PERMEABILITY OF HIGHLY-HEATED IRON BY GASES. 281 



the crystalline condit'on. The subsequent heating which I hare mentioned in 

 the gases escaping from the lava is doubtless also an indication of the heat 

 rendered sensible by the act of crystallising. 



When the eruptive matter, instead of having, like the lavas above spoken of, 

 a great tendency to crystallise, offers on the contrary, along with an excess of 

 silica, a tendency to solidify in the vitreous form, it constitutes obsidian. It 

 then imprisons and solidifies in some way the volatile substances held in solution 

 by it, and at the same time it retains a certain quantity of latent heat (which I 

 propose to call the latent heat of fusion) which gives it a minimum of density; 

 But it is remarkable, that if we proceed to heat this obsidian nearly up to its 

 melting point, it puffs up so that its volume increases enormously ; and yet, 

 this extreme porosity of substance, rendering it everywhere of excessive fria- 

 bility, and, as it were, papyraceous in texture, corresponds only to an insignifi- 

 cant loss — some thousandths — of its original weight. When thus once trans- 

 formed into pumice, it requires a very intense heat to soften it anew and melt 

 it. Is it not natural to suppose that the temperature to which the obsidian was 

 at first subjected, and which was relatively low, has only brought this glass to 

 a particular molecular state, which by permitting the stored-up heat to be dis- 

 engaged, has furnished the rest of the supply of caloric necessary to resoften the 

 substance and facilitate the expulsion of the gases ? Just as in the well-known 

 experiment of M. Regnault, the soft sulphur (that is, the vitreous sulphur, the 

 obsidian of sulphur) when raised to 92 or 93 degrees, suddenly sets free a cer- 

 tain quantity of heat, and raises the temperature of the thermometer in contact 

 with it to 110°. 



However this may be, let us revert to the Phlegraean plains which surround 

 Vesuvius. We shall find them to consist entirely of trachytes, obsidian, and 

 pumice, all which are beyond compare vitreous or vitrifiable substances. We 

 may therefore conceive that a relatively small elevation of temperature, much 

 inferior to that observed at each eruption of Vesuvius, on being applied in the 

 interior of the soil to the masses of obsidian, changes them into pnmice, with a 

 large increase of volume ; and from this there would result an immense force 

 which, breaking the overlying crust, would lift it up in a bubble-shaped heap, 

 projecting its fragments in all directions. Thus would be accounted for, as I 

 have already remarked, both the facts observed at Monte Nuovo in 1538, and 

 the production of the numerous craters of the Campagna. 



Lastly (and I need not say that I offer this conjecture with reserve), if we 

 notice the resemblance that exists between the map of the Phlegrcean plains and 

 that of the moon's surface, it is natural enough to believe that this latter owes 

 its form to action of the same nature, and it may not perhaps be inappropriate 

 to remark that a globe composed entirely of vitrified matter may thus have con- 

 densed and retained in solution within its own mass the gaseous elements which 

 originally surrounded it, and which , but for this circumstance, would have con- 

 stituted an atmosphere for it. And in applying this conception to our own 

 globe, is it not conceivable that the primitive graaitic crust, essentially rich in 

 silica (a substance of which I have proved the extreme fusibility) had condensed 

 before its solidification, a portion at least of the gases which form our atmos- 



VOL. IX. T 



