88 EFFECT'S of HEAT. 



accurately filled with metal, which plated the carbonate all 

 over without penetrating it in the leaft, fo that the metal was 

 eafily removed. The weight was reduced from forty to thirty- 

 fix grains. The fub fiance was very hard, and refifled the knife 

 better than any refult of the kind previoufly obtained ; its frac- 

 ture was cryftalline, bearing a refemblance to white faline 

 marble; and its thin edges had a decided femitranfparency, 

 a circumftance firfl obferved in this refult. 



On the 3d of March of the fame year, I made a fimilar 

 experiment, in which a pyrometer-piece was placed with- 

 in the barrel, and another in the muffle ; they agreed in indi- 

 cating 23 . The inner tube, which was of Reaumur's porce- 

 lain, contained eighty grains of pounded chalk. The carbo- 

 nate was found, after the experiment, to have loft 34- grains. 

 A thin rim, lefs than the 20th of an inch in thicknefs, of 

 whitifh matter, appeared on the outfide of the mafs. In other 

 refpects, the carbonate was in a very perfect fiate ', it was of a 

 yellowifh colour, and had a decided femitranfparency and 

 faline fracture. But what renders this refult of the greatefl 

 value, is, that on breaking the mafs, a fpace of more than the 

 tenth of an inch fquare, was found to be completely cryftal- 

 iized, having acquired the rhomboidal fracture of calcareous 

 fpar. It was white and opaque, and prefented to the view 

 three fets of parallel plates which are feen under three different 

 angles. This fubftance, owing to partial calcination and fub- 

 fequent abforption of moifhire, had loft all appearance of its 

 remarkable properties in fome weeks after its production ; but 

 this appearance has fince been reftored, by a frefh fracture, 

 and the fpecimen is now well preferved by being hermetically 

 kiclofed. 



III. 



