of more than three hundred feet. At the distance of about fifty feet from the bottom, the water impinges on a shelf, whence it falls into a dark pool, amidst naked perpendicular rocks. When the river is in flood, it descends in one unbroken sheet of above 380 feet in height. Wales and Cumberland can boast of many beautiful cascades on a smaller scale, particularly the Pistil y Cain and Pistil y Mouddach, in Merionethshire. The cataracts of the Dahl in Sweden, of the Staubach in the Alps, of Tivoli and Terni in Italy, the less considerable cascades of Mavraneria, the far-famed Styx of Homer, Hesiod, and Pausanias, in Arcadia, and the Fall of the Rhine at Schaufhausen, are amongst the most celebrated in Europe.

A cataract may exist in a comparatively flat country, from a sudden sinking in its level. The stupendous cataract of Niagara, in which a large and rapid river, 1650 feet in width, precipitates itself, by two channels, in one vast leap of 160 feet perpendicular, is an instance of such an occurrence. The celebrated cataract of Tequendama, near Bogota, though found by Humboldt to be far less lofty than had been supposed by Bouguer, is still of stupendous height, and one of the most magnificent in the world: a river half the breadth of the Seine at Paris is precipitated into a rocky gulf, at two bounds, to a depth of 540 feet.

The cataract of the river Shirawati, in the Indian province of Canara, exceeds in beauty and sublimity every

Physical Geography. waterfall which has been hitherto made known in Europe. The country around the village of Haliali, about three miles north-west of the fall, presents the richness of a tropical forest, mingled with cultivation. The traveller comes suddenly on the river. "A few steps more," says Dr Christie, "over huge blocks of granite, bring you to the brink of a fearful chasm, rocky, bare, and black, down into which you look to the depth of a thousand feet." The bed of the river is one fourth of a mile broad, in a direct line; but the edge of the fall is elliptical, with a sweep of about half a mile. This body of water rushes, at first for about three hundred feet, over a slope at an angle of 45°, in a sheet of white foam, and is then precipitated to the depth of eight hundred and fifty feet more, into a black abyss, with a thundering noise. It has therefore a depth of 1150 feet. In the rainy season the river appears to be about thirty feet in depth at the fall; in the dry season it is much lower, and is divided into three cascades of varied beauty and astonishing grandeur; but the smaller streams are almost dissipated in vapour before they reach the bottom. Dr Wilson succeeded in viewing this magnificent cataract from the bed of the river, from which it appears to much advantage.¹

Number of rivers. The number of considerable rivers which fall into the sea in different parts of the old continent amounts to 430, and those of the new continent may be estimated at 140, but this seeming disproportion is amply compensated by the vast dimensions of the latter.

Quantity of water they discharge. Several attempts have been made by philosophers to compute the quantity of water which rivers discharge into the ocean; but it is a problem scarcely admitting of any solution, beyond a probable approximation. From the observations of Father Riccioli on the discharge of the river Po, it has been calculated, that the water poured out by all the rivers of the globe is equal to forty-one cubic miles daily, or to 14,965 cubic miles annually. This computation is probably too high. The greatest estimate of the mean annual fall of rain, dew, &c. is not above thirty-four inches for the whole earth; and the superficies of the globe being 196,816,658 square miles, this will afford a mass of water equal to 105,614 cubic miles as the whole that is annually precipitated on the earth. But from this must be deducted all that falls on the ocean, which, if we consider the precipitation in the ratio of their surfaces, will give 38,271 cubic miles as the water that falls on the land. Of this quantity, however, at least two thirds are expended in irrigating the soil or in sustaining vegetable life, and are restored to the atmosphere by the process of evaporation, so that no more than 12,757 cubic miles will become the annual tribute of the rivers of the globe to the ocean. This estimate differs less from that founded on the tables of Cotte than some later speculations on this subject, in which we suspect that the mean annual rain, and the quantity returned by the rivers to the sea, are considerably underrated.

Length of their course. The length of the course of twenty-two of the principal rivers, and the area of their respective domains or basins, were calculated with much care, and given in a tabular form in the essay on Physical Geography in the last edition of the Encyclopædia Britannica. These elements we here insert, with the addition of a reduction of the proportional lengths into English miles, which has been obtained by multiplying them by 180, the distance between the remotest sources of the Thames and its embouchure at the Nore, that river having been assumed as unity in the first column of the table. The proportion of water discharged by each has been omitted, because the data on which that column was calculated did not seem to be sufficiently established.

A diagram exhibiting the lengths of the courses of the principal rivers on the globe will be found in Plate CCCCX.

We have noticed the usual courses of rivers; but sometimes their streams, after flowing through open channels, are engulfed, and afterwards re-appear at considerable distances from their occultation. Limestone districts afford many specimens of this phenomenon. In Derbyshire, the small streams named the Hamps and the Manifold sink into a chasm on a common near Ashburn, and re-appear in Islam grounds, three miles below. The stream which issues from the Cavern of the Peak is believed to be the same which disappears near the road from Chapel-in-Frith to Castleton; and the caves of Weathercote and Yordas in Yorkshire exhibit similar instances. The occultation of the Rhine has great celebrity, and attracts numerous visitors from Geneva. The Guadiana, one of the largest rivers in Spain, disappears near the village of Castillo de Cervera, and, after pursuing a subterranean course for twelve or fourteen miles, bursts again into day near the high road to Madrid, between Manzanares and Villaharta, at the spot named Ojos de Guadiana, or the Eyes of the Guadiana. No country surpasses Greece in the number of its subterranean streams, the peculiarities of which are often clothed in splendid mythological fictions by her ancient bards. The waters of many valleys in the Peloponnesus have no other outlets than Ugias, or chasms engulfing the streams, which are termed Kataothra by the modern Greeks. Such are the outlets of the valleys of Tegea, Mantinea, Ascea, Orchomenus, Alea, Stymphalus, and Peneus. A familiarity with such phenomena, and a poetical temperament, readily led the ancient Greeks to conceive still more distant secret communications; and the imagination which peopled every

¹ Jameson's Journal, vol. xxiv. for January 1838.

grove and animated every stream with presiding deities, could easily reconcile itself to the fabled loves of Alpheus and Arethusa, which led the enamoured river-god to pursue his coy nymph, beneath the bed of the ocean, from Greece to the shores of Sicily.

5. SPRINGS.—Springs derive their origin from water raised into the atmosphere by evaporation, falling down in the form of showers or mists on hills, and thence percolating through fissures in rocks, or through porous beds, until they appear at lower points of the earth's surface. Hence springs of any size are rarely found near the summits of mountains. When they occur in such situations, they probably arise from water collected on higher though distant mountains, finding its way, as in the two legs of a siphon, through a fissure open to the air only at its extremities. Springs are most numerous in mountainous districts, because hills readily arrest and condense clouds. The observation of Mr White of Selborne, on the condensation of moisture by a single tree, is a proof of the effect of even small elevations in collecting moisture from the atmosphere. Springs are directed to the surface by the position of the strata or fissures through which they flow. In flat countries, springs sometimes have a long subterranean course, especially when they encounter beds of plastic clay, through which their waters cannot penetrate. Copious perennial springs usually undergo very little alteration in their temperature throughout the year; and therefore very generally indicate the mean temperature of the climate where they occur. But water brought up from considerable depths in the earth, as in those wells which (from having been first introduced in Artois) are termed Artesian, has usually a higher degree of heat, which increases with their depth.

A higher temperature than the mean of the latitude is occasionally also observed in large springs, the source of which is perhaps not deep. Professor Trall has stated, that the largest spring in Great Britain, St Winfred's Well, at Holywell, has always a temperature several degrees above the mean heat of our climate, and that at different times he had found it to vary as much as 4° of Fahrenheit. This noble spring may be considered as the bursting out of a subterranean river; for it throws up, it is said, about twenty-one tons of water per minute, or 30,240 tons daily. It is worthy of remark, that this spring issues from a hill much disturbed by numerous mineral veins, and probably has its temperature also affected by surface water.

The Crawley Spring, from which Edinburgh is supplied with water, though of far inferior size, is yet very remarkable, as it yields 220 cubic feet, or more than a tun, per minute, of the purest water.

The most magnificent spring we have ever beheld is the far-famed Fountain of Petrarch, at Vaucluse. It rises within a cavern, at the foot of a vast semicircular precipice of compact limestone terminating a wild valley. In the beginning of autumn, by clambering over a heap of rubbish, the traveller may descend into the cavern, and will find himself on the brink of a gulf of the clearest water, of unfathomable depth, with the deep blue tint of the ocean out of soundings, rising with great force, and an unruffled surface, from the recesses of the mountain. It issues from the cavern through concealed channels, and forms at once the river Sorgue, capable at its very source of moving machinery, and almost immediately navigable for boats. The waters of this extraordinary spring never vary half a degree in temperature. When the melting of snow on the neighbouring Alps increases its sources, the cavern is entirely filled with water, and the stream rushes over the rugged bank at the mouth of the cavern with the tumult of a cascade.

The water of springs always contains air, and some saline ingredients. In good potable water the air usually equals $\backslash frac\{1\}\{2\}$th or $\backslash frac\{1\}\{3\}$th of its volume; of this one third is atmospheric air, and the rest carbonic acid gas. The saline contents of

good spring-water generally vary from $\backslash frac\{1\}\{5000\}$th to $\backslash frac\{1\}\{50000\}$th of their weight; and such water has a specific gravity of from 1.00016 to 1.00030. Such springs are termed soft, and answer well for culinary and other domestic purposes. When the solid ingredients exceed this proportion, the water becomes hard; and is less fitted for cooking or dissolving soap, in proportion to the quantity of foreign impregnation it contains. The springs of Upsal, analysed by Bergman, afford a good instance of a very soft water. The three purest contained no more than $\backslash frac\{1\}\{5000\}$th part of their weight of solid matter, and had a specific gravity of 1.0002. From a kanne (about 57,000 grains troy) of the Sandvik spring, he obtained 9.65 of solid contents, in the following proportions:

besides four cubic inches of carbonic acid gas, and two cubic inches of oxygen gas.

The same ingredients are found in most potable waters, and many also contain sulphate of lime. The springs which supply Liverpool have a specific gravity of from 1.0026 to 1.0027, and contain $\backslash frac\{1\}\{5000\}$ of solid matter. They are rather hard; but when the ingredients are in more considerable quantity, the spring becomes a mineral water.

Mineral springs, besides the substances already indicated, Mineral may contain one or more of the following substances: sulphuretted hydrogen, or nitrogen; sulphates of soda, ammonia, magnesia, alumine, iron, copper; nitrates of potassa, lime, and magnesia; muriates of potassa, ammonia, and baryta; more rarely free sulphuric acid, boracic acid, and borate of soda. Muriatic acid uncombined has been detected in the waters of the Pusambio in South America, which arise from the volcanic mountains of Purace, and are strongly impregnated with sulphuric and muriatic acids; and in Java sulphuric acid has been discharged by volcanoes in such quantities as to destroy vegetation. Boracic acid is deposited from several of the Tuscan springs, especially from the thermal waters of Sasso. The borate of soda is found in some of the lakes of Persia and Thibet. Mineral waters may be, 1. acidulous; 2. chalybeate; 3. sulphureous; or, 4. saline. The first generally owe their sparkling appearance and acid qualities to carbonic acid. The second contain either the carbonate or the sulphate of iron. The third owe their fetid smell to sulphuretted hydrogen, or to sulphurets of lime, or of some other alkaline substance. The fourth contain a notable portion of neutral salts, from which they have a disagreeable taste, and are hard waters.

The acidulous waters almost always contain some alkaline or earthy salt. Such are the waters of Tunbridge, Spa, and Pyrmont, Seltzer, and others. The chalybeates generally contain from one to two and a half grains of oxide of iron, or some salts, in each English pint, often combined with other salts. Of this kind are the British springs of Peterhead, Brighton, and the chalybeate sources of Harrowgate and Cheltenham, which last, however, are much mixed with other saline ingredients. These vary from 1.0007 to 1.0010 in specific gravity.

Our sulphureous waters are potent, and owe their qualities to sulphuretted hydrogen. The principal are, Strath-ous waters: peffer, near Dingwall; St Bernard's Well, at Edinburgh; Hartwell, near Moffat; Harrowgate, Leamington, and a less known well at Kilburne, near London. Their specific gravity ranges from 1.0042 to 1.0132, especially of those which, like Harrowgate and Leamington, contain much saline matter. Moffat spring contains very little saline impregnation, and its sulphuretted hydrogen amounts to 1.20 cubic inch in each English pint. The old well at Harrowgate contains much saline matter, and 1.75 cubic inch of that gas in the same quantity of water. The thermal wa-

Physical Geography. ters of Aix-la-Chapelle are very strongly impregnated; containing no less than 55 cubic inches of sulphuretted hydrogen.

Saline waters. Saline mineral waters contain a notable quantity of some of the salts already indicated as found in springs. The most noted in Great Britain are the waters of Dumblane, Pitcaithly, Cheltenham, and Leamington. Their specific gravity varies from 1.0047 to 1.0119. The waters of the two last contain respectively 80.5 and 110.38 grains of salts in each English pint. All these mineral springs are reputed remedies for various diseases, but are evidently unfit for culinary purposes. The source of their peculiar impregnations is the mineral beds through which they flow.

Thermal springs. Thermal Springs.—The heat of thermal waters has been ascribed, by Hutton, Laplace, and Bischoff, to the agency of a central heat, which is conceived to exist in the bowels of the earth. Some indeed have supposed that their temperature may be derived from the mutual chemical action of local accumulations of certain substances at more moderate depths in the earth. But the long-continued uniformity of their heat has given rise to a modification of this doctrine, which is ably supported by Dr Daubeny. He considers thermal springs as deriving their temperature from the same source as volcanoes, viz. the infiltration of water through the external crust of the earth, to the regions where he conceives the metallic elements of earthy and alkaline bodies to exist. The intense heat extricated during the oxidation of these elements will convert a portion of the water into steam, which, under compression, will attain a high temperature, act on various earthy matters, and communicate its heat to the subterranean waters which issue in thermal springs. The position of such springs in the vicinity of disturbances or upheavings of strata by igneous rocks, seems to favour this idea. Whatever may be the cause, we can now no longer doubt that there is within the earth a source of heat, the effect of which is perceived in the increased temperature as we descend in deep mines, and in the distances to which the effects of volcanoes are known to be propagated. The hypothesis of a central heat, unconnected with chemical changes, has been strongly insisted on, from the uniformity of the increments of temperature with the depth of borings for Artesian wells; and the result would lead to the conclusion, that the heat of the earth augments, at an average, one degree of Fahrenheit for every 100 feet of perpendicular depth; so that at the depth of three miles water would be boiling hot. But if we adopt the views of Daubeny, the temperature of particular springs will depend on the variable profundity of the unoxidized materials of the globe, and the supply of water. The temperature of thermal waters seems to be liable to little change in each spring; but in different waters it occurs in every degree, from a temperature very little exceeding the mean of the place, to the boiling point of water.

Table of the Temperature of some Thermal Springs.

Geyser. Iceland abounds with boiling springs, some of which are remarkable for their issuing out of a small mass of rock in the midst of a large river, as that figured by Sir George Mackenzie in the Reikiadalsaa; others are celebrated for

their intermissions and magnificent jets of boiling water, as the Geyser. These fountains are well known by the writings of Von Trol, Stanley, Mackenzie, Hooker, Henderson, and by two Danish publications, the first by Olafsen and Povelsen, the latter by Ohlsen. The two principal of these singular fountains are in a valley near Haukadal, about sixteen miles from Skalholt, the bottom of which contains numerous cavities filled with hot water. In the midst of a little mound formed by siliceous deposits from its waters, six or seven feet in height, opens a circular basin fifty-six feet in diameter in one direction, and forty-six in the other. The depth of this basin is three feet; but in its bottom is a round hole about ten feet wide, the termination of a pipe or fissure which sinks into the earth to the depth of seventy-eight feet. By this pipe the water occasionally retires, leaving the basin quite dry, and no steam is then perceived to issue from the pipe. Periodically the water begins to rise in the pipe, and again fills the basin to overflowing; the ground is shaken by hollow subterranean explosions; suddenly a prodigious column of boiling water is shot into the air with astonishing violence, and clouds of steam obscure the atmosphere. This is followed by successive jets, sometimes to the number of sixteen or eighteen in five minutes, the last of which is generally the most energetic, and attains an altitude of from 90 to 100 feet. On a second visit to the Great Geyser in 1815, Mr Henderson saw it rise to the height of 150 feet; and Lieutenant Ohlsen says that in 1804 he took the altitude of one eruption with a quadrant, and ascertained it to be 212 feet. These efforts are attended with a loud noise, and the ground trembles beneath the feet, whilst the velocity with which the jets and the accompanying steam are hurled into the air is astonishingly sublime. When stones are thrown into the pipe, they remain there until the succeeding jet projects them with great violence into the air, and they may be seen descending amid showers of boiling water. The intermissions of this extraordinary fountain are not absolutely regular; but there are generally four great series of eruptions from its orifice in twenty-four hours.

The jets from the Strokr, or New Geyser of Stanley, though smaller in diameter, are scarcely less magnificent; and, taking place more frequently, and with surprising violence, afford a high gratification to the traveller.

The mechanism of these Geyser must be simple, to have lasted for many ages. They are mentioned by Saxo in his History of Denmark; and this shows that some of them have existed for about 600 years. It is not easy to imagine an application and removal of a heating cause within the earth in such short intervals; but an arrangement somewhat resembling that represented in No. 13 would be ca-

No. 13.

pable of producing a Geyser. A cavity similar to that supposed by Mackenzie or by Lyell, in which water performs the office of a valve, from the form of the lower part of the canal communicating with the surface, might suffice; or its

effect might be increased by a simple valve formed by a mass of stone accidentally getting into the position represented in the figure. Indeed, an experiment of Mr Henderson on the Strockr proves the dependence of the force of its jets on some such mechanical obstruction. That gentleman found that the eruptions were accelerated by throwing a quantity of stones into the pipe, and the violence of the succeeding jets seemed to be in proportion to the load so applied; just as loading the valve of a steam-boiler will increase the elasticity of the steam. Henderson asserts that he was thus able to double the effect of the fountain, and to make it project the water to a height of 200 feet. The steam is no doubt produced by the agency of that volcanic fire which everywhere pervades Iceland, and is evidently in great activity under this secluded valley.

If the cavern in the imaginary section, No. 13, be supplied with water from a subterranean source, whilst steam is introduced into it from some more distant volcanic focus, the fountain will remain quiescent until the accumulation of steam, and the pressure of the vertical column of water, aided by the rude valve of rock, will augment its elasticity, until it become sufficient to overcome the resistance, and to project the water remaining in the pipe into the atmosphere.

The Geysers are not confined to the valley near Haukadal; several on a smaller scale exist at Reykum, issuing, at intervals of about an hour and a half, from a bank of hot sulphureous clay.

The waters of the Great Geyser hold a notable proportion of silex in solution, which no doubt is owing to their alkaline impregnation, and their high temperature. According to the analysis of Dr Black, in the Edinburgh Philosophical Transactions, vol. iii., an English gallon contains—

The silex is deposited in beautiful stalactitic forms in many of the adjacent pools, and the Geysers petrify plants, like other petrifying waters.

Cold intermitting springs have been well described by the late Mr Gough of Kendal, who divides them into two kinds; reciprocating, or such as have a constant but irregular flow; and true intermitting springs, in which the issue is entirely suspended for a time, and the water returns at intervals of greater or less regularity.¹ Of the former sort we have a good example in Giggleswick Well, in Yorkshire, in which the reciprocations are well marked. In it the water will sometimes rise and fall a foot in ten or fifteen minutes; at other times it will fall as much in four minutes, but usually requires double that time to regain its former height. The water bubbles out with force, so as to carry with it particles of sand. Another reciprocating well occurs near Torbay, called Lay-well. It is described, in an early volume of the London Philosophical Transactions, as ebbing and flowing sixteen times in an hour. A similar well, but on a far less scale, is St Anthony's Well on Arthur Seat. The cause seems to be in both instances the confinement and compression of air in tortuous cavities within the subterranean channel of the spring, which, from the position of the aperture, is alternately compressed by the accumulating fluid, until it acquire sufficient elasticity to drive before it the resisting column of water. It is easy to conceive a simple mechanism capable of this alternation, without any species of valve, except that formed by the water itself, which may be represented by the diagram No. 13, without its valve. Of the perfectly intermitting spring the most remarkable is the

Bolder-Born, in Westphalia, near Paderborn. After flowing for twenty-four hours, it entirely ceases for the space of six hours; it then returns with a loud noise, and mixed with air, in a stream sufficiently powerful to turn three mills very near its source. In Lowthorpe's Abridgment of the Philosophical Transactions (vol. ii.), it is said to disappear twice in twenty-four hours; and possibly its intermissions are not regular, but may depend upon the quantity of rain which has recently fallen. Dr Boué has described a powerful intermitting spring which occurs at Bihar, in the commune of Maybugyer, in Hungary. It issues, many times a day, from the foot of a mountain, after hollow subterranean noises, in such quantity as in two minutes to fill the channel of a considerable stream. Its returns are most frequent from Christmas to Midsummer, when they take place every quarter of an hour; but they are rarer during the other half of the year.²

The most probable manner in which such powerful effects are produced, seems to be the descent of streams of water, much mixed with air, into a cavern, the more direct communication of which with the surface is through a channel so connected with the cavern as to confine the air in its upper part, in the manner of a water-blast, until the increasing elasticity, from compression, enables the imprisoned air to dislodge the water accumulated between it and the outlet.

SECT. VII.—The Atmosphere.

The earth is surrounded by an invisible and highly elastic fluid, termed its atmosphere, which accompanies it in all its movements; being retained by the influence of gravitation toward the earth, though, from the repulsion between its own particles, it extends to a considerable distance from the surface. Its density diminishes as it recedes from the earth, and its expansibility is in the inverse ratio of the force by which it is compressed. Its density is farther affected by the influence of heat and cold, by the admixture of watery vapour it contains, by the decreasing force of terrestrial gravitation, and the increasing attraction of the heavenly bodies as its distance from the surface of the earth is augmented. Experiment and calculation have proved that the rate of diminution is such, that the atmosphere at the height of seven miles would only have a density $\backslash frac\{1\}\{4\}$th of what it has at the surface; and that if the altitudes be taken in an arithmetical series, the decrease in density will follow a geometrical ratio. Thus, if the density at the surface be assumed as unity, at seven miles it will be $\backslash frac\{1\}\{4\}$th, at fourteen miles it will be $\backslash frac\{1\}\{16\}$th, at twenty-one miles it will be $\backslash frac\{1\}\{64\}$th, and so on. This property of the air would lead to the idea of an indefinite extension of the atmosphere; but there is evidently a limit to this, probably arising from the gravitation of the particles of air; and the nicest observations on the duration of twilight, as well as some other phenomena in astronomy, almost warrant the inference, that the utmost limit of the atmosphere cannot exceed forty-five miles in altitude. See the article ATMOSPHERE.

Air is ponderable; 100 cubic inches of air at the temperature 60° Fahrenheit weigh 30.5 grains troy, at a mean barometrical pressure; and a perpendicular column of the whole atmosphere is just balanced by a mercurial column of about thirty inches in height. Hence the length of such a column is employed to ascertain the weight or pressure of the atmosphere; and the oscillations of the barometer have taught us, that this pressure varies in different stations on the earth's surface, and in the same place at different times. See the articles BAROMETER, ATMOSPHERE, and METEOROLOGY.

¹ Nicholson's Journal, eighth series, vol. v. p. 35.
VOL. XVII.

² Jameson's Journal, vol. xxxv. July 1835.

The height of the barometer forms an important part of a meteorological journal. Observation has made us acquainted, not only with the occasional, but also with the diurnal oscillations of the column; and exhibited the connection of the more considerable variations with changes in the weather. The trouble of continual attention to obtain the mean heights of the barometer at any place, have given rise to several contrivances for registering the maxima and minima during the absence of the observer. A very ingenious instrument for this purpose is described by Mr Keith, in the Philosophical Transactions of Edinburgh (vol. iv.). The open end of the instrument is a larger cylinder than the rest of the tube. The cistern is placed horizontally at the upper part, and an ivory float, supporting a kneed wire, is placed on the surface of the fluid, as in No. 14. The scale is above the opening of the tube, and the kneed wire moves two light indices, sliding on a fine wire, with sufficient friction to retain them in the position which they receive from the movements of the ivory float. The same gentleman caused the instrument, by a slight modification, to record the whole course of its observations through a given period. A perpendicular cylinder, divided into thirty-one compartments, is moved round its axis once a month by clock-work; a pencil attached to the float wire records the whole course of the oscillations; and a proper scale, marked on the cylinder, gives the value of the irregular curves thus traced. Once a month the cylinder is covered with a fresh paper, ruled into thirty-one columns, for another monthly series.

Professor Traill has given a simple and cheap form of a register barometer for maxima and minima.

A diagonal barometer registers the maximum height of the column. Before the tube is bent, a piece of iron wire is introduced into the closed end of the tube, which is then filled in the usual way. The wire being lighter than the mercury, will rise to the top of the column, is pushed before it on the rise of the barometer, and, on the fall, remains, by its own weight, at the point it had attained. A slight modification of Cassini's rectangular barometer marks the minima, by means of a piece of iron wire introduced at the open end of the tube, as in No. 15. To make the indications of both instruments as nearly equal as the construction will allow, the lower part of this tube is bent to less than a right angle.

It is obvious that the barometer affords the means of ascertaining the weight of the whole atmosphere. The pressure of the atmosphere varies from twenty-eight to thirty-one inches of the barometric column. A column of pure mercury, whose base is a square inch, and its height the mean between twenty-eight and thirty-one, or 29.5 inches, is found to weigh 14.5 pounds avoirdupois; therefore, the pressure of the whole atmosphere will be 14.5 pounds on each square inch of the earth's surface. And as the superficies of our globe is equal to 790,116,426,647,756,800 square inches, if we multiply this number by 14.5, we obtain, as the absolute weight of the whole atmosphere, 11,456,688,186,392,473,600 pounds; or, if we adopt the more common estimate of the mean barometric height at 30 inches, and the weight of that column at 15 pounds on the square inch, this will give 11,851,746,399,716,352,000

pounds, or 5,290,958,214,159,085.7 tons, as the mass of our atmosphere.

The variations of atmospheric temperature are ascertained by the thermometer. (See the articles METEOROLOGY and THERMOMETER.) The instrument, for this purpose, should be placed in the shade, protected against reflected heat, and also against that transmitted through walls, or other bodies to which it is attached. It is by no means easy, in a town, to obtain an eligible position for a measure of atmospheric temperature, on account of the many causes which affect the instrument. It ought to be protected by a glass case against rain; yet the air should circulate freely around it, and it ought to radiate its heat freely to the sky. The importance of obtaining accurate information on changes of the temperature of the atmosphere has given rise to numerous observations on the annual, monthly, daily, and hourly variations of the thermometer.

The labour and minute attention required in such investigations have induced ingenious men to devise methods of ascertaining mean temperatures from a few observations. It has been found that the mean of the thermometer during the month of April will, in our climate, nearly indicate the mean annual temperature; and that the mean of observations made about half past 8 A.M., at 2 P.M., at 9 P.M., and an hour before sunrise, will give the mean temperature of each day. To abridge such fatiguing observations, register thermometers have been devised. The best known of these are the thermometers of Six and of Rutherford. (See the article METEOROLOGY, Plate CCCLIV. fig. 6 and 7.) A modification of Six's thermometer, rendering springs to the indices unnecessary, has been constructed by Professor Traill, in which the indices slide in slightly inclined tubes, as will be easily understood from No. 16.

As in Six's thermometer, the expansions of alcohol are the measure of the temperature; the indices, of iron wire, enclosed in glass tubes, are moved by a short column of mercury; and they are adjusted for a fresh observation by bringing them to the extremities of the mercurial column with a magnet.

The atmosphere itself is invisible; but the column of air interposed between the eye and distant objects invests them with an azure hue. This tint depends upon the greater refrangibility of the blue rays of light than of those towards the other end of the spectrum. When a ray of white or undecomposed light enters our atmosphere, the red and yellow rays pass with little deviation from a rectilinear course; but the blue rays are dispersed in the air, and affect the colour of objects beheld through a long column of the atmosphere. As we ascend lofty mountains, the refractive power of the air diminishes with its density, and hence the blue tint is lost in the blackness produced by the want of refraction; until a traveller, on the ridges of the Alps, the Andes, or the Himalaya, will perceive the sky to assume a hue approaching to deep blue or black. Saussure has described this phenomenon, and invented the Cyanometer, an instrument for measuring the intensity of the colour of the sky. (See the article CYANOMETER.) The form will be best understood from No. 17.

With this instrument Saussure found a numerical value for the colour of the sky at different places. Thus it was at Geneva = 17°, at Chamouni = 19°, on Col du Géant = 31, and on Mont Blanc, at the height of 15,750 feet, it was = 39°. It also showed that the intensity of the colour varied at the same place at different times. The intensity of the colour is greatest at the zenith. It increases towards mid-day, more rapidly on mountains than on plains; and the difference between morning and mid-day is then also greatest; but the whole difference is most considerable in plains.

The following comparative tables of observations on Col du Géant, and at Geneva, illustrate these points. The first column shows the degrees above the horizon, the second the degree of the cyanometer.

At midnight, on the Col du Géant, the cyanometer indicated 51°.

Some years ago we endeavoured to give greater permanence to the tints by employing thin films of blue glass in constructing a cyanometer. The first degree was formed of a single film enclosed between plates of thin crown glass, and each succeeding degree consisted of an additional film.

To the refractive power of the atmosphere we owe the splendors of sunrise and sunset. When the sun is on the horizon his beams reach the eye through a long column of dense air; the more refrangible rays are dispersed, and the eye receives the red and orange rays undiluted, especially through the serene air of a summer evening. As the brilliant hues of sunset melt away, the blue tone which the atmosphere had acquired by the dispersion of the most refrangible rays, but which was lost in the splendour of sunset, gradually reappears, and, mingling with the ruddy gleam, sheds the purple light of evening on the softened landscape. As the sun descends more below the horizon the purple colour fades away, and gradually gives place to the "faint erroneous ray," that "flings half an image on the strained eye."

Atmospherical air was long considered as a simple substance, one of the elements of other matter; but the discoveries of Black, Cavendish, Priestley, Scheele, and Lavoisier, have proved it to be a compound, of which the principal constituents are oxygen and nitrogen gases, usually estimated at twenty-one volumes of the former to seventy-nine of the latter. Dr Black discovered that the atmosphere always contained carbonic acid, which was found to exist in slightly variable proportions; and it holds water suspended in the state of vapour, in proportions which are still more fluctuating. If we average the two last ingredients, the composition of the atmosphere may be stated as follows:

The manner in which these different ingredients are united has given rise to some controversy. The opinion once generally entertained was, that they were united by an affinity between their particles, like that which forms chemical compounds. The celebrated Dr Dalton first suggested the theory, that of the various elastic fluids constituting the atmosphere, the particles of one are neither retractive nor repellent of the particles of another; but that the pressure upon any one particle of such a mixture is entirely derived from particles of the same kind of matter. Each fluid, on this view, is considered as occupying the same space as it would in a vacuum; and each constituent of the atmosphere, under a mean barometrical pressure of thirty inches, as exerting its own separate pressure in the following proportions, depending on its own quantity and specific gravity. Thus, the

Soon after the composition of the atmosphere was detected, and the qualities of oxygen in supporting combustion, and respiration were understood, it was conceived that important modifications of disease might depend on or originate in the varying proportions of this essential ingredient in the air; and various methods were devised for analysing atmospheric air, and ascertaining the quantity of oxygen it contained. 1. Dr Priestley devised a method of doing this by the eudiometer, or measure of the purity of the air. His method consisted in adding, in a graduated tube, fifty-five volumes of nitric gas to a hundred of atmospheric air. The property of this gas is to abstract oxygen from the air, and to form nitrous acid, which is absorbed by water; and the amount of the diminution marked the quantity of oxygen in the air. 2. Gay-Lussac improved Priestley's eudiometer. Instead of adding the nitric gas in the graduated tube, he mixed a hundred volumes of the gas with as much of the air to be analysed, in a wide jar, and transferred this mixture to the graduated tube to ascertain the diminution; on dividing the residual air by four, he obtained the amount of the oxygen present in the air. 3. Scheele employed the sulphuret of potassa for the same purpose. 4. Seguin used the rapid combustion of phosphorus; and, 5. Volta exploded a mixture of hydrogen and air by means of an electrical spark. This is one of the most elegant and accurate of the eudiometers. He recommends a mixture of 200 volumes of hydrogen with 300 of air. On passing an

¹ Manchester Memoirs, vol. v. See the article ATMOSPHERE.

Physical Geography. electric spark through the wires in the top of the jar, the hydrogen is fired, instantly forms water with the oxygen of the included air, and, in ordinary cases, the diminution of volume of the mixture amounts to 195, of which one third is oxygen, or sixty-five parts of oxygen have been lost by 300 of air, or about twenty-one per cent.

6. Dr Hope's eudiometer consists of a graduated tube exactly ground to a small tubulated phial. The bottle may be filled, either with hydrosulphuret of potassa, or with a liquid proposed by Davy (which is formed by saturating sulphate of iron in solution with nitric gas.)

The air is in the tube, which is inserted in the neck of the phial; the instrument is inverted; the air, now in the phial, is agitated with the liquid for a few minutes, and on opening the tubulature of the phial, below the surface of the liquid, the amount of the oxygen absorbed will be indicated on the graduated tube. This very neat instrument has been modified by Dr Henry, who substituted a bottle of caoutchouc for the phial. Any of these instruments, especially 18 or 19, will show the quantity of oxygen in any given air; but it has been found that the quantity of oxygen of the air is little subject to variation, and that gas is equally present on mountains and in plains, in the town or in the country, except in confined situations, where fuel is burning, or many animals respiring. Air which is very deleterious to man has no sensible difference in the proportion of the oxygen from the purest atmosphere; and we must probably look to other causes than this for the unwholesomeness of the atmosphere.

The quantity of carbonic acid in the atmosphere is subject to slight variation, except in confined apartments. Dr Dalton found it only varying from $\backslash frac\{1\}\{1400\}$th to $\backslash frac\{1\}\{1000\}$th of the whole air, an uniformity which is surprising, considering how much carbonic acid is constantly formed during respiration and combustion. Perhaps this may be partly owing to our imperfect modes of ascertaining very minute changes in its quantity. If we are to give credit to the experiments of the younger Saussure, the quantity in the atmosphere is greater in summer than in winter, by 2-3 volumes in 10,000; but his experiments are not generally considered as satisfactory, in a case where the difference appears so great, and the point so difficult of determination. Even the experiments of Dalton are not considered as decisive; for lime-water does not appear to be capable of abstracting all the carbonic acid from air, as was remarked by Cavendish; and philosophers have generally considered the quantity as ranging from $\backslash frac\{1\}\{1400\}$th to $\backslash frac\{1\}\{1000\}$th of the whole atmosphere. In confined apartments, however, the quantity of carbonic acid is sometimes much increased, while the oxygen is consumed in the process of combustion and respiration; and this double vitiation of the air may give rise to fatal consequences, as was too well illustrated in the catastrophe of our unfortunate countrymen in the dungeon of the Black Hole in Calcutta in 1756, and in one which happened to persons confined in St Martin's Roundhouse in London in 1742. In the celebrated Grotto del Cane, the carbonic acid which issues from crevices in the floor fills the bottom of the cave to the level of the entrance, and then flows out by its superior gravity, so that the upper part is free of the noxious impregnations. A man enters with impunity, because his head is above the stratum of carbonic acid; but the air of the cave is speedily fatal to a dog, because the prone position of his air-passages brings his respiratory organs into contact with the gas. It is remarkable that these gases do not appear to vary considerably in their relative proportions in any part of the accessible atmo-

sphere. Air collected on the summit of Mont Blanc was analysed by the younger Saussure, and agreed in the proportions of the gases with that of Geneva; and the same uniformity was found in air from Spain by De Marti, from Egypt and France by Bertholet, from the coast of Guinea and England by Davy, from the summit of Teneriffe and the ridges of the Equinoctial Andes by Humboldt.

The quantity of aqueous vapour in the atmosphere is liable to far greater variations than that of the other constituents. Water in the state of vapour is invariably present; but the proportion is modified by the temperature, and by changes in the barometric pressure of the air. The process by which water finds its way into the atmosphere is one of the highest importance, and has obtained the name of

Spontaneous Evaporation.—If water be exposed to the air, it gradually disappears; and if this be carried on in closed vessels, the air is found to have taken up water, which may again be obtained by the action of certain substances that have a great affinity for water. The principal substances which have been employed for abstracting water from air are, sulphuric acid, quicklime, chloride of lime, and carbonate of potassa. If given proportions of any of these be enclosed in a vessel containing air, they absorb the moisture of that air; and the weight they thus acquire has been taken as the measure of the quantity of its water. The same is estimated by the quantity of water which is absorbed by a given volume of air thus artificially dried. Some very careful experiments on this subject were made by Saussure, who inferred that a cubic foot of air, at temperature 65° Fahrenheit, when fully saturated with moisture, holds eleven grains of water dissolved. Notwithstanding the acknowledged accuracy and patience of this philosopher, we are inclined to prefer the method of Dr Dalton (founded on the suggestion of Le Roy of Montpellier), of observing the temperature at which moisture begins to be separated from the air. This is termed the Dew-point of that atmosphere, and forms an important element in several interesting problems in meteorology. Dalton had shown, that the water existing in the atmosphere has precisely the same elasticity as it would assume in a vacuum at the same temperature; hence it is obvious that it exists in the air, not as water, but as elastic vapour. But the elasticity of vapour is demonstrated to depend on temperature; and, therefore, if we could measure the elasticity, we could determine the quantity of water in the atmosphere. This elasticity is measured by the Manometer, a barometer the cistern of which is enclosed in the vessel containing the air, and therefore only affected by the variations of elasticity of the included air. Deductions from these experiments have been given in a tabular form in the Manchester Society's Transactions.

In the article HYGROMETRY there is another table of great value, showing the elasticity, weight, and rate of expansion of aqueous vapour, in air saturated with moisture, at each degree of Fahrenheit's thermometer, from 0° to 100°, which was intended by the author of that treatise to facilitate hygrometrical investigations. In this table the elasticity of the vapour is calculated on a formula deduced by Biot from Dalton's experiments.

The quantity of water suspended in the air is liable to great variation, from several natural causes; and it is chiefly liable to be deposited or separated, on a diminution of the capacity of the air for moisture, by a reduction of its temperature. Warm air is capable of taking up, in the state of invisible vapour, more water than cold air. Thus a cubic foot of saturated air, at 32°, contains of watery vapour 2-350 grains; at 60°, 5-825 grains; at 70°, 7-941 grains. When the temperature of the air is diminished, as, for instance, by mingling with colder currents in the atmosphere, a quantity of water is separated; and, according to the sud-

deness of the change, will be suspended in the form of visible clouds, or precipitated in fogs or rain.

Changes in the density of the air also affect its capacity for retaining moisture. Evaporation is quicker, everything else remaining unchanged, in rare than in dense air. Saussure found, that water evaporated quicker on the Col du Géant, which is 11,275 feet above the sea, than at Geneva, which is but 1324 feet above that level, in the proportion of eighty-four to thirty-seven. Light air then speedily becomes charged with moisture, which slight changes in temperature are ready to precipitate; and hence a falling barometer is often the immediate precursor of rain.

When atmospheric air approaches its point of saturation with moisture, evaporation goes on slowly, at length ceases, and a slight diminution of temperature is sufficient to precipitate a portion of water on contiguous bodies; we then term the air damp or moist. This change is not always to be immediately recognised by our sensations; and as the change is important to the success of some processes in the arts, and also to health, philosophers have invented instruments for indicating the approaches of the air to saturation, which have been named

Hygrometers, or measures of moisture. The change in bulk which some vegetable and animal substances undergo, in moist air, has been employed for this purpose. The most common are the twisting and untwisting of a cord of flax or hemp by moisture and dryness; the movements of the natural twist in the beard of the wild oat, avena fatua; in the Indian grass, andropogon contortum, proposed by Captain Kater; or the arista of the seed of stipa pennata (a common inmate of our gardens, and often found in our fields), proposed by Dr Cumming of Denbigh. Unfortunately these substances, though indicating minute changes in the air, are incapable of forming comparable hygrometers, and gradually become less and less sensible to the influence of humidity. The same may be said of the animal substances which have been employed in the construction of such instruments. The most simple of the animal hygrometers is Wilson's, formed of a rat's bladder fixed to a glass tube, and filled with mercury to a certain point on the tube when the bladder is moist. As it dries, the contraction of the bladder raises the mercury in the tube. Still more delicate hygrometers were formed by Saussure and De Luc, on nearly the same construction. (See the article HYGROMETRY.) Saussure employed a fine human hair, freed from unctuous matter by boiling in a weak alkaline solution; and De Luc used a slip of whalebone, scraped extremely thin. These animal bodies lengthen by moisture; and, in both instruments, one end of the hygroscopic substance was fixed to the frame, while the other was attached to a fine flat wire lapped round an index made to move along a graduated arc by the varying length of the hygroscopic substance. In Saussure's instrument, the index was brought back by a small weight as a counterpoise; and the same was effected in De Luc's by a spiral gold-wire spring. Both instruments were graduated by placing them in a jar filled with air saturated with moisture, and then in air dried by quicklime, or some similar body; and the space between these extreme points was divided into 100°, or any given number of degrees.

The imperfection of these instruments, when made by the hands of the inventors with every care, rendered it difficult to have two that gave the same indications under similar circumstances; and the still more fatal objection, that they become altered by time in mobility and delicacy of indication. Although both Gay-Lussac and Biot have endeavoured to show the relation between the indications of the hair hygrometer and the quantity of moisture in the air, we cannot consider the instrument as entitled to confidence.

A very elegant hygrometer was invented by Daniel, founded on the determination of the temperature at which dew begins to be deposited; but for the description of it we must refer to the article HYGROMETRY, as well as for an ingenious substitute devised by Mr Adie of Edinburgh, and another by M. Pouillet. All these depend on the dew-point; and all are liable to the objection, that it is difficult to determine the precise degree at which dew begins to be formed, especially at low temperatures. The original idea of obtaining a measure of the hygroscopic state of the air from comparing the indications of a thermometer with its bulb moist and dry, is certainly due to Dr James Hutton, the celebrated geologist. This eminent philosopher made use of a single thermometer, which he carried enclosed in a glass case. This form of hygrometer, neglected for thirty years, has at length become an object of philosophic interest in this country and Germany; and two thermometers are now mounted on the same scale, and the indications of the wet and dry bulbs seen at the same time. This instrument has been termed a psychrometer. The most elegant form of an instrument on this principle is the hygrometer of Sir John Leslie, or double air-thermometer; one of the balls is covered with tissue paper, defended by a thin covering of blue silk, while the other is naked; but in order to obviate the effect of light, the dry ball is formed of pale-blue glass. This instrument is described in his original paper, and fully in the article METEOROLOGY of this work. When the covered ball is moistened with water, the evaporation produces a degree of cold in proportion to the dryness of the ambient air; and the air within that ball contracting, the elasticity of the air in the other ball pushes the coloured fluid in the stem towards the covered ball. This transference is marked on a scale, each degree of which is equivalent to $\backslash frac\{1\}\{100\}$th of a centigrade thermometrical degree. This hygrometer does not indicate the absolute dryness of the atmosphere, but only the degree of dryness it has after being reduced to the temperature of the humid ball. In the article to which we have just referred, there are tables given for correcting the indications of Leslie's hygrometer, and showing the point of saturation at different centesimal degrees, and the corresponding degrees of the hygrometer.

The instrument gives excellent comparative results; and two or three minutes are sufficient to bring the fluid to equilibrium, which marks the full effect of the evaporation; and no farther sinking of the hygrometer is produced by keeping the ball bedewed with moisture. In winter, in our climate, it ranges from 5° to 25°; in our summer from 15° to 55°. It has occasionally been observed as high as 80° or even 90°, and in thick fogs it falls to the beginning of the scale: it rises when the weather clears up after rain, and in long tracts of clear warm weather it remains high. It always falls before rain, and remains low during wet weather. When as low as 15°, the air feels damp; when it is between 30° and 40°, we should term the weather dry; from 50° to 60°, very dry; and from 70° upwards, it would be called intensely dry. In a warm inhabited apartment the hygrometer would not be below 50°; but when below 30°, that room would be felt to be uncomfortably damp.

The mean annual evaporation is an interesting subject, Annual for an approximation to which we are mainly indebted to evaporation. The experiments of Dr Dalton and Mr Hoyle, at Manchester, in 1796, 1797, and 1798. A cylindrical vessel of tinned iron, three feet deep and ten inches in diameter, was filled with soil; it was provided with one pipe near the bottom, and another one inch from the top. This was left in this state for a year, until it became covered with grass. Bottles were now applied to the end of each pipe, so as to collect both the surplus water which soaked through the earth, and that which ran off at the upper pipe. The soil

Physical Geography. was soaked with water on commencing the experiment; a rain gage of similar dimensions was placed near the cylinder of tin; and the evaporation was estimated by subtracting the quantity which passed into the bottles from the whole rain.

The mean evaporation for each month, deduced from the three years, was

giving 25.158 inches for the annual evaporation at Manchester. Dr Dalton considers that we ought to add five inches to this for the annual dew, which will raise the evaporation to about thirty inches.

From the nature of our climate, the mean annual evaporation over the globe cannot be estimated at less than thirty-four; and this would make the whole mass of water raised from the surface of the earth equal to 105,614.72 cubic miles, or equal to all the rain that falls upon the globe. The evaporation of any place may be ascertained by exposing freely to the air a given quantity of water in a shallow circular dish, accurately turned, so as to have 100 inches of superficies. It should be so placed as not to receive any rain: on measuring the remainder once a week or once a month, we shall obtain the evaporation.

No. 20.

There is reason to believe that the evaporation from the foliage of trees and from herbage is somewhat greater than from an equal area of water, and that the quantity from the surface of new-ploughed land is almost equal to what arises from water; but we have got no good comparative experiments on this subject.

Having considered the conversion of water into an elastic fluid, let us turn our attention to the mode by which it is re-converted into moisture.

1. Clouds.—Some have accounted for the suspension of so dense a fluid as water in air, on the supposition that the vapour in clouds existed in the shape of hollow spherules, filled with a lighter fluid than the atmosphere; but this suggestion would seem to be devised rather to meet the difficulty than to be supported by observation; and it is possible to explain the considerable durability of clouds without recourse to any such gratuitous assumption.

The experiments of Volta and Cavallo have shown, that when water is converted into vapour, not only is caloric absorbed, but electricity is also developed. The vapour acquires positive electricity, and the remaining fluid possesses the opposite electric state. By the process of evaporation much electric fluid must be carried into the atmosphere; and when the capacity of the vapour for electricity is diminished by its partial condensation by cold in the upper regions of the air, it may be conceived that the spherules of vapour may become surrounded with atmospheres of electric fluid, the mutual repulsions of which may prevent the coalescing of the particles of vapour into drops so heavy as to descend by their gravity on the earth, especially as a stratum of air charged with moisture is specifically lighter than dry air at the same temperature. The

form of clouds, and their different degrees of aggregation, may depend on the electric tension of the atmospheres surrounding the particles of which they are composed; or on a balance between the electric repulsions and the gravitation of the particles for each other. This view is favoured by the generally positively electrified state of the atmosphere when clouds are forming, the heavy falls of rain which accompany electric explosions when the discharge of the accumulated electricity allows the coalescing of the moisture into heavy drops, as well as by the peculiar odour of some kinds of fog, and the highly electric state such fog often exhibits. Beccaria, Romayne, Bennet, and Volta have remarked, that the atmosphere exhibits signs of electricity on the appearance of clouds and fogs. Beccaria always found the electricity of the air positive; and Romayne states, that mere diminution of volume in any substance is constantly attended with increased electrical tension. The various forms of clouds, their aggregation in deep masses piled on each other, or their dispersion in long radiated filaments, may be more easily explained on an electric theory than on any other supposition; and the accidental manner in which solar light is reflected from them to the eye of the spectator gives an infinite diversity to their appearance.

The vulgar have long remarked the association of certain forms of clouds with the state of the weather; and meteorologists have begun to appreciate the value of a descriptive language for indicating their varied appearances. No system of nomenclature has yet been applied to clouds superior to that proposed by Mr Luke Howard. He divides them into seven species. (See Plate CCCCXI.)

1. Cirrus.—A cloud resembling a lock of hair or a feather, composed of parallel, flexuous, or diverging fibres, unlimited in the direction of their increase.

2. Cumulus.—A cloud which increases from above, in dense, convex, or conical heaps.

3. Stratus.—An extended, continuous, level sheet of cloud, increasing from beneath.

4. Cirro-Cumulus.—A connected system of small roundish clouds, placed in close order or contact.

5. Cirro-Stratus.—An horizontal or slightly inclined sheet, attenuated at its circumference, concave downwards, or undulated. Groups having these characters.

6. Cumulo-Stratus.—A cloud in which the cumulus is mixed with cirro-stratus or cirro-cumulus; the cumulus flattened at top, and overhanging its base.

7. Nimbus.—A dense cloud, spreading out into a crown of cirrus, and passing beneath into a shower.¹

The cirrus is generally formed of white radiated streaks, pencilled on an azure sky. It is the most elevated species of cloud. Its motions are slow. It generally indicates a breeze, and often precedes a storm. Horizontal sheets of cirrus, with streamers pointing upwards, often indicate rain; while the depending fringes are the precursors of fine weather.

The cumulus is generally a dense cloud, moving near the surface of the earth. When large, this species often exhibits large intervening spaces of clear sky. It begins in the morning, and obtains its maximum about two P. M., and usually decreases before sunset, breaking up and disappearing before nightfall. It is the prognostic of settled weather.

The stratus is the lowest modification of cloud, often creeping along the ground, in calm evenings, from lakes and rivers, and rising toward the higher grounds. It often, at night, travels over plains, invests the summits of moderate elevations, but usually melts away before the morn-

¹ Philosophical Magazine; Rees' Cyclopædia; Nicholson's Journal, xxx.

ing sun. Few days are more serene than those ushered in by stratus.

Cirro-cumulus consists of dense rounded masses, in close position. It often forms beds at different altitudes in the air. It is transient in winter, and occurs in the intervals of summer showers. A frequent repetition of it is amongst the best prognostics of fine weather.

The cirro-stratus is narrow in proportion to its horizontal length. It seldom retains long the same form. It is often seen very high in the air, especially that variety which is composed of grouped masses, and is chiefly seen in summer evenings. When seen overhead, it has an uniform hazy appearance; but viewed on the horizon, it often seems very dense, from being viewed edgewise. It indicates rain or storms of snow when it is stationary. It often forms the seat of rainbows and halos when descending from the zenith. It often shrouds mountain summits, and in cold weather often descends into plains as soaking dense mist. The cumulo-stratus usually appears in an overcast sky, and may be regarded as a prognostic of rain. It appears to be formed by the successive blending of cumuli, inclosed by streaks of dark stratus. The thunder-cloud is one modification of this species. It often assumes a portentous size, and is contorted into singular forms. Its indications are not uniform. When it appears in the morning, the day is often dry, though overcast; but when much mixed with stratus, it indicates approaching rain, after some interval.

The nimbus is a shower seen in profile as it approaches the spectator from the horizon. The thunder-cloud, on the discharge of its electricity, generally passes into the nimbus.

Mists or Fogs are but the condensation of invisible vapour into minute drops of water. They are generally more stationary on the sea than on the land. They are less frequent in the day than the night, in hot than in cold climates; and they are frequent in approaching the poles. One of the most remarkable stationary fogs is on that part of the Atlantic called the Banks of Newfoundland. The warm waters of the gulf-stream, after sweeping along the shores of the United States, and warming the superincumbent air, are encountered at Newfoundland by the polar currents, and the colder air over them. The consequence of this intermixture of the aerial currents at different temperatures, is the diminished capacity of the mixture for moisture, and the consequent deposition of a portion of its vapour in the form of mist. When the cold producing condensation of atmospheric moisture is intense, instead of mist, a deposition of minute particles of ice may take place, and will cover with delicate crystallizations the hair and clothes of the traveller, or the twigs and branches of trees. When the air is saturated with moisture, sudden condensation assumes the form of

Rain.—Amongst the numerous speculations on the production of rain, one of the most ingenious is that proposed by Dr James Hutton. He evidently considered the water as dissolved by the air; but if we merely consider the term "complete saturation with moisture" as meaning the dew-point, and the quantity suspended by air as indicating the quantity of vapour existing in the atmosphere at a given temperature by its own elasticity, this phraseology will produce no error. Hutton ascribed rain to the intermingling of air saturated with moisture at different temperatures. His views were communicated to the Royal Society of Edinburgh in 1787; but, from his contenting himself with a mere enunciation of his theory, it has been scarcely noticed by meteorologists, until illustrated in the articles METEOROLOGY and HYGROMETRY. Sir John Leslie has shown, that the mere commixture of air at different temperatures would produce but a small part of the effect perceived. It is proved by experiment that the capacity of air for moisture, or the quantity of vapour in a given space, increases

in a much higher ratio than the temperature; and hence, if the space occupied by the mixture of masses of moist air, at different temperatures, have the mean temperature of the mixture, or one below it, a deposition of moisture or of rain will follow; but the whole effect would be smaller than what is often observed to take place, unless we suppose that two currents of air are moving, with considerable velocity, in opposite directions.

The objection to this simple view of the question is, that there are often heavy falls of rain when the clouds emitting it are evidently in very slow movement; and we have long believed, that without calling in the aid of electricity, as already stated when treating of the formation of clouds, no theory of rain will be satisfactory.

The hygrometric water in the air seems to owe its suspension to the repulsions among the electric atmospheres of its particles: when that electricity is withdrawn by any cause, the particles of water, no longer kept asunder by their electricity, coalesce, and descend by their gravity with a force proportional to their quantity and the reduction of their electric tension. If the electric fluid be slowly withdrawn, there is neither lightning nor any explosion; but if the accumulation has been great, and the abstraction sudden, we have the phenomena of thunder and lightning. On this view, evaporation is to be considered as the grand

mode by which the atmosphere acquires electricity, and the accumulation seems in general in proportion to the intensity and duration of the evaporation. Hence lightning is most common and vivid in warm latitudes; and there is seldom any considerable electric discharge from the atmosphere, without a concomitant fall of rain, which is usually proportional to the violence of the thunder-storm.

Rain falls in drops of a size varying from $\backslash frac\{1\}\{2\}$th to $\backslash frac\{1\}\{4\}$th of an inch in diameter; and it has been calculated that their ultimate velocity is in the duplicate ratio of their diameters. When formed high in the atmosphere, or passing through very cold strata of air, they congeal into hail or into snow.

To ascertain the quantity of rain which falls in any place in a given time, various species of ombrometers or rain-gages have been employed. The most usual form of the gage is that represented in No. 21. It consists of a cylindrical vessel of copper, usually $\backslash frac\{3\}\{4\}$th of an inch in diameter. Upon its upper part, a funnel of the same metal, with an upright brim exactly one foot in diameter, is fitted; a hollow vessel of metal, or a piece of cork, serves as a float, to support a rod of boxwood graduated in tenths of an inch. The water which falls into the funnel finds its way into the cylinder, and raises the float. The quantity is read off the rod at a stay placed across the funnel, and from the relative proportion of the cylinder and funnel, each $\backslash frac\{1\}\{10\}$th of an inch is equivalent to $\backslash frac\{1\}\{100\}$th of an inch of water entering the mouth of the funnel.

Another very convenient form is represented in No. 22, where the

Physical height of the column of water in the cylinder is seen in a Geography. glass tube connected with the cylinder below. A scale is applied, as in the figure; and when the observation is made, the water is let off by the stop-cock.

Mr Luke Howard's rain-gage is seen in No. 23. The funnel is fitted on the neck of a glass bottle, by means of an exterior tube, which is soldered to the funnel, and prevents the entrance of water into the vessel, except through the mouth of the funnel. The quantity of rain is ascertained by a graduated glass measure, the divisions of which correspond to 100ths of an inch, falling on the mouth of the funnel. This same measure will serve to ascertain the quantity evaporated in a given time, provided the mouth of the vessel, No. 20, is exactly equal to the mouth of the rain-gage.

There are considerable practical difficulties in the proper position of a rain-gage. It is found, that if one be placed on the ground, and another on the top of a building, that which is in the lowest position will almost always indicate more rain than the upper. Thus, the quantity collected in a gage on the top of York Minster, from February 1833 to February 1834, only equals 14.963 inches, while perfectly similar instruments placed on the top of the museum of that city, and on the ground, gave relatively 19.852, and 25.706 inches. The elevations of the first and second stations are 212 feet 10½ inches and 43 feet 8 inches above the surface. These differences are too considerable to be attributed to any thing but some imperfection in the instrument when much exposed to gales of wind; and it probably arises from eddies being formed round the rim of the funnel, which divert part of the water. On the whole, the most accurate estimate of the quantity of rain will be obtained from instruments placed on the ground, or near it. The larger the recipient surface, the more accurate will be the indications; and on this principle, General Sir Thomas Brisbane, president of the Royal Society of Edinburgh, collects the rain from the roof of his observatory (the area of which has been accurately ascertained), for measuring the quantity of rain which falls in his vicinity.

When the quantity of rain which falls at any place has been regularly kept for several years, it is easy to obtain the mean annual quantity at that place. The multiplication of such observations has proved, that the mean annual quantity is greatest in tropical climates, and diminishes as we recede from the equator. Thus, the mean annual rain at

In general the number of rainy days is least where the class of rain is the greatest, and the dampness of any climate must not be inferred from the annual quantity of rain.

A country may be involved in almost perpetual fogs, or have many days of drizzling rain, in which the number of inches of rain throughout the year may be very small.

The quantity of rain, however, depends upon other circumstances than the latitude. It is much influenced by the vicinity of mountains in every climate. Thus,

Yet in Demerary as much as six inches of rain have been observed to fall in twelve hours. Similar observations occur in our own country.

In Great Britain, it is generally found that much more rain falls on the western than on the eastern coasts. Thus,

The latter town lies in a sheltered nook, or perhaps its annual quantity would have been higher; for,

It is scarcely necessary here to dwell on various extraordinary substances which have sometimes fallen from the air. Showers of sulphur have been mentioned by different observers. Wormius records such a shower as occurring at Copenhagen in May 1646; but the same phenomenon occurred in May 1804; and the philosophers of our own age, on analyzing this deposit, recognised it to consist of a vegetable pollen, resembling the powder of lycopodium. This substance fell in the night time, and, being phosphorescent when observed, gave considerable alarm. At Bordeaux, in 1761, a shower of a yellow powder was observed, which, however, was recognised as the pollen of the pine forests in the neighbouring landes, carried up into the air by a violent gale. Of a similar nature, in all probability, was the yellow rain which covered the Lake of Zurich in 1677.

The supposed showers of blood which have been recorded in some ancient chronicles are now ascribed to the same source as the red snow of Greenland and the Alps, that is, to the red globules or seeds of the uredo nivalis; or to minute red insects, which are sometimes carried down in great quantity with rain. The appearance of drops of blood falling from the air has occasionally been produced by the red excrement of insects. That small frogs and fishes have occasionally descended with rain has been admitted, as such animals, and even much more ponderous matter, have been elevated into the atmosphere by whirlwinds.

The showers of stones which have been noticed by several ancient writers as happening in Italy, Greece, and Asia Minor, are now believed to belong to the class of facts described under the article METEOROLITE.¹ These are mi-

¹ See also Chladni, Mémoire, in Journal de Physique; Howard, in Phil. Trans. 1802; Bibliothèque Britannique, 1803; Nicholson's Journal, 1803, 1805, 1809, 1810; Annales du Muséum d'Hist. Nat. 1803; Izarn's Lithologie Atmosphérique, Edin. Phil. Journal, vol. 1.

neral bodies, which, however they find their way into the atmosphere, have undoubtedly fallen upon the earth; and, however remote the places in which they descend, they are all distinguished by one remarkable similarity, viz. their containing an alloy of iron and nickel, generally with twenty-five of the former to six or eight of the latter.

Dew.—The phenomena and cause of this deposition have engaged the attention of philosophers from the time of Aristotle; but their speculations present little that is satisfactory until the experimental investigations of Dr Wells. One of the most remarkable circumstances attending the deposition of dew is, that the bodies on which it is found have invariably a lower temperature than the ambient air, as was first pointed out by Dr Patrick Wilson of Glasgow, who ascribed the coldness to the deposition of dew; but Dr Wells has proved that the cooling of these bodies below the temperature of the air always precedes the formation of the dew, and is, in fact, the cause of this aqueous meteor. His theory is founded on Leslie's view of the heat which is sent off from the earth toward a clear sky; and on this simple principle Wells has been enabled to give a satisfactory explanation of the phenomena of dew.

The essential requisites for the deposition of dew are five in number, viz.

1. An atmosphere replete with moisture. As dew is a deposition of water from the air, it obviously must be in excess ere it is deposited. In Egypt, the south winds, blowing over extensive tracts of sandy desert, deposit no dew; but when the Etesian or north wind blows, loaded with moisture from the Mediterranean, the deposition of dew in lower Egypt is sufficient to soak through the garments of the inhabitants; and in the arid plains of New Castile the dews are far less copious than on the coasts of Andalusia.

2. There must be a considerable difference between the temperature of the earth in the night and the day. Other things being the same, the deposition is greatest when a cool evening succeeds a sultry day. In our climate, the dews are most abundant in spring and autumn, when the difference of temperature is greatest; but as the ratio of the increase of moisture in the air is greater than the increase of heat, hot climates have more copious dews than Great Britain, even though the difference between diurnal and nocturnal temperature be less in the former. Thus we have found the dews in the south of Spain, and in Morocco, far more copious than here; a short exposure about sunset in those countries completely soaking one through the clothes.

3. But the most essential requisite for the deposition of dew is, that the bodies upon which it forms should have their temperature considerably lower than that of the ambient air. In repeating Wilson's experiments, Dr Wells found that all bodies on which dew formed were colder than the air by $10^\{\backslash circ\}$ or $15^\{\backslash circ\}$. Thus, in clear weather, he observed that a thermometer laid on grass was $12^\{\backslash circ\}$ lower than one laid on mould, and sixteen and a half below one on a gravel walk. He always found that the coldness of the body preceded the deposition of dew; and this is illustrated by the deposition of moisture which immediately takes place on pouring into a glass cylinder, placed in a warm, moist room, very cold water.

4. A serene and cloudless sky is also very favourable to the formation of dew. Wells found, that although the atmosphere be in other respects favourable to the deposition, yet little or none is formed if the sky be veiled in clouds; and, in such cases, the temperature of objects on the surface was near that of the air. He points out how clouds act as a screen to intercept the emanations be-

tween the earth and the upper regions of the air, which are well known to be the abodes of perpetual congelation. The same effect was produced by interposing screens of an opaque material between the sky and the surface of the earth. Thus a thermometer, laid on a table, indicated a lower temperature than one placed on the ground under the table; and a lock of wool laid on the table collected ten grains of dew, whilst an equal weight of wool placed under the table gained only two grains. Fogs, acting also as screens, intercept the influence of the sky; and though in themselves moist, they are unfavourable to the deposition of true dew, which may be regarded as water separated from the inferior stratum of air; but fogs, like rain, are precipitated from the general mass of the superincumbent air.

The effect of clouds in preventing the radiation of heat is beautifully illustrated by carrying Professor Leslie's elegant invention, the Aethroscope, into the open air. Every cloud that passes the zenith affects this delicate instrument, which rapidly loses its heat when exposed to the clear sky; or the effect is still more conspicuous if the operator hold a parasol, and alternately extend the aethroscope beyond the shade and withdraw it under the parasol.

5. Serene and calm weather is favourable to the formation of dew; a fact which had been noticed by Aristotle. The influence of wind seems to depend on the rapid succession of aerial particles preventing the full influence of the cooling medium on the general temperature of an inferior stratum of the atmosphere. Every circumstance which favours radiation is favourable to the deposition of dew, such as a rough surface, and dark colour; proving the dependence of dew on the cooling of the objects on which it is deposited.¹

Dew often partakes of the sensible qualities of the honey and dies upon which it is found; hence dew-water is very apt to putrefy. Some substances found on plants have been erroneously confounded with dew. What is termed honey-dew generally owes its qualities to the saccharine exudation from the bodies of the insects called Aphides. The jelly-dew is believed to be the original form of a cryptogamian vegetable production, the Tremella nostoc of Linnaeus; a membranous, pellucid, greenish-yellow matter, about one or two inches in width, which is at first moist and soft to the touch, but dries into a blackish membrane.

Snow.—When the congelation of moisture takes place slowly, the flakes of snow assume very symmetrical figures, the varieties of which have been well represented in the work of Scoresby. They usually present modifications or combinations of the hexagonal prisms, often consisting of a star of six rays, formed of prisms united at angles of $60^\{\backslash circ\}$; and from these other prisms shoot, at similar angles, giving the whole a plumose appearance, of exquisite beauty, and often of surprising regularity.

The quantity of snow which falls in any place is not proportional to the height of its latitude. In many temperate countries, the successive falls of snow, during a winter, may often greatly exceed in depth what falls in the same period in countries between the $65^\{\backslash text\{th\}\}$ and $80^\{\backslash text\{th\}\}$ degrees of latitude. In fact, it bears some proportion to the evaporation in countries where the winter temperature is below the freezing point of water. Snow is extremely useful, in cold climates, as a loose and spongy covering of a white colour, in preventing the loss of terrestrial heat by radiation, and in defending the surface from the frigorific influence of very cold winds.

Hail differs from snow chiefly in its state of aggregation, and its far less regular crystalline form. It appears, indeed, to be the production of sudden and intense cold in the

¹ Wells's Essay on Dew, London, 1814.

² Scoresby's Arctic Regions, vol. i.

Physical higher regions of the atmosphere, and therefore often occur in warm climates, where snow is utterly unknown. A hail shower is often ushered in by a rattling noise, heard over head, before almost any hailstones have reached the ground. The ordinary size of hailstones with us varies from $\backslash frac\{1\}\{10\}$th to $\backslash frac\{1\}\{4\}$th of an inch in diameter. Occasionally they are much larger; as in the remarkable hail-storm described by Dr Neill, which occurred during a thunder-storm in the Orkneys on the 24th of July 1818.¹ Mingled with ordinary hail were enormous masses of ice, some as large as the egg of a goose; by which animals were killed, and individuals wounded. An enormous hailstone is recorded to have fallen, amongst many very large masses, at Handsworth House, near Birmingham, during a thunder-storm, on the 8th of June 1811. It consisted of a cuboidal mass $6\backslash frac\{1\}\{2\}$ inches in diameter; and, from the engraved sketch, it resembled a congeries of frozen balls about the size of walnuts. On the Continent of Europe, the large hail of summer is frequently exceedingly destructive to vineyards; and in some districts of France, where the damage has often been great from thunder-storms, accompanied by large hailstones, the ingenious device of the Paratonnerre has been employed to prevent the accumulation of electricity over these valleys. The contrivance consists in the erection of pointed electric conductors on the surrounding hills, by which the lightning is silently and gradually abstracted from the air, and dangerous accumulations over the included districts generally prevented. In hot climates, the descent of large masses of frozen water is still more common than in Europe, especially within the Calms off the western coast of Africa.

Winds.—The influence of mechanical causes in producing those aerial currents termed winds has been generally admitted, from the ages when they were poetically imagined as viewless forms, imprisoned or let loose from the caverns of Æolus, to the present day, when they are considered as the consequences of partial changes in the density of the atmosphere, arising from the unequal distribution of heat over the surface of the globe, or from alterations on its hygrometric and electric states. But the manner in which their force and direction are affected by these circumstances, in those sudden and local storms which frequently agitate the air, is difficult of explanation, although the general phenomena may admit of a very probable solution.

Register anemoscopes. The dependence of the weather on the direction and force of the winds has given rise to many inventions for indicating their variations. The direction is shown by the well-known vane, or simple anemoscope, the movements of which are referred to the cardinal points by a fixed index, or by a plate marked with the rhumbs of the mariner's compass. To save the trouble of frequent daily observations, the labour of the meteorologist may be materially abridged by the following methods of obtaining maxima and minima of the direction of the winds.

In the first, No. 24, the vane b, fixed to the axis a, causes the latter to revolve in a socket, in the centre of a writing slate marked with the rhumbs of the compass. The arm c, jointed to it, carries a pencil d, which records the progress

of the vane on the slate. The pressure of the pencil may be regulated by sliding the ball e on the arm; and, to diminish friction, the socket consists of a flint, with a smooth hollow cavity, and the axis is kept perpendicular by moving between friction rollers.

In the second, No. 25, the same effect is shown on a vertical dial, by converting the horizontal motion of the vane into a vertical movement of the index. This is produced by one bevelled wheel e on the axis of the vane, and a similar one c on the axis of the index f. The maxima and minima are shown by two moveable arms gg, concentric with the index, but pushed to either side by a stud on the lower surface of the index. These instruments, devised long ago by Traill, and published in Professor Jameson's Journal for 1837, sufficiently answer the object for which they were intended.

The varying force of the wind is ascertained by the anemometer. The principle of several such instruments is the direct pressure of the wind on a given surface, kept by a vane perpendicular to the direction of the breeze. Sometimes this pressure is measured by raising a weight; in some instruments by the degree of compression of a spiral spring; or, as in the ingenious anemometer of Mr Waddel of Leith, by the compression of a slender cylindrical bag filled with quicksilver, and connected with a vertical tube one foot in length, with a bore of $\backslash frac\{1\}\{20\}$th of an inch. The force of the wind is estimated by the height of the column of mercury. A practical difficulty arises from the quickness of the oscillations, which renders it not easy to read off the indications of the instrument during a smart gale.

The same objection in some measure applies to the elegant pocket anemometer of Dr Lind. It consists of two equal glass tubes, united below by a narrow connecting piece, as in No. 26, and having between them a scale of equal parts. The tubes are filled to the middle with water; and as the instrument is made to revolve freely round the iron spindle a, its body acts as a vane, to keep the orifice of the knee brass tube always directed to the wind. The breeze received at that aperture depresses the liquid in that limb, and raises it in the other to a certain point on the scale, according to the force of the breeze. When it is very strong, the detached piece of brass b is joined to the top, and an allowance made for the effect on the column. The sum of the two columns is the height of a column of water which the wind is capable of sustaining; and every body opposed to that wind will be pressed on with a force equal to the weight of a column of water of that height, and with a base equal to that of the surface exposed to the wind. In the original instrument the bore of both legs = $\backslash frac\{1\}\{10\}$ths of an inch, the connecting tube was $\backslash frac\{1\}\{10\}$ths, and the length of the whole was six inches. Lind gave the following table as its indications.

¹ Edinburgh Philosophical Transactions, vol. ix.

This instrument is, upon the whole, the most convenient anemometer we possess, and has the advantage of portability. The oscillations of the liquid might be farther checked by twisting spirally the tube of communication, as in the marine barometer.

A very ingenious anemometer has been constructed by Mr Richard Adie, late of Edinburgh, now of Liverpool. An index E is moved by the rising and falling of a cylinder A, closed at one end, and inverted in a larger cylinder G, containing water, on the principle of a gasometer for chemical experiments. The inverted vessel is suspended by a cord passing

over a pulley B fixed on the axis C of the index. The counterpoise is a weight W, the cord of which, wound on a fuse or snail D, equalizes the counterpoise in the different positions of the internal cylinder. The influence of the wind on the instrument is produced through a tube, as in the figure. One end terminates in the interior cylinder, and the other in a kned tube moving freely round by means of the attached vane. A very gentle breeze is sufficient to elevate the cylinder, and the revolution of the index marks the comparative force of different winds. This anemometer has been found to act well, and has been put up at several of the northern lighthouses.

The last anemometer we shall notice is that invented by Professor Whewell, who exhibited it to the British As-

sociation. It consists of a small wind-wheel of eight vanes, directed to the wind by an attached vane. By a train of Geography-toothed wheels and endless screws, the rapid revolutions of the wheel are converted into a slow vertical motion, by which a pencil is made to trace a downward line on the surface of a vertical cylinder having the same axis as the vane. The length of this line records the whole amount of the aerial current that passes over the place of observation, and the part of the circumference of the cylinder on which the trace appears indicates the direction of the winds. Three of these instruments are, by direction of the Association, on trial in different places, and the results will be communicated in the present year (1838).

In Smeaton's valuable paper upon the moving power of Force of wind and water (Philosophical Transactions, vol. li.) there the wind is the following table of the force of the wind, calculated by his friend Mr Rouse; which the celebrated engineer considers as entitled to confidence, especially in velocities below fifty miles per hour.

The third column is calculated according to the squares of the velocities of the wind, which Smeaton considers as generally correct.

Trade-winds.—In those portions of the Atlantic and the Pacific Oceans between the tropics, the wind has, through-out the year, a direction from the east. The following explanation agrees pretty well with the facts. The dense cool air of the polar and temperate regions naturally tends to displace the lighter warm air over intertropical regions, which, heated by the sun, produce vertical movements in the superincumbent atmosphere. Hence the ascending current of warm air between the tropics has its place supplied by the denser air from either side; and thus two currents are set in motion, from the poles towards the tropical regions. The velocity of every point on the earth's surface increases from the poles to the equator. But as the particles of these currents cannot at once acquire a velocity equal to the continually accelerating velocity of the parts of the earth's surface over which they arrive in this transit, the northern and southern currents will gradually seem to acquire a motion in an opposite direction to that of the rotation of the earth, that is, both will gradually decline to the west, assuming in the northern hemisphere the character of a N. E., and on the south side of the line of a S. E. wind; whilst both will become more easterly as they approach the equator. This effect of the augmenting velocity of the earth's surface, in approaching the equinoctial line, is increased by the continual movement of the point over

Physical Geography which the sun is vertical, and consequently his heat the greatest, to the west; and the effect of the gravitation of the sun and moon on the atmosphere, as was shown by D'Alembert, must directly tend to increase the force of the easterly wind. The influence of these luminaries on our atmosphere is corroborated by the observations of Bacon, Halley, and Gassendi, on the frequency of storms about the equinoxes, or at full and new moon, and the general occurrence, in calm weather, of light airs of wind at the time of high water. When these causes are not counteracted by the superior rarefaction of air over land heated by the sun's rays, the easterly winds blow with much regularity, as in the great oceans; and, from their important influence on navigation, they have obtained the denomination of trade-winds. They extend to about 30° on each side of the equator; but, from the less density of the atmosphere in the northern hemisphere, arising from the greater quantity of land to be acted on by the sun's rays in that half of the globe, the line which marks the blending of the S. E. and N. E. trade-winds is between two and three degrees north of the equator.

If the surface of the globe had been one expanse of water, the trade-winds would have blown all round the equator, and the only other winds known would have probably been N. and S. winds near the poles, gradually bending from E. to W. as they approached the tropics. But the air over the land being much more heated by the sun's rays than that over the sea, the vicinity of large islands or of continents causes interruptions to the regularity of the trade-winds. Thus, on the coasts of Africa, within the regular limits of the trade-winds, westerly gales are not unknown; and a southerly wind prevails along the western shores of that continent, from Cape Palmas to the Cape of Good Hope; a consequence of the form and climate of Northern Africa, which continues the southern polar current far beyond its limits in the free ocean.

A similar cause appears to give rise to the Calms on that coast, from 10° to 4° N., a tract avoided by seamen, on account of these calms, and sudden alternations, with tremendous storms of thunder and lightning, to which it is subject. The general direction of the trade-winds is affected by their approach to the shores of America; they are expanded, chiefly owing to the resistance they meet with from the land, and also by the additional rarefaction of the air re-acting upon that continent. Precisely similar disturbances from the position of land produce

Monsoons.

The Monsoons of the Indian Ocean. The eastern trade-wind blows regularly in that sea between the parallels of 10° and 30° S.; but from these parallels northwards, or on the sea between Sumatra and Eastern Africa, the course of the trade-winds is reversed for half the year, from April to October, during which period there is a constant wind from the S. W. During the other six months the regular trade-wind, from the N. E., resumes its course. These winds are termed the S. W. and N. E. monsoons, from a Malayan word signifying seasons. The latter is produced by the same causes as the general trade-wind. The other appears to be occasioned by the great rarefaction of the atmosphere over the extensive regions of Eastern Asia when the sun is north of the equator, and the dense air of the Indian Ocean rushes in to occupy the place of the ascending current.

This disturbance of the trade-wind is attended with very remarkable effects on the climate of the country where it prevails. It is the harbinger of the rainy season in India, which is ushered in by violent thunder-storms. A most striking picture of this change of monsoon is given by Mr Elphinstone in his Account of Ceylon. "The most remark-

able rainy season is that called in India the S. W. monsoon. It extends from Africa to the Malayan peninsula, and deluges all the intermediate countries, within certain lines of latitude, for four months in the year. In the south of India this monsoon commences about the beginning of June, but it gets later as we advance towards the north. Its approach is announced by vast masses of clouds, that rise from the Indian Ocean, and advance towards the N. E., gathering and thickening as they approach the land. After some threatening days, the sky assumes a troubled appearance in the evenings, and the monsoon in general sets in during the night. It is attended by such a thunderstorm as can hardly be imagined by those who have only seen that phenomenon in a temperate climate. It generally begins with violent blasts of wind, which are succeeded by floods of rain. For some hours lightning is seen almost without intermission; sometimes it only illumines the sky, and shows the clouds near the horizon; at other times it discovers the distant hills, and again leaves all in darkness; when, in an instant, it re-appears in vivid and successive flashes, and exhibits the nearest objects in the brightness of day. During all this time the distant thunder never ceases to roll, and is only silenced by some nearer peal, which bursts on the ear with such a sudden and tremendous crash as can scarcely fail to strike the most insensible heart with awe. (Malabar is the province most distinguished for the violence of the monsoon.) At length the thunder ceases; and nothing is heard but the continued pouring of the rain, and the rushing of rising streams. The next day presents a gloomy spectacle; the rain still descends in torrents, and scarcely allows a view of the blackened fields: the rivers are swollen and discoloured, and sweep down along with them the hedges, the huts, and the remains of the cultivation which was carried on during the dry season, in their beds.

"This lasts for some days, after which the sky clears, and discovers the face of nature changed, as if by enchantment. Before the storm, the fields were parched up, and, except in the beds of the rivers, scarce a blade of vegetation was to be seen. The clearness of the sky was not interrupted by a single cloud, but the atmosphere was loaded with dust: a parching wind blew, like a blast from a furnace, and heated wood, iron, and every solid material, even in the shade; and immediately before the monsoon this wind had been succeeded by still more sultry calms. But when the first violence of the monsoon is over, the whole earth is covered with a sudden but luxuriant verdure: the rivers are full and tranquil; the air pure and delicious; and the sky is varied and embellished with clouds. The effect of this change is visible on all the animal creation, and can only be imagined in Europe by supposing the depth of a dreary winter to start at once into all the freshness and brilliancy of spring. From that time the rain falls at intervals for about a month, when it comes on again with great violence, and in July the rains are at their height: during the third month they rather diminish, but are still heavy; and in September they gradually abate, and are often suspended till near the end of the month, when they depart amid thunders and tempests, as they came."

Sea and land breezes are occasioned by the same causes Sea and land breezes which produce the S. W. monsoon, and are chiefly experienced on tropical sea-coasts.

The sea-breeze commences about 10 A. M., and continues throughout the day, till towards 6 P. M., when it gradually dies away, and about eight in the evening is succeeded by a much fainter breeze from the land, especially in those islands which have considerable mountains. It is well

known in Jamaica, but less so in the smaller Antilles, and scarcely at all on the flat coasts of America. It continues during the night, and usually dies away before six or seven in the morning; but it is much less regular in its periods than the sea-breeze. The latter, too, occurs in places where the land-breeze is quite unknown. An admirable description of these is contained in the voyages of Dampier, and given at length in the article METEOROLOGY. The sea-breeze is more or less felt on the coasts of all warm countries. We have had occasion to observe with what regularity the sea-breeze sets in from the S. W. during the summer at Gibraltar, and it blows with considerable force throughout the greatest part of the day.

The explanation of the sea-breeze is the rarefaction of the air by the heat of the land, and the rushing in of denser air from the sea to supply its place; and as the influence of the sun decreases, this breeze dies away. The land-breeze is chiefly produced by the descent of air from mountainous regions, flowing, by its gravity, towards the sea, over which the atmosphere is less liable to suffer extreme rarefaction and condensation than the air over the land.

In temperate climates the winds are far less regular than in tropical countries. The general causes which produce them act with less force and constancy; and the disturbing causes, such as the vicinity of land, alteration in density from moisture and dryness, act with more uncontrolled energy. The most prevalent winds in Great Britain are from the S. W., especially on the western side of our island. The Royal Society registers give the following average for the winds throughout the year at London.

Or, with Professor Playfair, if we group the S. with the westerly, and the N. with the easterly winds, we shall have the following results, from the Edinburgh register of Playfair, and the Glasgow register of Dr Meek.

These determinations of the direction of the winds refer to the surface of the earth; but there can be no doubt that the direction of the aerial currents is often different in the upper regions of the air. The permanence of very elevated clouds, their occasional movements in directions opposite to those of the winds below, and the information obtained from aeronauts, all prove this point.

There are several remarkable modifications of the wind, which deserve notice in this place.

The Sirocco of Italy and Sicily is a S. E. or S. wind, which, heated on the sandy wastes of Arabia and Libya, becomes occasionally moist in its passage across the Mediterranean, and oppresses the inhabitants of Italy, Malta, and Sicily with excessive languor, and a sinking of their mental energies. When it sets in, it causes a very sudden and powerful rise of the thermometer, and is accompanied by a haze which obscures the pure sky of those southern countries; the sun appearing dimmed and shorn of his beams.

The Solano of Spain is only a modification of the Sirocco. We have experienced it as most oppressive on the eastern shores of Spain; and it is greatly detested by the natives, who gravely remark, that "no animals except a pig and an Englishman are insensible to the Solano." The Italian condemnation of a stupid work, "era scritto in

tempo del Sirocco," is not more pointed than the Spanish adage, "no rogar alguna gracia en tiempo de Solano," not to ask any favour during the Solano; and both proverbs sufficiently indicate the belief of the people of southern Europe in the disagreeable qualities of the S. E. wind. We have observed fine dust deposited from the Solano or Levanter at Gibraltar, and to this we partly attribute the haze with which it is accompanied. If this dust be brought with the Solano from Africa, it is less surprising than the following instance of dust carried by the easterly wind into the Atlantic. On the morning of the 19th January 1826, when the Clyde East Indiaman was on her voyage to London, in latitude 10. 40. N., longitude 27. 41. W., her rigging was observed to be covered with an impalpable powder of a brownish colour, and on unfurling the sails at two P. M. to catch the breeze, they emitted clouds of dust, which had lodged in them during a strong gale from the E. and N. E. In this case the nearest land in that direction was about seven hundred miles distant.

The Khamisin of Syria, the Samiel of the Turks, and the Si-Khamisin of the Arabs, is well described by Volney. It is a wind in or Simûn, no degree poisonous, except in as far as it is dangerous from its extreme heat and aridity; qualities it derives from blowing over sandy deserts intensely heated by the sun. Denon states, that it is loaded with impalpable sand, which penetrates into the closest packages. These winds derive their Syrian name from their prevalence during the fifty days about the equinoxes. They produce difficult respiration, a shrivelled skin, and a distressing sense of heat. Volney compares the effect of them to the sensation produced by the hot air from a baker's oven. They sometimes blow in squalls, and in Egypt usually continue for about three days at a time. In that period vegetation is withered, the inhabitants shut themselves up in their houses or their tents; and the Simûn is dangerous to persons of a full habit, from the sudden entrance of this heated and parched air into the lungs. In the sandy deserts it is greatly dreaded; but fortunately its extent is not considerable, and it seldom lasts long. It has been well described by Chardin, Sir John Malcolm, and Badhin.

The Harmattan of the west coast of Africa is of the same kind, and produced by the same causes. The extreme dryness of this wind has been mentioned by all observers. It usually blows from the Great Sahara towards the coast, from the N. E., and, like the Solano, Sirocco, and Simûn, is loaded with dust. The dust which fell on the vessel just mentioned was from a violent Harmattan. The only difference between the Sirocco and these other winds appears to be, that the Sirocco obtains some moisture in crossing the Mediterranean; but they all seem to be mere modifications of aerial currents produced by the same causes.¹

The Tornadoes, Typhoons of the Chinese Seas, the Ox-eye Tornado, of the Cape, and the Hurricanes of the West Indies, are violent and extraordinary agitations of the air, generally accompanied by thunderstorms, and usually of short duration. The destructive violence of the three first has been well described by several navigators; and of the hurricane which desolated Barbadoes in the year 1780 we have a short but good account by Sir B. Blane.²

The name of Whirlwind has been given to eddying currents, seemingly depending on electricity, which sometimes rage with surprising fury in particular spots, tearing up trees by the roots, and overthrowing buildings. The extent of the whirlwind is usually confined; and as it passes through a forest it has been observed to form a lane or long track of inconsiderable breadth. The singular columns of moving sand, described by Bruce as occurring in Africa, appear to have been the effect of whirlwinds, or probably analo-

¹ For an account of the Harmattan, see Norris in Philosophical Transactions, lxxi.; and Lind on Hot Climates.

² Edinburgh Philosophical Transactions, vol. i.

Physical Geography. gous to the meteor termed a Waterspout, which seems to be decidedly an electrical phenomenon.

A curious account of waterspouts is contained in Michaud's Mémoire in the Transactions of the Academy of Turin, which has been translated in Nicholson's Journal (vol. i. 4to series). The conical pillar would appear to consist of condensed vapour, which often has the deep indigo-blue tint of the cloud from which it proceeds. As this column descends to the sea, the spot below it becomes agitated, and, according to some observers, the waters are bodily drawn up towards the cloud. But this opinion seems to be doubtful; for it is stated that the water discharged on the bursting of a waterspout is always fresh, and therefore any water derived from the sea must have passed into the air in the form of vapour. Besides, they generally dissolve merely into mist on touching the land.

Iris. On the luminous meteors which appear in the air our space will not permit us to expatiate. The Iris is well known to be produced by the refraction of white light by the spherules of water in a cloud, as has been beautifully and satisfactorily demonstrated by Newton. The explanation of the single and the double iris will be found in every treatise on natural philosophy.

Aurora. The Aurora Borealis, and the similar luminous Southern Meteor, are still objects of doubtful origin. The brilliancy of the colours is subject to periodical increments and intervals of decline; and, after a considerable period of obscurity, it has, within the last few years, again begun to blaze forth with increasing splendour. We have seen all the prismatic colours present in an aurora; and many years ago, in the Orkney Islands, we were enabled at midnight to distinguish letters of a large print by its light. From the distance to which the same aurora has been seen, the meteor must often be very high above the earth, perhaps in the utmost limits of our atmosphere; but at other times it appears to be nearer to the earth, and we have persuaded ourselves that we have heard a noise accompanying the sudden coruscations of a brilliant aurora.

Halo and parhelion. The Halo, the Parhelion, and Paraselene, are optical phenomena produced by refraction of light by vapour, or minute spicula of ice floating in the air. Parhelia have been described by Scheiner, Gassendi, Delahire, Cassini, Halley, Muschenbroeck, and Epinus.¹

Lightning has been considered, with other electrical phenomena, under ELECTRICITY. The identity of lightning with common electricity, as excited by the machine, was so completely established by the celebrated Franklin, that no doubt has since existed on the subject; and the recent discoveries of Oersted and Seebeck have indicated a close affinity between the phenomena of electricity, heat, and magnetism. As the electrical state of the atmosphere is an important subject of inquiry, we beg leave to recommend to the philosophical traveller some instrument for ascertaining this quality of the atmosphere; and perhaps no better portable electroscope has yet been devised than the pocket instrument of Cavallo.

Bolis, St Elmo's fire. The Bolis, St Elmo's Fire, Castor and Pollux of the ancients, appear to be electric meteors. The former we have twice seen. The last time was on the 18th of September 1835, at eleven P.M., when at sea. The head appeared an

elliptical ball of the most intense, brilliant, bluish-white light; the train had a tint of azure mixed with reddish, which sufficiently distinguished it on the clear sky; and sparks proceeded from its tail. Its motion was nearly horizontal, from S.W. to N.E., at a considerable height. It was visible for five or six seconds, and disappeared in a dense cloud. It was observed by four gentlemen standing together on the deck, and we thought that it emitted a crackling noise as it proceeded, with a majestic uniform movement.

The more common meteor, the Shooting Star, as it is vulgarly termed, seems to be an electrical ball of fire, at a considerable height in the atmosphere. Its general course is from the zenith towards the horizon, which it seldom appears to reach.

The Ignis Fatuus, or Will o' the Wisp, is a meteor which apparently is connected with decaying animal or vegetable matter; and probably is produced by the disengagement of phosphuretted hydrogen. It is frequently seen in boggy grounds, as a flickering, unsteady light, in irregular motion. It sometimes plays over dunghills; and has occasionally, to the extreme terror of the ignorant, appeared as a lambent flame in church-yards. Dr Derham saw it once playing about a dead thistle. The most extraordinary meteor of this kind is that described by Signor Beccari, in the Philosophical Transactions (vol. xxxvi.). In the low grounds to the north and east of Bologna, ignes fatui of remarkable splendour were at that time seen almost every dark night. That in the fields of Bagnara, to the east of the city, was especially brilliant. It appeared and disappeared suddenly; it rose and fell, and often hovered about six feet from the ground; it sometimes contracted its volume, then expanded or divided into two; would again unite, and assume the form of a wave of flame; often it dropped bright scintillations, and occasionally threw as much light upon the road as a large torch.

The curious meteor described by Dr Shaw² as running along the ground, and playing on the ears and mane of his horse, has its counterpart, on a smaller scale, in other countries, and has often been witnessed in Britain. It seems in some instances to be an electrical meteor; but possibly it may also arise from the disengagement of phosphuretted hydrogen from living animals, or of a luminous and unctuous matter, like that of the glow-worm, or of phosphorescent marine animals.

A lambent flame has occasionally been observed to play around the heads of children, which has probably a similar origin; it has been finely described by Virgil, in his account of the glory that surrounded the head of the young Ascanius.

Ecce levis summo de vertice visus Iuli
Fundere lumen apex, tractuque innoxia molli
Lambere flamma comas, et circa tempora pasti.
Aen. H. 632.

SECT. VIII.—Climate.

The temperature of the atmosphere is derived from the heating of that portion of it in contact with the ground, which has received and absorbed the incident solar rays. Whilst the earth revolves on its axis, successive portions of its surface are illuminated and heated by the sun; but the excessive accumulation of heat in any particular place is prevented by the abstracting power of the air, and by the radiation from the earth's surface which takes place during the night. These causes alone, however, could not have secured the diffusion of warmth over a large portion of the earth. Had the axis of the earth been perpendicular to the plane

No. 28.

¹ See Priestley on Vision, and Muschenbroeck's Philosophy.

² Travels in Palestine and the East.

of its orbit, the same places would have had the sun always vertical; the equatorial regions would thus have been parched with intolerable heat; and much of what is now the fairest portions of the globe, the seats of literature and the arts, would have been doomed to sterility and desolation. The simple yet stupendous contrivance of the inclination of the axis to the plane of the orbit, by bringing successive portions of the northern and southern hemispheres alternately under the vivifying solar influence, has rendered the largest portion of the globe the abodes of animated beings; the fervour of a tropical climate is thus rendered less oppressive, and the limits of the temperate regions are greatly extended. The division into five Zones, and their denominations, give us a general idea of the temperature in each; but these are very inadequate to express the diversities of atmospheric temperature. Ancient geographers distributed the earth's surface into twenty-six parallel bands or Climates, which were supposed to indicate, with sufficient exactness, the differences of the heat of each region, until the invention of the thermometer gave more precise information.

We have already stated how multiplied and careful observations may point out the mean temperature of any place, and the contrivances which have been resorted to for abridging the labour of the meteorologist; and philosophers have endeavoured to supply by calculation what is wanting in observation. The most ingenious of these speculators was Mayer of Göttingen. From an examination of such registers of temperature as he could obtain, he concluded "that the mean temperatures decrease from the equator to the poles, as the squares of the sines of latitude;" and therefore, if the mean temperature at the level of the sea under the equator were ascertained, this formula would enable us to calculate the mean temperature of any given latitude. Mayer applied his formula to the equatorial temperature as determined by Bouguer, = to 24° of Reaumur; but this has been found by Humboldt to be too high. It should have been only 22° of that scale, which is equivalent to 27°·5 of the centigrade, and to 81°·5 of Fahrenheit. With these corrections, Mayer's formulae for the three thermometers, T signifying the temperature sought, will be

Unfortunately, however, later observations have shown that Mayer's formula is worse than useless in high latitudes, and indeed is only found tolerably consistent with observation in latitudes between 40° and 60° N. The thermometrical registers of Scoresby, Parry, Franklin, and Ross, have proved that the mean temperature of the parallel between 70° and 75° is very far below what Mayer's formula would assign, and is even far below what that empirical rule would give to the pole. In low latitudes, too, it is not to be depended on; and it is totally at variance with the recent but well-established fact of meridians of greatest cold. No attempt of this kind has yet been found reconcilable with observation. In the mean time, the celebrated Humboldt has performed an acceptable service to science, by a more critical examination of the accumulated observations of his predecessors and contemporaries. The results he has given in his essay on the Distribution of Heat over the Globe; and the representation of them upon a chart of Isothermal Lines is peculiarly valuable, as deductions from actual observation. An inspection of such a chart shows, that these lines are neither parallel to the equator nor to each other. They are higher in Europe than at 100° to the E. or W. Thus the isothermal line of temperature 54°·5, in France passes through lat. 45° 46'; in China (long. 116°

E.) it is in lat. 39° 54'; and in America, on the western coast (in long. 104° W.), it is in 44° 40'. Physical Geography.

The isothermal lines descend lower in America than in Europe. Thus the line of 32°, or the freezing point, in Europe, lies in Lapland between 66° and 68°, but in America it passes on the coast of Labrador in lat. 54°; and this difference, which is small near the tropic, increases much with the latitude. Thus,

In both continents, the most rapid decrease of temperature takes place between lat. 40° and 45°. This circumstance has justly been considered by Humboldt as producing a happy influence upon the civilization and industry of the nations inhabiting these parallels; because there slight changes in latitude produce considerable diversity in the vegetable productions, which may become objects of rural economy; and when contiguous countries differ much in their products, it stimulates the industry of each, and gives vigour to commercial intercourse; circumstances highly favourable to civilization. The range of the thermometer is not the same under the same isothermal lines, as representing the mean temperature of the year. It is greater in the New than in the Old Continent. Thus, Cincinnati in America, and St Malo in France, are both on the isothermal line of 54°; yet the difference between the mean heat of summer and winter is 40° at the former, and but 32° at the latter. Even though the difference between the mean temperature of the seasons increase with the latitude, under the same meridians, yet the difference at Philadelphia in America, in lat. 40°, is much greater than at Copenhagen in Europe, in lat. 56°.

The remarkable fact of the influence of longitude on temperature leads to the conclusion, that on each side of the equator there are two meridians, under which the mean temperature is lowest. These have been termed by Sir David Brewster the cold meridians, and their extremities are the poles of greatest cold. The position of these in the northern hemisphere may be approximated from recent investigations; and perhaps we shall not greatly err if we assign the longitude of 95° W. for the American, and of 100° E. for the Asiatic cold meridians. The apparent coincidence of the cold meridians with the general direction of Hansen's lines of no variation is perhaps more than accidental, when we reflect that there seem to have been, in former ages, migrations of the cold meridians eastward and westward, coincident, as far as we can judge from recorded changes of climate in northern countries, with similar migrations of the magnetic needle.

Elevation above the sea is one of the most important circumstances by which climate is modified. As we ascend elevation in the atmosphere the temperature diminishes, and under above the every latitude the summits of very lofty mountains penetrate into the abodes of perpetual congelation: such sum-

¹ Travail in Mémoires de la Société de Physique et d'Histoire Naturelle de Genève, tom. iv.

Physical Geography. mits are covered with perpetual snows. The altitude at which this takes place varies with the latitude, diminishing from the equator to the poles. In climates where there is a great difference between the heat of summer and of winter, this point will fluctuate between two limits, one of which is the upper, the other the lower limit of congelation. The distance between them is least at the equator, and increases as we recede from it. This circumstance is important; for it influences the formation of those stupendous glaciers, that excite the astonishment of the traveller, on the elevated regions of the temperate zone, and on the declivities of the arctic circle.

Sir John Leslie concluded, from some experimental investigations, that in moderate elevations we might estimate this decrease of temperature as we ascend, at 1° of Fahrenheit for every 300 feet of elevation; and he has given a formula for computing the point of congelation for every latitude; which, however, does not agree with observation in high parallels. The same remark applies to Kirwan's formula,¹ which presupposes a gradual depression of the point of congelation in receding from the equator. The requisite data for his calculations are the mean temperature at the level of the sea at the equator and at the given place, and the height of the equatorial point of congelation. Thus, if we take the point of perpetual congelation at Humboldt's estimate of 15,744 feet, and the mean temperature at the level of the sea there at 81°·5, the difference between that mean temperature and 32° will bear the same ratio to the equatorial point of congelation, as the excess of the mean temperature at the given place above the freezing point, to the point of congelation at the latter. Thus, if we wish to find the point of congelation in this latitude, 56°, we turn to a table of mean temperatures, and find it = 49° Fahrenheit.

On this formula, then, 5407 feet would be the point of perpetual congelation in our climate; and had we any mountain exceeding that elevation, its summit would be covered with perennial snow.

We apprehend, however, that no formula will apply to all latitudes; and we know that Kirwan's does not agree with observation in high latitudes.

Even when the latitude and elevation of two places do not differ greatly, they may suffer very different extremes of temperature. The chief agent in effecting this is the ocean. It has been found that the proximity of the sea, in all climates, affects the extremes of temperature. When cold air sweeps over the surface of the ocean, the caloric of the upper particles is diminished, their density is increased, and they sink, while their place is supplied by fresh particles from below, which becoming cooled in their turn, give rise to internal motions in the ocean. As the law of expansion during cooling from 39° to 32°, which forms so singular an exception in the case of fresh water, does not take place in sea-water, there is nothing to arrest these internal motions in the ocean:² and from its enormous mass, it is obvious that no partial application of cold, in this way, can sensibly reduce its mean temperature. Hence we find that the temperature of islands is not in winter so low as that of continents in the same parallels. Thus the severity of winter in Britain bears no proportion to that of continental regions under the same parallel. The Orkney Islands are in the same latitude as St Petersburg; yet, in the former, a frost of a week's duration is a comparatively rare occurrence, whilst at the latter, the wide, deep, and rapid Neva is for several months annually frozen over.

On the other hand, the ocean exerts a great influence in mitigating the excessive heat of tropical climates. It is true, that when the upper particles of its waters become heated, the internal motions cease; but the rarefaction produced sets a current in motion to the west (as becomes apparent in the Gulf Stream), the place of which is supplied by the influx of cooler waters from either pole. This is not all. The increased temperature is mightily diminished by increased evaporation; a process by which as much caloric is absorbed as would be capable of heating an equal weight of water 942° Fahrenheit. Hence tropical islands are less oppressively hot than continents in the same latitude. The summer temperature of the smaller Antilles, for instance, is lower than that of the same parallels in America, Africa, or Arabia. The intersection of continents by arms of the sea, or by vast collections of water, tends to mitigate the fervour of summer. The extended form of the old continent toward the east renders inequalities of temperature much greater in Eastern Asia than under the same parallels in Europe. Thus, at Pekin, in latitude 40° N., longitude 116° 20' E., the mean temperature of summer is 78°·8, and of winter 23°—a difference of not less than 55°·8, which gives rise to a frost of several months' duration in that part of China; yet Pekin is under the same parallel as the southern extremity of Naples, where frost is unknown, and of the central provinces of Spain, in which, though at an elevation of 2000 feet above the sea, ice is an extremely rare occurrence.

The face of a country, though less important than the circumstances already noticed, affects its climate. A region shrouded in primeval forests, or covered with swamps and marshes, will have a different temperature when cleared and drained. Marshes, by evaporation, rob the surface of much of its heat; and forests, by intercepting the sun's rays, the source of terrestrial warmth, and by the increased exhalant surfaces exposed to the air, undoubtedly have some effect on mean temperature. In hot climates forests tend to cool the air, and in frigid regions to prevent the loss of the earth's heat by chilling blasts. The clearing of forests may thus affect the temperature, and the draining of marshes increase the salubrity of particular regions; but the mere effect of cultivation can never be very considerable in changing a climate.

A combination of several of the causes already noticed as affecting climate have conspired to render the North American continent colder than the same parallels in Europe. Thus Nova Scotia and Canada, in latitude 47°, are in the same parallel as the central provinces of France and Upper Italy. In the former, the cold of winter has been observed as low as —40° Fahrenheit, or 52° below the freezing point of water; and the lakes and rivers of that part of America are bound up in impenetrable ice during four or five months of the year; while in the latter, in ordinary seasons, the severity of winter is scarcely felt. Labrador and England, each in latitude 53°, differ widely in mean temperature.

The climate of America, however, should not be compared with that of Europe, but with the eastern side of the old continent. The eastern sides of both, from causes which perhaps are not easily explained, are certainly colder than the

¹ Essay on Climate.

² Dr Hope, Transactions of the Royal Society of Edinburgh.

western; and when observations are multiplied on the western shores of America, it will probably be found that both continents approximate more nearly in mean temperature than the above comparisons would seem to indicate. In fact, the whole of Europe, as Humboldt has remarked, has an insular climate compared with the eastern parts of Asia and America; "and upon the same isothermal line the summers become warmer, and the winters colder, in proportion as we advance from the meridian of Mont Blanc to the east or to the west."¹ The eastern parts of China, and the Atlantic portions of the United States, have great differences between their summer and winter temperatures, or they are what Humboldt terms excessive climates; whilst all that we yet know of California, and of the country between it and the mouth of the Columbia, shows that the mean temperature of their winters and summers differs much less, and that they approach more nearly to the climate of Europe.

The distribution of the land in high latitudes is not unimportant as regards climate. Land is more easily heated than water by the sun's rays; and where the icy barriers are in contact with land, there is a greater probability of their being melted by the heat of summer, than where the ice extends over a large surface of uninterrupted ocean. This seems to be one cause of the greater extent of the southern than of the northern polar ices.

Had the globe been uniformly covered by water, the progression of climate would probably have been very regular. Indeed it might be shown, that the effect of the polar currents produced by equatorial rarefaction, supposing them to move with a velocity of two miles per hour, or to make three annual transfers from the pole to the equator, would be sufficient to produce the variations of temperature which are experienced at the level of the ocean.

On the whole, the peculiarities of climate are chiefly attributable to latitude, elevation, the distribution of land and water, proximity to the cold meridians, the humidity or dryness of the air; to which we may add, though less important, the prevalence of particular winds, the transparency or cloudiness of the sky, and the state of the general surface of any country.

SECT. IX.—Geographical Distribution of Plants.

Whoever considers the vast accessions made to our list of vegetable species within the last twenty years, and the immense tracts of the earth still unexplored by scientific men, will readily conceive that we are yet in no condition to pronounce even as to the probable amount of the existing plants. In Professor Lindley's Introduction to Botany, the following is an approximation to known species.

But if this afford a just estimate of the known species, it cannot give us any idea of the plants existing on the globe. Europe, with a surface $\backslash frac\{1\}\{4\}$th that of Asia and Africa, and in some respects less favourable to a varied vegetation, has a flora more numerous than either, and, till within a few years, superior to both united. This can only be attributed to the greater diligence with which the botanic riches of Europe

have been explored, and may lead us to suspect that we are yet far from a knowledge of all existing species. Physical Geography.

While botanists considered plants only as belonging to artificial systems, the progress of geographical phytology was inconsiderable. Tournefort, in ascending Mount Ararat, had observed at its foot the plants of Western Asia; but a little way up he recognised the vegetable forms of Italy; at a still higher level, those of Central France; next those of Sweden; and beyond them the flora of Lapland and the Alps.

Similar remarks were occasionally made by other botanists in other regions. They had determined with considerable accuracy the northern boundaries of some plants, especially those useful to man; but whilst the method of grouping plants was merely artificial, the physical laws which regulate their distribution were not perceived, and the circumstances which limit certain vegetable forms to particular situations were not comprehended. Humboldt's essay on the geography of plants appeared at Paris in 1807, and contained some generalizations on the succession of vegetable forms in Equinoctial America; but it is admitted on all hands, that the disquisitions of Dr Robert Brown, in the Supplement to Flinders' Voyage, published in 1814, led the way to still more extensive generalizations on the geographical distribution of plants, and first indicated the proportions that subsist in different places between the three great divisions of the vegetable kingdom. Since that period, the progress of vegetable geography has been rapid, chiefly through the labours of Von Humboldt, Brown, Wahlenberg, Decandolle, Von Buch, Parrot, Hornemann, and Schouw; and this improvement is ascribed, by the illustrious philosopher first named, to the adoption of the natural method of Jussieu. To none of these writers are we more indebted than to Von Humboldt. The ideas developed in his Essai sur la Géographie des Plantes, and in his Tableaux de la Nature, were expanded in the Prolegomena de Distributione Geographica Plantarum. His invaluable Mémoire on isothermal lines (Mém. d'Arcueil, and Edin. Phil. Journ. iii. iv. and v.), and, last of all, his New Enquiries into the Laws which regulate the distribution of Vegetable Forms (same Journal, vi.), contain an admirable summary of almost all we know on vegetable geography.

The primary divisions of the natural arrangement are into acotyledones, which include the ferns, fungi, lichens, mosses, sea-weeds, &c.; monocotyledones, including palms, rushes, sedges, grasses, liliaceous plants, &c.; dicotyledones, which comprehend all plants not contained in the other groups. The circumstances on which the geographical distribution of these different vegetable forms depend, may all be comprehended in climate and soil. The causes chiefly modifying climate are, latitude, elevation above the sea, longitude, moisture, and insolation; particularly the first two. We shall endeavour to point out very generally how vegetation is modified by these circumstances; and, first, of the effect of latitude.

In proceeding from the equator towards either pole, the temperature diminishes; but we have seen that the latitude on isothermal lines are not parallel to the equator, and the plants, vegetable forms which characterize them descend lower in the new than in the old continent. In equatorial regions the vegetation consists of dense evergreen forests, characterized by palms and arborescent ferns, mixed with Epidendra and rigid grasses; and the lower tribes of vegetable life, such as Fungi, Confervæ, and Musci, are very rare; but Musaceæ, Melastomaceæ, Scitamineæ, Myrtaceæ, Piperaceæ, and the larger Compositeæ, are very frequent. On receding from the equator these tropical forms give place to Rosaceæ, Conifereæ, and Amentaceæ; verdant meadows of soft

¹ Edinburgh Philosophical Journal, vol. iii.

Physical Geography. grasses intersperse the landscape, the epiphytal Orchidæ disappear, and are succeeded by plants whose fleshy roots draw their sustenance from the soil; the trunks of aged trees are clothed with mosses, parasitic Fungi invest decaying vegetables, and the waters abound with Conferve. On entering the frigid regions trees disappear; dicotyledonous plants become more and more rare; and Gramineæ, Mosses, and Lichens are the last retreats of vegetable life. The northern hemisphere, with which we are best acquainted, is by Humboldt divided into six isothermal bands.

1st band has a mean temperature above 77° F., and may be considered the natural region of Palms, of the Banana, and the Coffee-tree. It extends northwards in the old continent to lat. 32°, in the new to 23° 30'.

2d band, with a mean temperature ranging between 77° and 68°, is the proper region of the Citron and its varieties. In the old continent it reaches to 37° or 38°, in the new to 29°.

3d band, with a mean temperature from 68° to 59°, is the true region of the Olive and the Vine. In the old continent it extends to 43° 30'; in the new, and in Eastern Asia, to 32° or 33°.

4th band, with a mean temperature from 59° to 50°, produces the Vine, and in perfection Wheat, and the Oak. In Europe it extends to 52° 25', in America to 42° 25', in China to 40°.

5th band, with a mean temperature from 50° to 41°, is the region of various Cerealia, and of forests of Quercus Robur. In Europe it extends to lat. 60°, in America probably to 50°.

6th band, with a mean temperature from 41° to 32°, is the native region of the Pine, the Birch, and the Willow, in its lowest parallels; and in its higher, of alpine plants, Lichens, and Byssi. It reaches to the limits of perpetual congelation. In Europe it extends to lat. 71°, in Asia to the arctic circle, but in Eastern America only to 57° 8'.

But the effect of elevation above the sea is no less striking, and is capable of counterbalancing or of subverting the influence of a low latitude. Wherever lofty mountains occur in the torrid zone, their summits are veiled in perpetual snows; and the ascent presents the flora of a vast variety of climates. This is beautifully illustrated in Humboldt's Géographie des Plantes Équinoxiales, the chart of which forms Plate CCCCXII. in this volume. He divides Equinoxial America into eight regions.

1. The region of Palms, Scitamineæ, and Liliaceæ. This is the habitat of true Palms, of the Musa, Theophrasta, Mussenda, Pumeria, Casalpinia, Hymenæa, Cecropia, and Cusparia. It extends from the level of the sea, under the equator, to the elevation of 5700 feet. One species of tree found in this region extends to the height of 6800 feet.

2. The region of arborescent Ferns, and the true Cinchona. Gigantic Ferns first appear at the height of 1200 feet, and disappear at an elevation of 4900 feet. The medicinal Cinchona attain the height of 8800 feet.

3. The region of Oaks commences at 5200 feet, and reaches to

4. The region of Shrubs. Trees disappear entirely at a height of 10,800 feet, and are succeeded by shrubby plants.

5. The region of plants corresponding to the alpine plants of Europe. These are first found at 6100 feet above the sea, and continue to the great height of 12,600 feet. Amongst these the most conspicuous are the genera Gentiana, Lobelia, Ranunculus, &c.

6. The region of Grasses. Here the Gramineæ are the sole vegetable produce; and the principal genera are Avena, Agrostis, Dactylis, Panicum, Stipa, Javara. In this region snow regularly falls, and the plants enumerated continue to the vast elevation of 14,200 feet.

7. The region of the Hyperoxylon, Byssi, Hepaticæ, and Lichenes. These lowly and hardy forms of vegetable life

continue to the very verge of uninterrupted winter, which in the

8. Region covers the Equinoxial Andes, from the height of 15,700 feet, to their towering summits.

In the temperate climate of Europe, the changes produced by elevation, though less striking, are not less important on vegetation. M. Ramond, in examining the Pyrenees, found vegetable forms distributed into very regular zones.

1. The region of Oaks lies at the foot of the chain, and is chiefly characterized by Quercus Robur, which continues to the elevation of 4800 feet above the sea.

2. The zone of Beeches, characterized by the Fagus Sylvatica, commences at 1800 feet above the sea, and extends to 1400 feet beyond the oaks, rising to 5200 feet, or to a mile, in height.

3. The region of the Silver Fir, or Pinus Picea, and Yew. It commences at 4200 feet, and extends to the height of 6000 feet, or 600 beyond the region of Beeches.

4. The region of the Pinus Sylvestris and Pinus Pumilio, commencing at 6000 feet, and extending to an elevation of 6200 feet; and this is the extreme limit of trees.

5. Region of dry-leaved shrubs and creeping plants, such as Rhododendron, Daphne, Passerina, Globularia Repens, and two diminutive species of willow, Salix Herbacea and S. Reticulata. The limit of this last is uncertain, but it rises much beyond this region.

6. The region of alpine plants with perennial roots and naked stems; as Gentiana, Primula, Saxifraga, Ranunculus. The extreme limit of phenogamous plants is marked by Ranunculus Glacialis, Saxifraga Cespitosa, S. Oppositifolia, S. Androsacea, and S. Groenlandica. These elegant but hardy individuals, intermixed with Byssi and Lichens, reach to

7. The region of perennial snow.

As a further illustration, we may give the progress of vegetation within the arctic circle from the masterly sketch of Wahlenberg. In ascending the Lapland Alps, from the head of the Gulf of Bothnia, he found eight well-marked zones.

1. The region of the Spruce Fir had almost ceased before reaching the mountains. It had become a slender pole, with short, drooping branches. The Rubus Arcticus had ceased to ripen its fruit, and the pools were no longer ornamented with Arundo Phragmitis, Galium Boreale, and Carex Globularia. The upper limit of the Spruce Fir is 3200 feet below the line of perpetual snow. This is the true region of Tussilago Nivea; it has a mean temperature of 37° 5 of Fahrenheit.

2. Pinus Sylvestris, or Scotch Fir, is still found; but its stem is low, and its branches widely spread. In this region is the last appearance of Ledum Palustre, Salix Pentandra, and Veronica Serpyllifolia. Near the upper limit of the Scotch Fir the Phaca Alpina is found, and the berries of Vaccinium Myrtillus do not ripen well. The upper limit of this zone is 2800 feet below the line of snow; its mean temperature 36° 5 Fahrenheit. A little below this boundary barley ceases to ripen; but grazing farms, with turnips and potatoes, are cultivated, even 400 feet higher.

3. The zone of the Birch with short, thick stems and knotty branches; its lively green leaves are refreshing to the eye, but the tree is diminished, so as to be commanded from every eminence. The upper boundary of the Birch is 200 feet below the line of snow. The Alnus Incana, Prunus Padus, and Populus Tremula, had long before this disappeared; and Sorbus Aucuparia, Rubus Arcticus, and Erica Vulgaris, disappear near the upper limit of this region; but the drier spots produce Lichen Rangiferinus, Tussilago Frigida, and Pedicularis Sceptum Carolinum, to the upper boundary of the Birch.

4. The lower skirts of this region are covered with the

dark foliage of Betula Nana; Salix Glauca fringes the streams; every hill is covered with Arbutus Alpina, and variegated with Andromeda Cærulea and Trientalis Europea; while the bogs are ornamented with Andromeda Polifolia and Pedicularis Lapponica. This zone reaches to within 1400 feet of perpetual snow.

5. Brush-wood has disappeared; Salix Lanata is two feet high; Betula Nana creeps along the ground; the hills abound with Azalea Procumbens and A. Lapponica. No berries ripen here, except those of Empetrum Nigrum. This zone rises to within 800 feet of the line of snow; its mean temperature is 34° Fahrenheit.

6. Region where patches of snow remain all the year. Bare places still produce a few shrubby plants, as Empetrum Nigrum, Andromeda Tetragona, and A. Hypnoides, Gentiana Nivalis, G. Tenella, Draba Alpina, and Dryas Octopetala. This zone reaches to within 200 feet of perennial snow, which forms

7. Zone in which, wherever a patch of soil is visible, various Saxifragæ and Ranunculi appear. The upper limit of this region is 4200 feet above the sea; but on the higher mountains we have an

8. Region of pure snow. Yet even here, if a southern aspect chance to thaw for a few days any cleft in a rock, a few hardy specimens of Ranunculus Glacialis venture to bloom. The plant has been gathered even 500 feet above the last zone; and Umbilicated Lichens have been seen 2000 feet above the line of perennial snow. The mean temperature of the eighth region is 30° Fahrenheit.

In these instances, then, the effect of elevation is equivalent to latitude; but it must be recollected, that plants will not thrive equally in places with the same mean temperature. Some require a strong ephemeral heat. Hence, in judging of the aptitude of any place for rearing particular plants, we must compare the mean temperature of the summer, as well as of the whole year, before we decide. Thus we are enabled to explain why the Pistacio-nut ripens in Pekin, but will not ripen in France, where the isothermal line for the whole year is the same; but though the Chinese winter be more severe than that of France, the summer heat is far greater. Innumerable other instances might be adduced of the same fact. The moisture of a climate has much influence upon its vegetation. Water is the vehicle of the food of plants, and perhaps yields a great proportion of it. If moisture be deficient, plants die; but they require water in very different proportions. Those with soft, broad, smooth leaves, that grow rapidly, and have many cortical pores, require much water to maintain their vitality; on the other hand, plants with few cortical pores, with oily or resinous juices, and small roots, will generally thrive best in dry situations.

Insolation, or exposure to light, is necessary for most plants. The green colour of plants is only formed in light, as is shown by etiolation; and light appears to be the cause of certain movements which are remarked in the flowers of most plants, and in other parts of some delicately organized individuals, which open and close their leaves according to the degree of light. This last property is chiefly seen in tropical plants. Light appears to be necessary to the decomposition of carbonic acid, and the fixation of carbon in their tissues; and it is indispensable to the right performance of the function of reproduction.

The influence of soil on vegetables is seen in the preference which many plants have for a calcareous soil; some affect siliceous sands, others clay retentive of water; some plants thrive best in the clefts of slaty rocks; some delight

to dwell amid granitic rocks, and others on a saliferous soil. Earthy matters enter largely into the composition of some vegetables; and, in the epidermis of the Gramineæ, silica is invariably found. The presence of animal matter in soils is necessary to many plants, and is generally nutritive to all. Iron and copper are found in small quantities in some plants. The stations of particular plants have often been determined by these peculiarities of soil; and where a soil and climate are equally suitable for many social plants, we find them growing together, until the strongest obtains the mastery, and chokes the others. Thus Erica Vulgaris appears to have usurped in Europe a space once occupied by other genera, if we may judge by what generally happens on exterminating heath; for then other plants very speedily make their appearance, the seeds of which seem to have long preserved their vitality in the earth, and only to have wanted room to spring into visible existence.

A combination of these causes no doubt influences the distribution of particular species. Thus, in the old world, Ericæ and the Proteaceæ of Southern Africa are peculiar, and are replaced in Australia by Epacridæ and new genera of Proteaceæ. The Banksia, the Yucca, the Goodenia, and the leafless Acacia, are peculiar to the latter country. The Cinnamon, Nutmeg, and Clove are confined to the Indian islands; the Thea and Camellia are indigenous to China. America does not contain a single species of Ericæ, from one extremity to the other; nor has a Pæonia ever been found in it, except a solitary one observed by Douglas to the west of the Rocky Mountains. That mountain barrier divides two vegetations, almost as peculiar as those of two continents. On its eastern side the forests of North America are distinguished by the variety of their Oaks and Juglandes, the magnificent inflorescence of the Rhododendron, the Magnolia, the Azalea, and the humbler beauties of the Actæa and Vaccinium; all of which are utterly unknown on the western side of that ridge. America is the real habitat of the Cinchonaceæ and the Cacti, of the Fuschia, the Calathea, the Mustisia, and all the Bromeliaceæ.

One grand question in vegetable geography, first considered by Brown, has yet to be noticed, viz. the numerical relations that subsist between different vegetable forms. We may ask, which natural families of plants abound most over the world? and give in succession the comparative numbers that are known to exist. This interesting comparison has been attempted in Decandolle's grand work. But this information is less important and satisfactory than the investigation of the relation which the numbers of each family bear to soil and climate. The question thus becomes far more complex; and we are under the necessity of comparing the ascertained number in each division with the whole mass of phænogamous plants.

It is to this last mode of studying vegetable forms that the admirable inquiries of Humboldt are directed. We strongly recommend the whole of that paper to the perusal of our readers. Our limits will only permit us to give his table of the numerical results of his investigations of the distribution of certain extensive families which exert a marked character on the vegetable physiognomy of the countries where they occur.¹

In the following table the sign $\backslash nearrow$ indicates that the denominator of the fraction diminishes from the equator towards the pole; $\backslash nwarrow$ that it diminishes towards the equator; $\backslash leftarrow$ that it diminishes from the north pole and the equator to the temperate zone; $\backslash rightarrow$ that it diminishes towards the equator and the north pole.

¹ The Agaveæ of Humboldt correspond to the Acotyledones of other botanists; his Glumaceæ include the families of Monocotyledones indicated in the table.

"On comparing the two continents, we find in general in the new world, under the equatorial zone, fewer Cyperaceæ and Rubiaceæ, but more Compositæ; under the temperate zone, fewer Labiæ and Crucifereæ, and more Compositæ, Ericæ, and Amentaceæ, than in the corresponding zones in the old world. The families that increase from the equator to the pole, according to the method of fractional indications, are Glumaceæ, Ericæ, and Amentaceæ. The families which decrease from the pole to the equator are Leguminosæ, Rubiaceæ, Euphorbiaceæ, and Malvaceæ: the families that appear to attain their maximum in the temperate zone are Compositæ, Labiæ, Umbellifereæ, and Crucifereæ."

Botanists have attempted to divide the globe into botanical regions. The first successful attempt of this kind was made by Decandolle, who divides it into twenty regions, the designations of which are drawn from particular portions of the earth, marked by the peculiarity of their vegetable productions. But the most able and luminous exposition of this sort is the Phyto-Géographie of Professor Schouw. The designations of each kingdom or region are usually derived from some of its most characteristic vegetable productions; but where we are not sufficiently acquainted with the botany of certain districts, he uses a geographical designation. Schouw divides the earth into twenty-two botanical regions, of which our limits will only allow a very imperfect sketch.

1. Region of Saxifraga and Musci, or of the alpine arctic flora.—This includes all countries within the arctic circle.

2. Region of Umbellifereæ and Crucifereæ.—This includes the whole of Europe north of the Pyrenees not included in the last region, most of Siberia, and the Caucasus.

3. Region of Labiæ and Caryophyllæ, or Mediterranean flora.—It includes the south of Europe, north of Africa, the Canaries and Azores, as well as Western Asia.

4. Japanese Region includes Eastern Asia and Northern China.

5. Region of Asteræ and Solidagines.—The north-eastern part of the United States. It is marked by the varieties of its Oaks and Pines.

6. Region of Magnoliæ.—This comprehends the southern portion of the United States.

7. Region of Cacti, Melastomaceæ, and Piperaceæ.—This includes Lower Mexico, the West Indies, and the northern coast of South America.

8. Region of Cinchonaceæ.—This includes Bolivia and Columbia to a certain elevation, or Peru and New Granada.

9. Region of Escalloniæ, Calceolarieæ, Winterieæ, and Vaccinieæ.—It includes the highest parts of South America.

10. Chilian Region.—It is formed into a district region, because not above one half of its plants occur in the lower regions of South America.

11. Region of Arborescent Compositæ.—This includes Buenos Ayres, and the eastern division of temperate South America.

12. Antarctic Region.—This includes the southern part of the American Continent, the Magellanic region of Decandolle.

13. New Zealand Region.—It has many plants in common with New Holland, as Epacris, Melaleuca, and Myoporum; some common to South America, as Wintera, Ancestrum, and Weinmannia; some common to South Africa, as Gnaphalium, Xeranthema, and Mesembryanthemæ; some which seem peculiar, as Phormium, Areca Sapida, Dracena Indivisa.

14. Region of Epacrides and Eucalypti.—This embraces the southern parts of Australia and Van Diemen's Land. The families of Tremandreæ, Stackhouseæ, Proteaceæ, Eucalypti, Cassuarineæ, Diosmæ, Styliæ, &c. characterize it.

15. Region of Mesembryanthemum and Stopleæ.—This comprehends the south of Africa. Besides these plants, it is distinguished for the number and beauty of its Ericæ. It possesses several plants mentioned under 13, which are only found in it and Australia. It abounds in compositæ.

16. Region of Western Africa.—Its characteristic is Adansonia, the largest of known trees. Many of its Cyperaceæ are peculiar. It wants, or has very sparingly, the American Cacti, Palms, Peppers, and Passion-flowers.

17. Region of Eastern Africa.—Two thirds of its known

plants belong also to India. This region includes Madagapary, gascar.

18. Region of Scitamineæ.—This includes the flora of the Indian peninsula. It contains Carcunæ, Zedoariæ, Cardamona, &c.; and its Scitamineæ are more numerous than the same family in America.

19. Himalayan Region.—Tropical forms here disappear. It bears a strong resemblance in its vegetation to some parts of Europe. Orchideæ and Filices are numerous. It presents the Cerealia and Fruits of Europe.

20. Southern Chinese and Cochin-Chinese Region.—The flora of this region is little known, but appears peculiar.

21. Region of the Cassia and Mimosa.—This includes Persia and Arabia. The peculiar plants are Senna, Balsamodendron, Cadia, Caucanthus, Strænia, Coffea Arabica. Ferula Assafoetida belongs to it.

22. Polynesian Region.—Most of the genera are common to this region and India, or to America.

SECT. X.—Geographical Distribution of Animals.

The slightest acquaintance with zoology is sufficient to show that animals do not indiscriminately spread themselves over every part of the habitable earth. One species is found to be peculiar to a certain region, and sometimes is represented in another region by an analogous yet different animal. Difference of climate, and the greater or less facility of procuring subsistence, are often the probable causes of such limitations, when animals have the power of emigration; but in numerous instances we can trace the operation to no secondary cause. Why, for instance, should animals, incapable of crossing arms of the sea, be found in certain insular situations, and denied to other lands, seemingly as well fitted for their propagation and their maintenance?—why should the forms of organization which we find to prevail in one portion of the globe be represented in another by other forms, and by a very different though not less admirable structure? In this species of inquiry, one thing is demonstrable, that if, in every class of animals, we compare their organization with the intended residence and mode of life of each creature, with its necessities and the nature of its aliment, the most admirable adaptation of means to the end is everywhere perceptible; whether we consider the instinct which impels the ant and the hamster to lay up stores of food for the season of scarcity, or the bee at once to solve a delicate problem in practical geometry; whether we regard the structure of the proboscis of the gnat or of the elephant as suited to their wants; or whether we contemplate the form and position of the levers that move the limbs of animals, or the exquisite mechanism of the visual organ; all proclaim the handiwork of a wise and beneficent Creator.

To us it appears evident that nature has only distinguished animated beings into species, the limits of which are fixed by the capability of procreating fertile progeny; and that all our divisions into genera, orders, and classes are more or less artificial, though of high value in assisting the memory to recollect individuals, or to arrange newly-discovered species, and often as leading to interesting inferences respecting their habits and geographical distribution. When several species resemble each other in structure and habits, we agree to consider them as of one genus; when their resemblances are somewhat less numerous, we consider them as belonging to the same order; and when their similarity is confined to a few points, they are grouped in the same class. All these divisions are but artificial helps to our memory, whether we follow the sys-

tem of the illustrious Swede; or, with Cuvier, found distinctions more on the anatomical structure of the animals; or, with our sagacious countryman Macleay, attempt to reduce all recognized animals to quinary circular groups; or, with Swainson, resolve these circles into ternary associations; or, with Kirby, expand these groups into the magic number seven. Some arrangement is absolutely necessary in pursuing the subject of animal geography, when we consider the multitude of species; and perhaps no system is so convenient as that of Cuvier, from its very general reception, and its sufficiently according with the geographical distribution of species.

It is not easy to give any accurate idea of the real number even of known species of animals, from the difficulties thrown in the way of animal statistics, by the multiplication of species, and from confounding the effects of sex and age on some animals, and overlooking specific differences in others. Naturalists have differed widely on this subject. The estimate of Keferstein, as exhibiting the number of extinct as well as of living species, is interesting.¹

But this table contains no account whatever of living insects and arachnides. Mr Swainson² gives the following as the numbers known at a later period.

In the same work Swainson has given a conjectural estimate of the number of probably existing species.

Besides these, there exist innumerable hosts of Infusoria, in which the researches of Ehrenberg have detected remarkable peculiarities of structure; proving that, though minute, they possess a very complex organization, not less wonderful than the anatomy of the higher animals.

To these statements we may add, that M. Temminck has lately given the number of well-ascertained mammifera as 930, besides 140 species which he considers as doubtful; that the birds of our recent catalogues exceed 6000; and that Cuvier had enumerated 6000 species of fishes. With regard to insects, some naturalists have stated, that at least

¹ Die Naturgeschichte des Erzkörpers, Frnner 1834.

² ZOOLOGY, vol. ii. of the Cabinet Cyclopædia.

two species of insects feed on every species of plant; and therefore, that as the ascertained vegetable species amount to 86,000, there must be 172,000 species of insects.

The evident tendency of animals to be congregated in groups peculiar to certain divisions of the earth's surface, had long been familiar to zoologists; but it does not appear that any attempt was made to define the limits of those zoological divisions until the appearance of the essay of Mr Swainson, in the Cabinet Cyclopaedia. The ingenious author has there shown the claims of Europe to be considered as one of the grand zoological divisions of the globe, from the number of the types of natural families which it possesses. Swainson contends that the birds of any district afford a fairer criterion of the limits of a geographical distribution than any other class of animals. Quadrupeds he believes to be too much under the dominion of man, and liable to have their geographic limits disturbed by human interference; and the other classes of animals are either too numerous or too few to afford the means of determining the limits of such divisions; whilst birds, though seemingly fitted by nature to become wanderers, are surprisingly steady in their localities, and even in the limits of their annual migrations. These migrations are evidently caused by scarcity of food. Thus our swallows leave us when their insect food begins to fail, and they naturally pursue that route which is shortest, and affords subsistence by the way. The distance from the shores of the Baltic to Northern Africa is not half so great as that between England and America; and during the migration over land, the winged travellers find food and resting-places as they proceed to more genial climates.

The distribution of mammifera may probably be less characteristic than that of birds in Europe, where the long progress of civilization, and the spirit of commercial enterprise, have blended the races of the larger animals; so that it would be difficult, in most cases, to trace the native country of their original stock: but in most other parts of the world peculiar quadrupeds might be found sufficiently characteristic of each division.

The natural associations of animals have suggested another division of the earth into zoological kingdoms; and, did our limits permit, it would not be difficult to subdivide these into zoological provinces, each distinguished by some less considerable peculiarities of its animal productions. It is at once apparent, that these divisions will not correspond to the zones of the geographer, which do not coincide with marked differences in climate, nor with the supply of food, nor with the configuration of the dry land; circumstances most materially influencing the distribution of animals. The earth may be divided into fourteen zoological kingdoms; the first eight belong to the old, the six last to the new world.

OLD WORLD.

1. Palæonarectic Kingdom.—This division includes all Europe and Asia within the 60th degree of N. lat. Many of its mammifera and birds are common to it and the northern regions of America, and its peculiar animals of this class are but few. The Polar Bear it has in common with North America, but the Rein-Deer of Northern Europe now appears to be at least a variety differing from the Rein-Deer of North America; and certainly the Elk of the former should not be considered as the same species with the Moose-Deer of the latter. The Glutton, Gulo Borealis, seems to be confined to this region; as are the common Lemming and the Hamster. The great Sea-Otter, Lutra Marina of Steller, is found on both sides of the Northern Pacific, as low as 50° or perhaps 45° N. lat. The Cetacea, as might be expected, are common to the arctic seas of both worlds, and some of these occasionally wander southwards; but the true Greenland Whale, and the Narwal, are confined to the arctic seas. The birds that con-

stantly reside in the arctic regions are very few; but this region is penetrated by numerous birds well known in other parts of Europe and Asia, which resort to it as their breeding-places during its transient summer. These birds chiefly belong to the order of Natatores. A single Woodpecker, Tridactylia Hirsuta, is peculiar to the palæonarectic kingdoms; as are the Great Wood-Grouse, and several other species of the genus Tetrao. The Snowy Owl and Jer-Falcon belong to this region, though occasionally they are found in the north of Scotland, and occur also in North America. Yet these regions, seemingly so inhospitable, are not without numerous tenants. No part of creation is an absolute waste, destitute of animation. It is well known that the enormous number of ducks and other water-birds which frequent the streams and lakes of arctic countries, are attracted by the infinite swarms of insect food which abound in them during their short summer. Scoresby found numerous butterflies and other insects on the bleak shores of East Greenland; and he states that the Greenland seas teem with myriads of minute animalcules, which are the food of the Balæna Mysticetus, and which, in sea-water examined by him, existed to the number of 110,592 in each cubic foot, or 23,887,872 in every cubic fathom of that water.

2. Occidento-Caucasian Kingdom.—This division is equivalent to the European province of Swainson; but that designation appears to be not very appropriate to a region embracing also a part of Western Asia and of Northern Africa. The zoology of this division has been more completely explored than that of any other portion of the globe, and it presents to the naturalist types or characteristic species of a great number of genera in all classes of zoology. It contains many imported species, some of which are domesticated, or rendered subservient to the wants of man; so that it is difficult to trace its larger mammifera to their parent stock. Its most characteristic quadruped is the Aurochs, or Bos Urus. It is the animal next in size to the Rhinoceros, but is not the original stock, as has been alleged, of our domestic cattle. It formerly was abundant in Europe, but has fled from human persecution into the pathless forests of Lithuania, and the wilds of the Carpathians and the Caucasus. The Brown Bear, Ursus Arctos, appears strictly to be a denizen of this region, though it is also found in the palæonarectic kingdom. The Ibex, the Stag, and the Roe, are also among its characteristic animals; and though the Fallow Deer is said also to be found in Eastern Asia, it appears to be chiefly an inhabitant of the Occidento-Caucasian kingdom. The Horse and Ox, though both imported, nowhere attain to greater perfection in their most valuable qualities, though the former requires the occasional transfusion of Arabian blood to preserve its energies unimpaired. Its more considerable characteristic birds are, the noble Gypæte or Bearded Vulture of the Alps; the Imperial and Royal Eagles of Temminck, F. imperialis and chrysaetos; and several species of Falcons. The Red Grouse is found in no part of the world except the British islands. The Great and the Little Bustards are peculiar also to this division, which is particularly rich in many other genera of birds. It contains about 470 species of aves, of which 310 are inhabitants of Britain; in this list are seventeen Raptors, 117 Insessores, seventeen Rasores, sixty-six Grallatores, ninety-three Natatores, besides thirteen occasional visitors to our shores.

3. Orientalo-Caucasian Kingdom.—In this division there is included all Asia between the Altaian and Himalayan chains. It should most probably be subdivided into two provinces.

a. The Mongolian province, of which the characteristic Mammalia are the Yak, Bos Gruniens; the Argali, Ovis Ammon, the Onager or Wild Ass; the Wild Horse, or Equus Hemionus of Pallas. The Felis Onea also belongs to this region; as does the Ursus Thibetianus, the Thibetian Musk,

which is pretty widely diffused in this division; and the lively Jerboa. Its birds are little known to the naturalist, except the few which have been brought from Dauria.

b. Septentrio-Sinican Province.—The little we know of the zoology of China enables us only to state, that this division abounds with beautiful Phasianidae, particularly with the Golden and Silver Pheasants, which were originally brought from China. The Argus Polylectron is said to be found in the colder regions of Thibet and China; and the Impeyan Lophophorus, which we have received from the Himalayan Mountains, may probably also inhabit the mountainous districts of China.

4. The Austro-Asiatic Kingdom may be considered as bounded by the 30th degree of N. lat. and the equator. It includes Southern China, Cambogia, Siam, the Birman empire, Hindustan, and Ceylon. It may be characterized as the native region of the Tiger, the Panther, of the Asiatic Elephant, and of the Long-armed Apes; of the Peacock, the Giant Argus, and the Hornbills; of the Gangetic Crocodile, the Python, and Cobra de Capello. Its other remarkable animals include the One-horned Rhinoceros, the Hunting Leopard, the Malayan Bear, the Ursine Sloth; that curious animal described by Sir William Jones, the Slow Lemur, or Stenops of Illiger; the Short-tailed Pangolin or Scaly Manis, the Brahmins' Bull, the Urneh or Wild Buffaloe, and the less-known wild ox of India, named the Gour, as well as the Antelope Picta or Nilghau, A. Tragocamelus, and several other antelopes; the beautiful Spotted Axis, and several other deer; the Pigmy and Indian Musks. Among its characteristic birds, besides those already mentioned, we should not omit the Jungle Fowl, and the Bankiva Cock, which many naturalists suppose to be the stock of our domestic poultry. The genus Buceros is the representative of the Toucans of America, and has the nerves of its nose distributed in its upper mandible and bony casque; the Ciconia Argala, or gigantic Adjutant-bird, is the most remarkable of its waders.

5. Polynesian Kingdom.—In this division we include the Philippines, Borneo, the Moluccas, Celebes, Java, and Sumatra. It may be considered as the native country of the great Orang-Outan, and numerous Monkeys, among which the Proboscis Monkey and Cochin-China Monkey are the most remarkable; of enormous Bats, of the Galeopithecus or Flying Lemur, of the Indian Tapir, and of a species of Musk. It is also characterized by its very magnificent birds: among these we may notice the genera Iora, Calypsemena, Eurythraeus, Irena, Gracula, Lamprocoris, and its very splendid Lories. It is distinguished by the beauty and splendour of the genus Cinnyris, formerly included among the Creepers; and the elegance of its Columbidae. It also contains the Argus Pheasant or Polylectron, the Gigantic Argus, and the Lophyrus Cuvierii. Its seas abound in most beautiful and valuable shells, though it is comparatively poor in fluviatile Testacea.

6. Chaldeo-Arabian Kingdom.—The natural boundaries of this province are well marked by seas, the river Tigris, and by the cultivated part of Syria. It may be distinguished as the region in which the most valuable qualities of the Horse are developed, and where the Camel, and its variety the Dromedary, with the Gazelle or Antilope Dorcas, are the most conspicuous mammifera. Its ornithology is but little investigated; but the deserts are traversed by the Ostrich; its mountainous regions breed Pheasants and Doves; and one bird that frequents Arabia, and is greatly respected by the people, is a thrush, Turdus Seleucus of Forskal, which follows and destroys the innumerable swarms of locusts. The rarest and most beautiful shells are found in the adjacent seas; and the pearl-fisheries of the Persian Gulf have been long celebrated.

7. Australian Kingdom.—If we are permitted to include in this division New Holland, Papua, Van Diemen's

Land, and New Zealand, it has one of the richest and most peculiar fauna of any on the globe. It may be characterized as the native region of the Kangaroo, the Ornithorhynchus, and Ecidna, of the entire superb family of Birds of Paradise, of the Honey-suckers, of innumerable Parrots, of the Emeu, and Menura Superba. The absence of all the larger and pachydermatous quadrupeds was common to Australia, with the islands in the great Southern Ocean, at the time of their discovery, and is one reason for grouping those islands in this zoological kingdom. We have, besides the quadrupeds above mentioned, whole genera peculiar to this kingdom, particularly the Perameles, Phalanger, Petaurista, Phascolomys, and Hypsiprymnus. The following genera of birds are peculiar to that region: Seriulus, Podargus, Malurus, Ptilonorhynchus, Glaucopis, Pardalotus, Mellisuga, Menura, Dacelo, Megapodius, Seythrops, Dromicius, Ceriopsis, &c.

8. The African Kingdom includes that continent south of Atlas, with the islands of Madagascar and the Seychelles. It is the region of the Camelopard, the Hippopotamus, and the Two-horned Rhinoceros, of the African Lion, the Leopard, and African Elephant, of the Zebra and Quagga, of the Pongo or Black Orang, and of the larger Baboons. It abounds in a great variety of Antelopes. These animals, with few exceptions, are found over a large extent of Africa; but the Hyrax of the Cape, and the Tenrec of Madagascar, are more limited in their range. Of the variety and beauty of its birds, some idea may be formed by consulting the splendid works of Le Vaillant. The Occipital Vulture of Burchill, and numerous Eagles, are found in its southern regions. The genera Pogonius, Musophaga, Numida, Struthio, Anastomus, Touraco, Indicator, Pomerops, Centropus, contain its most remarkable birds; but the brilliancy of the Cinnyridae and Lamprocorinae is well known. Most of the African birds brought to Europe have rich plumage. The extinct species of bird, the Dodo, belonged to this division of the earth, as our early voyagers found it in the Mauritius and Isle of France. The Crocodile and Cayman appear to be diffused throughout this continent. The number of poisonous Serpents is considerable, and highly dangerous in its hotter regions; yet in Africa, as in other parts of the earth, by far the greater number of snakes are harmless. Africa is peculiarly rich in land Testacea, among which is the Achatina Zebra, the largest of land shells. Amongst its insects, we may notice the number of its locusts, and the prevalence of the Termes Bellicosus, an insect allied to the ant, which constructs a mansion of clay, equalling in size, and surpassing in solidity, the simple habitations of most of the natives of Africa to the south of the desert of Sahara.

NEW WORLD.

The zoology of North America has been so admirably in the new world treated by Dr Richardson in his memoirs in the Sixth Report of the British Association, that it ought to be studied by all interested in such researches, who will not fail to consult the splendid Fauna of North America, lately published by him and Mr Swainson. In most parts of America we meet with many new genera in every class of animals, and find few species that are not peculiar to the new world, with the exception of marine animals. The Quadrumana, which range in America between 29° on each side of the equator, have the peculiarity of either wanting the thumb on the fore extremity, or having it so placed, or imperfectly developed, as not to be a real opponent to the fingers; while the prehensile tails of many, and the hairy buttocks of all, showing that none of them sit erect, distinguish them from the individuals most resembling them in the old world. The Carnivora of America are almost all peculiar to it; indeed, with the exception of some of the marine carnivora, and a very few terrestrial species, all are specifically distinct from those found in other parts of the

Physical Geography. earth. The whole order of Marsupialia are either American or Australian; and the species of one zoological kingdom are unknown in the other. America exceeds every other country in the number of its Rodentia; and there is reason to doubt whether almost any of this order are common to the two worlds. The Edentata occur chiefly in South America, but all are peculiar to the new world. Though America appears at one period to have been rich in the number of its Pachydermata, as their fossil remains testify, yet at its discovery by Europeans there were not above four or five species of this order found in that continent, all peculiar to America, and only one, Dicotyles Torquatus, or Peccari, common to its northern and southern divisions. Among Ruminantia we consider the American Rein-Deer and the Elk as peculiar species; and the identity of the Ovis Montana of the Rocky Mountains and the Argali of Siberia as very doubtful. The order of Cetacea is probably common to both worlds, with the exception of the Manatus Americanus of Cuvier, which occurs on the coast of Florida, and seems also to occur in that of South America. Dr Richardson reckons, that out of 207 species of Mammifera found in North America, 169 are peculiar to that country; and if we take Temminck's estimate of this class to be correct, North America contains $\backslash frac\{4\}\{5\}$th of all known Mammalia. The same author gives the number of North American birds at 696 species, of which 54 are Raptores, 400 Insectores, 33 Rasores, 87 Grallatores, and 122 Nutatores. Except in the last two orders, the species common to the new and old worlds are few, and a great many genera are wholly American.

The Alligator, the Boa Constrictor, and the Rattlesnake, are all peculiar to America. The former and the latter are widely diffused over the American continent; and the Rattlesnake is found even as far north as Canada.

In the new world we begin with the

9. Neonartie Kingdom.—This includes all America between lat. 50° N. and the pole, with Greenland and the intervening islands. The characteristic Mammifera are the Musk Ox, the Black American Bear, the Occidental Wolf, the Wolverine, and, as we have already stated, the Rein-Deer and the Moose-Deer, with several species of Marmot, Squirrel, Lemming, and other animals allied to the genus Mus. The animal discovered by Scoresby in Greenland, the Mus Groenlandicus, has since been discovered in Novia Zembla. The Arctic Fox, Arctic Hare, and the Beaver, are common to both continents. The Mustela Erminea and Ursus Maritimus seem identical in both.

The Raptorial birds peculiar to this region are, several kinds of hawks and owls; of the genus Tetrao, T. Canadensis, T. Franklinii, T. Obscurus, and T. Leucurus. Most of the Nutatores are also found in the first zoological kingdom. The Cygnus Buccinator is peculiar to this region; and among the duck tribe we may notice as American species Anas Valisneria, A. Canadensis, and A. Huteinsonii.

10. Septentrio-American Kingdom.—In this region, on account of the considerable similarity of species, we would include all the British possessions south of lat. 50°, New Albion, and the other country west of the Rocky Mountains, between Queen Charlotte's Sound and New Mexico, and all the territories of the United States, as far as lat. 30° N. It may zoologically be characterized as the region of the Griesly Bear, the Bison, the Wapeti, the giant of the deer tribe, and of the Antelope Fureifer. It possesses one marsupial animal, Dipelphis Virginica, a species which ranges from the lakes of Canada to the intertropical regions of America. But its most distinguishing characteristic is the number of its Rodentia, amounting to not less than fifty-

three well-ascertained species, and only one of which, the Beaver, is found in the old continent.

Its birds are numerous, and among these the Wild Turkey is the most conspicuous and characteristic. Of the Raptores it has many falcons and hawks, and, among the rest, Washington's Eagle, a magnificent bird, which is found in Kentucky. Of its numerous Insectores, the greatest number are peculiar to North America. The Trochilidae first appear in this region. A considerable number of species are found in it, but only three range to the north of latitude 33°. They resemble in structure the Honey-eaters of the Australian kingdom; but we doubt the propriety of denominating the Humming-Birds suetorial; for, having dissected a considerable number of them, we invariably found their stomachs crammed with minute insects. To capture these is probably the object of their fluttering about flowers, the nectar of which they were supposed to sip. Their structure, too, assimilates them to insectivorous birds. The Rasores of this kingdom are all, except a single species of Tetrao, peculiar to America. Of this genus the most remarkable are Tetrao Cupido, T. Umbellus, T. Franklinii, T. Urophasianus, Oxyx Douglasii, O. Picta. Of the numerous Grallatores found in this region a considerable number occur also in Europe, and still more of the Nutatores; but of the former order, the Ardea Herodias, Platalea Alacia, Ciconia Maguari, are peculiar to America; as are of the latter order Pelicanus Americanus, and Rynchops Nigra. The Alligator Lucius abounds in the valley of the Mississippi. Two very peculiar reptiles are there found, the Syren Lacertina of Carolina, and Meopoma Gigantea. These have no animal at all resembling them in the old world, except the Proteus Anguinus of the subterranean lakes of Carniola. Among the numerous tortoises there may here be noticed T. Serpentina, T. Ferox, and T. Clausa. The serpents are numerous; the most remarkable are the Rattlesnakes, four, if not five, distinct species of which are to be found in this kingdom.

11. Equinoxial American Kingdom.—Using this term in an extended sense, we would include under it the regions between the parallels of 30° N. and S. of the equator, but exclusive of the elevated valleys and table-lands of Mexico and Bolivia, which, from the peculiarities of their climates, are entitled to be considered as separate zoological kingdoms.

The division under consideration is distinguished by the number of its Quadrupedana, all of which are furnished with tails, and many of them have that organ prehensile, or so formed as to constitute a sort of fifth hand. It is the region of the Jaguar, a beast of prey of vast strength and courage, so very unlike the account of its habits in the pages of Buffon, that it is now generally supposed that the French naturalist confounded it with another South American Felis, the Ocelot.¹ In this region, too, the Puma abounds, an animal that has far less claim to the name of lion of America than the Jaguar has to be termed the tiger of that continent. The Puma has a considerably larger range than its congeners now mentioned, being found in the woods of America from Brazil to Canada. We once possessed the large skin of a black Jaguar, killed in Brazil, which is now deposited in the museum of the Royal Institution in Liverpool. It is of a beautiful deep blackish-brown ground, on which, in certain lights, the still deeper ocellated marks of the Jaguar are visible. It also abounds with the Tapir, the Capybara, and the Agouti. The Orinoco and other rivers of Equinoxial America swarm with the Manati, Trichechus Amazonus, which wanders far from the sea, as does a species of Porpoise not yet ascertained. It is also distinguished by the splendour of the plumage of its birds, of which numerous genera are

¹ See Humboldt's Personal Narrative.

either wholly peculiar to Tropical America, or are almost unrepresented in other regions. Its Raptorial birds are often distinguished by their size. To this region belong the magnificent King Vulture, and a very numerous species of the same family as large as a turkey, the Vultur Uruba; the Destructor and Harpy Eagles, the giants of their tribe, are the tyrants of the lower provinces. The Insectores of Tropical America are very numerous. Among them the American genera Trogon, Galbula, Ampelis, Rupicola, Procnias, Nectarinia, and Trochilus, are remarkable for the magnificence of their plumage. The bell-bird, Cusma-rhynchus Carunculatus, is celebrated for the deep intonation of its simple note, which simulates the distant bell of a convent; the Troipiales are distinguished by the abrupt yet pleasing contrast of their colours. The enormous Goatsuckers, especially that of the cave of Caripe, and the Momots or Prionites, are peculiar to this region. The Toucans and Aracaris are conspicuous for the size and structure of their bills, in which the organs of smell are conspicuously developed. Tropical America also abounds with beautiful Psittacidae, among which the Ultramarine Parrot, the Scarlet and Blue, and the Blue and Yellow Macaws, are the most conspicuous.

The beautiful family of Craz or Pauxi, of Penelope, the singular Rhea, and Serpent-eater, are among its Gallinaceæ; the Boat-Bill, Canceroma, and the remarkable species Myteria Americana, and Palamedea Cornuta, the Scarlet Ibis, the Trumpeter or Psophia Crepitans, the Jacana, are among its waders.

This region abounds with snakes. Some, like the Boa Constrictor and Boa Cenchris, are remarkable for their enormous size; others, like the Canine Boa, Garden Boa, the Peruvian and Mourning Snakes, for the beauty or elegant pattern of their colours; others, like the Rattlesnake, Crotalus Durissus, or the redoubtable Bushmaster, Fer de Lance or Vipera Megara, dreaded for the virulence of their poison. The fluviatile fishes of this region are remarkable; but we can only here notice the Electric Gymnote, the Soldier Loricaria, and the Salmo Rhombeus; the latter the pest of the South American rivers. The Papilionidae and Phalanidae among its insects, are noted for the splendour of their colours and their size; and the singular Lantern-fly for the brilliancy of its phosphorescence. The wounds inflicted by the large Aranea Avicularia, or Bird-catching Spider, the Scolopendra Morsitans, which there grow to an enormous size, and by a small species of scorpion, are much dreaded. In these Arachnidae, dissection shows a tubular offensive weapon, and a poison apparatus resembling that connected with the poison-fangs of serpents.

12. Mexican Kingdom.—Its animal productions, though but imperfectly explored, would seem to justify the inference drawn from its peculiarities of climate, that this portion of America should be considered as a distinct zoological kingdom. It has been recognised as the point in which the faunas of North and South America meet. There the Wolf of a northern climate ranges the same forest as the Monkey of a tropical region; the Bunting and the Titmouse nestle near the Parrot and the Trogon; the Phalarope of the north searches for its food on the same beach with the Jacana and the Boat-bill of Brazil.¹ Lichtenstein has indicated several species of Weasels and Martens as peculiar to Mexico. The Bassaris Astuta of Lichtenstein is an intermediate genus between Viverra and Nasua. The Mexican Wolf is perhaps a peculiar species. Mr Swainson states, that out of 114 species of Mexican birds examined by him, sixty-seven, or more than one half, are peculiar to that country; yet, among so many species, there was but a single new genus, Ptilogomys, which unites the Tyrant Shrikes with the Ca-

terpillar-eaters; thirty-six species are common to Mexico and the United States, and eleven to Mexico and South America. The lakes of the valley of Mexico contain that singular animal the Azocotl of the Mexicans, the Siren Pisciformis of Shaw, which seems to be intermediate in appearance between the other Sirens and the Protei.

13. Bolivio-Chilian Kingdom.—The vast elevation of the greatest part of this region has strongly impressed its fauna with peculiarities which future research will extend. In the mean time, it is sufficiently characterized as the region of the Guanaco, the Alpaca, and Vicuña, three distinct species, which have sometimes been confounded under the name of Llama. They are the Camels of South America, and were almost the only mammalia subdued and domesticated by the ancient Peruvians. This is the peculiar region of the Condor, a bird which, though not so enormous in size as the earlier travellers reported, is still as large as the Læmergeyer of the Alps, the largest of the Raptors of the old world. It loves to dwell amid the snowy solitudes of the Andes, perched on pinnacles from 9700 to 1500 feet above the sea, whence it pounces on its prey in the subjacent valleys. The fauna of these countries is still imperfectly explored.

14. Austro-American Kingdom.—This district extends from lat. 30. S. to Cape Horn, and embraces the Pampas of Buenos Ayres and the south of Chili. Its fauna has been little explored. Amongst its quadrupeds are the numerous heads of wild horses and sheep, originally introduced by the Spaniards, but now spread over a great part of South America. The Antarctic and Chili Fox seem to be peculiar; and perhaps the Felis Colorolo and F. Guigna of Molina should be considered as belonging to it. On its coasts many species of seals are found, especially the Phoca Longicollis or Fur-seal, Ph. Falklandica, Ph. Flavescens, Ph. Leonina or Bottle-nosed Seal, Ph. Lupina, and Ph. Jubata, which is also found, it is said, in the seas of Kamtschatka. The extensive Pampas of this region are the chosen haunts of the Rhea, or American Ostrich; and on its southern coasts are found the huge, wingless, Patagonian Penguin, with the whole genus Pachyptila, judiciously separated by Illiger from the Petrels.

Such is an imperfect sketch of the distribution of animals over the globe. But the natural limitation of species has been in some measure affected by human agency. The domesticated animals have been by man imported from different parts of Asia into Europe, and finally into America. At the discovery of that continent it was without the horse, the cow, the sheep, the hog, the dog, and our common poultry, all which are spread over it in innumerable herds, and in some places have relapsed into the wild state in countries well suited for their subsistence. The same useful animals have been by Europeans, within the last half century, carried to the larger islands of the Pacific, where they were previously unknown. How many insects may have been propagated by the cargoes of our ships in distant lands, it is easier to conjecture than to estimate; how many have been imported with the cerealia, and other grain of Europe, into newly-discovered regions, it is impossible to say. Human agency has sometimes been the means of propagating in Europe disgusting or destructive species from foreign regions. Thus the commerce of the Dutch wafed the Teredo Navalis to the dyke-defended coasts of Holland, to the imminent hazard of that country; the Brown Rat and the Blatta, which now infest this country, are believed to be importations from the East Indies; and the white bug that now lays waste our orchards is stated to have reached us with American fruit-trees.

¹ Richardson, Report of the British Association, vol. v.; and Address, vol. vi.

Still the effect of human agency is confined, and the limitations imposed by nature upon animal migrations are generally preserved.

SECT. XI.—Varieties of the Human Species.

The identity of species in the whole human family is indicated in the sacred Scriptures, and is confirmed by the deductions of the physiologist and the anatomist. The true principle which serves to mark identity of species in the animal kingdom, is the propagation, not of hybrids, but of fertile progeny between a male and female of a different race. The offspring between individuals of all the known races of mankind are equally fertile with the progeny of a pair of the same nation; and, therefore, it can be no longer doubted that man is everywhere of the same species. There are, however, several well-marked varieties of the human species, that appear distributed in a peculiar manner over the earth; and hence the Geographical Distribution of Man becomes an object of our attention.

Man appears to possess a remarkable pliancy of constitution, and to be the only animal fitted by nature to inhabit every possible variety of climate. His animal existence is not less secured in the frozen regions of the arctic circle, or under the burning climate of the equator, than in the intermediate countries; and he may be said to be the only true denizen of the whole world.

The classification of Cuvier, in his Tableau Élémentaire, divided the human race into five varieties, viz. 1. the White Race; 2. the Lapland, Samoied, and Esquimaux Race; 3. the Mongolian Race; 4. the Negro Race; 5. the American Race. In his Regne Animal he seems inclined to confine them to three, of which the characteristics are very strongly marked, the White, the Yellow, and the Black Races; but he admits that the Malays and the Americans are not easily distinguishable from any of the three great varieties of mankind in the ancient continent. Blumenbach considers the varieties as five, viz. the Caucasian, the Mongolian, the Ethiopian, the Malayan, and the American; and this distribution, though not altogether free from objection, seems one of the best hitherto proposed.

1. Caucasian Variety, is so named, because the traditions respecting the diffusion of the human race seem to point out the region between the Caspian and Black Sea as the cradle of this race, from which there have been numerous radiations; and the most perfect specimens, in point of beauty of form, are found among the Georgians and Circassians, still inhabiting the native seats of their race. The characteristics of the Caucasian Variety are, the fair skin, the oval contour of the head; soft hair, varying from black to light brown or flaxen, wavy or slightly curled; eyes varying from blue to dark brown; cranium expanded; facial angle large; nose thin, and rather aquiline or straight; small mouth, perpendicular cutting teeth; lips gently recurved; chin full and rounded. In this race the intellectual qualities have been more strikingly displayed than in any other. Its principal branches may be still traced by certain, though often not very obvious, affinities in the structure of language. The most ancient offsets appear to have been that to the south and the east. The Southern Armenian or Syrian branch seems to have been the parent stock of the Assyrians, the Chaldeans, the Arabs, the Jews, the Phoenicians, and Abyssinians. The examination of the most ancient mummies would lead to the conclusion that the ancient Egyptians belonged to the same stock, though the Ethiopian countenances of some of their sculptures have sometimes led to a different conclusion. We believe, however, that these may be transcripts of venerated originals imported with the religion of Egypt from the Nubian Meröe.

The second branch appears to have given rise to the ancient Medes, the Persians, the Afghans, and the higher

castes of India whose language was the Sanscrit; whilst another portion of it, taking a westerly direction, spread itself over Asia Minor, and, under the name of Pelasgians, penetrated into Europe. It seems to have been preceded by an offset from the same branch, the Celts, who early penetrated to the western shores of Europe, where they were found, after the lapse of centuries, by the Teutones, a younger shoot from the same stock. The whole nations already mentioned, and the whole present inhabitants of Europe, with the exception of the Samoieds and Laps, are derived from the Caucasian Variety, the most scattered and most energetic of the human family; and the affinity in descent will account for the strong resemblances which have been traced between the languages of India, of ancient Media, and the principal European tongues. The Scythian branch of the Caucasian race may be considered as the ancestors of the Parthians, the Tartars of the Taurida, the Turcomans, the Finlanders, the Hungarians, and the tribes about the mouths of the Danube and to the north-east of the Euxine.

2. The Mongolian Variety seems to have had its original seat in the vicinity of the Altaiian chain of mountains. It is characterized by an olive complexion, and black eyes, the outer angles of which are pointed rather upward; their hair coarse, lank, and thin, their beard scanty; the head of a square form, with a low and narrow forehead, a broad and flattened face, high cheek-bones, flat nose, a wide mouth, and thick lips. Two branches of this race are represented by the Nomads Kalmuchs and Kalkas, whose ancestors spread devastation over a large portion of the earth, under their leaders Attila, Zenghiz Khan, and Timur. A third branch, the Mantschou, in later times had conquered China, of which they are still the rulers. The ancient Chinese, however, seem to have belonged to the same race; and the Japanese, the Coreans, and people of Thibet, probably are also offsets from the great Mongolian variety. The Esquimaux of America appear from their form to belong also to the Mongolian Variety; and may have passed into America on the arctic ice, just as rein-deer now do annually; or as the Tschutskoi have been frequently known to do in order to attack the American Indians. The Laps, the Samoieds, the Tschutskoi, and the Kamtschatkades, bear strong marks of their Mongolian origin, though some are disposed to regard them as derived from the Scythian branch of the Caucasian race.

3. The Ethiopian Variety owns Africa as its native region. This variety is distinguished by a black or very dark skin; black eyes; woolly, crisp, coarse hair, collected into little knots; a skull laterally compressed; a forehead low and narrow, a small facial angle, produced by the moderate projection of the maxillary bones, with an oblique position of the front teeth, and a small chin; the nose flat and simous; the lips, especially the upper, thick; the arms long; the legs often slightly bowed. Lawrence has satisfactorily shown that there is not one peculiarity of the negro race, which is not occasionally met with in some of the other races; and that there is no reason to believe that the negro is not of the same origin as the rest of mankind. Mr Browne, and other recent travellers, have stated, that the Furians and Nubians, though black, have often handsome elevated features; and we have seen a jetty negress, from the eastern shores of Africa, with features that might well have been a model for a Grecian statue of Juno in bronze. Those variations are probably from the intermixture of the Caucasian with the Negro race.

In the Malayan peninsula, in Luconia, and in Borneo, among the mountainous districts, are found a few scattered tribes of black men, who form the chief population of Papua or New Guinea. They have dark skins and woolly hair, and, if we may judge from the native of Papua brought to England by Sir Stamford Raffles, the Papous do not ma-

terially differ from the natives of the eastern coast of Africa, whence probably their ancestors have emigrated.

4. The Malayan Variety has its native seats in the Malayan peninsula and the adjacent islands. The inhabitants of Polynesia, New Holland, and New Zealand, are believed to belong to this race. It is far less characterized than the three races before noticed, and perhaps might be considered as produced by their intermixture. The perspicuous characteristics are a brown colour; sometimes almost approaching to white, as in the inhabitants of Tahiti and the Marquesas, at other times very swarthy, as in the New Zealander and Australian. The hair is black, thick, and generally slightly curled; the head is laterally compressed, the forehead rather narrow, but high; the bones of the face large; the nose full, broad at the apex, and passing gradually into the cheeks; the mouth large; the lips rather thick; the eyes sometimes have the Mongolian cut; the hands are small. But many of the islands are occupied by a very handsome race of inhabitants. The Malay language is widely scattered over the islands of the Pacific, which shows a common origin. The recent South Sea voyages show the vast distances to which design or accident has carried the people of one group of islands to another in their frail canoes, and point out how those islands may have been originally peopled.

5. The American Variety was spread over all that continent at its discovery in the end of the fifteenth century. This race, like the last, is not very distinctly characterized, but bears a very strong resemblance to some of the Mongolian tribes in the scantiness of the beard, which is general in both parts of that continent. The principal peculiarities of this race, besides the scanty beard, which they generally try entirely to eradicate, are a dark skin, with a tint of red, or what has, not very happily, been termed a copper-colour. The hair is black, lank, long, and coarse. The skull is very similar to the Mongolian, but the features are more prominent, especially the nose; the forehead is generally low and retreating; the eyes deeply set; and the face, across the cheek-bones, broad, the mouth large, the lips rather thick. Some of the native tribes singularly flatten the head by compression while in the infant state. This is particularly the case with a tribe on the Columbia. One of their skulls, in the College Museum of Edinburgh, is depressed in a most extraordinary degree, so that the frontal sinus seems almost as high as the vertex; and the head is posteriorly and laterally extended. This was the skull of a person of rank. It was procured by the late Dr Gardner, along with the skull of a slave of the same tribe, in which this flattening, no doubt esteemed a great beauty, had not been practised, but which exhibited, to our eyes, a much more handsome form than that of his master.

The peopling of America was long a problem of much difficulty; but it has of late years been elucidated in various respects. Strong affinities have been discovered between the languages of America and Eastern Asia, which confirm the inferences drawn from physiognomy. Sidi Melli-melli, Tunisian envoy to the United States in 1804, on seeing the deputies of the Cherokees, the Miamis, and Osages, assembled at Washington, instantly recognised their Tatar physiognomies; and Genet, when minister plenipotentiary of France, was also struck with the resemblance of the Indians of America to the Asiatic Tatars, with whose appearance he was familiar. The late researches of Dr Heckewelder, and other American archaeologists, have reduced the supposed infinite variety of North American tongues to three or four radical languages. According to this authority, all the North American native tongues may be traced to the

Floridan, the Lené-Lenapé, the Iroquois, and the Chepewyan; and the researches of Klaproth, and other German philologists, have strengthened the belief in the affinities between the native American languages and those of Eastern Asia.

The traditions of the Mexicans, and their hieroglyphical writings, trace their migrations from the north to the table-land of New Spain. Their mythology and their calendars strongly indicate an Asiatic origin, whilst their architecture and their pyramidal temples have something of an Indian or Egyptian character. The scanty annals of this singular people show, that migrations from the north-west had at different epochs reached the elevated table-land of Mexico.

The first occupiers, or aborigines, are named the Ol-mecs. How long they had been there does not appear; but they were driven out in the year 648 of our era by the Toltecks, a people who left their country, which they named Tlapallan, supposed to be in Eastern Asia, in the year 544. A pestilence seems to have weakened this nation in 1051; and in 1170 the Chichimacks migrated into Mexico, where, eight years afterwards, the Anahuatlachs also arrived. This last nation consisted of seven tribes, the principal of which were the Aztecks. This tribe called their native country Aztlan, which they left in 1054, according to Gama; and, uniting themselves with the Chichimacks and the remnant of the Toltecks, they founded the city of Tenochtitlan in 1325. It was this united people whom the Spaniards denominated Mexicans, and whose empire was extinguished at the capture of the capital, Tenochtitlan, by Cortez in 1521.

The Toltecks appear to have pushed colonies into South America, either by the Cordillera of the Andes, or by the eastern plains at their feet. When Gonzales Ximenez de Quesada arrived on the table-land of Bogota, he was surprised to find a population on that plateau in a state of considerable civilization. It consisted of four tribes, all clothed in cotton garments, and settled in agricultural communities. The principal tribe was the Muyseca, who possessed a calendar, the intercalations of which had as striking a similarity to that employed by nations on the banks of the Indus, as that of the Mexicans approached the calendar of the Mantechous; and it is believed that the Muyseca deduced their calendar from the motions of Jupiter in the ecliptic.

Several travellers have been struck with the resemblance of the Peruvian to the Malayan race; and it is possible that, by the isles of the Pacific, their ancestors reached the southern Cordillera. Their architecture has a sort of Egyptian character in the form of its apertures, and the squaring of its blocks of stone; but their theocracy had much resemblance to that of some nations of Central Asia.

Our limits preclude our pursuing this interesting subject further, or entering on the consideration of the alleged causes of the varieties of the human species. We must refer our readers to the Researches of Humboldt, to the Physiology of Lawrence, to L'Histoire Naturelle du Genre Humain of Virey, to the Dissertations of Blumenbach, Cuvier, and Klaproth, to the work of Dr Smith of New Jersey, to Prichard's History of Man, and to the article MAN in the present work. Suffice it to state, that whilst some attribute the varieties of the human family to the long-continued action of climate, food, and the external condition or habits of life, others have regarded them as the consequences of accidental peculiarities transmitted to the posterity of the individuals affected by them; and a third class have considered the three great varieties of the human race as directly derived from the individual qualities of the sons or Noah. But whatever view we take, it will not be found the less difficult to account for the marked national pecu-

¹ Humboldt, Recherches.

² Ibid.

Physical similarities of people confessedly derived originally from the Geography same stock.

In conclusion, we may remark, that the study of physical geography, both as it regards inorganic and animated nature, tends to exercise the intellect in tracing the connection between natural causes and their consequences, and to

expand and elevate the mind to the noblest of human contemplations,—the wisdom and beneficence of the Deity.

"I cannot go
Where universal Love not smiles around,
Sustaining all yon orbs, and all their sons;
From seeming evil still educating good,
And better thence again a better still,
In infinite progression."

EXPLANATIONS OF REFERENCES TO TABLES OF MOUNTAINS.
PLATE CCCCXII.