Sir Charles Lyell

Difficulty of interpreting the phenomena of drift before the glacial hypothesis was adopted — Effects of intense cold in augmenting the quantity of alluvium — Analogy of erratics and scored rocks in North America and Europe — Bayfield on shells in drift of Canada — Great subsidence and re-elevation of land from the sea, required to account for glacial appearances — Why organic remains so rare in northern drift — Mastodon giganteus in United States — Many shells and some quadrupeds survived the glacial cold — Alps an independent centre of dispersion of erratics — Alpine blocks on the Jura — Whether transported by glaciers or floating ice — Recent transportation of erratics from the Andes to Chiloe — Meteorite in Asiatic drift.

It will appear from what was said in the last chapter of the marine shells characterizing the boulder formation, that nine-tenths or more of them belong to species still living. The superficial position of "the drift" is in perfect accordance with its imbedded organic remains, leading us to refer its origin to a modern period. If, then, we encounter so much difficulty in the interpretation of monuments relating to times so near our own—if in spite of their recent date they are involved in so much obscurity—the student may ask, not without reasonable alarm, how we can hope to decipher the records of remote ages.To remove from the mind as far as possible this natural feeling of discouragement, I shall endeavour in this chapter to prove that what seems most strikingly anomalous, in the "erratic formation," as some call it, is really the result of that glacial action which has already been alluded to. If so, it was to be expected that so long as the true origin of so singular a deposit remained undiscovered, erroneous theories and terms would be invented in the effort to solve the problem. These inventions would inevitably retard the reception of more correct views which a wider field of observation might afterwards suggest.

The term "diluvium" was for a time the popular name of the boulder formation, because it was referred by some geologists to the deluge. Others retained the name as expressive of their opinion that a series of diluvial waves raised by hurricanes and storms, or by earthquakes, or by the sudden upheaval of land from the bed of the sea, had swept over the continents, carrying with them vast masses of mud and heavy stones, and forcing these stones over rocky surfaces so as to polish and imprint upon them long furrows and striÆ.

But no explanation was offered why such agency should have been developed more energetically in modern times than at former periods of the earth's history, or why it should be displayed in its fullest intensity in northern latitudes; for it is important to insist on the fact, that the boulder formation is a northern phenomenon. Even the southern extension of the drift, or the large erratics found in the Alps and the surrounding lands, especially their occurrence round the highest parts of the chain, offers such an exception to the general rule as confirms the glacial hypothesis; for it shows that the transportation of stony fragments to great distances, and the striation, polishing, and grooving of solid floors of rock, are here again intimately connected with accumulations of perennial snow and ice.

That there is some intimate connection between a cold or northern climate and the various geological appearances now commonly called glacial, cannot be doubted by any one who has compared the countries bordering the Baltic with those surrounding the Mediterranean. The smoothing and striation of rocks, and the erratics, are traced from the sea-shore to the height of 3000 feet above the level of the Baltic, whereas such phenomena are wholly wanting in countries bordering the Mediterranean; and their absence is still more marked in the equatorial parts of Asia, Africa, and America; but when we cross the southern tropic, and reach Chili and Patagonia, we again encounter the boulder formation, between the latitude 41° S. and Cape Horn, with precisely the same characters which it assumes in Europe. The evidence as to climate derived from the organic remains of the drift is, as we have seen, in perfect harmony with the conclusions above alluded to, the former habits of the species of mollusca being accurately ascertainable, inasmuch as they belong to species still living, and known to have at present a wide range in northern seas.

But if we are correct in assuming that the northern hemisphere was considerably colder than now during the period under consideration, owing probably to the greater area and height of arctic lands, and to the quantity of icebergs which such a geographical state of things would generate, it may be well to reflect before we proceed farther on the entire modification which extreme cold would produce in the operation of those causes spoken of in the sixth chapter as most active in the formation of alluvium. A large part of the materials derived from the detritus of rocks, which in warm climates would go to form deltas, or would be regularly stratified by marine currents, would, under arctic influences, assume a superficial and alluvial character. Instead of mud being carried farther from a coast than sand, and sand farther out than pebbles,—instead of dense stratified masses being heaped up in limited areas,—nearly the whole materials, whether coarse or fine, would be conveyed by ice to equal distances, and huge fragments, which water alone could never move, would be borne for hundreds of miles without having their edges worn or fractured; and the earthy and stony masses, when melted out of the frozen rafts, would be scattered at random over the submarine bottom, whether on mountain tops or in low plains, with scarcely any relation to the inequalities of the ground, settling on the crests or ridges of hills in tranquil water as readily as in valleys and ravines. Occasionally, in those deep and uninhabited parts of the ocean, never reached by any but the finest sediment in a normal state of things, the bottom would become densely overspread by gravel, mud, and boulders.

In the Western Hemisphere, both in Canada and as far south as the 40th and even 38th parallel of latitude in the United States, we meet with a repetition of all the peculiarities which distinguish the European boulder formation. Fragments of rock have travelled for great distances from north to south; the surface of the subjacent rock is smoothed, striated, and fluted; unstratified mud or till containing boulders is associated with strata of loam, sand, and clay, usually devoid of fossils. Where shells are present, they are of species still living in northern seas, and half of them identical with those already enumerated as belonging to European drift 10 degrees of latitude farther north. The fauna also of the glacial epoch in North America is less rich in species than that now inhabiting the adjacent sea, whether in the Gulf of St. Lawrence, or off the shores of Maine, or in the Bay of Massachusetts. At the southern extremity of its course, moreover, it presents an analogy with the drift of the south of Ireland, by blending with a more southern fauna, as for example at Brooklyn near New York, in lat. 41° N., where, according to MM. Redfield and Desor, Venus mercenaria and other southern species of shells begin to occur as fossils in the drift.

The extension on the American continent of the range of erratics during the Pleistocene period to lower latitudes than they reached in Europe, agrees well with the present southward deflection of the isothermal lines, or rather the lines of equal winter temperature. Formerly, as now, a more extreme climate and a more abundant supply of floating ice prevailed on the western side of the Atlantic.Another resemblance between the distribution of the drift fossils in Europe and North America has yet to be pointed out. In Norway, Sweden, and Scotland, as in Canada and the United States, the marine shells are confined to very moderate elevations above the sea (between 100 and 700 feet), while the erratic blocks and the grooved and polished surfaces of rock extend to elevations of several thousand feet.

Fig. 118.

• K. Mr. Ryland's house.
• h. Clay and sand of higher grounds, with Saxicava, &c.
• g. Gravel with boulders.
• f. Mass of Saxicava rugosa, 12 feet thick.
• e. Sand and loam with Mya truncata, Scalaria Groenlandica, &c.
• d. Drift, with boulders of syenite, &c.
• c. Yellow sand.
• b. Laminated clay, 25 feet thick.
• A. Horizontal lower Silurian strata.
• B. Valley re-excavated.

I described in 1839 the fossil shells collected by Captain Bayfield from strata of drift at Beauport near Quebec, in lat. 47°, and drew from them the inference that they indicated a more northern climate, the shells agreeing in great part with those of Uddevalla in Sweden.[134-A] The shelly beds attain at Beauport and the neighbourhood a height of 200, 300, and sometimes 400 feet above the sea, and dispersed through some of them are large boulders of granite, which could not have been propelled by a violent current, because the accompanying fragile shells are almost all entire. They seem, therefore, said Captain Bayfield, writing in 1838, to have been dropped down from melting ice, like similar stones which are now annually deposited in the St. Lawrence.[134-B] I visited this locality in 1842, and made the annexed section, fig. 118., which will give an idea of the general position of the drift in Canada and the United States. I imagine that the whole of the valley B was once filled up with the beds b, c, d, e, f, which were deposited during a period of subsidence, and that subsequently the higher country (h) was submerged and overspread with drift. The partial re-excavation of B took place when this region was again uplifted above the sea to its present height. Among the twenty-three species of fossil shells collected by me from these beds at Beauport, all were of recent northern species, except one, which is unknown as living, and may be extinct (see fig. 119.). I also examined the same formation farther up the valley of the St. Lawrence, in the suburbs of Montreal, where some of the beds of loam are filled with great numbers of the Mytilus edulis, or our common European mussel, retaining both its valves and purple colour. This shelly deposit, containing Saxicava rugosa and other characteristic marine shells, also occurs at an elevated point on the mountain of Montreal, 450 feet above the level of the sea.[135-A]

In my account of Canada and the United States, published in 1845, I announced the conclusion to which I had then arrived, that to explain the position of the erratics and the polished surfaces of rocks, and their striÆ and flutings, we must assume first a gradual submergence of the land in North America, after it had acquired its present outline of hill and valley, cliff and ravine, and then its re-emergence from the ocean. When the land was slowly sinking, the sea which bordered it was covered with islands of floating ice coming from the north, which, as they grounded on the coast and on shoals, pushed along such loose materials of sand and pebbles as lay strewed over the bottom. By this force all angular and projecting points were broken off, and fragments of hard stone, frozen into the lower surface of the ice, had power to scoop out grooves in the subjacent solid rock. The sloping beach, as well as the floor of the ocean, might be polished and scored by this machinery; but no flood of water, however violent, or however great the quantity of detritus or size of the rocky fragments swept along by it, could produce such long, perfectly straight and parallel furrows, as are everywhere visible in the Niagara district, and generally in the region north of the 40th parallel of latitude.[135-B]

By the hypothesis of such a slow and gradual subsidence of the land we may account for the fact that almost everywhere in N. America and Northern Europe the boulder formation rests on a polished and furrowed surface of rock,—a fact by no means obliging us to imagine, as some think, that the polishing and grooving action was, as a whole, anterior in date to the transportation of the erratics. During the successive depression of high land, varying originally in height from 1000 to 3000 feet above the sea-level, every portion of the surface would be brought down by turns to the level of the ocean, so as to be converted first into a coast-line, and then into a shoal; and at length, after being well scored by the stranding upon it of thousands of icebergs, might be sunk to a depth of several hundred fathoms. By the constant depression of land, the coast would recede farther and farther from the successively formed zones of polished and striated rock, each outer zone becoming in its turn so deep under water as to be no longer grated upon by the heaviest icebergs. Such sunken areas would then simply serve as receptacles of mud, sand, and boulders dropped from melting ice, perhaps to a depth scarcely, if at all, inhabited by testacea and zoophytes. Meanwhile, during the formation of the unstratified and unfossiliferous mass in deeper water, the smoothing and furrowing of shoals and beaches is still going on elsewhere upon and near the coast in full activity. If at length the subsidence should cease, and the direction of the movement of the earth's crust be reversed, the sunken area covered with drift would be slowly reconverted into land. The boulder deposit, before emerging, would then for a time be brought within the action of the waves, tides, and currents, so that its upper portion, being partially disturbed, would have its materials re-arranged and stratified. Streams also flowing from the land would in some places throw down layers of sediment upon the till. In that case, the order of superposition will be, first and uppermost, sand, loam, and gravel occasionally fossiliferous; secondly, an unstratified and unfossiliferous mass, for the most part of much older date than the preceding, with angular erratics, or with boulders interspersed; and, thirdly, beneath the whole, a surface of polished and furrowed rock. Such a succession of events seems to have prevailed very widely on both sides of the Atlantic, the travelled blocks having been carried in general from the North Pole southwards, but mountain chains having in some cases served as independent centres of dispersion, of which the Alps present the most conspicuous example.

It is by no means rare to meet with boulders imbedded in drift which are worn flat on one or more of their sides, the surface being at the same time polished, furrowed, and striated. They may have been so shaped in a glacier before they reached the sea, or when they were fixed in the bottom of an iceberg as it ran aground. We learn from Mr. Charles Martins that the glaciers of Spitzbergen project from the coast into a sea between 100 and 400 feet deep; and that numbers of striated pebbles or blocks are there seen to disengage themselves from the overhanging masses of ice as they melt, so as to fall at once into deep water.[136-A]

That they should retain such markings when again upraised above the sea ought not to surprise us, when we remember that rippled sands, and the cracks in clay dried between high and low water, and the foot-tracks of animals and rain-drops impressed on mud, and other superficial markings, are all found fossil in rocks of various ages.

On the other hand, it is not difficult to account for the absence in many districts of striated and scored pebbles and boulders in glacial deposits, for they may have been exposed to the action of the waves on a coast while it was sinking beneath or rising above the sea. No shingle on an ordinary sea-beach exhibits such striÆ, and at a very short distance from the termination of a glacier every stone in the bed of the torrent which gushes out from the melting ice is found to have lost its glacial markings by being rolled for a distance even of a few hundred yards.

The usual dearth of fossil shells in glacial clays well fitted to preserve organic remains may, perhaps, be owing, as already hinted, to the absence of testacea in the deep sea, where the undisturbed accumulation of boulders melted out of very large bergs may take place. In the Ægean and other parts of the Mediterranean, the zero of animal life, according to Prof. E. Forbes, is approached at a depth of about 300 fathoms. In tropical seas it would descend farther down, just as vegetation ascends higher on the mountains of hot countries. Near the pole, on the other hand, the same zero would be reached much sooner both on the hills and in the sea. If the ocean was filled with floating bergs, and a low temperature prevailed in the northern hemisphere during the glacial period, even the shallow part of the sea might have been uninhabitable, or very thinly peopled with living beings. It may also be remarked that the melting of ice in some fiords in Norway freshens the water so as to destroy marine life, and famines have been caused in Iceland by the stranding of icebergs drifted from the Greenland coast, which have required several years to melt, and have not only injured the hay harvest by cooling the atmosphere, but have driven away the fish from the shore by chilling and freshening the sea.

If the cold of the glacial epoch came on slowly, if it was long before it reached its greatest intensity, and again if it abated gradually, we may expect to find the earliest and latest formed drift less barren of organic remains than that deposited during the coldest period. We may also expect that along the southern limits of the drift during the whole glacial epoch, there would be an intimate association of transported matter of northern origin with fossil-bearing sediment, whether marine or freshwater, belonging to more southern seas, rivers, and continents.

That in the United States, the Mastodon giganteus was very abundant after the drift period is evident from the fact that entire skeletons of this animal are met with in bogs and lacustrine deposits occupying hollows in the drift. They sometimes occur in the bottom even of small ponds recently drained by the agriculturist for the sake of the shell marl. I examined one of these spots at Geneseo in the state of New York, from which the bones, skull, and tusk of a Mastodon had been procured in the marl below a layer of black peaty earth, and ascertained that all the associated freshwater and land shells were of a species now common in the same district. They consisted of several species of Lymnea, of Planorbis bicarinatus, Physa heterostropha, &c.

In 1845 no less than six skeletons of the same species of Mastodon were found in Warren County, New Jersey, 6 feet below the surface, by a farmer who was digging out the rich mud from a small pond which he had drained. Five of these skeletons were lying together, and a large part of the bones crumbled to pieces as soon as they were exposed to the air. But nearly the whole of the other skeleton, which lay about 10 feet apart from the rest, was preserved entire, and proved the correctness of Cuvier's conjecture respecting this extinct animal, namely, that it had twenty ribs like the living elephant. From the clay in the interior within the ribs, just where the contents of the stomach might naturally have been looked for, seven bushels of vegetable matter were extracted. I submitted some of this matter to Mr. A. Henfrey of London for microscopic examination, and he informs me that it consists of pieces of small twigs of a coniferous tree of the Cypress family, probably the young shoots of the white cedar, Thuja occidentalis, still a native of North America, on which therefore we may conclude that this extinct Mastodon once fed.

Another specimen of the same quadruped, the most complete and probably the largest ever found, was exhumed in 1845 in the town of Newburg, New York, the length of the skeleton being 25 feet, and its height 12 feet. The anchylosing of the last two ribs on the right side afforded Dr. John C. Warren a true gauge for the space occupied by the intervertebrate substance, so as to enable him to form a correct estimate of the entire length. The tusks when discovered were 10 feet long, but a part only could be preserved. The large proportion of animal matter in the tusk, teeth, and bones of some of these fossil mammalia is truly astonishing. It amounts in some cases, as Dr. C. T. Jackson has ascertained by analysis, to 27 per cent., so that when all the earthy ingredients are removed by acids, the form of the bone remains as perfect, and the mass of animal matter is almost as firm, as in a recent bone subjected to similar treatment.

It would be rash, however, to infer from such data that these quadrupeds were mired in modern times, unless we use that term strictly in a geological sense. I have shown that there is a fluviatile deposit in the valley of the Niagara, containing shells of the genera Melania, Lymnea, Planorbis, Valvata, Cyclas, Unio, and Helix, &c., all of recent species, from which the bones of the great Mastodon have been taken in a very perfect state. Yet the whole excavation of the ravine, for many miles below the Falls, has been slowly effected since that fluviatile deposit was thrown down.

Whether or not, in assigning a period of more than 30,000 years for the recession of the Falls from Queenstown to their present site, I have over or under estimated the time required for that operation, no one can doubt that a vast number of centuries must have elapsed before so great a series of geographical changes were brought about as have occurred since the entombment of this elephantine quadruped. The freshwater gravel which incloses it is decidedly of much more modern origin than the drift or boulder clay of the same region.[138-A]

Other extinct animals accompany the Mastodon giganteus in the post-glacial deposits of the United States, among which the Castoroides ohioensis, Foster and Wyman, a huge rodent allied to the beaver, and the Capybara may be mentioned. But whether the "loess," and other freshwater and marine strata of the Southern States, in which skeletons of the same Mastodon are mingled with the bones of the Megatherium, Mylodon, and Megalonyx, were contemporaneous with the drift, or were of subsequent date, is a chronological question still open to discussion. It appears clear, however, from what we know of the tertiary fossils of Europe—and I believe the same will hold true in North America—that many species of testacea and some mammalia, which existed prior to the glacial epoch, survived that era. As European examples among the warm-blooded quadrupeds, the Elephas primigenius and Rhinoceros tichorinus may be mentioned. As to the shells, whether fresh water, terrestrial, or marine, they need not be enumerated here, as allusion will be made to them in the sequel, when the pliocene tertiary fossils of Suffolk are described. The fact is important, as refuting the hypothesis that the cold of the glacial period was so intense and universal as to annihilate all living creatures throughout the globe.

That the cold was greater for a time than it is now in certain parts of Siberia, Europe, and North America, will not be disputed; but, before we can infer the universality of a colder climate, we must ascertain what was the condition of other parts of the northern, and of the whole southern, hemisphere at the time when the Scandinavian, British, and Alpine erratics were transported into their present position. It must not be forgotten that a great deposit of drift and erratic blocks is now in full progress of formation in the southern hemisphere, in a zone corresponding in latitude to the Baltic, and to Northern Italy, Switzerland, France, and England. Should the uneven bed of the southern ocean be hereafter converted by upheaval into land, the hills and valleys will be strewed over with transported fragments, some derived from the antarctic continent, others from islands covered with glaciers, like South Georgia, which must now be centres of the dispersion of drift, although situated in a latitude, agreeing with that of the Cumberland mountains in England.

Not only are these operations going on between the 45th and 60th parallels of latitude south of the line, while the corresponding zone of Europe is free from ice; but, what is still more worthy of remark, we find in the southern hemisphere itself, only 900 miles distant from South Georgia, where the perpetual snow reaches to the sea-beach, lands covered with forests, as in Terra del Fuego. There is here no difference of latitude to account for the luxuriance of vegetation in one spot, and the absolute want of it in the other; but among other refrigerating causes in South Georgia may be enumerated the countless icebergs which float from the antarctic zone, and which chill, as they melt, the waters of the ocean, and the surrounding air, which they fill with dense fogs.

I have endeavoured in the "Principles of Geology," chapters 7. and 8., to point out the intimate connexion of climate and the physical geography of the globe, and the dependence of the mean annual temperature, not only on the height of the dry land, but on its distribution in high or low latitudes at particular epochs. If, for example, at certain periods of the past, the antarctic land was less elevated and less extensive than now, while that at the north pole was higher and more continuous, the conditions of the northern and southern hemispheres might have been the reverse of what we now witness in regard to climate, although the mountains of Scandinavia, Scotland, and Switzerland, may have been less elevated than at present. But if in both of the polar regions a considerable area of elevated dry land existed, such a concurrence of refrigerating conditions in both hemispheres might have created for a time an intensity of cold never experienced since; and such probably was the state of things during that period of submergence to which I have alluded in this chapter.

Alpine erratics.—Although the arctic regions constitute the great centre from which erratics have travelled southwards in all directions in Europe and North America, yet there are some mountains, as I have already stated, like those of North Wales and the Alps, which have served as separate and independent centres for the dispersion of blocks. In illustration of this fact, the Alps deserve particular attention, not only from their magnitude, but because they lie beyond the ordinary limits of the "northern drift" of Europe, being situated between the 44th and 47th degrees of north latitude. On the flanks of these mountains, and on the Subalpine ranges of hills or plains adjoining them, those appearances which have been so often alluded to, as distinguishing or accompanying the drift, between the 50th and 70th parallels of north latitude, suddenly reappear, to assume in a more southern country their most exaggerated form. Where the Alps are highest, the largest erratic blocks have been sent forth, as, for example, from the regions of Mont Blanc and Monte Rosa, into the adjoining parts of France, Switzerland, Austria, and Italy, while in districts where the great chain sinks in altitude, as in Carinthia, Carniola, and elsewhere, no such rocky fragments, or a few only and of smaller bulk, have been detached and transported to a distance.

In the year 1821, M. Venetz first announced his opinion that the Alpine glaciers must formerly have extended far beyond their present limits, and the proofs appealed to by him in confirmation of this doctrine were afterwards acknowledged by M. Charpentier, who strengthened them by new observations and arguments, and declared, in 1836, his conviction that the glaciers of the Alps must once have reached as far as the Jura, and have carried thither their moraines across the great valley of Switzerland. M. Agassiz, after several excursions in the Alps with M. Charpentier, and after devoting himself some years to the study of glaciers, published, in 1840, an admirable description of them, and of the marks which attest the former action of great masses of ice over the entire surface of the Alps and the surrounding country.[140-A] He pointed out that the surface of every large glacier is strewed over with gravel and stones detached from the surrounding precipices by frost, rain, lightning, or avalanches. And he described more carefully than preceding writers the long lines of these stones, which settle on the sides of the glacier, and are called the lateral moraines; those found at the lower end of the ice being called terminal moraines. Such heaps of earth and boulders every glacier pushes before it when advancing, and leaves behind it when retreating. When the Alpine glacier reaches a lower and warmer situation, about 3000 or 4000 feet above the sea, it melts so rapidly that, in spite of the downward movement of the mass, it can advance no farther. Its precise limits are variable from year to year, and still more so from century to century; one example being on record of a recession of half a mile in a single year. We also learn from M. Venetz, that whereas, between the eleventh and fifteenth centuries, all the Alpine glaciers were less advanced than now, they began in the seventeenth and eighteenth centuries to push forward so as to cover roads formerly open, and to overwhelm forests of ancient growth.

These oscillations enable the geologist to note the marks which they leave behind them as they retrograde, and among these the most prominent, as before stated, are the terminal moraines, or mounds of unstratified earth and stones, often divided by subsequent floods into hillocks, which cross the valley like ancient earth-works, or embankments made to dam up the river. Some of these transverse barriers were formerly pointed out by Saussure below the glacier of the Rhone, as proving how far it had once transgressed its present boundaries. On these moraines we see many large angular fragments, which, having been carried along on the surface of the ice, have not had their edges worn off by friction; but the greater number of the boulders, even those of large size, have been well rounded, not by the power of water, but by the mechanical force of the ice, which has pushed them against each other, or against the rocks flanking the valley. Others have fallen down the numerous fissures which intersect the glacier, where, being subject to the pressure of the whole mass of ice, they have been forced along, and either well rounded or ground down into sand, or even the finest mud, of which the moraine is largely constituted.

As the terminal moraines are the most prominent of all the monuments left by a receding glacier, so are they the most liable to obliteration; for violent floods or debacles are often occasioned in the Alps by the sudden bursting of what are called glacier-lakes. These temporary sheets of water are caused by the damming up of a river by a glacier which has increased during a succession of cold seasons, and, descending from a tributary into the main valley, has crossed it from side to side. On the failure of this icy barrier, the accumulated waters are let loose, which sweep away and level all transverse mounds of gravel and loose boulder below, and spread their materials in confused and irregular beds over the river-plain.

Another mark of the former action of glaciers, in situations where they exist no longer, is the polished, striated, and grooved surfaces of rocks already alluded to. Stones which lie underneath the glacier and are pushed along by it, sometimes adhere to the ice, and as the mass glides slowly along at the rate of a few inches, or at the utmost two or three feet, per day, abrade, groove, and polish the rock, and the larger blocks are reciprocally grooved and polished by the rock on their lower sides. As the forces both of pressure and propulsion are enormous, the sand, acting like emery, polishes the surface; the pebbles, like coarse gravers, scratch and furrow it; and the large stones scoop out grooves in it. Another effect also of this action, not yet adverted to, is called "roches moutonnÉes." Projecting eminences of rock are smoothed and worn into the shape of flattened domes, where the glaciers have passed over them.

Although the surface of almost every kind of rock, when exposed in the open air, wastes away by decomposition, yet some retain for ages their polished and furrowed exterior; and, if they are well protected by a covering of clay or turf, these marks of abrasion seem capable of enduring for ever. They have been traced in the Alps to great heights above the present glaciers, and to great horizontal distances beyond them.

There are also found, on the sides of the Swiss valleys, round and deep holes, with polished sides, such holes as waterfalls make in the solid rock, but in places remote from running waters, and where the form of the surface will not permit us to suppose that any cascade could ever have existed. Similar cavities are common in hard rocks, such as gneiss, in Sweden, where they are called giant caldrons, and are sometimes 10 feet and more in depth; but in the Alps and Jura they often pass into spoon-shaped excavations and prolonged gutters. We learn from M. Agassiz that hollows of this form are now cut out by streams of water, which flow along the surface of glaciers, and then fall into fissures which are open to the bottom. Here, forming a cascade, the stream cuts a round cavity in the rock with the gravel and sand, which it either finds there or carries down with it, and causes to rotate; and, as it usually happens that the glacier is advancing, a locomotive cascade is produced, which converts the first circular hole into a deep groove.

Another effect of a glacier is to lodge a ring of stones round the summit of a conical peak which may happen to project through the ice. If the glacier is lowered greatly by melting, these circles of large angular fragments, which are called "perched blocks," are left in a singular situation near the top of a steep hill or pinnacle, the lower parts of which may be destitute of boulders.

Alpine blocks on the Jura.—Now some or all the marks above enumerated,—the moraines, erratics, polished surfaces, domes, striÆ, caldrons, and perched rocks, are observed in the Alps at great heights above the present glaciers, and far below their actual extremities; also in the great valley of Switzerland, 50 miles broad; and almost everywhere on the Jura, a chain which lies to the north of this valley. The average height of the Jura is about one third that of the Alps, and is now entirely destitute of glaciers, yet it presents almost everywhere similar moraines, and the same polished and grooved surfaces, and water-worn cavities. The erratics, moreover, which cover it, present a phenomenon which has astonished and perplexed the geologist for more than half a century. No conclusion can be more incontestible than that these angular blocks of granite, gneiss, and other crystalline formations, came from the Alps, and that they have been brought for a distance of 50 miles and upwards across one of the widest and deepest valleys of the world, so that they are now lodged on the hills and valleys of a chain composed of limestone and other formations, altogether distinct from those of the Alps. Their great size and angularity, after a journey of so many leagues, has justly excited wonder; for hundreds of them are as large as cottages; and one in particular, celebrated under the name of Pierre À Bot, rests on the side of a hill about 900 feet above the lake of Neufchatel, and is no less than 40 feet in diameter.

It will be remarked that these blocks on the Jura offer an exception to the rule before laid down, as applicable in general to erratics, since they have gone from south to north. Some of the largest masses of granite and gneiss have been found to contain 50,000 and 60,000 cubic feet of stone, and one limestone block near Devens, which has travelled 30 miles, contains 161,000 cubic feet, its angles being sharp and unworn.[143-A]

Von Buch, Escher, and Studer have shown, from an examination of the mineral composition of the boulders, that those on the western Jura, near Neufchatel, have come from the region of Mont Blanc and the Valais; those on the middle parts of the Jura from the Bernese Oberland; and those on the eastern Jura from the Alps of the small cantons, Glaris, Schwytz, Uri, and Zug. The blocks, therefore, of these three great districts have been derived from parts of the Alps nearest to the localities in the Jura where we now find them, as if they had crossed the great valley in a direction at right angles to its length: the most western stream having followed the course of the Rhone; the central, that of the Aar; and the eastern, that of the two great rivers, Reuss and Limmat. The non-intermixture of these groups of travelled fragments, except near their confines, was always regarded as most enigmatical by those who adopted the opinion of Saussure, that they were all whirled along by a rapid current of muddy water rushing from the Alps.

M. Charpentier first suggested, as before mentioned, that the Swiss glaciers once reached continuously to the Jura, and conveyed to them these erratics; but at the same time he conceived that the Alps were formerly higher than now. M. Agassiz, on the other hand, instead of introducing distinct and separate glaciers, imagines that the whole valley of Switzerland was filled with ice, and that one great sheet of it extended from the Alps to the Jura, when the two chains were of the same height as now relatively to each other. Such an hypothesis labours under this difficulty, that the difference of altitude, when distributed over a space of 50 miles, gives an inclination of no more than two degrees, or far less than that of any known glaciers. It has, however, since received the able support of Professor James Forbes, in his excellent work on the Alps, published in 1843.

In the theory which I formerly advanced, jointly with Mr. Darwin[143-B], it was suggested that the erratics may have been transferred by floating ice to the Jura, at the time when the greater part of that chain, and the whole of the Swiss valley to the south, was under the sea. At that period the Alps may have attained only half their present altitude, and may yet have constituted a chain as lofty as the Chilian Andes, which, in a latitude corresponding to Switzerland, now send down glaciers to the head of every sound, from which icebergs, covered with blocks of granite, are floated seaward.[144-A] Opposite that part of Chili where the glaciers abound is situated the island of Chiloe, 100 miles in length, with a breadth of 30 miles, running parallel to the continent. The channel which separates it from the main land is of considerable depth, and 25 miles broad. Parts of its surface, like the adjacent coast of Chili, are overspread with recent marine shells, showing an upheaval of the land during a very modern period; and beneath these shells is a boulder deposit, in which Mr. Darwin found large travelled blocks. One group of fragments were of granite, which had evidently come from the Andes, while in another place angular blocks of syenite were met with. Their arrangement may have been due to successive crops of icebergs issuing from different sounds, to the heads of which glaciers descend from the Andes. These icebergs, taking their departure year after year from distinct points, may have been stranded repeatedly, in equally distinct groups, in bays or creeks of Chiloe, and on islets off the coast, so as afterwards to appear, some on hills and others in valleys, when that country and the bed of the adjacent sea had been upheaved. A continuance in future of the elevatory movement, in the region of the Andes and of Chiloe, might cause the former chain to rival the Alps in altitude, and give to Chiloe a height equal to that of the Jura. The same rise might dry up the channel between Chiloe and the main land, so that it would then represent the great valley of Switzerland. In the course of these changes, all parts of Chiloe and the intervening strait, having in their turn been a sea-shore, may have been polished and scratched by coast-ice, and by innumerable icebergs running aground and grating on the bottom.

If we apply this hypothesis to Switzerland and the Jura, we are by no means precluded from the supposition that, in proportion as the land acquired additional height, and the bed of the sea emerged, the Jura itself may have had its glaciers; and those existing in the Alps, which had at first extended to the sea, may, during some part of the period of upheaval, have been prolonged much farther into the valleys than now. At a later period, when the climate grew milder, these glaciers may have entirely disappeared from the Jura, and may have receded in the Alps to their present limits, leaving behind them in both districts those moraines which now attest the former extension of the ice.[144-B]Meteorites in drift.—Before concluding my remarks on the northern drift of the Old World, I shall refer to a fact recently announced, the discovery of a meteoric stone at a great depth in the alluvium of Northern Asia.

Erman, in his Archives of Russia for 1841 (p. 314.), cites a very circumstantial account drawn up by a Russian miner of the finding of a mass of meteoric iron in the auriferous alluvium of the Altai. Some small fragments of native iron were first met with in the gold-washings of Petropawlowsker in the Mrassker Circle; but though they attracted attention, it was supposed that they must have been broken off from the tools of the workmen. At length, at the depth of 31 feet 5 inches from the surface, they dug out a piece of iron weighing 17¹/₂ pounds, of a steel-grey colour, somewhat harder than ordinary iron, and, on analysing it, found it to consist of native iron, with a small proportion of nickel, as usual in meteoric stones. It was buried in the bottom of the deposit where the gravel rested on a flaggy limestone. Much brown iron ore, as well as gold, occurs in the same gravel, which appears to be part of that extensive auriferous formation in which the bones of the mammoth, the Rhinoceros tichorhinus, and other extinct quadrupeds abound. No sufficient data are supplied to enable us to determine whether it be of Post-Pliocene or Newer Pliocene date.

We ought not, I think, to feel surprise that we have not hitherto succeeded in detecting the signs of such aËrolites in older rocks, for, besides their rarity in our own days, those which fell into the sea (and it is with marine strata that geologists have usually to deal), being chiefly composed of native iron, would rapidly enter into new chemical combinations, the water and mud being charged with chloride of sodium and other salts. We find that anchors, cannon, and other cast-iron implements which have been buried for a few hundred years off our English coast have decomposed in part or entirely, turning the sand and gravel which enclosed them into a conglomerate, cemented together by oxide of iron. In like manner meteoric iron, although its rusting would be somewhat checked by the alloy of nickel, could scarcely ever fail to decompose in the course of thousands of years, becoming oxide, sulphuret or carbonate of iron, and its origin being then no longer distinguishable. The greater the antiquity of rocks,—the oftener they have been heated and cooled, permeated by gases or by the waters of the sea, the atmosphere or mineral springs,—the smaller must be the chance of meeting with a mass of native iron unaltered; but the preservation of the ancient meteorite of the Altai, and the presence of nickel in these curious bodies, renders the recognition of them in deposits of remote periods less hopeless than we might have anticipated.