Mallet asserts that in many cases the clouds of smoke observed do not consist of true smoke like that produced when wood or vegetable matters are incompletely burned, but is only ordinary air mixed with sulphurous acid gas, and various other gases. But not only fire and smoke are seen at times coming out of fissures in the earth. A thing still more frequently thrown out is water, which often spouts forth along with great quantities of mud, sand, and the finely ground fragments of earthy materials generally. Among many other instances where the emission of water from the crevices was particularly noticeable, may be mentioned the earthquakes at Jamaica in 1687 and 1692. Here the water, in some places, was thrown out of the ground to considerable heights in the air. Mallet calls attention to the fact that the waters of springs collect in reservoirs consisting either of fissures or crevices of the rocks, of small width but great depth, which are vertical or inclined to the horizon, or in reservoirs that are formed of extended beds of sand or gravel. Now, when the earthquake waves moving horizontally over the surface produce movements that squeeze these fissures together, the water in the fissures is spurted out in high jets, and carries with it the finely divided rock or sand formed by the rubbing together of the rock surfaces. In the case of the reservoirs consisting of beds of sand or gravel, lying between impervious layers, if, during an earthquake motion, the land areas are suddenly lowered, the water rushing into the cavity thus left will afterwards be shot out with considerable force, when the land is suddenly raised again. Where there are no direct openings in the ground the water will burst through the crust in the shape of great vertical jets, thus forming a circular hole, broken or fractured at its edges. Water jets of this character were especially numerous during the earthquake of Calabria in 1783. In a swampy plain, known as Rosarno, many of these circular wells or openings about the size of an ordinary carriage wheel, though in some cases much larger, were to be seen crowded together. The appearance of the openings are represented in Fig. 40. Some of these were filled with water, but the greater number were dry and filled with loose sand. These latter, when examined by digging, were shown to be funnel-shaped, as seen in Fig. 41. As seen, the margins of the wells exhibit a series of cracks or crevices extending radially outward from the centre. Their origin is evident. As the water was violently expelled by the squeezing motion of the upper and lower impervious strata, it shot upwards, thus producing the funnel-shaped tube. At the same time the force of the eruption was sufficiently great to produce the radial fissures or fractures at the sides. [Illustration: FIG. 40. CIRCULAR HOLLOW FORMED BY CALABRIAN EARTHQUAKE] [Illustration: FIG. 41. SECTION OF CIRCULAR HOLLOW FORMED BY CALABRIAN EARTHQUAKE]