Procuring Crystallisations in the Dry Way. 313 



of ttis new kind of dissolvent, or by the simple cooling of 

 the solution made at a high temperature. It is under the 

 first of these two points of view that he has undertaken a 

 series of experiments, the result of which he has laid before 

 the Academy. 



M. Ebelmen immediately thought of the known dissolvents 

 of this kind, such as boracic acid, phosphoric acid, the alka- 

 line borates and phosphates, which dissolve, as is known, a 

 considerable number of oxides when they are in a state of 

 fusion, and have the property besides of volatilising slowly 

 at a high temperature, whence it might be supposed that the 

 solution would leave the dissolved substances under crystal- 

 line forms. The experiment completely succeeded, although 

 made in circumstances by no means favourable ; for it was 

 carried on in the porcelain ovens of the royal manufactory, 

 where they reach the maximum temperature very slowly, 

 and stop it almost suddenly ; so that there are only five or six 

 hours for the evaporation of the dissolvent, during which 

 crystallisation took place. It followed from this that only 

 very small crystals could be obtained, and that, in many 

 cases, none could be obtained at all. It is likewise proper to 

 add, that the experiment is quite novel ; and we do not know, 

 with respect to these dissolvents, the degree of solubility of 

 the different substances. 



In spite of these obstacles, M. Ebelmen has made a series 

 of experiments, the results of which are of high interest. 

 On the one hand, by dissolving alumina in borate of soda 

 or in boracic acid in fusion, he has made this substance 

 crystallise, and has obtained the mineral known by the name 

 of corundum, with all its characters. It is true that the crys- 

 tals obtained are small, but pei'fectly formed, and possess all 

 the characters of infusibility, great hardness, lustre, ci'vstal- 

 lisation, optical properties, and properties of composition, 

 that we discern in the identical natural substances. 



Thus the general problem is solved — infusible substances 

 present no greater difticulties than the otliers ; and besides, 

 we are certain that we can henceforth obtain the crystallisa- 

 tion of a substance in a much more perfect manner than in 

 all tlie experiments hitherto made. 



