264 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



was inhabited by white men, and geological conditions indicate that 

 at least 200 years have elapsed since the last show of activity. 



If a volcano is assumed to be a slowly cooling system maintained 

 ])y residual heat from a local hot zone below the surface, then Avhat 

 manner of mechanism is it which can account for the sudden de- 

 A elopment in such a cooling system, of violent explosive activity suf- 

 ficient in volume to extend over several years, and of such intensity 

 as to cover the surrounding country with boulders of considerable 

 size, without displaying evidence of high initial temperature or con- 

 tributory chemical reactions? 



It has been a long search to find an explanation which appeared 

 properly to account for an outbreak of this kind, but a suggestion 

 has at last been found, partly as a result of theoretical reasoning 

 and partly from experience in the laboratory, which appears to 

 account for it completely and at the same time to go far toward 

 elucidating the entire problem of volcanism.^ Freed from chemical 

 terminology the mechanism is simply this : A silicate solution in 

 its liquid state can take up water in solution in considerable quan- 

 tity. A simple solution of silica and potash when heated under 

 pressure in the laboratory is capable of taking up as much as 12.5 

 per cent of water in solution. A rock magma in the earth is just 

 such a silicate solution, although more complicated in character, 

 and is entirely competent to carry 5 or G per cent of water in solu- 

 tion under appropriate conditions." If it should happen that the 

 lava beneath a volcano carries such quantities of Avater in solu- 

 tion, then all the phenomena of volcanism become appreciably 

 clearer, for in the 8,000 or more analyses of crystalline rocks of 

 igneous origin which have been gathered together by Washington 

 and published by the United States Geological Survey ^^ there is 

 none containing more than I14 per cent of water and less than 1 

 per cent is usual. This must mean that in the process of crj'stal- 

 lization of the rock from the magma the wat«r content is for the 

 most part discharged. Should it happen that this discharge of 

 water takes place in a closed space then immense pressures might 

 develop and explosive activity of tremendous intensity might result. 

 Furthermore if the amount of participating magma Avere large and 

 the rate of crystallization considerable, such activity might continue 

 for long periods of time. 



Beneath the surface, magma basins are of course inaccessible to 

 direct observation, but two facts have been noted in the study of 



" G. W. Moroy. " Tho Development of Pressure In Magmas as a Result of Crj'stalliza- 

 tion." Journ. AVash. Acad. 12, lilO, 1922. Arthur L. Day and E. T. Allen, op. cit. 



"> Ct. .T. W. .Tudd ct nl. "The eruption of KraUatoa and subsequent phenomena," p. 36 

 (Rept. of Krakatoa Com., Roy. Soc. London, 1888). 



'» II. S. AA'ashiuKton, "Chemical Analyses of Igneous Rocks, etc." U. S. Geol. Surv. 

 Trof. Taper 90, 1917. 



