summers and winters of the whole Earth become more or less strongly contrasted, as the excentricity of its orbit increases and decreases. Hence during increase of the excentricity, the epochs of moderately contrasted seasons and epochs of strongly contrasted seasons, through which alternately each hemisphere passes, must grow more and more different in the degrees of their contrasts; and contrariwise during decrease of the excentricity. So that in the quantity of light and heat which any portion of the Earth receives from the sun, there goes on a quadruple rhythm: that of day and night; that of summer and winter; that due to the changing position of the axis at perihelion and aphelion, taking 21,000 years to complete; and that involved by the variation of the orbit’s excentricity, gone through in millions of years. * * * * * § 95. Those terrestrial processes whose dependence on the solar heat is direct, of course exhibit a rhythm that corresponds to the periodically changing amount of heat which each part of the Earth receives. The simplest, though the least obtrusive, instance is supplied by the magnetic variations. In these there is a diurnal increase and decrease, an annual increase and decrease, and a decennial increase and decrease; the latter answering to a period during which the solar spots become alternately abundant and scarce: besides which known variations there are probably others corresponding with the astronomical cycles just described. More obvious examples are furnished by the movements of the ocean and the atmosphere. Marine currents from the equator to the poles above, and from the poles to the equator beneath, show us an unceasing backward and forward motion throughout this vast mass of water—a motion varying in amount according to the seasons, and compounded with smaller like motions of local origin. The similarly-caused general currents in the air, have similar annual variations similarly modified. Irregular as they are in detail, we still see in the monsoons and other tropical atmospheric disturbances, or even in our own equinoctial gales and spring east winds, a periodicity sufficiently decided. Again, we have an alternation of times during which evaporation predominates with times during which condensation predominates: shown in the tropics by strongly marked rainy seasons and seasons of drought, and in the temperate zones by corresponding changes of which the periodicity, though less definite, is still traceable. The diffusion and precipitation of water, besides the slow alternations answering to different parts of the year, furnish us with examples of rhythm of a more rapid kind. During wet weather, lasting, let us say, over some weeks, the tendency to condense, though greater than the tendency to evaporate, does not show itself in continuous rain; but the period is made up of rainy days and days that are wholly or partially fair. Nor is