to recede in a contrary direction; or whether these principles cooperate, or unequally oppose each other, as has been ingeniously contended, I shall not take upon me to decide.  It is sufficient to say that such an effect appears to be the first general law of the tropical winds.  Whatever may be the degree of the sun’s influence upon the atmosphere in his transient diurnal course, it cannot be doubted but that, in regard to his station in the path of the ecliptic, his power is considerable.  Towards that region of the air which is rarefied by the more immediate presence of the heat, the colder and denser parts will naturally flow.  Consequently from about, and a few degrees beyond, the tropics, on either side, the air tends towards the equator; and, combining with the general eastern current before mentioned, produces (or would, if the surface were uniform) a north-east wind in the northern division, and a south-east in the southern; varying in the extent of its course as the sun happens to be more or less remote at the time.  These are denominated the trade-winds, and are the subject of the second general observation.  It is evident that, with respect to the middle space between the tropics, those parts which at one season of the year lie to the northward of the sun, are, during another, to the southward of him; and of course that an alteration of the effects last described must take place, according to the relative situation of the luminary; or in other words, that the principle which causes at one time a north-east wind to prevail at any particular spot in those latitudes must, when the circumstances are changed, occasion a south-east wind.  Such may be esteemed the outline of the periodical winds, which undoubtedly depend upon the alternate course of the sun northwards and southwards; and this I state as the third general law.  But although this may be conformable with experience in extensive oceans, yet, in the vicinity of continents and great islands, deviations are remarked that almost seem to overturn the principle.  Along the western coast of Africa and in some parts of the Indian seas, the periodical winds, or monsoons as they are termed in the latter, blow from the west-north-west and south-west, according to the situation, extent, and nature of the nearest lands; the effect of which upon the incumbent atmosphere, when heated by the sun at those seasons in which he is vertical, is prodigious, and possibly superior to that of any other cause which contributes to the production or direction of wind.  To trace the operation of this irregular principle through the several winds prevalent in India, and their periodical failures and changes, would prove an intricate but, I conceive, by no means an impossible task.* It is foreign however to my present purpose, and I shall only observe that the north-east monsoon is changed, on the western coast of Sumatra, to north-west or west-north-west by the influence of the land.  During the south-east monsoon the wind is found to blow there, between that point and south.  Whilst the sun continues near the equator the winds are variable, nor is their direction fixed till he has advanced several degrees towards the tropic:  and this is the cause of the monsoons usually setting in, as I have observed, about May and November, instead of the equinoctial months.