But although, as already mentioned, the Interglacial hypothesis in its simpler form has many supporters in this country, I do not think that the above scheme in its entirety has yet found any adherents among British glacialists. Usually, when beds supposed to be of interglacial age have been described by other workers, it has been implied that only a single interval of milder conditions was in mind; and even in the exceptional cases where several different boulder-clays separated by sand and gravel have been held to represent as many different epochs of glaciation, it is rare that any attempt has been made, except by Prof. Geikie himself, to classify the supposed events in accordance with the scheme. I suppose that most field-workers have felt, like myself, that while some part of the classification might possibly be sustained, this finished arrangement of the admittedly imperfect evidence was too artificial to be accepted with confidence, and that it was inadvisable to allow one's self to be hampered, in an inherently difficult task, with further difficulties that, after all, might, like "the word Bearbaiting, be "carnal and of man's creating.' On the other hand, partly, no doubt, from the persuasive manner in which its author has presented his case and his courteous readiness to meet objections, but still more from the vast extent of the field drawn upon for the argument, the scheme has aroused less active criticism than it has, in my opinion, deserved. The critic has shrunk from the magnitude of the task of testing it in all its parts, while to pick out the local flaws in any particular part has seemed invidious.

In taking this scheme as the basis of my examination into the evidence, I am aware that the local limitations which I have set myself will be held to impair the validity of my conclusions. But as there is at present in every glaciated country the same confusion of opinion on the Interglacial problem as in our own, and the same discussion upon the fundamental value of the evidence, it appears to me that we can find strong justification for considering our own problem on its separate merits. And the necessity for a re-sifting of the British evidence is the more urgent since it is frequently taken for granted in the discussions abroad that there is a well-established glacial sequence in Britain, which can be called in to support the argument for other lands.

The Interglacial Problem in Other Countries.

It will serve to illustrate the condition of the problem in other countries if I refer briefly to some of the literature which happens to have come under my notice, though can rarely claim sufficient knowledge of the foreign work to discuss its value.

Norway. In Norway there appears to be no direct evidence for interglacial epochs, though the existence of one such epoch is supposed to be indicated by a change in the direction of ice-flow, and by the presence of an arctic flora at the base of the Danish peat-mosses which is absent in Norway. By Dr. A. M. Hansen the superficial deposits are classed as follows:-preglacial: proteroglacial interglacial: deuteroglacial: and postglacial.

Sweden,-In Sweden, and, I believe, also in Denmark, the Interglacial hypothesis is generally accepted, at least to the extent of one epoch of deglaciation, but is strenu1 "Period" in original; op. cit.: probably misprints for "Epoch." A. M. Hansen, "The Glacial Succession in Norway." Journ. Geol.. vol. ii. (1894), pp. 123–144.

46

ously opposed by Dr. N. O. Holst, who states his conviction, based on the result of his observations in Greenland, that the so-called interglacial sands and gravels and the upper moraine of Sweden represent the residual products of the ice-sheet that laid down the "lower moraine as a ground-moraine. He also embraced the drifts of North Germany in this explanation.1

Germany.-In Germany, the discussion on the "Interglacialismus is still in active progress. The idea of one interglacial epoch, corresponding to the " Helvetian" of Prof. J. Geikie's scheme, is widely entertained; and some geologists, influenced largely by evidence in the Alps, think that an earlier interglacial stage (="Norfolkian"), preceded by a stage of glaciation (=" Scanian "), may have to be admitted, though the German evidence is acknowledged to be imperfect. But Prof. Geikie's interpretation of the North German drifts, on which he seeks to establish the Neudeckian Interglacial" and the "Mecklenburgian Glacial" epochs, is strongly and authoritatively opposed. In a searching criticism of these views Dr. K. Keilhack, of the Prussian Geological Survey, states that no reason has been found, by himself or his colleagues, for the proposed separation of the upper drifts into these separate epochs; and he remarks that, on similar grounds, the so-called last glacial epoch would have to be divided into four if not five epochs, so that even the most fanatical advocate for as many glacial periods as possible would be terrified.' Prof. Geikie, in his reply to this criticism," brings forward the British evidence to establish the case especially weak, and we in this country had expected that in Germany. But, as we shall see, this evidence is the stronger proof lay in Germany, While the supporters of the Interglacialismus thus uncertain how much of the scheme they will accept, there are other geologists in Germany who repudiate the singleness hypothesis in its entirety, and hold for the " of the Ice-Age." Among these I may mention Prof. E. Geinitz, whose vigorous attack has been supported by Dr. W. Wolff, in a useful summary of the discussion, which contains many references to the literature."

are

enormous

Russia.-In Russia, again, opinion is divided, and the evidence brought forward in favour of the Interglacial idea has been adversely criticised by Mr. S. Nikitin, of the Russian Geological Survey, who considered that, whatwestward, ever may have been the conditions farther oscillations of the ice-margin would suffice to explain the facts observed in this outer portion of the glaciated area. The Alps. In the Alps there appears to be definite evidence for several periods of advance of with intervening glaciers from the mountain valleys, periods of great recession, and these are supposed to correspond to glacial and interglacial epochs in Northern Europe; but there has been much difference of opinion respecting this evidence and its interpretation. By Profs. A. Penck and E. Brückner, who have systematically investigated the phenomena, the ice-movements are held to indicate four separate epochs of glaciation, with three, or perhaps four, warm interglacial epochs.' Not having yet found an opportunity to make myself sufficiently acquainted with the evidence, I may not fully recognise its importance; but it appears to me that the factors governing the glaciation of this Alpine region may have been very different from those that controlled the lowland

324-339

4Die Einheitlichkeit der quartären Eiszeit." Neues Jahrb. j. Mineralogie, &c., xvi. (1902), pp. 1-98, and other papers. 5 Zur Kritik der Interglacial-Hypothese." Naturwiss. Wochenschrift. Neue Folge, Bd. ii. No. 26 (1903), 14 pp.

6 "Sur la constitution des dépôts quaternaires en Russie, &c." Rep. Congrès Internat, d'Archéologie," Moscou, 1892.

7 Die Alpen im Eiszeitalter.' Leipzig (19015), not yet complete; for convenient summary see "Glazialexkursion in die Ostalpen." No. 12 of "Guides to Excursions of the Geological Congress," Vienna, 1903.

glaciation.

And although it is certain that the great extension of the Alpine glaciers was due to the same glacial conditions that gave rise to the lowland ice-sheets of Northern Europe, I do not regard it as a necessary consequence that advances and retreats of the ice should occur simultaneously in both regions. Variation in the relative amount of snowfall over the glaciated areas during the course of the Glacial Period, for which there is much evidence, would be likely to produce great effects in the high-lying reservoirs of the Alps; and at the latitude of this region we should expect rapid recession of the lowlevel glaciers in response to diminished supply. To distinguish between the effects of oscillations in precipitation and of oscillations in temperature under such conditions must be peculiarly difficult.

North America.-In North America, where both the drifts and their literature attain gigantic proportions, the state of opinion is closely analogous to that among ourselves. It is agreed by all that during the Glacial Period there were very extensive oscillations in the borders of the ice-sheets; and by some geologists some of the stages of recession are supposed to represent mild epochs of actual "deglaciation"; while others, fewer in number, among whom Mr. Warren Upham and Dr. G. F. Wright have been the most active, regard these stages as of minor consequence, and advocate the essential unity of the glaciation. And between the two extremes stand the great majority of the workers in American glacial geology, who refrain from expressing positive opinions, but mostly lean toward the idea of at least one great interruption in the glaciation. Some of the suggested schemes of classification are fully as elaborate and complex as that proposed for Europe, but it seems to be recognised that these are only of local value. Prof. T. C. Chamberlin and his fellow-workers in the North-Central States have, however, adopted a sequence based on the successive advance of different ice-lobes, which is believed to be of wider application; and Prof. Chamberlin has tentatively suggested that some of these divisions may have their counterpart in the European scheme, but is careful to show that the correlation must at present remain entirely hypothetical,2 especially as the proposed American grouping may itself require modification.

It is well established that the American ice-sheets, like their European equivalents, radiated from several distinct centres that attained their maximum influence consecutively, and not simultaneously. Of these the "Laurentide" and the "Keewatin" sheets had their radiants over comparatively low ground east and west of Hudson Bay, while the Cordilleran sheet spread outward from the Western Mountains. In his general discussion of the glacial phenomena of North-Western Canada, Mr. J. B. Tyrrell concludes that the Cordilleran sheet had reached its greatest extent and had retired before the boulder-clay of the Keewatin sheet was laid down; and that the Keewatin sheet, in turn, had gone south to its farthest limit, and had retired for many hundreds of miles-more than half-way to its gathering ground-before the Laurentide sheet had reached its greatest extension.

If these conclusions be accepted, they must imply that at least in some cases the recession of the ice-lobes was due to causes acting locally, and not to mild interglacial periods affecting the whole hemisphere. The phenomena of invasion by successive ice-lobes in the peripheral regions might thus be readily explained without recourse to the Interglacial hypothesis.

Most of the detailed evidence brought forward in America to support the Interglacial idea is as fragmentary and unconvincing as that of our own country. But there is one notable exception, to which I must particularly refer, as it has been investigated by a Research Committee of the Association, and has, moreover, come under my

1 eg, "The Diversity of the Glacial Period in Long Island," by A. C. Veatch. Journ. Geol., vol. xi. (1903), pp. 762-776.

2 Classification of American Glacial Deposits." Journ. Geol., vol. iii. (189), pp. 270-277, and in J. Geikie's "Great Ice Age," 3rd. ed., chap. xli. See also Chamberlin and Salisbury's recent text-book, " Geology: Earth History," vol. iii. chap. xix. (London, 1906).

3The Glaciation of North-Central Canada." Journ. Geol., vol. vi. (1898), pp. 147-161; and "The Genesis of Lake Agassiz," ibid., vol. iv. (1896), pp. 811-815.

personal observation. In this case the interglacial deposit, first described by Dr. G. J. Hinde, are magnificently eXposed in cliff sections at Scarboro' Heights, on the shores of Lake Ontario, near Toronto. When I visited these sections under the guidance of my friend Prof. A. P. Coleman, in 1897, they impressed me strongly, inasmuch as they afforded the kind of evidence for which one had sought in vain in Britain. The section around Scarboro Heights reveals a great mass of fossiliferous stratified deposits, more than 180 feet thick, consisting in the lower part of slightly peaty clays, and in the upper part of sands; and these deposits are overlain by a complex series of boulder-clays, with intercalated beds of sand and gravel, attaining a thickness of at least 200 feet. The fossiliferous clays are the lowest beds seen in the cliff section, but beds belonging to the same series, that are exposed in the Don Valley, on the outskirts of Toronto, are underlain by a few feet of boulder-clay, so that it seems to be beyond question that the Scarboro' beds were deposited in an interval between two epochs of glaciation. In their upper part these beds contain a flora and fauna indicating a col climate, but in their lower portion some of the plants and freshwater shells no longer exist so far north as Canada and are therefore considered to denote a climate warmer than that of the present day. On this and other evidence it is clear that during the course of the Glacial Period the whole of the district was for a considerable time released from the ice-sheets which previously and afterwards covered it. Moreover, in the opinion of Prof. Coleman, some of the plants and shells of the warm-climate beds denote conditions that would be incompatible with the persistence of ice-sheets anywhere in Canada and it this be so, then we here have proof for at least one interglacial epoch. But I still permit myself to feel doubt regarding this last-mentioned deduction, as the shells and plants in question, which have their present habitat in the Middle United States, even yet endure winters of consider. able severity; and there are certain factors in the composition of the beds and their altitude above Lake Ontario that justify caution. It is, however, mainly from my knowledge of this "Toronto formation," and of the Kirmington section in England, presently to be discussed, that I still maintain an undecided attitude in respect to the Interglacial hypothesis in its simpler form.

Further support to the probability of an interglacial epoch has been adduced from the history of the great lakes which formerly existed in the Interior Basin of the Western States. It has been shown by the researches of G. K. Gilbert in the "Lake Bonneville" basin and of I. C. Russell in that of "Lake Lahontan," that there were two separate epochs, during which these enormous basins were filled with water, and an intervening arid epoch. during which they were dried up. The region is one in which the actual glacial phenomena are restricted to the mountain valleys; but as it seems evident that the lakes were associated in some way with the Glacial Period, the two stages of extension are supposed to represent two distinct epochs of glaciation, separated by a long interglacial drought. The correlation, however, has difficulties which are very impartially discussed by Gilbert and Russell; and it will not admit of more than one interglacial episode.

The Interglacial Problem in the British Islands. Let us now consider the application of the Interglacial hypothesis to our own land.

The task of following up the evolution of Prof. Geikie's scheme through its various phases, though instructive, is very confusing-one might even say irritating-by reason of the continual changes of correlation which its author has suggested in sorting out the British drift deposits into this orderly sequence. Our East Coast bouider-clays, for example, were at one time held to cover four glacial epochs.

1 Prof. A. P. Coleman, Reps. British Assoc. for 1898, pp. 522-29: = 1899, pp. 411-414; for 1900, pp. 328-40; also (summary and discussion) "Glacial and Interglacial Beds near Toronto," Journ. Grol val u (1901), pp. 285-310.

The Duration of the Toronto Interglacial Period. Amerint Geologist, vol. xxix. (1902). p. 79.

3 "Lake Bonneville." Monogr. U.S. Geol. Survey, vol. i. (150)). 4 "Lake Lahontan." Monogr. U.S. Geol. Survey, vol. xi. (188)

and their associated gravels to mark three mild interglacial epochs; and all except the first glaciation were supposed to be represented in the boulder-clays of Lancashire and Cheshire.' Then, somewhat vaguely, it was allowed that perhaps there were only three separate glaciations on the east coast, with a minor episode of recession of the ice-margin; and the Lancashire and Cheshire boulder-clays were correlated with the two later of these glacial epochs.2 But subsequently we are reduced in the eastern district to two epochs of glaciation, with one mild interval, of which the equivalents are all recognised also in the north-west of England.'

While these and other similar changes may show a laudable desire of their author to keep pace with the growth of definite information, I cannot help feeling that they also show the premature character of the whole scheme, and a flexibility in it that justifies suspicion. Moreover, in spite of these frequent changes in the correlation and this local lopping off of glacial and interglacial episodes, we find, with surprise, that the number of separate epochs in the classification has not diminished, but has actually increased, by regrowth in fresh places. This, again, may betoken the inherent vitality of the scheme, in which case it will gain strength from every readjustment; but it must certainly also denote the weakness of its original basis. In considering its application to this country we will begin by glancing at the evidence for the two earliest epochs of the classification.

Norfolkian" (First

** Scanian" (First Glacial) and "
Interglacial) Epochs.

It is acknowledged that the First Glacial Epoch is not represented in Britain by any boulder-clay or other evidence of land glaciation, but is based mainly upon the supposed existence of a great Baltic glacier which overflowed the southern part of the Scandinavian peninsula from southeast to north-west, a direction differing widely from that of the later ice-sheets. This glaciation of Scania is supposed to have been contemporaneous with the deposition of the Chillesford Clay and Weybourn Crag of Norfolk, which contain a marine fauna indicative of cold conditions. The Forest Bed series of Norfolk, with its temperate land fauna and flora, is then interpreted as the product of a mild interglacial epoch ("Norfolkian ") intercalated between the "Scanian " glaciation and the more severe Saxonian glaciation which followed; and it is implied that during this mild stage the earlier ice-sheet vanished.

So far as I can gather, the recognition of the "Scanian" ice-sheet rests on dubious grounds, being based chiefly on the disputed supposition that the lower boulder-clay of North Germany is not the equivalent of the lower boulderclay of Sweden, but of a subsequent Swedish boulder-clay. For the Norfolkian" disappearance of the first Swedish ice-sheet no direct evidence is forthcoming, since it is acknowledged that no interglacial deposits representing this stage have been found in Sweden. But the Norfolk Forest Bed is here brought into the argument to prove the 'deglaciation "--so that the Scandinavian geologist is invited to accept the "First Interglacial Epoch mainly on the supposed strength of the British evidence, while the British geologist is expected to acknowledge the First Glacial Epoch on the supposed strength of the Swedish evidence. This method of argument might have weight if the evidence afforded by either region were perfectly definite. But in the present instance the conclusion that the Forest Bed represents an interglacial episode is not acceptable to the observers who have the fullest knowledge of the Norfolk sections, Mr. Clement Reid pointing out that the enclosing of the North Sea by the union of Britain with the southward continental land affords an adequate explanation of the apparent climatal discrepancy between the fauna of the sea and that of land; while Mr. F. W.

1 "Great Ice Age," 2nd ed (1877), p. 393. "Prehistoric Europe" (1881), pp. 263-266.

"Great Ice Age," 3rd. ed. (1894), chaps. xxv. and xxvi., and Journ. Geol. (supra cit.).

"The Pliocene Deposits of Britain." Mem. Geol. Survey (1890), pp. 186-190.

Harmer shows the probability of the transport of southern relics into this old estuarine deposit by river-drifting.'

It has, indeed, been long recognised that the marine Pliocene deposits of eastern England present us with an intelligible chain of evidence for the gradual and uninterrupted approach of the Glacial Period; and to break this chain will require stronger reasons than have yet been adduced. From the Coralline Crag, with seas warmer than at present, to the Red Crag and Norwich Crag, with a northern element steadily gaining ground in the fauna, we pass upward to the Chillesford Clay and Weybourn Crag, wherein this element becomes predominant. Then follows the period of slight elevation indicated by the Forest Bed, wherein, along with its temperate-climate fauna, such northern forms as the musk ox and glutton are associated; and finally we gain just a glimpse of truly arctic conditions in the Leda myalis bed and the Arctic freshwater bed, immediately before the advent of the great ice-sheet that relentlessly blotted out both land and sea.

"Saxonian "

(Second Glacial), "Helvetian (Second Interglacial), and "Polandian" (Third Glacial) Epochs. Regarding the glacial severity of the ensuing stage-the "Saxonian Epoch of Prof. Geikie's scheme-all are agreed; and from this stage onward to the close of the "Glacial Period as usually understood, or to the close of the "Polandian Epoch of the proposed classification, our difficulties of interpretation arise not from lack of evidence, but rather from its superabundance and local intricacies.

It happens, fortunately, that the great bulk of our British drifts, with the exception only of those in certain mountainous districts, are now included by Prof. Geikie within the two above-mentioned glacial epochs and the intervening Helvetian Interglacial Epoch.' Therefore, in dealing more particularly with the deposits assigned to these three epochs in certain typically glaciated districts, we shall bring under consideration a considerable portion of the drifts of our islands, and shall obtain results which can be applied to many other areas in which the structure of the glacial deposits is essentially similar. The first district to be considered shall be that which lies nearest us; and in discussing the drifts of East Yorkshire I propose to interweave some personal opinions that I have deduced from the facts, which will afterwards be given wider application.

EAST YORKSHIRE DRIFTS.-The long cliff-sections between the Humber and the Tees constitute one of the best exposures of lowland drifts in Britain, or even in Europe. They fortunately include some deposits which reveal the conditions prevailing in the neighbouring part of the North Sea basin just before the great glaciation; and they therefore enable us without interruption to continue the history begun in East Anglia.

The old cliff of chalk and the marine beach at its foot which lie buried at Sewerby, on the southern side of Flamborough Head, under sheets of boulder-clay and gravel, prove to us that at the very beginning of glacial times the North Sea still held possession of its basin, and with a surprisingly slight difference from its present level. A few far-transported stones in the old beach denote that ice-floes sometimes drifted southward into Holderness Bay; while the bones of animals in the shingle, and in the blown sand which overlies it, prove that among the denizens of the neighbouring land were the elephant (E. antiquus), rhinoceros (R. leptorhinus), hippopotamus (H. amphibius), and bison. This fauna is frequently considered to be proof of mild conditions of climate; but from the mode of its occurrence in this and other places, I can find no reason to doubt that these animals inhabited the country, perhaps as seasonal migrants, until the time that it was actually covered by the encroaching ice-sheets. And here I may note my opinion, that throughout the discussion of our glacial deposits too much weight has been allowed to the deductions regarding climate based upon scanty indications afforded by the ancient fauna and flora.

1 "The Later Tertiary History of East Anglia." Proc. Geol. Assoc., vol. xvii. (1902), p. 449.

We know little regarding the range of adaptability possessed by the forms in the past, and can judge only from their present habitat, which is generally governed by many other factors besides climate; moreover, it is granted that species already established, when subjected to gradual change, will persist for long in circumstances that would have effectively barred their introduction. In the Upper Zambesi Valley last year I was more impressed with the cold of the nights than with the heat of the days; and even at that latitude the sturdy hippopotamus in his nocturnal raids must experience a temperature occasionally descending below freezing-point.

It took us long to break away from the established conviction that the fossil elephant and rhinoceros could not have existed in a cold climate; and the same conviction still lingers with respect to their companion, the hippopotamus. But the far-travelled stones in the Sewerby beach and in the beaches of the same age in the south of Ireland are evidence that the British seas were already cold enough to carry ice-floes while these large mammals still tenanted the land.

The next event indicated by the Sewerby section is a slight elevation of the land. Then the traces of an increasingly rigorous climate become conspicuous, for the sand-dunes which had been banked against the old cliff are covered by chalky rubble containing a few land shells1; and this material, like the corresponding "head" which covers the ancient beaches of the south of Ireland and the south-west of England, appears to represent the frostsplintered rock washed down from the rock slopes during the season of thaw.

According to my reading of the evidence, it was during this time that the bed of the North Sea was gradually filled by a great ice-lobe that spread southward and outward along the basin, slowly but irresistibly churning up and dragging forward the old sea-floor as part of its ground-moraine. When it impinged upon the rising ground of eastern Britain the progress of this sheet was arrested and part of its burden left in the form of the lowest boulder-clay-the "Basement Clay of Yorkshire and the "Cromer Till" of Norfolk. In Yorkshire this boulderclay frequently includes huge transported masses of Secondary strata, which still maintain their identity, in some cases even to their bedding planes; and along with these we sometimes find patches of the material of the old sea-floor which have similarly escaped destruction. More frequently the preexisting deposits from which the boulder-clay has been derived have been thoroughly kneaded together, and fragments of Pleistocene shells are then scattered through its mass, along with fossils derived from the Secondary and older rocks.

In adopting the hypothesis that the Basement boulderclay represents the ground-moraine of an ice-sheet we may consider briefly the probable conditions under which this "East British ice-lobe " was accumulated. Whether the elevation subsequent to the stage represented by the infraglacial beaches was sufficient to drain off the shallow seas around our islands is uncertain, but it must, at any rate, have restricted their area and rendered them still shallower; and it is unlikely that there was then any southward connection of the North Sea with the English Channel. The climate by this time had become such that permanent snow-caps could accumulate in the northern parts of our country at elevations not much above present sea-level. Indeed, I am inclined to think that the climate may have been actually colder at this time than during any of the later phases of the Glacial Period, and that the stage of maximum glaciation lagged considerably behind the stage of minimum temperature. Under these conditions, with the snowfall on the uplands always slowly drawing away in ice-streams to the basins, and there accumulating, it is inevitable that the enclosed basins would eventually become ice-covered, any open water within them being in time obliterated, either directly by the encroaching glaciers, or indirectly by the packing of bergs and floes, until the basins themselves possessed a surface upon which the snowfall could accumulate. Thus the basins became

1 Lamplugh, Proc. Yorks. Geol Sec., vol. xv. (1903), pp. 91 95.

great reservoirs of ice, in which the supplies from the surrounding uplands received important augmentation by direct accretion of snowfall;-reservoirs, moreover, containing a substance sufficiently rigid not to require retaining walls; so that, in time, the surface of the ice within the basins rose higher than many parts of the rim. The general movement of the mass within its reservoir then became dependent mainly upon its own configuration, and only secondarily upon the shape of the solid ground.

These conditions in the North Sea basin had their parallel in the basin of the Irish Sea, in which the " West British ice-lobe " was developed; and on the low interior plain of Ireland, where the similar though smaller "Ivernian" sheet held possession.

Now, the crux of the Interglacial problem, so far as the British Islands are concerned, lies in the question whether these huge reservoirs, after their first filling, were completely emptied during the supposed interglacial epoch of warmth named by Prof. Geikie the " Helvetian," and were afterwards refilled for the later Polandian glaciation, in which, on the evidence of the upper boulder-clays, it is generally agreed that ice-sheets from the basins again closed in upon the land. It is this one interglacial or "middle glacial epoch only that most of the British supporters of the hypothesis have demanded, and have attempted to establish in the East Yorkshire sections.

"

For my own part, although I have sought long and carefully for evidence of this great interglacial episode in the Yorkshire drifts, and at first with the belief that such evidence must surely be somewhere forthcoming, my search has not only failed to bring to light any adequate proof of its reality, but has yielded many facts which I cannot explain otherwise than by recognising that the ice-lobe continued to occupy the basin of the North Sea during the deposition of the beds claimed as interglacial, though its margin had for a time shrunk considerably within its earlier limits.

The "Purple" Boulder-Clays and Stratified Drifts.—The drifts overlying the Basement Clay in East Yorkshire consist of a complex and very variable series, in which bands of boulder-clay predominate in some places and lenticular sheets of well-stratified material in others. In the cliff-sections of the Holderness plain certain bands of boulder-clay, known as the Upper and Lower Purple Clays, are persistent for many miles; but when the series approaches the rising ground of the Wolds the individuality of the beds is lost, and they are often replaced entirely by irregular mounds of sand and gravel.

2

I began work on these sections with the then-prevalent idea that every separate band of boulder-clay above the Basement Clay might indicate a separate glacial epoch, and that warm interglacial epochs might be represented by the partings of sand and gravel between these boulderclays; and the object of one of my early papers was to show that more of these divisions were present than had found place in the scheme of classification then in vogue. But after struggling for a time under an ever-increasing load of epochs I was compelled, in tracing the separate bands northwards, to recognise, as my friend Mr. J. R. Dakyns had previously recognised, that the whole series underwent protean changes, the boulder-clays sometimes splitting into numerous shreds amid thick sheets of sand and gravel, at other times merging into a single mass to the exclusion of all stratified material, and not rarely presenting a passage from uncompromising " till "to stratified gravel, sand, and clay. Hence I was driven to conclude that stratified and unstratified drift must often have been forming simultaneously at places very little distance apart; and on finding, also, that the whole of the deposits between the Basement Clay and the Upper or Hessle Clay were not only knit together in this fashion, but were similarly interwoven with the top and bottom of these boulder-clays, I had finally to abandon the Interglacial hypothesis altogether so far as the coast-sections were concerned. I mention this experience in order to show that my present

1 "On the Divisions of the Glacial Beds in Filey Bay." Proc. Varks. Geol. Sec., vol. vii. (1879), pp. 167-177. "Glacial Beds at Bridlington." Ibid., vol. vii. (1879), pp. 123–127.

scepticism respecting the Helvetian Interglacial Epoch is based, not upon any preconceived objection to the idea, but upon the failure of the hypothesis when I have put it to the test in this and other districts; and I find also that my experience in this particular runs parallel with that of many other investigators of the so-called "middle glacial" deposits of England.

Marine Detritus in Glacial Gravels. From certain characters of the moundy gravels on Flamborough Head and in Holderness, such as their rudely linear arrangement, their indifference to the contours, and their relation to the middle or Purple boulder-clays, it appears most probable that they represent the material deposited along the margin of the ice-sheet by the surface-waters flowing from it and from the adjacent land.' From the occurrence of more or less fragmentary marine shells in them, the gravels were, however, originally supposed to be of marine origin, and this view is still upheld by some geologists. It is the same question in which so many of the so-called "middle glacial sands and gravels of the British Islands are involved, and upon which there has been so much discussion. If it be permissible for me to reiterate the well-known argument by which the presence of marine shells in gravels of glacial origin is explained it may be outlined as follows.

Since the basins around our islands are known to have been occupied by the sea at the beginning of the Glacial Period, and since these basins were afterwards filled by ice-lobes, which, as we have seen, moved outward in many places upon the land, dragging with them much of the material of the old sea-floor, it is inevitable that a certain amount of marine detritus will occur in the deposits formed by the ice or derived from its melting. Just as we find shells, and sometimes even transported masses of marine deposits, intact in the Basement Clay, so we find marine relics likewise, though usually more scattered and less perfect, in the gravels derived from the same icesheet. This deduction is consistent with our knowledge of existing glaciers and ice-sheets; thus, Sir Archibald Geikie has recorded the presence of sea-shells in the moraine of a Norwegian glacier 2; Profs. E. J. Garwood and J. W. Gregory have found an excellent illustration of the same phenomenon in one of the Spitzbergen glaciers; and Prof. R. D. Salisbury, in describing the characteristic upturning of the layers, of ice at the end of one of the glacial lobes which descends into a shallow bay in North Greenland, gives the following instructive note on the conditions which he observed: "Here the upturning of the layers brought up shells from the bottom of the bay, and left them in marginal belts where the upturned layers outcropped. These shells were mingled with other sorts of débris. In one case their quantity could have been measured by some such unit as the wagon-load."

In our islands, as Prof. P. F. Kendall has clearly shown in discussing the drifts of Western England, it is only where the ice-lobes have passed over portions of the preexisting sea-floors that we find marine remains in the drift deposits; while in other places, at the same or lower elevations, where there is proof that the ice-flow was from the land, such remains are invariably absent.

The occurrence of these shells in a few places at high elevations, all explicable by consideration of the geographical circumstances, gave rise to the idea of a great mid-glacial submergence, and upon this idea the hypothesis of a mild interglacial epoch has mainly hinged. In Prof. Geikie's latest scheme this supposed submergence is, indeed, reduced to moderate limits, but it is still the essential factor in the argument.

The same idea of a moderate degree of submergence, accompanied by temperate conditions of climate, has been

applied by Mr. Clement Reid' to the shelly gravels of Holderness. Mr. Reid has also proposed to include the buried cliff-beds of Sewerby in the same interglacial stage; but as the gravels rise to nearly 100 feet above the level of the old beach in northern Holderness, and are separated from it by the Basement boulder-clay, I am sure that this

correlation cannot be sustained.

These Holderness gravels are supposed to be absent from the coast sections, and it is suggested that they may lie below sea-level in this quarter; but this is not very probable, as they are found at an elevation of 50 feet within a few miles of the coast in southern Holderness, and the Basement boulder-clay rises well above sea-level in the cliffs at Dimlington. It is true that the gravels of the coast sections afford no support to the idea of a mild interglacial submergence, and are evidently of similar origin with the rest of the glacial deposits, but I can see no other reason against their correlation with the gravels of the neighbouring interior. Except in two or three limited tracts, the shells in the Holderness gravels are as fragmentary, and nearly as scanty, as in the moundy gravels of Flamborough Head, which from their character and position cannot be of marine origin. Even at the exceptional places referred to, where the fossils are more plentiful, there is a mixture of forms, including an abundance of the freshwater shell Corbicula fluminalis, which seems to denote their derivation from preexisting local deposits; and in the new section at Burstwick, described by Mr. T. Sheppard, these shelly gravels revealed the same close association with the boulder-clay that is so frequently displayed in the glacial gravels of the coast sections.

The Kirmington Section.-There is, however, one case known to me in the east of England, and only one, in which an undoubtedly contemporaneous fauna occurs in beds intercalated with the boulder-clay series.3 At Kirmington, in North Lincolnshire, a brickyard is worked in a deposit of estuarine clay lying in the middle of a broad shallow valley which cuts across the Chalk Wolds about eight miles south of the Humber. Recent investigation by a Research Committee of the Association, in which I took an active share, has shown, somewhat undescends to present sea-level, and that the estuarine warp expectedly, that the surface of the chalk at this place is underlain by more than 60 feet of drift, consisting of sand and chalky gravel, with two thick bands of tough these beds were proved was insufficient to show precisely clay containing far-travelled stones. The boring in which whether the stony clays possessed the distinguishing features of true till, but there can be no doubt as to their glacial character, since we know of no deposits of this kind in the east of England except those of glacial age.. At the base of the estuarine warp, at 65 feet above Ordnance datum, we found a thin seam of silt and peat containing a few freshwater shells and plant remains, which, like the very scanty fauna of the overlying warp, give no precise indication of climatal conditions, though suggesting that the climate was cooler than at present. The estuarine bed is overlain by a coarse gravel of rolled flints, and in one part of the section this gravel is covered by 3 or 4 feet of red clay with far-travelled stones, resembling the Upper boulder-clay or Hessle Clay of Holderness. The character and fauna of the warp show that it must have been laid down between tide-marks, and we therefore gain an exact measure of the sea-level at the time of its accumulation, and also, I think, of the highest limit of marine submergence in this part of England during any stage of the Glacial Period.

The position of the deposit, at the fringe of the great sheet of drift, which covers the lowland east of the Wolds

1 "The Geology of Holderness." Mem. Geol. Survey (1885). "On another Section in the so-called Interglacial Gravels of Holderness." Proc. Yorks. Geol. and Polytech. Soc., vol. xiii. (1895), PP. 1-14. 3 The freshwater deposit which I found some years ago at Bridlington, and at first thought to be probably intercalated with the boulder-clay, proved on fuller exposure to lie above the boulder-clay, with which it had become entangled by later disturbance. See Geol. Mag., dec. ii., vol. vi. (1879), p. 393; and Proc. Yorks. Geol. and Polytech. Soc., vol. vii. (1981). P. 389. 4 Rep. British Assoc. for 1904, PP. 272-4.