INTRODUCTION. 7 



At the base of the South Wales Coal-field, under the Farewell Rock — not the 

 Millstone Grit of this district, though the name Farewell Rock is used in this sense 

 by the Lancashire miners — is a marine band which also resembles in its fauna the 

 Redesdale Ironstone even more closely than the Pennystone of Coalbrookdale. 

 These beds are known as the " Rosser Veins." About 40 per cent, of the 

 Brachiopoda and Mollusca are common to the two deposits. 



The raolluscan fauna of the marine beds in the Canister series of the 

 Lancashire, Yorkshire, and North Midland Coal Measures differs markedly from 

 those of the marine bands of South Wales and Coalbrookdale, the characteristic 

 Canister fossils, Aviculopecten papyracoiis and Goniatites Listen being conspicuously 

 absent. The latter is stated by Salter to occur, however, in the South Wales bed 

 ('Mem. Geo). Surv.,' " Iron Ores of South Wales," p. 221, pi. i, fig. 36), but does 

 not seem to be present at Coalbrookdale. 



These interesting recurrences of a fauna characteristic of older rocks in beds 

 of a more recent date cannot but point to the fact that, during the time occupied 

 by the deposit of the intervening rocks, certain species must have survived in 

 some outside area, where a definite set of conditions favorable to their survival 

 obtained, or only altered so gradually that, as the shore rose or sank, those forms 

 for whose existence a certain depth of water was necessary would not feel the 

 change, and would be able to slowly migrate with the varying change of the level 

 of the floor of the sea; while others, more plastic, became modified into new 

 species by accumulated variations in successive generations. 



Such an alternation and repetition of conditions, bringing back with them 

 faunas which had once lived, died, and. been emtombed at the same place, but on a 

 lower level, renders the value of palseontological zones of small moment in the 

 matter of the correlation of the various beds of the same period in diS*erent 

 districts ; but, on the other hand, reasoning from the bathymetrical distribution 

 of the recent representatives of palaeozoic forms, strong evidence of the method 

 and conditions under which the f ossiliferous beds were deposited, may be deduced. 

 The Carboniferous rocks of Great Britain can have been at no time laid down at 

 any great distance from land. The older rocks of Devon and Cornwall on the 

 south. Middle and North Wales, and the Lake District, and the high lands of 

 Galloway on the west, and due north the Highlands of Scotland, were dry land 

 during the Carboniferous epoch. That these points of high land could not have 

 been isolated islands in the Carboniferous sea is evident from the littoral deposits 

 of slates and sandstones of Devonshire in the south, and the Scoto-Northumbrian 

 area in the north, so well developed along the line of the receding or advancing 

 shores, which must have been the result of surface denudation from large 

 land-areas, and this condition of things would necessitate a large and complicated 

 system of river-drainage. While in the main the general movement during 



