39 Elements of Mineralogy. 



and fome detached fragments of copper, 

 antimonial and arfenical ores. 



If we confider the immenfe quantity of 

 matter thrown up at different periods by vol- 

 canic mountains, without leffening their ap- 

 parent bulk *, we muft conclude the feat of 

 thefe fires to be feveral miles, perhaps him* 

 dreds of miles, below the level of the fea ; 

 and as iron makes from T to J of all thefe 

 ejections, we may infer that the interior parts 

 of the earth confift chiefly of this rcietal, its 

 ores, or ftones that contain it, whofe greater 

 or lefler dephlogiftication in different parts 

 may be the caufe of the variation of mag- 

 netic direction. 



The origin of thefe fires is not eafily ac- 

 counted for. It is well known that mar- 

 tial pyrites, being moiflened, will acquire 

 heat j but that this heat fhould burn: into 

 actual flame, the concourfe of open air is ab- 

 folutely requifite ; however, if we fuppofe 

 the heated pyrites to have been in contact 

 with black wad and petrol, we may fappofe 

 the flame to arife, as we fee it produced by 

 art from the deficcation of that fubftance, and 

 its mixture with the mineral oil, Thaj; 

 ore when heated affords dephlogifticated air, 



* Mr. Gerhard computes that Vefwlus has ejected from 

 the year 70 to the year 1783, 309658161 cubic feet. 2 Mi" 

 ntral. Gefcb, .87. 



Of 



