274 H. S. WASHINGTON PERSISTENCE OF VENTS AT STROMBOLI 



in lavas is derived from the wall rocks; for^ apart from other considera- 

 tions^ the discussion of which would lead us too far afield, were the lava 

 non-aqueous and the walls water-bearing (both being, as in this case, of 

 essentially the same chemical composition), it is to be expected that the 

 walls would have a lower fusion temperature than that of the lava, so that 

 the size of the vent would be continuously enlarging. 



■i. Practically all observers who have reported on the subject have noted 

 a complete absence of synchronism in the activity of the several vents. 

 Bergeat and Wegner have given very considerably detailed data, and their 

 observations were fully confirmed by those of Kozu and myself. This 

 points clearly to the independence of the several vents for a very great 

 distance down — that is to say, it may be assumed that the conduits do not 

 branch or diverge from a central main conduit, but that they are inde- 

 pendently continuous down to the magma reservoir. 



Another example of such independence of action is furnished by the 

 cascades of lava which burst up through the talus surrounding Halemau- 

 mau and at considerable heights (for example, 40 feet) above its level. ^' 

 Here we have not only independence of action, but a very distinct differ- 

 ence in the prevailing internal pressure. These sustained pressure differ- 

 ences in neighboring conduits offer unexceptionable proof of their com- 

 plete independence one of another. 



It will be noted that in this conclusion I differ with Perret,*^ who holds 

 the view that the several vents are the openings of interconnected con- 

 duits, which diverge or ramify from a central one. Against this view 

 the lack of synchronism and the different type of activity at each seem to 

 me to be conclusive arguments. 



5. Judging from the size of the vents, the diameters of the Stromboli 

 conduits must be relatively small — that is, measurable in a few tens of 

 meters. The diameters of the different vents vary considerably. Those 

 of the Torreone and Zolfo have Ijeen estimated at different times and by 

 different observers at from 20 to 100 meters, Ijeing apparently generally 

 from 30 to 50. The Central and Famarole vents would seem to be con- 

 stantly and consistently smallei'. While the diameter of the main Scarp 

 vent, number 1 on Ricco's plans, is apparently much greater, it must be 

 remembered that this represents not the size of the vent itself, but that 

 of the funnel-like crater formed by its violent explosive activity, in the 

 loose material of this part of the terrace. Presumably the mouths of the 

 vents are somewhat larger than their conduits, because of the mechanical 

 tearing away of the edges and walls by the explosions, though this is com- 

 pensated for in part by the adhesion to the walls of the blobs thrown up 



*^ Cf . Day and Shepherd: Bull. Geol. Soc. Am., vol. 24, 1913, p. 601, and plate 26. 

 *8 F. A. rcrrct : Am. Jour. Sci., vol. xlii, 1916, p. 447. 



