G2 
260° C, with an elastic force of 119 atmospheres; whereas water 
appeared to expand to nearly four times its volume, and required a tem- 
perature near that at which zinc melts (328° C, Daniel). When in this 
highly expanded state, the liquids were very mobile, and seemed much 
more compressible than under other circumstances ; for they did not burst 
the tube, if too much had been sealed up in it, until after their normal volume 
would have been decidedly greater than its capacity. No one could fail 
to see that these phenomena have much in common with what occurs at a 
lower temperature in the case of the liquid inclosed in sapphire, and that 
they are of great importance in connexion with the origin of fluid-cavities. 
Since they become full of liquid at a comparatively low temperature, it 
was not unreasonable to suppose that the minerals in which they occur 
must have been formed where the heat was scarcely above that of the 
atmosphere; but these facts seem to show that the occurence of such 
fluid cavities is quite reconcilable with a very high temperature; for it is 
obvious that if, at a great depth below the surface, heated, highly com- 
pressed gaseous carbonic acid were inclosed in growing crystals, it might 
condense on cooling so as to more or less completely fill the cavities with 
the liquid acid. 
If the same principles could be applied in the case of water, we should 
be led to infer that it could not exist in a liquid state at a higher tempe- 
rature than that of dull redness, corresponding closely with what Mr, 
Sorby deduced from the fluid-cavities in some volcanic rocks. In that 
case, according to Cagniard-Latour, the liquid when condensed would 
occupy only one-fourth part of the cavity, audit would scarcely be likely to 
contain any fixed salt in solution; whereas the fluid- cavities in the 
minerals of ejected blocks are often two-thirds full of what seems to have 
been a supersaturated solution of alkaline chlorides. The phenomena now 
under consideration should certainly be borne in mind in studying volcanic 
action ; and it is possible that some cavities now containing water may 
have been formed by the inclosure of very highly compressed steam. In 
some cases the requisite pressure would be enormous, and other facts seem 
to show that it was more generally caught up in a liquid state. 
The cavities in emerald are very interesting in connexion with this subject, 
and also furnish strong evidence against the opinion that the liquid was not 
present when the crystals were formed, but penetrated into the fluid- 
cavities at a subsequent period, and either filled vacant spaces or removed and 
Ann. de Chimie, 1S22, t. xxi. pp. 127 & 178 ; t. xxii. p. 410. 
