The islands form scarcely 1/22d of the continental masses, which are so unequally divided that they consist of three times more land in the northern than in the southern hemisphere; the latter being, therefore, pre-eminently oceanic. From 40 degrees south latitude to the Antarctic pole the earth is almost entirely covered with water. The fluid element predominates in like manner between the eastern shores of the Old and the western shores of the New Continent, being only interspersed with some few insular groups. The learned hydrographer Fleurieu has very justly named this p 289 vast oceanic basis, which, under the tropics, extends over 145ºdegrees of longitude, the 'Great Ocean', in contradistinction to all other seas. The southern and western hemispheres (reckoning the latter from the meridian of Teneriffe) are therefore more rich in water than in any other region of the whole earth.
These are the main points involved in the consideration of the relative quantity of land and sea, a relation which exercises so important an influence on the distribution of temperature, the variations in atmospheric pressure, the direction of the winds, and the quantity of moisture contained in the air, with which the development of vegetation is so essentially connected. When we consider that nearly three fourths of the upper surface of our planet are covered with water,* we shall be less surprised at the imperfect condition of meteorology before the beginning of the present century, since it is only during the subsequent period that numerous accurate observations on the temperature of the sea at different latitudes and at different seasons have been made and numerically compared together.
[footnote] *In the Middle Ages, the opinion prevailed that the sea covered one seventh of the surface of the globe, an opinion which Cardinal d'Ailly ('Imago Mundi', cap. 8) founded on the fourth apocryphal book of Esdras. Columbus, who derived a great portion of his cosmographical knowledge from the cardinal's work, was much interested in upholding this idea of the smallness of the sea, to which the misunderstood expression of "the ocean stream" contributed not a little. See Humboldt, 'Examen Critique de l'Hist. de la Geographie', t. i., p. 186.
The horizontal configuration of continents in their general relations of extension was already made a subject of intellectual contemplation by the ancient Greeks. Conjectures were advanced regarding the maximum of the extension from west to east, and Dicaearchus placed it, according to the testimony of Agathemerus, in the latitude of Rhodes, in the direction of a line passing from the Pillars of Hercules to Thine. This line, which has been termed 'the parallel of the diaphragm of Dicaearchus', is laid down with an astronomical accuracy of position, which, as I have stated in another work, is well worthy of exciting surprise and admiration.*
[footnote] *Agathemerus, in Hudson, 'Geographi Minores', t. ii., p. 4. See Humboldt, 'Asie Centr.', t. i., p. 120-125.
Strabo, who was probably influenced by Eratosthenes, appears to have been so firmly convinced that this parallel of 36 degrees was the maximum of the extension of the then existing world, that he supposed it had some intimate connection with the form of the earth, and therefore places under this line the continent whose existence p 290 he divined in the northern hemisphere, between Theria and the coasts of Thine.*
[footnote] *Strabo, lib. i., p. 65, Casaub. See Humboldt, 'Examen Crit.', t. i., p. 152.
As we have already remarked, one hemisphere of the earth (whether we divide the sphere through the equator or through the meridian of Teneriffe) has a much greater expansion of elevated land than the opposite one: these two vast ocean-girt tracts of land, which we term the eastern and western, or the Old and New Continents, present, however, conjointly with the most striking contrasts of configuration and position of their axes, some similarities of form, especially with reference to the mutual relations of their opposite coasts. In the eastern continent, the predominating direction — the position of the major axis — inclines from east to west (or, more correctly speaking, from southwest to northeast), while in the western continent it inclines from south to north (or, rather, from south-southeast to north-northwest). Both terminate to the north at a parallel coinciding nearly with that of 70ºdegrees, while they extend to the south in pyramidal points, having submarine prolongations of islands and shoals. Such, for instance, are the Archipelago of Tierra del Fuego, the Lagullas Bank south of the Cape of Good Hope, and Van Diemen's Land, separated from New Holland by Bass's Straits. Northern Asia extends to the above parallel at Cape Taimura, which, according to Krusenstern, is 78 degrees 16', while it falls below it from the mouth of the Great Tschukotsehja River eastward to Behring's Straits, in the eastern extremity of Asia — Cook's East Cape — which, according to Beechey, is only 66 degrees E.*
[footnote] *On the mean latitude of the Northern Asiatic shores, and the true name of Cape Taimura (Cape Siewere-Wostotschnoi), and Cape Northeast (Schalagskoi Mys), see Humboldt, 'Asie Centrale', t. iii., p. 35, 37.
The northern shore of the New Continent follows with tolerable exactness the parallel of 70 degrees, since the lands to the north and south of Barrow's Strait, from Boothia Felix and Victoria Land, are merely detached islands.