232 THE ERUPTION OF VESUVIUS IN APRIL, 1906. 



northeast. This cut has a bottom several meters wide, which is very 

 much shattered and which will probably fall into the crater. The 

 minimum lowering of the mountain has been determined to be 103 

 meters, but tlie dimensions frequently change, the crater enlarging 

 at the expense of its edges. The vertical walls of the crater present 

 admirable sections showing the alternation of beds of fragmental 

 material, ash, and lava, traversed by vertical or oblique dikes that 

 characterize the internal anatomy of the cone. 



The crater of 1906 is remarkably like that of 1822, but its dimen- 

 sions are less. That also had a notch on one side which, like that of 

 the present crater, was located in line with fissures. 



The wide and deep crater thus formed is a true caldera, and the 

 phenomena of its production have a strong bearing on the mode of 

 formation of this kind of volcanic mechanism. It is the more in- 

 structive in that the eruption is not an exceptional one. Accounts 

 of the eruptions of 1631, 1761, 1779, 1839, 1850, and 1872 show that 

 the same phenomena have been repeated with greater or less in- 

 tensity. They constitute the characteristic traits of a special type of 

 the great Vesuvian eruptions. 



The caldera of Vesuvius is comparable to that of the Soufriere of 

 St. Vincent and to the crater of Mount Pelee previous to the erup- 

 tion of 1902. The V-shaped notch, which played so fatal a role in 

 the destruction of St. Pierre, is the equivalent of that opened in the 

 eruption of Vesuvius. It is noteworthy to see this feature, caused by 

 the existence of fissures traversing the cone, repeat itself in diifer- 

 ent volcanoes. 



Dry avalanches. — With the exception of the material that fell 

 upon Ottajano and its surroundings, the products of eruption other 

 than the fine dust have been thrown but a short distance. The sides 

 of the crater have been covered several meters thick with a bed of 

 blocks of all sizes up to several meters in diameter, minute fragments, 

 and fine dust. The dry avalanches have been produced either by 

 direct ejection of material from the crater or by the loosening of 

 material already accumulated on the slopes, and the latter method 

 has apparently been predominant. In the most active phase ava- 

 lanches could be seen to detach themselves from the summit, but it 

 was impossible to observe their point of departure. 



The avalanches roll along the surface, followed by a train of light 

 dust, and are easily distinguishable from the dense, clear-cut Peleean 

 clouds, which expand vertically in the course of their downward 

 progress. The mechanism of the avalanches is easily understood. 

 The profile of the cone is irregular, the slope increasing at a short 

 distance from the sunnnit. It is at this level that the loosening of 

 material takes place, either by a disturbance of equilibrium or by 

 earth movements caused by the explosions or by the shock of solid 



