During the thirty months' interval between my two visits, the gradual elevation of the fire-lake continued quite uni formly, as attested by occasional photographs. By its fre quent overflows it had built itself up to a height of fully fifty feet above the previous main floor of Kilauea, so that it formed an extremely low truncated cone, surmounted by the level lake, to the edge of which visitors daily approached.
About March, 1894, a recession began, which ended in a final collapse of activity. The lake soon sank some hundreds. of feet, carrying with it the sides of a circular pit, about 1,400 feet in diameter, and central to the original 2,40o-foot pit. When I saw it in the following September, the fire-lake was not less than five hundred feet below the rim. During the evening, masses of rock frequently crashed in, driving heavy surges of fire far up the talus. There was a good deal of steam-cloud slowly rising, charged with sulphur. During my previous visit, all vapour had seemed to be absent, and I made the circuit of the pit without encountering sulphur. Sub sequent photographs had also indicated the absence of vapor from the lake.
I now have to add an important observation. To my great surprise, at this last visit, I perceived that the three fountains above described were in full activity and in the same relative position as before, although during the thirty months the level of the lake had risen three hundred and fifty feet and had then fallen five hundred feet. By what system of supply-ducts such
fountains had been so long maintained was a mystery con cealed in the fire-depths. But the fact of a marvelous steadi ness and uniformity of action was obvious. For a long period a uniform and gentle outpour of effervescence had been main tained. It has persisted for two years and a half, throughout all the immense changes.
I submit as the unavoidable conclusion that the source of supply for this five years' outpour of gently effervescing lava was in an interior magma which itself contained the impelling force in its own originally occluded gases. For its activity this source was wholly independent of any encounter with water to generate steam. Expanding steam evidently had no part in that steady, quiet, persistent activity in the fire-lake of Kilauea.
I would add that the exceptionally quiet and uniform activ ity of Kilauea seems to render it one of the most important of all volcanoes for study. I regret to say that since the col lapse nearly eight years ago no lava has appeared in the crater, except a small quantity last June, which has again gone out of sight.