The exact sequence of events which leads up to an eruption is not known, but a definite conception of the principles involved may perhaps be secured by a definite case. Suppose a geyser-tube filled with water, and heated at its lower end. As the water is heated below, convection tends to distribute the heat throughout the column of water above. If convection were free, and the tube short, the result would be a boiling spring; but if the tube is long, and especially if convection is impeded, the water at some level below the surface may be brought to the boiling-point earlier than that at the top. Under these circumstances if even a little water in the lower part of the tube is converted into steam, the steam will raise the column of water above, and it will overflow. The overflow relieves the pressure on all parts of the column of water below the surface. If before the overflow there was any considerable volume of water essentially ready to boil, the relief of pressure following the overflow might allow it to be converted into steam suddenly, and the sudden conversion of any considerable quantity of water into steam would cause the eruption of all the water above it (Figs. 213 and 214). The height to which the water would be thrown would depend upon the amount of steam, the size and straightness of the tube, etc. It is clear that everything which impedes convection in the geyser tube will hasten the period of eruption, since impeded circulation will have the effect of holding the heat down, and so of bringing the water at some level below the top more quickly to the boiling-point. It follows that anything which chokes up the tube, or which increases the viscosity of the water, or its surface tension, would hasten an eruption.[119] Geysers often build up crater-like basins or cones (Figs. 214 to 217) about themselves, the cone being of material deposited from solution. In the Yellowstone Park the precipitation of the matter in solution (chiefly silica) is partly due to cooling and partly to the algæ which abound even in the boiling water, and the brilliant colors of the deposits about the springs are attributable to these plants. When the water from any geyser or hot spring ceases to flow the plants die and the colors disappear. The details of the surface of the deposits about geysers and hot springs are often complicated, and frequently very beautiful (Fig. 218). [Illustration: +Fig.+ 214.—“Giant” Geyser, Yellowstone Park, in eruption. Shows also the cone. (Wineman.)] [Illustration: +Fig.+ 215.—Cone (or crater) of Castle Geyser, Yellowstone Park. (Detroit Photo. Co.)] [Illustration: +Fig.+ 216.—Cone (or crater) of Grotto Geyser, Yellowstone Park. (Detroit Photo. Co.)] [Illustration: +Fig.+ 217.—Cone of Giant Geyser, Yellowstone Park. (Wineman.)]