After upwards of twenty years' study and observation of meteoric electricity, I am enabled to prove that atmospheric electricity is never manifested without rain, hail or snow, and that manifestations of light are always accompanied by thunder--manifestations of light (_lampi_), thunder and rain being most closely connected. We may have rain without manifestations of light, but never the latter without rain or hail. I cannot here repeat what I have demonstrated in other memoirs; I can only say that the lightnings of Vesuvius, erroneously believed to be not accompanied by thunder, are really not accompanied by rain, but are induced by the descent of ashes and lapilli.[6] GENERAL CONCLUSIONS. We may conclude from what I have stated: 1. That by the assiduous study of the central crater, and the indications afforded by the "Apparatus of Variations" and the "Electro-Magnetic Seismograph," we can obtain precursory signals of eruptions; and that the other premonitory signs pointed out by our ancestors, such as the drying up of wells, either only happen occasionally or are mere coincidences, such as those of the coincidence of a dry or a rainy season, the prevalence of certain winds, etc.[F] 2. That the fumaroles of the lavas are communications between the external surface of the lava, hardened and more or less cooled, and the interior lava still pasty, or at least incandescent. 3. That from the lava, while flowing, there is no escape of acid vapours, neither from the fumaroles at the first period of their existence, but these, if they last long enough, arrive at an acid period. 4. That hydrochloric is the first acid that appears, combined afterwards with sulphurous acid, and, still later, with sulphuretted hydrogen. 5. That vigorous lava streams may have eruptive fumaroles. (See Translator's Note 2 to p. 94.) 6. That the sublimations follow a certain order in their appearance. In the neutral period we get sea-salt mixed with some metallic oxides, the first of which is oxide of copper. But in the great lavas, chloride of iron appears simultaneously with the acid period. Hydrochloric acid transforms the oxides into chlorides, which, in their turn, change into sulphurets or sulphates on the appearance of sulphurous acid. 7. That the acids, by attacking the scoriæ, create new chlorides and sulphates, which are thus not products merely of sublimation. 8. That micaceous peroxide of iron--so common and abundant near the eruptive mouths--is very scarce and rare on the lavas, unless conveyed there from the craters. 9. That chloride of iron--so manifest on the fumaroles of the great lavas--is only found in small eruptions close to the discharging mouths. 10. That the frequency of chloride of iron in the lavas of great eruptions masks the order of transformation of the other products. 11. The fumaroles at the summit of Vesuvius present even greater gradations, for they often emit carbonic acid or pure watery vapour.