GENER.AL DISCUSSION 275 



from the boiling lava below, as was well seen by ns in August, 1914, at 

 the Torreone vent. 



Examples of such narrow conduits are furnished by the pit craters 

 which occur in Puna and elsewhere on the island of Hawaii.*^ One of 

 these, on the slopes of Hualalai, is described by Brigham as about 25 feet 

 in diameter, vertical, and with an estimated depth of about 1,800 feet, 

 I am informed by Doctor Day that the walls of some of these Hawaiian 

 pipes are glazed as by fusion. 



The small size— about 10 meters in diameter — of the vent which opened 

 in the floor of the crater of Vesuvius^^ on July 5, 1913, also indicates that 

 its conduit is of similar order of magnitude. 



Before discussing the origin of these vents their situation must be 

 described, as it has an important bearing on the problem. The present 

 crater terrace represents a late site of activity within an earlier and 

 larger crater, the remains of which are seen in the Serra Vancori to the 

 south and ridges to the northeast and southwest. The relations are much 

 like those of Vesuvius and Somma, but on a smaller scale. The present 

 crater differs, however, from Vesuvius and nearly all other similar intra- 

 sommal craters or cones in being asymmetrical. The northwest side 

 plunges steeply down to the sea in the so-called Sciarra. 



This peculiar configuration has been variously interpreted, as is brought 

 out by the discussion of Bergeat. With him we may dismiss the ideas 

 that the Sciarra represents the original Vancori crater (as it may be 

 called), or that it is an erosion valle}^, or Scrope's suggestion that it is 

 due to an explosion which blew off half the old Vancori cone. The ex- 

 planation of Bergeat,^^ that it is due to faulting and subsequent subsidence 

 of a 1)1 ock on tbe northwest, is the most rational one. 



Although not mentioned by Bergeat, the submarine contour lines lend 

 strong probability to this view. As seen on a hydrographic chart, or 

 even as shown on Bergeat's small scale map (plate IT), the gradient of 

 the sea-bottom is far steeper on the northwest side of the island (that is, 

 along the Sciarra) than elsewhere, and very close to the shore becomes 

 almost precipitous. It may be added that this line of sharp declivity is 

 continued along the whole northwest edge of the ^Eolian Islands as far 

 as Alicudi. 



But, whatever be the cause of the disappearance of the nortliwestern 

 part of the old Vancori cone, we may reasonably suppose that the Sciarra 

 is merely a talus heap, made up of the material ejected from the vents 



49 W. T. Bi-igham : Mem. Bishop Mus., vol. ii, 1909, pp. 11, 97. 



=0 A. Malladm : Rend. Accad. Sci. Nap., 1914, p. 10 ; Boll. Soc. Geog., 1914, pp. 45 ff. 



Washington and Day : Bull. Geol. Soc. Am., vol. 26, 1915, p. 379. 

 "1 Bergeat : Op. cit., p. 27. 



