in the Cavities of Minerals. 507 



fear, however, of bursting so rare and interesting a cavity, 

 has prevented me from making this experiment. The cavity 

 contains a great number of crystals of different forms, not one 

 of which melts with heat, and almost all of which possess 

 double refraction. When I first submitted this cavity to the 

 microscope, there were Jive small crystals lying between D and 

 the vacuity V; one a flat prism, another a hexagonal plate, a 

 third amorphous, and a fourth and fifth two irregular halves 

 of a hexagon. Upon the first application of heat, one or two 

 of these crystals leapt from their resting place, and darted to 

 the opposite side of the cavity. In a iew seconds the others 

 quitted their places one after another, performing the most 

 rapid and extraordinary rotations. One crystal joined an- 

 other, and, at last, four of them thus united revolved with such 

 rapidity as completely to efface their respective shapes. They 

 then separated on the withdrawal of the heat, and took the 

 position which their gravity assigned them. On another oc- 

 casion, a long flat prism performed the same rotation round 

 its middle point; and I have repeated the experiment so often, 

 in showing it to others, that the small crystals have been driven 

 between the inclined edges of the cavity, from which I cannot 

 extricate them. I have succeeded, however, in conducting a 

 fine octohedral crystal, truncated on its edges and angles, into 

 the arena at D, where I have just seen it perform its rotation, 

 as indicated by the concentric circles on the right-hand of D. 



In seeking for the cause of so extraordinary a phaenomenon, 

 we are reminded of the rotations of camphor and other vola- 

 tile substances ; but in this case no gas or matter of any kind 

 could be thrown off without becoming visible in the fluid. The 

 pyro-electricity of topaz next suggests itself as a moving power; 

 but though it might produce attractions and repulsions, we 

 cannot see how it could turn a crystal upon its axis. The 

 experiments of Libri and Fresnel, on the repulsions which 

 heated bodies exert upon each other at sensible distances, 

 afford us as little aid. They may enable us to account for 

 the mere displacement of the crystals by the application of 

 heal, or for their sudden start from their places of rest, but 

 they do not supply us with a force fitted to give and to sustain 

 a rapid rotatory movement. 



I have already had occasion to state, that the cavities often 

 burst when too much heat is applied to the specimen. This 

 generally takes place by a separation of the laminae, which 

 fly off in splinters; but when the burst cavity is large and 

 insulated, a piece of the solid crystal is scooped out on its 

 weakest side. Sometimes a great number of cavities explode 

 at the same time, and when they are small, or exist in a part 



