"Most people would have searched above its original point of egress. As for me, after a month of study, observation, and reasoning, I sought for and found it fifty meters lower down. And this was the explanation of the matter: I told you before that it was first necessary to determine the origin, nature, and direction of the fissures in the granite which enabled the water to spring forth. It was easy for me to satisfy myself that these fissures ran from the plain toward the mountain and not from the mountain toward the plain, inclined like a roof undoubtedly, in consequence of a depression of this plain which in breaking up had carried along with it the primitive buttresses of the mountains. Accordingly, the water, in place of descending, rose up again between the different interstices of the granitic layers. And I then discovered the cause of this unexpected phenomenon.
"Formerly the Limagne, that vast expanse of sandy and argillaceous soil, of which you can scarcely see the limits, was on a level with the first table-land of the mountains; but owing to the geological character of its lower portions, it subsided, so as to tear away the edge of the mountain, as I explained to you a moment ago. Now this immense sinking produced, at the point of separating the earth and the granite, an immense barrier of clay of great depth and impenetrable by liquids. And this is what happens: The mineral water comes from the beds of old volcanoes. That which comes from the greatest distance gets cooled on its way, and rises up perfectly cold like ordinary springs; that which comes from the volcanic beds that are nearer gushes up still warm, at varying degrees of heat, according to the distance of the subterranean fire.
"Here is the course it pursues. It is expelled from some unknown depths, up to the moment when it meets the clay barrier of the Limagne. Not being able to pass through it, and pushed on by enormous pressure, it seeks a vent. Finding then the inclined gaps of granite, it gets in there, and reascends to the point at which they reach the level of the soil. Then, resuming its original direction, it again proceeds to flow toward the plain along the ordinary bed of the streams. I may add that we do not see the hundredth part of the mineral waters of these glens. We can only discover those whose point of egress is open. As for the others, arriving as they do at the side of the fissures in the granite under a thick layer of vegetable and cultivated soil, they are lost in the earth, which absorbs them.