THE CONTINENTAL GLACIERS OF THE “ICE AGE”

Earlier cycles of glaciation.—Our study of the rocks composing the outermost shell of the lithosphere tells us that in at least three widely separated periods of its history the earth has passed through cycles of glaciation during which considerable portions of its surface have been submerged beneath continental glaciers. The latest of these occurred in the yesterday of geology and has often been referred to as the “ice age”, because until quite recently it was supposed to be the only one of which a record was preserved.

This latest ice age represents four complete cycles of glaciation, for it is believed that the continental ice developed and then completely disappeared during a period of mild climate before the next glacier had formed in its place, and that this alternation of climates was no less than three times repeated, making four cycles in all. At nearly or quite the same time ice masses developed in northern North America and in northern Europe, the embossments of the ice domes being located in Canada and in Scandinavia respectively (Fig. 322). There appears to have been at this time no extensive glaciation of the southern hemisphere, though in the next earlier of the known great periods of glaciation—the so-called Permo-Carboniferous—it was the southern hemisphere, and not the northern, that was affected (Fig. 323 and Fig. 304, p.276). From the still earlier glacial period our data are naturally much more meager, but it seems probable that it was characterized by glaciated areas within both the northern and the southern hemispheres.

Contrast of the glaciated and nonglaciated regions.—Since we have now studied in brief outline the characteristics of the existing continental glaciers, we are in a position to review the evidences of former glaciers, the records of which exist in their carvings, their gravings, and their deposits.

An observant person familiar with the aspects of Nature in both the northern and southern portions of the central and eastern United States must have noticed that the general courses of the Ohio and Missouri rivers define a somewhat marked common border of areas which in most respects are sharply contrasted (Fig. 324). Hardly less striking is the contrast between the glaciated and the nonglaciated regions upon the continent of Europe (Fig. 325).

It is the northern of the two areas which in each case reveals the characteristic evidences of glaciation, while there is entire absence of such marks to the southward of the common border. Within the American glaciated region there is, however, an area surrounded like an island, and within this district (Fig. 324) none of the marks characteristic of glaciation are to be found. This area is usually referred to as the “driftless area”, and occupies portions of the states of Wisconsin, Illinois, Minnesota, and Iowa. Even better than the area to the southward of the Ohio and Missouri rivers, it permits of a comparison of the nonglaciated with the drift-covered region.

The “driftless area.”—Within this district, then, we have preserved for our study a landscape which remains largely as it was before the several ice invasions had so profoundly transformed the general surface of the surrounding country. Speaking broadly, we may say that it represents an uplifted and in part dissected plain, which to the south and east particularly reveals the character of nearly mature river erosion (Fig. 177, p. 170). The rock surface is here everywhere mantled by decomposed and disintegrated rock residues of local origin. The soluble constituents of the rock, such as the carbonates, have been removed by the process of leaching, so that the clays no longer effervesce when treated with dilute mineral acid.

Wherever favored by joints and by an alternation of harder and softer rock layers, picturesque unstable erosion remnants or “chimneys” may stand out in relief (Fig. 326). Furthermore, the driftless area is throughout perfectly drained—it is without lakes or swamps—since all valleys are characterized throughout by forward grades. The side valleys enter the main valleys as do the branches a tree trunk; in other words, the drainage is described as arborescent. In so far as any portions of a plane surface now remain in the landscape, they are found at the highest levels (plate 16 A). The topography is thus the result of a partial removal by erosion of an upland and may be described as incised topography. Nowhere within the area are there found rock masses foreign to the region, but all mantle rock is the weathered product of the underlying ledges.

Characteristics of the glaciated regions.—The topography of the driftless area has been described as incised, because due to the partial destruction of an uplifted plain; and this surface is, moreover, perfectly drained. The characteristic topography of the “drift” areas is by contrast built up; that is to say, the features of the region instead of being carved out of a plain are the result of molding by the process of deposition (plate 16 B). In so far as a plane is recognizable, it is to be found not at the highest, but at the lowest level—a surface represented largely by swamps and lakes—and above this plain rise the characteristic rounded hills of various types which have been built up through deposition. The process by which this has been accomplished is one easy to comprehend. As it invaded the region, the glacier planed away beneath its marginal zone all weathered mantle rock and deposited the planings within the hollows of the surface (Fig. 327). The effect has been to flatten out the preËxisting irregularities of the surface, and to yield at first a gently undulating plain upon which are many undrained areas and a haphazard system of drainage (Fig. 328). All unstable erosion remnants, such as now are to be found within the driftless area, were the first to be toppled over by the invading glacier, and in their place there is left at best only rounded and polished “shoulders” of hard and unweathered rock—the well-known roches moutonnÉes.

The glacier gravings.—The tools with which the glacier works are never quite evenly edged, and instead of an in all respects perfect polish upon the rock pavement, there are left furrowings, gougings, and scratches. Of whatever sort, these scorings indicate the lines of ice movement and are thus indubitable records graven upon the rock floor. When mapped over wide areas, a most interesting picture is presented to our view, and one which supplements in an important way the studies of existing continental glaciers (Fig. 334, p.308, and Fig. 336, p. 312).

It has been customary to think of the glacier as everywhere eroding its bed, although the only warrant for assuming degradation by flow of the ice is restricted to the marginal zone, since here only is there an appreciable surface grade likely to induce flow. Both upon the advance and again during the retreat of a glacier, all parts of the area overridden must be subjected to this action. Heretofore pictured in the imagination as enlarged models of Alpine glaciers, the vast ice mantles were conceived to have spread out over the country as the result of a kind of viscous flow like that of molasses poured upon a flat surface in cold weather. The maximum thickness of the latest American glacier of the ice age has been assumed to have been perhaps 10,000 feet near the summit of its dome in central Labrador. From this point it was assumed that the ice traveled southward up the northern slope of the Laurentian divide in Canada, and thence to the Ohio river, a distance of over 1300 miles. If such a mantle of ice be represented in its natural proportions in vertical section, to cover the distance from center to margin we may use a line six inches in length, and only ¹/₁₀₀ of an inch thick. Upon a reduced scale these proportions are given in Fig. 329. Obviously the force of gravity acting within a viscous mass of such proportions would be incompetent to effect a transfer of material from the center to the periphery, even though the thickness should be doubled or trebled. Yet until the fixed glacial anticyclone above the glacier had been proven and its efficiency as a broom recognized, no other hypothesis than that of viscous flow had been offered in explanation. The inherited conception of a universal plucking and abrasion on the bed of the glacier is thus made untenable and can be accepted for the marginal portion only.

Not only do the rock scorings show the lines of ice movement, but the directions as well may often be read upon the rock. Wherever there are pronounced irregularities of surface still existing on the pavement, these are generally found to have gradual slopes upon the side from which the ice came, and relatively steep falls upon the lee or “pluck” side. If, however, we consider the irregularities of smaller size, the unsymmetrical slopes of these protruding portions of the floor are found to be reversed—it is the steep slope which faces the oncoming ice and the flatter slope which is upon the lee side. Such minor projections upon the floor usually have their origin in some harder nodule which deflects the abrading tools and causes them to pass, some on the one side and some upon the other. By this process a staple-shaped groove comes to surround the nodule, leaving an unsymmetrical elevated ridge within, which is steep upon the stoss side and slopes gently away to leeward.

Younger records over older—the glacier palimpsest.—Many important historical facts have been recovered from the largely effaced writing upon ancient palimpsests, or parchments upon which an earlier record has been intentionally erased to make room for another. In the gravings upon the glacier pavement, earlier records have been likewise in large part effaced by later, though in favorable localities the two may be read together. Thus, as an example, at the great limestone quarries of Sibley, in southeastern Michigan, the glaciated rock surface wherever stripped of its drift cover is a smoothly polished and relatively level floor with striÆ which are directed west-northwest. Beneath this general surface there are, however, a number of elliptical depressions which have their longer axes directed south-southwest, one being from twenty-five to thirty feet long and some ten feet in depth (Fig. 330). These boat-shaped depressions are clearly the remnants of an earlier more undulating surface which the latest glacier has in large part planed away, since the bottoms of the depressions are no less perfectly glaciated but have their striÆ directed in general near the longer axis of the troughs. Palimpsest-like there are here also the records of more than one graving.

The dispersion of the drift.—Long before the “ice age” had been conceived in the minds of Agassiz and his contemporaries, it had been remarked that scattered over the North German plain were rounded fragments of rock which could not possibly have been derived from their own neighborhood but which could be matched with the great masses of red granite in Sweden well known as the “Swedish granite.” Buckland, an English geologist, had in 1815 accounted for such “erratic” blocks of his own country, here of Scotch granite, by calling in the deluge of Noah; but in the late thirties of the nineteenth century, Sir Charles Lyell, with the results of English Arctic explorers in mind, claimed that such traveled blocks had been transported by icebergs emanating from the polar regions. A relic of Buckland’s earlier view we have in the word “diluvium” still occasionally used in Germany for glacier transported materials; while the term “drift” still remains in common use to recall Lyell’s iceberg hypothesis, even though the original meaning of the term has been abandoned. Drift is now a generic term and refers to all deposits directly or indirectly referable to the continental glaciers.

In general the place of derivation of the glacial drift may be said to be some point more distant from and within the former ice margin at the time when it was deposited; in other words, the dispersion of the drift was centrifugal with reference to the glacier.

Wherever rocks of unusual and therefore easily recognizable character can be shown to occur in place and with but limited areas, the dispersion of such material is easy to trace. The areas of red Swedish and Scotch granite have been used to follow out in a broad way the dispersion of drift over northern Europe. Within the region of the Great Lakes of North America are areas of limited size which are occupied by well marked rock types, so that the journeyings of their fragments with the continental glacier can be mapped with some care. Upon the northern shore of Georgian Bay occurs the beautiful jasper conglomerate, whose bright red pebbles in their white quartz field attract such general notice. At Ishpeming in the northern peninsula of Michigan is found the equally beautiful jaspilite composed of puckered alternating layers of black hematite and red jasper. On Keweenaw Peninsula, which protrudes into Lake Superior from its southern shore, is found that remarkable occurrence of native copper within a series of igneous rocks of varied types and colors. Fragments of this copper, some weighing several hundreds of pounds each and masked in a coat of green malachite, have under the name of “drift” or “float” copper been collected at many localities within a broad “fan” of dispersal extending almost to the very limits of glaciation (Fig. 331).

Some miles to the north of Providence in Rhode Island there is a hill known as Iron Hill composed in large part of black magnetite rock, the so-called Cumberlandite. From this hill as an apex there has been dispersed a great quantity of the rock distributed as a well marked “bowlder train” within which the size and the frequency of the dispersed bowlders is in inverse ratio to the distance from the parent ledge (Fig. 332). Similar though less perfect trains of bowlders are found on the lee side of most projecting masses of resistant rocks within the area of the drift.

Large bowlders when left upon a ledge of notably different appearance easily attract attention, and have been described as “perched bowlders.” Resting as they sometimes do upon a relatively small area, they may be nicely balanced and thus easily given a pendular or rocking motion. Such “rocking stones” are common enough, especially among the New England hills (plate 17 B). Many such bowlders have made somewhat remarkable peregrinations with many interruptions, having been carried first in one direction by an earlier glacier to be later transported in wholly different directions at the time of new ice invasions.

The diamonds of the drift.—Of considerable popular, even if not economic, interest are the diamonds which have been sown in the drift after long and interrupted journeyings with the ice from some unknown home far to the northward in the wilderness of Canada. The first stone to be discovered was taken by workmen from a well opening near the little town of Eagle in Wisconsin in the year 1876. Its nature not being known, it remained where it was found as a curiosity only, and it was not until 1883 that it was taken to Milwaukee and sold to a jeweler equally ignorant of its value, and for the merely nominal sum of one dollar. Later recognized as a diamond of the unusual weight of sixteen carats, it was sold to the Tiffanys and became the cause of a long litigation which did not end until the Supreme Court of Wisconsin had decided that the Milwaukee jeweler, and not the finder, was entitled to the price of the stone, since he had been ignorant of its value at the time of purchase.

An even larger diamond, of twenty-one carats weight, was found at Kohlsville, and smaller ones at Oregon, Saukville, Burlington, and Plum Creek in the state of Wisconsin; at Dowagiac in Michigan; at Milford in Ohio, and in Morgan and Brown counties in Indiana. The appearance of some of the larger stones in their natural size and shape may be seen in Fig. 333.

While the number of the diamonds sown in the drift is undoubtedly large, their dispersion is such that it is little likely they can be profitably recovered. The distribution of the localities at which stones have thus far been found is set forth upon Fig. 334. Obviously those that have been found are the ones of larger size, since these only attract attention. In 1893, when the finding of the Oregon stone drew attention to these denizens of the drift, the writer prophesied that other stones would occasionally be discovered under essentially the same conditions, and such discoveries are certain to continue in the future.

Tabulated comparison of the glaciated and nonglaciated regions.—It will now be profitable to sum up in parallel columns the contrasted peculiarities of the glaciated and the unglaciated regions.

Unglaciated Region	Glaciated Region
TOPOGRAPHY
The topography is destructional; the remnants of a plain are found at the highest levels or upon the hill tops; hills are carved of a high plain; unstable erosion remnants are characteristic.	The topography is constructional; the remnants of a plain are found at the lowest levels in lakes and swamps; hills are molded above a plain in characteristic forms; no unstable erosion remnants, but only rounded shoulders of rock.
DRAINAGE
The area is completely drained, and the drainage network is arborescent.	The area includes undrained areas,—lakes and swamps,—and the drainage system is haphazard.
ROCK MANTLE
The exposed rock is decomposed and disintegrated to a considerable depth; it is all of local derivation and hence of few types—homogeneous; the fragments are angular; soils are leached and hence do not contain carbonates.	No decomposed or disintegrated rock is “in place”, but only hard, fresh surface; loose rock material is all foreign and of many izes and types—heterogeneous; rock bowlders and pebbles are faceted and polished as well as striated, usually in several directions upon each facet; soils are rock flour—the grist of the glacial mill.
ROCK SURFACE
Rock surface is rough and irregular.	Rock surface is planed or grooved, and polished. Shows glacial striÆ.

Unassorted and assorted drift.—The drift is of two distinct types; namely, that deposited directly by the glacier, which is without stratification, or unassorted; and that deposited by water flowing either beneath or from the ice, and this like most fluid deposited material is assorted or stratified. The unassorted material is described as till, or sometimes as “bowlder clay”; the assorted is sand or gravel, sometimes with small included bowlders, and is described as kame gravel. To recall the parts which both the glacier and the streams have played in its deposition, all water-deposited materials in connection with glaciers are called fluvio-glacial.

Till is, then, characterized by a noteworthy lack of homogeneity, both as regards the size and the composition of its constituent parts. As many as twenty different rock types of varied textures and colors may sometimes be found in a single exposure of this material, and the entire gamut is run from the finest rock flour upon the one hand to bowlders whose diameter may be measured in feet (Fig. 335).

In contrast with those derived by ordinary stream action, the pebbles and bowlders of the till are faceted or “soled”, and usually show striations upon their faces. If a number of pebbles are examined, some at least are sure to be found with striations in more than one direction upon a single facet. As a criterion for the discrimination of the material this may be an important mark to be made use of to distinguish in special cases from rock fragments derived by brecciation and slickensiding and distributed by the torrents of arid and semiarid regions.

Inasmuch as the capacity of ice for handling large masses is greater than that of water, assorted drift is in general less coarse, and, as its name implies, it is also stratified. From ordinary stream gravels, the kame gravels are distinguished by the form of their pebbles, which are generally faceted and in some cases striated. In proportion, however, as the materials are much worked over by the water, the angles between pebble faces become rounded and the original shapes considerably masked.

Features into which the drift is molded.—Though the preËxisting valleys were first filled in by drift materials, thus reducing the accent of the relief, a continuation of the same process resulted in the superimposition of features of characteristic shapes upon the imperfectly evened surface of the earlier stages. These features belong to several different types, according as they were built up outside of, at and upon, or within the glacier margin. The extra-marginal deposits are described as outwash plains or aprons, or sometimes as valley trains; the marginal are either moraines or kames; while within the border were formed the till plain or ground moraine, and, locally also, the drumlin and the esker or os. These characteristic features are with few exceptions to be found only within the area covered by the latest of the ice invasions. For the earlier ones, so much time has now elapsed that the effect of weathering, wash, and stream erosion has been such that few of the features are recognizable.

Marginal and extra-marginal features are extended in the direction of the margin or, in other words, perpendicular to the local ice movement; while the intra-marginal deposits are as noteworthy for being perpendicular to the margin, or in correspondence with the direction of local ice movement. Each of these features possesses characteristic marks in its form, its size, proportions, surface molding and orientation, as well as in its constituent materials. It should perhaps be pointed out that the existing continental glaciers, being in high latitudes, work upon rock materials which have been subjected to different weathering processes from those characteristic of temperate latitudes. Moreover, the melting of the Pleistocene glaciers having taken place in relatively low latitudes, larger quantities of rock dÉbris were probably released from the ice during the time of definite climatic changes, and hence heavier drift accumulations have for both of these reasons resulted.

Marginal or “kettle” moraines.—Wherever for a protracted period the margin of the glacier was halted, considerable deposits of drift were built up at the ice margin. These accumulations form, however, not only about the margin, but upon the ice surface as well; in part due to materials collected from melting down of the surface, and in part by the upturning of ice layers near the margin (see ante, p. 277).

An important rÔle is played by the thaw water which emerges at the ice margin, especially within the reËntrants or recesses of the outline. The materials of moraines are, therefore, till with large local deposits of kame gravel, and these form in a series of ridges corresponding to the temporary positions of the ice front. Their width may range from a few rods to a few miles, their height may reach a hundred feet or more, and they stretch across the country for distances of hundreds or even thousands of miles, looped in arcs or scallops which are always convex outward and which meet in sharp cusps that in a general way point toward the embossment of the former glacier (Fig. 334, p. 308, and Fig. 336). These festoons of the moraines outline the ice lobes of the latest ice invasion, which in North America were centered over the depressions now occupied by the Laurentian lakes. There was, thus, a Lake Superior lobe, a Lake Michigan lobe, etc. With the aid of these moraine maps we may thus in imagination picture in broad lines the frontal contours of the earlier glaciers. At specially favorable localities where the ice front has crossed a deep valley at the edge of the Driftless Area, we may, even in a rough way, measure the slope of the ice face. Thus near Devils Lake in southern Wisconsin the terminal moraine crosses the former valley of the Wisconsin River, and in so doing has dropped a distance of about four hundred feet within the distance of a half mile or thereabouts (Fig. 337).

The characteristic surface of the marginal moraine is responsible for the name “kettle” moraine so generally applied to it. The “kettles” are roughly circular, undrained basins which lie among hummocks or knobs, so that the surface has often been referred to as “knob and basin” topography (plate 17 C).

Kames.—Within reËntrants or recesses of the ice margin the drift deposits were especially heavy, so that high hills of hummocky surface have been built up, which are described as kames. Most of the higher drift hills have this origin. They rarely have any principal extension along a single direction, but are composed in large part of assorted materials. In contrast with other portions of the morainal ridges they lack the prominent basins known as kettles. Other kames are high hills of assorted materials not in direct association with moraines and believed to have been built up beneath glacier wells or mills (p. 278).

Outwash plains.—Upon the outer margin of the moraine is generally to be found a plain of glacial “outwash” composed of sand or gravel deposited by the braided streams (Fig. 308, p. 280) flowing from the glacier margin. Such plains, while notably flat (Fig. 338), slope gently away from the moraine. Between the outwash plain and the moraine there is sometimes found a pit, or fosse (Fig. 309, p. 281), where a part of the ice front was in part buried in its own outwash (Fig. 339).

Pitted plains and interlobate moraines.—Where glacial outwash is concentrated within a long and narrow reËntrant, separating glacial lobes, strips of high plain are sometimes built up which overtop the other glacial deposits of the district. The sand and gravel which compose such plains have a surface which is pitted by numerous deep and more or less circular lakes, so that the term “pitted plain” has been applied to them. The surface of such a plain steadily rises toward its highest point in the angle between the ice lobes. Though consisting almost entirely of assorted materials, and built up largely without the ice margins, such gently sloping pitted platforms are described as interlobate moraines. Upon a topographic map the course of such an interlobate moraine may often be followed by the belts of small pit lakes (see Fig. 336).

Eskers.—Intra-morainal features, or those developed beneath the glacier but relatively near its margin, include the “serpentine kame”, esker, or, as it is called in Scandinavia, the os (plural osar) (Fig. 340). These diminutive ridges have a width seldom exceeding a few rods, and a height a few tens of feet at most, but with slightly sinuous undulations they may be followed for tens or even hundreds of miles in the general direction of the local ice movement (Fig. 341). They are composed of poorly stratified, thick-bedded sands, gravels, and “worked over” materials, and are believed to have been formed by subglacial rivers which flowed in tunnels beneath the ice. Inasmuch as the deposits were piled against the ice walls, the beds were disturbed at the sides when these walls disappeared, and the stratification, which was somewhat arched in the beginning, has been altered by sliding at both margins. As already stated, eskers have not a general distribution within the glaciated area, but are often found in great numbers at specially favored localities. Formed as they are beneath the ice, it is believed that many have their materials redistributed so soon as uncovered at the glacier margin, because of the vigorous drainage there. They are thus to be found only at those favored localities where for some reason border drainage is less active, or where the ice ended in a body of water.

Drumlins.—A peculiar type of small hill likewise found behind the marginal moraine in certain favored districts has the form of an inverted boat or canoe, the long axis of which is parallel to the direction of ice movement, as is that of the esker (Fig. 342). Unlike the esker, this type of hill is composed of till, and from being found in Ireland it is called a drumlin, the Irish word meaning a little hill (Fig. 343). Drumlins are usually found in groups more or less radial and not far behind the outermost moraine, to which their radiating axes are perpendicular. The manner of their formation is involved in some uncertainty, but it is clear that they have been formed beneath the margin of the glacier, and have been given their shape by the last glacier which occupied the district.

The mutual relationships of nearly all the molded features resulting from continental glaciation may be read from Fig. 344.

The shelf ice of the ice age.—Shelf ice, such as we have become familiar with in Antarctica as a marginal snow-ice terrace floating upon the sea, no doubt existed during the ice age above the Gulf of Maine (see Fig. 324, p. 298), and perhaps also over the deep sea to the westward of Scotland. Though the inland ice probably covered the North Sea, and upon the American side of the Atlantic the Long Island Sound, both these basins are so shallow that the ice must have rested upon the bottom, for neither is of sufficient depth to entirely submerge one of the higher European cathedrals.

Character profiles.—All surface features referable to continental glaciers, whether carved in rock or molded from loose materials, present gently flowing outlines which are convex upward (Fig. 345). The only definite features carved from rock are the roches moutonnÉes, with their flattened shoulders, while the hillocks upon moraines and kames, and the drumlins as well, approximate to the same profile. The esker in its cross sections is much the same, though its serpentine extension may offer some variety of curvature when viewed from higher levels.

Reading References for Chapter XXII

General:—

James Geikie. The Great Ice Age. 3d ed. London, 1894, pp. 850, maps 18.

Chamberlin and Salisbury. Geology, vol. 3, 1906, pp. 327-516.

Frank Leverett. The Illinois Glacial Lobe, Mon. 38, U. S. Geol. Surv., 1899, pp. 817, pls. 34; Glacial formations and Drainage Features of the Erie and Ohio Basins, Mon. 41, ibid., 1902, pp. 802, pls. 25; Comparison of North American and European Glacial Deposits, Zeit. f. Gletscherk., vol. 4, 1910, pp. 241-315, pls. 1-5.

Former glaciations previous to Ice Age:—

A. Strahan. The Glacial Phenomena of Paleozoic Age in the Varanger Fjord, Quart. Jour. Geol. Soc., London, vol. 53, 1897, pp. 137-146, pls. 8-10.

Bailey Willis and Eliot Blackwelder. Research in China, Pub. 54, Carnegie Inst. Washington, vol. 1, 1907, pp. 267-269, pls. 37-38.

A. P. Coleman. A Lower Huronian Ice Age, Am. Jour. Sci. (4), vol. 23, 1907, pp. 187-192.

W. M. Davis. Observations in South Africa, Bull. Geol. Soc. Am., vol. 17, 1906, pp. 377-450, pls. 47-54.

David White. Permo-Carboniferous Climatic Changes in South America, Jour. Geol., vol. 15, 1907, pp. 615-633.

Driftless and drift areas:—

T. C. Chamberlin and R. D. Salisbury. Preliminary Paper on the Driftless Areas of the Upper Mississippi Valley, 6th Ann. Rept. U. S. Geol. Surv., 1885, pp. 199-322, pls. 23-29.

R. D. Salisbury. The Drift, its Characteristics and Relationships, Jour. Geol., vol. 2, 1894, pp. 708-724, 837-851.

R. H. Whitbeck. Contrasts between the Glaciated and the Driftless Portions of Wisconsin, Bull. Geogr. Soc., Philadelphia, vol. 9, 1911, pp. 114-123.

Glacier gravings:—

T. C. Chamberlin. The Rock Scorings of the Great Ice Invasions, 7th Ann. Rept. U. S. Geol. Surv., 1888, pp. 147-248, pl. 8.

The dispersion of the drift:—

R. D. Salisbury. Notes on the Dispersion of Drift Copper, Trans. Wis. Acad. Sci., etc., vol. 6, 1886, pp. 42-50, pl.

N. S. Shaler. The Conditions of Erosion beneath Deep Glaciers, based upon a Study of the Bowlder Train from Iron Hill, Cumberland, Rhode Island, Bull. Mus. Comp. ZoÖl. Harv. Coll., vol. 16, No. 11, 1893, pp. 185-225, pls. 1-4 and map.

William H. Hobbs. The Diamond Field of the Great Lakes, Jour. Geol., vol. 7, 1899, pp. 375-388, pls. 2 (also Rept. Smithson. Inst., 1901, pp. 359-366, pls. 1-3).

Glacial features:—

T. C. Chamberlin. Preliminary Paper on the Terminal Moraine of the Second Glacial Epoch, 3d Ann. Rept. U. S. Geol. Surv., 1883, pp. 291-402, pls. 26-35.

G. H. Stone. Glacial Gravels of Maine and their Associated Deposits, Mon. 34, U. S. Geol. Surv., 1899, pp. 489, pls. 52.

W. C. Alden. The Delaven Lobe of the Lake Michigan Glacier of the Wisconsin Stage of Glaciation and Associated Phenomena. Prof. Pap. No. 34, U. S. Geol. Surv., 1904, pp. 106, pls. 15; The Drumlins of Southeastern Wisconsin, Bull. 273, U. S. Geol. Surv., 1905, pp. 46, pls. 9.

W. M. Davis. Structure and Origin of Glacial Sand Plains, Bull. Geol. Soc. Am., vol. 1, 1890, pp. 196-202, pl. 3; The Subglacial Origin of Certain Eskers, Proc. Bost. Soc. Nat. Hist., vol. 35, 1892, pp. 477-499.

F. P. Gulliver. The Newtonville Sand Plain, Jour. Geol., vol. 1, 1893, pp. 803-812.

---