318 Intelligence and Miscellaneous Articles. 



furnished with minerals, such as those of Saint Gothard, Sweden, 

 Finland, or the United States, we find that the introduction of chlo- 

 rides, mixed with fluorides and sometimes with sulphuretted com- 

 pounds, accounts for the formation of their most characteristic 

 minerals. In this explanation we must include the rich deposits of 

 red oxide of zinc with franklinite in New Jersey, as well as different 

 masses of specular iron, and protoxide of iron, which have also been 

 produced in the limestone. 



We find magnesian compounds, such as spinel, chondrodite, mica, 

 pyroxene, amphibole, warwickite, and serpentine accumulated, some- 

 times in large quantities, in limestones which contain no magnesia. 

 This fact, hitherto unexplained, would only be a consequence of the 

 different chemical affinities of lime and magnesia, for throughout our 

 experiments we see chloride of magnesium precipitated by lime ; 

 and when these bases are both in the presence of chloride of silicium 

 or aluminium, the lime yields ite oxygen, and the magnesia remain- 

 ing in the state of oxide enters by preference into an oxidized com- 

 bination with the regenerated silica or alumina. The same principle 

 explains the presence of magnesia, with exclusion of lime, in the 

 protoxidized irons. Are we to attribute to the same cause the prepon- 

 derance of magnesia over lime in the elements of granite and in ser- 

 pentine ? 



The mode in which quartz and the silicates are connected princi- 

 pally with the granite rocks, has long been a difficulty in all the 

 hypotheses upon the formation of the strata called primitive. Now 

 we find in our experiments that quartz crystallizes at the same time, 

 or even later than the silicates, at a temperature scarcely exceeding 

 a cherry-red heat, and consequently infinitely below its point of fusion. 



Is not this also the cause which appears sometimes to protect 

 quartz from the influence of the basic silicates or of the aluminates, 

 as in granite when it envelopes crystals of cymophane instead of 

 forming a double silicate, like emerald or euclase ? 



If mica still gives off, by heat, fluorides of silicium, boron or 

 lithium, shall we venture to affirm that the granitic pastes have not 

 also originally contained chlorides of silicium, boron or aluminium ? 

 These are deficient, it is true, amongst the vapours collected in the 

 present day in the neighbourhood of the apertures of volcanos, where 

 they are decomposed and precipitated by aqueous vapour in contact 

 with the atmosphere, but where they very probably contribute to the 

 formation of the silicates, already considered by the best observers 

 as products of volatilization. Do we not also find chlorine in con- 

 siderable quantity in certain masses, such as the zirconian syenite of 

 Norway, and the rock of Ilmen (miascite), where this body is prin- 

 cipally combined in elaolite, and where it appears to have carried 

 with it zirconium, tantalium, and all that series of rare elements 

 which belqng to these rocks ? 



It has never been shown that the presence of a certain quantity of 

 water is an obstacle to such reactions at a high temperature, since we 

 find silica and alumina separate, in an anhydrous state, from a watery 



