THE KATMAI REGION, ALASKA 603 



obstructions, the bottom layer of ejecta should be made up largely 

 of material from such a source. As a matter of fact this bottom 

 layer is essentially pumiceous and actually appears to be more 

 nearly free of foreign material and more n.early of rhyolitic composi- 

 tion than any layer above it. The view that the close of a period of 

 activity at a volcanic vent is attended by the formation of a plug 

 of lava which seals up the conduit and that the renewal of activity 

 necessitates the clearing out of such a plug, finds little to support 

 it here. Nor, I think, does a consideration of events in certain 

 other explosive eruptions lead to views different from those 

 expressed for Katmai. Many instances might be cited in which 

 for months previous to a paroxysmal eruption manifestations have 

 occurred, such as outbursts of gas and ashes, that can hardly be 

 looked upon otherwise than as indicating a quite direct connection 

 between the surface and the subterranean activity. It seems not 

 unusual, too, for lava to appear in the crater and remain com- 

 paratively quiet for a certain period before explosive inflation 

 occurs. The great eruption at Krakatoa in 1883 seems to have 

 followed such a course.^ At Pelee also there were premonitory 

 symptoms, consisting of an increased evolution of vapors, at times 

 mixed with cinders; later the moderately explosive (though 

 immensely destructive) ejection of the nuees ardentes, accompany- 

 ing the rise of lava in the crater.^ A somewhat similar course of 

 events may be found in Koto's description of the eruption of 

 Sakura-jima.^ By what means a volcanic vent can remain suffi- 

 ciently open to permit a free escape of vapors, without allowing 

 magma to issue, and what conditions finally bring this period to a 

 close and cause a body of gas-charged, actively corrosive magma 

 to appear are matters whose explanation presents many difficulties. 

 No theory of volcanism that I have seen appears at all adequate 

 to account for the phenomena. Indeed, some of the fundamental 

 concepts of current theories seem irreconcilable with them. 



Objection may be raised to the somewhat novel idea that has 

 been presented here, of a state of unstable equilibrium of the 



^ J- W. Judd, "The Eruption of Krakatoa and Subsequent Phenomena," Report 

 of the Krakatoa Committee of the Royal Society, pp. 11-20. 



= A. Lacroix, La Montague Pelee et ses Eruptions, pp. 35-39. 

 3B. Koto, The Great Eruption of Sakura-jima in 1914, pp. 56-82. 



