have fallen from the Atmofphere. 35^ 



believe that the volcanoes which burn at its furface are in a 

 ftate of tranquil inflammation, maintained by a very thin 

 ftratu-a of atmofpheric air, very dilferent from that of our 

 volcanoes in a ftate of eruption ; and, confequently, that they 

 cannot be endowed with a very ftrong proiettile force : for 

 on the earth this force is produced by the expanfibility of 

 gas comprelfed by the weight of the atmofphere. But gales 

 difengaged from the lunar volcanoes ought to expand with- 

 out any obftacle, and without any explofion where there is 

 fcarcely any atmofphere. Befides, the exiftence of thefe 

 gafes is doubtful, fince their produ6lion would foon have 

 formed around the moon an atmofphere of greater denfity 

 and extent, which is contrary to obfervation. This explana- 

 tion therefore muft be abandoned, or we (hould be forced to 

 admit a fcries of phaenomena, not only foreign but contrary 

 to what we daily obferve : a folution more (imple, and de- 

 duced from fads, which daily take place before our eyes, 

 will perhaps appear more proper for folving the problem. 

 It is well known that there are no metals which may not 

 be volatilized by heat: it is known alfo that hydrogen gas 

 exercifes its diffolving power on charcoal, iron, and the mod 

 fixed fubftances. It we fuppofe that, during the burning 

 procel's of volcanoes, or the Icfs perceptible but more affi- 

 duous procefs of the decompofition of organic bodies, iron 

 and nickel are fublimatcd by heat and elevated with the 

 hydrogen gas which difl'olves them, this gafeous folution 

 will foon reach the upper regions of the atmofphere. There, 

 as I have fliown, with fome degree of probability, in my 

 Conje£lures on the Dimhjutiou of the Waters, &^c. exift:^, 

 and is continually renewed, a ftratum of hydrogen produced 

 by the continual decompofition of the water, and which 

 caufes the greater part of the phajnomena that accompany 

 thunder and the aurora borealis. During fi;orms, that is to 

 fay, when the equilibrium is refiored with an explofion be- 

 tween the eleftricity of the earth and that of the atmofphere, 

 what muft take place? The hydrogen inflames and exhibits 

 fome of thofe luminous meteors, the exiftence of which, ac- 

 cording to conftant tradition, ought, as appears, to precede 

 the formation of ftoncs. The gas in burning abandons the 

 metal it has diffolvcd, and reduces that which was in the 

 hate of oxide. The ftrong heat produced at this moment 

 fuzes the metal, and molecular attraftion collecls it into 

 inalils of greater or lefs fize, which, when they fall to the 

 earth, retain for fome time a portion of the caloric developed 

 during their formation. The furface only, whicb, by ira- 

 verfing the atmofphere in a ftate of ignition, may have ab- 

 A a 3 forbcd 



