250 II. S. AVASHINGTON PERSISTENCE OF VENTS AT STROxMBOLI 



sistence in location, small size and verticality of conduit, contiguit}^ with 

 independence of action, and others to be mentioned later. The only 

 hypothesis that seems to be competent to explain these characters, which 

 preclude any deep-seated or extensive explosive activity in the formation 

 of the conduits, is Daly's ^^gas-fluxing" hypothesis. 



As this persistence in location is one of the least well known or recog- 

 nized characters of volcanoes, we shall first take up the consideration of 

 this feature of the Stromboli vents, discussing later the application of 

 this and other features of the volcanic vents to the explanation of the 

 mechanism of volcanic action. 



Bergeat^ pointed out this persistence as regards one of the vents and 

 thinks it probable for another, but does not discuss the cause or bearing 

 of the phenomenon. 



Perret also, in a recently published paper,^ calls attention to the per- 

 sistence in location of several of the vents in spite of profound changes 

 in the configuration of the crater terrace floor, due to violent eruptions, 

 such as those of 1907 and 1912. He, however, thinks that the feeding 

 conduits are divergent from a central one at a comparatively small depth. 



Search through the somewhat extensive literature dealing with this 

 volcano (with the aid of Bergeat's very complete bibliography as far as 

 1899) has revealed much seemingly conclusive evidence of this per- 

 sistence ; and the establishment of the fact is so important, because of the 

 bearing of the phenomenon on certain volcanological problems, that it 

 seems worth while to present quite fully the documentary data. 



The evidence is partly that of verbal descriptions and partly that of 

 plans, sketches, and views of the crater terrace. The first line of evidence 

 has been summarized by Bergeat and will be only incidentally touched 

 on here. The second line of evidence, only alluded to by Bergeat, is in 

 many respects much more striking and conclusive. As many of the 

 sketches and plans are published in works which are not readily accessible, 

 and as the inspection of a consecutive series of views and plans at dif- 

 ferent dates adds to the cogeucy of the argument, a quite complete set of 

 reproductions (except of recent photographs) will be given here.* 



The crater terrace of Stromboli is so well known and accounts of it 

 are so numerous that a detailed description is hardly called for here. It 

 oifers exceptionally favorable conditions for observations bearing on the 

 possible persistence of volcanic vents. In the first place, it has been the 



2 A. Bergeat: Die iiolischen Inseln. MiinchcD, 1899, p. 34. 



3 F. A. Perret : Am. .Tour. Sci., vol. xlii. 1916, p. 447. 



* My thanks are due Dr. L. H. Adams for his kindness in making- some of the neces- 

 ry photofii-aiihs. 



