in the Cavities of Minerals. 15 



paz belonging to Mr SIVRIGHT. In examining this specimen, I 

 observed a very remarkable cavity, of the form shewn in Fig. 11., 

 where A, B and C are three separate portions of the new fluid, 

 insulated by the interposition of the second fluid DEF. The 

 first portion A of the new fluid had four vacuities V, X, Y, Z, 

 while the other two portions B, C, had no vacuity. Having of- 

 ten succeeded in making the vacuities pass from one branch of a 

 cavity to another branch, I did not doubt that the vacuities of 

 the portions B and C had passed over the second fluid into the 

 portion A. In order to determine this, I took an accurate draw- 

 ing of all the phenomena at a temperature of 50, as represented 

 in Fig. 11., and I carefully watched the changes which took place, 

 by raising the temperature to 83. The new fluid at A gra- 

 dually expanded itself, till it filled up ah* the four cavities V, X, 

 Y, Z ; but as the portions B, C, had no cavities for this purpose, 

 they could only expand themselves, by pushing back the sup- 

 posed second fluid DEF. This actually happened. The second 

 fluid quitted entirely the edge of the cavity at F. The two 

 portions of new fluid B, C, were immediately united into one ; 

 and the second fluid having retreated to its new limit mnn'o, and 

 being itself but slightly expansible, h'ke common fluids, its other 

 limit necessarily advanced to p q r. This experiment, which I 

 have often repeated, and shewn to others, involves one of those 

 rare combinations of circumstances, which Nature sometimes pre- 

 sents to us, in order to lay open some of the most mysterious of 

 her operations. Had the portions B, C, of the new fluid been 

 accompanied, as is usual, with their vacuities, the interposed se- 

 cond fluid would have remained immoveable between the two 

 equal and opposite expansions : but from the accidental circum- 

 stance of these vacuities having passed over into the other branch 

 A of the cavity, the second fluid is placed in a sort of unstable 

 equilibrium, and, h'ke the arms of a lever, it yields to every varia- 

 tion of the power and of the resistance. 



