balance of evidence is against the existence of any connecting strait, and in favour of the view that the Pennine range then formed a continuous and lofty chain of hills reaching from Derbyshire to the Scottish border, so that the western Dyassic lake was entirely isolated from the waters of the inland sea to the east.

The

Neither can I see that there is any strong evidence for the existence of the east and west barrier-ridge through Cheshire, which is supposed by Professor Hull to have divided the western area into two distinct lakes. differences between the Lancastrian and Salopian deposits are really unimportant, and are quite compatible with their having been formed in different parts of the same lake. The width of the ridge, as shown on Professor Hull's map,' is so small, that it is not likely to have been a permanent barrier, but it may have been a subaqueous ridge separating the lake into two basins, much as the Mediterranean is divided into two basins by the ridge between Sicily and Africa.

This lake then appears to have extended from Warwickshire and the Malvern Hills to the Firth of Clyde, a distance of 280 miles, with an extreme width of 100 miles, so that it was about the size of the modern Lake Huron. It spread over the counties of Warwick, Worcester, Stafford, Salop, Cheshire, and Lancashire, and over the eastern part of the Irish Sea. It encircled the Lake District, which must have risen as a rocky island out of its waters; it covered the valley of the Eden and Solway Firth, and arms of it ran up the valleys of the Nith and Annan, and probably for some distance up the Firth of Clyde, while westward a gulf extended into Ireland as far as Armagh and Dungannon in Tyrone. How far it reached westward and southward beyond Anglesey we have no means of

knowing, but there is no reason to suppose it extended far, for Wales and Ireland were doubtless connected by a tract of high mountainous land.

The slopes surrounding this lake seem in most places to have been steep, and the rivers running into it were consequently rapid, carrying down quantities of sand, and in some regions large stones and boulders, as in the breccias of the Midland counties. The size of the transported boulders in these breccias, their angularity, and the occasional striation of their surfaces, suggested to Professor Ramsay that they had been carried by floating ice. Many of the fragments can be identified with Welsh rocks, and it is highly probable that the Welsh mountains were then much more lofty than at present, and that snow may have accumulated on them in sufficient quantity to form glaciers. Some of these may have reached the level of the lake, and torrential streams bursting from others may have been equally active in carrying down the rock-fragments quarried from the frosty regions above.

On the northern borders of the lake volcanic forces came into play, and lava-flows with beds of volcanic ash were interbedded with the lacustrine sandstones (see p. 104).

From the thickness of the mechanical deposits in this north-western lake, and the rarity of magnesian limestones, we may infer that many streams and rivers ran into it, bringing a constant supply of fresh water and preventing the formation of chemical deposits. The northeastern lake, on the other hand, seems to have suffered from evaporation and concentration; possibly, also, the waters poured into it contained a larger proportion of salts in solution. At any rate, there are good reasons for regarding the Dyassic dolomites as direct chemical deposits, although the process of precipitation cannot be imitated in our laboratories.

The analogy between the conditions of the modern Cas

pian Sea and those which appear to have prevailed in the sea of the Magnesian Limestone has been pointed out by Sir A. Ramsay. Just as the Caspian is believed to have been originally connected with the Arctic Ocean, and as its fauna is really a marine assemblage, so also the inland sea which in Dyassic times stretched from England into Germany seems to have been isolated from the main oceans of the period, the introduction of new species and genera being thus prevented, so that the fauna was only a dwarfed and modified remnant of Carboniferous life.

Mr. E. Wilson suggests the following as the probable sequence of events in the north-eastern basin.1 After indicating the formation of the basement sands, he says"After a time the waters would become sufficiently saturated to cause dolomitic materials to be thrown down to some extent, which, commingled with the sand and mud, as also with the large supplies of ferrous carbonate likewise brought down by the rivers, would give rise to the bluecoloured plant-bearing dolomitic sandstones and shales of the Marl Slate series. During this stage, mechanical deposition predominated, on the whole, over chemical precipitation. . . . Somewhat suddenly (however) this state of things came to an end. Chemical precipitation now began to predominate, and the formation of the white and yellow Dolomites commenced." In the north this precipitation continued uninterruptedly, but the intercalations of shale and marl in the southern part of the area point to the inflowing of large rivers from the land which then existed over the east of England. It was clearly these southern

2

1 "The Permian Formation of N.E. England, Midland Nat.," vol. iv. 2 The change, as Mr. Wilson suggests, was probably due to progressive subsidence, which, by diminishing the altitude of the surrounding land, while increasing the area of evaporation, would tend to promote condensation of the lake-waters, and the consequent precipitation of mineral matter in solution.

rivers that brought down the argillaceous material of the Marl Slates and Middle Marls, derived doubtless from a wide surface of Coal-measures then undergoing destruction, while the streams from the west, being smaller but more rapid, may be credited with the introduction of the Yorkshire "quicksands."

I

CHAPTER VIII.

TRIASSIC PERIOD.

1. Stratigraphical Evidence.

HE lie of the Triassic rocks is different from that of

THE

any of the Paleozoic systems. In England their outcrop is nearly continuous from the southern to the northern coasts; in Gloucestershire it is very narrow, but it broadens out over the Midland counties, and stretches northward over tracts of considerable width on either side of the Pennine range. The one tract meets the sea in Durham, the other in Lancashire, but both must originally have extended much farther north, for detached areas of Trias occur in Cumberland and Dumfries, also in the north-east of Ireland, in the Inner Hebrides, and in the north-east of Scotland on the Moray Firth.

The Triassic strata are everywhere unconformable to the rocks on which they rest; they extend far and wide beyond the edges of the Dyassic beds, and run up many of our wider valleys as if the principal hill-ranges of England were then already in existence, as indeed they doubtless were. The Coal-measures and older Palæozoic rocks had been bent into troughs, basins, and ridges, and had suffered enormously from erosion and detrition before the Triassic beds were deposited upon them, so that in most parts of the country these beds rest upon a surface of erosion which had been previously formed across the tilted edges of the Paleozoic rocks.