Some suppose that all the water that the earth drinks in from rain is sent out again into the rivers. They set down by way of proof the fact that there are fewest rivers in the localities where there is least frequent rain. On that account, they say, the deserts of Ethiopia are destitute of streams, and few springs are found in the interior of Africa, because there is always a blazing sky and almost perpetual summer. Therefore there are ugly stretches of sandy waste, without tree and without inhabitant, sprinkled at rare intervals by showers that they immediately swallow up. On the other hand, it is well known that there are abundant streams and rivers in Germany and Gaul and next to them in Italy, because they enjoy a moist climate, and even the summer is not without rainfall. VII A great deal can obviously be urged in reply to this. First of all, 1 as a diligent digger among my vines, I can affirm from observation that no rain is ever so heavy as to wet the ground to a depth of more than 10 feet. All the moisture is absorbed in the upper layer of earth without getting down to the lower ones. How, then, can rain, which merely damps the surface, store up a supply sufficient for rivers? The greater part of it is carried off at once into the sea by river-channels. But a small portion is absorbed by the ground, and even that is not retained. For the ground is either dry and so uses up at 2 once the water poured into it; or else it is saturated and throws off what of the rainfall it does not require. This is the reason why rivers do not rise with the first rainfall; the thirsty ground absorbs it all. And then, again, how are we to explain the fact that some rivers burst out from rocks and mountains? What contribution can be made to them by rains that are carried down over the bare crags and have no earth into which to sink? Besides, wells sunk in the very driest localities 3 to a depth of 200 or 300 feet reveal rich springs of water at a depth to which rain water does not penetrate. One may be sure there is no rain water there nor any gathering of moisture, but living (= spring) water as it is usually called. The opinion in question is disproved by this other argument, too; some springs well up in the very summit of a mountain. It is plain, therefore, that the water in them is forced up or forms on the spot, since all the rain water runs off. VIII