The ancients conceived that of the five zones into which they divided the surface of the globe, two only were habitable, supposing that the heat between the tropics, and the cold within the polar circles, were too intense to be supported by mankind. The falsehood of this reasoning has been long evinced; but the particular comparisons of the heat and cold of these various climates has as yet been very imperfectly considered. However, enough is known safely to determine this position, that all places between the tropics are far from being the hottest on the globe, as many of those within the polar circles are far from enduring that extreme degree of cold to which their situation should seem to subject them: that is to say, that the temperature of a place depends much more upon other circumstances than upon its distance from the pole, or its proximity to the equinoctial. This proposition relates to the general temperature of places, taking the whole year round; and in this sense it cannot be denied that the city of London, for instance, enjoys much warmer seasons than the bottom of Hudson's Bay, {170} which is nearly in the same latitude with it, but where the severity of the winter is so great that it will scarcely permit the hardiest of our garden plants to live. And if the comparison be made between the coast of Brazil and the western shore of South America, as, for example, betwixt Bahia and Lima, the difference will be still more considerable; for though the coast of Brazil is extremely sultry, yet the coast of the South Seas in the same latitude is perhaps as temperate and tolerable as any part of the globe, since in ranging along it we did not once meet with so warm weather as is frequent in a summer's day in England: which was still the more remarkable as there never fell any rains to refresh and cool the air. The causes of this temperature in the South Seas are not difficult to be assigned, and shall be hereafter mentioned. I am now only solicitous to establish the truth of this assertion, that the latitude of a place alone is no rule whereby to judge of the degree of heat and cold which obtains there. Perhaps this position might be more briefly confirmed by observing, that on the tops of the Andes, though under the equinoctial, the snow never melts the whole year round: a criterion of cold stronger than what is known to take place in many parts far removed within the polar circle.