96 On the Fluids contained in crystallized Minerals. 



visible. The crystals were completely immersed in the fluid, 

 together with a few minute globules of air, which had come out 

 of the cavity. The crystals daily increased in bulk, with a 

 corresponding diminution of the fluid ; but a fortnight elapsed 

 before the crystaUization of the fluid was complete. Even then 

 a slight degree of moisture could be observed on the surface 

 of the crystals, and also on the space included within them, and 

 this moisture still remains. Some of the globules, which are ex- 

 tremely small, still remain in a fluid state. When the crystals 

 attained such a size as to come near the surface of the fluid, the 

 edges of the upper surface of some of them gradually rose above 

 it, and these have now the form of an inverted four-sided pyra- 

 mid, a form which is often assumed by muriate of soda, when 

 slowly crystallizing. 



The elasticity of the globule of air in the cavities of all the 

 crystals I have yet examined, is evidently great, for whenever 

 a rent was formed, the globule, however small before, always ex- 

 panded to such a degree, as to expel the whole of the fluid. In 

 the cavity of fluor-spar above mentioned, the globule of air ex- 

 panded to more than the size of the cavity, for a part of it even 

 escaped along with the last portions of the fluid. Indeed I 

 have repeatedly found the elasticity of the air to be so great, 

 that when a direct opening was suddenly formed into some ca- 

 vities of sulphat of barytes, the whole of the fluid was blown out 

 in an instant, not a trace of it being left behind. 



I have observed a very curious property of the globule of air 

 in the fluid cavities of various minerals. These globules, when- 

 ever they are moveable, always occupy the upper part of the ca- 

 vity in which they occur ; but if the end of a heated wire be 

 made to touch the surface of a crystal next the under side or 

 end of a cavity, the globule of air immediately descends to it, 

 and that, too, with a rapidly accelerating motion. On removing 

 the wire, the globule immediately ascends to its former posi- 

 tion, but with a uniform motion. Perhaps you can afi^ord an 

 explanation of this phenomenon. I am, &c. 



Edinburgh, / 



