60 
Temperature. 
Volume. 
29 
30 
31 
32 
139 
150 
174 
217- 
The apparent expansion of the liquid is doubtless to some extent in- 
creased by the condensation of the gas, as the space occupied by it is 
diminished. When in the highly expanded condition this liquid appears 
to be remarkably elastic. Berthelot has shown, in his paper on forced 
dilatation*, that the force with which liquids adhere to the interior of a 
glass tube is sufficient to prevent their contraction to the normal volume, 
if they have been heated so as to expand and quite fill the tube, arid then 
cooled to a temperature below that requisite to fill it. This fact must 
always be borne in mind in studying fluid-cavities, and explains why the 
bubbles, as it were, hesitate to return, and then make their appearance with 
a sudden start. Such a forced dilatation is very remarkable in the case 
described ; for though it was requisite to raise the temperature to 32? C. 
to fill the cavity, no vacuity was formed until it fell to 31° ; and therefore 
it seems as if the force of adhesion were sufficient to stretch it to considerably 
more than its normal bulk, even perhaps to the extent of one-fifth or one- 
fourth. Moreover, in the case shewn in fig. 1 . , the liquid expanded so 
as to fill the cavity at about 30° C. ; and yet it can be heated up to 42° 
without bursting it, though, even if the expansion did not continue to 
increase, and were the same for each degree as from 31° to 32°, the normal 
volume would be about four times that of the cavity, — which in any case 
seems only to be explained by supposing that its elasticity is most remark- 
ably great, more like that of a gas than of a liquid. There was no decided 
evidence of its passing into a gaseous state, as does occur when cavities 
contain a less amount of liquid. 
Simmler* has shewn that the physical properties of the liquid in topaz, 
as observed by Brewster, agree more nearly with those of liquid carbonic 
acid than with those of any other known substance. Dana, in his ' Minera- 
logy' (5th edition, 1S6S, p. 761,) calls it Brewsterlinite, and says that its 
composition is unknown. The facts at Simmler' s command were not in all 
respects satisfactory, since the amount of expansion given by Brewster 
was from ! 0° to 2G*7°0., whereas that of liquid carbonic acid observed 
* Annales de Chimie ser. 3 t. xxx. p. 232. 
