M. Gay-Luisac on Vokcmos. 87 



oxv*Teii. The perchloiide of iron is very volatile ; it crystallizes 

 on cooling into very small light flakes, which instantly fall into 

 deli(}uescence on exposure to the air. It heats so strongly with 

 water, that I should not be surprised, if, in a large mass, and 

 with a proportional quantity of water, it should become incan- 

 descent. I make this observation in order to suggest to my 

 readers, that if silicium and aluminium really existed in the 

 bowels of the earth in the state of chloride, they might jn-o- 

 duce a much higher temperature upon coming in contact with 

 water, since their aflinity for oxygen is much greater than 

 that of iron. 



If, as can hardly be doubted, sulphurous acid be really dis- 

 engaged from volcanos, it is very difficult to foi"m an opinion 

 of its true origin. Whence should it derive the oxygen neces- 

 sary to its formation, unless it be the result of the decomposi- 

 tion of some sulphates by the action of heat; and of the affi- 

 nity of their bases for other bodies ? This opinion appears to 

 me to be the most probable ; for I cannot conceive, from what 

 is known of the properties of sulphur, that it is an agent in 

 volcanic fires. 



Klaproth and M. Vauquelin have conjectured that the co- 

 lour of basalt might be ascribed to carbon ; but, to confute 

 tliis supposition, we need only remark, that when a fusible mi- 

 neral, even if it contain less than ten hundi'edths of oxide of 

 iron, is heated to a high temperature in a crucible made of 

 clay and pounded charcoal [crcusct hrasque), a considerable 

 quantity ot" iron is produced, as Klaproth has shown in the first 

 volume of his Essays. Messrs. Gueniveau and Berthier assert, 

 moreover, that there I'emains no more than from three to four 

 hundredths of oxide of iron in the scoriae of highly heated fur- 

 naces. Nov/, as lava contains a large pi'oportion of iron, and 

 as the basalt which has been analysed contains from fifteen to 

 twenty-five liundredths of the same substance, it is not pro- 

 bable that carbon could exist in the presence of so large a 

 quantity of iron without reducing it*. 



Is it not possible that if hydrogen be disengaged from vol- 

 canos, metallic iron, the oxides of which have the property 

 of reducing at a high temperature, may be found in lava? 

 It is at least certain that it does not contain iron in the state of 

 peroxide ; for lava acts powerlully on a magnetized bar, and the 

 iron it contains ajipears to be at the precise degree of oxida- 

 tion which alone is determinable by water; that is to say, in 

 the state of deutoxide. I have already shown, that if hydro- 

 gen be mixed with many times its volume of aqueous vapour, 

 It becomes incapable of reducing oxides of iron. 



* When these reflections were read before the Academy of Sciences, 

 M. Vauquelin observed that lie had found carbon in the ashes ejected by 

 the last eruption of V^esuvius. — Ann. de Cliim. toni. xxiii. p. 195. The 



