or cliff, against which the limestones and their associated strata were laid down.1 Lastly, can we gather any evidence regarding the northern limits of the Carboniferous sea and the probability of continental land existing in that direction? There is, in the first place, good reason to believe that land existed outside the north-west portion of Ireland: in Galway and Mayo the Carboniferous limestones are everywhere bordered by conglomerates, and it is stated that these are on the horizon of the Upper Limestones; that near Oughterard the conglomerates and sandstones graduate eastward through shales into limestones along the line of strike, and that in other localities the limestones themselves contain pebbles up to the size of a bean. The northern part of Donegal seems to have formed part of this land, for the Carboniferous rocks occurring in South Donegal and North Tyrone seem to pass northward into shales and sandstones. On the west coast of Scotland, in the district of Morvern (Argyleshire), there is a small tract of Carboniferous sandstone let down by faults among the older rocks. This is believed to be of Coal-measure age, and was probably laid down in a gulf which penetrated the northern mass of land between Donegal and the Highlands of Scotland; we may suppose, therefore, that a similar gulf existed in Lower Carboniferous times, but a smaller and less extensive one.2 There is every reason to suppose that the Highlands were land throughout the Carboniferous period. I am informed by Mr. H. M. Cadell that the oil-shales of the Lothians are mostly replaced by grits and sandstones in the north of Fife, and that everything indicates an approach to land in that direction. Indeed, the persistent recurrence 1 "Mem. Geol. Surv. Scotland, Expl. of Sheet 24," p. 17. 2 By mistake it has been carried too far north in the map, Plate IV. of shallow and fresh-water conditions throughout the Scottish Carboniferous series proves the neighbourhood of land, and leads to the conclusion that the area of the Devonian Lake Caledonia was converted at this period into a land-locked gulf which stretched north-eastward into continental land, and was only connected with the more open sea by narrow channels between the mainland and the islands above indicated. We may reasonably suppose that many rivers emptied themselves into this gulf, especially at its north-eastern end, and the nature of the deposits indicates that the subsidence was at times more than counterbalanced by the amount of material brought down by the rivers, so that the eastern part of the gulf was sometimes silted up and converted into tracts of low swampy land, enclosing large sheets of water, which were sometimes fresh, sometimes brackish, and only occasionally invaded by the sea. 2. Geographical Restoration. The evidence for most of the coast-lines delineated on Plate IV. has been amply discussed, and it only remains to show reason for the lines to the east and south of England, and to give some account of the conjectured extent of the continent which lay to the north of the British Carboniferous sea. The entire absence of Carboniferous rocks over the whole of the Scandinavian peninsula, except the extreme south of Sweden, renders it highly probable that this area formed part of the northern continent, and was united to the Scottish Highlands; the southern border of this land seems to have crossed the centre of Denmark, and a prolongation of this line would strike the coast of Yorkshire. Professor Hull suggests that it trended south-westward and joined that of the land which lay over the midland counties of England, but there is really no evidence for this, and I think it is more likely to have had an outline such as that shown on the map, broken into a series of bays and estuaries, like that in which the Scotch measures seem to have been formed. As there was evidently land to the north-west both of Scotland and Ireland, and as this was probably united to the Scoto-Scandinavian land, it is evident that the greater part of the North Atlantic must have been land at this time. When, moreover, we remember that no rocks of Carboniferous Limestone age have yet been found in any part of Iceland or Greenland, nor in any part of northern Canada, nor in the Arctic regions south of Grinnell Land (where it does occur), the facts seem to be greatly in favour of Professor Hull's view that at this period a large continent occupied the whole area of the North Atlantic, and extended from Finland on the east to the Rocky Mountains on the west. We cannot attempt to define its southern border across what is now the Atlantic Ocean; this, of course, must ever remain a matter of pure speculation, but it is puerile to make this a ground of objection to Professor Hull's hypothesis.' Those who have adopted the theory of the permanence of continents and of oceans throughout all geological time are naturally biassed against the existence of an Atlantic continent at any period, but those who think that continents may last long without being absolutely permanent see no reason why large parts of the Atlantic region should not have been land more than once in the course of geological time. Professor Green has published a representation of Lower Carboniferous geography which differs in some important particulars from Professor Hull's, and from the restoration I have attempted in Plate IV. There are two 1 66 Physical History of the British Isles," 1880, p. 37, and "Trans. Roy. Dub. Soc.," ser. ii. vol. iii. p. 305. |