128 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [Jan. 21, 



At the time, this surprised me greatly ; but the fact is quite in 

 harmony with the statements of Sir H. De la Beche. All the sec- 

 tions of the Silurian system south of the Severn are made on the 

 hypothesis that the mountains on the north-west side of the lines of 

 section are older than the Silurian rocks. Now, as a general rule, 

 this hypothesis is not correct, and it in some measure vitiates the 

 base-lines of the several sections, and so destroys a part of their 

 meaning. But the section at the east side of the Berwyns, ending 

 with Craig y Glyn, is appealed to by Mr. Murchison as proving the 

 low position of the Asaphus Buchii and the Llandeilo flag. For 

 several years I myself put this interpretation upon the section in 

 question. But since a doubt has arisen about the age of the Llan- 

 deilo flag, Mr. Salter has re-examined the specimens both from 

 Craig y Glyn on the east flank of the Berwyns, and also from one or 

 two localites near Grat Arrenig and in the Rhiulas limestone; and he 

 now retracts his identification of any of the fragments with Asaphus 

 Buchii. The only fragments he can identify belong to Asaphus 

 tyrannus, which certainly has a very low range among the fossili- 

 ferous slates. The evidence therefore supplied by my sections in 

 North Wales gives us no help in determining the age of the Llan- 

 deilo flag. Mr. Salter also examined during the past summer one 

 or two sections of the Silurian series of South Wales. I cannot give 

 his remarks in detail ; but I may state the result of them. They go 

 to prove either that the Llandeilo flag is not inferior to the Caradoc 

 sandstone {e.g. at Builth), or that it is associated with the upper 

 part of it, a part containing several fossils of the Wenlock shale or 

 limestone. Hence, coupling these remarks with what has been stated 

 by Sir H. De la Beche, I should class the Caradoc sandstone and 

 Llandeilo flags of South Wales, the Caradoc sandstone of the Mai- 

 verns, that on the south-east side of the Berwyns, and lastly, the 

 fossiliferous bands of Glyn Ceiriog and Mathyrafal, all in one group, 

 and compare it with that of the Coniston limestone ; perhaps in- 

 cluding with the Coniston limestone also the Coniston flags*. If 

 this view be correct, we cease to be surprised at finding the Llan- 

 deilo flags among the great folds and undulations of the Upper Si- 

 lurian rocks of South Wales. The Calymene Blumenhachii ranges 

 from the Ludlow and Wenlock limestone down to the rocks under 

 the Caradoc sandstone, &c. The Trinucleus Caractaci is a most 

 abundant Caradoc sand fossil, yet it ranges into the Wenlock shales 

 under Wenlock Edge ; and were I to seek for the Asaphus Buchii 



* Of 45 species of Trilobites and shells found in the Llandeilo flags — 



3 are Wenlock species exclusively. 



10 common to Wenlock, and the Glyn Ceiriog and Mathyrafal series. 



10 Mathyrafal and Glyn Ceiriog. 



1 Coniston only. 



1 Irish Wenlock — Kerry. 



5 found in the Caradoc. 



30, leaving 15 as peculiar to it. 

 Only 9 Coniston species are contained in the Llandeilo series. The additional 

 species were added by myself (39 is SirR.M.'s number) at Builth last year. — J.W.S. 



