THEORY of the EARTH. 25-5 



forming fome part in the ftruclure of this earth, deferves to be 

 confidered. 



THE nature of granite, as a part of the ftruclure of the 

 earth, is too intricate a fubjedl to be here confidered, where we 

 only feek to prove the fufion of a fubftance from the evident 

 marks which are to be obferved in a body. We fhall, therefore, 

 only now confider one particular fpecies of granite ; and if this 

 mall appear to have been in a fluid ftate of fufion, we may b'e 

 allowed to extend this property to all the kind. 



THE fpecies now to be examined comes from the north coun- 

 try, about four or five miles weft from Portfoy, on the road 

 to Huntly. I have not been upon the fpot, but am informed 

 that this rock is immediately connected or continuous with the 

 common granite of the country. This indeed appears in the 

 fpecimens which I have got ; for, in fome of thefe, there is to 

 be perceived a gradation from the regular to the irregular fort. 



THIS rock may indeed be confidered, in fome refpects, as a 

 porphyry ; for it has an evident ground, which is feld-fpar, in 

 its fparry ftate ; and it is, in one view, diftinctly maculated with 

 quartz, which is tranfparent, but fomewhat dark-coloured *. 



CONSIDERED as a porphyry, this fpecimen is no lefs fingular 

 than as a granite. For, inftead of a filiceous ground, maculated 

 with the rhombic feld-fpar, which is the common ftate of por- 

 phyry, the ground is uniformly cryftallized, or a homogeneous 

 regular feld-fpar, maculated with the tranfparent filiceous fub- 

 ftance. But as, befides the feld-fpar and quartz, which are the 

 eonftituent parts of the ftone, there is alfo mica, in fome 

 places, it may, with propriety, be termed a granite. 



THE fingularity of this fpecimen cdnfifts, not in the nature 

 or proportions of its eonftituent parts, but in the uniformity 

 of the fparry groxmd, and the regular fhape of the quartz mix- 

 ture. This filiceous fubftance, viewed in one direction, or 

 longitudinally, may be confidered as columnar, prifmatical, 



oc 



* Plate II, fig. i. 2. 3. 



