VIII.

VIII.

Recent Researches in the Glacial Geology of the Continent.[AA]

[AA] Presidential Address to the Geological Section of the British Association, Newcastle, 1889.

THE President of this section must often have some difficulty in selecting a subject for his address. It is no longer possible to give an interesting and instructive summary of the work done by the devotees of our science during even one year. So numerous have the students of geological science become--so fertile are the fields they cultivate--so abundant the harvests they reap, that one in my present position may well despair of being able to take stock of the numerous additions to our knowledge which have accumulated within the last twelve months. Neither is there any burning question which at this time your President need feel called upon to discuss. True, there are controversies that are likely to remain unsettled for years to come--there are still not a few matters upon which we must agree to differ--we do not yet see eye to eye in all things geological. But experience has shown that as years advance truth is gradually evolved, and old controversies die out, and so doubtless it will continue to be. The day when controversies shall cease, however, is yet, I hope, far in the future; for should that dull and unhappy time ever arrive, it is quite certain that mineralogists, petrologists, palæontologists, and geologists shall have died out of the world. Following the example of many of my predecessors, I shall confine my remarks to certain questions in which I have been specially interested; and in doing so I shall endeavour to steer clear, as far as I can, of controversial matters. My purpose, then, is to give an outline of some of the results obtained during the last few years by Continental workers in the domain of glacial geology.

Those who are not geologists will probably smile when they hear one declare that wielders of the hammer are extremely conservative--that they are slow to accept novel views, and very tenacious of opinions which have once found favour in their eyes. Nevertheless, such is the case, and well for us that it is so. However captivating, however imposing, however strongly supported by evidence a new view may appear to be, we do well to criticise, to sift the evidence, and to call for more facts and experiments, if such are possible, until the proofs become so strong as to approach as near a demonstration as geologists can in most cases expect such proofs to go. The history of our science, and indeed of most sciences, affords abundant illustration of what I say. How many long years were the views of sub-aërial erosion, as taught by Hutton and Playfair, canvassed and controverted before they became accepted! And even after their general soundness had been established, how often have we heard nominal disciples of these fathers of physical geology refuse to go so far as to admit that the river-valleys of our islands have been excavated by epigene agents!

If, as a rule, it takes some time for a novel view to gain acceptance, it is equally true that views which have long been held are only with difficulty discarded. Between the new and the old there is a constant struggle for existence, and if the latter should happen to survive, it is only in a modified form. I have often thought that a history of the evolution of geological theories would make a very entertaining and instructive work. We should learn from it, amongst other things, that the advance of our science has not always been continuous--now and again, indeed, it has almost seemed as if the movement had been retrograde. Knowledge has not come in like an overwhelming flood--as a broad majestic river--but rather like a gently-flowing tide, now advancing, now retiring, but ever, upon the whole, steadily gaining ground. The history I speak of would also teach us that many of the general views and hypotheses which have been from time to time abandoned as unworkable, are hardly deserving of the reproach and ridicule which we in these latter days may be inclined to cast upon them.

As the Scots proverb says: "It is easy to be wise behindhand." It could be readily shown that not a few discarded notions and opinions have frequently worked for good, and have rather stimulated than checked inquiry. Such reflections should be encouraging to every investigator, whether he be a defender of the old or an advocate of the new. Time tries all, and each worker may claim a share in the final establishment of the truth.

Perhaps there is no department of geological inquiry that has given rise to more controversy than that which I have selected for the subject of this address. Hardly a single step in advance has been taken without vehement opposition. But the din of contending sides is not so loud now--the dust of the conflict has to some extent cleared away, and the positions which have been lost or maintained, as the case may be, can be readily discerned. The glacialist who can look back over the last twenty-five years of wordy conflict has every reason to be jubilant and hopeful. Many of those who formerly opposed him have come over to his side. It is true he has not had everything his own way. Some extreme views have been abandoned in the struggle; that of a great Polar ice-sheet, for example, as conceived of by Agassiz. I am not aware, however, that many serious students of glacial geology ever adopted that view. But it was quite an excusable hypothesis, and has been abundantly suggestive. Had Agassiz lived to see the detailed work of these later days, he would doubtless have modified his notion and come to accept the view of large continental glaciers which has taken its place.

The results obtained by geologists who have been studying the peripheral areas of the drift-covered regions of our Continent, are such as to satisfy us that the drifts of those regions are not iceberg-droppings, as we used to suppose, but true morainic matter and fluvio-glacial detritus. Geologists have not jumped to this conclusion--they have only accepted it after laborious investigation of the evidence. Since Dr. Otto Torell, in 1875, first stated his belief that the Diluvium of north Germany was of glacial origin a great literature on the subject has sprung up, a perusal of which will show that with our German friends glacial geology has passed through much the same succession of controversial phases as with us. At first icebergs are appealed to as explaining everything--next we meet with sundry ingenious attempts at a compromise between floating-ice and a continuous ice-sheet. As observations multiply, however, the element of floating-ice is gradually eliminated, and all the phenomena are explained by means of land-ice and "Schmelz-wasser" alone. It is a remarkable fact that the iceberg hypothesis has always been most strenuously upheld by geologists whose labours have been largely confined to the peripheral areas of drift-covered countries.

In the upland and mountainous tracts, on the other hand, that hypothesis has never been able to survive a moderate amount of accurate observation. Even in Switzerland--the land of glaciers--geologists at one time were of opinion that the boulder-clays of the low-grounds had a different origin from those which occur in the mountain-valleys. Thus, it was supposed that at the close of the Pleistocene period the Alps were surrounded by great lakes or by gulfs of some inland sea, into which the glaciers of the high valleys flowed and calved their icebergs--these latter scattering erratics and earthy débris over the drowned areas. Sartorius von Waltershausen[AB] set forth this view in an elaborate and well-illustrated paper. Unfortunately for his hypothesis no trace of the supposed great lakes or the inland sea has ever been detected: on the contrary, the character of the morainic accumulations, and the symmetrical grouping and radiation of the erratics and perched blocks over the foot-hills and low-grounds, show that these last have been invaded and overflowed by the glaciers themselves. Even the most strenuous upholders of the efficacy of icebergs as originators of some boulder-clays, admit that the boulder-clay or till, of what we may call the inner or central region of a glaciated tract is the product of land-ice. Under this category comes the boulder-clay of Norway, Sweden, and Finland, and of the Alpine Lands of central Europe, not to speak of the hilly parts of our own islands.

[AB] "Untersuchungen über die Klimate der Gegenwart und der Vorwelt," etc.--_Natuurkundige Verhandelingen v. d. Holland. Maatsch. d. Wetensch. te Haarlem_, 1865.

When we come to study the drifts of the peripheral areas, it is not difficult to see why these should be considered to have had a different origin. They present certain features which, although not absent from the glacial deposits of the inner region, are not nearly so characteristic of such upland tracts. I refer especially to the frequent interstratification of boulder-clays with well-bedded deposits of clay, sand, and gravel; and to the fact that these boulder-clays are often less compressed than those of the inner region, and have even occasionally a silt-like character. Such appearances do seem at first to be readily explained on the assumption that the deposits have been accumulated in water opposite the margin of a continental glacier or ice-sheet--and this was the view which several able investigators in Germany were for some time inclined to adopt.

But when the phenomena came to be studied in greater detail, and over a wider area, this preliminary hypothesis did not prove satisfactory. It was discovered, for example, that "giants' kettles"[AC] were more or less commonly distributed under the glacial deposits, and such "kettles" could only have originated at the bottom of a glacier. Again, it was found that pre-glacial accumulations were plentifully developed in certain places below the drift, and were often involved with the latter in a remarkable way. The "brown-coal formation" in like manner was violently disturbed and displaced, to such a degree that frequently the boulder-clay is found to underlie it. Similar phenomena were encountered in regions where the drift overlies the Chalk--the latter presenting the appearance of having been smashed and shattered--the fragments having often been dragged some distance, so as to form a kind of friction-breccia underlying the drift, while large masses are often included in the clay itself. All the facts pointed to the conclusion that these disturbances were due to tangential thrusting or crushing, and were not the result of vertical displacements, such as are produced by normal faulting, for the disturbances in question die out from above downwards.

Evidence of similar thrusting or crushing is seen in the remarkable faults and contortions that so often characterise the clays and sands that occur in the boulder-clay itself. The only agent that could produce the appearances, now briefly referred to, is land-ice, and we must therefore agree with German geologists that glacier-ice has overflowed all the drift-covered regions of the peripheral area. No evidence of marine action in the formation of the stony clays is forthcoming--not a trace of any sea-beach has been detected. And yet, if these clays had been laid down in the sea during the retreat of the ice-sheet from Germany, surely such evidence as I have indicated ought to be met with. To the best of my knowledge the only particular facts which have been appealed to, as proofs of marine action, are the appearance of bedded deposits in the boulder-clays, and the occasional occurrence in the clays themselves of a sea-shell. But other organic remains are also met with now and again in similar positions, such as mammalian bones and freshwater shells. All these, however, have been shown to be derivative in their origin--they are just as much erratics as the stones and boulders with which they are associated. The only phenomena, therefore, that the glacialist has to account for are the bedded deposits which occur so frequently in the boulder-clays of the peripheral regions, and the occasional silty and uncompressed character of the clays themselves.

[AC] These appear to have been first detected by Professor Berendt and Professor E. Geinitz.

The intercalated beds are, after all, not hard to explain. If we consider for a moment the geographical distribution of the boulder-clays, and their associated aqueous deposits, we shall find a clue to their origin. Speaking in general terms, the stony clays thicken out as they are followed from the mountainous and high-lying tracts to the low-grounds. Thus they are of inconsiderable thickness in Norway, the higher parts of Sweden, and in Finland, just as we find is the case in Scotland, northern England, Wales, and the hilly parts of Ireland. Traced south from the uplands of Scandinavia and Finland, they gradually thicken out as the low-grounds are approached. Thus in southern Sweden they reach a thickness of 43 metres or thereabout, and of 80 metres in the northern parts of Prussia, while over the wide low-lying regions to the south they attain a much greater thickness--reaching in Holstein, Mecklenburg, Pomerania, and west Prussia, a depth of 120 to 140 metres, and still greater depths in Hanover, Mark Brandenburg, and Saxony. In those regions, however, a considerable portion of the diluvium consists, as we shall see presently, of water-formed beds.

The geographical distribution of the aqueous deposits, which are associated with the stony clays, is somewhat similar. They are very sparingly developed in districts where the boulder-clays are thin. Thus they are either wanting, or only occur sporadically in thin irregular beds, in the high-grounds of northern Europe generally. Further south, however, they gradually acquire more importance, until in the peripheral regions of the drift-covered tracts they come to equal and eventually to surpass the boulder-clays in prominence. These latter, in fact, at last cease to appear, and the whole bulk of the diluvium, along the southern margin of the drift area, appears to consist of aqueous accumulations alone.

The explanations of these facts advanced by German geologists are quite in accordance with the views which have long been held by glacialists elsewhere, and have been tersely summed up by Dr. Jentzsch.[AD] The northern regions, he says, were the feeding-grounds of the inland-ice. In those regions melting was at a minimum, while the grinding action of the ice was most effective. Here, therefore, erosion reached its maximum--ground-moraine or boulder-clay being unable to accumulate to any thickness. Further south melting greatly increased, while ground-moraine at the same time tended to accumulate--the conjoint action of glacier-ice and sub-glacial water resulting in the complex drifts of the peripheral area. In the disposition and appearance of the aqueous deposits of the diluvium we have evidence of an extensive sub-glacial water-circulation--glacier-mills that gave rise to "giants' kettles"--chains of sub-glacial lakes in which fine clays gathered--streams and rivers that flowed in tunnels under the ice, and whose courses were paved with sand and gravel. Nowhere do German geologists find any evidence of marine action. On the contrary, the dovetailing and interosculation of boulder-clay with aqueous deposits are explained by the relation of the ice to the surface over which it flowed.

Throughout the peripheral area it did not rest so continuously upon the ground as was the case in the inner region of maximum erosion. In many places it was tunnelled by rapid streams and rivers, and here and there it arched over sub-glacial lakes, so that accumulation of ground-moraine proceeded side by side with the formation of aqueous sediments. Much of that ground-moraine is of the usual tough and hard-pressed character, but here and there it is somewhat less coherent and even silt-like. Now a study of the ground-moraines of modern glaciers affords us a reasonable explanation of such differences. Dr. Brückner[AE] has shown that in many places the ground-moraine of the Alpine glaciers is included in the bottom of the ice itself. The ground-moraine, he says, frequently appears as an ice-stratum abundantly impregnated with silt and rock-fragments--it is like a conglomerate or breccia which has ice for its binding material. When this ground-moraine melts out of the ice--no running water being present--it forms a layer of unstratified silt or clay, with stones scattered irregularly through it. Such being the case in modern glaciers, we can hardly doubt that over the peripheral areas occupied by the old northern ice-sheet boulder-clay must frequently have been accumulated in the same way. Nay, when the ground-moraine melted out and dropt here and there into quietly-flowing water it might even acquire in part a bedded character.

[AD] _Jahrb. d. königl. preuss. geologischen Landesanstalt für 1884_, p. 438.

[AE] "Die Vergletscherung des Salzachgebietes, etc.": _Geographische Abhandlungen herausgegeben v. A. Penck_, Band i. Heft 1.

The limits reached by the inland-ice during its greatest extension are becoming more and more clearly defined, although its southern margin will probably never be so accurately determined as that of the latest epoch of general glaciation. The reasons for this are obvious. When the inland-ice flowed south to the Harz and the hills of Saxony it formed no great terminal moraines. Doubtless many erratics and much rock-rubbish were showered upon the surface of the ice from the higher mountains of Scandinavia, but owing to the fanning-out of the ice on its southward march, such superficial débris was necessarily spread over a constantly-widening area. It may well be doubted, therefore, whether it ever reached the terminal front of the ice-sheet in sufficient bulk to form conspicuous moraines. It seems most probable that the terminal moraines of the great inland-ice would consist of low banks of boulder-clay and aqueous materials-the latter, perhaps, strongly predominating, and containing here and there larger and smaller angular erratics which had travelled on the surface of the ice. However that may be, it is certain that the whole region in question has been considerably modified by subsequent denudation, and to a large extent is now concealed under deposits belonging to later stages of the Pleistocene period.

The extreme limits reached by the ice are determined rather by the occasional presence of rock-striæ and _roches moutonnées_, of boulder-clay and northern erratics, than by recognisable terminal moraines. The southern limits reached by the old inland-ice appear in this way to have been tolerably well ascertained over a considerable portion of central Europe. Some years ago I published a small sketch-map[AF] showing the extent of surface formerly covered by ice. On this map I did not venture to draw the southern margin of the ice-sheet in Belgium further south than Antwerp, where northern erratics were known to occur, but the more recent researches of Belgian geologists show that the ice probably flowed south for some little distance beyond Brussels.[AG] Here and there in other parts of the Continent the southern limits reached by the northern drift have also been more accurately determined, but, so far as I know, none of these later observations involves any serious modification of the sketch-map referred to.

[AF] _Prehistoric Europe_, 1881.

[AG] See a paper by M. E. Delvaux: _Ann. de la. Soc. géol. de Belg._, t. xiii. p. 158.

I have now said enough, however, to show that the notion of a general ice-sheet having covered so large a part of Europe, which a few years ago was looked upon as a wild dream, has been amply justified by the labours of those who are so assiduously investigating the peripheral areas of the "great northern drift." And perhaps I may be allowed to express my own belief that the drifts of middle and southern England, which exhibit the same complexity as the Lower Diluvium of the Continent, will eventually be generally acknowledged to have had a similar origin. I have often thought that whilst politically we are happy in having the sea all round us, geologically we should have gained perhaps by its greater distance. At all events we should have been less ready to invoke its assistance to explain every puzzling appearance presented by our glacial accumulations.

I now pass on to review some of the general results obtained by continental geologists as to the extent of area occupied by inland-ice during the last great extension of glacier-ice in Europe. It is well known that this latest ice-sheet did not overflow nearly so wide a region as that underneath which the lowest boulder-clay was accumulated. This is shown not only by the geographical distribution of the youngest boulder-clay, but by the direction of rock-striæ, the trend of erratics, and the position of well-marked terminal moraines. Gerard de Geer has given a summary[AH] of the general results obtained by himself and his fellow-workers in Sweden and Norway; and these have been supplemented by the labours of Berendt, E. Geinitz, Hauchecorne, Keilhack, Klockmann, Schröder, Wahnschaffe, and others in Germany, and by Sederholm in Finland.[AI] From them we learn that the end-moraines of the ice circle round the southern coasts of Norway, from whence they sweep south-east by east across the province of Gottland in Sweden, passing through the lower ends of Lakes Wener and Wetter, while similar moraines mark out for us the terminal front of the inland-ice in Finland--at least two parallel frontal moraines passing inland from Hango Head on the Gulf of Finland through the southern part of that province to the north of Lake Ladoga.

Further north-east than this they have not been traced; but, from some observations by Helmersen, Sederholm thinks it probable that the terminal ice-front extended north-east by the north of Lake Onega to the eastern shores of the White Sea. Between Sweden and Finland lies the basin of the Baltic, which at the period in question was filled with ice, forming a great Baltic glacier which overflowed the Öland Islands, Gottland, and Öland, and which, fanning-out as it passed towards the south-west, invaded, on the south side, the Baltic provinces of Germany, while, on the north, it crossed the southern part of Scania in Sweden and the Danish islands to enter Jutland.

[AH] _Zeitschrift d. deutsch. geolog. Ges._, Bd. xxxvii., p. 177.

[AI] For papers by Berendt and his associates see especially the _Jahrbuch d. k. preuss. geol. Landesanstalt_, and the _Zaitschr. d. deutsch. geol. Ges._ for the past few years. Geinitz: _Forsch. z. d. Lands- u. Volkskunde_, i. 5; _Leopoldina_, xxii., p. 37; I. _Beitrag z. Geologie Mecklenburgs_, 1880, pp. 46, 56. Sederholm: _Fennia_, I. No. 7.

The upper boulder-clay of those regions is now recognised as the ground-moraine of this latest ice-sheet. In many places it is separated from the older boulder-clay by interglacial deposits--some of which are marine, while others are of freshwater and terrestrial origin. During interglacial times the sea that overflowed a considerable portion of north Germany was evidently continuous with the North Sea, as is shown not only by the geographical distribution of the interglacial marine deposits, but by their North Sea fauna. German geologists generally group all the interglacial deposits together, as if they belonged to one and the same interglacial epoch. This perhaps we must look upon as only a provisional arrangement. Certain it is that the freshwater and terrestrial beds which frequently occur on the same or a lower level, and at no great distance from the marine deposits, cannot in all cases be contemporaneous with the latter. Possibly, however, such discordances may be accounted for by oscillations in the level of the interglacial sea--land and water having alternately prevailed over the same area. Two boulder-clays, as we have seen, have been recognised over a wide region in the north of Germany. In some places, however, three or more such boulder-clays have been observed overlying one another throughout considerable areas, and these clays are described as being distinctly separate and distinguishable the one from the other.[AJ] Whether they, with their intercalated aqueous deposits, indicate great oscillations of one and the same ice-sheet--now advancing, now retreating--or whether the stony clays may not be the ground-moraines of so many different ice-sheets, separated the one from the other by true interglacial conditions, future investigations must be left to decide.

[AJ] H. Schröder: _Jahrb. d. k. preuss. geol. Landesanstalt für 1887_ , p. 360.

The general conclusions arrived at by those who are at present investigating the glacial accumulations of northern Europe may be summarised as follows:--

1. Before the invasion of northern Germany by the inland-ice the low-grounds bordering on the Baltic were overflowed by a sea which contained a boreal and arctic fauna. These marine conditions are indicated by the presence under the lower boulder-clay of more or less well-bedded fossiliferous deposits. On the same horizon occur also beds of sand, containing freshwater shells, and now and again mammalian remains, some of which imply cold and others temperate climatic conditions. Obviously all these deposits may pertain to one and the same period, or more properly to different stages of the same period--some dating back to a time when the climate was still temperate, while others clearly indicate the prevalence of cold conditions, and are therefore probably somewhat younger.

2. The next geological horizon in ascending order is that which is marked by the Lower Diluvium--the glacial and fluvio-glacial detritus of the great ice-sheet which flowed south to the foot of the Harz Mountains. The boulder-clay on this horizon now and again contains marine, freshwater, and terrestrial organic remains--derived undoubtedly from the so-called pre-glacial beds already referred to. These latter, it would appear, were ploughed up and largely incorporated with the old ground-moraine.

3. The interglacial beds which next succeed contain remains of a well-marked temperate fauna and flora, which point to something more than a mere partial or local retreat of the inland-ice. The geographical distribution of the beds, and the presence in these of such forms as _Elephas antiquus_, _Cervus elephas_, _C. megaceros_, and a flora comparable to that now existing in northern Germany, justify geologists in concluding that the interglacial epoch was one of long duration, and characterised in Germany by climatic conditions apparently not less temperate than those that now obtain. One of the phases of that interglacial epoch, as we have seen, was the overflowing of the Baltic provinces by the waters of the North Sea.

4. To this well-marked interglacial epoch succeeded another epoch of arctic conditions, when the Scandinavian inland-ice once more invaded Germany, ploughing through the interglacial deposits, and working these up in its ground-moraine. So far as I can learn, the prevalent belief among geologists in north Germany is that there was only one interglacial epoch; but, as already stated, doubt has been expressed whether all the facts can be thus accounted for. There must always be great difficulty in the correlation of widely-separated interglacial deposits, and the time does not seem to me to have yet come when we can definitely assert that all those interglacial beds belong to one and the same geological horizon.

I have dwelt upon the recent work of geologists in the peripheral areas of the drift-covered regions of northern Europe, because I think the results obtained are of great interest to glacialists in this country. And for the same reason I wish next to call attention to what has been done of late years in elucidating the glacial geology of the Alpine Lands of central Europe--and more particularly of the low-grounds that stretch out from the foot of the mountains. Any observations that tend to throw light upon the history of the complex drifts of our own peripheral areas cannot but be of service. It is quite impossible to do justice in this brief sketch to the labours of the many enthusiastic geologists who within recent years have increased our knowledge of the glaciation of the Alpine Lands. At present, however, I am not so much concerned with the proofs of general glaciation as with the evidence that goes to show how the Alpine ground-moraines have been formed, and with the facts which have led certain observers to conclude that the Alps have endured several distinct glaciations within Pleistocene times.

Swiss geologists are agreed that the ground-moraines which clothe the bottoms of the great Alpine valleys, and extend outwards sometimes for many miles upon the low-grounds beyond, are of true glacial origin. Now these ground-moraines are closely similar to the boulder-clays of this country and northern Europe--like them, they are frequently tough and hard-pressed, but now and again somewhat looser, and less firmly coherent. Frequently also they contain lenticular beds, and more or less thick sheets of aqueous deposits--in some places the stony clays even exhibiting a kind of stratification--and ever and anon such water-assorted materials are commingled with stony clay in the most complex manner. These latter appearances are, however, upon the whole best developed upon the low-grounds that sweep out from the base of the Alps. The only question concerning the ground-moraines that has recently given rise to much discussion is the origin of the materials themselves. It is obvious that there are only three possible modes in which those materials could have been introduced to the ground-moraine: either they consist of superficial morainic débris which has found its way down to the bottom of the old glaciers by crevasses; or they may be made up of the rock-rubbish, shingle, gravel, etc., which doubtless strewed the valleys before these were occupied by ice; or, lastly, they may have been derived in chief measure from the underlying rocks themselves by the action of the ice that overflowed them.

The investigations of Penck, Blaas, Böhm, and Brückner appear to me to have demonstrated that the ground-moraines are composed mostly of materials which have been detached from the underlying rocks by the erosive action of the glaciers themselves. Their observations show that the regions studied by them in great detail were almost completely buried under ice--so that the accumulation of superficial moraines was for the most part impossible; and they advance a number of facts which prove positively that the ground-moraines were formed and accumulated under ice. I cannot here recapitulate the evidence, but must content myself by a reference to the papers in which this is fully discussed.[AK] These geologists do not deny that some of the material may occasionally have come from above, nor do they doubt that pre-existing masses of rock-rubbish and alluvial accumulations may also have been incorporated with the ground-moraines; but the enormous extent of the latter, and the direction of transport and distribution of the erratics which they contain cannot be thus accounted for, while all the facts are readily explained by the action of the ice itself, which used its sub-glacial débris as tools with which to carry on the work of erosion.

[AK] Penck: _Die Vergletscherung der deutschen Alpen._ Blaas: _Zeitschrift d. Ferdinandeums_, 1885. Böhm: _Jahrb. d. k. k. geol. Reichsanstalt_, 1885, Bd. xxxv., Heft 3. Brückner: _Die Vergletscherung d. Salzachgebietes, etc._, 1886.

Professor Heim and others have frequently asserted that glaciers have little or no eroding power, since at the lower ends of existing glaciers we find no evidence of such erosion being in operation. But the chief work of a glacier cannot be carried on at its lower end, where motion is reduced to a minimum, and where the ice is perforated by sub-glacial tunnels and arches, underneath which no glacial erosion can possibly take place; and yet it is upon observations made in just such places that the principal arguments against the erosive action of glaciers have been based. If all that we could ever know of glacial action were confined to what we can learn from peering into the grottoes at the terminal fronts of existing glaciers, we should indeed come to the conclusion that glaciers do not erode their rocky beds to any appreciable extent. But as we do not look for the strongest evidence of fluviatile erosion at the mouth of a river, but in its valley--and mountain-tracks, so if we wish to learn what glacier-ice can accomplish, we must study in detail some wide region from which the ice has completely disappeared. When this plan has been followed, it has happened that some of the strongest opponents of glacial erosion have been compelled by the force of the evidence to go over to the other camp. Dr. Blaas, for example, has been led by his observations on the glacial formations of the Inn valley to recant his former views, and to become a formidable advocate of the very theory which he formerly opposed. To his work and the memoirs by Penck, Brückner, and Böhm already cited, and especially to the admirable chapter on glacier-erosion by the last-named author, I would refer those who may be anxious to know the last word on this much-debated question.

The evidence of interglacial conditions within the Alpine lands continues to increase. These are represented by alluvial deposits of silt, sand, gravel, conglomerate, breccia, and lignites. Penck, Böhm, and Brückner find evidence of two interglacial epochs, and maintain that there have been three distinct and separate epochs of glaciation in the Alps. No mere temporary retreat and re-advance of the glaciers, according to them, will account for the various phenomena presented by the interglacial deposits and associated morainic accumulations. During interglacial times the glaciers disappeared from the lower valleys of the Alps--the climate was temperate, and probably the snow-fields and glaciers approximated in extent to those of the present day. All the evidence conspires to show that an interglacial epoch was of prolonged duration. Dr. Brückner has observed that the moraines of the last glacial epoch rest here and there upon löss, and he confirms Penck's observations in south Bavaria that this remarkable formation never overlies the morainic accumulations of the latest glacial epoch. According to Penck and Brückner, therefore, the löss is of interglacial age. There can be little doubt, however, that löss does not belong to any one particular horizon. Wahnschaffe[AL] and others have shown that throughout wide areas in north Germany it is the equivalent in age of the Upper Diluvium, while Schumacher[AM] points out that in the Rhine valley it occurs on two separate and distinct horizons. Professor Andreæ has likewise shown[AN] that there is an upper and lower löss in Alsace--each characterised by its own special fauna.

[AL] _Abhandl. z. geol. Specialkarte v. Preussen_, etc., Bd. vii. Heft 1; _Zeitschr. d. Zeutsch. geol. Ges._, 1885, p. 904; 1886, p. 367.

[AM] _Hygienische Topographie von Strassburg i. E._, 1885.

[AN] _Abhandl. z. geol. Specialkarte a. Elsass-Lothringen_, Bd. iv. Heft 2.

There is still considerable difference of opinion as to the mode of formation of this remarkable accumulation. By many it is considered to be an aqueous deposit; others, following Richthofen, are of opinion that it is a wind-blown accumulation; while some incline to the belief that it is partly the one and partly the other. Nor do the upholders of these various hypotheses agree amongst themselves as to the precise manner in which water or wind has worked to produce the observed results. Thus, amongst the supporters of the aqueous origin of the löss, we find this attributed to the action of heavy rains washing over and rearranging the material of the boulder-clays.[AO] Many, again, have held it probable that löss is simply the finest loam distributed over the low-grounds by the flood-waters that escaped from the northern inland-ice and the _mers de glace_ of the Alpine lands of central Europe. Another suggestion is that much of the material of the löss may have been derived from the denudation of the boulder-clays by flood-water, during the closing stages of the last cold period. It is pointed out that in some regions, at least, the löss is underlaid by a layer of erratics, which are believed to be the residue of the denuded boulder-clay. We are reminded by Klockmann[AP] and Wahnschaffe[AQ] that the inland-ice must have acted as a great dam, and that wide areas in Germany, etc., would be flooded, partly by water derived from the melting inland-ice, and partly by waters flowing north from the hilly tracts of middle Germany. In the great basins thus formed there would be a commingling of fine silt material derived from north and south, which would necessarily come to form a deposit having much the same character throughout.

[AO] Laspeyres: _Erläuterungen z. geol Specialkaret v Preussen_, etc., _Blatt. Gröbzig, Zörbig, und Petersberg_.

[AP] Klockmann: _Jahrb. d. k. preuss. geol. Landesanstalt für 1883_, p. 262.

[AQ] Wahnschaffe: _Op. cit._, and _Zeitschr. d. deutsch. geol. Ges._, 1886, p. 367.

From what I have myself seen of the löss in various parts of Germany, and from all that I have gathered from reading and in conversation with those who have worked over löss-covered regions, I incline to the opinion that löss is for the most part of aqueous origin. In many cases this can be demonstrated, as by the occurrence of bedding and the intercalation of layers of stones, sand, gravel, etc., in the deposit; again, by the not infrequent appearance of freshwater shells; but, perhaps, chiefly by the remarkable uniformity of character which the löss itself displays. It seems to me reasonable also to believe that the flood-waters of glacial times must needs have been highly charged with finely-divided sediment, and that such sediment would be spread over wide regions in the low-grounds--in the slackwaters of the great rivers and in the innumerable temporary lakes which occupied, or partly occupied, many of the valleys and depressions of the land. There are different kinds of löss or löss-like deposits, however, and all need not have been formed in the same way.

Probably some may have been derived, as Wahnschaffe has suggested, from denudation of boulder-clay. Possibly also, some löss may owe its origin to the action of rain on the stony clays, producing what we in this country would call "rain-wash." There are other accumulations, however, which no aqueous theory will satisfactorily explain. Under this category comes much of the so-called _Berglöss_, with its abundant land-shells, and its generally unstratified character. It seems likely that such löss is simply the result of sub-aërial action, and owes its origin to rain, frost, and wind acting upon the superficial formations, and rearranging their finer-grained constituents. And it is quite possible that the upper portion of much of the löss of the lower-grounds may have been re-worked in the same way. But I confess I cannot yet find in the facts adduced by German geologists any evidence of a dry-as-dust epoch having obtained in Europe during any stage of the Pleistocene period. The geographical position of our Continent seems to me to forbid the possibility of such climatic conditions, while all the positive evidence we have points to humidity rather than dryness as the prevalent feature of Pleistocene climates.

It is obvious, however, that after the flood-waters had disappeared from the low-grounds of the Continent, sub-aërial action would come into play over the wide regions covered by the glacial and fluvio-glacial deposits. Thus, in the course of time, these deposits would become modified,--just as similar accumulations in these islands have been top-dressed, as it were, and to some extent even rearranged. I am strengthened in these views by the conclusions arrived at by M. Falsan--the eminent French glacialist. Covering the plateaux of the Dombes, and widely spread throughout the valleys of the Rhone, the Ain, the Isère, etc., in France there is a deposit of löss, he says, which has been derived from the washing of the ancient moraines. At the foot of the Alps, where black schists are largely developed, the löss is dark grey, but west of the secondary chain the same deposit is yellowish, and composed almost entirely of silicious materials, with only a very little carbonate of lime. This _limon_ or löss, however, is very generally modified towards the top by the chemical action of rain--the yellow löss acquiring a red colour.

Sometimes it is crowded with calcareous concretions, but at other times it has been deprived of its calcareous element and converted into a kind of pulverulent silica or quartz. This, the true löss, is distinguished from another _lehm_, which Falsan recognises as the product of atmospheric action--formed, in fact, _in situ_, from the disintegration and decomposition of the subjacent rocks. Even this lehm has been modified by running water--dispersed or accumulated locally, as the case may be.[AR]

[AR] Falsan: _La Période glaciaire_, p. 81.

All that we know of the löss and its fossils compels us to include this accumulation as a product of the Pleistocene period. It is not of post-glacial age--even much of what one may call the "remodified löss" being of late Glacial or Pleistocene age. I cannot attempt to give here a summary of what has been learned within recent years as to the fauna of the löss. The researches of Nehring and Liebe have familiarised us with the fact that, at some particular stage in the Pleistocene period, a fauna like that of the alpine steppe-lands of western Asia was indigenous to middle Europe, and the recent investigations by Woldrich have increased our knowledge of this fauna. At what horizon, then, does this steppe-fauna make its appearance? At Thiede Dr. Nehring discovered in so-called löss three successive horizons, each characterised by a special fauna. The lowest of these faunas was decidedly arctic in type; above that came a steppe-fauna, which last was succeeded by a fauna comprising such forms as mammoth, woolly rhinoceros, _Bos_, _Cervus_, horse, hyæna, and lion.

Now, if we compare this last fauna with the forms which have been obtained from true post-glacial deposits--those deposits, namely, which overlie the younger boulder-clays and flood-accumulations of the latest glacial epoch, we find little in common. The lion, the mammoth, and the rhinoceros are conspicuous by their absence from the post-glacial beds of Europe. In place of them we meet with a more or less arctic fauna, and a high-alpine and arctic flora, which as we all know eventually gave place to the flora and fauna with which Neolithic man was contemporaneous. As this is the case throughout north-western and central Europe, we seem justified in assigning the Thiede beds to the Pleistocene period, and to that interglacial stage which preceded and gradually merged into the last glacial epoch. That the steppe-fauna indicates relatively drier conditions of climate than obtained when perennial snow and ice covered wide areas of the low-ground goes without saying, but I am unable to agree with those who maintain that it implies a dry-as-dust climate, like that of some of the steppe-regions of our own day.

The remarkable commingling of arctic- and steppe-faunas discovered in the Böhmer-Wald[AS] by Woldrich shows, I think, that the jerboas, marmots, and hamster-rats were not incapable of living in the same regions contemporaneously with lemmings, arctic hares, Siberian social voles, etc. But when a cold epoch was passing away the steppe-forms probably gradually replaced their arctic congeners, as these migrated northwards during the continuous amelioration of the climate.

[AS] Woldrich: _Sitzungsb. d. kais. Akad. d. W. math. nat. Cl._, 1880, p. 7; 1881, p. 177; 1883, p. 978.

If the student of the Pleistocene faunas has certain advantages in the fact that he has to deal with forms many of which are still living, he labours at the same time under disadvantages which are unknown to his colleagues who are engaged in the study of the life of far older periods. The Pleistocene period was distinguished above all things by its great oscillations of climate--the successive changes being repeated and producing correlative migrations of floras and faunas. We know that arctic and temperate faunas and floras flourished during interglacial times, and a like succession of life-forms followed the final disappearance of glacial conditions. A study of the organic remains met with in any particular deposit will not necessarily, therefore, enable us to assign these to their proper horizon. The geographical position of the deposit, and its relation to Pleistocene accumulations elsewhere, must clearly be taken into account. Already, however, much has been done in this direction, and it is probable that ere long we shall be able to arrive at a fair knowledge of the various modifications which the Pleistocene floras and faunas experienced during that protracted period of climatic changes of which I have been speaking. We shall even possibly learn how often the arctic, steppe-, prairie-, and forest-faunas, as they have been defined by Woldrich, replaced each other. Even now some approximation to this better knowledge has been made. Dr. Pohlig,[AT] for example, has compared the remains of the Pleistocene faunas obtained at many different places in Europe, and has presented us with a classification which, although confessedly incomplete, yet serves to show the direction in which we must look for further advances in this department of inquiry.

[AT] Pohlig: _Sitzungsb. d. niederrheinischen Gesellschaft zu Bonn_, 1884; _Zeitschr. d. deutsch. geolog. Ges._, 1887, p. 798. For a very full account of the diluvial European and northern Asiatic mammalian faunas by Woldrich, see _Mém. de l'Acad. des Sciences de St. Pétersbourg_, vii^e sér., t. xxxv., 1887.

During the last twenty years the evidence of interglacial conditions both in Europe and America has so increased that geologists generally no longer doubt that the Pleistocene period was characterised by great changes of climate. The occurrence at many different localities on the Continent of beds of lignite and freshwater alluvia, containing remains of Pleistocene mammalia, intercalated between separate and distinct boulder-clays has left us no other alternative. The interglacial beds of the Alpine Lands of central Europe are paralleled by similar deposits in Britain, Scandinavia, Germany, and France. But opinions differ as to the number of glacial and interglacial epochs--many holding that we have evidence of only two cold stages and one general interglacial stage. This, as I have said, is the view entertained by most geologists who are at work on the glacial accumulations of Scandinavia and north Germany. On the other hand, Dr. Penck and others, from a study of drifts of the German Alpine Lands, believe that they have met with evidence of three distinct epochs of glaciation, and two epochs of interglacial conditions. In France, while some observers are of opinion that there have been only two epochs of general glaciation, others, as, for example, M.

Tardy, find what they consider to be evidence of several such epochs. Others again, as M. Falsan, do not believe in the existence of any interglacial stages, although they readily admit that there were great advances and retreats of the ice during the Glacial period. M. Falsan, in short, believes in oscillations, but is of opinion that these were not so extensive as others have maintained. It is, therefore, simply a question of degree, and whether we speak of oscillations or of epochs, we must needs admit the fact that throughout all the glaciated tracts of Europe, fossiliferous deposits occur intercalated among glacial accumulations. The successive advance and retreat of the ice, therefore, was not a local phenomenon, but characterised all the glaciated areas. And the evidence shows that the oscillations referred to were on a gigantic scale.

The relation borne to the glacial accumulations by the old river alluvia which contain relics of palæolithic man early attracted attention. From the fact that these alluvia in some places overlie glacial deposits, the general opinion (still held by some) was that palæolithic man must needs be of post-glacial age. But since we have learned that all boulder-clay does not belong to one and the same geological horizon--that, in short, there have been at least two, and probably more, epochs of glaciation--it is obvious that the mere occurrence of glacial deposits underneath palæolithic gravels does not prove these latter to be post-glacial. All that we are entitled in such a case to say is simply that the implement-bearing beds are younger than the glacial accumulations upon which they rest. Their horizon must be determined by first ascertaining the relative position in the glacial series of the underlying deposits. Now, it is a remarkable fact that the boulder-clays which underlie such old alluvia belong, without exception, to the earlier stages of the Glacial period. This has been proved again and again, not only for this country but for Europe generally.

I am sorry to reflect that some twenty years have now elapsed since I was led to suspect that the palæolithic deposits were not of post-glacial but of glacial and interglacial age. In 1871-72 I published a series of papers in the _Geological Magazine_ in which were set forth the views I had come to form upon this interesting question. In these papers it was maintained that the alluvia and cave-deposits could not be of post-glacial age, but must be assigned to pre-glacial and interglacial times, and in chief measure to the latter. Evidence was led to show that the latest great development of glacier-ice in Europe took place after the southern pachyderms and palæolithic man had vacated England--that during this last stage of the Glacial period man lived contemporaneously with a northern and alpine fauna in such regions as southern France--and lastly, that palæolithic man and the southern mammalia never revisited north-western Europe after extreme glacial conditions had disappeared. These conclusions were arrived at after a somewhat detailed examination of all the evidence then available--the remarkable distribution of the palæolithic and ossiferous alluvia having, as I have said, particularly impressed me.

I coloured a map to show at once the areas covered by the glacial and fluvio-glacial deposits of the last glacial epoch, and the regions in which the implement-bearing and ossiferous alluvia had been met with, when it became apparent that the latter never occurred at the surface within the regions occupied by the former. If ossiferous alluvia did here and there appear within the recently glaciated areas it was always either in caves, or as infra- or interglacial deposits. Since the date of these researches our knowledge of the geographical distribution of Pleistocene deposits has greatly increased, and implements and other relics of palæolithic man have been recorded from many new localities throughout Europe. But none of this fresh evidence contradicts the conclusions I had previously arrived at; on the contrary, it has greatly strengthened my general argument.

Professor Penck was, I think, the first on the Continent to adopt the views referred to. He was among the earliest to recognise the evidence of interglacial conditions in the drift-covered regions of northern Germany, and it was the reflections which those remarkable interglacial beds were so well calculated to suggest that led him into the same path as myself. Dr. Penck has published a map[AU] showing the areas covered by the earlier and later glacial deposits in northern Europe and the Alpine Lands, and indicating at the same time the various localities where palæolithic finds have occurred, and in not a single case do any of the latter appear within the areas covered by the accumulations of the last glacial epoch.

[AU] _Archiv für Anthropologie_, Bd. xv. Heft 3, 1884.

A glance at the papers which have been published in Germany within the last few years will show how greatly students of the Pleistocene ossiferous beds have been influenced by what is now known of the interglacial deposits and their organic remains. Professors Rothpletz[AV] and Andreæ,[AW] Dr. Pohlig[AX] and others, do not now hesitate to correlate with those beds the old ossiferous and implement-bearing alluvia which lie altogether outside of glaciated regions.

[AV] Rothpletz: _Denkschrift d. schweizer. Ges. für d. gesammt. Nat._, Bd. xxviii. 1881.

[AW] Andreæ: _Abhandl. z. geolog. Specialkarte v. Elsass-Lothringen_, Bd. iv. Heft 2, 1884.

[AX] Pohlig: _op. cit._

The relation of the Pleistocene alluvia of France to the glacial deposits of that and other countries has been especially canvassed. Rothpletz, in the paper I have cited, includes these alluvia amongst the interglacial deposits, and in the present year (1889) we have an interesting essay on the same subject by the accomplished secretary of the Anthropological and Archæological Congress which met recently in Paris. M. Boule[AY] correlates the palæolithic cave- and river-deposits of France with those of other countries, and shows that they must be of interglacial age. His classification, I am gratified to find, does not materially differ from that given by myself a number of years ago. He is satisfied that in France there is evidence of three glacial epochs and two well-marked interglacial horizons. The oldest of the palæolithic stages of Mortillet (Chelléenne) culminated according to Boule during the last interglacial epoch, while the more recent palæolithic stages (Moustérienne, Solutréenne, and Magdalénienne) coincided with the last great development of glacier-ice. The Palæolithic age, so far as Europe is concerned, came to a close during this last cold phase of the Glacial period.

[AY] Boule: _Revue d'Anthropologie_, 1889, t. 1.

There are many other points relating to glacial geology which have of late years been canvassed by Continental workers, but these I cannot discuss here. I have purposely indeed restricted my remarks to such parts of a wide subject as I thought might have interest for glacialists in this country, some of whom may not have had their attention directed to the results which have recently been attained by their fellow-labourers in other lands. Had time permitted I should gladly have dwelt upon the noteworthy advances made by our American brethren in the same department of inquiry. Especially should I have wished to direct attention to the remarkable evidence adduced in favour of the periodicity of glacial action. Thus Messrs. Chamberlin and Salisbury, after a general review of that evidence, maintain that the Ice Age was interrupted by one chief interglacial epoch and also by three interglacial sub-epochs or episodes of deglaciation. These authors discuss at some length the origin of the löss, and come to the general conclusion that while deposits of this character may have been formed at different stages of the Glacial period, and under different conditions, yet upon the whole they are best explained by aqueous action. Indeed a perusal of the recent geological literature of America shows a close accord between the theoretical opinions of many Transatlantic and European geologists.

Thus as years advance the picture of Pleistocene times becomes more and more clearly developed. The conditions under which our old palæolithic predecessors lived--the climatic and geographical changes of which they were the witnesses--are gradually being revealed with a precision that only a few years ago might well have seemed impossible. This of itself is extremely interesting, but I feel sure that I speak the conviction of many workers in this field of labour when I say that the clearing up of the history of Pleistocene times is not the only end which they have in view. One can hardly doubt that when the conditions of that period and the causes which gave rise to these have been more fully and definitely ascertained we shall have advanced some way towards the better understanding of the climatic conditions of still earlier periods. For it cannot be denied that our knowledge of Palæozoic, Mesozoic, and even early Cainozoic climates is unsatisfactory. But we may look forward to the time when much of this uncertainty will disappear. Meteorologists are every day acquiring a clearer conception of the distribution of atmospheric pressure and temperature and the causes by which that distribution is determined, and the day is approaching when we shall be better able than we are now to apply this extended meteorological knowledge to the explanation of the climates of former periods in the world's history.

One of the chief factors in the present distribution of atmospheric temperature and pressure is doubtless the relative position of the great land- and water-areas; and if this be true of the present, it must be true also of the past. It would almost seem, then, as if all one had to do to ascertain the climatic conditions of any particular period, was to prepare a map depicting with some approach to accuracy the former relative position of land and sea. With such a map could our meteorologists infer what the climatic conditions must have been? Yes, provided we could assure them that in other respects the physical conditions did not differ from the present. Now there is no period in the past history of our globe the geographical conditions of which are better known than the Pleistocene. And yet, when we have indicated these upon a map, we find that they do not give the results which we might have expected. The climatic conditions which they seem to imply are not such as we know did actually obtain. It is obvious, therefore, that some additional and perhaps exceptional factor was at work to produce the recognised results.

What was this disturbing element, and have we any evidence of its interference with the operation of the normal agents of climatic change in earlier periods of the world's history? We all know that various answers have been given to such questions. Whether amongst these the correct solution of the enigma is to be found, time will show. Meanwhile, as all hypothesis and theory must starve without facts to feed on, it behoves us as working geologists to do our best to add to the supply. The success with which other problems have been attacked by geologists forbids us to doubt that ere long we shall have done much to dispel some of the mystery which still envelopes the question of geological climates.