contained in crystallized Minerals. 9^ 



were nearly hemispherical, and seemed of considerable density, 

 those from other cavities spread out to a considerable extent, m-, 

 dicating less tenacity, and a greater attraction, between the par- 

 tides of the fluid and the surface on which they spread. The 

 globules, too, form different cavities, crystallized with very dif- 

 ferent degrees of rapidity. Several minutes elapsed, before the 

 dense hemispherical globules of one cavity began to crystallize. 

 The crystallization then went on slowly, and was not com- 

 pleted until after the lapse of twenty-four hours ; whereas most 

 of the thin flattened globules from other cavities, crystallized al- 

 most the instant after they reached the surface. The dense he- 

 mispherical globules seemed to lose very little by evaporation ; 

 but the thin flattened globules seemed to sustain a very con^-. 

 derable loss by that process. 



In the instance first mentioned, each globule of the fluid 

 formed only one crystal ; but in all the others, each globule gave 

 forth a considerable number of crystals. These were always 

 arranged in a curve, immediately within the circumference of 

 the globule. Sometimes the crystals were aggregated together, 

 sometimes they were more or less detached, and sometimes a 

 detached crystal or two formed within the curve. The whole 

 of the crystals had the same form, that of a right prism with 

 a rhombic base. 



Since, therefore, the cavities, in sulphat of barytes, evidently 

 contain the matter of that substance in a fluid state, it seems 

 fair to infer, that the cavities in other crystallized minerals may 

 contain their own matter in a similar state. This I have late- 

 ly ascertained to be the case with fluor-spar. About two months 

 ago, I succeeded in forming a partial rent in a crystal of that 

 mineral, containing a cavity with a fluid and moveable globule 

 of air. The instant the rent took place, the air began to ex- 

 pand, and continued to do so, until the whole of the fluid was 

 expelled from the cavity. The fluid appeared on the line of the 

 rent, in the form of twelve distinct globules. These were tena- 

 cious, and of a hemispherical form. One of them was much 

 larger than all the rest put together. For several hours after 

 the fluid came out, there was no appearance of crystallization ; 

 but, next morning, a number of cubical crystals, aggregated in a 

 curve within the margin of the largest globule, were distinctly 



