... ut in mare de terris venit umor aquai, in terras itidem manare ex aequore salso; percolatur enim virus, retroque remanat materies umoris et ad caput amnibus omnis confluit, inde super terras redit agmine dulci. --_De Rer. Nat._ vi. 633. The origin of underground water is regarded by Seneca as due to three causes. The earth itself contains moisture which it forces out at the surface; it includes also air which in the darkness of the subterranean wintry cold is condensed into moisture; by the principle of interchangeability, whereby one element passes into another, the earth in its interior resolves itself into moisture. If it be urged that the rivers are too vast to draw their supplies from these sources, the ready answer comes that the internal reservoir is quite spacious enough for the purpose, and that it might as well be matter of surprise that, with all the winds that constantly blow, the supply of air does not fail, or that a single wave of the sea should be left to follow so many breakers. If the questioner, still unsatisfied, should demand to know how water is produced, he is met with the query how air is produced on earth. There are in nature four elements, and he is not entitled to ask where one of them comes from. Each is a fourth part of nature, and it is obvious that what has an element as its source cannot fail. Hence the philosopher in pronouncing water to be an element has given it enough, and more than enough, of strength. In short, rain may give rise to a torrent, but not a river flowing steadily between its banks. Heavy rains will swell such a river, but cannot produce it. Having, as he believed, cleared the ground in this way, Seneca proceeds to consider the distribution of water within the earth. He opines that as in our body, so in the earth, there are channels by which both air and liquids flow. He states his conviction that the earth contains not only veins of water, but also large streams, and in a later part of the volume he speaks of both underground rivers, huge lakes, and a hidden sea from which rivers at the surface are supplied (154, 233, 235). He is aware that some of these subterranean reservoirs contain fish, about which he has some incredible tales to tell. He makes mention of rivers that sink underground and reappear, as if a matter for great astonishment. But examples of it may be found in many limestone districts, where the solution of the rock by underground water has given rise to tunnels, passages, and caverns into which, when their roofs give way, surface streams may be engulfed, to break out again from other openings at lower levels (141). The author concludes this part of his argument by asking if anybody is ignorant that there are some standing waters which have no bottom, whence, he contends, it is shown that this water is the perpetual source of large rivers.