This method of determining the gene ral form of the earth is much less liable to error and irregularity, than the mea surement of the lengths of degrees in. various parts, since the accidental varia tions of curvature produced by local dif ferences of density, and even by superfi cial elevations, may often produce con siderable errors in the inferences which might be deduced from these measure ments. For example, a degree measured at the Cape of Good Ilr:pe, in latitude 33° south, was found to be longer than a degree in France, in latitude. 56 north, and the measurements in Austria, in North America, and in England, have all exhibited signs of similar irregularities. There appears also to be some difference in the length of degrees under the same latitude, and in different longitudes. We may, however, imagine a regular elliptic spheroid to coincide very nearly with any small portion of the earth's surface, although its form must be somewhat dif ferent for different parts : thus for the greater part of Europe, that is, for Eng land, France, Italy, and Austria, if the measurements have been correct, this oscillating spheroid must have an ellip ticty of The earth is astronomically divided into zones and into climates. The tor rid zone is limited by the tropics, at the distance of 23° 28' on each side of the equator, containing all such places as have the sun sometimes vertical, or imme. diately over them: the frigid zones are within the polar circles, at the same dis tance from the poles, including all places which remain annually within the limit of light and darkness, for a whole diurnal rotation of the earth, or longer : the tern perate zones between these, have an uninterrupted alternation of day and night, but are never subjected to the sun's vertical rays. At the equator, th ere fore, the sun is vertical at the equinoxes, his least meridian altitude is at the sol stices, when it is 66° 32', that is, more than with us at midsummer ; and this happens once on the north, and once on the south side of the hemisphere. Be tween the equator and the tropics he is vertical twice in the year, when his de clination is equal to the latitude of the place, and his least meridian altitudes, which are unequal between themselves, are at the solstices. At the tropics, the meridian sun is vertical once only in the year, and at the opposite solstice, or the time of midwinter, his meridian altitude is 4', as with us in April and the be ginning of September. At the polar cir cles the sun describes, on midsummer day, a complete circle, touching the north or south point of the horizon ; and in mid winter he shews only half his disc above it, for a few minutes, in the opposite point; that is, neglecting the elevation produced by refraction, which, in these climates especially, is by no means inconsiderable. At either pole, the corresponding pole of the heavens being vertical, the sun must annually describe a spiral, of which each coil is nearly horizontal, half of the spiral being above the horizon, and half below ; the coils being much more open in the middle than near the end.
The climates, in the astronomical sense of the word, are determined by the dura tion of the longest day in different parts of the earth's surface; but this division is of no practical utility, nor does it furnish any criterion for judging of the climate in a meteorological sense.
The natural division of the surface of the globe is into sea and land ; about three-fourths of the whole being occupied by water, although probably no where to a depth comparatively very considerable, at most of a few miles on an average. The remaining fourth consists of lands, elevated more or less above the level of the sea, interspersed, in some parts, with smaller collections of water, at various heights, and in a few instances, some• what lower than the general surface of the main ocean. Thus the Caspian Sea is said to be about three hundred feet lower than the ocean; and in the interior parts of Africa there is probably a lake equally depressed.
We cannot observe any general sym metry in this distribution of the earth's surface ; excepting that the two large continents of Africa and South America have some slight resemblance in their forms, and that each of them is terminat ed to the eastward by a collection of numerous islands. The large capes pro
jecting to the southward have also a similarity with respect to their form, and the islands near them ; to the west the continents are excavated into large bays, and the islands are to the east : thus Cape Horn has the Falkland Islands ; the Cape of Good Hope, Madagascar ; and Cape Comorin, Ceylon, to the east.
The great continent, composed of Eu rope, Asia, and Africa, constitutes about a seventh of the whole surface of the earth : America about a sixteenth ; and Australasia, or New South Wales, about a fiftieth ; or in hundreth parts of the whole, Europe contains two ; Asia, seven; Africa, six; America, six ; and Austra lasia, two: the remaining seventy-seven being sea ; although some authors assign seventy-two parts only out of one hun dred to the sea, and twenty-eight to the land.
These proportions may be ascertained with tolerable accuracy by weighing the paper made for covering a globe, first entire, and then cut out according to the terminations of the different countries ; or, if still greater precision were requir ed, the greater part of the continents might be divided into known portions of the whole spherical surface, and the remaining irregular portions only weigh ed.
The general inclinations and levels of the continents are discovered by the courses of their rivers. Of these the prin cipal are, the river of Amazons, the Sene gal, the Nile, the river St. Lawrence, the Hoangho, the river La Plata, the Jenisei, the Mississippi, the Volga, the Oby, the Amur, the Oronooka, the Gan ges, the Euphrates, the Danube, the Don, the Indus, the Dnieper, and the Dwina ; and this is said to be nearly the order of their magnitudes. But if we class them according to the length of country through which they run, the order will, according to Major Bennel's calculation, be some what different ; taking the length of the Thames for unity, he estimates that of the River Amazons at 154; the Klan Kew, in China, 15f ; the Hoangho, l3i ; the Nile, 12i; the Lena, 11 ; the Amur, 11; the Oby, 10i ; the Jenisei, 10 ; the Ganges, its companion the Burampooter, the river of Ava, and the Volga, each 9/ ; the Euphrates, 8i; the Mississippi, 8 ; the Danube 7; the Indus, 53-; and the Rhine, 5f.
We may form a tolerable accurate idea of the levels of the ancient continent, by tracing a line across it in such a direction as to pass no river, which will obviously indicate a tract of country higher than most of the neighbouring parts. Begin ning at Cape Finisterre, we soon arrive at the Pyrenees, keeping to the south of the Garronne and the Loire.
After taking a long turn northwards, to avoid the Rhine, wet come to Swisser land, and we may approach very near to the Mediterranean, in the state of Genoa, taking care not to cross the branches of the Po. We make a circuit in Swisser land, and pass between the sources of the Danube, and of the branches of the Rhine, in Swabia. Crossing Franconia, we leave Bohemia to the north, in order to avoid the Elbe ; and coming near to the borders of Austria, follow those of Hungary to the south of the Vistula. The Dnieper then obliges us to go northwards through Lithuania, leaving the Don whol ly to the right ; and the Volga, to pass still further north, between Petersburg and Moscow, a little above Bjelosero. We may then go eastwards to the boun dary of Asia, and thence northwards to Nova Zembla. Hence we descend to the west of the Oby, and then to the east of the branches of the Volga, and the other inland rivers flowing into the lake Aral and the Caspian Sea. Here we are situat ed on the widely extended elevation of India, in the neighbourhood of the sour ces of the Indus ; and, lastly, in our way from hence towards Kamschatka, we leave the Jenisei and Lena on the left, and the Ganges, the Kiang Kew, the Hoangho, and the Amur, to the right.