Formation of Minerals in the Humid Way. 329 



formation peculiar to each class of metalliferous beds : we might even 

 hope, in thus tracing to their origin, step by step, in the same order 

 of systematic experiments, we may finally arrive at the crystallised 

 rocks which associate themselves to these beds by methods and phe- 

 nomena of continuity which it is impossible to mistake. 



4. On the Artificial Formation, by the Humid Method, of Corundum 

 and Diaspore. By M. H. de Sanarmont. 



I have recently presented to the Academy experiments upon the 

 artificial production, by the humid method, of many mineral species, 

 under the influence of heat combined with strong pressure, and I 

 have sought thus to explain the formation of a particular class of 

 metalliferous deposits produced principally by liquid agents. 



Other minerals belong to another class of repositories, wherein the 

 influence of gaseous agents appears to have predominated, and where 

 the water, which has been the chief agent in these phenomena must 

 have operated principally in the state of vapour. We must not al- 

 ways expect to find between these two kinds of formation a clearly- 

 defined line of demarcation. The substances which mineralise the 

 thermal springs, perhaps arise from great depths beneath, under the 

 form of volatile compounds, the liquid and gaseous agents may have 

 combined in very variable proportions, and the phenomena must 

 needs have exhibited themselves in many intermediate conditions be- 

 tween their two extreme limits. It is, besides, very difficult, even in 

 a chemical point of view, to conceive a very different mode of action 

 from watei* filling the same space in a liquid state at a very high tem- 

 perature, and that saturating it in the state of vapour enormously 

 compressed. il/^ 



Many mineral compounds then will arise, almost indifferently in 

 either of these circumstances, and I now desire to communicate the 

 first results of those experiments which I have undertaken on this 

 point, though they are still very incomplete. • i^ 



If we heat very strongly a hydrochloric solution existing 'iin> ail 

 oxide under the formula of R^ 0^, or R 0-, the acid even in excess 

 becomes free in solution, and the oxide is separated. The complete 

 precipitation, moreover, corresponds to a temperature which appears 

 to depend upon the state of dilution, and the excess of the acid. 



I have thus obtained anhydrous compounds, and sesquioxides <tf 

 iron and of chrome, the titanic and stanic acids, pulverulent and 

 amorphous, the latter only presenting some indications of crystal- 

 lisation. Alumina, on the contrary, crystallises, in suitable circum- 

 stances, anhydrous in the state oWcorundumy aiKl bydrated in the 

 state of dia$po7'e,A^o v.^lt^o'^-^ dyum « oJ niodJ ^^t^'idvibtM viiwvj>'Mi -i 



The corMn<iMmobtllined^i>y heating *n'acMl«okitionprofc«re4 from 

 hydrochloric of alumina, at a temperature which must equal that of 

 350 degrees, is a white crystallised sand, which scratches the emerald 



