At one of the deeper ends of the cave, water dropped continually from the crevices of the roof; a fact which Professor Smyth attributed to the slow advance of the summer wave of heat through the superincumbent rock, which was only now reaching the inner recesses of the loose lava, and liquefying the results of the past winter.

There would seem to be immense infiltration of meteoric water on the Peak; for, notwithstanding the great depth of rain which falls annually in a liquid or congealed form, the sides of the mountain are not scored with the lines of water-torrents.
Though occurring in lava, this cavern is quite different from lava-tunnels, such as the Surtshellir, which are recognised formations, produced by the cooling of the terminal surface-crust of the stream of lava, and the subsequent bursting forth of the molten stream within.
This, on the contrary, proved to be a smooth dome-shaped cave, running off into three contracting lobes or tunnels which might be respectively 70, 50, and 40 feet long, and were all filled to a certain depth with water: in the smoothness of the interior surfaces, Professor Smyth believed that he detected the action of highly elastic gases on a plastic material.
The astronomer takes exception to the term 'underground glacier'[120] which had been applied to this cavern.

He represents that the mountain is abundantly covered each winter with snow, in the neighbourhood of the ice-cave, which is nearly within the snow-line, and the stores of snow thus accumulated in the cave have no great difficulty in resisting the effects of summer heat, since all radiation is cut off by the roof of rocks.

The importance of this protection may be understood from the fact that in the middle of July the thermometer at this altitude gave 130 deg.

in the sun, but fell to 47 deg.