214 Transactions of the Boyal 3Iicroseopical Society. 



presents the greatest number of cavities with these little cubic 

 crystals, whose faces are sometimes covered with parallel striations 

 answering to the cleavage ^ ( go oo ). It would be easy to prove 

 that these cavities were formed and filled with the substances now 

 found in them at the very moment of the crystallization of the 

 quartz. By raising the temperature to about 100° C. we did not 

 succeed in expanding the liquid ; so that it is not liquid carbonic 

 acid, but rather a saturated aqueous solution, as Sorby showed. 

 The little cubic crystals gave rise naturally to the idea that the 

 cavities are filled with a supersaturated solution of sodium chloride ; 

 their form and the parallel striae which cover their faces call at once 

 to mind the crystals of this same salt. 



Following the example of Zirkel, Vogelsang, and Behrens, we 

 investigated the nature of these microscopic crystals by spectral 

 analysis. We carefully removed from the grains of quartz ex- 

 tracted from the diorite all the feldspar which could possibly remain 

 attached to them. They were hardly put in the Bunsen flame 

 when they slightly decrepitated, the cavities broke open, and the 

 ray D appeared; this experiment repeated several times always 

 gave us the same result. However, in order to be more assured of 

 the exactness of our research, we wished to confirm it by an entirely 

 diSerent method. Some fragments of quartz reduced to a fine 

 powder were put in a test tube of distilled water ; when the grains 

 of quartz had subsided we poured in a few drops of silver nitrate, 

 the water became slightly milky, and presented the opalescent tint 

 which characterizes the silver chloride. 



Thus we think we can affirm that our experiments demonstrate 

 that these cubes are crystals of sodium chloride, and the hquid in 

 the cavities is a saturated aqueous solution of this salt. This result 

 is not astonishing, if we reflect upon the analogy existing between 

 the plutonic and volcanic rocks. The latter almost always, as is 

 well known, show traces of this salt, and often are impregnated 

 with it. We will now endeavour to find the temperature at which 

 this water was enclosed, and therefore that of the rock at the very 

 moment of its crystalhzation. We take as the ground of our calcula- 

 tion the experiments made on the solubility of sodium chloride in 

 water. It has been observed that the solubility of this salt in- 

 creases directly as the temperature. The cubic crystals contained 

 in the cavity having been deposited by the liquid while it was 

 cooHng. 



The micrometric measurements of the cavity, of which we have 

 just spoken, furnished the elements for our calculation. The 

 volume of the water was found 0,0000002198687 mm., that of 

 the salt 0,000000098003. We had then only to ascertain to 

 what temperature we should raise the volume of water to make 

 it dissolve this volume of salt. On calculation we obtained for our 



