10 



SCIENCE. 



[N. S. Vol. III. No. 53. 



tile sandstone formation would be sufR- 

 cient, when the overljdng strata gave way, 

 to hurl their fragments out upon the sur- 

 souuding plain. As the data were quite 

 indefinite the computation could result only 

 in a rough approximation, and there is no 

 need to weary you with its details, but it 

 served to show that the assumed cause was 

 of the same order of magnitude as the re- 

 sult accomplished. The idea of applying 

 such a test needs no specific explanation, 

 because quantitative tests of this particular 

 type are among the most familiar resources 

 of investigation. Whenever a tentative 

 theory involves the application of force or 

 the expenditure of energy the investigator 

 (or his critic) habitually asks whether the 

 assumed cause affords a sufficient amount 

 of force or of energy. 



Practically the same conclusion was 

 reached in a more satisfactory way by 

 studying the accounts of other natural ex- 

 plosions where steam was the agent. At 

 several epochs in its history the top of 

 Mount Vesuvius has been torn away by a 

 sudden convulsion. In Java the summit 

 of Mount Tomboro was blown away, with 

 the production of a great crater which now 

 contains a lake, and a similar catastrope 

 occurred on the slope of Mount Pepandajau. 

 The great explosion of Krakatoa, in 1883, 

 demolished several volcanic islands and 

 created others, reconstructing the topog- 

 raphy of a district in the Straits of Sunda. 

 On Julj' 15, 1888, a great opening was torn 

 in the Japanese mountain Kobandai, the 

 summit and part of one side being removed. 

 The last mentioned instance is the most 

 available for comparison because the agency 

 of steam distinctly appeared, and because 

 the history of the event has been admirablj^ 

 reported by two Japanese geologists. Profs. 

 Sekiya and Kikuchi, of Tokio. * 



* The Eruption of Bandai-san. Trans. Seismologi- 

 cal Soc. of Japan, Vol. XIII., (1890), pp. 139-222. 

 9 plates. 



There were in this case about twenty ex- 

 plosions, all occurring within the space of 

 one or two minutes. A cloud of rock frag- 

 ments ascended to a height of 4,000 feet. 

 The greater number, moving obliquely away 

 from the mountain side, fell upon its lower 

 slope, down which they rolled as an ava- 

 lanche for a distance of five miles, over- 

 whelming several villages and transforming 

 a fertile plain into a rocky desert. In other 

 directions fell showers of stoues, and a cloud 

 of dust descending more slowly. The re- 

 sulting crater, less regular in form than the 

 subject of our study, was nineteen times as 

 capacious, and from its bottom fierce jets of 

 steam issued for weeks and even months. 

 Kobandai is a volcanic peak, and although 

 it had been quiescent for ten centuries 

 there can be little doubt that the steam it 

 evolved was generated by volcanic heat. 



The competency of volcanic steam for the 

 production of a crater is thus shown by a 

 parallel instance, and the onlj^ conspicuous 

 difference between the Japanese case and 

 the' Arizonian lies in the fact that in the 

 one the disrupted rock was volcanic and in 

 the other it was not. This diiference seems 

 unessential, for in neither case was there 

 au eruption of liquid rock; the ancient 

 lavas of Kobandai had been cold for ages, 

 and their relation to the catastrophe was 

 wholly passive. Moreover, the manifesta- 

 tion of volcanic energy is no more excep- 

 tional on the Arizona plateau than in the 

 Bandai district. The little limestone crater 

 is in the midst of a great volcanic district. 

 (Fig. 1, Page 3, and Plate 1, Fig. 5.) 

 The nearest volcanic crater is but ten miles 

 distant, and within a radius of fifty miles 

 are hundreds of vents from which lava 

 has issued during the later geologic periods. 



In following this line of thought I have 

 but reversed the logical route by which Mr. 

 Johnson probably reached his theory, veri- 

 fjdng the theorj^ by recomparison with its 

 source. 



