THEORY of the EARTH. 247 



diflblved flate, that particular ftru&ure of the ftone, by means 

 of which the fpar had been admitted, muft appear at prefent 

 upon an accurate examination. 



THIS, however, is not the cafe, and we may reft the argu- 

 ment here. The fepta reach not the circumference j the furface 

 of the ftone is folid and uniform in every part j and there is 

 not any appearance of the fpar in the argillaceous bed around 

 the ftone. 



IT, therefore, neceflarily follows, that the contraction of the 

 iron-ftone, in order to form fepta, and the filling of thefe cavi- 

 ties with fpar, had proceeded pan paffu ; and that this opera- 

 tion muft have been brought about by means of fufion, or by 

 congelation from a ftate of fimple fluidity and expanfion. 



IT is only further to be obferved, that all the arguments 

 which have been already employed, concerning mineral con- 

 cretions from a {imply fluid ftate, or that of fufion, here take 

 place. I have feptaria of this kind, in which, befides pyrites, 

 iron-ore, calcareous fpar, and another that is ferruginous and 

 compound, there is contained filiceous cryftals j a cafe which is 

 not fo common. I have them alfo attended with circumftances 

 of concretion and cryftallization, which, befides being extreme- 

 ly rare, are equally curious and interefting. 



THERE is one fact more which is well worth our attention, 

 being one of thofe which are fo general in the mineral regions. 

 It is the cryftallizations which are found in clofe cavities of the 

 moft folid bodies. 



NOTHING is more common than this appearance. Cavities 

 are every where found clofely lined with cryftallizations, of 

 every different fubftance which may be fuppofed in thofe 

 places. Thefe concretions are well known to naturalifts, and 

 form part of the beautiful fpecimens which are preferved in the 

 cabinets of collectors, and which the German mineralifts have 

 termed Drufen. I {hall only particularize one fpecies, which 

 may be defcribed upon principle, and therefore may be a pro- 

 pen 



