"_No well below the sea level becomes dry of itself_," even in seasons of extreme and continued drought. But the contents do not vary with the tides, the rise of which is so trifling that the distance from the ocean, and the slowness of filtration, renders its fluctuations imperceptible. On the other hand, the well of Potoor, the phenomena of which indicate its direct connection with the sea, by means of a fissure or a channel beneath the arch of magnesian limestone, rises and falls a few inches in the course of every twelve hours. Another well at Navokeiry, a short distance from it, does the same, whilst the well at Tillipalli is entirely unaffected as to its level by any rains, and exhibits no alteration of its depths on either monsoon. ADMIRAL FITZROY, in his _Narrative of the Surveying Voyages of the Adventure and Beagle_, the expedition to which Mr. Darwin was attached, adverts to the phenomenon in connection with the fresh water found in the Coral Islands, and the rise and fall of the wells, and the flow and ebb of the tide. He advances the theory propounded by Darwin of the retention of the river-water, which he says, "does not mix with the salt water which surrounds it except at the edges of the land. The flowing tide pushes on every side, the mixed soil being very porous, and causes the water to rise: when the tide falls, the fresh water sinks also. _A sponge full of fresh water placed gently in a basin of salt water, will not part with its contents for a length of time if left untouched_, and the water in the middle of the sponge will be found untainted by salt for many days: perhaps much longer if tried."--Vol. i. p. 365. In a perfectly motionless medium the experiment of the sponge may no doubt be successful to the extent mentioned by Admiral Fitzroy; and so the rain-water imbibed by a coral rock might for a length of time remain fresh where it came into no contact with the salt. But the disturbance caused by the tides, and the partial intermixture admitted by Admiral Fitzroy, must by reiterated occurrence tend in time to taint the fresh water which is affected by the movement: and this is demonstrable even by the test of the sponge; for I find that on charging one with coloured fluid, and immersing it in a vessel containing water perfectly pure, no intermixture takes place so long as the pure water is undisturbed; but on causing an artificial tide, by gradually withdrawing and as gradually replacing a portion of the surrounding contents of the basin, the tinted water in the sponge becomes displaced and disturbed, and in the course of a few ebbs and flows its escape is made manifest by the quantity of colour which it imparts to the surrounding fluid.]