All around the earth within the tropics, over the land and over the sea, there is a belt of almost daily rains, varying in width, north and south, in different sections, but averaging about five hundred miles. This belt of daily rains is formed at and by the meeting of N. E. and S. E. trades, and travels north and south with them, as they do with the sun, _encircling the globe_. By this narrow belt a portion of the earth's surface, an average of some 35 deg. of latitude, is supplied with moisture. Wherever it is situated at any given period, the tropical rainy season exists; and when it is absent in its northern or southern transit, the dry season prevails. Southern Mexico is within the range of this moving belt, and in its course to the northward with the sun, in our summer from May to October, it arrives over, and covers that country with a rainy season. When the sun returns to the south, taking with it the trades and this belt of tropical rains, that portion of Mexico is without rain, and dry, and so continues until the rainy belt returns in the following year. While the belt is over Southern Mexico it is nearly all _precipitation_, and there is little _evaporation_; while that belt is _absent_ it is all _evaporation_, with little or no _rain_. Surely this is not consistent with the prevailing belief of simple evaporation, ascent to a colder stratum, commingling, and condensation, and rain. Southern Mexico at least is not supplied by mere evaporation from its surface, and must therefore form an exception to that belief, and to the Huttonian theory. But we shall recur again to the peculiarity of distribution within the tropics. Turn now for a brief space to Northern Mexico, Southern New Mexico, and Southern California. In Northern Mexico, Southern New Mexico, Utah, and California, between the parallels of 28 deg. and 32 deg., and particularly west of the mountain ranges, we find an almost rainless region, sterile and worthless, resembling that which is found upon nearly the same parallels of north latitude in Northern Africa, Egypt, Arabia, Beloochistan, Afghanistan, and North-western India; and in corresponding latitudes south of the Equator, in Peru, a portion of Southern Africa, and the northern and middle portions of New Holland. Why Northern Mexico and the other countries named are thus sterile and comparatively rainless, we shall see hereafter, when we examine critically the machinery of distribution as it operates within the tropics. It is the fact that it is thus sterile and rainless to which we desire to call attention in this place. Mr. Bartlett thus describes it: