336 Analysis of' Scientific Books and Memoirs. 



reason to conclude, that the crystalline particles of which they are com* 

 posed were not formed as the substance cooled ; that few or no lavas are 

 ever reduced naturally to complete fusion (none in fact but the glassy la- 

 vas, obsidian, pearlstone, &c) ; but that they consist of crystalline par- 

 ticles of various sizes, which, when the rock is solid, contain very minute 

 portions of water mechanically combined with their substance,* that is, 

 intervening between the parallel plane surfaces of the crystals. In this 

 case, any continued accession of caloric to a mass of such rock confined 

 beneath the crust of the earth, and already at an intense temperature, 

 must sooner or later so increase the expansive force of the confined water, 

 as to reduce more or less of it to vapour, breaking through or heaving up- 

 wards the confining crust, and causing the lava to intumesce and rise out- 

 wardly in a state of imperfect liquefaction through any fractures which 

 this violent expansive effort may create in the overlying beds. The li- 

 quidity of lava consists, under this idea, not in its absolute fusion, but in 

 the mobility afforded to its component crystals by the intervention of 

 more or less of highly elastic vapour between the opposite facets, and the 

 degree of liquidity will therefore depend on the quantity of vapour gene- 

 rated through the substance, and the size of the component crystals. But 

 at a certain term in the relative proportion of these circumstances, the va- 

 pour will become so abundant as to enable a part of it to unite into 

 bubbles, which, by their inferior specific gravity, are urged to rise up- 

 wards, and escape from the surface of the liquefied mass in which they are 

 formed. " The quantity of vapour discharged in this manner consists 

 therefore at all times of the surplus of that which has been generated in 

 the lava beyond what is necessary to communicate to its component crys- 

 tals the degree of mobility required for the union of this surplus vapour 

 into parcels or bubbles, and the rise of these, when formed, to the surface." 

 The explosions of all volcanic eruptions are produced by the rapid ascent 

 and escape of such bubbles, collecting as they rise through the lava into 

 prodigious volumes of vapour ; and the remainder of these parcels of va- 

 pour, which are prevented from thus escaping by the superficial indura- 

 tion of the exposed masses of lava, occasions the cellular and cavernous 

 structure of such rocks, both on the large and small scale. The consoli- 

 dation of liquefied lava, under these circumstances, takes place, not only 

 by loss of temperature, but also, and chiefly, by the immediate escape of 



* This, at least, is the temporary assumption of the author, who, in a later part 

 of the work, observes, that he inclines to suppose the water itself may be generated, 

 together with other fluids, by the volatilization of a superficial pellicle of the proxi- 

 mate crystals, and the combination of the oxygen and hydrogen set free by this 

 process, through the intense temperature pervading the mass. Such a supposition, 

 if not supported, is perhaps not opposed by the present state of chemical knowledge ; 

 and would explain all the phenomena of lavas, as well as the idea of a mechanical 

 interposition. In short, the existence and general dissemination of water, or rather 

 steam, in lavas, is a positive fact, susceptible of direct and incontrovertible proof; 

 and it is indifferent to the purpose of the author how, or at what time, we suppose 

 it to be produced there. 



