238 THE ERUPTION OF VESUVIUS IN APRIL, 1906. 



Although it was not possible to make observations upon the dry 

 fuma roles of the crater, I have found in the Vulcan ian breccias 

 nnportant material torn from the deep parts of the crater. A large 

 block contains geodes of colorless and limpid or bluish and opales- 

 cent potassium chloride (sylvite), cubes of which are often more 

 than 2 centimeters in diameter. In some geodes the sjdvite is cov- 

 ered with crystals of sodium chloride. A. Scacchi has shown that 

 the pure chloride of sodium is extremely rare on Vesuvius, and that 

 it is always accompanied by potassium chloride, which is generally 

 predominant. Up to the present time, however, these minerals have 

 been found on Vesuvius only as incrustations, stalactites, and more 

 rarely as small crystals. Never till now have crystals been observed 

 so comparable in size and perfection to those of Stassfurth, which 

 were formed under entirely ditt'erent conditions. The crystals of 

 sodium chloride do not contain potash; the sylvite contains 2.66 

 ])er cent of sodium chloride, existing doubtless as an impurity, 

 because the index of refraction is exactly that for pure sylvite. I 

 have found also crystals of a new mineral, just described by John- 

 ston Lavis — a chloride of potash and manganese with a little soda 

 (chloromanganokalite). There is, further, a small amount of mag- 

 nesium and calcium chlorides and undetermined sulphates. The 

 block impregnated with these salts is a fresh vesicular leucotephrite, 

 and there is no doubt but that the chlorides were formed by subli- 

 mation. A notable quantity of manganese chloride accompanies 

 alkaline chlorides. The great size of the crystals shows they were 

 formed at a high temperature, free from disturbance. 



SECONDARY FUMAROLES. 



These products emitted by a lava flow are determined by the 

 decreasing temperature of the magma in the course of cooling, and 

 are constituents of the fluid magma at the time of its ejection. 



If a limited portion of magma is violently expelled instead of 

 being quietly poured out, it cools quickly and loses at once all the 

 volatile products which do not remain imprisoned. Under conditions 

 of slow cooling, however, fragmental lava may behave like the con- 

 tinuous rock of a flow, the nature of the products of emanation being 

 limited only by the temperature at the time of ejection. Breccias 

 accumulated in a short time, like those of Mount Pelee, offer the best 

 conditions for the slow cooling of a new magma transported to a 

 distance. They were formed of coarse and fine material of poor 

 conductivity for heat, and contained some blocks of high temperature. 

 The release of volatile products, slow and tranquil when the cooling 

 of the breccia was undisturbed, rapid and violent when it was 

 hastened by rain water, determined the production of fumaroles, first 



