var'iuus parts of Scotland, 443 



which happens to suit any particular hypothesis which we may wish 

 to support. The accurate description of mineralogical species, must 

 be the base of all geological reasonings, but if v/e intermix charac- 

 ters derived from geological circumstances with true mineralogical 

 characters, we set out upon z.petitio prlnc'ipil^ and end by reasoning in 

 a circle. In the case of limestone, the chemical properties and well 

 known popular characters of which have given it a very decided and 

 constant name, we act rightly. The name of the species is retained, 

 be its geological situation what it may, and its geological accidents 

 are distinguished by the addition of that name which suits its posi- 

 tion, whether primitive, transition, or floetz. But on the other 

 hand, a mineral equally common, whose characters too are suf- 

 ficiently familiar, is distinguished by two names, although 

 even more identical in structure than the different tribes of 

 limestone, and it is called clay slate, or graywacke slate, for no 

 other reason than that in the one case it is associated with the rocks 

 called primitive, and in the other with those which go by the name of 

 transition. It Is obviously necessary that the same practice should be 

 adopted in this case as in the former, and that, adhering to the name 

 of the species, we should distinguish its geological position, if re- 

 quired, by the superaddition of the corresponding term. We shall 

 thus have primitive clay slate, and transition clay slate, if we find It 

 necessary to retain these geological distinctions. 



It is almost superfluous to quote instances of the existence of genuine 

 clay slate occurring together with well characterized graywacke slate, 

 as they may be seen in all the slate counties of England, and among 

 others remarkably In Cornwall. I trust that the geologists who 

 have attended to these rocks, will see, with me, the necessity of 

 adopting this distinction ; a distinction, without which all accuracy 



3 K 2 



