the Chemical Theory of Folcanos. 21 



from 47.6 parts of water requisite for the oxidation of the 

 metallic bases in 100 parts of lava, would suppose the presence 

 of about 84 parts of sulphur. In fact, were lava to be formed 

 in this way, and were the hydrogen separated to be united with 

 sulphur, the quantity of the latter ought to be 84 per cent, of 

 that of the lava generated. Applying this result for calculat- 

 ing the bulk of sulphur requisite for producing the lava ejected 

 from Skaptar Jojcul, we find this bulk about 1.4 times as much 

 as that of Mont Blanc, for the specific gravity of sulphur is 

 1.99, and that of lava 3.2. Indeed, such a considerable quan- 

 tity of sulphur can scarcely be assumed in the focus of a vol- 

 cano. But even let us suppose it, on the other hand it must 

 be asked, how comes it that the earthy and alkaline metals, 

 having a strong affinity to sulphur, are not rather united with 

 this substance \ Though these combinations do not take place 

 at a common temperature, yet they were to be expected, as soon 

 as heat is evolved by the commencing oxidation of the metal- 

 lic bases, at the expense of water or atmospheric air. There- 

 fore, the advocates of the chemical theory cannot suppose the 

 presence of sulphur in the focus itself, where these metals are 

 imagined to be, but in higher places, where the hydrogen 

 arising from beneath would unite with it. Granting even the 

 sulphur to be in these higher places, the formation of sulphu- 

 retted hydrogen, nevertheless, cannot be supposed, as it is 

 known that hydrogen and sulphur cannot be united directly, 

 and scarcely even by melting and volatilizing sulphur in hy- 

 drogen gas. 



Finally, it must be well considered that, according to the 

 chemical theory, the decomposition of water by metallic prin- 

 ciples would be the first phasis of volcanic action ; that, conse- 

 quently, the second one would be the combination of the hy- 

 drogen separated with sulphur. These two phases of volcanic 

 action would of course precede by a long period the eruptions 

 of lava. But volcanic phenomena present, in general, the 

 contrary of these suppositions. Disengagements of sulphu- 

 retted hydrogen gas do not take place before the issuing of 

 lava, but, for the most part, after all other volcanic phenomena 

 having ceased. It is well known that these disengagements 

 take place, in general, when a volcano is in tlie state of a sol- 



