154 Scientific Proceedings, Royal Dublin Society. 



ized during an eruption is hardly credible until a few facts convince 

 us that such is the case. I have seen fumarole chimneys having 

 in a short time their whole interior glazed by a mixture of chlorides, 

 one to three centimeters thick, and from the intense heat as trans- 

 parent as an ice covering, which was, without doubt, the result of 

 sublimation, and not decomposition, as the rocks upon which it was 

 deposited were quite unaltered. Another proof of the large amount 

 of saline substances ejected by a volcano is the quantity met with 

 in the falling ashes during a lava eruption. The outburst in 1872 

 produced an ash asserted by Prof. Palmieri 1 to be poorer in soluble 

 constituents than any other since 1855, yet it contained from 4 to 

 9 per cent, of saline matter, chiefly sodic chloride. As this erup- 

 tion was lateral, the principal part of the ash was derived from the 

 crater edges and chimney walls, which would tend to lower the 

 amount of soluble portion. 



It was observed in the eruption of 1855 2 that the alkaline 

 chlorides were only evolved sometime after the lava had been cool- 

 ing — that is to say, saline crusts only formed around the f umaroles 

 at a late date ; and I have noticed the same thing. Scacchi sup- 

 posed that it may be a spontaneous rise in temperature in the lava 

 in cooling, similar to that developed in phosphate of lead, nitrate 

 of copper, or argentic 3 iodide when passing from the amorphous to 

 the crystalline condition. Or again, to their early union with 

 other elements of the lava. This may possibly be so, the combi- 

 nation being broken up by a lowering of temperature (?), leaving 

 the chlorides free to be sublimed. It seems to me that the chlorides 

 must be continually escaping, but that they are not deposited until 

 the scoria and fumarole sides are cooled enough to allow such to 

 occur. The liquids included in cavities in crystals are generally 

 solutions of chlorides or sulphates. 



There is little doubt that these saline materials must form a 

 very important constituent of the magma ; but whether they play 

 much part as a solvent medium for certain minerals is a thing yet 

 to be experimentally verified, though one is inclined to think that 

 they really do perform a very important function in that way. 



i Annali del Reale Osserv. Meteor. Vesuviano, 1874, p. 73. 



2 Guarini, Palmieri, Scacchi. Mem. s. Incend. Vesuviano del mere di 

 1355, &c, pp. 141, 143, and 149. 



3 G. F. Rodwell, Phil. Trans. E. S., Part iii., p. 1134. 



