244 AN ENQUIRY INTO THE [Ch. XII. 



The granitic masses of Cornwall, therefore, are found to 

 consist of alternating beds or layers of distinct rocks, after 

 the same manner as the primary slates ; and like these, also, 

 some beds are very compact and massive, others are tabular 

 and lamellar ; but the latter do not pass into a perfect schis- 

 tose structure, as in the granite of Arran. We have not 

 such frequent opportunities of tracing these beds longitu- 

 dinally as those of the slate groups ; and it may be conceded, 

 that the latter often run a longer course ; but, it must also 

 be admitted, that, in many instances, the granitic beds are 

 more persistent than some of the primary strata. 



But it may be objected, that Cornwall is but a small part 

 of the world ; and that it is, therefore, unfair to attach too 

 much importance to a single example, which may prove an 

 exception to the general rule. This argument is not started 

 merely for the sake of extending the discussion, for it has 

 been actually advanced by one who deservedly holds a dis- 

 tinguished rank among geologists, accompanied by a re- 

 commendation to visit other granitic countries. To argue 

 on too narrow a basis has certainly been the besetting sin of 

 our science; we have, therefore, no intention of disputing 

 the correctness of such advice : but, in reply to the sugges- 

 tion, that Cornwall may be an exception, we would venture 

 to express an opinion, that the works of Nature are so con- 

 sistent, and in such unity with each other, that they do not, 

 like human laws and human productions, admit of exceptions, 

 or similar contrivances for surmounting difficulties; and 

 when our systems require them, it is a proof of their imper- 

 fection. If, for instance, we find the granite of Cornwall, 

 however insignificant in extent when compared with that of 

 the whole world, to have a certain structure (disclosed by 

 advantages for examination, which no other known mass of 

 granite possesses), are we to conclude that this is an excep- 

 tion, because at variance with our previous knowledge? 

 Ought we not rather to suspect the correctness of our data, 

 to ascertain, in the first place, whether there be in fact any 



