[Illustration: FIG. 2.--FLAKES OF VOLCANIC ASH. Magnified about 100 diameters. A, flake with a branching rib; B, fragment of a broken hollow sphere of glass; C, fragment with drawn-out tubular vesicles; D and E, plain fragments of broken pumice bubbles. (From American Geologist, April, 1893.)] The explosive eruptions which give rise to showers of this kind of ash, or dust, are due to fusion and superheating of subterranean masses of rocks charged with more or less moisture. A part of this moisture escapes in the form of steam at the time of an eruption. But the viscidity of the ejected material prevents much of the steam from passing off, and such of the lava as cools most rapidly retains a certain quantity in solution, as it were. Obsidian is a rock which has been made in this way. It often contains much of the original water, which will cause it to swell up into a stony froth when fused. This volcanic dust has the same property. If one small particle of it be heated on a piece of platinum foil it is seen to swell up into a compound bubble of glass (Fig. 3). It is evident that this is due to the expansive force of the heated included moisture, to which the viscid half-molten glass readily yields. At the time of the eruption which produced this dust, subterranean heat was applied to the moisture-bearing rock until this was superheated to such an extent that the weight of the overlying material was insufficient to hold the water from expanding into steam. Then there was a tremendous explosion, and the molten magma was thrown up with such a force that it was shattered into minute droplets, in the same way as water does when it is thrown forcibly into the air. Being thus released from pressure, the steam inside of each little particle of the heated glass caused it to swell out into a tiny bubble. As this kept on expanding it was cooled, the thin glass wall of the bubble congealed, and finally burst from the pressure of the steam within. This is the reason why the little dust particles are thin, mostly triangular, and often slightly concave flakes with sharp angles. Sometimes the angles appear rounded, as if the fragments had been viscid enough to creep a little after the bubble burst. The study of one single little grain of dust, barely visible to the naked eye, thus makes clear the nature of a catastrophe which must have shaken a whole mountain, and which left its traces over a quarter of a continent. [Illustration: FIG. 3.--A PARTICLE OF VOLCANIC ASH SWELLED UP BY FUSION. Magnified 100 diameters.]