In all that has been said above it has been assumed that the quickest-moving or equatorial belt of the earth is also the hottest, and consequently that over which the air has the greatest tendency to rise.  But, although this is generally true, it is not, by any means, universally so.  The variations, however, which are observed to occur in those places where the circumstances form an exception to the general rule, tend strongly to confirm the theory of Hadley.  The monsoons of India, as I shall presently show, are examples of this; but the most striking instance with which I am personally acquainted occurs in the Pacific Ocean, between the Bay of Panama and the Peninsula of California, from latitude 8 deg. to 22 deg. north.  If the huge continent of Mexico were taken away, and only sea left in its place, there can be no doubt but the ordinary phenomena of the Trade-winds would be observable in that part of the Pacific above mentioned.  Cool air would then be drawn from the slow moving parallels lying to the northward, towards the swift moving latitudes, near the equator, in order to supply the place of the rarefied air removed to the higher regions of the atmosphere, and, of course, north-easterly breezes would be produced.  But when the sun comes over Mexico, that vast district of country is made to act the part of an enormous heater, and becomes a far more powerful cause of rarefaction to the superincumbent air than the ocean which lies between it and the equator.  Accordingly, the air over Mexico, between the latitudes of 10 deg. and 30 deg., is more heated than that which lies over the sea between the line and latitude 20 deg.; and as the coolest, or least heated, that is, the most dense fluid, always rushes towards the place lately occupied by the hottest and most buoyant, the air from the equator will be drawn towards the coast of Mexico, the great local source of heat and rarefaction.

But as this equatorial air is of course impressed with a more rapid eastern velocity than those parts of the earth which form the southern shores of Mexico, a westerly wind must be produced by the relative difference in these two motions.  At that particular season of the year when the sun is in high southern declination, Mexico is not exposed to his perpendicular rays.  The equatorial regions are then more heated than Mexico, and accordingly we actually find north-easterly breezes nearly as they would be if Mexico were out of the way, and quite in accordance with our theory.

In like manner, in the Atlantic, when the sun is far to the north, the great deserts of the western angle, or shoulder of Africa, become as vehemently heated, or more so, perhaps, than Mexico, and this draws the air from the equator, so as to produce the south-westerly winds I have already spoken of in the troublesome range called the Variables.