to the last named. This is a drumlin, a 66 name now used to designate the class of glacial accumulations which Professor Hitchcock originally called 'lenticular hills,"" Dr. Wright, whose words we have just quoted, states that they abound in the neighbourhood of Boston, and largely give character to the scenery of the three north-eastern counties of

1

Massachusetts. They have been noticed in Scotland by Professor J. Geikie, and in Ireland by Mr. Maxwell Close.

According to a description given by Mr. Warren Upham, "these hills vary in size from a few hundred feet to a mile in length, with usually half to two-thirds as great width. Their height, corresponding to their area, varies from 25 to 200 1 "The Ice Age in North America," chap. xi.

feet. But whatever may be their size and height, they are singularly alike in outline and form, usually having steep sides, with gently sloping, rounded tops, and presenting a very smooth and regular contour. . . . The trend, or direction of the longer axis, of these lenticular hills is nearly the same for all of them comprised within any limited area, and is approximately like the course of the striæ or glacial furrows marked upon the neighbouring ledges." According to Dr. Wright, these hills resemble in structure the lower portions of the till. They are only imperfectly if at all stratified, and are formed of very compact clay filled with foreign and finely striated stones. They occur at various altitudes above sea-level in America, even up to 1500 feet on the tract between the Merrimack and Connecticut Rivers.

Drumlins are attributed to the action of ice by most geologists who have discussed their origin, though some think them to have been shaped by the sea. By the majority, however, they are believed to have been formed above water. Some suppose

them to mark the site of ancient "moulins," on a scale adequate to the vanished ice-sheet, when so much débris was carried down by the plunging water that it accumulated in huge heaps at the bottom, which afterwards, when the course of the stream had been diverted by the opening of new crevasses, were modified in shape by the pressure and sculpture of the moving ice. Some regard these drumlins as

morainic masses, which, when the ice-sheet has subsequently advanced, have met with a somewhat similar treatment, so that they represent the original contour of the till after the first ice-sheet had melted away. Others again compare drumlins with the sandbanks in rivers, and look upon them as masses of unstrátified drift, slowly and locally accumulated under the irregularly moving ice-sheet, in places where more material was brought than could be carried away.

To the objection, that these hypotheses are incompatible with the erosive action generally attributed to ice, their advocates reply that this action was confined to mountain regions, where the valleys were steep, or to broken hilly tracts, where the ice must have flowed with comparative rapidity, but irregularly; while in the open lowlands and broad valleys, where it would advance with diminished but more equable motion, it would be powerless either to abrade or to erode, and thus accumulation might take place beneath it. This principle is not improbable, but if it be adopted, it will be found to create difficulties in regard to the hypothesis which attributes the larger lake basins to the excavatory action of ice; for these often lie in valleys where the slope of the ground is slight, and in districts where large masses of till have accumulated.1

1 For instance, according to Dr. Wallace (Fortnightly Review), the ice only begins to excavate at the foot of the mountains; that is to say, where the slope becomes comparatively gentle.

CHAPTER II

ICE-WORK IN GREAT BRITAIN AND IRELAND— THE DEPOSITS AND THEIR SIGNIFICANCE

OVER no small part of the lowlands in Britain, as far south as the northern margin of the Thames valley, deposits occur which are generally admitted to be indicative of a temperature considerably lower than at the present time, even if they are not the direct products of land-ice. Somewhat similar deposits reappear in a limited area on the Sussex coast. Contemporary beds very likely occur in the intervening districts; but as they nowhere exhibit characteristics distinctly glacial, all correlation can be only conjectural. Some traces of ice action probably remain in the south-west of England, but here also the evidence is less conclusive than in the central and northern regions. Among the glacial deposits of the lowlands, the most characteristic is a boulder clay. This contains numerous pieces of rock, ranging in form from rounded to angular, but as a rule more or less subangular, embedded in clay, sometimes loamy or sandy, but generally stiff and tenacious. Those of smaller size commonly are more or less worn, and are often