On the European continent, marine Pliocene beds are found in Belgium and Italy. The former deposits greatly resemble our Crags, whilst the latter are of interest on account of the mixture of volcanic beds with marine sediments in Sicily, showing that the formation of Etna commenced in Pliocene times. Various deposits formed in inland basins are found in France and Germany, but the most remarkable occur in the Vienna basin, where Caspian conditions prevailed over large areas, and the ordinary strata alternate with chemical deposits of which the best-known are the celebrated rock salt masses of Wieliczka, near Cracow. At the same time volcanic activity was rife to the south of the Carpathian mountains. Other deposits, which are partly referable to the Pliocene period, occur in Greece at Pikermi, and in India in the Siwalik hills; these are celebrated for their remarkable mammals, as are the Pliocene strata of the Western territories of North America. The occurrence of marked earth-movements since Pliocene times is indicated by the nature of the deposits of Barbadoes, where radiolarian cherts have furnished two echinids which are described by Dr Gregory as deep-sea forms. These beds were once referred to the Miocene period, but there is good reason for assigning them to a later date, and correlating them with the Pliocene beds of other areas, in which case there must have been a considerable uplift in this region since Pliocene times, a fact of great theoretical importance.

The climatic conditions of Pliocene times show steady fall of temperature. The early Pliocene beds of Britain were deposited during the prevalence of warmer temperatures than those which now exist in the same area, but during later Pliocene times, the temperature was at first similar to that now prevailing, and afterwards distinctly colder, and we find in the upper Pliocene beds the remains of organisms of a northern type. In the uppermost deposit of the Cromer 'Forest' Series, the arctic birch and arctic willow indicate the commencement of the cold which culminated in the succeeding 'Great Ice Age.'

The flora and fauna. Little need be said of the Pliocene fossils: the flora approaches that of present times, and the invertebrates are in most cases specifically identical with those now living. The vertebrates alone differ markedly from living forms, being chiefly of extinct species, and in many cases belonging to extinct genera. It is interesting to find that the mammalian fauna of Pliocene times resembles the existing fauna of the area in which the beds are found, a fact long ago observed by Darwin. Thus the European Pliocene mammals are like existing European forms, whilst in Australia the mammalian terrestrial fauna consists of Marsupials, and in South America there are Edentata of Pliocene age[108].

CHAPTER XXVII.

THE PLEISTOCENE ACCUMULATIONS.

Classification. The term Pleistocene, as used here, is approximately equivalent to the expressions 'Glacial Period' and 'Great Ice Age' of some writers; but I have adopted it in preference to these expressions, because it may eventually be possible to define the Pleistocene period in such a manner as to give the term a strictly chronological meaning, whereas the other terms indicate the existence of climatic conditions which must have ceased in some areas sooner than in others. At present, climatic change gives us the best means for separating the accumulations formed subsequently to the Pliocene period over large parts of the Eurasian land-tract, and the most convenient division of these continental accumulations is to refer them to three periods, viz.:—

The Forest Period (in which we are now living).
The Steppe Period.
The Glacial Period.

Some of the accumulations which were formed during the Steppe period are included in the Pleistocene period by many writers, but I prefer to treat of them as post-Pleistocene.

In the present state of our knowledge of the glacial deposits any attempt to make a classification applicable over very wide areas is doomed to failure, and the very principles upon which the classification should be based are a subject of disagreement. The most promising basis for classification is founded on alternate recession and advance of land-ice, though the proofs that advance takes place simultaneously over very wide areas are not yet forthcoming. Dr J. Geikie in the last edition of his work The Great Ice Age adopts four periods of glaciation, with intervening periods of recession, and this division accords with the observations of many foreign geologists. In order to understand the method of classification upon this basis, a few words concerning glacial deposits in general will not be out of place. Glacial accumulations may be divided into three classes:—(i) true glacial accumulations, formed on, in, and under the ice, and left behind upon its recession, (ii) marine glacial deposits, laid down in the sea, when floating ice is extensively found on its surface, and (iii) fluvio-glacial deposits, laid down by streams which come from the ice. The two former indicate glacial conditions, while the occurrence of fluvio-glacial deposits overlain by true glacial deposits indicates an advance of land-ice, for the fluvio-glacial deposits are accumulated in front of those which are truly glacial. Accordingly if we find alternations of glacial and fluvio-glacial deposits on a large scale, we may fairly infer the alternation of periods of great glaciation with others when the ice diminished, or in other words of glacial and interglacial periods. There is, however, in many cases great difficulty in distinguishing glacial deposits from marine glacial ones, while some of the true glacial deposits formed in the ice (englacial deposits) cannot readily be distinguished from those of fluvio-glacial origin. Furthermore, as the terminal moraines of land-ice often rest upon other true glacial deposits, it is often difficult to know whether we are dealing with the products of one or two glaciations over limited areas. The test of superposition is often applicable, and one is enabled to obtain some clue as to the relative order of events. In England at least three periods of glaciation seem to be indicated by the glacial deposits. On the east coast the Cromer Forest Series is succeeded by the Cromer Till, and in Yorkshire the Basement Clay occupies a similar position with regard to the overlying glacial accumulations to that of the Cromer Till. Whether these deposits be marine or terrestrial, and the evidence is not yet sufficient to settle this question to the satisfaction of all geologists, there is no doubt that they are glacial. Above them, in East Anglia, lies the Contorted Drift, the origin of which is still a moot point, and it is overlain by the great Chalky Boulder Clay, which extends far and wide over East Anglia, the Midland Counties and into Yorkshire. Evidence has been adduced to connect this with the till or boulder clay which spreads over the upland districts of the north of England at the foot of the main hill-systems. This set of deposits indicates a second glaciation. As the upland till is often ploughed out by glaciers which have left their traces in the form of moraines in our upland regions, we seem here to have evidence of a third glaciation, which naturally leaves no traces in the southern districts, and the exact age of this cannot be ascertained in the absence of fossil evidence, though we may provisionally refer it to the Pleistocene period.

Another attempt has been made to classify the glacial deposits, on the supposition that there have been periods of elevation and depression of the land during Pleistocene times. Some writers advocate one interglacial period when the land was depressed to an extent of 1400 and perhaps 2000 feet, while others have advocated the occurrence of a number of such interglacial marine periods. The evidence for the supposed oscillations is furnished by the existence of shell-bearing sands associated with boulder clays at high levels, the best known being on Moel Tryfan in Caernarvonshire, near Macclesfield in Cheshire, and near Oswestry in Shropshire. As many geologists believe that these shells have been carried to their present position by ice in a way which it is not our province to discuss here, we may dismiss this method of classification as based upon events which cannot be proved to have occurred. In the present state of our knowledge, it is indeed best to avoid, as far as possible, classifications which are intended to be applicable over wide regions, and to devote our attention to local details, gradually piecing together the evidence which is obtained as the result of exhaustive examination of each separate area[109].

The foregoing remarks will convince the student that any attempt to show the distribution of land and sea during any part of the glacial period is not likely to meet with general acceptance, as so much depends upon the terrestrial or marine origin of the deposits of the lowlands, and the mode of formation of the shell-bearing drifts of high levels. The occurrence of elevation to a greater height than that which our country at present possesses during portions at any rate of the glacial period has been inferred on general grounds, but direct evidence in favour of it is furnished by the existence of a number of ancient valleys on the land around our coasts, whose floors are often considerably below sea-level, while the valleys are now completely filled up with glacial accumulations, except where they have been partially re-excavated by streams which for some distance run above the courses of the ancient streams.

The climatic conditions of glacial times can only be briefly touched upon in this place. If the periods of advance can be proved to be contemporaneous over wide areas, this points to alternations of colder and warmer periods, or at any rate of drier and wetter periods, though local advance may be due to a number of causes. It must be borne in mind that with the temperature remaining the same, advance of ice can be brought about by increased precipitation of aqueous vapour in the form of snow.

The question of the cause of the glacial period is one that only indirectly affects the stratigraphical geologist until he has accumulated sufficient evidence to indicate the cause. It must suffice to observe that the extremely plausible hypothesis of Croll (for which the student should consult Dr Croll's Climate and Time) does not explain the apparent gradual lowering of climate throughout Tertiary times till the cold culminated in the Pleistocene period, and the student will do well to remain in suspense concerning the cause of the Ice Age until further evidence has been brought to bear upon it.

The glacial flora and fauna. The glacial deposits naturally yield few traces of life, except those which have been derived from other deposits, and we are dependent for our information concerning the fauna and flora of the glacial period upon the remains furnished by the interglacial deposits. Unfortunately it is very hard to ascertain which deposits are interglacial, and many which have been claimed as such are either preglacial or postglacial. The meagre evidence which we possess points to the existence of an arctic fauna or flora in Britain during the prevalence of this glacial period. A question which has received much attention of recent years is that of the existence of preglacial or interglacial man, on which much has been written. The existence of man in glacial times is probable, but it is the opinion of many of those who are most competent to form a judgment, that it has not been proved in the only conclusive way, namely, by the discovery of relics of man in deposits which are directly overlain by glacial deposits, or which at any rate are demonstrably older than glacial deposits[110].

CHAPTER XXVIII.

THE STEPPE PERIOD.

The occurrence of a period marked by dry climate over wide areas of the Eurasian continent, and possibly also in North America, is evidenced by the widespread distribution of an accumulation known as loess, concerning the origin of which there has been much difference of opinion, though that it was formed subsequently to the glacial period seems to be generally admitted, inasmuch as it is largely composed of rearranged glacial mud. The formation of the loess as a steppe-deposit was first advocated by Baron von Richthofen, and his views were supported by Nehring after study of the loess-fauna. Richthofen's explanation of the loess as due to the spread of dust by wind in a dry region is becoming widely accepted, and it necessitates the widespread occurrence of steppe conditions, as the loess has a very extensive geographical range, and may be truly regarded as the normal continental deposit of Eurasia during the period immediately succeeding the glacial period. In our own country, as the sea cannot have been far distant during these times the normal loess is not found, but several accumulations occur, which on stratigraphical and palæontological grounds must be regarded as synchronous with the formation of the loess. These are certain rubble-drifts of the southern counties, the older river-gravels of southern England, and some of the older cave deposits of various parts of England. It is doubtful whether any classification into minute subdivisions can be adopted for them, though Prof. Boyd Dawkins has advocated their separation into an older age of River Drift Man, and a newer period of Cave Man, on account of the evidences of a lower state of civilisation afforded by examination of the River Drift implements when compared with those fashioned by Cave Man. Roughly speaking, the Steppe period corresponds with the period during which Palæolithic man existed, at any rate in north-west Europe, and we may speak of the Steppe period as the Palæolithic period, without asserting that Palæolithic man necessarily disappeared at the time when the climate changed and caused the replacement of Steppe conditions by others favourable to forest-growth.

Description of the accumulations. The loess consists of unstratified calcareous mud or dust, with a peculiar vertical fracture, and is interesting rather on account of the nature of its fossils and of its distribution than for its lithological characters. As it is not found in Britain it is not necessary to say much about it, but merely to refer to the published descriptions[111].

The British deposits require some notice, as their characters and mode of occurrence are of some significance. Along the south coast are deposits of coarse rubble which have yielded some organic remains, which have been described by Mr Clement Reid[112], who also discusses their origin. The rock, also known as the Elephant Bed, consists of angular fragments of flint and chalk, and seems to have been produced by streams which were able to flow over the surface of the chalk when it was frozen. Many other similar deposits in the south of England, which are found on the open surface, may have had a similar origin.

The Palæolithic river-gravels are found at various distances above present river-levels, and are the surviving relics of alluvial deposits which were laid down when the rivers ran at a higher level than they now do. That they are newer than the main glacial drifts of the region in which they occur is indicated by the frequent presence in them of boulders derived from the drift. Their antiquity is shown by the physical changes which have occurred since their deposition (there having been sufficient time since then to allow of the excavation of some river-valleys to a depth of over one hundred feet beneath their former level), and also by the character of the included mammals which will presently be referred to. The deposits vary in coarseness, like those of modern alluvial flats, from the coarse gravels of the river-beds to the fine loams and marls of the flood-plains. They are found, in Britain, with their typical mammalian remains, south-east of a line drawn from the mouth of the Tees to the Bristol Channel.

The cave-deposits have a wider distribution than those which have just been noticed, being also found to the north-west of the above-mentioned line in Yorkshire, and in North and South Wales. In the south of England they are found as far east as Ightham in Kent, and in a westerly direction to Torquay and Tenby. The Ightham deposits occur in fissures and consist of materials which were apparently introduced from above by river action[113]. The cave-deposits of limestone areas are sometimes found in fissures, but at other times in caverns with a fairly horizontal floor, on which the various accumulations lie in order of formation. The deposits vary in character and may be divided into three groups, though accumulations of intermediate character are found; the first group consists of cave-earths and cave-breccias—formed by weathering of the limestone, and the retention of the insoluble residue, as a more or less ferruginous mud, mixed with angular fragments of limestone, and with the remains of creatures which inhabited the caves; the second group consists of true deposits laid down under water, as gravels, sands, and laminated clays; while the third is composed of limestone deposited from solution in water, in the form of stalagmite[114].

The organic contents of the Palæolithic period are of much interest, and it is desirable to discuss their character before making further observations upon the physical conditions of the period.

The Palæolithic flora and fauna. The plants of some of the earlier deposits of the age we are considering show the prevalence of cold conditions during their accumulation, for instance the Arctic birch and Arctic willow are found in the accumulations beneath the implement-bearing Palæolithic deposits of Hoxne in Suffolk[115]. The invertebrate fauna consists essentially of the remains of molluscs. The loess molluscs are chiefly pulmoniferous gastropods which lived upon the land, though swamp forms are occasionally associated with them. The palæolithic river-gravels have yielded numerous land- and freshwater-molluscs of living species, though some which are abundant in the British gravels are now extinct in Britain, e.g. Cyrena (Cobicula) fluminalis and Unio littoralis. Marine deposits of this age are occasionally found, as at March, in Cambridgeshire, where the fauna closely resembles that of our present sea-shores.

The vertebrate remains are much more remarkable, and it is not quite clear that the association of forms whose living allies now live under widely different conditions has been satisfactorily explained. The river-gravels and cave-deposits contain remains of temperate forms, as the bison, and brown bear, associated with those of northern forms, as the mammoth, woolly rhinoceros, glutton, reindeer, and musk ox, and also with those whose living allies are inhabitants of warmer regions, like the lion, hyæna, and hippopotamus. One of the most remarkable creatures is the sabre-toothed lion or Machairodus, remains of which have been discovered in Kent's Cavern, Torquay, and in the caves of Cresswell Crags, Derbyshire.

The loess fauna consists of characteristic steppe animals, such as the jerboa, Saiga antelope and steppe-porcupine, and it is interesting to find an indication of this fauna in the Ightham fissures.

The first undoubted relics of mankind are found in the Palæolithic deposits, which are very widely spread over the Eurasian continent. They consist mainly of implements of bone and stone, the latter being chipped, but never ground or polished, though both bone and stone implements are frequently ornamented with engraved figures. The cave-deposits have furnished implements of a higher type than those usually found in the river-drifts, but the latter are also found in caverns in deposits beneath those containing the higher type, hence the division of the period into two minor periods, that of river-drift man, and that of cave-man[116].

There are several questions of interest connected with the Palæolithic fauna, three of which deserve some notice here. The absence of the relics of the Palæolithic mammalia and of the human implements in the river-gravels north-west of the line drawn between the Tees and Bristol Channel, and the presence of the mammalian remains in the caverns of that area requires some explanation. One such explanation assumes that the relics were destroyed in the open country to the north-west of that line, owing to glaciation, but it is not by any means universally accepted.

Another difficulty which in the opinion of some writers has not been fully cleared up is the mixture of apparently southern forms like the Hippopotamus, with others of northern character like the Musk ox, under such conditions as to show that the creatures lived in the British area contemporaneously. Seasonal migration might account for it, but the wide belt of overlap of apparent northern and southern forms requires something more, though secular changes of climate might shift the belt of seasonal overlap from one place to another, causing the entire belt of overlap to extend over a considerable distance.

The third, and perhaps most important difficulty is the abrupt change from the Palæolithic type of implement to the Neolithic type, characteristic of the next period. Some implements, as those of the kitchen-middens of Denmark, and those found at Brandon and Cissbury in this country, have been appealed to as intermediate in character, but evidence has been brought forward to show that each set is truly Neolithic, the one being the implements of the lowly fisher-folk who lived contemporaneously with the makers of the highly finished polished implements of Denmark, while the others are unfinished implements thrown away during the manufacture on account of flaws or accidental fractures. The difficulty is increased when we take into account the great physical and faunistic changes which occurred between Palæolithic and Neolithic times.

The country was undoubtedly more elevated than it is at present during portions if not during the whole of Palæolithic times, as shown by the appearance of the great mammals in Britain, the discovery of their remains beneath sea-level, and especially the occurrence of remains in the caverns of rocky islands such as those of the Bristol Channel, where they could not possibly have existed unless the present islands were connected with the mainland.

The fossils of the times between the Glacial period and the Neolithic period indicate variations of climatic conditions. Upon this point I cannot do better than quote the words of Sir John Evans in his Presidential Address to the British Association at Toronto[117]. "At Hoxne the interval between the deposit of the Boulder clay and of the implement-bearing beds is distinctly proved to have witnessed at least two noteworthy changes in climate. The beds immediately reposing on the clay are characterised by the presence of alder in abundance, of hazel, and yew, as well as by that of numerous flowering plants indicative of a temperate climate very different from that under which the Boulder clay itself was formed. Above these beds characterised by temperate plants, comes a thick and more recent series of strata, in which leaves of the dwarf Arctic willow and birch abound, and which were in all probability deposited under conditions like those of the cold regions of Siberia and North America.

"At a higher level, and of more recent date than these—from which they are entirely distinct—are the beds containing the Palæolithic implements, formed in all probability under conditions not essentially different from those of the present day."

CHAPTER XXIX.

THE FOREST PERIOD.

Subsequently to Palæolithic times, the physical conditions over Eurasia changed greatly, and at the commencement of Neolithic times the conditions were favourable for the growth of forests over wide regions of that continent. At the commencement of the Forest period the physical conditions were very much the same as they are at present, though minor changes have of course taken place since then, including probably a submergence of large parts of Britain to a depth of about fifty feet beneath its former level, as indicated by the existence of Neolithic submerged forests round many parts of our coast-lines.

The Forest period may be best subdivided for local purposes by reference to the civilisation of mankind at different times, and in this way we obtain the following divisions:

Historic Iron age.
Prehistoric Iron age.
Bronze age.
Neolithic age.

A classification may also be based upon changes in the flora. In Denmark the peat deposits of this age are divisible into five layers, characterised by different dominant forms of trees. These are as follows in descending order:

In our own country the forest growth has been much interfered with by man, but the lower fenland peat gives a good example of the material formed by forest growth. It is not necessary to touch on the various accumulations which are now being formed in different parts of our island, except to remark that the deposits of the Forest period give indications of earth-movements on a small scale, which is well seen in the fenland, where the forest peat is covered in places by a "buttery clay" with Scrobicularia piperata indicating submergence, and above this is a marsh peat.

The flora and fauna of the Forest period are practically those of the present day, though the larger forms of mammalia have disappeared one by one. The Irish elk and Bos primogenius probably became extinct early in the period, while as far as Britain is concerned the wolf, bear, and beaver have disappeared within historic times.

The relics of man deserve passing notice. The Neolithic period is characterised by the absence of metal instruments, though those made of stone were much more highly finished than those of Palæolithic times, and were often ground and polished. The first metal which was largely worked was bronze, which gradually replaced stone, though stone was extensively used in the Bronze age, as indicated by the imitation of bronze implements in stone. The Bronze age in turn was replaced by the Prehistoric iron age; at first, when iron was scarce, bronze implements were merely tipped with iron, but ultimately the one metal was practically replaced by the other.

The date of the Palæolithic period is unknown; no approximate date can be satisfactorily assigned to it, but various calculations, founded on different data, have been made as to the age of the Neolithic period, and several of them agree in placing it at about 7000 years from the present time.

It will be seen that no sudden and violent change marks the incoming of the human race, which to the geologist is but one of a large number of events which have followed each other in unbroken sequence, and accordingly the thread of the story where abandoned by the geologist is taken up by the antiquary, and passed on by him to the historian[118].

CHAPTER XXX.

REMARKS ON VARIOUS QUESTIONS.

There are many problems connected with geology which can only be solved by detailed study of the stratified rocks, and when solved the principles of the science will be more fully elucidated. In the present state of our knowledge some of these problems are ripe for discussion, others can merely be indicated, while others again have probably remained hidden, though it will be the task of the geologist of the future to clear them up. Among the many questions which demand knowledge of stratigraphical geology for their right understanding are the following, which will be briefly considered in this chapter:—the changes in the position of land and sea in past times, and the growth of continents; the replacement of a school of uniformitarianism by one of evolutionism; and the duration of geological time.

Changes in the position of land and sea. Certain physicists have arrived at the conclusion that the general position of our oceans and continents was determined at a very early period in the earth's history, and that the changes which have occurred in their position since then have been comparatively insignificant. The wide extent of land over which stratified rocks are distributed at once indicates that from the point of view of the geologist the changes have been very important, and it is worth inquiring whether they are not sufficiently important to prove that the primitive oceans and continents have undergone so much alteration as to be unrecognisable. Some authorities, while recognising the great changes which have occurred in the relative position of land and sea during those periods of which geologists have direct information, suppose that the changes took place to a large degree in certain 'critical areas' bordering the more stable areas of permanent ocean on the one side and permanent land on the other.

In discussing the question of general permanence of land and ocean regions it will be convenient to commence with a study of the present land areas, and at the outset we may take into consideration the present distribution of marine sediment over different parts of the land, using the last edition of M. Jules Marcou's geological map of the world for the purpose[119]. A glimpse at this map indicates that more than half of the land areas are occupied by rocks which are as yet unknown (many of which may be marine sediments), or by crystalline schists of which the mode of origin has not yet been fully explained, so that a large part of Central Asia, the interior of Africa, and of South America may have existed as land from very early times, and the same may be said of smaller portions of Europe and North America. Actual observation of a geological map therefore indicates the possibility that about half of the land surfaces may have existed as such through very long periods, but though there is a possibility of this, the probability is not very great. The unknown regions, as remarked above, may consist to a considerable extent of marine sediments, and the existence of isolated patches of late Palæozoic and of Mesozoic strata in the heart of Central Asia, points to the submergence of much wider regions than those in which these isolated patches have been found. Again, the character of the sediments when they abut against the crystalline schists frequently proves that these sediments once extended further over the crystalline schists, and have since been removed by denudation, so that even if we assume that the crystalline schists are all of very early date, and not necessarily formed in any case from marine sediments, we cannot suppose that all the area occupied by them has existed as land for long periods of time. On the other hand, the major part of Europe and North Africa, extensive tracts in Asia, the greater part of Australia, a very large part of North America and considerable tracts of South America give proofs of having been occupied by the oceans in Palæozoic and later times.

It may be answered that most of these regions containing marine sediments occur in critical areas, which have undergone a certain amount of oscillation owing to earth-movements, and that the interior parts of the great continental masses have been practically stationary. But if these lands had been land-areas through geological ages they must have been acted upon by the agents of subaerial denudation, throughout these ages, and long ago reduced to peneplains[120] unless the action of these subaerial agents was counteracted by that of elevating forces, but if these forces were sufficient to counteract the action of subaerial denudation through countless ages, they were also sufficient to raise extensive tracts of land above sea-level, and materially to alter the distribution of land and sea, and if elevation could go on to this extent, why not also depression?

Proceeding a step further, and examining the character of the sediments as well as their geographical distribution, we find further evidence of great crust-movements. It has been urged that deep-water sediments do not occur amongst the strata found on the continents,—that there are no representatives of the abysmal deposits of recent ocean floors amongst the strata of the geological column[121], but the researches of the last two decades have brought to light foraminiferal and radiolarian deposits, pteropodal deposits, and possibly deep-sea clays, which are comparable with those in process of formation at great depths in existing oceans, and though the proofs of their deep-sea origin are not always as full as might be desired in the case of the older rocks[122], we can speak with greater certainty when we examine those of Tertiary age, and if the deep-sea accumulations of this late date can be uplifted above sea-level, this is much more likely to have occurred with those of past times. When a deposit like the radiolarian rock of Barbadoes, the deep-water character of which has been conclusively proved, can be elevated into land since Miocene or possibly Pliocene times, it is evident that the crust-movements have been sufficient to produce the most profound changes in the distribution of land and sea during the long ages which are known to us. Another argument against the occurrence of extensive changes has been derived from an examination of those islands which are spoken of as oceanic islands. Strictly speaking an oceanic island is one in which the present fauna and flora give indications of their introduction by transport across intervening sea, and no indications of the existence of forms of life which inhabited it when it was once united to a continent; it may be inferred with a considerable degree of certainty that these islands have been isolated for long periods of time. It has been stated that these oceanic islands never contain marine sediments of any considerable degree of antiquity, and that there are therefore no traces of former continents over those wide tracts of ocean which are occupied by oceanic islands. The evidence is of a negative character. The islands would be less likely to exhibit ancient sediments than continents, for being near the ocean, they would be readily submerged, and the older deposits masked by newer ones, though this need not necessarily account for the entire absence of ancient rocks amongst them. The danger of the argument lies in the fact that we do not yet know how far these old rocks really are absent, as the geology of the oceanic isles has not been fully explored from this point of view, and already several cases of the asserted presence of ancient rocks on these islands have been recorded.