280 PERMEABILITY OF HIGHLY-HEATED IRON BY GASES, 



The oldest fact known of gas held in solution by substances in a state of ig- 

 neous fusion, is that which occurs in the ebullition of silver. The similar phe- 

 nomena which litharge gives at the instant of melting, were explained in the 

 same way by M. Thenard ; and the admirable researches of M. F61ii Le Blanc, 

 left no doubt in this respect. Lastly, the curious experiments which my brother 

 brought forward at the meeting on Dec. 14, give a most direct proof that 

 vitreous bodies in fusion possess the property of absorbing and of subsequently 

 disengaging gaseous substances, obtained from the surrounding medium ; 

 and in that case, the gas was of a combustible nature. It was natural, and 

 long ago it occurred to me, to connect with this singular property of lithoid 

 substances in fusion many facts which have been observed in recent lavas 

 and volcanic eruptions. The lavas which issue from volcanoes form two 

 distinct varieties, from our present point of view. The first, being rich in silica 

 and very readily fusible, easily assume the vitreous condition on cooling, and 

 then form obsidian ; the others, which are of more common occurrence (dole- 

 rites, amphigenites, basalts) contain generally not more than 50 per cent, of 

 silica, and moat of them are rich in lime. To fix our ideas by an example, the 

 neighbourhood of Naples presents both these varieties of rocks, — the old trachytes 

 and the pumiceous tufas of the Phlegrcean plains on the one hand, and the am- 

 phigenitic masses of the Somme and Vesuvius on the other. The lavas of the 

 volcano, whatever may have been their rate of cooling, are always crystalline, 

 with some very rare exceptions of very small extent, which are gubvitreous or 

 imperfectly crystalline. The volatile matters, such as steam, metallic chlorides, 

 hydro-sulphuric acid, &c,, which they contain, and which must have been dis- 

 solved in the highly-heated medium where they were fused, disengage them- 

 selves successively in the order I have explained, in proportion as the interior 

 work of crystallisation went slowly on ; precisely as at the instant of the ebulli- 

 tion of silver there is an escape of oxygen, or, as in another class of phenomena, 

 the air held in solution in water is separated from it at the instant of freezing. 

 The act of crystallising causing a large and sudden increase of density, there 

 results at that instant a corresponding disengagement of latent heat, and I do 

 not hesitate to assign to this cause the subsequent heating of the lava of 1855, 

 observed by M. Scacehi, and verified by M. Albert Gaudry and myself. Similar 

 facts did not escape older observers, for Serrao, after having proved the occur- 

 rence of this in the lava of 1737, remarks, that "the lavas must contain within 

 themselves some cause which develops heat, and brings them back to incan- 

 descence- when they have been already completely cooled (on the surface)." 



The flames which have been often observed in Vesuvius, and in particular by 

 Leopoldo i'illa, could be attributed only to the combustion of gases given out 

 during the eruption ; but, at the last eruption of December, 1861, I was fortu- 

 nate enough to put beyond doubt the fact that combustible gases are disengaged 

 from the incandescent lava in the act of cooling, and the exact analyses made 

 on my return, by MM. Le Blancj Fouqud, and myself, have proved that they con- 

 sisted of a mixture of light carburetted hydrogen and hydrogen. It is then na- 

 tural to admit that the incandescent matter was surrounded, in the furnace from 

 which it proceeds, by an atmosphere of this nature, that it became impregnated 

 •with it while in a liquid state, and again set it free in its progressive passage to 



