ranged in parallel banks, mounds, and ridges, the longer axes of which coincide with the trend of glaciation. Over wide plains, on the other hand, it rises and falls in long, gentle swellings. This varying configuration is undoubtedly original-it is not the result of subsequent subaërial erosion. In mountain-valleys the ice-flow, subject to no deflection, must have proceeded continuously in one direction, and its groundmoraine, we may suppose, would thus tend to accrete more or less regularly. In the broader lowland tracts, however, as in the lower reaches of Nithsdale, Teviotdale, and Tweeddale, the same uniformity of conditions did not exist. Each of these broad depressions was occupied by mers de glace, formed by the confluence of ice-flows streaming out from various ice-sheds. Under such conditions the movement of the united currents could not be so equable, and in consequence of variations in the pressure of the ice, and in the lines of most rapid motion, the ground-moraine would tend to heap up in banks or ridges, the longer axes of which would necessarily coincide with the direction of ice-flow.'

The "drumlins" and "drums" of Ireland and Scotland appear to be represented in Sweden by certain banks of boulder-clay, which are described by De Geer as a novel kind of radical moraines. He recognises their strong resemblance to the drumlins of New England (Geol. Fören. Forh., 1895, p. 212). Drumlins occur in the Island of Rügen, but they would seem to be rare in North Germany. Recently Dr. K. Keilhack has observed them in Neumark (Jahrb., d. königl. preuss. geol. Landesanstalt für 1893, 1895, p. 190). They have been recognised also in the low grounds of Switzerland by Dr. Früh (Jahresbericht d. St. Gallischen Naturwissensch. Ges., 1894-95). It is probable, however, that the lenticular mounds and banks of till known under the name of drumlins have not all been formed in the same way. Thus the short lenticular

Once more, over the peripheral areas of the inland ice, as in the great plains of Germany, the influence exerted by the confluence of ice-flows just referred to would no longer be felt, at least to the same extent. When the ice had fairly escaped from uplands and hilly ground all minor movements would merge in one continuous broad outflow, the ground-moraine, as a result, being spread out more or less uniformly.

Looked at broadly, Northern Europe displays a central region of glacial erosion and a peripheral area of glacial accumulation. In the former, as in the Scandinavian peninsula, Finland, and the more elevated portions of the British Islands, bare rock is conspicuous over wide districts, while glacial accumulations, confined for the most part to hollows and depressions, attain as a rule no great thickness. Outside of such areas of special erosion, on the other hand, as in the low grounds of England and the plains of Northern Europe, naked rock appears only at intervals, while morainic materials and fluvio-glacial deposits reach their greatest development.

Under the ice-sheet rock-grinding and rock-shattering were carried on side by side. No doubt the boulder-clays frequently rest upon a smoothed and

drumlins of South Galloway appear to owe their origin to glacial erosion. They are the relics of the sheet of boulder-clay which accumulated under the last general mer de glace that overwhelmed Scotland. At a later stage the Southern Uplands supported local ice-sheets and large glaciers which, flowing out upon the adjacent low grounds, ploughed into and greatly denuded the old boulder-clay. The drumlins of this region are, in short, simply roches moutonnées, composed sometimes entirely of boulder-clay, at other times partly of boulder-clay and partly of solid rock.

striated surface, but just as frequently the groundrock is shattered, crushed, and jumbled, and the débris mixed up with the overlying till. Such phenomena are not confined to any particular area. Examples of finely smoothed and of jumbled rocksurfaces may often be seen in one and the same quarry or other opening. The latter, however, are best developed in places where the ground-rock tended to yield most readily to the pressure of ice. Massive crystalline rocks are perhaps oftener smoothed than shattered below till; but again and again their jointed structure has led to their ready disruption, boulder-clay has been squeezed into their crevices, and numerous blocks, some of large size, have been torn out and enclosed in the till. The result of this infraglacial disruption, however, is better seen in the case of bedded rocks, especially when the dip of the strata has happened to coincide with the direction of ice-flow. In such cases the boulder-clay has often been forced in between the bedding-planes, and broad ledges and reefs of rock have been wedged up and forced out of place. Not only so, but in the case of chalk and certain Tertiary formations, the pressure of the ice-sheet has not infrequently squeezed the rocks into folds and flexures of such a character that the disturbance and contortion have sometimes been attributed to subterranean action. Superficial curving, flexing, and displacement of the kind referred to are met with both in high and low-lying regions; but as the more yielding strata are best developed within

the latter, it is there that we meet with the most. striking evidence of infraglacial disruption and quarrying.

From the various facts above referred to we are justified in concluding that glacier-ice is a most effective agent of erosion. It not only abrades, rubs, smooths, and polishes, but crushes, folds, disrupts, and displaces rock-masses, the amount of disturbance being in proportion to the resisting power of the rocks and the pressure exerted by the ice. Other things being equal, more crushing and displacement will be effected under a massive ice-sheet than under a small valley-glacier. It is obvious, therefore, that during the prolonged existence of an ice-sheet, transport and accumulation must result in very considerable modifications of the surface. The central area of dispersion becomes gradually lowered by the abstraction of rock-débris which is carried forward and accumulated over the peripheral area occupied by the mer de glace. Hence it is that in the former region ground-moraines are seldom very thick, and usually consist of local materials. As they are followed outwards, however, they gradually attain a greater depth, and are more widely spread, the local materials becoming more and more mixed with far-travelled detritus, until eventually the latter begins to predominate. The depth attained by the ground-moraines in the plains of Europe is often great, individual sheets of boulder-clay often exceeding one hundred feet in thickness.

Such boulder-clays, however, are not the only evidence of glacial erosion. With them are frequently associated beds of gravel and sand and laminated clay, consisting exclusively of erratic materials. These are admittedly the products of infraglacial wateraction; the materials have been derived principally, if not exclusively, from the washing and sifting of infra- and intra-glacial detritus. Extensive beds of such aqueous accumulations underlie the groundmoraines in some places, and in other places separate one mass of ground-moraine from another. Great mounds, banks, and sheets of the same character, which obviously are similar in origin to the fluvioglacial detritus of the Alpine Vorländer, fringe the margins of the ground-moraines, and sweep over wide areas in North Germany and Russia. All these, therefore, must be taken account of if we would form an adequate conception of the amount of erosion effected by the mers de glace of the Ice Age.

The diluvial deposits of North Germany necessarily vary in thickness. Sometimes they are only a few feet, at other times they exceed 200 yards. Dr. Wahnschaffe has collected the results of numerous borings made in those regions, from which we learn that in East Prussia they range in thickness from 20 feet up to 490 feet, in West Prussia from 20 feet to 360 feet, in Posen from 35 feet to 240 feet, in Brandenburg from 30 feet to 670 feet, in Mecklenburg from 6 feet to 430 feet; in the province of Saxony a depth of 400 feet has been noted. Mr. Amund Hel