A little consideration will show that it is a by no means unreasonable supposition that variations in atmospheric pressure may exercise a very important influence in bringing about volcanic outbursts. Changes in the barometer to the extent of two inches within a very short period are not uncommon occurrences. A very simple calculation will show that the fall of the mercury in the barometer to the extent of two inches indicates the removal of a weight of two millions of tons from each square mile of the earth's surface where this change takes place. Now, if we suppose, as we have good ground for doing, that under volcanic areas vast quantities of superheated water are only prevented from flashing into steam by the superincumbent pressure, a relief of this pressure to the extent of two millions of tons on every square mile could scarcely fail to produce very marked effects. The way in which explosions in fiery coal-mines generally follow closely upon sudden falls in the atmospheric pressure is now well known; and coal-mine explosions and volcanic outbursts have this in common, that both result from the sudden and violent liberation of subterranean gases. There are not a few apparently well-authenticated accounts of volcanic and earthquake phenomena following closely on peculiar atmospheric conditions, and the whole question of the relation of the volcanic forces to atmospheric pressure, as Spallanzani himself so long ago pointed out, is deserving of a most careful and rigorous investigation. [Sidenote: SUPPOSED TIDAL EFFECTS.] There is one other consideration which has frequently been urged as worthy of especial attention, in dealing with the question of the exciting causes of volcanic outbursts. If volcanoes were, as was at one time almost universally supposed, in direct communication with a great central mass of liquefied materials, or even if any large reservoirs of such liquids existed beneath volcanic districts, as others have imagined, then the different mobility of the solid and liquid portions of the earth's mass would give rise to tidal effects similar to those occurring in the surface waters of the globe. Under such circumstances, volcanic outbursts, like the tides, would be determined by the relative positions of the sun and moon to our globe. It is certain, however, that no very direct relation has yet been established between the lunar periods and those of volcanic outbursts, though recent close observations upon the crater of Vesuvius, by Professor Palmieri, do seem to lend support to the view that such relations may exist.