-tii-J /•'. .1. Perret Lava h'nij>tion of Siromboli. 



The products of this second explosion were more compact 

 than those of the Erst — fresh lava masses of a beautiful steel- 

 gray luster, and conglomerate bowlders of great general density 

 and up to a cubic meter in size. 



In connection with the crater emanations a series of phenom- 

 ena, of great importance to volcanology were so well revealed 

 by this eruption that the writer feels that he should not close 

 this paper without reference to them. 



From a crater in moderate activity, the usual gaseous emana- 

 tion in fair weather takes the form of a light cream-colored 

 vapor, which rarely fills the whole area of the crater cavity, 

 hut rises as a column of smaller diameter. At other times, 

 however, a dense mass of pure white vapor fills the entire 

 orifice, apparently pouring out of the crater in enormous 

 volume and giving the impression of great activity. But it is 

 often noticed that these two widely differing appearances may 

 occur during the same state of actual eruptive intensity, and 

 even that one condition may be replaced by the other in a few 

 minutes and without any increase or decrease of volcanic 

 activity. 



Further observation reveals the fact that the condition de- 

 pends upon the weather, the first condition prevailing on a 

 fair day and dry state of the air, while a humid wind produces 

 the second. The usual explanation — simple, and applicable to 

 mere fumarolic emanation — is that the water vapor in the 

 volcanic exhalations is absorbed in a dry atmosphere but con- 

 denses in contact with moist and already saturated air. But, 

 to apply this explanation in the case of an active volcano 

 would be to assume that these great volumes of water vapor 

 are being continuously exhaled from the volcano even when the 

 visible " panache " is a slender column rising in the center of a 

 great basin, as is so often the case. 



At Strom bob, on JSTov. 27, the weather in the morning was 

 clear, but threatening change. From Bunta Labronzo the crater 

 could be seen emitting light columns of creamy vapor from the 

 open months while the most easterly mouth was obstructed, 

 and emitted nothing. At noon a chill, moisture-laden wind 

 suddenly supervened, and immediately there poured from the 

 craters — including the obstructed mouth — a dense column of 

 pure white vapor, without the least increase in explosive 

 activity having occurred. 



The present writer has long suspected the atmosphere of 

 being the source of this water vapor, but while it is easy to 

 conceive of a condensation from humid air by a cold mountain 

 peak — as in the case of Alpine "cloud banners" — it was 

 more difficult to account for it over a hot crater. It is here, 

 nevertheless, that we have to seek the explanation of the 

 phenomenon, which is merely due to the nucleation by the 



