the Chemical Theory of Volcano*. 17 



of lava. 100 parts of atmospheric air, containing 23.32 parts 

 of oxygen by weight, 100 parts of lava will require 180 parts 

 of air by weight to be produced. Lava being 3.2 times heavier 

 than water, and water being 770 times heavier than atmo- 

 spheric air, lava is consequently 2464 times heavier than the 

 latter. Therefore, the production of 1 volume of lava will 

 require 4435 volumes of air. In my former papers, I have 

 mentioned that the violent eruption in the low country of 

 Skaptar Jokul in Iceland, in 1783, brought up a mass of lava 

 so considerable as to surpass in magnitude the bulk of Mont 

 Blanc. According to that calculation, such a mass, if consti- 

 tuted like the lava examined by Dr Kennedy, would have 

 required a quantity of atmospheric air equal to the bulk of 

 4435 times the size of Mont Blanc, had that lava been ge- 

 nerated by oxidation of the metals contained in it. Thus 

 a bulk of nitrogen gas of 3503 times as great as that of the 

 mountain mentioned, must have been evolved during the for- 

 mation of that mass of lava. 



It is beyond all probability to suppose caverns of such an 

 enormous extent in the interior of the earth for containing 

 such huge quantities of atmospheric air. Therefore, the ad- 

 vocates of the chemical theory will be obliged to assume very 

 long periods for the formation of such masses of lava, during 

 which the passages between the internal caverns and the ex- 

 ternal atmosphere must have been always open. Thus, they 

 must suppose that atmospheric air must enter these caverns 

 and quit them, after having been deprived of its oxygen, and 

 that this entrance and exit of air took place unceasingly 

 during the whole time of these actions. But it is obvious 

 that these streams of air must be supposed to exist in such 

 proportions as to support these actions, as soon as melted 

 masses require to be kept in the liquid ignited state. There 

 can, therefore, be no question that very considerable quantities 

 of nitrogen gas ought to appear in the neighbourhood of such 

 subterraneous actions on the surface of the earth. Are we 

 to suppose that the scanty exhalations of nitrogen gas from 

 thermal springs connected with volcanos proceed from such 

 actions ? Such exhalations are quite insignificant when com- 

 pared with the formation of the masses of lava alluded to. 



VOL. XXX. NO. MX. JANUARY 1840. B 



