6 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [NoV. 5, 



neighbourhood the rocks are so mineralized as to lose all distinctive 

 characters. If the slate- rocks near Bodmin belong to a very old part 

 of the Devonian system, it is no doubt possible that the coarse 

 quartzites of Pydar Downs may be still older : but I did not visit 

 them during the past summer, as I learnt from my friend Mr. Whit- 

 ley, of Truro, that he had examined them from end to end, without 

 discovering a single fossil ; and I still believe them to be Devonian. 



On the whole, therefore, after the excursions of the past summer, 

 I retain the same opinion of the age and grouping of the slate-rocks of 

 Cornwall that I had in the year 1836, with the very limited exception 

 of the two headlands of the south coast, called Nare Head and the 

 Dodman. But if such were the facts, how came it to pass that the 

 Devonian system was not at once established by Sir R. I. Murchison 

 and myself in the year 1836, when we determined the great culm- 

 trough of North Devon to be Carboniferous ? The reason is well 

 known. We sent a good series of the fossils of the Petherwin and 

 Barnstaple Groups to London. They were examined and named, 

 and every species was called Silurian. How this mistake originated 

 I do not now inquire, but I suspect that our package of fossils had 

 been misplaced, or mixed with fossils of an older age. However this 

 may be, the mistake inflicted on me the labour of a considerable part 

 of two summers. During a portion of the summers of 1 837 and 1838 

 (in the hopes of clearing away this difficulty, and never for a single^ 

 moment suspecting any great mistake in the determination of the 

 fossils of these groups) I traversed many parts of Devon and 

 some parts of Cornwall again and again, seeking for faults where 

 they were not to be found, and for anticlinal and synclinal lines where 

 nature had never formed them ; and at the end of the summer of 

 1838, I returned with the conviction that the first section of Devon- 

 shire, made by Sir R. T. Murchison and myself conjointly, was essen- 

 tially right, and that Cornwall bore the same relation to Devon which 

 I had supposed on my return from that county in 1836. Were 

 then the sections and fossils to continue in a permanent contradiction 

 to one another 1 If so, the Devonian System would never have been 

 heard of ; but on re-examining the fossils in 1 838, it turned out that all 

 the species of the Barnstaple Group had been wrongly named ; and 

 that so far from being Silurian, the only doubt respecting them was, 

 whether they might not be called Carboniferous rather than Devonian. 

 Thus the physical, and fossil evidence were brought into harmony : 

 and I may add, from this example, that no good classification either 

 of subdivisions or systems, or of subordinate formations, ever can be 

 attempted without a previous determination of the physical groups. 

 The study of fossils, based on ascertained physical groups, may pro- 

 duce, and often does produce, some modification of our lines of de- 

 marcation ; but the evidence of sections must ever remain as the 

 primary basis of geology. When a system has been well made out, 

 and its groups of fossils determined, we may then make use of com- 

 parative groups of fossils freely, and with very small risk of mistake. 

 But to begin vdth fossils, before the physical groups are determined, 

 and through them to establish the nomenclature of a system, would 



