In addition to brief discussions of the older theories of geyser action, a more extensive résumé of the modern but lesser known theory of Thorkelsson is presented.
Eruption patterns (i.e., the observed changes with respect to time during several periods) of Crater Hills Geyser, Sawmill Geyser, and Tardy Geyser, are cited as indicative of the presence, in these geysers, of the spring gases so important to Thorkelsson's theory. On the basis of the eruption patterns, subterranean connections are concluded to exist between Sawmill and Tardy geysers and between Grand and Turban geysers. An interesting Bernoulli effect, previously undescribed in the literature, was concluded to cause Grand's surface level to lower during the early portion of Turban's eruptions. Continuous time-temperature curves for several cycles of Jewel Geyser indicated, for the depths observed, that temperatures were often higher during quiescent periods than at the beginning of the eruption. Spasm Geyser possibly illustrates an influx of relatively cool water which terminates the eruption. Grotto and Castle geysers indicate that a cause of termination of the eruption, other than a rapid influx of cooler water, probably exists.
The usual methods of securing temperature-depth data are found unsatisfactory; some new methods are suggested. Temperature-depth curves of Chinaman Spring are taken to indicate not only the importance of convection currents in the upper parts of geyser wells but also the efficacy of constrictions and narrow tubes in inhibiting convectional cooling. Numerous temperature-depth curves of active and inactive geysers indicate that boiling conditions were rarely existent in the lower depths examined. Superheated waters were often located at the surface of the pools of the numerous geysers and hot springs examined.