As the smoke of the fluid lava is perfectly neutral, that is, neither acid nor alkaline, so the fumaroles at the first period of their existence with sublimations of sea-salt, mixed frequently with oxide of copper either in black powder or in shining laminæ, ought also to be neutral. But if the fumarole continues active, hydrochloric acid issues with the smoke, and often some time after sulphuric acid. Then the sublimations turn first yellow, then green, and more rarely azure. The chemical reactions show that these sublimations are chlorides or sulpho-chlorides, and sometimes sulphides, and they afford reactions, indicative of soda, magnesia, copper, lead, and traces of other substances, not excluding ammonia, which I must speak of separately. This, I have observed, is the general law with the fumaroles of the tranquil lavas, which occur with long and moderate eruptions--for instance, the lavas of 1871, and even those of 1872, preceding the 26th April. But in the great lavas of the great conflagrations of Vesuvius, chloride of iron more or less in combination with all the other substances above mentioned changes the appearance of the sublimations. The fumaroles in the lava of the 26th April frequently indicated chloride of iron. Sulphuretted hydrogen, by reaction of sulphurous acid, is decomposed, and sulphur sublimed, having a particular aspect, collects on the scoriæ. This is never found but in fumaroles of the smaller lavas; it was therefore absent in those of 1871, but frequently occurred in those of 1872.