TRANSACTIONS OF THE SECTIONS. 



33 



thing's, a theoretical continuity across this breach, having some real and true sig- 

 nificance. The general character of this view may readily be seen by a glance at 

 fig. 2, m -which Dr. Andrews's curves are shown by continuous lines (not dotted), 

 and cun-ed reflex junctions are shown by dotted lines connecting those of Dr. An- 



Zero x 

 I I I Islol 



j-'oi- Vol mive^ 



Fi^e s s cc 7-es 



17 5 1 I I loui 



drewss curves which are abruptly interrupted at their boiling- or condensing-points 

 of pressure. It is to be understood that each curve relates to one constant tempe- 

 rature, and that pressures are represented by the horizontal ordinates, and coitc- 

 spondrng volumes, of one mass of carbonic acid constant throughout aU the curves 

 are represented by the vertical ordinates. The author points out that, by experi- 

 ments of Donny, Dufotu", and others*, we have already proof that a continuation 

 of the curve for the liquid state past the boiling stage for some distance, as shown 

 dotted in fig. 2, from a to some point 6 towards/, would correspond to states already 

 attained. He thinks we need not despair of practically realizing the physical con- 

 ditions corresponding to some extension of the gaseous curve such as from c to «f 

 in the figure. The overhanging part of the curve from e to / he thinks may re- 

 present a state in which there woidd be some kind of unstable equilibrium ; and 

 so, although the curve there appears to have some important theoretical significance, 

 yet the states represented by its various points would be tmattainable through oh. 

 any ordinary mass of the fluid. It seems to represent conditions of coexistent tem- 

 perature, pressure, and volume, in which, if all parts of a mass of fluid were placed, 

 it would be in equilibrium, but out of which it would be led to rush, partly into 

 the rarer state of gas, and partly into the denser state of liquid, by the slightest 

 inequality of temperature or of density in any part relatively to other parts. 



* Donny, Ann. de Chimie, 1846, 3rd series, vol. xvi. p. 167 ; Dufour, Biblioth^que 

 Universelle, Archives, 1861, vol. xii, 



1871. 3 



