prevented each other’s nearer approach.  That island, however, does not experience the vicissitude of land and sea breezes, being too small, and too lofty, and situated in a latitude where the trade or perpetual winds prevail in their utmost force.  In sandy countries, the effect of the sun’s rays penetrating deeply, a more permanent heat is produced, the consequence of which should be the longer continuance of the sea-breeze in the evening; and agreeably to this supposition I have been informed that on the coast of Coromandel it seldom dies away before ten at night.  I shall only add on this subject that the land-wind on Sumatra is cold, chilly, and damp; an exposure to it is therefore dangerous to the health, and sleeping in it almost certain death.

Soil.

The soil of the western side of Sumatra may be spoken of generally as a stiff, reddish clay, covered with a stratum or layer of black mould, of no considerable depth.  From this there springs a strong and perpetual verdure of rank grass, brushwood, or timber-trees, according as the country has remained a longer or shorter time undisturbed by the consequences of population, which, being in most places extremely thin, it follows that a great proportion of the island, and especially to the southward, is an impervious forest.

Unevenness of surface.

Along the western coast of the island the low country, or space of land which extends from the seashore to the foot of the mountains, is intersected and rendered uneven to a surprising degree by swamps whose irregular and winding course may in some places be traced in a continual chain for many miles till they discharge themselves either into the sea, some neighbouring lake, or the fens that are so commonly found near the banks of the larger rivers and receive their overflowings in the rainy monsoons.  The spots of land which these swamps encompass become so many islands and peninsulas, sometimes flat at top, and often mere ridges; having in some places a gentle declivity, and in others descending almost perpendicularly to the depth of a hundred feet.  In few parts of the country of Bencoolen, or of the northern districts adjacent to it, could a tolerably level space of four hundred yards square be marked out.  I have often, from an elevated situation, where a wider range was subjected to the eye, surveyed with admiration the uncommon face which nature assumes, and made inquiries and attended to conjectures on the causes of these inequalities.  Some choose to attribute them to the successive concussions of earthquakes through a course of centuries.  But they do not seem to be the effect of such a cause.  There are no abrupt fissures; the hollows and swellings are for the most part smooth and regularly sloping so as to exhibit not unfrequently the appearance of an amphitheatre, and they are clothed with verdure from the summit to the edge of the swamp.  From this latter circumstance it is also evident that they