condition should exist in all geysers; neither is it at all necessary 
Fig. 10.—Arti-
ficial Geyser. in order to explain the phenomenon of an eruption. To prove beyond question the truth of his theory, Bunsen constructed an artificial geyser. The apparatus (Fig. 10) consisted of a tube of tinned sheet-iron about ten feet long, expanded into a dish above for catching the erupted water. It may or may not be expanded below for the convenience of heating. It was heated, also, a little below the middle, by an encircling charcoal chauffer, to represent the point of nearest approach to the boiling-point in the geyser-tube. When this apparatus was heated at the two points, as shown in the figure, the phenomena of geyser-eruption were completely reproduced; first, the violent explosive simmering, then the overflow, then the eruption, and then the state of quiescence. In Bunsen's experiment, the eruptions occurred about every thirty minutes.
According to Bunsen, a geyser does not find a cave, or even a perpendicular tube, ready made, but, like volcanoes, makes its own tube. Fig. 11 is an ideal section of a geyser-mound, showing the manner in which, according to this view, it is formed. The irregular line, b a c, is the original surface, and a the position of a hot spring. If the spring be not alkaline, it will remain an ordinary hot spring; but, if it be alkaline, it will hold silica in solution, and the silica will be deposited about the spring. Thus the mound and tube are gradually built up. For a long time the spring will not be eruptive, for the circulation will maintain a nearly equal temperature in every part of the tube—it 
Fig. 11.—Ideal Section of a Geyser-Tube
(according to Bunsen). may be a boiling, but not an eruptive spring. But, as the tube becomes longer, and the circulation more and more impeded, the difference of temperature between the upper and lower parts of the tube becomes greater and greater, until, finally, the boiling-point is reached below, while the water above is comparatively cool. Then the eruption commences. Finally, from the gradual failure of the subterranean heat, or from the increasing length of the tube repressing the formation of steam, the eruptions gradually cease. Bunsen found geysers in every stage of development—some playful springs without tubes; some with short tubes, not yet eruptive; some with long tubes, violently eruptive;
