The main task of studying the region in detail, Dutton left to his associate on the Geological Survey, J. S. Diller. Diller had visited Crater Lake as early as 1883, in company with Everett Hayden, and had spent several days exploring parts of the rim and Wizard Island. Between 1883 and 1886, he had been engaged in a study of the Lassen Peak district and had made reconnaissance trips to many other volcanoes in the southern Cascades. He was therefore especially well qualified to undertake the survey of Crater Lake. His first published accounts appeared in 1897, but his final report, supplemented by petrographic descriptions of the rocks by H. B. Parton, did not appear until 1902.3 It was a splendid piece of work, all the more to be praised when we recall how he was hampered by lack of time, by difficulties of access, and by the scarcity of suitable places to camp.
It would occupy too much space here to review Diller’s findings in full; many of them are referred to below. His main conclusion concerning the origin of Crater Lake should, however, be emphasized. Like Dutton, he could not subscribe to the view that explosion by itself was the cause. He realized that in order to feed the powerful glaciers which once spread down the sides of the old volcano, the summit must have been far higher than the present rim of the caldera, and he supposed that it may even have exceeded 14,000 feet in height. Approximately 17 cubic miles of material must therefore have disappeared to form the caldera. But if this volume of rock had simply been removed by explosion, a tremendous layer of debris should be found close to the rim. No such layer was to be seen. Accordingly he concluded that the top of the old volcano must have been engulfed. He supposed that engulfment was brought about, as in the case of the calderas of Hawaii, by withdrawal of support from beneath the top of the volcano following copious eruptions of lava far down the flanks, and though he searched in vain for the products of such eruptions, he did not despair that some day they would be found.
More than thirty years elapsed before the next important contribution was made to an understanding of the geology of Crater Lake. In 1934, B. N. Moore4 published an account of the great pumice sheet that encircles the lake. He was the first to recognize the dual character of the pumice deposits. One type is heavily charged with large lumps, is typically unsorted, and contains many charred logs. This type he realized to be the product of glowing avalanches or nuées ardentes like those erupted from Mount Pelée during the devastating activity of 1902. The other type of pumice is spread over hill and valley as a continuous mantle. This deposit is far more extensive, covering an area of more than 5000 square miles. It is granular and well sorted, and with increasing distance from the lake it becomes thinner. Clearly this pumice was ejected high above the volcano and was drifted northeastward by the prevailing winds, to fall in showers along the path.
Another notable advance in the interpretation of Crater Lake was made by W. W. Atwood, Jr.5 who in 1935 published the results of glacial studies commenced four years earlier. Atwood was the first to realize that during the slow growth of the volcano glaciers repeatedly advanced and retreated on its flanks. He revealed the story of a long struggle between the erosional and constructional processes that shaped the ancestral Mount Mazama.
In 1936, J. E. Allen6 examined the internal structure of the dacite flows of Llao Rock, Grouse Hill, and Cleetwood Cove, and determined the manner of extrusion. At the same time he proved that the Cleetwood lava, which extends from the caldera rim to the water’s edge, is not really a “backflow,” as Diller had supposed, but is the filling of an inclined feeder laid bare by the collapse of Mount Mazama.
Finally, in the same year, W. D. Smith and C. R. Swartzlow7 cast doubt on the idea that Crater Lake was formed by engulfment. On the basis of evidence offered by new exposures, they maintained that Diller had been misled concerning the volume of coarse ejecta around the rim. In their opinion, there seemed ample fragmental material to account for the formation of the caldera by devastating explosions. They thought that much of the shattered top of the volcano was blown far beyond the confines of the park, that much lies concealed beneath the ubiquitous mantle of pumice, and that much has been removed by erosion; and it appeared to them that the shape of the caldera itself was indicative of explosion. The challenge of their conclusions fostered new interest in the problem of Crater Lake.
