Cambridge Natural Science Manuals.
Geological Series.

THE PRINCIPLES
OF
STRATIGRAPHICAL GEOLOGY

London: C. J. CLAY AND SONS,
CAMBRIDGE UNIVERSITY PRESS WAREHOUSE,
AVE MARIA LANE.
AND
H. K. LEWIS,
136, GOWER STREET, W.C.

Leipzig: F. A. BROCKHAUS.
New York: THE MACMILLAN COMPANY.
Bombay: E. SEYMOUR HALE.

THE PRINCIPLES
OF
STRATIGRAPHICAL GEOLOGY

BY

J. E. MARR, M.A., F.R.S.

FELLOW AND LECTURER OF S. JOHN'S COLLEGE, CAMBRIDGE,
AND UNIVERSITY LECTURER IN GEOLOGY.

CAMBRIDGE:
AT THE UNIVERSITY PRESS.
1898
[All Rights reserved.]

Cambridge:
PRINTED BY J. & C. F. CLAY,
AT THE UNIVERSITY PRESS.

[PREFACE.]

The present work has been written in order that students may gain by its perusal some idea of the methods and scope of Stratigraphical Geology. I believe that this idea can be obtained most satisfactorily, if a large number of the details connected with the study of the stratified rocks are omitted, and I have accordingly given very brief accounts of the strata of the different Systems.

The work is intended for use in conjunction with any book which treats of the strata of the Geological Column at considerable length; some of these books are mentioned on pages [124], [125].

J. E. M.

Cambridge,
November, 1898.

[CONTENTS.]

ADDENDA ET CORRIGENDA.

[p. 38], line 15 from bottom: for 'joining' read 'jointing'

[p. 208], line 6 from bottom: for 'Dr' read 'Messrs Medlicott and'

[p. 214], line 15 from bottom: after 'Permo-Carboniferous Strata' insert 'through the Permian'

[p. 217], last line of footnote: for 'Dr' read 'Messrs Medlicott and'

[p. 217], insert a second footnote: 'For information concerning the Permian volcanic rocks see Sir A. Geikie's Ancient Volcanoes of Great Britain.'

[p. 235], insert a footnote: 'A good account of the British Jurassic rocks will be found in Mr H. B. Woodward's Memoir on "The Jurassic Rocks of Britain." Mem. Geol. Survey, 1893—.'

[p. 250], top line: for 'Gardiner' read 'Gardner'

[Trancriber's Note: Above corrections were made to the text.]

[CHAPTER I.]

INTRODUCTION.

It is the aim of the Stratigraphical Geologist to record the events which have occurred during the existence of the earth in the order in which they have taken place. He tries to restore the physical geography of each period of the past, and in this way to write a connected history of the earth. His methods are in a general way similar to those of the ethnologist, the archæologist, and the historian, and he is confronted with difficulties resembling those which attend the researches of the students of human history. Foremost amongst these difficulties is that due to the imperfection of the geological record, but similar difficulty is felt by those who pursue the study of other uncertain sciences, and whilst this imperfection is very patent to the geologist, it is perhaps unduly exaggerated by those who have only a general knowledge of the principles and aims of geology.

The history of the earth, like other histories, is a connected one, in which one period is linked on to the next. This was not always supposed to be the case; the catastrophic geologist of bygone times believed that after each great geological period a convulsion of nature left the earth's crust as a tabula rasa on which a new set of records was engraved, having no connexion with those which had been destroyed. Careful study of the records of the rocks has proved that the conclusions of the catastrophists were erroneous, and that the events of one period produce their impression upon the history of the next. Every event which occurs, however insignificant, introduces a new complication into the conditions of the earth, and accordingly those conditions are never quite the same. Although the changes were no doubt very slow, so that the same general conditions may be traced as existent during two successive periods, minor complications occurred in the inorganic and organic worlds, and we never get an exact recurrence of events. Vegetable deposits may now be in process of accumulation which in ages to come may be converted into coal, but the general conditions which were prevalent during that Carboniferous period when most of our workable coal was deposited do not now exist, and will never exist again. The changes which have taken place and which are taking place show an advance from the simple to the more complex, and the stratigraphical geologist is confronted with a problem to which the key is development, and it is his task to trace the development of the earth from the primitive state to the complex condition in which we find it at the present day.

Our general ignorance of the events of the earliest periods of the history of the earth will be emphasised in the sequel, and it will be found that the complexity which marks the inorganic and organic conditions which existed during the deposition of the earliest rocks of which we have detailed knowledge points to the lapse of enormous periods of time subsequent to the formation of the earth, and previous to the deposition of those rocks. The imperfection of the record is most pronounced for that long period of time, but in this respect the geologist is in the same condition as the student of human history, for the relics of the early stone age prove that man in that age had attained a fairly high state of civilisation, and the gap which separates palæolithic man from the first of our race is relatively speaking as great as that which divides the Cambrian period from the commencement of earth-history. Nevertheless, human history is a science which has made gigantic strides towards the solution of many problems connected with the development of man and civilisation, and similarly geology has advanced some way in its task of elucidating the history of our globe.

The task of the stratigraphical geologist is two-fold. In the first place, he must establish the order of succession of the strata, for a correct chronology is of paramount importance to the student of earth-lore. The precautions which must be taken in making out the order of deposition of the rocks of any area, and correlating those of one area with those of another will be considered in the body of the work. When this task is completed, there yet remains the careful examination of all the information supplied by a study of the rocks of the crust, in order to ascertain the actual conditions which existed during the deposition of any stratum or group of strata. In practice, it is generally very difficult to separate these two departments of the labour of the stratigraphical geologist, and the two kinds of work are often done to a large extent simultaneously, or sometimes alternately. Frequently the general succession of the deposits comprising an important group is ascertained, and at the same time observations made concerning the physical characters of the deposits and the nature of their included organisms, which are sufficient to afford some insight into the general history of the period when these deposits were laid down; a more detailed classification of the same set of deposits may be subsequently made, and as the result of this, more minute observations as to the variations in the physical and biological conditions of the period are possible, which permit us to write a much more concise history of the period. So great has been the tendency to carry on work in a more and more detailed manner, that it is very difficult if not impossible to tell when any approach to finality is reached in the study of a group of strata in any area. Roughly speaking, we may state that our knowledge of a group of strata is obtained by three processes, or rather modifications of one process. The general order of succession is established by the pioneer, frequently as the result of work carried on through one or two seasons. Subsequently to this, a more minute subdivision of the rocks is possible as the result of labours conducted by one or more workers who are enabled to avail themselves of the work of the pioneer, and our knowledge of the rocks is largely increased thereby. But the minutiæ, often of prime importance, are supplied by workers who must spend a large portion of their time in the area where the work lies, and it is only in districts where work of this character has been performed, that our knowledge of the strata approaches completion. The strata of the Arctic regions, for example, have in many places been examined by pioneers, but a great deal remains to be done in those regions; the main subdivisions only have been defined in many cases, and our information concerning the physical history of Arctic regions in past times is comparatively meagre. To come nearer home—a few miles north of Cambridge lies the little patch of Corallian rock at Upware; it has been frequently visited, and a large suite of organic remains extracted from it, but no one has devoted the time to the collection of remains from this deposit which has been devoted to that of some other formations presently to be mentioned, and accordingly our knowledge of the fauna of that deposit is far from complete. Contrast with this the information we possess of the little seam known as the Cambridge Greensand, from which organic remains have been sedulously collected during the extensive operations which have been carried on for the extraction of the phosphatic nodules which occur in the seam. The suite of relics of the organisms of that period is accordingly far more perfect than in the case of many other beds, and indeed the large and varied collection of relics of the vertebrata of the period which furnish much information of value to the palæontologist would not have been gathered together, had not this seam been so carefully worked, and an important paragraph in the chapter bearing on the history of this period would have remained unknown to us. Again, two little patches of limestone of the same age, one in central England and the other in the island of Gothland, have been the objects of sedulous inquiry by local observers, and we find again that our knowledge of the physical history of the period, as regards these two regions, is exceptionally perfect. Special stress is laid upon this point, for in these days, when every county possesses its learned societies whose members are desirous of advancing in every possible way the progress of science, it is well to insist upon the importance of this detailed work which can only be done by those who have a large amount of time to devote to the rigorous examination of the rocks of a limited area.

[CHAPTER II.]

ACCOUNT OF THE GROWTH AND PROGRESS OF STRATIGRAPHICAL GEOLOGY.

The history of the growth of a science is not always treated as an essential part of our knowledge of that science, and many text-books barely allude to the past progress of the science with which they deal. The importance of a review of past progress has, however, attracted the attention of many geologists, and Sir Charles Lyell, in his Principles of Geology, gave prominence to an historical sketch of the rise and progress of the science. Historical studies of this nature have more than an academic value; the very errors made by men in past times are useful as warnings to prevent those of the present day from going astray; the lines along which a science has progressed in the past may often be used as guides to indicate how work is to be conducted in the future; but perhaps the greatest lesson which is taught by a careful consideration of the rise and progress of a study is one which has a moral value, for he who pays attention to the growth of his science in past times, gains a reverence for the old masters, and at the same time learns that a slavish regard for authority is a dangerous thing. This is a lesson which is of the utmost importance to the student who wishes to advance his science, and will prevent him from paying too little attention to the work of those who have gone before him, whilst it will enable him to perceive that as great men have fallen into error through not having sufficient data at their disposal, he need not be unduly troubled should he find that conclusions which he has lawfully attained after consideration of evidence unknown to his predecessors clash with those which they adopted. Want of this historic knowledge has no doubt caused many workers to waste their time on work which has already been performed, but it has also led others to withhold important conclusions from their fellow-workers because they were supposed to be heterodox. In an uncertain science like geology one of the great difficulties is to keep an even balance between contempt and undue respect for authority, and assuredly a scientific study of the past history of a science will do much to enable a student to attain this end. It will be useful, therefore, at this point to give a brief account of the rise and progress of the study of stratigraphical geology, so far as that can be done without entering into technical details, at the same time recommending the student to survey the progress of this branch of our science for himself, after he has mastered the principles of the subject, and such details as are the property of all who have studied the science from the various text-books written for advanced students.

William Smith, the 'Father of English Geology,' is rightly regarded as the founder of stratigraphical geology on a true scientific basis, but like all great discoverers, his work was foreshadowed by others, though so dimly, that this does not and cannot detract from his fame. It is desirable, however, to begin our historical review at a time somewhat further back than that at which Smith gave to the world his epoch-making map and memoirs.

Before the eighteenth century, stratigraphical geology cannot be said to have existed as a branch of science—the way had not been prepared for it. Data had been accumulated which would have been invaluable if at the disposal of open-minded philosophers, but with few exceptions prejudice prevented the truth from becoming known. There were two great stumbling-blocks to the establishment of a definite system of stratigraphical geology by the writers of the Middle Ages, firstly, the contention that fossils were not the relics of organisms, and, secondly, when it was conceded that they represented portions of organisms which had once existed, the assertion that they had reached their present positions out of reach of the sea during the Noachian Deluge. For full details concerning the mischievous effects of these tenets upon the science the reader is referred to the luminous sketch of the growth of geology in the first four chapters of Sir Charles Lyell's Principles of Geology.

The disposition of rocks in strata, and the occurrence of different fossils in different strata, was known to Woodward when he published his Essay toward a Natural History of the Earth in 1695, and the valuable collections made by Woodward and now deposited in the Woodwardian Museum at Cambridge, show how fully he appreciated the importance of these facts, though he formed very erroneous conclusions from them, owing to the manner in which he drew upon his imagination when facts failed him, maintaining that fossils were deposited in the strata according to their gravity, the heaviest sinking first, and the lightest last, during the time of the universal deluge. The following extracts from Part II. of Woodward's book, show the position in which our knowledge of the strata stood at the end of the seventeenth century: "The Matter, subsiding ..., formed the Strata of Stone, of Marble, of Cole, of Earth, and the rest; of which Strata, lying one upon another, the Terrestrial Globe, or at least as much of it as is ever displayed to view, doth mainly consist.... The Shells of those Cockles, Escalops, Perewinkles, and the rest, which have a greater degree of Gravity, were enclosed and lodged in the Strata of Stone, Marble, and the heavier kinds of Terrestrial Matter: the lighter Shells not sinking down till afterwards, and so falling amongst the lighter Matter, such as Chalk, and the like ... accordingly we now find the lighter kinds of Shells, such as those of the Echini, and the like, very plentifully in Chalk.... Humane Bodies, the Bodies of Quadrupeds, and other Land-Animals, of Birds, of Fishes, both of the Cartilaginous, the Squamose, and Crustaceous kinds; the Bones, Teeth, Horns, and other parts of Beasts, and of Fishes: the Shells of Land-Snails: and the Shells of those River and Sea Shell-Fish that were lighter than Chalk &c. Trees, Shrubs, and all other Vegetables, and the Seeds of them: and that peculiar Terrestrial Matter whereof these consist, and out of which they are all formed, ... were not precipitated till the last, and so lay above all the former, constituting the supreme or outmost Stratum of the Globe.... The said Strata, whether of Stone, of Chalk, of Cole, of Earth, or whatever other Matter they consisted of, lying thus each upon other, were all originally parallel: ... they were plain, eaven, and regular.... After some time the Strata were broken, on all sides of the Globe: ... they were dislocated, and their Situation varied, being elevated in some places, and depressed in others ... the Agent, or force, which effected this Disruption and Dislocation of the Strata, was seated within the Earth."

Woodward's writings no doubt exercised a direct influence on the growth of our subject, but the indirect effects of his munificent bequest to the University of Cambridge and his foundation of the Chair of Geology in that University were even greater, for as will be pointed out in its proper place, two of the occupants of that chair played a considerable part in raising stratigraphical geology to the position which it now occupies.

The discoveries which were made after the publication of Woodward's book and before the appearance of the map and writings of William Smith are given in the memoir of the latter author, written by his nephew, who formerly occupied the Chair of Geology at Oxford[1]. It would appear that the fact that "the strata, considered as definitely extended masses, were arranged one upon another in a certain settled order or series" was first published by John Strachey in the Philosophical Transactions for 1719 and 1725. "In a section he represents, in their true order, chalk, oolites, lias, red marls and coal, and the metalliferous rocks" of Somersetshire, but confines his attention to the rocks of a limited district.

[1] Memoirs of William Smith, LL.D. By J. Phillips, F.R.S., F.G.S. 1844.

The Rev. John Michell published in the Philosophical Transactions for 1760 an "Essay on the Cause and Phænomena of Earthquakes," but Prof. Phillips gives proofs that Michell, who in 1762 became Woodwardian Professor, had before 1788 discovered (what he never published) the first approximate succession of the Mesozoic rocks, in the district extending from Yorkshire to the country about Cambridge. Michell's account was discovered written by Smeaton on the back of a letter dated 1788. The following is the succession as quoted in Phillips' memoir (p. 136):

The order of succession of the Cretaceous, Jurassic, Triassic and Permian beds will be readily recognised as indicated in this section, though the discovery of the detailed succession of the Jurassic rocks was reserved for Smith.

In the year 1778, John Whitehurst published An Inquiry into the Original State and Formation of the Earth, containing an Appendix in which the general succession of the strata of Derbyshire is noted. The main points of interest are that the author clearly recognised the 'toad-stones' of Derbyshire as igneous rocks, "as much a lava as that which flows from Hecla, Vesuvius, or Ætna," though he believed that they were intrusive and not contemporaneous, and he also foreshadows the distinction between the solid strata and the superficial deposits,—"we may conclude," he says, "that all beds of sand and gravel are assemblages of adventitious bodies and not original strata: therefore wherever sand or gravel form the surface of the earth, they conceal the original strata from our observation, and deprive us of the advantages of judging, whether coal or limestone are contained in the lower regions of the earth, and more especially in flat countries where the strata do not basset."

Werner, who was born in 1750, exercised more influence by his teaching than by his writings. His ideas of stratigraphical geology were somewhat vitiated by his theoretical views concerning the deposition of sediment from a universal ocean, in a definite order, beginning with granite, followed by gneiss, schists, serpentines, porphyries and traps, and lastly ordinary sediments. He recognised and taught that these rocks had a definite order "in which the remains of living bodies are successively accumulated, in an order not less determinate than that of the rocks which contain them[2]." The limited value of Werner's stratigraphical teaching is accounted for by Lyell, who remarks that "Werner had not travelled to distant countries; he had merely explored a small portion of Germany, and conceived and persuaded others to believe that the whole surface of our planet, and all the mountain-chains in the world, were made after the model of his own province," and the author of the Principles justly calls attention to the great importance of travel to the geologist. Those who cannot travel extensively should at any rate pay special attention to the works published upon districts other than their own, and even at the present time, the writings of some British workers are apt to be marked by some of that 'insularity' which our neighbours regard as a national characteristic.

[2] Cuvier's Eloge.

It is now time to turn directly to the work of William Smith, who, of all men, exercised the most profound influence upon the study of stratigraphical geology and may indeed be regarded as the true founder of that branch of the science. The memoir of his life which was before mentioned is all too short to illustrate the methods of work which he followed, but in it we can trace his success to three things:—firstly, his 'eye for a country,' to use a phrase which is thoroughly understood by practical geologists, though it is hard to explain to others, inasmuch as it epitomises a number of qualifications of which the most important are, a ready recognition of the main geological features from some coign of vantage, an intuitive perception of what to note and what to neglect, and the power of storing up acquired information in the mind rather than the note-book, so that one may use it almost unconsciously for future work; secondly, ability to draw conclusions from his observations, and thirdly, and perhaps most important of all in its ultimate results, a facility for checking these conclusions by means of further observations, and dropping those which were clearly erroneous, whilst extracting the truth from those which contained a germ of truth mixed with error.

Besides writers referred to above "some foreign writers, in particular Scilla and Rouelle, appear to have made very just comparisons of the natural associations of fossil shells, corals, &c. in the earth, with the groups of similar objects as they are found in the sea, and thus to have produced new proofs of the organic origin of these fossil bodies; but they give no sign of any knowledge of the limitation of particular tribes of organic remains to particular strata, of the successive existence of different groups of organization, on successive beds of the antient sea. Mr Smith's claim to this happy and fertile induction is clear and unquestionable[3]." We get a clue to the manner in which he arrived at his view in the following passage[4]:—"Accustomed to view the surfaces of the several strata which are met with near Bath uncovered in large breadths at once, Mr Smith saw with the distinctness of certainty, that 'each stratum had been in succession the bed of the sea'; finding in several of these strata abundance of the exuviae of marine animals, he concluded that these animals had lived and died during the period of time which elapsed between the formation of the stratum below and the stratum above, at or near the places where now they are imbedded; and observing that in the successively-deposited strata the organic remains were of different forms and structures—Gryphites in the lias, Trigoniæ in the inferior oolite, hooked oysters in the fuller's earth,—and finding these facts repeated in other districts, he inferred that each of the separate periods occupied in the formation of the strata was accompanied by a peculiar series of the forms of organic life, that these forms characterized those periods, and that the different strata could be identified in different localities and otherwise doubtful cases by peculiar imbedded organic remains[5]."

[3] Memoir of William Smith, p. 142.

[4] Ibid. p. 141.

[5] The work of Smith which directly bears upon the establishment of the law of identification of strata by included organisms is published in two treatises, entitled:—

(i) Strata identified by Organized Fossils, 4to. (intended to comprise seven parts, of which four only were published), commenced in 1816.

(ii) A Stratigraphical System of Organized Fossils, compiled from the original Geological Collection deposited in the British Museum. 4to. 1817.

William Smith seems to have recognised intuitively the truth of a law which was but dimly understood before his time,—the law of superposition, which may be thus stated: "of any two strata, the one which was originally the lower, is the older." This may appear self-evident but it was certainly not so. As the result of this recognition he established the second great stratigraphical law, with which his name will ever be linked, that strata are identifiable by their included organisms.

Before Smith's time, geological maps were lithological rather than stratigraphical, they represented the different kinds of rocks seen upon the surface without regard to their age; since Smith revolutionised geology, the maps of a country composed largely of stratified rocks are essentially stratigraphical, but partly no doubt on account of adherence to old custom, partly on economic grounds, the majority of our stratigraphical maps are lithological rather than palæontological, that is the subdivisions of the strata represented upon the map are chosen rather on account of lithological peculiarities than because of the variations in their enclosed organisms. It is hardly likely that Government surveys will be allowed to publish palæontological maps, which will be almost exclusively of theoretical interest, and it remains for zealous private individuals to accomplish the production of such maps. When they are produced, a comparison of stratigraphical maps founded on lithological and palæontological considerations will furnish results of extreme scientific interest.

Turning now from Smith's contributions to the science as a whole, we may now consider what he did for British geology. His geological map was published in 1815 and was described as follows:—"A Geological Map of England and Wales, with part of Scotland; exhibiting the Collieries, Mines, and Canals, the Marshes and Fen Lands originally overflowed by the Sea, and the varieties of Soil, according to the variations of the Substrata; illustrated by the most descriptive Names of Places and of Local Districts; showing also the Rivers, Sites of Parks, and principal Seats of the Nobility and Gentry, and the opposite Coast of France. By William Smith, Mineral Surveyor." The map was originally on the scale of five miles to an inch. In 1819 a reduced map was published, and in later years a series of county maps. He also published several geological sections, including one (in 1819) showing the strata from London to Snowdon.

The student should compare Smith's map of the strata with one published in modern times in order to see how accurate was Smith's delineation of the outcrop of the later deposits of our island.

The following table, taken from Phillips' memoir, p. 146, is also of interest as showing the development of Smith's work and the completeness of his classification in his later years, and as illustrating how much we are indebted to Smith for our present nomenclature, so much so that as Prof. Sedgwick remarked when presenting the first Wollaston Medal of the Geological Society to Smith, "If in the pride of our present strength, we were disposed to forget our origin, our very speech would bewray us: for we use the language which he taught us in the infancy of our science. If we, by our united efforts, are chiselling the ornaments and slowly raising up the pinnacles of one of the temples of nature, it was he who gave the plan, and laid the foundations, and erected a portion of the solid walls by the unassisted labour of his hands."[6]

[6] The reader may consult an interesting paper by Professor Judd, on "William Smith's Manuscript Maps," Geological Magazine, Decade IV. vol. IV. (1897) p. 439.

Comparative View of the Names and Succession of the Strata.

Table drawn up in 1799.			Table accompanying the map, drawn up in 1812.		Improved table drawn up in 1815 and 1816 after the first copies of the map had been issued.
			London Clay		1	London Clay
			Clay or Brick-earth		2	Sand
					3	Crag
			Sand or light loam		4	Sand
1	Chalk		Chalk		5	Chalk Upper Lower
2	Sand		Green Sand		6	Green Sand
			Blue Marl		7	Brick Earth
			Purbeck Stone, Kentish Rag and Limestone of the vales of Pickering and Aylesbury, Iron Sand and Carstone		8	Sand
					9	Portland Rock
					10	Sand
					11	Oaktree Clay
					12	Coral Rag and Pisolite
					13	Sand
3	Clay		Dark Blue Shale		14	Clunch Clay and Shale
					15	Kelloway's Stone
			Cornbrash		16	Cornbrash
4	Sand and Stone				17	Sand and Sandstone
5	Clay
6	Forest Marble		Forest Marble Rock		18	Forest Marble
					19	Clay over Upper Oolite
7	Freestone		Great Oolite Rock		20	Upper Oolite
8	Blue Clay
9	Yellow Clay
10	Fuller's Earth				21	Fuller's Earth and Rock
11	Bastard ditto and Sundries
12	Freestone		Under Oolite		22	Under Oolite
13	Sand				23	Sand
					24	Marlstone
14	Marl Blue		Blue Marl		25	Blue Marl
15	Blue Lias		Blue Lias		26	Blue Lias
16	White Lias		White Lias		27	White Lias
17	Marlstone, Indigo and Black Marls
18	Red Ground		Red Marl and Gypsum		28	Red Marl
19	Millstone		Magnesian Limestone		29	Redland Limestone
			Soft Sandstone
20	Pennant Street
21	Grays		Coal Districts		30	Coal Measures
22	Cliff
23	Coal
			Derbyshire Limestone		31	Mountain Limestone
			Red and Dunstone		32	Red Rhab and Dunstone
			Killas or Slate		33	Killas
			Granite, Sienite and Gneiss		34	Granite, Sienite and Gneiss

The above table contains a very complete classification of the British Mesozoic rocks, one of the Tertiary strata which is less complete, and a preliminary division of the Palæozoic rocks into Permian (Redland Limestone), Carboniferous (Coal Measures and Mountain Limestone), Devonian (Red Rhab and Dunstone) and Lower Palæozoic (Killas).

Since Smith's time the main work which has been done in classification is a fuller elucidation of the sequence of the Tertiary and Palæozoic Rocks, and this we may now consider.

The Mesozoic rocks are developed in Britain under circumstances which render the application of the test of superposition comparatively simple, for the various subdivisions crop out on the surface over long distances, and the stratification is not greatly disturbed. With the Tertiary and Palæozoic Rocks it is otherwise, for some members of the former are found in isolated patches, whilst the latter have usually been much disturbed after their formation.

Commencing with the Tertiary deposits we may note that "the first deposits of this class, of which the characters were accurately determined, were those occurring in the neighbourhood of Paris, described in 1810 by MM. Cuvier and Brongniart.... Strata were soon afterwards brought to light in the vicinity of London, and in Hampshire, which although dissimilar in mineral composition were justly inferred by Mr T. Webster to be of the same age as those of Paris, because the greater number of fossil shells were specifically identical[7]." It is to Lyell that we owe the establishment of a satisfactory classification of the Tertiary deposits which is the basis of later classifications. Recognising the difficulty of applying the ordinary test of superposition to deposits so scattered as are those of Tertiary age in north-west Europe, he in 1830, assisted by G. P. Deshayes, proposed a classification based on the percentage of recent mollusca in the various deposits. It may be noted, that although this method was sufficient for the purpose, it has been practically superseded, as the result of increase of our knowledge of the Tertiary faunas, by the more general method of identifying the various divisions by their actual fossils without reference to the number of living forms contained amongst them. The further study of the British Tertiary rocks was largely carried on by Joseph Prestwich, formerly Professor of Geology in the University of Oxford.

[7] Lyell, Students' Elements of Geology. 2nd Edition, p. 118.

Amongst the Palæozoic rocks, it has been seen that the Permian, Carboniferous and some of the Devonian beds were recognised as distinct by Smith, though a large number of deposits now known to belong to the last named were thrown in with other rocks as 'killas.' The Devonian system was established and the name given to it in 1838 by Sedgwick and Murchison, largely owing to the palæontological researches of Lonsdale. An attempt was subsequently made to abolish the system, but the detailed palæontological studies of R. Etheridge finally placed it upon a secure basis. The establishment of the Devonian system cleared the way for the right understanding of the Lower Palæozoic rocks, which Sedgwick and Murchison had commenced to study before the actual establishment of the Devonian system, and to these workers belongs the credit of practically completing what was begun by William Smith, namely, the establishment of the Geological Sequence of the British strata. The controversy which unfortunately marked the early years of the study of the British Lower Palæozoic Rocks is well-nigh forgotten, and in the future the names of Sedgwick and Murchison will be handed down together, in the manner which is most fitting.

Our account of the growth of British Stratigraphical Geology is not yet complete. In 1854, Sir William Logan applied the term Laurentian to a group of rocks discovered in Canada, which occurred beneath the Lower Palæozoic Rocks. Murchison shortly afterwards claimed certain rocks in N.W. Scotland as being of generally similar age, and since then a number of geologists, most of whom are still living, have proved the occurrence of several large subdivisions of rocks in Britain, each of which is of pre-Palæozoic age.

The above is a brief description of the growth of our knowledge of the order of succession of the strata which is the foundation of Stratigraphical Geology. A sketch of the manner in which the knowledge which has been obtained has been applied to the elucidation of the earth's history of different times would require far more space than can be devoted to it in a work like the present, but some idea of it may be gained from a study of the later chapters of the book. It will suffice here to remark, that to Godwin-Austen we owe the foundation of what may be termed the physical branch of Palæo-physiography, which is concerned with the restoration of the physical conditions of past ages, while Cuvier and Darwin have exerted the most influence on the study of Stratigraphical Palæontology.

[CHAPTER III.]

NATURE OF THE STRATIFIED ROCKS.

The present constituents of the earth which are accessible for direct study are divisible into three parts. The inner portion, consisting of rocks, is known as the lithosphere; outside this, with portions of the lithosphere projecting through into the outermost part, is the hydrosphere, comprising the ocean, lakes, rivers, and all masses of water which rest upon the lithosphere in a liquid condition. The outermost envelope, which is continuous and unbroken is the atmosphere, in a gaseous condition. It is well known that some of the constituents of any one of these parts may be abstracted from it, and become a component of either of the others; thus the atmosphere abstracts aqueous vapour from the hydrosphere, and the lithosphere takes up water from the hydrosphere, and carbonic anhydride from the atmosphere.

The nebular hypothesis of Kant and Laplace necessitates the former existence of the present solid portions of the lithosphere in a molten condition, and accordingly the first formed solid covering of the lithosphere, if this hypothesis be true, must have been formed from molten material, or in the language of Geology, it was an igneous rock. Consequently, the earliest sedimentary rock was necessarily derived directly from an igneous rock, with possible addition of material from the early hydrosphere and atmosphere, and all subsequently formed sedimentary rocks have therefore been derived from igneous rocks (with the additions above stated) either directly, or indirectly through the breaking up of other sedimentary rocks which were themselves derived directly or indirectly from igneous rocks. The observations of geologists show that this supposition that the materials of sediments have been directly or indirectly obtained for the most part from once-molten rocks is in accordance with the observed facts, and so far their observations testify to the truth of the nebular hypothesis. This being the case, the study of the petrology of the igneous rocks is necessary, in order to arrive at a true understanding of the composition of the sedimentary ones. The igneous rocks are largely composed of four groups of minerals, viz.—quartz, felspars, ferro-magnesian minerals, and ores. Of these the quartz (composed of silica) yields particles of silica for the formation of sedimentary rocks; the felspars, which are double silicates of alumina and an alkali or alkaline earth, being prone to decomposition furnish silicate of alumina and compounds of soda, potash, lime, &c. The ferro-magnesian minerals (such as augite, hornblende and mica) may undergo a certain amount of decomposition, and yield compounds of iron, lime, &c. We may also have fragments of any of these minerals, and of the ore group in an unaltered condition. The composition of a sedimentary rock which has undergone no alteration after its formation will therefore depend upon the character of the rock from which it was derived, the chemical changes which take place in the materials which compose it, before they enter into its mass, and the mechanical sorting which they undergo prior to their deposition.

In the above passage the terms igneous rock and sedimentary rock have been used, and it is necessary to give some account of the sense in which they were used.

An igneous rock is one which has been consolidated from a state of fusion. It is not necessary to discuss here the exact significance of the word fusion, and whether certain rocks which are included in the igneous division were formed rather from solution at high temperature than from actual fusion. This point is of importance to the petrologist, but to the student of stratigraphical geology the term igneous rock may be used in its most comprehensive sense. These igneous rocks were consolidated either upon the surface of the lithosphere or in its interior.

The other great group of rocks is one to which it is difficult to apply a satisfactory name. They have been termed by different writers, sedimentary, stratified, derivative, aqueous, and clastic, but no one of these terms is strictly accurate. The term sedimentary implies that they have settled down, at the bottom of a sheet of water for instance. It can hardly be maintained that limestones formed by organic agency, like the limestones of coral reefs, are sedimentary in the strict sense of the term, and an accumulation like surface-soil can only be called a sediment by straining the term. Stratified rocks are those which are formed in strata or layers, but many of the rocks which we are considering do not show layers on a small scale, and igneous rocks (such as lava-flows) are also found in layers, though such layers are not true strata in the sense in which the term is used by geologists; the term stratified is perhaps the least open to objection of any of those named above. Derivative implies that the fragments have been derived from some pre-existing rock, but as there are many ways in which fragments of one rock may be derived from another, the term is too comprehensive. Aqueous rocks should be formed in water, and most of the class of rocks which we are considering have been so formed, but others such as sand-dunes and surface-soil have not. (The term Aerial or Æolian has been suggested to include these rocks which are thus separated from the Aqueous rocks proper; the objection to this is that the origin of these rocks is closely connected with that of the true Aqueous rocks, and moreover the group is too small to be raised to the dignity of a separate subdivision.) Lastly, the name clastic has been given, because the rocks so called are formed by the breaking up of pre-existing rocks. There are two objections to this name. In the first place, some rocks included under the head clastic are formed by solution of material and its consolidation from a state of solution by chemical or organic agency, though we may perhaps speak of rocks being broken up by chemical as well as by mechanical action. The most important objection is that many clastic rocks are formed by the breaking up of rocks subsequently to their formation, and it has been proposed that rocks of this nature should be termed cataclastic, while those which are formed by the breaking up of pre-existing rocks upon the earth's surface should be termed epiclastic; another group formed of materials broken up within the earth, and accumulated upon its surface as the result of ejection of fragmental material from volcanic vents being termed pyroclastic. This classification is scientific, and under special circumstances is extremely useful, but the older terms have been used so generally, and with this explanation their use is so unobjectionable, that they may be retained, and the term stratified will be generally used to indicate all rocks which are not of igneous origin or formed as mineral veins in the earth's interior.

The division of rocks into three great groups, the Igneous, Stratified and Metamorphic (the latter name being applied to those rocks which have undergone considerable alteration since their formation), is objectionable, since we have metamorphic igneous rocks as well as metamorphic stratified ones. The most convenient classification is as follows:—

It must be distinctly understood that all geological phenomena must be taken into account by the stratigraphical geologist. The upheaval of strata, the production of jointing and cleavage in them, their intrusion by igneous material, their metamorphism, give indications of former physical conditions equally with the lithological characters of the strata, and their fossil contents. Nevertheless it is not proposed to give a full account of the various phenomena displayed by rocks; the student is referred to Text-books of General Geology for this information. It will be as well here, however, to point out in a few words the exact significance of the existence of strata in the lithosphere.

The formation of strata and their subsequent destruction to supply material for fresh strata are due to three great classes of changes. Beginning with a portion of lithosphere composed of rock, it is found that rock is broken up by agents of denudation, as wind, rain, frost, rivers and sea. These agents perform their function mainly upon the portion of the lithosphere which projects through the hydrosphere to form land, and the land is the main area of denudation. The materials furnished by denudation are carried away, and owing to gravitation, naturally proceed from a higher to a lower level, often resting on the way, but if nothing else occurs, ultimately finding their way to the sea, where they are deposited as strata. The sea is the principal area for the reception of this material, and it is there accordingly that the bulk of stratified rock is formed. If nothing else occurred, in time the whole of the land would be destroyed, and the wreckage of the land deposited beneath the sea as stratified rock. As it is there is a third class of change, underground change, causing movements of the earth's crust (to use a term which can hardly be defined in few words but which is generally understood), and as the result of the relative uplift of portions of the earth's crust, the stratified rocks formed beneath the oceans are raised above its level, giving rise to new masses of land, which are once more ready for destruction by the agents of denudation. This cycle of change (all parts of which are ever proceeding simultaneously) is one of the utmost importance to the stratigraphical geologist.

Stratification is the rock-structure of prime importance in stratigraphical geology, and a few words must here be devoted to its consideration, leaving further details to be dealt with hereafter. The surface of the ocean-floor is, when viewed on a large scale, so level, that it may be considered practically horizontal, and accordingly in most places the materials which are laid down on the ocean-floor give rise to accumulations which at all times have a general horizontal surface (when the ocean-slopes depart markedly from horizontality the deposits tend to abut against these slopes rather than to lie with their upper surfaces parallel to their original angle). A practically horizontal surface of this character may give rise to a plane of stratification (or bedding-plane) in more than one way. A pause may occur during which there is a cessation of the supply of material, so that the material which has already been accumulated has sufficient time to become partially consolidated before the deposition of fresh material upon it. In this way a want of coherence between the two masses is produced, along the plane of junction, which after consolidation of the deposits causes an actual divisional plane along which the two deposits may be separated. This is a plane of stratification. The pause may be produced in various ways, sometimes between successive high tides, at others as the result of physical changes which may have taken ages to happen. Again, after material of one kind has been deposited, say sand, some other substance such as clay may be accumulated on its upper surface, giving rise to a plane of stratification between two deposits of different lithological characters. If this occurs alone, there may be actual coherence between the two strata, so that it is erroneous to speak of a plane of stratification as if it were always one along which one deposit could be readily split from the other, though as a general though by no means universal rule, change from one kind of deposit to another is also marked by want of coherence between the two. The material between two planes of stratification forms a stratum or bed, though if the deposit be very thin it is known as a lamina, and the planes are spoken of as planes of lamination (no hard and fast line can be drawn between strata and laminæ; several of the latter usually occur in the space of an inch).

A stratum will have its upper and lower surface apparently parallel, though not really so, for no stratum extends universally round the earth, and many of them disappear at no great distance when traced in any direction. Parts of one stratum may be composed of different materials from other parts when traced laterally, thus one stratum may be found composed essentially of sand in one place, of mud in another, and of a mixture of the two in an intervening locality. Whatever be the composition of a stratum it dies out eventually, owing to the coming together of the upper and lower bounding planes of stratification. The stratum is thickest at some spot, from that spot it becomes thinner in all directions, until it disappears at last by the coalescence of the bounding-planes. This is spoken of as thinning-out. Strata, then, consist of lenticular masses of rock, separated from the underlying and overlying strata by planes of stratification. The shape of the lenticle may vary immensely, the thickness bearing no definite relationship to the horizontal extent. Some strata, many feet in thickness, may thin out and disappear completely in the course of a few yards, whilst others an inch or two in thickness may be traced horizontally for many miles. We often find thin strata of coal and limestone, extending for great distances, strata of mud thinning out more rapidly, and sandstones still more rapidly, but no universal rule connecting rapidity of thinning-out with composition of the strata can be laid down.

Having seen what a stratum is, it now remains to speak of the composition of the stratified rocks. They have been classified according to their composition, and according to their origin. According to composition they have been divided into:

Arenaceous rocks, composed essentially of grains of sand.

Argillaceous rocks, composed essentially of particles of mud.

Calcareous rocks, composed essentially of particles of carbonate of lime.

Carbonaceous rocks, composed largely of hydrocarbon compounds.

Siliceous rocks, composed essentially of silica not in the form of grains;

whilst according to their origin they have been separated into:—

Mechanically-formed rocks, composed of fragments derived from other rocks by mechanical fracture.

Chemically-formed rocks, composed of particles which have been chemically deposited from a state of solution.

Organically-formed rocks, composed of materials which have been derived from a state of solution or from the gaseous condition by the agency of organisms.

Whichever classification be adopted (and each is useful for special purposes), it must be noted that no hard and fast line can be drawn between one division and another. A rock may be partly arenaceous and partly calcareous, composed of a mixture of sand and lime, and the same rock may similarly be partly mechanically and partly organically formed, the sand being due to mechanical fracture, and the lime to the agency of organisms, and so with the other divisions.

As many of the changes which have occurred in past times have been concerned in destruction and obliteration, whilst deposition is the cause of preservation, the study of deposits is peculiarly adapted for testing the truth of the grand principle of geology that the changes which have taken place in past times are generally speaking similar in kind and in intensity of action to those which are in progress at the present day, and a study of the modern deposits is specially important as throwing light upon the characters of those which have been formed in past times. It will be abundantly shown in the sequel that the deposits of the strata are in general comparable in all essential respects with those which are being formed at present, and accordingly they give most valuable indications as to the nature of the physical and other conditions under which they were laid down. The desert sand, the precipitate of the inland sea, the reef-limestone and many another deposit can thus be detected by an examination of their lithological characters, combined with consideration of other kinds of evidence. The petrology of the sedimentary rocks is still in its infancy, though much has already been done, but it offers a wide field of inquiry to the field-geologist and worker with the microscope[8].

[8] The student will do well to consult The Challenger Report by Messrs Murray and Renard (1891), for information concerning many modern sediments, and Harker's Petrology for Students Section D, for general information on the Petrology of the Stratified Rocks.

[CHAPTER IV.]

THE LAW OF SUPERPOSITION.

In a previous chapter this law was given as follows: "Of any two strata, the one which was originally the lower is the older;" the general truth of the law depends upon the fact that except under very exceptional circumstances the strata are deposited upon the surface of the lithosphere, and not beneath it. There are occasions where strata may be deposited beneath the lithosphere, but as a general rule the geologist will not be misled by such occurrences. In caverns, accumulations often occur which are newer than the strata over the cavern roof, and so long as caverns are formed in ordinary sedimentary rocks, no great difficulty will result from this exception to the law of superposition. When caverns occur beneath masses of land ice, the order of superposition may be misleading. A deposit may be formed on the surface of the ice, and subsequently to this a newer deposit may be laid down in a sub-glacial or englacial cavern; upon the melting of the ice the newer deposit would be found with the older one resting upon its surface.

Apart from these exceptional cases, the law as stated holds good, but the reader will notice the insertion of the word 'originally' which requires some comment.

A geologist speaks of one bed lying upon another not only when the beds are horizontal, but when they are inclined at any angle, until they become vertical, so that until beds have been turned through an angle of 90° by earth-movement the test of superposition is applicable, but when they have been turned more than 90°, the stratum which was originally lower rests upon that which was originally above it, and in the case of these inverted strata, the test of superposition is no longer applicable. It was formerly supposed that cases of inversion were comparatively rare and local, and that the test of superposition could therefore be generally applied with confidence, but it is now known that though this is generally true of such strata as have been subjected only to those widespread, fairly uniform movements which are spoken of as epeirogenic or continent-forming, where the radius of each curve is very long, inversion is a frequent accompaniment of the more local orogenic or mountain-forming movements, where the radius of a curve is short. Though orogenic movements are limited as compared with those of epeirogenic character, they often affect large tracts of country, in which case the apparent order of succession of the strata need not be the true one, and examples of inversion may be frequent[9].

[9] For a discussion of the principles of mountain-building see Heim, A., Untersuchungen über den Mechanismus der Gebirgsbildung, and Lapworth, C. "The Secret of the Highlands," Geological Magazine, Decade II. vol. x. pp. 120, 193, 337.

It is not easy to lay down any definite rules for detecting inverted strata, where the top of an inverted arch is swept off by denudation or the bottom of an inverted trough concealed beneath the surface, beyond stating that if an easily recognised set of beds is obviously repeated in inverse order, inversion must have occurred, though even then it may not be clear which side of the fold shows the beds in original and which in inverted sequence. Suggestions are frequently made that ripple-marks and worm-tracks may be utilised in order to discover inversion, for the well-formed ripple-marks will appear convex on the upper surface of a bed which is not inverted, and we may note concave casts of these ripple-marks on the under surface of the overlying bed, whilst worm-tracks will appear concave on the upper surface, and their casts convex on the lower surface of the succeeding bed under similar conditions. In the case of inversion the occurrences will be the exact opposite to these. Unfortunately ripple-marks and worm-tracks may, as will appear in the sequel, be simulated by structures produced in quite a different way, and unless the observer is certain that he is confronted with true ripple-marks and worm-tracks he may be seriously misled. The geologist must take into account all the evidence at his disposal, when he is dealing with cases of possible inversion, but oftentimes he will after due consideration of all the phenomena be left in doubt unless he is able to supplement his observations on the succession of the strata by evidence derived from the included fossils.

The test of superposition is most apt to be misleading when the strata have been affected by the faults known as reversed faults or thrust-planes.

Reference to text-books will show that a fold consists of two parts, the arch and the trough, and that the two are connected by a common-, middle-, or partition-limb. In the case of an inverted fold, an S-shaped or sigmoidal structure is the result ([Fig. 1 A]).

Fig. 1.

A. A sigmoidal fold, showing a bed xx in an overfold with arch (a), trough (t) and common limb c.

B. A similar bed xx affected by a thrust-plane tt which replaces the common limb.

Here the portions of any bed (xx) which occur in the arch or trough are in normal position, and have not been moved round through an angle of 90°, whilst the portion which occurs in the common limb c has been moved round through an angle greater than 90° and is inverted, so that its former upper surface now faces downwards. In Fig. 1 B the common limb is replaced by a reversed fault, or thrust-plane, and the inverted portion of the bed seen in the common limb is therefore absent. An observer, applying the test of superposition, might suppose that the position of the bed x on the left-hand side of the figure was a different bed from the portion which is seen on the right-hand side, instead of belonging to the same bed, and in this way, if a number of parallel thrust-planes affected one bed or a set of beds, he might be led to infer the occurrence of a great thickness of strata where there was in reality a slight thickness, or even one bed only repeated again and again by faulting. It is quite certain that exaggerated estimates of the thickness of deposits have frequently been made owing to the non-recognition of the occurrence of repetition as a consequence of the existence of thrust-planes.

Where thrust-planes are suspected, it is well to look for some of the following features:

(a) The strata of a country affected by thrust-planes often crop out as lenticular masses, thinning out rapidly in the direction of the strike[10]. This is true of beds thrown into sharp folds whether or not inverted, but the lenticles will be wider in a direction at right angles to that of the strike as compared with their length when inversion has not occurred. It is also true of beds which were originally deposited as lenticles, such as many massive sandstones, and as almost any kind of deposit may be formed originally as a lenticle, the test by itself is by no means sufficient as a proof of thrusting, though it is suggestive.

[10] For definitions of the terms dip, strike, outcrop and allied expressions, the reader is referred to a Text-Book of Geology.

(b) The surfaces of the strata are often affected by the striations known as slickensides, and the joint-faces of gently inclined beds are also frequently marked by slickensides which often run in a nearly horizontal direction.

(c) A parallel structure presenting the appearances characteristic of the mechanically-formed features of a foliated rock is often developed, and one or more of certain accompanying phenomena will probably be found, which will be noticed more fully in a later chapter.

(d) Extension or stretching of the rocks will have been frequently produced, causing rupture, and the resulting fissures are usually filled with mineral-veins, though this occurrence is by no means characteristic of rocks which have been affected by thrust-planes.

(e) Chemical changes may have occurred which have resulted in the reconstitution of some of the rock-constituents, which may crystallise where pressure is least, thus we often find rocks which have undergone movements of the type we are considering marked by the existence of sericitic films upon the surfaces.

Another reservation must be made when considering the law of superposition. The test is only applicable for limited areas. Suppose we find a deposit of clay a resting upon another deposit of limestone b in the south of England, and can prove that the apparent succession is the true one, that is, that there has been no inversion; it is clear that the test of superposition is applicable in that area. Now, we may be able to trace the two deposits continuously across the country, one as a clay, the other as a limestone; so that when we reach the north of England we find the clay a still reposing upon the limestone b. The test of superposition is applicable in that area also, the clay of the northern area being newer than the limestone of the same region. But, for reasons which will ultimately appear, it by no means follows that the clay of the north is newer than the limestone of the south, although the two deposits are continuously traceable with the same lithological characters; it may have been formed simultaneously with the limestone of the south, or even before it. Something more, therefore, than the test of superposition is necessary in order to make out the relative ages of continuous deposits in a wide region, and this is still truer in the case of deposits which are discontinuous, whether separated from one another by the sea, or by outcrops of older or newer rocks.

A few words of warning may be added with reference to the detection of bedding-planes. A bedding-plane is one which separates two beds, and its existence is determined during the deposition of the beds. Many other planes are formed in rocks subsequently to their deposition, and it is not always easy to distinguish these from true bedding-planes. That even experienced observers may be led astray is shown by the fact that, of recent years, it has been proved that great masses of rock have been claimed as of sedimentary origin, and their apparent order of succession noted, which are in truth naught but irregular masses of intrusive igneous rocks affected by divisional planes which simulate bedding, produced in the rocks subsequently to their consolidation. Joints, faults, and cleavage-planes may all at times simulate planes of bedding, and it is frequently very difficult to distinguish them in the limited exposures with which a geologist has oftentimes to deal. It is easier to make suggestions for distinguishing bedding-planes from other planes which simulate them, than to apply the suggestions in practice, and the student of field geology will find that experience is the only guide, though after years of experience he may be confronted with cases where the evidence is insufficient to convince him that he is dealing with planes of stratification and not with some other structure.

From what has been remarked, it will be inferred that the test of superposition though of prime importance to the geologist is frequently insufficient to enable him to ascertain the true order of succession of the strata, and he is compelled to supplement this test by some other. There are several useful physical tests which may frequently be applied. Thus, if a rock a contains fragments of another rock b, under such circumstances as to show that the fragments of b were included in a during its deposition, it is clear that b is older than a. Here again, it will be found from what appears in a later chapter that the student is confronted with difficulties when actually examining rocks, for fragmental rocks of cataclastic origin, where the fragments have been formed as the result of fracture produced by earth-movements subsequently to the deposition of the rock, simulate epiclastic rocks in which the fragments were introduced during the accumulation of the deposits to so surprising a degree as sometimes to baffle the most experienced observer. Not only are the fragments of these cataclastic rocks broken up, but they may be further rounded so as to imitate in a remarkable manner the water-worn pebbles of an epiclastic conglomerate. Again, an older series of rocks may have had structures impressed upon them as the result of changes subsequent to their formation, and before the formation of a newer set which the latter therefore do not exhibit. Jointing, cleavage, and various metamorphic phenomena may thus be exhibited by the older rocks, but great care is required in applying this test, especially with a limited thickness of rocks, as one set may not exhibit the structures not because they were not in existence when the structures were developed, but because their nature is such that they were incapable of receiving or retaining the structures. For instance a mass of grit which is older than a mass of clay-slate may not be cleaved, because, although subjected to the pressure which produced the cleavage, it was of a nature not adapted to the development of cleavage structure.

On the whole, application of tests dependent upon physical features of rocks, does not often supplement to any great extent the information supplied by ascertaining the order of superposition, and in all cases, where possible, every other kind of information should be supplemented, by that which is acquired after examination of the included organisms of the strata.

[CHAPTER V.]

THE TEST OF INCLUDED ORGANISMS.

The second great law of the Stratigraphical Geologist is that fossiliferous strata are identifiable by their included organisms, in other words, that we can tell the geological age of deposits by examination of the fossils contained in them, though the determination of age must be given in more general terms in some cases than in others. Considerable misconception has arisen concerning the value of fossils as indices of age, and it is necessary therefore to discuss the significance of the law of identification of strata by their included organisms at some length.

The comparison between fossils and medals has frequently been made and fossils have well been styled the "Medals of Creation"; and the significance of fossils as guides to the age of deposits may perhaps be made clearer if we pursue this comparison some way. In the first place there is clear indication of a gradual increase in the complexity of organisation of the fossils as one passes from the earlier to the later rocks, and accordingly the general facies of a fauna is likely to furnish a clue to the age of the rocks in which it is found, even though every species or even genus represented in the fauna was previously unknown to science. So an antiquary versed in the evolution of art or metallurgy, might detect the general age of a medal with whose image and superscription he was not acquainted. He would know that a medal struck in iron was formed subsequently to the bronze age, or that one formed of palladium appertained to the present century. But quite apart from any theoretical knowledge, an antiquary would find as the result of accumulated experience that certain medals are characteristic of certain periods; he would learn that the denarius is characteristic of a different period from that indicated by the coin of the Victorian era, even though he had no knowledge of the technicalities of numismatics. The same is the case with the geologist. He may not be influenced by any knowledge of the evolution of faunas and floras, but actual work amongst the rocks will show him that the trilobite is not found with the belemnite or the ichthyosaur with the elephant, save under exceptional circumstances, which only prove the rule, as for instance when worn bones of ichthyosaurs are washed from their proper strata into gravels with elephant-bones.

It must be distinctly understood that the determination of fossils as characteristic of different periods is solely made as the result of experience. No à priori reasoning may give one indication of the actual range in time of a species or genus; no one can say why Discina has a long range in time, whilst that of the closely related Trematis is very limited. This being the case, the greater the mass of evidence which is accumulated as to the range of a fossil, the greater will be the value of that fossil as a clue to the age of the deposit in which it is found. This is so important, that it requires more than mere notice. If a fossil is found in abundance in a group of strata B in any one area, and is not found in an underlying group A or overlying group C in that area after prolonged search, we may confidently speak of the fossil as characteristic of the strata B in that area, though without further work, the value of the fossil as a clue to age in other areas would be unproved. It may nevertheless happen, that after more prolonged search in A or C, in the original area a few specimens of the fossil which has been spoken of as characteristic of B may be found in one or other of them, in small quantity. The value of the fossil as one characteristic of B will be slightly diminished, though only slightly, as it is not likely to turn up in numbers in the strata A or C after the prolonged search. Should the fossil be found also to be characteristic of the strata B in areas other than the original one, it becomes of more than local value, and if, after much study it is found to characterise the same strata over wide areas, the cumulative evidence now obtained will render the fossil peculiarly important to the stratigraphical geologist. The detection of characteristic fossils is not quite so simple as might be supposed from the above remarks, for examination of the position of one fossil will not prove the contemporaneity of beds in different places, to prove this, all the evidence at our disposal must be considered, for reasons which will be presently pointed out.

As the result of accumulated knowledge, we can now compile lists of characteristic fossils of the major subdivisions of the strata, which are of world-wide utility and as our knowledge increases, we are enabled to subdivide the strata into minor divisions of more than local value.

What is a fossil? Before discussing the value of fossils as aids to the stratigraphical geologist, it may be well to make a few observations as to what constitutes a fossil. It is difficult to give any concise definition, and as is often the case in geology, an explanatory paragraph is of more value than a mere definition. The term fossil was originally applied to anything dug up from the rocks of the earth's crust, and was used with reference to inorganic objects as well as organic remains, for instance minerals were spoken of as fossils. It is now applied essentially though not exclusively to relics of former organisms, though one still reads of fossil rain-drops, fossil sun-cracks, and so on. Furthermore, the relics need not necessarily be parts of the organism, the track of a worm or a bird's nest if embedded in the strata would be termed a fossil. It is generally agreed that no sharp line can be drawn between recent and fossil organic remains which is based upon the degree of mineralisation (or as it was sometimes termed petrifaction) of the relics, for many true fossils have not undergone mineralisation, subsequent to their entombment.

It has been suggested that the name fossil should be applied to organic remains which have been entombed by some process other than human agency, but this restriction is undesirable. The stone-implement of the river gravels is as genuine a fossil as the ammonite extracted from the chalk, and the human relics of very recent date may give information of a character quite similar to that supplied by other remains, for instance, the occurrence of moa-bones in New Zealand in accumulations below those containing biscuit-tins and jam-pots has been used as a geological argument pointing to the extinction of the moa before the arrival of Europeans in New Zealand. The biscuit-tin here serves all the purposes of a fossil, and there is no valid reason why it should not be spoken of as such.

This statement brings one to consider another method which has been adopted in order to separate fossil organisms from recent ones, namely the time-test. This again is inapplicable, for no line can be drawn between the shell which was buried in yesterday's tidal deposit and that which has lain in the strata through geological ages, and each may be equally useful to the geologist.

Whilst, then, we can give no definition of fossil which is likely to meet with general acceptance, the term can be so used, as not to give rise to any doubts as to its meaning, and it is generally applicable to any organic relics which have been embedded in any deposit or accumulation by any agent human or otherwise.

Mode of occurrence of fossils. It will not be out of place to say a few words as to the way in which fossils are found in strata, as beds are often inferred to be unfossiliferous, because of ignorance of methods which should be pursued in searching for organic relics. It is unnecessary to dilate upon the actual modes of preservation of organisms, which is treated of fully in other works. In the first place, it is rash to assert that any deposit is unfossiliferous because no fossils have been found in it, even after prolonged search. The Llanberis slates had been eagerly searched for fossils for many years without result, but that the search was not exhaustive was proved by the discovery of trilobites in them some years ago. Seekers after fossils are rather prone to confine their attention to strata which are already known to be fossiliferous than to pay much attention to those which have hitherto yielded no organic remains.

Some kinds of deposits are more often fossiliferous than others. Limestones which are frequently largely of organic origin, are often rich in remains, and muddy deposits more frequently furnish fossils than those of a purely sandy nature. The difference in the yield is not necessarily due to the original inclusion of more remains in one rock than in another, but is often caused by the obliteration of former relics owing to changes which have taken place in the rocks subsequently to their deposition. No sedimentary rock must be regarded as unfossiliferous, however unfitted it appears for the preservation of fossils. The writer has seen fossils, not only in coarse conglomerates, rocks which frequently contain no traces of organisms, but in deposits composed largely of specular iron ore, and even in intrusive igneous rocks, though in the latter case, the inclusion of fossils was due to circumstances which cannot have occurred with frequency.

In sandy strata, the substance of the fossils has often been completely removed, leaving hollow casts, which may be almost or quite unrecognisable. In these circumstances, much information may be obtained by taking impressions of the casts in modelling wax or some other material. The importance of this process may be judged from the results it yielded to Mr Clement Reid in the case of the fossils of the Pliocene deposits occurring in pipe-like hollows in the Cretaceous rocks of Kent and the discovery of the remarkable reptiles described by Mr E. T. Newton from the Triassic sandstones of Elgin.

In argillaceous rocks which have been affected by the processes producing cleavage, the fossils may be distorted beyond recognition or owing to the difficulty of breaking the rocks along the original planes of deposition, may remain invisible. Under such circumstances, small nodules of sandy or calcareous nature may sometimes be found included in the argillaceous deposits and may perhaps yield fossils. Oftentimes, also, where the argillaceous rock is in close proximity to a harder rock, such as massive grit, the argillaceous rock in close contiguity to the hard rock may escape the impress of cleavage-structure, and fossils may be readily extracted from rocks in this position when not obtainable from other parts of the deposit. It was under these circumstances that the trilobites alluded to above were obtained from the Llanberis slates.

The fossils of calcareous rocks are often very obvious, but difficult to extract, as they break across when the rock is fractured. They are frequently obtainable in a perfect condition when the rock is weathered. Occasionally they may be extracted from certain argillaceous limestones if the limestone be heated to redness, and suddenly plunged into cold water. Fossils are often found in a state which enables them to be readily extracted when a limestone is coarsely crystalline, though they cannot be extracted in a perfect condition when the same limestone is in a different state.

Many microzoa, which are invisible in rocks, even when viewed through a lens, may be found in microscopic sections of calcareous and silicious rocks, and plant structures may be detected under similar circumstances in the case of carbonaceous rocks.

Various special methods of extracting fossils from rocks have been described by different writers, many of which are very complex, and require much time. The mechanical action of the sand-blast and the solvent action of various acids as hydrochloric and hydrofluosilicic have been found of use upon different occasions[11]. The various processes which have been utilised in order to extract and develop fossils can, however, be best learned by information obtainable from curators of palæontological collections, and by actual experience, and there is yet much information to be acquired as to the manner of extracting fossils from various kinds of rocks.

[11] For information concerning use of acids see especially Wiman, C. "Ueber die Graptoliten," Bull. Geol. Inst., Upsala, No. 4, vol. II. Part II.

Relative value of fossils to the Stratigraphical Geologist. It has been hinted above that no general rule as to the relative value of fossils as guides to the age of strata can be laid down, and that the ascertainment of their relative value is largely the result of actual experience. It may be noted, however, that organisms which possess hard parts are naturally more important to the geologist than those which do not, as few traces of the latter are preserved in the fossil state, and even when preserved are usually too obscure to be of much practical use. Of the organisms which do possess hard parts, different groups have been utilised to a different degree, and one group will be more or less important than another, according to the use to which it is applied. Groups of organisms which have a long range in time are naturally useful for the identification of large subdivisions of the strata, whilst those which have had a shorter range are valuable when separating minor subdivisions.

Again, as the bulk of the sedimentary deposits has been formed beneath the waters of the ocean, relics of marine organisms are naturally more useful than those of freshwater ones. Other things being equal, the more easily the organism is recognisable, and the more abundant are its remains, the greater its value to the stratigraphical geologist, and as the remains of invertebrates are usually found in greater quantities and in more readily recognisable condition than those of the vertebrates, they have been used more extensively as indices of age. Of the invertebrates, the mollusca are often very abundant, their remains are adapted for preservation, and their characteristics have been extensively studied, and accordingly they have been and are of great use to the geologist. Of other groups, the graptolites, corals, echinids, brachiopods, and trilobites have been very largely utilised. The Lower Palæozoic strata have been divided into numerous groups, each characterised by definite forms of graptolites, and a similar use has been made of the ammonites in the case of the Mesozoic rocks. It is not to be inferred that these groups of organisms are naturally more useful than other groups, on account of the extent to which they have been used; we can merely state that they have been proved to be useful as the result of prolonged study; when other groups have received equal attention, they may well be found to be equally useful for the purposes which we have in view.

Contemporaneity and Homotaxis. From what has been already stated, it will be recognised that the ages of the various fossiliferous rocks of the geological column[12] in any one area can be identified with greater or less degree of certainty by reference to their included organisms, the various subdivisions being marked by the possession of characteristic fossils, and it will be naturally and rightly inferred that the greater the number of characteristic fossils of any one deposit, the more certain is the identification of that deposit. In practice, geologists are wont to ascertain the age of the strata after consideration of all the fossils found therein, some of which may be actually characteristic whilst many may come up from the strata below, or pass into those above. Having ascertained the order of succession and fossil contents of the strata in various regions, it is the task of the geologist to compare the strata of these two regions, and this task is fraught with considerable difficulty. Much controversy has arisen as to the degree of accuracy with which strata of remote regions can be correlated, and the subject is one which requires full consideration.

[12] Although the rocks do not always lie on one another in regular succession, it is often convenient to speak of them as though they did, and as though a column of strata could be carved out in any region consisting of horizontal bands of deposit one above another. We speak of such an ideal arrangement as constituting a 'geological column.'

Suppose that a series of strata which we will call A, B, and C is found in any one area, each member of which contains characteristic fossils which enable it to be recognised in that area, and we will further suppose that in another area a series of strata A´, B´, and C´ is discovered, of which A´ has the fauna of A in the former area, and similarly B´ the fauna of B, and C´ that of C.

It cannot be assumed that the stratum A is therefore contemporaneous with A´, B with B´, and C with C´, but on the other hand, it must not be assumed that they are not contemporaneous. This is a statement which requires some comment. It has been urged that if the deposits A and A´ in different localities contain the same fauna, this is a proof that the two are not contemporaneous, for some time must have elapsed in order to allow of the migration of the organisms from one area to another, it being justifiably assumed that they did not originate simultaneously in the two areas. But everything depends on the time taken for migration as compared with the period of existence of the fauna. If the former was extremely short as compared with the latter it may be practically ignored, for we might then speak of the strata as contemporaneous, just as a historian would rightly speak of events in the same way which occurred upon the same afternoon, though one might have happened an hour before the other. Let us then glance at the evidence which we have at our disposal, which bears upon this matter.

The objection to identification of strata with similar faunas as contemporaneous was urged by Whewell, Herbert Spencer, and Huxley, and the latter suggested the term Homotaxis or similarity of arrangement as applicable to groups of strata in different areas, in which a similar succession of faunas was traceable, maintaining that though not contemporaneous the strata might be spoken of as homotaxial. Huxley went so far as to assert that "for anything that geology or palæontology are able to show to the contrary, a Devonian fauna and flora in the British Islands may have been contemporaneous with Silurian life in North America, and with a Carboniferous fauna and flora in Africa[13]," a statement which few if any living geologists will endorse. If the statement be true, and the fauna which we speak of as Devonian, when present be always found (as it is) above that which we in Britain know as Silurian and below that which we term Carboniferous, the faunas must have originated independently in the three centres, and disappeared before the appearance of the next fauna, or having originated at the same centre, each must have migrated in the same direction, spread over the world, and become extinct as it reached the point or line from which it started. Suppose for instance a fauna A originates at the meridian of Greenwich, migrates eastward, and dies out again when it once more reaches Greenwich, that B and C do the same, at a later period, then the fauna B will always be found above A and C above B, but if B did not become extinct when it reached the Greenwich meridian, it would continue its eastward course, and C having in the meantime started on its first round, the fossils of the fauna B would be found both above and below those of C. It will be shown below that cases of recurrence do occur, but nowhere do we find a Silurian fauna above a Devonian one, or a Devonian one above one belonging to the Carboniferous, nor is the fauna of a great group of rocks found in one region above the fauna of another group, and in another region below the same. And this is true not only of the faunas of one major division, such as those of the Silurian and Carboniferous periods, but also of the faunas of many minor subdivisions into which the large ones are separated, for instance we do not find the Llandovery fauna of the Silurian period which in Britain is found below the Wenlock fauna embedded elsewhere in strata above the Wenlock. I have simplified the statement by assuming that the faunas are identical in the different localities, and exactly similar throughout the whole thickness of the containing strata, which is naturally not the case, but the additional complexity does not conceal the truth of what has been stated. In the absence of actual inversion of well-marked faunas, only one explanation is possible, namely, that the time for migration of forms is so short as compared with the entire period during which the forms existed, that it may be practically ignored, and the strata containing similar faunas may be therefore spoken of truthfully as contemporaneous and not merely homotaxial[14].

[13] Huxley, T. H. "Geological Contemporaneity and Persistent Types of Life," being the Anniversary Address to the Geological Society for 1862; reprinted in Lay Sermons, Addresses and Reviews.

[14] For fuller discussion of this matter see a paper by the Author 'On Homotaxis,' Proc. Camb. Phil. Soc., vol. VI. Part II. p. 74.

Apparent anomalies in the distribution of fossils. There are several occurrences which have tended to augment the distrust frequently felt concerning the value of fossils as indices of the age of the beds in which they occur, which may be here considered.

Though the greater number of fossil remains belonged to organisms which lived during the time of accumulation of the deposits in which they are now embedded, this is by no means universally the case, and the occurrence of remanié fossils, which have been derived from deposits more ancient than the ones in which they are now found is far from being a rare event. The existence of remains of this nature in the superficial drifts and river-gravels of our own country has long been recognised, and no one would suppose that the Gryphæa and other shells furnished by these gravels had lived contemporaneously with the species of Corbicula, Unio and other molluscs which are part of the true fauna of the gravels. In this case the water-worn nature of the remains is a good index to their origin, but in other cases, it is by no means an infallible guide, for we sometimes find on the one hand that remains of organisms proper to the deposits in which they occur are water-worn, whilst on the other the relics of remanié fossils are not. The now well-known gault fossils of the Cambridge Greensand at the base of the chalk were not always recognised as having been derived from older beds, and there are certain fossils found in nodules in the Cretaceous rocks of Lincolnshire, which still form a subject for difference of opinion, for while some writers maintain that they belong to the deposits in which they are now found, others suppose that the nodules have been washed out of earlier beds.

Occasionally we find forms which occurring in a set of beds A in an area, are absent from the overlying beds B, and appear again in the succeeding deposits C. Such cases of recurrence are by no means rare, though many supposed instances of recurrence have been recorded as the result of stratigraphical or palæontological errors. The best examples have been noted by Barrande among the Lower Palæozoic deposits of Bohemia. The stage D of Bohemia consists of five 'bandes' or subdivisions, the lowest (d 1), central (d 3) and uppermost (d 5) divisions are mainly argillaceous, whilst the second (d 2) and fourth (d 4) are essentially arenaceous. Some of the forms found in d 1, d 3 and d 5 have not been found in d 2 and d 4. The best-known example is the trilobite Æglina rediviva. It is clear that this and other forms did not become extinct during the deposition of the strata of d 2 and d 4, though they may have disappeared temporarily from the Bohemian area, or else lingered on in such diminished numbers that their remains have not been discovered. The range of the organism is in fact right through the deposits of the stage D, and the discontinuity of distribution is not a real anomaly; it may be compared to some extent with cases of discontinuous distribution in space. It is needless to remark that the whole fauna does not disappear for a time and then reappear, but only a few out of the many forms which compose it. The comparative rarity of examples of recurrence after long intervals is an indication that the palæontological record as it is termed is not so imperfect as some suppose, for if our knowledge of fossils were very imperfect, we should expect cases of apparent recurrence to be common, as the result of the non-detection of fossils in the intermediate beds. One of the most marked cases of apparent recurrence known some years ago was the reappearance of a genus of trilobite Ampyx in Ludlow rocks, found in the Bala rocks, but not in the Llandovery or Wenlock strata. It has since been discovered in Llandovery beds, and its eventual discovery in beds of Wenlock age may be regarded as certain. A supposed case of recurrence which would have been remarkable, that of the disappearance of Phillipsia in Ordovician rocks, its entire absence in those of Silurian age, and its reappearance in the Devonian, has broken down, for the supposed Ordovician form has been shown to belong to an entirely different group of trilobites from that containing the genus Phillipsia, and it has been therefore renamed Phillipsinella.

Many apparent anomalies of distribution have been explained as due to migration, but it is doubtful whether any one of these supposed anomalies is actual and not due to errors in determining the position of the beds or the nature of their included fossils. Some of the supposed anomalies have already been shown to be due to error, and the others will almost certainly be cleared up. In speaking of anomalies of distribution, the geologist can only be guided by experience as to what constitutes an anomaly. For instance the existence of a complete fauna in any one place in the beds of a system above that to which it is elsewhere confined would be regarded as anomalous and as probably due to error, whilst the reappearance of several forms in beds of a system higher than that in which they had hitherto been found, could hardly be considered as an anomaly. A geologist would suspect the statement that after the disappearance of an Ordovician fauna in an area and its replacement by a Silurian fauna, the Ordovician fauna reappeared for a time, but would not regard the statement that a Cenomanian fauna partly reappeared in the Chalk Rock with surprise.

The existence of a Silurian fauna in Ordovician times was maintained by Barrande in the case of the Bohemian basin. Lenticular patches of Silurian rocks having the lithological characters of the Silurian strata are found in the Ordovician beds of that region, and they contain fossils specifically identical with those of the Silurian rocks. Barrande explained this appearance as due to the existence of a fauna in other regions resembling the Silurian fauna of Bohemia, during the Ordovician period, when the normal Ordovician fauna of Bohemia inhabited that area. He supposed that in parts of the basin, when favourable conditions arose, colonies of the foreign fauna settled for a time, but did not get a permanent footing in the basin until the commencement of Silurian times. The theory of colonies has now been rejected for the Bohemian area, and the phenomena shown to be due to repetition of strata by folding and faulting, but it is a theory which is again and again advocated in order to explain apparently anomalous phenomena in other areas, and these apparent anomalies which are so explained, must be regarded with grave suspicion.

The various complexities alluded to in the foregoing pages increase the difficulty experienced by the geologist in correlating strata in different areas by their included organisms, but no one of them disproves the possibility of making these correlations, which can be carried on to a greater or less extent according to the nature of the faunas.

A good deal of misconception has arisen concerning the geographical distribution of former faunas, owing to the tendency to compare them exclusively with the littoral faunas of the present day. These littoral faunas have a comparatively limited geographical distribution, the forms of one marine province often differing considerably from those of an adjoining one, and still more widely from one which is remote, so that anyone confronted with the relics of faunas from the existing Australian and European seas, would find no indications furnished by identity of species that the faunas were contemporaneous. Recent researches have shown, however, that the creatures whose remains are deposited at some distance from the coast-line have a much stronger resemblance to one another than the littoral organisms have, if the fauna of two distant areas be compared. It is still a moot point which will be discussed in a later chapter, how far the deep-sea deposits of modern times are represented amongst the strata of the geological column by deposits of similar origin. But it is certain that many of the ancient strata are not littoral deposits, and it will be found that it is by comparison of the faunas of the deeper-water deposits that the geologist correlates the strata of remote regions: where shallow water deposits are formed, the faunas differ markedly in different regions, and these shallow-water forms can only be correlated owing to their occurrence between deeper-water strata. Thus if strata A, B and C be found in one area, and the fauna of A and C are deep-water forms, those of B being shallow-water forms, and in another area beds A´ contain the same fauna as A, and C´ the same fauna as C whilst the fauna of B´ is different from that of B, we can nevertheless correlate the strata B and B´ (if they be conformable with the underlying and overlying beds), because of the identity of age of the associated beds in the two areas. It will possibly be found that the strata A and C can be further subdivided into A₁, A₂, ... &c. C₁, C₂, ... by the existence of minor faunas, which are comparable in the two cases, but such subdivisions may not be established in the case of the beds B and B´.

To take actual examples:—The Llandovery beds of Dumfriesshire can be subdivided into several minor divisions each of which can be recognised in the Lake District of England, and to a large extent in Scandinavia and elsewhere, for the deposits in these areas are of deep-water character, and the sub-faunas of the subdivisions are similar in the different areas, but the Llandovery rocks of the Welsh borderland are shallow-water deposits, with a different fauna from that of the deep-water deposits of this age, and can only be stated to be contemporaneous with the Llandovery rocks elsewhere, because the deeper-water faunas of the underlying Bala rocks and overlying Wenlock rocks of the Welsh borders are respectively similar to those of the Bala and Wenlock rocks of the other regions. The shallow-water Llandoveries of the Welsh borders have only been separated into two divisions, upper and lower, and have not been split up into a number of subdivisions, each characterised by a sub-fauna, and each comparable with one of the subdivisions of Dumfriesshire, Lakeland and the other regions where the deep-water facies is found.

It will be seen that though the principle of William Smith that strata can be recognised by their included organisms has been extended since his time, and shown to apply to far smaller subdivisions of the strata than was suspected, the method of application is the same, and is more or less successful according to the amount of evidence which is accumulated in support of it.

[CHAPTER VI.]

METHODS OF CLASSIFICATION OF THE STRATA.

Earth-history like human history is the record of an unbroken chain of events. The agents which have produced geological phenomena have been in operation since the earth came into existence. Accordingly a perfect earth-history would be written as a continuous narrative, just as would a complete history of the human race. The historian of man finds it not only convenient but necessary to divide the epoch of which he is writing into periods of time, and so does the geologist, and in each case the division is necessarily more or less arbitrary. It is true that in writing the history or geology of a country, marked events stand out which form a convenient means of making divisions, but the marked events occurring in one country are not likely to take place simultaneously with those of another country, and consequently a classification of this character is only locally applicable.

The classification which is at present used by geologists was originally founded upon definite principles, and although our principles of classification have, as will appear, been somewhat altered subsequently, it has been found more convenient to modify the original classification than to adopt a new one in its entirety.

The largest divisions into which the strata of the geological column were separated were instituted because of the supposed extinction of faunas, and sudden or rapid replacement by other faunas of an entirely different character. This supposed rapid extinction and replacement is now known to have been only apparent and due to observation in restricted areas, and it is doubtful whether the three great divisions founded upon them are not rather mischievous than useful, as tending to disseminate wrong notions.

Moreover there is considerable diversity of opinion as to the terms to be adopted. The rocks were formerly divided into Primary, Secondary, and Tertiary. Owing chiefly to the use of the term Primary in another sense, the alternative titles Palæozoic, Mesozoic and Cainozoic (or Cænozoic) were suggested, and though the term Primary has been definitely abandoned in favour of Palæozoic, the words Secondary and Tertiary are used extensively as synonyms of Mesozoic and Cainozoic. It was soon perceived that the period of time included in the Palæozoic age was much longer than the combined periods of Secondary and Tertiary ages, and it was proposed to group the latter under one title Neozoic, whilst another suggestion was to split the Palæozoic age into an earlier Proterozoic and later Deuterozoic division. The interest excited by the advent of man is probably the cause of the attempt to establish a Quaternary division, which some hold to be a minor subdivision of the Tertiary, whilst others would separate it altogether. The terms Palæozoic, Mesozoic (or Secondary) and Cainozoic (or Tertiary) are now used so generally that any attempt to abolish them would be doomed to failure, but it must be remembered that they are purely arbitrary expressions, and the other terms which are not in general use, might be dropped with advantage.

The other subdivisions have been used somewhat loosely, and although an attempt has been made by the International Geological Congress to restrict certain names to subdivisions of varying degrees of value, it will probably be found best to allow of a certain elasticity in the use of terms, merely agreeing that they shall be used as nearly as possible with the signification assigned to them by the Congress. According to this classification, and apart from the division into Palæozoic, Mesozoic and Cainozoic, the strata of the geological column are grouped into Systems, which are subdivided into Series, and the series are further split up into Stages. A number of chronological terms were also suggested, of equivalent importance, thus the beds of a system would be deposited during a Period, those of a series during an Epoch, and those of a stage during an Age[15].

[15] The chronological words have been used so loosely that it is doubtful whether any good will come of trying to restrict their use, and Sir A. Geikie has pointed out the confusion which would arise if the term group be employed for the largest divisions (Palæozoic, &c.). The terms System, Series and Stage may well be employed in the senses suggested by the Congress.

The rocks of the Geological Column were originally divided into systems, owing to the occurrence of marked physical and palæontological breaks between the rocks of two adjacent systems, except in cases where a complete change occurred locally in the lithological characters of the rocks of two systems which were in juxtaposition: it is necessary to consider for awhile the nature of these breaks.

The most apparent physical break is where the rocks of one set of deposits rest unconformably upon the rocks of another one, indicating that the older set has been uplifted and to some extent eroded before the deposition of the strata of the newer set. This uplift and erosion signifies a change from oceanic to continental conditions in the area in which unconformity is found on a large scale, and accordingly a long period of time would elapse during which the continental surface would not receive deposits, so that the highest rocks of the underlying system would be considerably older than the lowest rocks of the one which succeeds it. Such a break may be obviously utilised for purposes of classification, but as some areas of the earth's surface must have been occupied by the waters of the ocean when other regions formed land, deposit in some areas must constantly have occurred simultaneously with denudation in others, and any classification founded upon the existence of unconformities will therefore have a purely local value.

Another, and less apparent physical break, which will also be locally applicable, may be due to the depression of an area to so great a depth that little or no deposit was formed upon the ocean floor there during the period of great depression; but as a break of this character is difficult to detect, the existence of unconformities has alone been practically utilised as a means of separating strata into systems owing to marked physical change, except in the cases where the lithological character of the strata completely changes, as between the Triassic and Jurassic rocks of England.

Fig. 2.

Palæontological breaks or breaks in the succession of organisms are in many cases, the result of physical breaks, and accordingly it is often possible to separate one set of strata from another by the existence of a combined physical and palæontological break between them. It is by no means necessary however that a physical break should be accompanied by a break in succession of the organisms, and the latter may also occur without the former. It was once maintained that a palæontological break was due to the complete and sudden extinction of a fauna and its entire replacement by a new one, but this is far from true, and accordingly the breaks differ in degree. Study of the strata shows that when the succession is not to any extent interrupted, the species do not appear simultaneously, but come in at different horizons, and they disappear in the same way. In [Figure 2] let A represent a set of conformable strata ab ... k, and suppose the vertical lines represent the ranges of the various species found in these strata. It will be seen that of 27 species whose range is shown only 2 pass through the whole thickness, so that the fauna of k is very different from the fauna of a, nevertheless the fauna of each stratum is closely similar to that of the underlying as well as to that of the overlying stratum, and though most of the species of k are different from those of a, this need not be the case with the genera. The fauna of the set of strata would contain every species whose range is represented, and for convenience' sake it might be said to be composed of sub-faunas, one of which occurs in each division ab ..., but the separation into sub-faunas would be artificial and merely for convenience' sake, for there is no break between any two sub-faunas. Turning now to B ([Fig. 2]), an attempt is made there to show what happens when there has been a physical break, resulting in the denudation of the strata ghik, and the deposition of another set op ... unconformably upon those deposits of the earlier set which have not been denuded. As the result of this we note, first, that the relics of organisms which existed in the area during the deposition of ghik, and were entombed in those strata, are destroyed by the processes of denudation, and a large number of organisms which lived long after the deposition of f, and disappeared not simultaneously but at different times during the period when denudation was in operation, seem to become extinct simultaneously at the top of f, though, if we could visit an area which was receiving sediment during the period of denudation, we should find them dying out in the rocks of that region at different levels. Furthermore, whilst denudation is going on, a longer or shorter period of time elapses, during which the upheaved area receives no deposit, and accordingly no organisms which lived during that period are preserved in the upheaved area. During this time a set of deposits lmn may have been laid down elsewhere, and besides the gradual disappearance of some of the organisms of ab ... k, there will have been a gradual appearance of new species. When the upheaved area is once more submerged, a new set of deposits op ... is accumulated in it, and the species which gradually appeared in adjoining regions will now migrate to it, and will seem to come in simultaneously at the bottom of o; accordingly we may find that there is not a single species which passes through from f to o and the palæontological break in this area is complete, though it is clear that it only implies local change, and that we may and indeed must find intermediate forms in other regions which fill up the gap.

As an illustration of the local character of a palæontological break we may cite the case of the Carboniferous and Permian systems of Britain. These rocks are separated from one another in our area by a physical and palæontological break, but in parts of India, and other places, we find a group of rocks now known as the Permo-Carboniferous rocks which contain a fauna intermediate in character between those of the Permian and Carboniferous systems, and a study of this fauna shows that the hiatus which exists locally is filled by the species contained in the Permo-Carboniferous rocks.

A palæontological break may, like a physical one, result from depression of the ocean-floor to so great a depth, that no organisms are preserved there during the period of great depression, and the remarks made concerning a depression of this nature when speaking of physical breaks will apply here also.

A local palæontological break may result owing to physical changes without the production of an unconformity in the area, or its submergence to a great depth, or if an unconformity is found, the break may be more marked owing to other physical changes. The difference between the Upper and Lower Carboniferous faunas is very marked in England, where the Upper Carboniferous beds were deposited under physical conditions different from those of the Lower Carboniferous, and accordingly the corals, crinoids and other open-water animals which flourished in Lower Carboniferous times are rare or altogether absent in the higher rocks. Where the change of conditions did not occur to a great extent as in parts of Spain and North America, the similarity between the two faunas is much more pronounced. Again, there is an unconformity between the Cretaceous and Eocene beds of England, which is accompanied by a palæontological break, but this break is more pronounced owing to difference of physical conditions, for we find abundance of gastropods in the lower Tertiary beds, and a rarity of these shells at the top of the chalk of England, though where physical conditions were favourable for the growth of gastropods, their shells are found in the higher strata of chalk age, and the palæontological break is not so apparent.

A palæontological break may occur also as the result of climatic change, though actual instances of this occurrence are much more difficult to detect owing to the general absence of any evidence of climatic change other than that supplied by the organisms themselves. Still, when no physical break exists, and the lithological characters of a group of sediments remain constant throughout, indicating the prevalence of similar physical conditions through the period of deposition of the sediments, if the fauna suddenly changes, there must have been cause for the change, and in the absence of any other cause which is likely to produce the change, alteration of the character of the climate may be suspected.

It follows from the observations which have been made, that although the rocks of the Geological Column may be divided into systems owing to the existence of physical and palæontological breaks, and this classification may be and has been applied generally, the line of demarcation between the rocks of two systems will be a purely conventional one, where there is no break, and, to avoid confusion, that line when once drawn should be adopted by everyone, unless good cause can be shown for its abandonment.

The subdivision of systems into series has been conducted in a manner generally similar to that in which large masses of strata have been grouped into systems, with the exception that actual breaks need not occur. The subdivision was usually made on account of marked differences in the lithological characters or fossil contents of the rocks of the various series, and frequently the lithological characters as well as the fossil contents are dissimilar; taking the rocks of the Silurian system of the typical Silurian area as an example, we find the Llandovery rocks largely arenaceous, the Wenlock rocks largely calcareo-argillaceous, and the Ludlow rocks argillaceo-arenaceous, whilst the fauna of the Wenlock rocks differs from that of the Llandovery rocks below and also from that of the Ludlow rocks above. The Llandovery, Wenlock and Ludlow therefore constitute three series of the Silurian system, but the lines of demarcation between these series are nevertheless conventional, for it has been suggested that a more natural division, as far as the British rocks are concerned, could be made by drawing a line, not as at present at the base of the Ludlow, but in the middle of that series as now defined, and uniting the Lower Ludlow beds with the Wenlock strata to form a single series.

The same process as that adopted in the case of series has been essentially pursued in subdividing these into stages. Each stage is usually different from that above and below in its lithological characters, fossil contents, or both, though the difference is usually less in degree than that which has been utilised for the demarcation of series. A stage is often, though not always, composed of deposits of one kind of sediment, and is furthermore frequently characterised by the possession of one or, it may be, two, three or more characteristic fossils. Thus the Wenlock series is divided in the typical area into Woolhope limestone, Wenlock shale, and Wenlock limestone, and the very names given to these stages indicate that each is largely composed of one kind of material. Their fossils are also to some extent different, though the difference between them is not likely to be of so marked a nature as that which exists between the faunas of separate series.

It will be seen that the system differs from the series and the series from the stage in degree rather than in kind, and no hard line can be drawn between divisions of different degrees of magnitude. It follows therefore that frequently a mass of sediment which one author will consider sufficiently important to constitute a system will be defined by another as a series, and similarly a series of one writer may become a stage of another.

The student of Stratigraphical Geology will find the expression 'fossil zone' occurring over and over again in geological literature, and as the term has been used somewhat vaguely by many writers and is apt to be misunderstood, it will be useful to notice the expression at some length.

Strictly speaking the term zone (a belt or girdle), when applied to distribution of fossils, should refer to the belt of strata through which a fossil or group of fossils ranges. Generally speaking, the expression is used in connexion with one fossil; thus we speak of the zone of C[oe]nograptus gracilis, the zone of Cidaris florigemma and the zone of Belemnites jaculum, though sometimes it is used with reference to more than one species, as the zone of Micrasters and the Olenellus zone. The term has been used not of a belt of strata but of a group of organisms[16], and zones defined as "assemblages of organic remains of which one abundant and characteristic form is chosen as an index," but if it be agreed that the term should be applied to strata and not to organisms this might be modified and the definition run:—'Zones are belts of strata, each of which is characterised by an assemblage of organic remains of which one abundant and characteristic form is chosen as an index.'

[16] See H. B. Woodward, "On Geological Zones," Proc. Geol. Assoc., vol. XII. Part 7, p. 295, and vol. XII. Part 8, p. 313.

It has been objected that the subdivision of strata into zones has been pushed too far, but this is merely because in the establishment of zones, workers find it easier to work out the successive zones where the strata are thin and presumably deposited with extreme slowness, than where they are much thicker and have been rapidly accumulated, and accordingly, as the subdivision of strata into zones is a recent event, geological literature contains many more references to thin zones than to those of great thickness. Where an abundant and characteristic form (which is chosen as an index) of an assemblage of organic remains ranges through a great thickness of deposit, there is no objection to speaking of the whole as a zone, and it cannot be divided. To give some idea of the variations in the thickness of strata through which these abundant and characteristic forms will range, I append a list of the zones of graptolites which have been established amongst the Silurian rocks of English Lakeland and the thickness of each (which in the case of the thicker deposits is naturally only approximate):—

It must not be supposed that each of the subdivisions in the above list is of equal importance, and has occupied approximately the same length of time for its formation, but a study of the strata proves by various kinds of evidence that the deposits in which the characteristic forms range through a small thickness of rock were on the whole deposited much more slowly than where the range is continuous through a great thickness of deposit.

The geological systems, as originally founded, were not very accurately separated from one another except locally. A comprehensive view of the characters of a system was taken, and accordingly the lines of demarcation between the same systems adopted by workers in different countries were by no means necessarily at or near the same geological horizon. As the result of more recent work, the establishment of fossil zones has been growing apace, and though many of these are seen to have only local significance, it is found as the result of experience that many of them are widely spread and occur in the same order in different localities; accordingly the remarks that have been made concerning the contemporaneity of strata apply to these zones also. After a study of this kind, a much more accurate comparison of strata is possible, and correlation of strata can be carried on to a much greater extent than when the systems were only roughly subdivided by reference to breaks, differences of lithological character, and general comparison of the faunas; accordingly whilst largely retaining the old names, the old method of classification is being partly superseded, and the included faunas alone are utilised to establish accurate correlations of the strata in various parts of the world. How far this correlation can be carried on remains to be seen, for the work though well advanced has by no means reached completion, and predictions as to the ultimate issue are useless without the experience by means of which only the work can be done. The difference between the methods of classification is well shown by an examination of the old and new divisions of the chalk. It was formerly roughly divided mainly by lithological characters into Chalk Marl, Lower Chalk without flints, Middle Chalk with few flints and Upper Chalk with many flints, but no two observers would probably agree as to where the deposit with few flints ceased and that with many commenced. The chalk is now separated on palæontological grounds into Cenomanian, Turonian, Senonian and Danian, and the superiority of the new method to the old is practically shown by the abandonment of the old classification except for very rough purposes, and the general acceptance of the new one. Many other examples might be given, but this one will suffice. In the case of some of the systems, the Carboniferous for example, the old classification founded upon lithological characters is largely extant, and it has been inferred therefore that no accurate subdivisions of the Carboniferous rocks can be made by reference to the faunas, owing to the rapidity with which the deposits were accumulated. It is by no means certain because the work has not been done that it cannot be done, and the experience obtained from a study of other strata in which subdivisions have been established by reference to the fauna would lead one to suppose that the non-establishment of subdivisions of the Carboniferous strata is due to our want of knowledge rather than to their non-existence.

The establishment of a classification on palæontological lines by no means does away with the necessity for local classifications on a lithological basis, and it has already been remarked that important results will follow from a comparison of the classifications of sediments founded on the two lines, results which have hitherto largely escaped our attention owing to the existence of a cumbrous classification attained by the application sometimes of one method, at other times of the alternative one.

[CHAPTER VII.]

SIMULATION OF STRUCTURES.

Although it is easy to give an account of the structures which are of importance to the student of the stratified rocks, actual observation of these structures is frequently attended with difficulties owing to the close imitation of one structure by another, and the past history of the science shows that erroneous conclusions have been reached again and again on account of the incorrect interpretation of structures.

Simulation of organisms has frequently been the cause of error. Inorganic substances take on the form of organisms with various degrees of closeness. The dendritic markings produced by efflorescences of oxide of manganese are familiar to all, and as the name implies, they simulate, to some extent, plant remains. More complex chemical changes have resulted in the production of rock-masses in which, not the outward form alone but, the internal structure of organisms is reproduced with more or less approach to fidelity, as the rocks which contain the supposed organisms described as Eozoon bohemicum, E. bavaricum, and, we may add, E. canadense. Mechanical changes in rocks subsequent to their formation may also cause the simulation of organisms by inorganic substances. Prof. Sollas has given reasons for considering the structure described as Oldhamia to be inorganic, and in the Carboniferous Sandstones of Little Haven, Pembrokeshire, every stage in the formation of tubular bodies resembling worm-tubes, as the result of complex folding of the strata, may be observed, whilst in other cases we find imitation of worm-tracks, as has been observed before.

It is when one inorganic structure is simulated by another that the stratigraphical geologist is most likely to be led astray, and accordingly it is worth noting some cases where this has occurred, as a warning, for it must not be supposed that the cases here noted are the only ones which are likely to occur.

It has been seen that the existence of bedding-planes is of prime importance to the geologist, and their detection is a matter of supreme moment. Under ordinary circumstances there is no great difficulty in distinguishing bedding-planes from other planes, but the importance of discovering them is often greatest when the difficulty is most pronounced. In rocks which have undergone no great amount of disturbance the planes of stratification are often marked by their regular parallelism, the separation of layers having different lithological characters by these planes, the arrangement of the longer axes of pebbles parallel to them, and the occurrence of fossils and also of rain-prints, ripple-marks and other structures produced during deposition, upon the surfaces of the strata, but none of these appearances is necessarily conclusive, especially in areas where the rocks have been subjected to orogenic movements. In regularly-jointed rocks, jointing may well be mistaken for bedding, and there is often great difficulty in discriminating between bedding and cleavage, especially when the exposures of rock are of small extent. Fossils may be dragged out along planes at an angle to the true bedding, pebbles will be compressed by cleavage so that their longer axes do not remain parallel to the bedding-planes but now lie parallel to the superinduced planes of cleavage, and a structure closely resembling 'ripple-mark' may be produced on planes other than those of original bedding, as the result of puckering. The alternation of rocks having different lithological characters may also be misleading. Intrusion of dykes along cleavage-planes, followed by decomposition of the dyke-rock causing it to resemble a sediment, and formation of mineral veins along the same planes, may give rise to an apparent succession of rocks of different lithological characters which could easily mislead an observer and cause him to mistake the cleavage-planes for planes of stratification. In rocks which have undergone great lateral pressure, the beds of different lithological character may be folded in such a way as to give very erroneous ideas of the true dip of the rock on a large scale. In [Fig. 3] the dip of the rocks in a small exposure might appear to be in the direction indicated by the unfeathered arrow, whilst the true dip of the strata as a whole, leaving the minor foldings out of account, is in the direction of the feathered arrow, at the inclination represented by the dotted line. The minor folds in a case like that represented may extend upwards for scores or even hundreds of feet, so that an error as to the direction and amount of dip may be made, even if the observer faces a cliff of considerable height.

Fig. 3.

False-bedding on a large scale may be a cause of error. In the Penrith Sandstone of Cumberland, the planes of deposition are often found dipping in one direction in a large quarry, but inspection of a wider area shows that this is not the true dip of the beds as a whole, but merely a local dip due to deposition on a slope, and any one attempting to calculate the total thickness of the beds by reference to these divisional planes might be seriously led astray. A reference to [Fig. 4] will explain this. The lines AA´, BB´ are the true bedding-planes cut across in the section, whilst the lines sloping to the right from xx are only lines of false-bedding on a large scale. An exaggerated estimate of the thickness of the deposit would be made by measuring the thickness of each of these stratula from A to A´ and adding these thicknesses together, whereas the actual thickness of the middle bed is the distance between A and B or A´ and B´.

Fig. 4.

When rocks have been affected by thrust-planes, the simulation of bedding may be carried out to a very full extent. Not only do the major thrust-planes resemble bedding-planes but the minor thrusts produce an appearance of divisional planes separating stratula or laminæ, and a close approximation to false-bedding is the result. To this structure Prof. Bonney has given the name 'pseudo-stromatism[17].' It may be developed in rocks of all kinds, whether possessing original planes of stratification or not, and as a result of its existence the geologist may be seriously misled, not merely by mistaking the direction of the strata, but also the nature of the rock, for we may find it produced in an unstratified glacial till, and in a massive igneous rock, and in each case the resulting rock will resemble a sedimentary deposit, and of course the observer may be confirmed in his erroneous opinion by the formation of apparent fossils, ripple-marks or other objects which he might expect to discover in sediments. As illustrative examples, reference may be made to a number of schistose rocks, in which the planes of discontinuity (which are in truth planes of foliation) have been taken for bedding-planes and the rocks claimed as sedimentary though they are in reality igneous; for instance many of the rocks of the Laurentian of Canada, of the Hebridean of the North West Highlands, and some of the ancient rocks of Anglesey.

[17] Bonney, T. G., Quart. Journ. Geol. Soc., vol. XLII. Proc. p. 65.

A foliated structure may, as is now well known, be simulated by a structure developed in a rock prior to its consolidation. The similarity of flow structure of some lavas to the foliated structure of a schist was long ago pointed out by Darwin and Scrope, and recent work has proved that parallel structure due to differential movement prior to consolidation may be developed in plutonic rocks, as shown by Lieut.-General McMahon in the Himalayan granites, and by Lawson amongst the plutonic rocks of the Rainy Lake Region; and as the foliated structure may be mistaken for original stratification the same may occur, and has occurred, when dealing with this flow-structure.

This is not the place to discuss the truth of the old theory of progressive metamorphism, in which it was maintained that a gradual passage could be traced between ordinary sediments and plutonic rocks, but it may be pointed out that much of the evidence which was relied upon to prove the theory was fallacious and due to the confusion of the parallel structure set up in plutonic rocks prior to, or subsequent to, consolidation, with original stratification. Recent study of metamorphic rocks has proved that the parallel structures developed in the rocks of an area which has undergone metamorphism may be produced by three distinct processes; they may be original planes of deposition, or formed in a solid rock subsequently to its formation, or in an igneous rock before its consolidation, and although it is sometimes possible to separate the structures produced by these processes, this is not always the case[18]. When a plutonic rock contains large phenocrysts and an eye-structure is developed in it, it may simulate a conglomerate, the rounded phenocrysts being taken for pebbles[19]. Still closer simulation of an epiclastic conglomerate may be produced in other ways and will be referred to immediately.

[18] It must be noticed that the rock in which parallel structure is produced before consolidation, if it undergoes no further change, though often associated with metamorphic rocks, is not itself metamorphic. The term gneiss applied to these rocks is a misnomer, unless the term be used even more vaguely than it is at present.

[19] See Lehmann, Untersuchungen über die Entstehung der Altkrystallinischen Schiefergesteine mit besonderer Bezugnahme auf das Sächsische Granulitgebirge, Plate XI. fig. 1.

We have already seen that the existence of unconformities has been utilised in the demarcation of large divisions of strata in various regions, and whether they be utilised in this manner or not, their detection is a matter of importance to the stratigraphical geologist, as they afford information concerning the occurrence of great physical changes during their production. These unconformities may also be closely simulated by structures produced in very different manner.

The occurrence of an unconformity implies the denudation of one set of beds before the deposition of another set upon them, and accordingly the denuded edges of the lower set will somewhere abut against the lower surface of the lowest deposit or deposits of the overlying set[20]. The existence of an unconformity may often be detected in section, but when the unconformity is upon a large scale this may not be possible, but it will be discovered by mapping the strata and will be apparent on a map owing to the deposits of the lower set of beds abutting against the others. This is well seen where the Permian rocks of Durham, Yorkshire, and Nottinghamshire rest upon different members of the underlying Carboniferous series, and will be noticed on any good geological map of England. But a similar effect may be caused by a fault, so that mere inspection of a map or even of the strata in the field and discovery of one set of beds ending off against another does not prove unconformity. When the fault is a normal one, with low hade (that is, having a fissure approaching the vertical position), the outcrop of the fault-fissure will approximate to a straight line if the fault has a straight course, even if the ground be very uneven, whereas, if the plane of unconformity has not been tilted to a high angle from its original horizontal position, it will crop out in a sinuous manner across uneven ground, in a way similar to that of beds which are nearly horizontal, so that though the general trend of the outcrop of the plane of unconformity may be fairly straight, its deviation from a straight line will be frequent and marked, as seen in the case of the Permian unconformity above referred to. But if the unconformable junction has been highly inclined its outcrop will resemble that of a normal fault, or if the fault be a thrust-plane with high hade, the outcrop of this will resemble that of an unconformable junction which has not been greatly tilted from its original horizontal position. In these cases we require more evidence before we can decide whether we are dealing with an unconformable junction or a faulted one.

[20] An unconformity may be simulated or an actual unconformity rendered apparently more important, as the result of underground solution of the underlying strata subsequently to the deposition of the upper set upon them, and any insoluble materials in the underlying strata may be left as an apparent pebble-bed at the base of the upper beds. This is seen at the junction of the Tertiary beds with the chalk near London. Subterranean water has dissolved the upper part of the chalk, increasing the unconformity which naturally exists between chalk and Tertiary beds, and the insoluble flint of the dissolved chalk is left as a layer of 'green-coated flint' at the base of the Tertiary deposits.

The lowest deposits of the newer set of strata lying above an unconformity have probably been laid down in water near the shore-line. As the unconformity, if large, implies elevation above the sea-level, the deposits first formed after this elevation has ceased, and depression commenced, will necessarily be littoral in character and possibly of beach-formation, and accordingly we often find that an unconformity is marked by the existence of an epiclastic conglomerate immediately above the plane of unconformity and, although this need not be continuous, it is usually found somewhere along the line of junction. The conglomeratic base of the Lowest Carboniferous strata when they repose upon the upturned edges of the Lower Palæozoic rocks of the dales of West Yorkshire is well known, and may be cited as an example. The association of conglomerates with unconformities is indeed so frequent that its possible occurrence will always be suspected and sought by the geologist. Unfortunately the result of recent observation is to show that along thrust-planes of which the outcrop simulates those of unconformable junctions, the difficulty of discrimination may be increased by the existence of cataclastic rocks which bear a close resemblance to epiclastic conglomerates, and which may be and have been styled conglomerates. It is well known that fragments of the adjoining rocks are knocked into a fault-fissure during the occurrence of the movements which cause the fault, to constitute a fault-breccia, and as the result of the abrasion of these fragments by chemical or mechanical agency, the angular fragments may become rounded and converted into rounded pebble-like bodies, when the rock is changed into a fault-conglomerate. [Fig. 5], from a photograph kindly supplied by Prof. W. W. Watts, shows a stage in the formation of a conglomerate of this nature from a fault-breccia; the fragment on the right remains angular, whilst those on the left have become much more rounded. The illustration is from a case described by Mr Lamplugh occurring in the slaty rocks of the Isle of Man, and Mr Lamplugh's paper[21] furnishes the reader with references to other examples of the production of similar rocks. No general rule can be laid down for distinguishing the true from the apparent unconformity, for the attendant phenomena will differ in each case; but if a fault-conglomerate should be suspected, the observer should try to ascertain whether fragments of a newer rock are imbedded in an older one, which sometimes occurs; he should note the existence of extensive slickensiding along the plane of junction and along planes of faulting, though the existence of these, implying as it does the occurrence of differential movement along the plane, does not prove that the movement was necessarily great, or that it did not take place along a plane of original unconformity; above all, he should look for structures such as mylonitic structure, pseudo-stromatism, development of new minerals, crushing out and stretching of fossils and fragments and, in short, for any structure which is familiar to him as a result of orogenic movements.

[21] Lamplugh, G. W., "On the Crush-Conglomerates of the Isle of Man," Quart. Journ. Geol. Soc., vol. LI. p. 563.

Fig. 5.

The effects of thrusting not only give rise to appearances suggestive of unconformity, but naturally also to a simulation of overlap. The thrust-planes are often parallel to original bedding-planes for some distance, but must cut across them sooner or later, producing lenticular masses which might be supposed to be due to the thinning out of beds as the result of cessation of deposition in a lateral direction.

Attention has already been directed to the deceptive appearance of great thickness of strata which is due to repetition of one stratum or set of strata by a series of thrust-planes, so that there is no actual inversion of any part of a bed. When masses of limestone are affected in this way, the thrust-planes may become sealed up, as the result of chemical change, and a compact irregular mass of limestone devoid of any definite divisional planes may be the consequence, and beds of grit sometimes exhibit the same feature to some extent.

Enough has been said to show that simulation of one structure by another has frequently occurred in rocks in so marked a degree as to render mistakes easy; and that these examples of 'mimicry' in the inorganic world are particularly frequent in rocks which have been subjected to great orogenic movements. The student will do well to acquaint himself with the macroscopic and microscopic structures which may be taken as characteristic of the rocks which have been thus affected, some of which can usually be detected with ease, and when he discovers them he may suspect that many phenomena which appear explicable in one way were in reality produced in a different one, for it is frequently very true of a region in which the rocks have been violently squeezed, stretched and broken that 'things are not what they seem.'

[CHAPTER VIII.]

GEOLOGICAL MAPS AND SECTIONS.

The writer does not propose to give an account of the intricacies of geological mapping, for their right consideration requires a separate treatise[22]; all he desires is to call attention to some of the uses of geological maps as a means of conveying information. A geological map may be looked upon as an attempt to express as far as possible in two dimensions phenomena which possess three dimensions; this can be done to some extent on the actual surface of the map, by conventional signs, still more fully, by supplementing the map with sections; but best of all by a geological model, which is cut across in various directions in order to show the underground structure as well as that of the surface.

[22] The student is recommended to consult in particular, Appendix I. "On Geological Surveying" in The Student's Manual of Geology, by J. B. Jukes (Third Edition, Edited by A. Geikie), p. 747, and Outlines of Field Geology, by Sir A. Geikie (Macmillan and Co.).

The ordinary geological map is one which shows the outcrop of the strata, subdivided according to age, as they would be seen upon the surface of the earth after stripping off the superficial accumulations, and it is to be feared that the term 'geological map' is associated in the minds of most students with a map of this character and of no other. Nevertheless, a great many most important observations other than those connected with the order of succession of the strata are capable of representation upon a geological map, and the possession of a large number of maps of any area upon the geology of which a person is engaged—each map to be used for recording observations of a particular kind—will save much writing in note-books and, what is of more importance, will allow him to compare observations which have been made at different times at a glance, instead of causing him to search through a series of note-books. Still, however well furnished with maps, the geologist will find a note-book essential[23].

[23] As a result of some experience, the writer recommends every student to acquire some skill in the use of the pencil, and if to such a degree that he can combine artistic effect with accuracy, so much the better. An acquaintance with photography is invaluable: often the possession of a camera would enable a section to be recorded, which is otherwise lost to science.

The earliest geological maps represented the variations in the surface soils, or at most the general lithological characters of the rocks which by their decay furnished the materials for the soils. We have seen that the first chronological map was due to William Smith, and most subsequent English geological maps have been based upon his map of the strata of England and Wales. The order of succession of the strata is represented in these maps to some extent by the use of arrows to indicate the direction of dip of the strata, though this is not an unerring guide where strata are reversed, and accordingly the addition of a legend at the side of the map may be looked upon as essential to the correct understanding of the map itself. The legend is usually in the form of a section of a column, the strata being arranged in right order, the oldest at the base and the newest at the summit, the colours by which the strata are indicated being similar to those placed upon the map. Other information besides the mere order of succession of the strata may appear in the legend; thus their relative and actual thicknesses can be indicated if the column is drawn to some definite scale, and a brief description of the lithological characters of the rocks may well be appended to the side of the column. On the actual maps it is customary to exhibit the outcrop of the junctions of all igneous rocks as well as of the sedimentary ones: the nature of the metamorphism which sedimentary rocks have undergone at the contact with igneous ones may be and often is indicated by suitable signs; the position of faults is shown, and often also that of metalliferous veins, the nature of the ore in the latter being further indicated in some suitable manner, as by giving the recognised symbol for the metal; and in many maps an attempt is made to show the variations in dip and strike of the cleavage-planes.

The Geological Survey of the United Kingdom publishes two sets of maps, one showing the 'solid geology' and the other the 'superficial geology.' It is easier to understand these terms than to define them, for in Britain there is a sharp line between the two everywhere except near Cromer. The maps showing the superficial geology represent gravels, glacial drifts and other incoherent accumulations of geologically recent origin, which to a greater or less extent mask the strata below which are usually composed of more or less solidified material. The maps showing the solid geology display the outcrops of these strata, though it is usual to insert alluvium upon these maps, as it is often impossible to trace the junction-lines of the strata below it. Attention has already been directed to the fact that these maps of solid geology, though chronological, that is, having the strata represented according to age, are founded largely upon lithological differences, rather than upon included organisms; and it has been stated that for theoretical purposes two sets of chronological maps, one founded upon lithological differences, the other upon difference of fossil organisms, would be extremely valuable.

Other phenomena are often best represented upon separate maps, for if all observations are crowded upon one map the result will be very confusing. Special glacial maps showing the contour of the country, with the portions between the contour lines coloured differently according to altitude, say the country between sea-level and 500 feet light green, that between 500 and 1000 dark green, that between 1000 and 1500 light brown and so on, exhibiting the direction of all observed glacial striae, the distribution of boulders so far as it is possible, and any other glacial phenomena which can be noted upon the map, will be valuable to the student of glaciation[24].

[24] For examples see Tiddeman, R. H., "Evidence for the Ice-Sheet in North Lancashire and the adjacent parts of Yorkshire and Westmorland," Quart. Journ. Geol. Soc., vol. XXVIII. pl. XXX., and Goodchild, J. G., "Glacial Phenomena of the Eden Valley" &c., Quart. Journ. Geol. Soc., vol. XXXI. pl. II.; and for a map of distribution of boulders, Ward, J. C., "Geology of the Northern Part of the English Lake District" (Mem. Geol. Survey), pl. IV.

Various structural features may be well displayed on separate maps. The trend of the axes of folds will be useful, and may be accompanied by other information of cognate character[25]; maps of the distribution of joint planes may be given in combination with those showing the folding of the strata if it be desired to exhibit the relationship between these; or with the physical features of the country, if the dependence of physical features upon joint structure be under consideration[26]. Much information concerning cleavage may be acquired from a map showing anticlinal and synclinal axes of cleavage[27], or the actual strike of the cleavage over different parts of a map may be represented, and its relationship to the geological structure of the district exhibited[28].

[25] See Bertrand, M., "Sur le Raccordement des Bassins Houillers du nord de la France et du sud de l'Angleterre," Annales des Mines, Jan. 1893, Plate 1.

[26] See Daubrée, A., Études Synthétiques de Géologie Expérimentale, 1^ère Partie, Plates III.-VI., for an example of the latter, which is also interesting as showing the utility of a map on transparent paper super-posed on another, when illustrating the connexion between two sets of structures.

[27] Ward, J. C., Geology of the Northern Part of the English Lake District, Plate IX.

[28] Harker, Alfred, "The Bala Volcanic Series of Caernarvonshire" (Sedgwick Essay for 1888), Fig. 5.

Maps exhibiting changes in physical geography appertain to the geologist as well as to the geographer. The position of ancient beaches, former lakes, representation of the changes in the courses of rivers and kindred phenomena may be shown upon maps, and will prove useful[29].

[29] For examples of maps of this kind, see Kjerulf, Th., Die Geologie des südlichen und mittleren Norwegen.

A perusal of the maps to which reference has been made above will give the student some notion of the extent to which maps may be utilised to represent geological structures, and may suggest other methods by which they may be utilised.

A geological section is usually drawn in order to exhibit the lie of the rocks, as it would be seen if a vertical cutting were made in that part of the earth's crust which is under consideration. The character of the section will depend upon circumstances. The Geological Survey of Great Britain issues two kinds of sections which are usually spoken of as vertical sections and horizontal sections, though each is in truth a vertical section; but whereas in the former the horizontal distance represented is small as compared with the thickness of the strata, in the latter the rocks of a considerable horizontal extent of country are exhibited in the section, and the section is not carried down to a great depth below the earth's surface. There is no essential difference between the two kinds of section, and often sections are drawn which cannot be definitely classed as belonging to either kind, but in extreme cases the vertical section is a representation of the order of succession as it would appear if the rocks were horizontal, no matter how disturbed they may be in reality; whereas the horizontal section represents the strata as they actually occur, with all the folds and faults by which they are affected. The accompanying figure ([Fig. 6]) represents a horizontal section on the left side of the figure with a vertical section of the same rocks on the right side.

Fig. 6.

Vertical sections are extremely useful when it is desirable to compare variations in the strata over wide extents of country: this can be done by drawing a series of columns of the strata, each showing in vertical section the lithological characters and thicknesses of the strata in one place, whilst the relationship between the strata of two different places may be indicated by joining the beds of the same age by dotted lines as shown in [Fig. 7][30].

Fig. 7.

[30] It is useful to adopt conventional symbols for the representation of strata of different lithological characters, and so far as possible to adhere to the same kind of symbol for any one kind of deposit. Those which are generally in use, are rough pictorial representations of the characters of the deposits, as shown in [Fig. 7]. The conglomerate is indicated by circular marks representing cross-sections of the pebbles, a breccia by triangular marks signifying that the fragments are angular and not rounded; a sandstone is indicated by dots to represent the grains of sand; a mud, clay or shale by continuous or broken horizontal lines, which reproduce the appearance of the planes of lamination so frequent in beds of this composition; a limestone is usually marked by the use of regular horizontal lines illustrating the pronounced bedding, with vertical lines at intervals to represent the regular jointing which occurs in so many limestones: the nature of the bedding may be further shown by drawing the lines comparatively far apart when the limestone is a thick-bedded one, nearer together when it is thin-bedded. Igneous rocks are represented by crosses or irregular V-shaped marks, illustrating the absence of stratification and presence of joints.

Volcanic ashes are sometimes represented by dots, at other times by signs somewhat similar to those which are used for true igneous rocks. Sedimentary rocks which are composed of more than one kind of material may be further shown by a combination of two symbols, thus the existence of a sandy clay may be shown by means of a combination of horizontal lines and dots, and so with other combinations. The practical geologist should become accustomed to the use of these symbols in his note-book; if used, they will save much writing.

These symbols are used in some of the later illustrations to this book.

The horizontal section is one which is in constant use by the practical geologist: the results of the first traverse of a district may be jotted down in his note-book in the form of a horizontal section (with accompanying notes), and the written memoir on the geology of any district composed largely of stratified rocks will almost certainly require illustration by means of these sections. Perhaps nothing more clearly marks the careful observer than the nature of the sections which he makes, and geological literature is too frequently marred by the publication of slovenly sections. A badly drawn section not only offends the eye, it may and frequently does convey inaccurate information.

Fig. 8.

In the above figure ([Fig. 8]) taken from Sir Henry de la Beche's "Sections and Views Illustrative of Geological Phænomena," Plate II., the lower drawing represents a section drawn to true scale, while that above shows one which is exaggerated. The student who saw this would infer that the uppermost beds on the left side of the upper section rested unconformably upon the dotted beds beneath, and once abutted against them in that portion of the figure where the beds have been removed by denudation in the deep valley, whereas an examination of the section drawn to true scale shows that the unconformity does not exist (although there is one at the base of the deposits marked by dots), and that there is room for the higher deposits to pass above those marked by dots at the place where the former have been removed by denudation. Whenever possible, horizontal sections should be drawn to true scale, the vertical heights being on the same scale as the horizontal distances. Sections which are so drawn represent the nature of the surface of the country as well as the relationship of the strata, and often illustrate in a marked degree the influence which the character of the strata has exerted upon the nature of the superficial features of a country. If it be impossible to draw a section in which the elevations and horizontal distances are represented upon a true scale, the former ought to be drawn on a scale which is a multiple of the latter; thus the vertical heights may be shown on 2, 3, or 4 or more times the scale chosen for the horizontal distances; when this is done, it will often be necessary to show the strata with an exaggerated dip, and accordingly the exaggerated section loses some of its value, though if vertical and horizontal scales bear some definite proportion it will still be more valuable than a rough diagram which is not drawn to any scale.

Section-drawing cannot be satisfactorily accomplished without some practice, and the student is strongly advised to acquire the art of drawing good sections; the writer can assert as the result of considerable experience in the conduct of examinations of all kinds, that slovenly sections are the rule in candidates' papers, and good sections very rarely appear. Study of the six-inch maps and horizontal sections (drawn on the same scale) of the Geological Survey of the United Kingdom will enable the student to familiarise himself with admirable sections, and it should be his aim to produce sections like these. He is recommended to take some of these six-inch maps which show contour-lines as well as the disposition of the strata, and to draw sections on the scale of six inches to the mile, vertical and horizontal, exhibiting the proper outline of the ground and the arrangement of the strata, and afterwards to compare them with the published sections. The sections should be drawn as far as possible at right angles to the general strike of the strata. Some datum-line is taken for the base of the section (say sea-level) and offsets drawn vertically from this where the section crosses a contour-line or recorded height. The height is marked on these offsets; thus if a recorded height of 2700 feet (just over half a mile) occurred on the line of section a height of somewhat over three inches is marked on the offset, and so with the other points where the section crosses contours or recorded heights. By joining these points on the offsets, giving the connecting lines curves similar to those which are likely to occur in nature, the general character of the surface of the ground is represented. The geology of the district is next shown. Wherever a dip is marked on the map, the direction and amount of dip is shown by a short line on the section, and where dips are not actually seen along the line of section, the dips which are nearest to that line on the map must be considered, and marked on the section. The lines of junction between the various deposits shown by different colours upon the map are inserted on the section as short lines, the inclination being judged by study of the nearest dips; faults and igneous rocks must be marked off, and any indication of the hade of the fault or the slope of the edges of the igneous rock which the map affords will be taken into account. The section will then appear somewhat as shown in the following figure:

Fig. 9.

and sufficient indication of the trend of the rocks will be obtained to shew that they form portions of curves which may then be filled in as shown in [Fig. 10] and the section will be complete.

Fig. 10.

It will be noticed that the small dyke of igneous rock on the right of the main dyke is joined to it lower down, though no indication of this is given along the line of section; but the requisite information for this and evidence of the existence of the small dyke proceeding from the left-hand side of the main one may be obtained by the study of the rocks in a valley on one side or other of the line of section.

After the student has become conversant with the nature of geological maps and sections, and has read Sir A. Geikie's Outlines of Field Geology, he should on no account omit to learn something of the art of making geological maps, by going into the field and attempting to produce a map, for the art of geological surveying does not come naturally to any one, and some acquaintance with the methods of surveying is a necessity to everyone who wishes to make original geological observations, though all cannot expect to afford the time and acquire the skill necessary for the production of maps vying with the detailed maps of the Government Survey. Before actually attempting to draw lines on a map on his own account, he will do well to tramp over a portion of a district with the published geological map in his hands, selecting a country which is not characterised by great intricacy of geological structure, and he can then attempt to represent the geology of another portion of the same district without consulting the published map. Of all the districts of Britain with which he is acquainted the writer believes that the basin of the river Ribble, in the neighbourhood of the town of Settle in the West Riding of Yorkshire, is best adapted for studying field geology in the way suggested above, for the main geological features are marked by extreme simplicity, and the exposures are good, whilst the presence of an important fault-system and of a great unconformity relieve the area from monotony. Anyone who stands on the summit of Ingleborough or Penyghent will grasp the main features of a portion of the district without any difficulty, for it lies beneath his feet like a geological model, and when the student has mastered and mapped in the leading features, he can find bits of country with geology of varying degrees of complexity amongst the Lower Palæozoic rocks of the valleys which run down to Ingleton, Clapham, Austwick and Settle.

The biologist is supplied with laboratories at home and abroad, where he may study his science under the best conditions. Would that some munificent person would found, in a district like that referred to above, a geological station where Cambridge students would have the means of acquiring a knowledge of field-geology under conditions more favourable than those presented by the flats around the sluggish Cam!

[CHAPTER IX.]

EVIDENCES OF CONDITIONS UNDER WHICH STRATA WERE FORMED.

The establishment of the order of succession of the strata, and the correlation of strata of different areas merely pave the way for the geologist. To write the history of the earth during various geological ages, he has to ascertain the physical and climatic conditions which prevailed during the successive geological periods, and to study the various problems connected with the life of each period. In the present chapter an attempt will be made to illustrate the methods which have been pursued in order to write to the fullest degree which is compatible with our present knowledge, the earth-history of various ages of the past. In making this attempt, the physical and climatic conditions may be first considered, and their consideration followed by that of the changes in the faunas, though it will frequently be necessary to refer to one set of conditions as illustrative of the other.

It will be assumed here that the great principle of geology, that the modern changes of the earth and its inhabitants are illustrative of past changes, is rigidly true. Reference will be made to this principle in a later chapter, but it is sufficient to state here that the study of the sediments which have been deposited from the commencement of Lower Palæozoic times to the times in which we now live bear the marks of having been formed under physical conditions, which, in the main, are similar in kind to those which prevail upon some part of the surface of the lithosphere at the present day.

One of the most important inferences of the stratigrapher relates to the existence of marine or terrestrial conditions over an area at any particular time, and we may, in the first place, consider the evidence which supplies us with a clue to this subject.

It has been previously stated that the ocean is essentially the theatre of deposition, the land that of destruction, and accordingly, the presence of deposit as a general rule indicates the evidence of marine conditions during the formation of those deposits, though this is not universally the case. Again, as denudation is practically confined to the land areas, and the shallow-waters at their margins, unconformity on a large scale gives evidence of the existence of terrestrial conditions in the area in which it is developed, during its production. Accordingly a mass of deposit separated from deposits above and below by marked unconformities shows the alternation of terrestrial conditions (during which the unconformity was produced) and marine conditions (during which the deposits were laid down). The deposits formed after an unconformity has been developed will naturally be of shallow-water character, as will also be those of the period immediately preceding the incoming of conditions which will cause the occurrence of another unconformity, and between these two shallow-water periods will occur a period when deeper-water conditions probably prevailed. We can therefore not only divide the history of any particular area into a series of chapters, of which every two successive ones will describe a continental period and a marine one, but each marine period may be divided into three phases—a shallow-water phase at the commencement, an intermediate deeper-water phase, and a shallow-water phase at the end. These phases are frequently complicated by the occurrence of a host of minor changes, but on eliminating these, the effects of the three great phases are shown by study of the nature of the strata, and their recognition does much to simplify the detailed study of the stratigraphical geology of various parts of the earth's surface.

In discriminating between terrestrial conditions and marine ones, the existence of unconformities is of great importance in marking terrestrial conditions and is often the only available evidence, for no accumulations or deposits formed on the land may be preserved to testify to the terrestrial conditions[31]. When terrestrial deposits and accumulations do occur, they are extremely important, and it is necessary to allude to the points wherein they differ from marine deposits.

[31] The term terrestrial is used above in opposition to marine, to include the conditions prevalent above sea-level. The term continental would be better if it did not exclude insular conditions. Accordingly deposits formed in rivers, and fresh-water and salt-water lakes are spoken of as terrestrial.

Apart from organic contents, the mechanically formed deposits of rivers and lakes resemble in general characters the shallow-water deposits of the ocean, though they are usually less widely distributed. It is the accumulations which have actually been formed as æolian rocks, or those which have been laid down as chemical precipitates in salt-lakes which, by study of lithological characters, furnish the most convincing evidence of their terrestrial origin.

Many æolian accumulations may be looked upon as soils, if the term soil be used in a special sense to refer to the accumulations which are produced as the result of the excess of disintegration over transportation in an area, whilst others are due to transport which has not been sufficiently effective to carry the material to the sea. When the weathered material accumulates above the weathered rock, it depends chiefly upon climate whether the disintegrated rock becomes mingled with much decayed organic matter forming humus. If this organic matter exists in quantity, the probability is that the accumulation is a terrestrial one, though this is by no means necessarily the case, for under exceptional circumstances a good deal of humus may be deposited in the sea, as beneath the mangrove-swamps which line the coasts of some regions, and to go further back, in the case of the Cromer Forest series of Pliocene times, or some coals, such as the Wigan Cannel Coal of the Carboniferous strata.

In addition to the work of water, which affects both land and sea-deposits, the land is especially characterised by the operations of wind and frost upon it, for these produce results which may frequently serve to differentiate a land-accumulation from a deposit laid down beneath sea-level. The effect of wind in rounding the grains of sand which are blown by it is well-known, and samples of the 'millet-seed' sands of desert regions are preserved in most museums. The greater rounding which characterises wind-borne as compared with water-borne sand grains is due, in great measure, to the greater friction between the grains when carried by the air than when swept along by the water. Under favourable circumstances water-worn grains may become rounded, especially when agitated by gentle currents sweeping over a shoal[32]; but a large mass of sand, in which most of the grains have undergone much rounding so as to give rise to 'millet-seed' sand, will nevertheless be probably formed by wind-action except where a marine deposit is formed of material largely derived from an earlier æolian one. The effect of frost is to split rocks into fragments which are more or less angular before they are subjected to water-action. The broken fragments are prone to collect on slopes as screes, and as any scree-material falling into the sea is likely to become rounded except under conditions which rarely prevail, the existence of much scree-material in a rock suggests its terrestrial origin. Glaciers gave rise to terrestrial moraines, which may occasionally be identified as land-accumulations by mere inspection of their physical characters, but all geologists are aware of the difficulties with which they are confronted when they attempt to discriminate between terrestrial and marine glacial deposits.

[32] Cf. Hunt, A. R., "The Evidence of the Skerries Shoal on the wearing of Fine Sands by Waves," Trans. Devon. Assoc., 1887, vol. XIX. p. 498.

The existence of much material amongst the stratified rocks which has been precipitated from a state of solution is an indication of the terrestrial origin of the rocks, which were laid down on the floors of the inland seas, separated more or less completely from the open ocean; for the waters of the ocean are capable of retaining in solution all of the material which is brought down to them, and accordingly precipitates of carbonate of lime, rock-salt, gypsum and other compounds formed from solution, are only formed on a large scale in inland lakes, though they may be formed to some extent when the water of a lagoon is only slightly connected with that of the open ocean, and the evaporation is great, for instance in the lagoons of coral reefs. Certain physical features often mark the deposits of chemical origin, cubical or hopper-crystals of rock-salt may be dissolved, and the hollow afterwards filled with mud, so that the rock surfaces are sometimes marked with pseudomorphs of mud after rock-salt. Sun-cracks and rain-prints impressed on the rock are not actual indications of terrestrial origin of the rocks on which they are found, for the shallow-water muds of an estuary may be deposited in the sea and yet exposed to the action of the air at low tide, but they mark very shallow-water deposits which have been exposed to the atmosphere immediately after their formation if not during the time they were formed, and they frequently occur amongst the deposits of inland lakes.