308 OEIGIN OF CRYSTALLINE EOCKS. [XIII. 



hydrous carbonate slowly separates ; second, by the action of 

 bicarbonate of lime in solution, which, with sulphate of mag- 

 nesia, gives rise to gypsum ; this first crystallizes out, leaving 

 behind a much more soluble bicarbonate of magnesia, which 

 deposits the hydrous carbonate in its turn. In this way, for 

 the first time, in 1859, the origin of gypsums and their inti- 

 mate relation with magnesian limestones were explained.* 



It was, moreover, shown that, to the perfect operation of this 

 reaction, an excess of carbonic acid in the solution during the 

 evaporation was necessary to prevent the decomposing action 

 of the hydrous mono-carbonate of magnesia upon the already 

 formed gypsum. Having found that a prolonged exposure to 

 the air, by permitting the loss of carbonic acid, partially inter- 

 fered with the process, I was led to repeat the experiment in a 

 confined atmosphere, charged with carbonic acid, but rendered 

 drying by the presence of a layer of desiccated chloride of cal- 

 cium. As had been foreseen, the process under these conditions 

 proceeded uninterruptedly, pure gypsum first crystallizing out 

 from the liquid, and, subsequently, the hydrous magnesian 

 carbonate. t This experiment is instructive, as showing the 

 results which must have attended this process in past ages, 

 when the quantity of carbonic acid in the atmosphere greatly 

 exceeded its present amount. (Ante, pages 43, 47, and 91.) 



As regards the hypotheses put forward to explain the supposed 

 dolomitization of previously formed limestones by an epigenic 

 process, I may remark that I repeated very many times, under 

 varying conditions, the often-cited experiment of Von Morlot, 

 who claimed to have generated dolomite by the action of sul- 

 phate of magnesia on carbonate of lime, in the presence of water 

 at a somewhat elevated temperature under pressure. I showed 

 that what he regarded as dolomite was not such, but an admixt- 

 ure of carbonate of lime with anhydrous and sparingly soluble 

 carbonate of magnesia ; the conditions in which the carbonate 

 of magnesia is liberated in this reaction not being favorable to 

 its union with the carbonate of lime to form the double salt 



* See the recent conclusions of Ramsay, noticed ante, page 92. 

 f Canadian Naturalist, new series. 



