Geology

To those who support as well as to those who deny the doctrine of the permanence of oceanic basins and continental plateaux, India, the land of paradoxes, provides striking illustrations of these diametrically opposed opinions, a circumstance which suggests that the real truth lies somewhere between the extreme positions taken up by two classes of equally sincere naturalists. Those who think that the main orographical features originally developed on the solidified crust have never been seriously modified recognize in the main Peninsula an example of solid land which has neither been folded nor disturbed since the earliest geological times. Those who hesitate to recognize any limits to the mobility of the earth’s crust quote the Himalayas as an example of an area in which marine deposits containing Nummulites, and therefore no older than the London Clay, have been raised to an elevation of 20,000 feet within the Tertiary period.

Within the limits of the Indian Empire we have, therefore, two utterly dissimilar areas, unlike in geological history and equally unlike in the physical features which are the direct outcome of the geological past. In the Peninsula we have one of the few masses of land which have withstood all tendencies to earth-folding for as long as the palaeontological record stretches back. In the Himalayan region, on the other hand, the folding of the crust has produced, during the latest geological epoch, the grandest of our mountain ranges.

Except in marginal strips which show temporary and local trespasses of the sea on the coast, not a single marine fossil is found throughout the whole extent of peninsular India. The orographical features of this area are the outcome of the differential erosion of an old land surface, where the shallow open valleys, with rivers near their base-level of erosion, and the gently undulating plains are due to the toning down of the rocks by long exposure to the weather.

A very different state of things is disclosed in the land lying to the west, north, and east of the great Indo-Gangetic alluvial belt: in Sind, Baluchistan, the Himalayan belt, Assam, and Burma we have abundant evidence of repeated immersions beneath the ocean. In this area the directions of the mountain chains are determined by comparatively young rock-folds, while the region having been but lately elevated, its rivers are swift and torrential, cutting down their beds so rapidly that the valley sides are steep, with loosened material, always ready to slide off in destructive landslips.

In attempting to express these two distinct geological stories in European terminology we find that our simplest and most easily translated characters are preserved in the marine fossiliferous strata, while it is practically impossible to correlate directly the land and fresh-water formations which are so largely developed on the Peninsula with their equivalent stages in the European standard scale.

The reasons for this contrast are simple. Conditions of life are much more uniform, and facilities for migration much more perfect, in the ocean than on land. On land areas there is a greater variety of physical features, and a greater diversity, therefore, of climate and other conditions which affect the distribution of living beings. Consequently, when such areas are cut off from one another by impassable physical barriers, the intermingling of plants and animals is prevented, and evolution proceeds at independent rates in the separated areas, attaining corresponding stages at quite different times. As an example of the errors which would arise if we compared the fresh-water and land fossils of widely separate areas with one another, we have, in the existing indigenous mammalian fauna of the isolated Australian continent, a stage of evolution about equivalent to that which characterized Europe in Jurassic times. This want of correspondence during the same period of living forms in widely-separated land areas is one of great importance to the Indian geologist, who has had the point most strikingly brought home to him in his attempt to determine the age of the great coal-bearing system in India. The luxuriant growths of ferns, horsetails, cycads, and conifers which flourished in the great river valleys of the old Gondwana continent did not make their appearance in Europe until well on in the Mesozoic era, yet, from other evidence, we know that the lowest coal measures in India were being formed during Upper Palaeozoic times.

On the other hand, among marine fossils, especially such freely migrating forms as cephalopods, we have in general a tendency to wide geographical distribution with a very limited vertical range. The recognition, therefore, of the species of marine fossils in Indian formations permits a more precise correlation of Indian strata with those of Europe than is possible in the case of the fresh-water strata. The marine formations enable us to fix the chief landmarks in Indian geological history, and, having established these, we can consider the associated fresh-water and unfossiliferous rocks.

Before settling down to the description of the Indian rocks, there is one more stratigraphical principle of which the reader should be reminded: it is necessary to explain how it is that, in our attempts to express Indian stratigraphy in European terminology, we never attain full success.

In consequence of the way in which most areas have been alternately immersed below the sea to receive deposits of sediment, and raised to the denuding action of atmospheric agents, the sedimentary record in any country is marked by interruptions at irregular intervals in the scale. These interruptions or ‘breaks’ are not on the same horizons for all parts of a large area, for one part may be below water and receiving sediment when another is exposed and being cut into by the weather. Thus the dominant breaks in the Indian stratigraphical scale can only by an infinitesimal chance be strictly contemporaneous with those in Europe. In employing such breaks in the succession to define the upper and lower limits of stratigraphical units, we obtain series of strata which cannot correspond precisely to the units in the European scale; and for these series we are driven, therefore, to employ, in the first instance, local names which may cover parts of two or more European units. In Southern India, for instance, we have a series of strata answering generally to the Upper Cretaceous of Europe; but the four natural subdivisions in India do not correspond precisely with the European subdivisions. The lower part of the Utatur stage of Southern India corresponds to the Cenomanian of Europe, while its uppermost beds contain a Turonian fauna. The lower beds of the next succeeding stage, the Trichinopoly beds, are Turonian, while its higher beds are Senonian (see Bibliography, paper No. 1).¹ This circumstance, which increases the difficulties of correlation, is no more than one expects from the teaching of physical geography. The changes which occur during the processes of sedimentation are local: one area is receiving the fine detritus carried out by a large river, another is covered with sand brought down by a swifter stream, while a third is being buried in foraminiferal ooze, or supports a coral reef in the deeper and clearer water of the ocean. All have their characteristic forms of life, and these differ again, in both lithology and fossil-contents, from the beds produced contemporaneously in lakes and river valleys. The commencement and end of sediment in one area do not coincide with those in another; and, as a final difficulty in the way of precise correlation over large areas, many animals migrate from one region to another, their remains being found in one area and those of their descendants in another at a higher horizon in the stratigraphical scale. When the reader, therefore, finds the Indian geologist hesitating over the naming of his stratigraphical systems in India, he should remember that the hesitation is not due always to imperfect knowledge, but to consciousness of the fact that no unit in the stratigraphical formations of India corresponds exactly with the stages defined in the European scale. Nevertheless, our nearest approach to precision in correlation will be among the marine formations, and these, consequently, are used as reference horizons in our attempt to express the data of Indian geological history in terms familiar to European students.

The datum line in stratigraphy is the base of the Cambrian system, the so-called Olenellus zone, characterized in various parts of the world by remains of this genus, or its near relations, belonging to the extinct order of Crustacea known as trilobites. Below this line there are many thousand feet of strata without determinable fossil remains, and generally quite unfossiliferous; above it are piled the great fossil-bearing systems preserving the records of evolution among animals and plants through the Palaeozoic, Mesozoic, and Cainozoic eras to the present day.

Fortunately in India we have a trace of this datum line preserved in the Salt Range of the Punjab, where, although the Trilobites preserved are not exactly like the well-known Olenellus, there are forms which must have been close relations of it, and we can safely assume that these beds, referred to in more detail below as the Neobolus beds, are equivalent to the Lower Cambrian of the European scale.

To the ages preceding the date at which the Neobolus beds were formed we refer:—

(a) The great mass of crystalline schists which are exposed over half the Peninsula, forming the old floor on which the unaltered sediments were laid down;

(b) The great thicknesses of unfossiliferous strata known by such local names as Gwaliors, Cuddapahs, and Vindhyans.

The ages following the Lower Cambrian period have left their records in India in two groups:—

(c) During the Palaeozoic era deposits were formed in the extra-peninsular area with fossil remains referable to one or other of the well-known systems of Europe from the Cambrian to the Carboniferous. No records of this era have been preserved on the Peninsula.

(d) From Permo-Carboniferous times to the present day we have a double history: a record of life and events on the stable Peninsula, and a series of deposits formed in the adjoining ocean whose bed was afterwards upheaved to constitute the extra-peninsular parts of India.

Indian rocks thus fall naturally into four great groups: two below the Olenellus datum line without fossils, and two above the horizon at which the oldest recognizable fossils occur. The arrangement and chief divisions of these four groups are shown in the accompanying table (p. 55).

The oldest is a group of highly folded and foliated, immeasurably old, crystalline schists, gneisses, and plutonic rocks having the typical characters of the Archaean group of Europe and America, with which they can be correlated with sufficient safety to warrant the employment of the same group name.

The next is a great group of unfossiliferous strata lying with marked unconformity on the Archaean gneisses and schists, separated from the latter by a great physical ‘break,’ which is unmistakable throughout the Peninsula. This group is here distinguished as the Purāna (old), and it includes such isolated systems as the Cuddapahs, the Gwaliors, and the Vindhyans in the Peninsula—rocks which are sometimes locally folded, but never foliated, and often practically undisturbed. To what extent the Purāna group is represented in the unfossiliferous systems of the outer Himalayas it is impossible to say; for the only correlation data being lithological, the records have been mutilated by the folding of the Himalayan range. This group corresponds to much of what in America has been known as the Algonkian—rocks which lie between the base of the fossiliferous Cambrian and the eroded edges of the Archaean schists. Whether the younger stages of the Purāna group were formed before Cambrian times is not known; for on the Peninsula, where the Upper Vindhyans are the youngest strata in the Purāna group, they come into relation with no fossiliferous rocks older than the Permo-Carboniferous of Europe.

The nomenclature and grouping of the fossiliferous strata require explanation. In Europe the corresponding fossiliferous scale is divided very unequally into the Palaeozoic, Mesozoic, and Cainozoic groups, the inferior and superior limits being fixed at positions showing pronounced physical and palaeontological ‘breaks.’ As the evolution of animals and plants has been continuous since life first appeared on the earth, palaeontological breaks have obviously only a local significance, and the chief gaps in the European scale do not correspond to those which are noticeable in Indian strata. If the scale had been divided into groups according to Indian data, the lowest group would have its superior limit at an horizon corresponding approximately to the Permo-Carboniferous of Europe, for at about this stage there was a pronounced revolution in the physical features of the Indian area. The second stage would commence with the conglomerate which was formed in a period of great cold, giving rise to ice-sheets which have left their marks in a boulder-bed below the Productus limestone formation of the Punjab Salt Range, and in the Talcher series which forms the lowest stage of the great coal-bearing Gondwana system. There is no break at a higher stage so pronounced and widespread as this Upper Palaeozoic gap. Local changes in the physical geography of the extra-peninsular area are recorded in many localities, but there is a fairly continuous history of evolution from the time when the remarkable brachiopod Productus invaded the Indian seas to the coral banks and oyster beds of modern times. The Indian fossiliferous strata thus fall naturally into two great groups which are approximately equal in value. The lower and older group, which may be conveniently distinguished as the Dravidian, is about equivalent to that portion of the European Palaeozoic which includes the Cambrian, Ordovician, Silurian, Devonian, and Carboniferous systems; while the upper and younger, distinguished as the Aryan group, includes all strata from the Permo-Carboniferous system to the present day.

There is no trace of any fossiliferous strata on the Peninsula of India which can be referred to the Dravidian group; and such unfossiliferous strata as the Cuddapah and Vindhyan systems are, in the absence of fossil evidence, grouped with the Purāna strata, and regarded as more ancient than the Cambrian. In the extra-peninsular area we have members of the Dravidian group in the Salt Range, where there are strata of about the same age as the Cambrian of Europe; in the Central Himalayas, where Cambrian, Silurian, Devonian, and Carboniferous rocks are preserved; and in Burma, where representatives of the Silurian and Devonian have been definitely recognized by characteristic fossils.

The younger or Aryan group is represented on the Peninsula by the great fresh-water Gondwana system, which was followed by marginal encroachments of the sea in Upper Jurassic and Lower Cretaceous times; and by the deposition of subaerial and lacustrine formations in the Upper Cretaceous, with the great outflow of basic lava, covering the Deccan by volcanic eruptions, which continued until the commencement of the Tertiary period.

In the extra-peninsular area there is a great development of fossiliferous rocks, ranging through Permian, Mesozoic, and Tertiary times. Most of these are of marine origin, for the great central sea, Tethys, extended from Europe over most of this area throughout the Mesozoic era. It was only in the Tertiary period that the Himalayan region emerged and gradually drove back the ocean, until, in Miocene times, it was restricted to Baluchistan on one side and Burma on the other. It was driven back still farther in Pliocene times, when the mastodon and other mammals now extinct roamed through the jungles of Burma and the Himalayan foothills, leaving their bones to be buried in the rapidly accumulating river sands.

II. Pre-Cambrian History of India

In dealing with the unfossiliferous rocks formed in pre-Cambrian times, the geologist is in a predicament similar to that in which the historian finds himself when dealing with legendary periods for which no written records exist. The subject demands either a full description of the numerous difficulties which arise from want of precise data, or the briefest of statements consistent with the paucity of definite conclusions to which the geologist is able to point.

In the pre-Cambrian history of India there are two great and well-marked divisions, the records of the Archaean being immeasurably older than those of the Purāna era. Among the Archaean group are rocks which were presumably formed by the ordinary processes of mechanical sedimentation, yet after their formation there was time enough for them to be folded into great mountain ranges, and then cut down to a base-level of erosion, before the Purāna sediments were laid down on their upturned and denuded edges.

What this ‘break’ represents in the geological time-scale we have not the slightest idea; but it is probably no exaggeration to say that the lapse of time since Olenellus flourished in the Cambrian seas is small compared with that between the formation of the lava-flows which are now folded up in the Kolar gold-field, and the deposition of the basement sediments of the Cuddapah area. In gauging geological time the intervals of no record are as important as those of continuous sedimentation; and if there is one break in stratigraphical history that is universal, it is this which cuts off the Archaean crystalline schists from all subsequent rock groups. This Eparchaean interval in peninsular India is as well marked as it is in the Great Lakes region of North America, both areas having escaped folding movements since the deposition of the old unfossiliferous rocks. In these two areas, therefore, the circumstance of foliation alone is sufficient to mark off the Archaean from the Purāna group. But in areas which have been highly disturbed since Purāna times the old and the very old rocks have all been foliated, and it is now impossible to distinguish one from another. This summary permits only of brief reference to the chief characters of the two great groups, which are here taken in order of age.

More than half the area of the Peninsula is occupied by exposures of the old crystalline rocks, which must have obtained their present characters at great depths, being afterwards exposed to the surface by denudation of the superficial rocks. The crystalline rocks now exposed at the surface form only a fraction of the whole, for large areas are covered by the mantle of younger sediments and lava-flows; and it is in consequence of its position with regard to the ordinary sedimentary rocks that the old crystalline group is sometimes spoken of as the ‘fundamental complex.’

The fundamental complex in India agrees in essential respects with that of other countries, for instance America, where this group, on account of its great age—greater than that of any known fossiliferous rocks—was first named Archaean. Some of the rocks forming this complex are masses of deep-seated igneous origin; others presumably originated as sandstones, shales, limestones, lava-flows, and other forms of superficial deposits, which became metamorphosed by close-folding, by depression to great depths in the crust, and by further injection of igneous material.

In some cases the gneisses and schists present so markedly the chemical and mineralogical characters of igneous rocks that one does not hesitate to regard them as merely plutonic masses deformed by pressure and movement. These are often spoken of as ‘orthogneisses’ and ‘orthoschists.’ Others retain the essential chemical characters of well-known sediments, and differ from them merely in mineral character and texture due to metamorphism. These are known as ‘paragneisses’ and ‘paraschists.’ But there is a considerable fraction of the Archaean group in India, as elsewhere, whose precise origin is doubtful: some of these indeed appear to be the result of the intimate mingling of igneous injections and pre-existing rocks.

The Archaean group in India may thus be divided conveniently as follows:—

4. Dharwarian.

Eruptive Unconformity

III. Granites of Bihār (‘dome-gneiss’), granites of North Arcot in Southern India, anorthosites of Bengal, charnockite series of the Madras Presidency, norites of Coorg, pegmatites, &c.

2. Schistose and

3. Gneissose rocks.

Among the gneisses and schists it is practically impossible to distinguish a succession in time, even locally; and we are reduced to a system of classification which separates the distinctly eruptive types from those of doubtful origin, and from the masses in which there are signs of the commingling of eruptive and probably sedimentary material in a way far too intimate for differentiation. The following main divisions represent the chief types noticeable in the field:—

(1) Well-banded gneisses and schists, among which are alternations of bands of dissimilar lithological types, presenting the characters which one would expect from the metamorphism of a formation consisting of shales, sandstones, limestones, and lava-flows. The areas in which the mica-bearing pegmatites of Nellore and Hazāribāgh occur form good illustrations of this division of the Archaean group (2).

(2) More massive gneisses, such as might result from the more complete metamorphism of group (1), and the inclusion of more eruptive material which has absorbed and become intimately mixed with the pre-existing rocks.

(3) Definite eruptive types, deformed by earth pressures, with a foliation structure often in conformity with the associated gneisses and schists of groups (1) and (2). The rocks of this group generally show a family character over considerable areas. As examples we have the elaeolite-syenites and associated rocks of Coimbatore (3), the great granite masses of North Arcot and Salem, the norites of Coorg, the anorthosites of Bengal, the charnockite series which forms the larger hill masses in Southern India (4), and the so-called ‘dome-gneiss’ which rises as bosses in the midst of group (1) in the mica-belt of Bihār (2).

The rocks of group (3) are in general younger than those of the two preceding groups of crystalline rock, having attained their present position by eruptive transgression. As to groups (1) and (2), not only are we unable to determine their relative ages, but we are by no means certain that they are older than the Dhārwārs.

The rocks of the Dhārwār system are generally quartzites and fissile schists, chloritic, talcose, micaceous, and hornblendic. The quartzites often include much iron ore, and all grades are found between a quartzite with a few crystals of magnetite or hematite, and beds of almost pure micaceous iron-ore. Among the chloritic and talcose schists there are at times beds of pot-stone, and even the finer grades of steatite, indicating, probably, the derivation of some of the material from the alteration of the ferromagnesian, peridotic rocks of igneous origin. Similarly, too, among the hornblendic schists relics of true diabatic structures are often preserved, and many of the beds are doubtless the result of the alteration of basic lava-flows, while others suggest ash-beds. In fact, all through these beds there is abundant evidence of igneous action, which is no more than one would expect to have been the case in the earlier days of the earth’s history. Limestones are comparatively rare in the Dhārwār system, but they occur occasionally, and so, too, do true conglomerates which, notwithstanding the difficulty with which they are distinguished from autoclastic or crush conglomerates, may be taken as evidence of real water action in Dhārwār times (5).

The Dhārwārs have attracted a special interest on account of the valuable minerals they include: iron ores in great richness and purity in the Central Provinces and Bellary, copper ores disseminated at a particular horizon in Singbhūm, and gold in the quartz reefs of Kolār are examples well-known.

Upon the weathered surfaces of the highly folded Dhārwārs and the associated gneisses and schists of the Archaean group, enormous thicknesses of sediments were deposited in peninsular India. These rocks being devoid of fossils, isolated occurrences cannot with certainty be correlated, and consequently local names have been freely used to distinguish them. In Southern India we have the Cuddapah system, amounting to 20,000 feet in thickness with several unconformities, covered, also unconformably, by a thin series of strata distinguished as the Kurnools. Other examples of rocks of this class occur near Kalādgi in the Southern Marāthā country in the valley of the Bhima, near Pakhal in the Godāvari valley, in the valley of the Pengangā, in parts of the Mahānadi valley, and in Chotā Nāgpur. Farther west, in Central India, are series of old rocks distinguished as the Gwalior and Bijāwar series, and finally there is the great Vindhyan system, all being unfossiliferous.

A general survey of these old rocks reveals a lithological contrast between the lower beds, in which ferruginous jaspers and porcellanites are common, and the higher beds in which the rocks are shales, limestones, and sandstones, more nearly resembling materials formed by later common processes of sedimentation. The lower beds are also remarkable for the inclusion of basic lava-flows, which are conspicuous in the Gwalior series and in the Cheyar division of the Cuddapahs. The older division is thus represented by the typical Bijawars, the Gwaliors, and the lower half of the Cuddapahs; while the younger division includes the original Vindhyans, the Bhima series, the upper part of the Cuddapahs, and the Kurnools. Local difficulties must naturally occur in drawing a line between the older and the younger systems, and in the precise classification of isolated exposures of rocks belonging to the Purāṇa group; but such difficulties are the natural result of the conditions of sedimentation generally, deposition in one area, with a continuous record of beds, being contemporaneous with erosion and consequent unconformity in an adjoining district.

In Southern India a great development of the Purāṇa strata has been preserved in the Cuddapah and Kurnool Districts, forming a basin of sedimentary rocks that cover some 14,000 square miles. The strata within this basin have been divided into two very unequal groups, on account of a marked unconformity between the lower 20,000 feet, distinguished as the Cuddapahs, and the upper 1,200 feet, known as the Kurnool series.

Cuddapahs are divided into four series, separated from one another by unconformities; and it is highly probable that the lowest series in the Cuddapahs, in which we find the peculiar ferruginous jaspers and porcellanous beds, are the equivalents of the so-called Bijāwars and Gwaliors in Central India, while the upper series of the Cuddapahs and the associated Kurnools, in which normal sedimentary rocks occur, correspond generally with the Vindhyans.

The Vindhyan system is conspicuously displayed along the great escarpment of the Vindhyan range from which the rock system derives its name, stretching from Ganurgarh hill in Bhopal territory east-north-eastwards to the ancient fort of Rohtasgarh (24°37’ N.; 83°55’ E.). The rocks of this system are prominently sandstones, with subordinate bands of shale and limestone. Three of the massive sandstones, known as the Upper Bandair, Upper Rewah, and Upper Kaimur respectively, stand out conspicuously and determine the leading features of the main Vindhyan area. The system has been divided very unequally into a lower and an upper division; and the lower division includes large quantities of material that appears to have been ejected from volcanoes, producing beds of siliceous materials which, when very fine-grained, have a characteristically porcellanous aspect, and when in coarser fragments resemble some of the old European greywackes. The Vindhyan system, notwithstanding the apparent suitability of some of the shales and limestones for the preservation of fossil remains, has so far yielded no recognizable structures of organic origin, and its subdivisions are based on purely lithological characters (6). The system is remarkable for including rocks in which diamonds are found. These have been obtained in a band at the base of the Rewah stage in the State of Panna, the horizon being apparently about the same as the Banganapalle beds in the Kurnool series of Southern India. The most important product of the system is, however, its resources in lime and building stone, which are referred to in the chapter on Mines and Minerals (Vol. III, chap. iii).

Unfossiliferous sedimentary rocks occur in Upper Burma below the Lower Silurian of the Northern Shan States. Their age is unknown; but they had been folded and greatly denuded before the Lower Silurian beds were deposited on their upturned edges, and their general strike of folding coincides with the schists which underlie them in the Ruby Mines District. The schists in their turn pass into gneisses with which are found various forms of granulites and crystalline limestones; and this apparently gradual succession is similar to the order so frequently seen in the Himālayas, an association of beds among which it is almost impossible to make out a time-scale to distinguish upper from lower, or older from younger.

The unfossiliferous rocks which occur so prominently in the outer hills south of the crystalline, snow-covered peaks of the Himālayas, illustrate the difficulty of distinguishing between upper and lower in a group of highly folded, often inverted, unfossiliferous rocks, and the impossibility of correlating one area with another, which combine to throw doubt on any systematic grouping of these strata.

Attempts have been made to distribute the various occurrences of these rocks over the recognized stratigraphical scale. Slates and quartzites in the Dhaola Dhār region have been referred to as Silurian, while associated volcanic rocks have been classed as Carboniferous, Permian, and Triassic. The well-known Blaini boulder-bed of the Simla area has been regarded as the equivalent of the Tālcher and Salt Range boulder-beds of Permian age, while the strata apparently above and apparently below have been relegated to higher horizons on the one hand and to the Carboniferous and lower on the other. The weak points in the arguments for and against these correlations are freely admitted by their authors, and they must be regarded, consequently, as mere attempts to draw the simplest natural inference from a few significant features in lithology and succession. There is one important feature, however, which has not received its full share of recognition: the fact that only unfossiliferous rocks occur south of the snowy range and crystalline axis, while fossiliferous beds varying from Cambrian to Tertiary extend along the whole length of the range on the Tibetan side, suggests an original difference between the two areas. Those on the south have a significant proximity to the peninsular unfossiliferous Purāṇa group, whose age is regarded as pre-Cambrian, mostly or wholly. That such old rocks extended far beyond their present peninsular limits is highly probable, and they may be represented in the enormous thicknesses of pre-Cambrian Vaikritas of the Central Himālayan and Tibetan zone of the range. Had the unfossiliferous rocks of the outer Himālayas been formed during Palaeozoic and Mesozoic times, it is hardly likely that they would have uniformly escaped the inclusion of fossils which are so abundant in the rocks of corresponding age in the Salt Range, on the north-western frontier, in the Central Himālayas, and in Burma. It seems far more natural to suppose that a northern extension of the Purāṇa group has become involved in the Himālayan folding, while the beds of the Peninsula have remained undisturbed: the relations of the younger Gondwānas of the Darjeeling area to those of the Peninsula are precisely parallel to this. The suggestion is thus offered that the great masses of unfossiliferous rocks, well-known in hill stations like Simla and Naini Tāl, the Attock slates farther west and the Buxa series farther east, should be referred to the Purāṇa group instead of being correlated with Palaeozoic systems.

II. Cambrian and Post-Cambrian History of India

A. THE DRAVIDIAN ERA

The oldest fossiliferous strata we know in India are found in the Salt Range of the Punjab, where beds are exposed with fossils whose nearest relatives occur in the lower division of the Cambrian, the oldest fossiliferous system of Europe.

The undoubtedly Cambrian beds are found lying on a formation of peculiar marl with beds of rock-salt and gypsum, possibly of Tertiary age, similar to the salt deposits of the Kohāt area which will be referred to in describing the Tertiary system of India. The occurrence of such old beds, lying over masses of much younger material, is due possibly to the former having been thrust bodily over the salt-marl formation during the process of earth-folding.

The Cambrian strata of the Salt Range may be conveniently divided into the following series, which overlie one another in conformable sequence:—

4. Bhāganwalla, or Salt-pseudomorph series.

5. Jutāna, or Magnesian sandstone series.

6. Khussak, or Neobolus beds.

7. Khewra, or Purple sandstone series.

The purple sandstones are quite devoid of recognizable fossils, but they frequently reveal the ripple-marks originally produced on the sandy shore on which they were formed. They graduate into the overlying dark-coloured shales and cream-coloured dolomitic rocks of the Neobolus series, which, by the fossils they contain, indicate the prevalence of a deep sea. The Neobolus series nowhere exceeds 150 feet in thickness, but where best developed it is capable of subdivision into five zones, in the uppermost of which the principal fossils have been found. Besides the brachiopod genus Neobolus, which has given its name to the series, the most interesting form is a new genus of the peculiar Palaeozoic crustacean sub-class of trilobites recently distinguished by the name Redlichia. This form was until lately mistaken for the well-known Olenellus, a trilobite characteristic of the lowermost Cambrian (Georgian) strata of Europe and America. The nearest relatives of Redlichia, as well as of the associated animal remains, occur, however, in the Lower Cambrian elsewhere, and the Neobolus beds may consequently be regarded as the homotaxial equivalents of these old rocks (7).

The next series above is a sandy dolomite with intercalated argillaceous layers, and among the few and imperfect fossils they contain is one resembling the peculiar mollusc Stenotheca, Salter, found in the Lower Cambrian rocks of America. These beds gradually pass into the next higher, and uppermost, of the Cambrian strata preserved in the Salt Range, which are remarkable for the preservation of sandy models of cubic crystals, evidently pseudomorphs of salt crystals left by the evaporation of salt water before the deposition of succeeding layers of sediment. Here closes the first chapter in the geological history of the Salt Range; and between these rocks of Cambrian age and the next higher, which are not older than Upper Carboniferous, we have no record of sedimentation in this area, though the interval was wide enough for the deposition of three great systems of strata elsewhere, and for the evolution of several new classes of plants and animals.

For an imperfect record of Indian geological history during the great interval between the Cambrian and the Permo-Carboniferous rocks of the Salt Range we must turn to other areas, the best known being a zone of folded strata in the Central Himālayas, near the frontier of Tibet in the border tracts of Spiti and Kumaun.

The earlier records in this zone have been obliterated by metamorphism. The Vaikritas and Haimantas rock distinguished as the Vaikrita system pass gradually up (Upper Cambrian). The fossil remains have escaped destruction. These less-altered old strata are, on account of the snow-clad mountains which they form, named the Haimanta (snowy) system. The base of this system is fixed at an horizon of conglomerates which is exposed only in the Kumaun end of the zone. The conglomerate series is overlaid by greenish phyllites, slates, quartzites, and grits with obscure fossils. In places these are altered by granitic intrusions. The uppermost division of the Haimantas consists of alternating beds of quartzite and shale, with narrow bands of dolomitic limestone, which become more important at the summit. The shaly beds include several trilobites of the family Olenidae, indicating an Upper Cambrian (Potsdamian) age. The dolomitic limestones are covered by red slates and reddish-brown dolomites, over which a conglomerate marks the unconformity separating the Cambrian from the succeeding Ordovician strata.

The Ordovician is represented in the Kumaun area by a coral limestone, while in Spiti this stage is probably not preserved, the lowest of the Silurian beds being red grits and quartzites, with overlying shales and limestones, which contain fossils, like the coral Halysites catenularia, Lmck., indicating an Upper Silurian (Gothlandian) age. The Gothlandian beds are overlaid by a grey limestone, which becomes reddish-brown on weathering, but has not yielded fossils sufficiently well-preserved to determine its exact age, though from its position it is probably Devonian. Then follows in conformable succession a formation which first develops into a red and then shades off into a white quartzite, named the Muth quartzite.

Over the Muth quartzite, where the complete sequence is displayed, as in the Lipak valley of Eastern Spiti, we find, in order, grey limestones with numerous brachiopods of Upper Carboniferous age; alternating beds of limestone, shale, and quartzite, with a thin band of conglomerate; and a thick covering mass of limestone with flaggy sandstones and shales, containing brachiopods and fragments of a trilobite belonging to the genus Phillipsia, Portl.; finally, shales distinguished as the Fenestella shales, with numerous specimens of this and other Bryozoa resembling some from the Zewan beds of Kashmir, and thus probably of uppermost Carboniferous or Permo-Carboniferous age (8).

With these beds we approach the close of the second chapter in the geological history of Northern India; for at about this horizon, corresponding to the Upper Carboniferous of England, there is an important break in the deposition, and new conditions are introduced by a widespread conglomerate which forms the base of the division of marine sediments distinguished as the Aryan group.

The only other areas within British India where fossiliferous rocks of older Palaeozoic age have been found are in Chitrāl to the west and in Burma far away to the east. In Chitrāl Devonian fossils have been found in a limestone exposed on the right bank of the Chitrāl river, immediately opposite Reshun, where it seems to overlie a red sandstone and a still lower conglomerate. The best preserved of these fossils include corals and brachiopods, of which the corals show affinities with forms found in the Upper Silurian of England, while the brachiopods include forms like Orthis striatula, Schloth., Spirifer extensus, Sow., S. disjunctus, Sow., Athyris concentrica, v. Buch., and Atrypa aspera, Schloth., which have Devonian affinities, some of them resembling Devonian brachiopods found in Southern China (9).

A similar series of rocks, amounting to more than 2,000 feet in thickness in Hazara, has, on account of its relations to the Triassic rocks in that area, been generally referred to as the infra-Trias. These beds are, however, almost certainly identical with the rocks just described in Chitrāl: the succession in both areas consists of a coarse conglomerate at the base, resting unconformably on a great slate series and overlaid by red or purple sandstones and shales, with limestone above as the principal member of the series. Fossils have not, however, been found in the Hazara rocks.

Rock formations of Lower Palaeozoic age cover considerable areas in the Shan States and in Karenni. They are, however, so thoroughly covered with a coat of decomposition products, and often so concealed by the thick undergrowth of jungle, that precise information as to their structures and palaeontology is not easily acquired. But some calcareous shales and limestones have yielded Echinosphaerites, Wahl., one of the peculiar stalked cystoids so characteristic of the Ordovician system in Europe and America. In higher beds Orthoceras, trilobites, and graptolites have been found, and, with the last-named, a form of Tentaculites closely resembling the Ordovician form T. elegans, Barr.

Devonian beds have also been recognized containing, with other fossils, the unmistakable and characteristic coral, Calceola sandalina, Lam. The predominating rock in this system is a limestone, distinguished as the Maymyo limestone, which extends, almost without interruption, from the neighbourhood of Maymyo to the Salween river in the Northern Shan States.

B. THE ARYAN ERA

The changes in physical geography which occurred towards the end of the Carboniferous period are marked by a widespread conglomerate in Spiti and the Bhot Mahāls of Kumaun. This conglomerate forms the basement bed of a great series of strata which were laid down successively, without a sign of interruption or break, throughout a period corresponding to the whole of the Permian period and the succeeding Mesozoic era of Europe. The beds thus formed have been preserved in a zone lying to the north of the crystalline snowy peaks of the Central Himālayas near the boundary between Tibet and North-western India. Recent observations in Tibetan territory to the north of Sikkim show the eastward continuation of the younger fossiliferous strata of this series, which rest abruptly on the northern flanks of the crystalline axis, and probably cover the older beds which happen to be exposed in Spiti and Kumaun.

The form of physical revolution which gave rise to this great series of strata appears to have been an eastern trespass of the Eurasian ocean, whose southern shore apparently coincided with the present line of snowy peaks, while an arm stretched into the Punjab as far as the Salt Range. Thus commenced what may conveniently be distinguished as the Aryan era in Indian geological history. The great central ocean above referred to, known to geologists as Tethys, flowed over a belt stretching across Central Asia, leaving deposits in which the fossil contents of places so widely separated as Burma, China, the Central Himālayas, Siberia, and Europe show the marked affinities due to free migration in the ocean.

The very complete scale of conformable strata preserved in the Central Himālayas attains a thickness of some 7,000 feet from the basement conglomerate below the Permian calcareous sandstone to the top of the Chikkim shales of Cretaceous age. The Permian ‘Productus shales’ pass up gradually into beds which introduce Triassic conditions through successive zones characterized by the genera of ammonites, Otoceras, Ophiceras, and Meekoceras. These are followed by beds with fossils so unmistakably characteristic of the Trias that the different stages recognized in Europe can be approximately defined and divided into zones. The Triassic rocks are followed by about 2,800 feet of Jurassic strata, among which occur the well-known Spiti shales, now known as far west as Hazara and as far east as Sikkim. The Spiti shales are covered by the Giumal sandstone and the unfossiliferous Chikkim series, which resembles the flysch deposits so frequently found in Cretaceous and Lower Tertiary formations (9).

One of the most interesting features in connexion with the geology of the Central Himālayas is the occurrence in the Kumaun section of numerous blocks of older rocks, mainly limestones, lying on the Spiti shales and the Giumal series without apparently any regularity of distribution. They are weathered into picturesque crags, rising in abrupt pinnacles with precipitous walls, and on account of their composition (often brightly-coloured, semi-crystalline limestones) they stand in striking contrast to the more sombre shales and sandstones forming the undulating country around. At first sight they recall the Kippen and lambeaux de recouvrement, or blocs exotiques, of the Alpine regions in Europe whose origin has been the subject of much controversy. In Europe these exotic blocks have been supposed to obtain their abnormal positions by being shorn off from highly crushed anticlinal folds, or by the removal of all but these fragments of enormous recumbent folds, or by faulting (10). Whether any or all of the theories employed to account for the Alpine exotic blocks are satisfactory appears to be far from settled; but the Himalayan examples seem to admit of a very simple explanation. They are always associated with igneous rocks, which are often amygdaloidal and otherwise generally agree in character with igneous rocks of surface (volcanic) origin. In these lava-flows the exotic blocks are buried by the hundred and are of various sizes. They are not only older than the Jurassic and Cretaceous strata they rest on, but belong to a facies either palaeontologically or lithologically foreign to the rocks of the same age in the Central Himālayas. They have thus come from a distance; and there being no signs of volcanic necks in the neighbourhood, the basic lava-flows in which they are embedded must have come from afar, like the flows characteristic of the so-called fissure eruptions, carrying with them their load of stratified rocks torn off from various horizons of the formations occurring in the area of eruption. It is assumed, for several reasons which need not be discussed here, that these exotic blocks belong to formations occurring farther north in Tibet, but political reasons have hitherto prevented the exploration of that country, and as a consequence this theory of the origin of the exotic blocks must remain for a time unverified: to establish it firmly it would be necessary to trace the blocks to their source, and to show that the fragments of basic lava in which they are embedded are outliers of lava-flows farther north. As a first simple inference, however, from the facts so far available, the explanation just offered is, among the theories which have been considered, the one that offers least difficulty.

The Central Himālayan area is exceptional in possessing such a complete and unbroken succession of strata. As a general rule we find the Permian and Triassic rocks linked together, as in the Salt Range, on the North-western frontier, and in Kashmir, while in other areas the Upper Jurassic and Cretaceous rocks are associated with one another. It will therefore be convenient to notice the other occurrences in approximately natural sub-groups, beginning with the Permo-Trias of the Salt Range, where we have the nearest approach to the remarkably complete succession of the Central Himālayas.

The first chapter in the geological history of the Salt Range closed, as already explained, in the Cambrian period, between which and the uppermost Carboniferous no traces of sedimentation have been preserved in that area (11). The second chapter opens with the remarkable boulder-bed which rests unconformably on the Cambrian strata. In its essential characters, as well as in its stratigraphical position, this boulder-bed corresponds to, and is probably contemporaneous with, the Talcher boulder-bed at the base of the Gondwana system in the Peninsula. On it, therefore, we have two great systems of strata developed: that in the Peninsula was formed in the river-valleys of the old Gondwana continent, while the beds in the Salt Range, now to be described, represent the deposits which were laid down at about the same time in the adjoining ocean.

The boulder-bed of the Salt Range, like that of the Tālcher series, has the peculiarity of being composed of a fine-grained silty matrix with included boulders of varying size up to several cubic feet. Many of these are faceted and striated in a manner which agrees with the general characters of the formation in pointing to a glacial origin; and several of them prove to be identical with the peculiar felsitic lavas we find in the Mallāni series, on the western flanks of the Arāvalli range, about 750 miles to the south. The glacial origin of the beds is shown also by the exposure of typically ice-scratched surfaces on the rocks they rest upon.

The fossils found in the beds immediately associated with the boulder-bed show a Carboniferous facies, having noticeable affinities with forms occurring in the Upper Carboniferous marine beds of Australia, to which area the same great ocean apparently extended. Among the identical species in these two widely separated areas are Eurydesma globosum, E. ellipticum, E. cordatum, Conularia laevigata, C. tenuistriata, Pleurotomaria nuda, and Martinopsis darwini.

The beds overlying the boulder-bed introduce a change in the physical geography which commenced with the retreat of the Australian ocean, the development of an area of internal drainage unfavourable to life and to the preservation of organic remains resulting in the deposition of about 400 feet of red and purple sandstones and shales with gypseous bands. These beds are known as the Speckled Sandstone series.

Further developments in the local physical geography resulted in the gradual encroachment of a western ocean which opened up marine communication with the European area, and gave rise to the formation of a system of fossiliferous rocks, mainly limestones, which, on account of the abundance of a genus of brachiopods, is known as the Productus Limestone. The lower beds are sandy and coaly in the east, but become more calcareous towards the west, that is, as we pass out to the deeper sea; and as we ascend in the series we find the encroachment of the sea more completely established, with the production of purer limestones. This lowest division of the Productus limestones is distinguished as the Amb series; and some of its fossils show great affinities with those of the Gshelian stage of Russia and the Fusulina limestones of the Carnian Alps, strata which are regarded as Upper Carboniferous in age. Many of the species are found also in the Artinskian (Permo-Carboniferous) and in even younger stages in Europe; but the fact that most of the fossils belong to the class of brachiopods, whose species, on account of their stationary habits, have a wide vertical range, prevents the more precise correlation which would have been possible if the animals had belonged to migratory forms like the cephalopods, which we shall find to predominate in the Triassic beds overlying the Productus limestones. Among the brachiopods characteristic of the Amb beds, or Lower Productus series, are Productus lineatus, Waag., P. cora, d’Orb., P. spiralis, Waag., P. semireticulatus, Schl., Athyris roysii, Leo., Spirifer marcouii, Waag., S. alatus, Schl., Martinia glabra, Mart., Reticularia lineata, Mart., Orthis pecosi, Marcou, and Richtofenia sinensis, Waag., the last-named being one of the peculiar aberrant forms of brachiopods which also characterize beds of this age in Southern China, an area to which this great Eurasian ocean extended in Permo-Carboniferous and Permian times.

The middle division of the Productus limestones forms a large and conspicuous fraction of the whole formation, being characterized by the prevalence in it of more exclusively Permian fossils. Its younger age is also marked by the appearance of forms, like the lamellibranch Oxytoma, Meek, not known elsewhere below the Trias, and of Nautilus peregrinus, which has a near relative in the Jurassic strata of Europe. There are three well-marked palaeontological zones in this middle division. The lowest of these is characterized by the survival of the foraminiferal genus Fusulina, Fisch., which, with its relative Schwagerina, Möll., attained an enormous development in Carboniferous and Permian limestones elsewhere. The central zone is distinguished by including the peculiar brachiopod Lyttonia nobilis, Waag. The occurrence of the cephalopod Xenodiscus (Xenaspis) carbonarius in the uppermost zone of this division of the Productus limestone formation indicates a greater affinity of the series with the Triassic beds than would be supposed from the fossils mentioned above, and it is possible that a re-examination of this interesting series of beds will place it on a level with the Zechstein of Europe.

The upper division of the Productus limestones shows still more the approach of the conditions characteristic of Mesozoic times by the appearance of several forms of true ammonites with complicated sutures. Prominent among these species are Cyclolobus Oldhami, Waag., Medlicottia Wynnei, Waag., and Euphemus indicus, Waag., which predominate in successive zones from below upwards in this order.

As in the Central Himālayas, so in the Salt Range, there is a perfectly gradual passage from strata which are unquestionably Permian up to beds which contain an essentially Triassic fauna. To draw a line, therefore, exactly corresponding to the base of the Trias in Europe is as difficult as it is unimportant: the main point to establish is the fact that the perfectly conformable passage is accompanied by the gradual replacement of typical Palaeozoic forms by characteristic Mesozoic fossils. Within a few feet of the beds containing the highest remains of the genus Bellerophon, we meet with limestone containing traces of ammonites; and from this horizon up for over 200 feet, in a typical section near the village of Chideru (32°33’ N.; 71°50’ E.) in the western part of the Salt Range, we find beds in which the characteristic Triassic ammonite Ceratites, Haan, occurs so abundantly that its name has been employed to distinguish the whole series (12).

According to the predominating rock, this series can be divided into four lithological stages and five palaeontological zones, as follows:—

Of these, zones 1 and 2 are characterized by the frequent occurrence of the genus Meekoceras, Hyatt, and zones 3, 4, and 5 by the abundance of fossils belonging to the genus Hedenstroemia, Waag. Palaeontologically, therefore, the beds are capable of division into two stages which correspond approximately to the beds in the Trias of the Central Himālayas, where Meekoceras is found associated with Ophiceras and Otoceras in a series of limestones and shales, which are in turn covered by further beds in which Ceratites is well represented with, as in the Salt Range, Flemingites flamingianus, Waag. With the Ceratite beds, which are approximately equivalent to the Lower Trias of Europe, the second chapter of the Salt Range closes; and all younger records, representing the Muschelkalk, the Upper Trias, and the two lower divisions of the Jurassic system, have, if they ever existed, been completely removed.

Exposures of beds belonging to different parts of the Permo-Trias so well displayed in the Salt Range occur at different points farther north. In the Bannu District, for instance, a boulder-bed with scratched and faceted boulders, like the well-known occurrence of the Salt Range, occurs covered with limestone containing fossils of Permian age. In this area the Triassic Ceratite limestones also follow in conformable succession, while the Trias is likewise represented in Hazara. Permian limestones with fossils like those of the Productus limestone series have been found in Chūra and the Bāzār valley, but the Palaeozoic rocks are apparently covered up by younger formations south of the Safed Koh (13).

The Permo-Triassic series of Kashmir have a special interest, on account of the occurrence of remains of the Lower Gondwana plant Gangamopteris associated with those of typical Permo-Carboniferous fishes and labyrinthodonts (16).

Perhaps the nearest approach to the remarkably complete succession which we have in the Central Himālayas is furnished in Hazāra, where a sequence, complete but for a local unconformity between the two lowest systems, extends from Triassic through Jurassic and Cretaceous to the great Nummulitic or Eocene formation. The Triassic rocks of Hazāra rest unconformably on the Devonian or so-called infra-Trias, consisting of some 50 to 100 feet of acid felsitic material (probably of volcanic origin and associated with a polisitic hematite) at the base, followed by a limestone formation of from 500 to 1,200 feet containing Megalodon and other characteristic fossils (14).

In the Tenasserim Division of Southern Burma there are limestones from which a few fossils have been obtained, having affinities with the Carboniferous limestone of Sumatra and less intimate relations to some forms occurring in the Productus limestone of the Salt Range. So far as they go, the fossils indicate approximately a Permo-Carboniferous age. These limestones are associated with a series of shale and sandstone beds, distinguished as the Moulmein series, which have a total thickness of about 5,000 feet and rest on another series, the Mergui series, which consists of some 12,000 feet of unfossiliferous sandstones, grits, and shales. Limestones similar to those which are fossiliferous in Tenasserim are found east of the Salween river, and farther north in Karenni, where they have yielded a number of fossils, chiefly brachiopods, like Athyris, Productus, and Spiriferina, belonging to species closely related to forms known in the Productus limestone of the Salt Range. South-west of Hsipaw (Thibaw) also, in the Northern Shan States, limestones have been found with the form Fusulina, which is so common in the Carboniferous and Permian formations elsewhere. This area, only recently visited by the Geological Survey, promises a geological record which, from its geographical position between the standard stratigraphical scale of Northern India and the Palaeozoic formations of Sumatra and adjoining areas, will be of unusual value as an index to the physical geography of the Indian region in Palaeozoic times.

In the gorge of the Subansiri river in Assam numerous boulders of limestone and sandstone have been found, including fossils of Lower Productus limestone affinities, but these may possibly have been brought from the Tibetan plateau (15).

The Jurassic system is well represented in parts of Western and North-western India. The so-called ‘massive limestone’ of Baluchistan, which is several thousand feet thick and forms many conspicuous peaks, such as the Takatu north of Quetta, and the Takht-i-Sulaiman, is of about the same antiquity as the oolite of England, the fossils from its uppermost strata being of Callovian age. The massive limestone rests conformably on a great thickness of shales and limestones corresponding in age to the Lias of England. A somewhat similar succession is characteristic of the North-West Frontier Province, where, too, there are coverings of Cretaceous and Tertiary strata, with generally unimportant interruptions in the stratification.

A traverse of the country between Ali Masjid in the Khyber Pass and the British frontier at Shinawari covers representatives of every system from the Tertiary to the Carboniferous, and some older altered rocks of probably Lower Palaeozoic age. From the Bara valley south to the Sāmāna range, the Palaeozoic rocks are covered, the beds exposed being either Mesozoic or Tertiary in age, thrown into a series of folds with a tendency to inversion over to the south. This gives the northern slopes of the ranges a comparatively gentle inclination, while their southern scarps are steep and rocky. An example is furnished by the Cretaceous and Jurassic rocks which form the heights of Dargai, rendered famous by the engagement of October 20, 1897, when the British suffered the tactical disadvantage of having to carry the scarp face (13).

In Northern Hazāra there is an exposure of beds precisely similar in lithological character and fossil contents to the remarkable Spiti shales of the Central Himālayas, and they are covered, too, by flysch-like beds identical in character with the Giumal sandstones of Spiti and Kumaun (14). On the other hand, Jurassic rocks in Southern Hazāra present a facies quite unlike that of the Spiti shales, being more calcareous and sandy, and generally more like the Jurassic rocks of the Salt Range (14).

Remains of formations deposited in Upper Jurassic and Lower Cretaceous times are found exposed in the region west and north-west of the Arāvalli range—in Cutch, in the Rājputāna desert near Jaisalmer and Bikaner, and in the western part of the Punjab Salt Range. These have been most completely examined in Cutch, where they attain a development of over 6,000 feet, ranging from a stage about equivalent to the Bathonian (Middle Jurassic) of Europe, through the Upper Jurassic, to the Neocomian, without any decided unconformity. These rocks have been divided into four series distinguished by local names as follows:—

The great mantle of sand which has spread over Rājputāna during recent times effectually conceals large areas of rocks, patches of which here and there peep through; but, being isolated, they cannot be grouped with certainty except where they are fossiliferous. In the neighbourhood of Jaisalmer, however, there are highly fossiliferous limestones which include many forms identical with those characteristic of the Chāri series of Cutch, and these are overlaid by other series consisting of sandstones and limestones which have yielded fossils resembling those of the Cutch Katrol series. We thus have proof that the sea extended so far eastwards during Upper Jurassic times.

From Jaisalmer to the Salt Range, where we find marine Jurassic rocks again, is about 350 miles due north, the whole country between being completely covered with recent alluvial accumulations. In the Salt Range, we have, as already described, an uninterrupted succession from the Permo-Carboniferous to Lower Triassic times; then an interruption occurs, and the next formation preserved is of Middle or Upper Jurassic age. These strata are developed in the western part of the Salt Range extending to the Indus, being exposed again farther west to the Chichali (Maidani) hills (32°51’ N.; 71°11’ E.) and in the Sheikh Budin hills (32°18’ N.; 70°49’ E.). Small patches of coal occur near the base of the series; and for the rest it consists of an alternation of conglomeratic sandstones, shales, and limestones, the last-named being especially developed in the western exposures. Two less usual formations are a bed of hematite and layers of a peculiar golden oolite, similar to that well-known in Cutch. So far as they have been examined, the fossils correspond with those of the Chāri and probably also of the higher series of Cutch.

The narrow encroachments of the sea which took place in Jurassic times on the Peninsula were extended during the Cretaceous period, and especially in Cenomanian times, when profound changes occurred in the physical geography of the earth. Relics of this Cretaceous transgression of the ocean, preserved on the Coromandel coast of Madras and in the Narbadā valley, are well-known examples which may be conveniently selected for special notice.

The best studied of these marine formations is that represented by three patches on the Coromandel coast, where, by a very narrow accident of relative level between sea and land, we find highly fossiliferous rocks which have made a contribution of inestimable value to our knowledge of marine zoology in Upper Cretaceous times. Situated as a sort of half-way stage between the Pacific and the Atlantic areas, this coast, which was immersed to a very small depth, formed a home and final resting-place for many animals which migrated under stress of competition from one region to the other, marking out their route, like the old East Indiamen, by wrecks on the southern coast of India, in Natal, and on the west coast of Africa. The fossil remains include many forms which appear to have flourished from Brazil right around the oceanic belt to British Columbia, together with others which modified themselves to develop species peculiar to the conditions in various parts of the Cretaceous sea. Supposing a few feet of elevation in Cretaceous times, and no shells would have stranded on the shelving beaches of the Coromandel coast; a similar amount of depression in recent times would have hidden the deposits beyond the reach of the geologist. As it is, the small patches of strata on the eastern coast of Madras form a little museum of Cretaceous zoology, in which nearly a thousand species of extinct animals have been recognized; and, by the inclusion of many cosmopolitan forms, they permit the correlation of these rocks with those in many parts of Europe; Syria; the north-western borders of India; North, West, and South Africa, and Madagascar; Brazil; the Eastern, Central, and Western States of the American Union; British Columbia; Japan; Sakhalin; Borneo; and Australia (1).

The highly fossiliferous Cretaceous rocks of the Coromandel coast form three small patches separated from one another by the alluvium of the Vellar and Penner rivers, and by the sub-recent Cuddalore sandstones. The largest of the three patches is in Trichinopoly District; the other two are west of Cuddalore and north-west of Pondicherry respectively. They rest on the eroded surface of the old gneiss, or unconformably on the Upper Gondwana beds.

The fossils show a range in age from the lowest beds corresponding with the upper greensand (Cenomanian) to the Danian. The strata are divided as follows into four stages:—

The lowest of these stages, distinguished by the name Utatür, is only partially covered by the younger beds, and is exposed as a wide band along the western border of the Trichinopoly Cretaceous patch. Its base is generally a coral-reef limestone; but the principal part of the series consists of fine silts, calcareous shales, sandy clays, sands, grits, and some conglomerates. The fossils include fragments of cycadaceous woods, often bored by molluscs, with a rich assemblage of marine forms generally indicating, by their nature and mode of preservation, a littoral habit. Near the base of the Utatür stage occurs the common lower Cenomanian ammonite Schloenbachia inflata, Sow., a species characteristic of this horizon in Europe, West Africa, Brazil, Australia, California and elsewhere; but more especially of the Cretaceous rocks of the Atlantic province. The majority of the fossils are, however, near relatives of, or identical with, those found in the Pacific province. In the higher beds there is a rich Acanthoceras fauna, including the world-wide form A. rhotomagense, Brong., found also at various places in Europe, Syria, the Caucasus, Persia, and the regions bordering the north-west of India, Japan, South, West, and North Africa, and Madagascar: in this zone, also, Turrilites costatus, Brong., and Alectryonia carinata, Lam., agree in indicating a middle and upper Cenomanian age.

The uppermost zones of the Utatür stage mark the commencement of Turonian conditions by the appearance of the characteristic lamellibranch Inoceramus labiatus, Schloth., and of ammonites related to the European form Mammites nodosoides.

There is a slight break between the Utatür and the next overlying Trichinopoly stage, shown by a stratigraphical unconformity, as well as by a considerable change in the fauna. Among the Trichinopoly beds there occur typical relatives of Pachydiscus peramplus, Mant., an ammonite characteristic of the lower chalk of England, and of the corresponding horizon in many other parts of the world, ranging over the Atlantic province, and in the Pacific area as far as Japan. The higher Trichinopoly beds mark the incoming of a lower senonian fauna among the gastropods and lamellibranchs as well as among the ammonites. Thus the genus Schloenbachia, Neum., represented in the lowest Utatür beds by S. inflata, Sow., is here represented by the tricarinata type.

Then follows the Ariyalur stage, covering a large area on the east side of the Trichinopoly patch. A point of peculiar interest in connexion with this series is the occurrence in it of a tooth and some ill-preserved bones of the dinosaurian Megalosaurus, resembling M. Bucklandii, a well-known form from the Stonesfield slate, belonging to the bathonian stage or great oolite of England, and thus much older than these rocks in Southern India. As this genus is not known above the neo-comian in Europe, we have another example of the class, so well illustrated by the Gondwana fossils, showing the different rates of development which occur in widely separated land areas, cut off by sea or by other physical barriers from one another. The most important cephalopods in the Ariyalur beds are the upper senonian species of Pachydiscus and Baculites, B. vagina, Forb., being especially characteristic. Apparently this series is also represented in the Pondicherry area, where, as in Trichinopoly, the next and highest stage is also preserved.

The Ninniyur beds are intimately related to those of the Ninniyur stage, but contain a fauna sufficiently characteristic to permit their correlation with the danian stage of the Upper Cretaceous in Europe. Thus the disappearance of ammonites and other characteristic Mesozoic forms foreshadows the faunal characters which distinguish the approaching Tertiary period. The characteristic form, Nautilus danius, Schloth., makes its appearance, and the only other genus that requires mention to ensure the Mesozoic character of the beds is the gastropod Nerinea, Defr., whose name is used to distinguish the highest beds in the Pondicherry area. Thus closes the most complete fragment of Mesozoic history in peninsular India, the only record we have of the life in the seas washing the Coromandel coast when the Mesozoic era was approaching its close in Europe. The Cretaceous sea, which left such perfect samples of its inhabitants on the Coromandel coast, also stretched north-eastwards as far as Assam; and there, on the margin of a mass of old rocks which formed a part of the peninsular crystalline gneisses, it deposited limestones, sandstones, and shale beds, containing numbers of fossils identical with the more completely studied formations in Southern India.

The rocks of corresponding Cretaceous age on the western coast are known as the Bāgh beds, which occur in the Narbadā valley and separate the Deccan trap-flows from the underlying Archaean gneisses. Some forty species of marine animals have been identified in these beds, including a few cosmopolitan forms which show a specific identity with those in the Coromandel beds, but many of them are distinct types manifesting a greater affinity with Cretaceous fossils from Arabia, Palestine, and Europe, areas which were covered by the same great ocean. In Kāthiāwār some sandstones in the neighbourhood of Wadhwān resemble the Bāgh beds in lithology, in the few imperfectly preserved fossils which they have yielded, and in their position unconformably below the Deccan trap-flows.

Above the youngest member of the Vindhyan system there is a great gap of unknown width in the geological history of peninsular India; and it is probable that much of the record has been destroyed by denudation, for its next chapter commences with a formation deposited on a land surface when India was part of a large continent exposed to the weather.

The oldest rocks after the Vindhyans are distinguished as the Talcher series, which form the lowermost stage in a great system of sub-aerial and fresh-water deposits known as the Gondwana system. In Gondwana times India, Africa, Australia, and possibly South America, had a closer connexion than they appear to have at present. Although probably at no time forming a continuous stretch of dry land, they were sufficiently connected to permit of the free commingling of plants and land animals. At different parts of this great southern continent there occur peculiar boulder-beds whose special characters appear to be best explained as the result of ice action. The boulders of this peculiar formation of the Tälcher series vary from mere pebbles to blocks weighing many tons, generally well rounded and rarely scratched, lying often in a matrix of fine silt, a matrix which would not exist if the boulders had reached their present positions by rolling in rapid streams. The formation in New South Wales which is taken to be the equivalent of the Tälcher boulder-bed has a similar structure, with large and sometimes striated boulders embedded in a fine, silty matrix; and in this case the tranquil conditions under which the formation was laid down are shown by the inclusion of numerous delicate Fénestellae and undisturbed bivalves lying in the silt. The age of the Australian formation is fixed by the associated Upper Carboniferous marine beds, and this testimony agrees with that of the boulder-bed of the Salt Range already referred to (ante, page 70). In Kashmir, beds have been found containing Gangamopteris (a typical Lower Gondwana plant), associated with fish and labyrinthodont remains related to those of the Permo-Carboniferous in Europe ¹.

The lowermost beds of the Gondwāna system are thus fixed by indirect evidence as Upper Carboniferous or Permo-Carboniferous in age. Later on it will be shown that the uppermost stages of this system are associated with marine deposits of oolitic, or possibly neocomian, age. We thus have a great system of strata ranging from the Carboniferous, through Permian and Triassic times, to the period during which the well-known oolites of Europe were formed. The fossil contents of this system give a record of the natural history of the great southern continent of Gondwāna, which differs in a remarkable and most interesting way from that of the northern hemisphere. Allusion has already been made to this in the Introduction; and after a brief description of the subdivisions of the Gondwāna formations, the question will be referred to in greater detail, though the few pages to which this chapter is necessarily limited are insufficient for a full discussion of a subject which possesses such an important bearing on palaeontology. The reader who wishes for more information is referred to the memoirs cited at the end of the chapter.

The Gondwāna rocks are preserved as small patches let down, mostly by faulting, into the great crystalline mass of the Peninsula. Originally they must have covered a much wider area; but as the Peninsula has been exposed ever since to the free action of weathering agents, the Gondwana formations have been cut into like the older formations, and the coal-measures thus preserved in India now form but a fraction of those that once existed. Isolated patches of Gondwana rocks, including coal beds, have been involved in the folded extra-peninsular area, in the Darjeeling District, and in Northern Assam. The string of Gondwana patches which determines the direction of the river Dâmuda includes our most valuable deposits of coal. Their faulted, parallel boundaries and general east-west alignment suggest the action of the same earth movements as occurred in structural lines parallel to the subterranean ridge of high specific gravity running across India to the south of this line—the great depression of the Gangetic valley and, farther afield, the main axis of folding in the Himālayan region. All these phenomena are probably connected, though not necessarily contemporaneous.

The lowest subdivision has already been referred to as the Talcher series. The rocks of this series are generally soft sandstones and peculiar silty shales, often of a greenish hue, which break up in a most characteristic way into small angular fragments. The peculiar characters of the Talcher rocks permit their ready recognition; and though a comparatively thin formation, probably not exceeding 800 feet in thickness, they are developed, with all their peculiar characteristics, over an enormous area on the Peninsula.

In the upper layers only a few plant remains have been found; but the Tälchers generally, notwithstanding their lithological suitability for the preservation of delicate fossils, are remarkably devoid of signs of life, a feature which is consistent with the evidences of great cold indicated by the glacial boulder-beds near the base of the series.

The next younger beds are grouped together under the name Dâmuda series, and these are subdivided in Bengal into the following stages:—

3. Rāniganj stage.

4. Ironstone shale stage.

5. Barākar stage.

The Barākars are recognizable in many other areas; but the upper stages cannot be identified with certainty, and local names like Kamptee, Bijori, and Motur are used to distinguish the Dāmuda beds above the Barākars in other coalfields. It is in the Dāmuda series that the most valuable Indian coal-seams occur. The associated rocks are all sandstones and shales, which sometimes attain a thickness of 10,000 feet. The iron-stone shale stage is so called on account of the lenses of clay-ironstone which, as in the Rāniganj coalfield, sometimes occur in sufficient abundance to supply a valuable iron-ore. All these stages are in general conformity with one another, though the upper may be found to overlap the lower.

There is generally, however, a slight unconformity between the uppermost stage of the Dâmuda series and the next, which is distinguished as the Pânchet series. The Pânchets are characterized by the absence of coal-seams, being composed of micaceous sandstones, often of a greenish colour, with bands of red clay. The series is well-known on account of the reptilian and amphibian fossil bones it has yielded, besides a few fossil plants which show more pronounced affinities with those of the Dâmudas than with the higher beds.

The whole of the foregoing series—Tâlcher, Dâmuda, and Pânchet—make up the lower division of the Gondwâna system, being cut off from the Upper Gondwânas by a marked stratigraphical break, accompanied by a contrast in fossil contents. The plants of the Lower Gondwâna beds include many equisetaceous forms, while those of the Upper Gondwânas show a prevalence of cycads and conifers; the species of common genera of ferns, as well as other orders, are quite distinct in the two divisions.

The Upper Gondwānas have a lower series, distinguished as the Rājmahāl series in Bengal and as the Mahādevas in the central parts of the Peninsula. The Mahādevas attain a thickness of 10,000 feet in the Sātpurā area, most of the rocks being sandstones and unfossiliferous. The Rājmahāls, on the other hand, have yielded a number of fossil plants, and are interesting, too, on account of the great sheets of basaltic lava interstratified with the shaly and sandy sediments, attaining a thickness of over 2,000 feet. The Rājmahāl lava-flows are often amygdaloidal like those of the Deccan trap series, the cavities yielding agates and zeolites of considerable variety and beauty.

Rocks of Upper Gondwana age occur at various places along the east coast of the Peninsula. In some cases marine fossils have been found associated with the plant-bearing beds, and these have helped to fix the position of the Upper Gondwana in the standard scale of marine strata. More pronounced evidence as to the age of the upper limit of the Gondwana is afforded by the occurrence of plant-bearing beds in the so-called Umia series of Cutch, whose age has been already referred to as about equivalent to the neocomian of Europe (ante, page 76).

A series between the Rajmahâl horizon and the Umia series deserves special mention on account of the animal remains which it has yielded. Its beds occur in the Godâvari valley, and have been named the Kota-Mâleri series from two villages near which they are developed. In the lower or Mâleri stage fossil remains of the remarkable fish Ceratodus, species of which are still living in Australian waters, and of the reptilian genera Hyperodapedon and Parasuchus, have been found with numerous coprolites; while in the Kota stage a greater variety has been found, including the crustacean Estheria, and several forms of fish and reptiles, with plant remains which indicate the position of these beds in the Gondwâna sequence.

At the time when the Glossopteris flora flourished on the great southern continent of Gondwana-land, Lepidodendron, Sigillaria, and Calamites were conspicuous among the forests of the northern hemisphere. But the separation of the two great continents was not sufficiently complete to prevent the southward extension of some members of the Lepidodendron flora to Africa and South America; and the fact that typical members of the Upper Palaeozoic Lepidodendron flora, as it is known in Europe and North America, have now been found associated with the Glossopteris flora in South America and in South Africa proves beyond question that the two were coexistent ².

The predominant flora of the Lower Gondwana system, in which Glossopteris and Gangamopteris are prominent genera, has much closer affinities with the Mesozoic plants of Europe than with the plants of the Upper Palaeozoic coal-measures. This fact at first seemed inconsistent with many other evidences pointing to an Upper Palaeozoic age for the Lower Gondwānas; but the explanation offered by the earlier members of the Geological Survey of India, though for many years a stumbling-block to European palaeontologists, has received conclusive support in recent times. A flora closely resembling that of the Indian Gondwānas was found represented also in Australia, South and East Africa, Argentina, and Brazil. In most of these places, too, the formations in which the fossil plants occurred were associated with a boulder-bed having the peculiarities of that at the base of the Indian Gondwānas, and regarded as the result of ice action by the Indian geologists. With a boulder-bed of Permo-Carboniferous age at the base, and a marine intercalation of Jurassic and neocomian forms near the summit of the Gondwana system, we have an inferior and a superior limit of the time-scale over which to distribute its various series. A considerable fraction of the lowest beds must represent the Permo-Carboniferous, Permian, and Triassic periods; and yet the plants they contain show, when compared with European fossils, a predominating Rhaetic and Jurassic facies. The remarkable agreement between the Glossopteris (Gondwana) flora of India and the fossil plants of similar formations in Australia, Africa, and South America can only be explained on the assumption that these lands, now separated by the ocean, once constituted a great southern continent.

That India and the southern and central parts of Africa were once united into one great stretch of nearly continuous dry land is proved by overwhelming evidence ³. In the first place, besides the remarkable correspondence among the plants which flourished during Upper Palaeozoic times in India, South Africa, and the portions of East Africa which have been explored, there is an agreement between the peculiar generalized labyrinthodonts and reptiles of which remains are found in the Panchet series of India and in corresponding beds in South Africa. So far as this evidence goes, it points either to a complete land connexion, or to an approximation sufficiently close to permit free migration of land animals and plants.

A study of the distribution of Jurassic cephalopods indicates the existence of a tropical sea to the north-west of this supposed land barrier, and of a cold sea to the south-east. The separation was not, however, sufficiently complete to prevent the migration of species from the Cutch area, which we presume to have been on the north-west side of the barrier, to the Lower Godāvari, which was probably on its south-eastern shores. But a shallow strait in Upper Jurassic or Lower Cretaceous times would be sufficient to account for the small amount of commingling thus indicated by the occurrence of identical species on opposite sides of India.

The Upper Cretaceous fossils demonstrate the existence of the land barrier more completely. The marine Cretaceous beds near Bāgh, in the lower Narbadā valley, contain fossils which, especially the echinoderms, show striking resemblances to those of the Cretaceous beds of Syria, North Africa, and Southern France, all patches of rocks deposited in the great ocean of which the modern Mediterranean is a shrunken relic.

But the Bâgh beds differ in facies from the Cretaceous beds of the Trichinopoly area, since in the latter, though there are many forms that had a world-wide distribution in Cretaceous times, types related to the fossils of the Pacific province preponderate, as shown by numerous correspondences with South Africa, Borneo, Japan, Sakhalin, Chili, California, Vancouver, and even as far as Queen Charlotte Islands. We have thus a contrast between the Mediterranean-Atlantic Cretaceous province and that of the Pacific, and this contrast is preserved in the Bâgh beds on the west of India and the Trichinopoly formations on the east coast. The evidence goes even farther, for in Assam, Trichinopoly, and South Africa the Cretaceous beds show a distinctly littoral character, indicating the fact that the old Mesozoic coast-line on the east was not far from a line joining these places.

There is still another piece of evidence as to the existence of the old Indo-African continent, all the more striking because it belongs to an entirely different field of observation. It is found that between the Seychelles, which are connected by comparatively shallow waters with Madagascar and Africa, and the Maldives, which are on the Indian continental platform, there exists a submarine bank, preventing the ice-cold Antarctic currents that characterize the greater depths in the South Indian Ocean from extending into the Arabian Sea, which has thus a higher temperature than the water at corresponding depths to the south of this bank. We have here the remains of the old continent, depressed sufficiently to cut off India from South Africa, but still enduring as a bank between the great abyssal depressions to the north-west and the south-east.

Finally, the modern distribution of animals is explained by this occurrence of a Mesozoic Indo-African continent, and in turn furnishes further evidence in favour of the conclusion already based on palaeontological data. Within the part of India south of the Gangetic plain are numbers of genera and species not found in other parts of the Indo-Malayan region, which have near relatives in Madagascar and South Central Africa. These, distinguished as the Dravidian constituent of the Indian fauna, are comparatively low forms, mostly reptiles, batrachians, and invertebrates, with only one mammalian genus, Platacanthomys. The likeness between the Dravidian fauna of the Indian Peninsula and some forms in Madagascar can only be accounted for by this supposition of an ancient land communication, while the amount of divergence they show is no more than would be expected from independent evolution, since the separation occurred in early Tertiary times ⁴.

The great revolutions in physical geography, which took place towards the end of the Cretaceous and during early Tertiary times, resulted in the break-up of the old Gondwāna continent, and were followed by the rise of the Himālayan range. These orogenic movements appear to have been caused, or accompanied, by igneous action on an unusually grand scale. The great masses of basic lava covering more than 200,000 square miles in peninsular India remain as a fragment of the enormous flows which must have spread over that area, and probably over a very much larger portion of the old Gondwāna continent to the west and south, now buried under the Indian Ocean or removed by denudation. Among similar phenomena in other parts of the world, at or near the same period, may be mentioned the great basaltic flows of North-western Europe with their associated granites, gabbros, and other intrusive rocks, and the Laramie series in the United States which very closely parallels the case of the Deccan trap of India. Besides the Deccan trap, other intrusive and extrusive igneous rocks made their appearance at about the same time in parts of extra-peninsular India. Burma contains intrusives of basic and ultra-basic rocks cutting through the early Tertiary strata, and now remaining as conspicuous masses of serpentine. In the North-western Himālayas similar rocks, accompanied by volcanic ashes and probably also by lavas, were formed during and subsequent to lower eocene times, while in Baluchistan even more extensive series of eruptions have been detected. Finally, with this great period of earth-movement we must connect much of the granite which is so prominent in the Central Himālayas and contributes to the great core of the range.

The most extensive and best known of the instances of eruptive activity which characterized the close of the Mesozoic, and the opening of the Cainozoic, era is naturally the Deccan trap. The great lava-flows which make by far the chief part of this formation constitute the plateau of the Deccan, concealing all older rocks over an area of 200,000 square miles, filling up the old river valleys, and levelling the surface of the country. Subsequent denudation has carved these lava-flows into terraces and flat-topped hills, with, as in the seaward face of the Sahyādri or Western Ghāt range, steep scarps, rising to about 4,000 feet and indicating a part only of the original thickness of the accumulated lavas, ashes, and beds of interstratified marl. The trap-rock is usually a form of olivine basalt or augite-andesite, rarely porphyritic, but often vesicular with amygdala of beautiful zeolites, calcite, and agate.

At the base of the flows are beds in which limestones of lacustrine origin predominate. These beds, known as the Lameta series, were laid down unconformably on all the older formations, even on the youngest members of the Gondwāna system, while they were themselves exposed to local denudation before the lavas spread over and protected them from the weather.

Among the few fossils which have been found in the Lameta series are the bones of a large dinosaur, Titanosaurus indicus, Lyd., allied to some Lower Cretaceous and Upper Jurassic reptiles in Europe. The occurrence of this form in strata which are certainly not older than Upper Cretaceous agrees with the evidence of the Megalosaurus from the Ariyalur stage of Southern India, in pointing to the backward state of evolution among Indian reptiles in Upper Mesozoic times—one more among the many evidences from Indian geology to prove that correlation of strata by land animals often contradicts the evidence of marine forms.

In making an attempt to fix the position of the Deccan trap in the European stratigraphical scale, the chief point to guide our judgement is the fact that we are limited below by the cenomanian (Bagh) beds and above by the Nummulitic rocks of Cutch, while in Baluchistan what appears to correspond to the base of the series is associated with marine strata of about senonian age, and in Sind the upper flows have spread out over beds regarded as equivalent to the oldest Tertiary in Europe. The eruptions thus probably began at about the time of the formation of the upper chalk of England, and finished before the remarkable foraminiferal genus Nummulites made its appearance and spread throughout the great Eurasian central ocean. During this interval, which geologically is a very short one, there was time for the accumulation of lava-flows which amounted to not less than 6,000 feet in thickness in some places, with intervals of rest sufficient for lakes, stocked with fresh-water mollusca, to form on the cold surfaces of several of the lava-flows. So this remarkable accumulation of volcanic materials has remained until today, with its original horizontality of bedding but slightly disturbed. Except on its northwestern fringe, where it was bent down with the subsiding land to the north to suffer the encroachment of the early Tertiary sea, it has remained exposed to the weather, which has carved the great lava-flows to produce the characteristic scenery of the Deccan plateau.

Until the Deccan trap has been dissected out by the weather in the way in which the Tertiary basalts of North-western Europe have been cut up, we shall have very little visible evidence of the masses of ultra-basic rocks which almost certainly lie below. But it is just possible that portions of these ultra-basic rocks have been squirted into the early Tertiary rocks of the North-western frontier; and the numerous masses of olivine-rock exposed in Mysore and the Madras Presidency may even have had a similar origin, though it is also conceivable that the latter are as old as the Cuddapah lava-flows which, like the Deccan trap, once extended far beyond their present limits. These dunite-masses in Madras can be dismissed with a very few words. The majority of them are almost pure olivine, though at times they contain enstatite and chromite, and locally pass, by concentration of other minerals, into various forms of picrite. But the chief feature of interest in connexion with these rocks is their frequent, almost constant, decomposition into magnesite without ordinary serpentinous alteration. The original olivine-rocks must have been attacked by water and carbonic acid of deep-seated origin, probably originally contained in the magma; and, with the formation of magnesite, chalcedonic silica is also separated. The so-called Chalk Hills near Salem (11° 39’ N.; 78° 10’ E.) form a well-known instance of these peridotite eruptions, being so named because of the abundance of dazzling white magnesite.

Passing on to Burma, we find numerous and large masses of peridotite which were erupted in early Tertiary times. Unlike those of Madras, whose age, it should be remembered, is unknown, the Burma peridotites are always much serpentinized. One instance, interesting because of its connexion with the valuable mineral jadeite, may be taken as an example. In Upper Burma, in the vicinity of Tawmaw (25°44’N.; 96°14’E.), serpentinous rocks are found piercing strata of miocene age. Microscopic examination of the rock shows that large quantities of the original olivine have escaped hydration, but most of the mineral has been altered to serpentine. The jadeite occurs in the masses of serpentine, standing out, when exposed, by its white colour against the dark-green serpentine ⁵.

Serpentinous masses, presumably of the same age as that intrusive in the miocene rocks of Upper Burma, are found also as irregular bosses and dikes at various places mainly on the eastern side of the Arakan Yoma, where they are intruded into rocks of the Chin series. The serpentine, with chromite, found in the Andaman Islands probably belongs to the same series of eruptions.

Baluchistan was the scene of the grandest and most interesting manifestations of igneous action during this period. With the beds of volcanic ash which are found below the hippuritic limestone (Cretaceous), and at different stages to the middle eocene, there are certain basic intrusions which, with the ash beds, were formed before the folding of the rocks, and have consequently suffered the usual deformations. Either as a cause or as an accompaniment of the folding movements, great intrusions of granophyric rock—granites and more basic types—were forced into the Nummulitic limestone and associated rocks some time after the close of the eocene period. Then followed the injection of dikes and sills of dolerite before the pliocene strata were deposited. But this did not close the volcanic action in this interesting area: lavas and ashes were ejected and further material injected into the pliocene (Siwalik) rocks, while eruptive activity persisted on to recent times in Baluchistan as well as in Persia, and some of the volcanoes are still active, though showing signs of senility ⁶.

With a brief remark on the granitic rocks, because of their possible connexion with the Himālayan granite, we must leave this attractive section of Indian geology. It is interesting to note that the granites found cutting the limestones, which are full of Nummulites, often show the peculiar granophyric structure so characteristic of the similar early Tertiary rocks of North-western Europe, that they pass in the same way into more basic types also with micrographic structures, and that they are similarly traversed by basaltic dikes. But besides these peculiar features they are in places porphyritic, and otherwise recall some granites in the Central Himālayan zone.

While the Deccan trap was being poured out on the Peninsula of India, at the time when the typical Cretaceous fauna of Europe was gradually giving way to the forms which mark the distinctly Tertiary formations, deposition was going on in the seas washing the west coast of India; and as a result we have preserved, in parts of Sind and Baluchistan, sediments which contain fossils with affinities both to the Cretaceous and to the Tertiary types. The exact side of the dividing line on which a particular formation should be placed can be decided only by detailed examination of its fossil contents.

In one of these cases, which alone there is space to mention, we have a series of beds in which some highly fossiliferous olive-coloured shales contain large quantities of a peculiar globose species of lamellibranch, Cardita beaumonti, which gives its name to the formation. Associated with this form are reptilian remains with Mesozoic affinities, besides corals and echinoids of mixed Cretaceous and Tertiary types which have yet to be critically examined. The association, as an interbedded flow, of amygdaloidal trap with the Cardita beaumonti beds gives one fixed point for the age of a portion of the Deccan trap.

The Tertiary system which forms the southern fringe of the Himālayas is divided as follows:—

Sirmūr series

• Sabāthu stage
• Dagshai stage
• Kasauli stage

Murree beds

Siwālik series

• Lower Siwālik or Nāhan stage
• Middle Siwālik
• Upper Siwālik

A review of these Tertiary deposits shows a general passage from marine beds at the base to the great river deposits of the Siwālik, which are essentially similar in origin to the modern alluvia deposited by the rivers emerging from the Himālayan valleys on to the plains of Hindustān.

The lowest or Sabāthu stage of the Sirmūr series consists of a highly disturbed set of grey and red gypseous shales, with layers of limestone and sandstone, in which the fossils indicate a marine origin and an age equivalent to the Nummulitic beds.

The Dagshai stage, with its hard, grey sandstones and bright-red clays, follows conformably above the Sabāthu beds, and in turn passes up into the Kasauli stage, which is essentially a sandstone formation in which the clay beds are distinctly subordinate in quantity.

On reaching the Kasauli stage all evidences of marine action disappear, and the deposits seem to have been formed in fresh water, the sea having then permanently retreated from the plains of Northern Hindustan, while the conditions favourable to the formation of the great thicknesses of sandstones, clays, and conglomerates which mark the Upper Tertiary, or Siwalik, series were developed.

Marine conditions prevailed in Lower Tertiary (Nummulitic) times along the foot of the Himālayas, as far east at any rate as Garhwāl, and the deposits of marine Sabāthu beds can be traced at intervals north-westward to Jammu, while Nummulitic rocks occur also in the Salt Range; over various parts of the North-West Frontier Province, covering up large tracts of older rocks; at the back of the zone of crystalline, and now generally snow-covered, peaks in the far parts of Kashmir and Ladākh; on the Tibetan border in Spiti and Kumaun; and away to the far east in the region of Tibet north of Sikkim. Still farther east, in Assam and Burma, Nummulitic rocks occur in numerous places. Within the limits of this chapter it is possible to refer to a few only of the remarkable features of this widespread series of deposits. The guiding line throughout is generally the occurrence of the remarkable foraminifer Nummulites, which, on account of the way it spread itself throughout Europe and Central Asia in early Tertiary times, is as useful in marking a stratigraphical horizon as the freely migrating cephalopods of the Mesozoic group.

In Sind, where the Tertiary marine rocks have attained an exceptional development, the following subdivisions are recognized:

The lower portion of the Rānikot stage is poor in fossils, consisting of pyritous and carbonaceous shales and soft variegated sandstones with gypsum. The upper beds, however, are rich in marine fossils, among which in the two uppermost zones are Nummulites of different species which persist thence, through the Kīrthar, to the Lower Nāri stage.

The great thickness of beds represented by the Rānikot and Lower Kīrthars has not been preserved in Baluchistan, where a series of beds distinguished by the name Ghāzij rest directly, and with distinct unconformity, on the Cardita beaumonti beds. The Ghāzij passes locally into flysch-like material (the Khojak shales), which apparently accumulated with rapidity and produced a great thickening of the beds without much change in their fossil contents. Above the Ghāzij-Khojak stage we find a limestone formation, known as the Spīntangi stage, which is represented in parts of Baluchistan, and caps the scarp of the Kirthar range between Baluchistan and Sind. The great thickness of Khojak, Ghāzij, and Spīntangi beds in Baluchistan represents merely the Upper Kirthar of Sind. Between the middle eocene Spīntangi beds of Baluchistan and the Lower Nāri (upper eocene) there is a distinct unconformity, corresponding approximately to the bartonian of England. But this gap is not apparent in Sind, the yellow or brown Nāri limestone following the white limestones of the Kirthar stage with seeming conformity. With the Lower Nāri end the Nummulites, and the limestones in which they occur are succeeded by a great thickness of comparatively barren sandstones.

Representatives of the Nummulitic series, which are so well developed in Sind and Baluchistan, occur also in Cutch and Kāthiāwār, in Surat and Broach.

The Lower Tertiary rocks of the Kohāt region are remarkable for the valuable deposits of rock-salt which occur at their base, and which, in default of contradictory evidence, are assumed to be of Tertiary age, though the base of the salt-bearing series is not exposed. The salt, and its associated gypsum, shales, and sands, in the Kohāt region present certain characteristics which distinguish the formation from the salt-marl occurring so mysteriously below the Cambrian beds of the Salt Range in the Punjab. The colours are generally grey instead of reddish, and the potash and magnesium minerals of the Salt Range are not found in the Kohāt area.

The Lower Tertiary rocks of the Kashmir and Ladakh area deserve special mention on account of the associated peridotites and basic igneous lavas and ashes—a set of rocks which, placed in this stratigraphical position, suggests a genetic connexion with the great Deccan trap eruptions and some of the basic eruptives of Baluchistan of about the same period.

The Nummulitic rocks of Assam are of importance on account of the economic value of their limestones, coal-beds, and mineral oil, which are referred to in detail in the chapter on Mines and Minerals (Vol. III).

In miocene times the sea was driven back, and marine strata of this age are consequently restricted to areas nearer the present coast lines. In Sind there is a fine display of marine miocene beds in the Kirthar range, where the series is cut through by the Gaj river and is named the Gaj series in consequence. In Cutch beds of corresponding age are well developed, while far away on the other side of the peninsular mass there are relics of the miocene sea in Upper Burma, distinguished as the Yenangyaung series.

The earliest records of the Tertiary history of Burma are still sealed up in a great thickness of flysch-like shales and limestones which, occurring in the forest-clad and almost inaccessible Arakan hills, have only been superficially examined. Above these, on both flanks of the zone of older rocks which stretches from Cape Negrais northwards to Manipur, we find marine beds, the Bassein series, of upper eocene age, showing that a great part of this area was covered by a shallow sea. This sea became, in lower miocene times, silted up by sand and mud, with included organic remains, which afterwards gave rise to the thin coal-beds and petroleum-bearing sands of the Prome stage. Then followed a further inroad of the miocene sea, with its corals, echinoderms, molluscs, crustacea, and fish, many of whose direct descendants are living to-day in the Indian and Pacific Oceans. These, and the deposits formed in the estuaries of the rivers which poured their contents into the miocene sea of Burma, constitute the Yenangyaung stage. Then followed the changes which, after a local denudation of the Yenangyaungian sediments, resulted in the deposition of 20,000 feet of sandstones in river valleys that formed the home in pliocene times of many remarkable mammals and reptiles, contemporaneous with, and in many cases similar to, the animals whose remains have made the Siwalik series so famous. The folding of these pliocene rocks, distinguished as the Irrawaddy system, into a north and south series of anticlines and synclines introduced the modern physical conditions of Burma, and determined the disposition of the great valleys of the Irrawaddy, Sittang, and Salween, whose sediments are in places burying, while in others the rivers are cutting away, the deposits produced by the great rivers which drained this area in pliocene times. The southerly extension of the Irrawaddy series is buried under the delta of the river, and possibly even under the Andaman Sea, where a longitudinal depression forms the submarine continuation of the Irrawaddy basin, and comes to the surface in the Andaman and Nicobar Islands, where, besides unfossiliferous rocks similar to those of the Chin series of the Arakan hills, there are younger soft limestones, clays, and coral sands whose precise age is not known.

Table of Tertiary Formations in Burma.

The Chin series has not, so far, yielded any fossils, and beyond Nummulites very little has been obtained from the Bassein series; but one of its fossils, Velates Schmiedeliana, Chemn., is a gastropod of great interest on account of this further evidence of its wide distribution and consequent value as a means for determining the geological horizon. From France, this gastropod ranges through Italy, Egypt, Persia, Cutch, Sind, and Western Burma, being a widely distributed inhabitant of the great Mediterranean sea which stretched as a belt across this area in early Tertiary times.

But the chief interest to the student of natural history lies in the rich molluscan fauna of the Yenangyaung series of the miocene Pegu system. These beds, well exposed in the anticline near Yenangyat (21°6’ N.; 94°51’ E.) and southwards to Minbu (20°10’ N.; 94°53’ E.), have yielded 167 species of Pelecypoda and Gastropoda, of which 30 per cent. are either identical with or closely related to species now living in the Indian Ocean, and 19 per cent. have near relations still living in the Western Pacific. While no species among this assemblage of molluscs is identical with any found in the miocene beds of Europe, 14 per cent. have their nearest relatives in the well-known eocene beds near Paris. These facts indicate an easterly migration of many of the molluscan animals during Tertiary times, the descendants of the eocene sea of Europe living in the miocene sea in India and Burma, and contributing, by further movement eastward, to the fauna now living in the seas of Japan, China, the Philippines, and Australia. The miocene beds of Burma have thus yielded forms which constitute connecting links between living forms in the Pacific and closely related extinct species which lived in the eocene sea of Europe. In some of these cases the living species are, so far, only known east of Singapore, having apparently, during the continuation of this easterly migration after miocene times, become extinct as far as the Indian Ocean is concerned, while their descendants have passed on to the Pacific ⁷.

The name Siwālik, now applied to the fringing foothills of the Himālayas in the United Provinces and the Punjab, is also used to indicate a great system of river deposits remarkable for its wealth of vertebrate fossil remains. The deposits of sands, clays, and conglomerates are essentially similar to those formed in modern times by the Himālayan rivers; and their relations to the modern alluvium show that they were produced in the same way, and were then caught up in the folding movements by which the Himālayas, rolling out as a mighty rock-wave towards the south, rose as the greatest mountain range in the world.

The most interesting and, for stratigraphical purposes, the most important among the fossil remains found in the Siwālik are those of vertebrate animals, especially of the mammalian class. The exact horizon of many of the specimens was not recorded by the collectors; and it is consequently not certain whether the apparent mixture of forms having relatives in the oligocene, miocene, and pliocene strata of Europe correctly represents the life in the jungles of the Himālayan foothills, or whether the order of succession was the same as in Europe. The general facies of the fauna, however, shows predominating pliocene affinities, on the whole newer than the fossils of the Manchhar beds in Sind, for which an upper miocene age is accepted on stratigraphical as well as palaeontological evidence.

A remarkable feature in connexion with the Siwālik vertebrate fauna is the abundance of the larger mammals, and the predominance of true ruminants over the artiodactyle ungulates. Out of sixty-four genera of mammals which have been identified among the Siwalik fossils, thirty-nine have species still living, while twenty-five are now extinct. Among the reptiles only two out of twelve genera are extinct, while all the birds and fishes whose remains have been examined belong to living genera. The impoverishment in variety of large mammals since pliocene times is a feature of considerable interest, as it is not peculiar to India and is supposed to be due to the effects of the glacial epoch. We have now but a single species of elephant in India to compare with the eleven species which lived at the foot of the Himalayas in pliocene (Siwalik) times, while the two species of Bos now living in India are all that are left of the six which formerly lived in the Siwāliks.

Very pertinent evidence as to the age of the Upper Siwālik is obtained in Burma, where the basin of the Irrawaddy contains a great system of beds, chiefly composed of yellow sands, which attain in some places a thickness of 20,000 feet and rest, with slight unconformity, on marine beds whose miocene age is placed beyond doubt by their fossil contents. Two features of special interest in connexion with these beds may be mentioned: one is the common occurrence at various horizons of pieces of silicified, exogenous wood; the other is the abundance in the two lower zones of vertebrate remains agreeing very closely in character with many of the remarkable forms which have made the Siwālik so famous.

The vertebrate fossils are, from the geological point of view, the more interesting. The specimens so far collected appear to belong to twenty-six species, of which only thirteen have been specifically determined, and eleven of these are identical with forms known in the Upper Siwālikas. The undetermined species belong to genera which are all known in the Siwālikas; so there is a sufficiently complete correspondence to justify us in regarding the Irrawaddy system as the equivalent of the Upper Siwālikas. As the stratigraphical position of the Irrawaddy system shows it to be of pliocene age, we thus have a confirmation of the conclusion which has been arrived at by a comparison of the Siwālik fauna with the pliocene fossils of other lands.

It is probable that the rocks known as the Tipam sandstones in North-eastern Assam are of the same age as the Siwālik sandstones and the Irrawaddy series, but no unquestionable fossils have been found in them.

Since pliocene times, when the Himālayas finally rose as a great barrier between India and the rest of Asia, considerable changes have taken place in the physical geology of the Peninsula.

country. In the two great areas of folding which meet the eastern and western extremities of the Himālayan range volcanic action has persisted down to recent times. On the east we have Barren Island, Narcondam, and Puppa, representing the northern extension of the line which in the region of Sumatra, Java, and the Sunda Islands has been so remarkable for its volcanic activity, while on the west, in the Irānian region of folding, we have volcanoes like Koh-i-Sultan, Koh-i-Tafdān, and Basman Koh now settling down to the solfataric stage.

Earthquakes tend generally to be more frequent in the regions of extra-peninsular India, where the rocks have been recently folded, than in the more stable Peninsula; and the areas which have recently come into prominence in this connexion are the Province of Assam and the Kangra valley in the Punjab Himālayas. In the former tract the most violent earthquake on record occurred on June 12, 1897. The known extent of the area over which the shock was distinctly felt was about 1,200,000 square miles. Within the epifocal area of 10,000 square miles, which was situated in Western Assam and Eastern Bengal, alterations have occurred in the heights and relative positions of the hills, in addition to the usual phenomena of earth-fissures, sand-eruptions, small faults, and the destruction of buildings. The violence of the movements is shown by the fracture of upright stones, indicating, in the case of short stones which were broken and overturned, a modified form of projection, while in others there was distinct rotation by the action of a vorticose motion in the ground. In the alluvial areas the effects were especially conspicuous, vibrations being noticed in the distant and detached alluvial area of Ahmadabād, though the earthquake was not noticed over the rocky ground to the east for about a hundred miles. In the Assam-Bengal alluvial area the river channels were narrowed, railway lines were bent into sharp curves and bridges compressed, while fissures and sand-vents opened in myriads. Ever since the great earthquake of 1897 the same area has been disturbed by small shocks, more than 5,000 being recorded during the following year ⁸.

The Kāngra earthquake occurred on April 4, 1905, at an early hour in the morning, in consequence of which it resulted in a great loss of human life, estimated at about 20,000. Its epifocal area lay about a curved NW.-SE. line, some 160 miles long, extending from the neighbourhood of Kāngra, through Kulū, to near Mussoorie. This line corresponds to a fault or chain of faults, which, emerging near the surface in the Kangra valley, caused the greatest destruction near the north-west end of the line, with an intensity of shock diminishing to the south-east, where the focus was deeper below the surface. The area of extensive damage to masonry buildings was only about 5,800 square miles, as compared with 150,000 square miles of similar damage in the Assam earthquake of 1897. But on account of the great depth of the focus at its south-easterly end, the waves spread out over a wide area, and serious damage was caused over about 27,000 square miles, while the shock was sensibly felt over an area nearly as large as that disturbed in 1897, being recognized as an earthquake as far west as Quetta, as far south as Surat in Bombay and False Point in Bengal, and as far east as Lakhimpur in Assam.

Within India proper there have been local changes in the relative level of land and sea within recent geological times, in some cases connected with earthquakes, as in the case of the earthquake of Cutch in 1819 when a part of the Rann was submerged, and in the Assam hills, among which alterations of level and horizontal distance were detected by measurements after the great earthquake of 1897. The Andamans and Nicobars have been isolated from the Arakan coast by submergence at a probably recent date.

On the east side of Bombay Island trees have been found imbedded in mud about 12 feet below low-water mark, while a similarly submerged forest has been described on the Tinnevelly coast. On the other hand, there is evidence to show that a part of the coast of Tinnevelly has risen and driven back the sea in the neighbourhood of Kāyal. Again, the accumulations of thick masses of old alluvium in the rocky basins of the Narbadā and Tāpti rivers indicate changes in the relative levels of the upper or eastward, and lower or westward, parts of these basins.

The clays and sandstones of uppermost pliocene or of pleistocene age which are found in the Narbadā valley have sometimes been referred to as the older alluvium of the Narbadā—a misleading expression, as, although they were formed under fresh-water conditions, they could not have been deposited in their present position in a rock basin by the Narbadā river as it now exists. They include remains of mammalian specifically, and sometimes generically, distinct from forms now living, and among them bones of a hippopotamus now represented only in Africa. The molluscs in these deposits belong to known living species of fresh-water habit; and the rocks therefore cannot be older, probably, than pleistocene, though some of the mammals are identical in species with those in the pliocene Siwalik series.

Recently, among the older alluvium of the higher part of the Godāvari valley, in the Nāsik District of Bombay, remains of extinct vertebrates have been found, including a skull of Elephas namadicus, Falc. and Caut., of exceptional size. Remains of Hippopotamus and Bos namadicus have recently been obtained in wells 80 feet below the bed of the Ganges near Allahābād. These, like the vertebrate remains found many years ago in the Jumna valley, indicate a pleistocene age.

One of the most interesting among sub-recent and recent formations is the calcareous freestone, largely used for building purposes in the Bombay Presidency, which is quarried from deposits that occur near Porbandar and other places on the Kāthiāwār coast. The rock consists largely of the remains of minute foraminifera, with small quantities of sand grains which have been transported by the wind from the sea-shore. Deposits of this nature attain thicknesses of 200 feet, showing their characteristic false-bedding, near Junāgarh, which is 30 miles from the coast; but the foraminifera are carried much farther inland, being found as far as Bikaner in the Rājputāna desert. The rounded and small shells of the foraminifera which make up such a large part of the Porbandar stone are often mistaken for oolitic grains, which also occur in the deposit ⁹.

The most important and extensive among the deposits of very young age in India are the great alluvial accumulations on the confluent plains of the Indus, Ganges, and Brahmaputra. Throughout the great Indo-Gangetic alluvial area a sandy micaceous and calcareous clay forms the prevailing material, the older alluvium being distinguished by the nodular segregations of carbonate of lime, called kankar, used largely as a source of lime and as road metal. These alluvial deposits have been penetrated by borings in two places below the sea level. The boring at Calcutta reached a depth of 481 feet without signs of either a rocky bottom or marine beds, while fragments of fresh-water shells were found as low as 380 feet below the surface, and coarse pebble beds were met throughout the lowest section of the borehole, showing that the present site of Calcutta was near the margin of the river valley which has undergone depression accompanying the accumulation of alluvial material. The boring at Lucknow extended to nearly 1,000 feet below sea-level, with no further sign of an approach to the bottom than that shown by the appearance of coarse sand near the end of the hole.

Besides the deposits formed by the great rivers on the plains of India, Assam, and Burma, there are interesting river deposits at higher levels, like those of the upper Sutlej valley in Hundes, which have yielded numerous vertebrate fossil remains; the karewa deposits of the upper Jhelum in Kashmir; the so-called tanr lands of Nepāl, in which beds of peat and phosphatic clay occur; and the similar deposits in Manipur and farther east in the Chindwin valley of Burma.

The sand-dunes of the coast of Orissa, the teris of Tinnevelly and Travancore on the coast, the accumulations of blown sand on the banks of the Kistna, Godāvari, and Cauvery, the great gravel slopes which form the dāman fringes of the Baluchistan hills, the finer loess of the plains, the extensive accumulations of the Potwar, the great desert sand deposit of Rājputāna and Sind, and the peculiar black soil or regar so widely distributed over the Deccan must be passed by with a mere mention.

The rust-coloured caps which frequently cover the rocks in moist tropical climates, and have been known for a century under the name laterite, have long been a puzzle to geologists. In its typical form this material has a vesicular or scoriaceous appearance, on which account it has been supposed by some to have a volcanic origin. Occasionally it has a polisitic structure, and is often mottled through irregular distribution of the ferric hydrate stain. There is hardly a doubt about the fact that most, and probably all, real laterites are formed by the subaerial decomposition of the rocks on which they lie, and that the peculiar structures they show are the result of molecular segregation among these products. For a long time laterite was regarded as merely a ferruginous clay, formed by the decomposition of the aluminous and ferromagnesian silicates in the rocks which are attacked by the weather; but analyses recently made show that much of the silica has been removed during the process of rock decomposition, and that the alumina, instead of being retained as a hydrous silicate such as we get in a clay, is often present as a simple hydrate of alumina, being stained red with the corresponding hydrates of iron, and mechanically mixed with other substances set free during the processes of rock-weathering. Analyses show a great variation in composition; but there is a general tendency among laterites to differ from the rocks from which they are derived by the concentration of alumina, iron oxide, and titania, while the silica, alkalies, and alkaline earths are carried away by the atmospheric waters. The fact that this peculiar form of rock-weathering is characteristic of, and practically confined to, moist tropical climates has given rise to the suggestion that the alteration of the fresh rock is effected by the action of some organism, which grows at the surface of the rock and possesses the power of breaking up the rock silicates. The separated silica is removed in solution, while the hydrated alumina and iron oxide remain behind, and, by their segregative power, cement the other products into a mass with the peculiar structures which characterize laterite ¹⁰.

Laterite may become broken off and carried to lower levels by the action of streams, and when re-deposited at lower levels may become cemented again into a compact mass by the segregative action of the hydrates, including sand-grains of quartz and other minerals. Thus there are high-level laterites, resting on the rocks at whose expense they have been formed, and low-level laterites, formed in the usual way of detrital deposits.

Laterites are not merely modern formations; several old land surfaces show traces of lateritic deposits. On the old surface of India, for instance, which was overwhelmed and covered by the Deccan trap in uppermost Cretaceous times, laterites existed, and are sometimes now exposed where the weathering agents have cut away the protecting layer of trap. At the base of the Tertiary rocks north-east of Surat, and at a few other places, are rocks having such a perfect resemblance to modern laterite that there is little or no doubt that the conditions for the formation of this peculiar material existed in early eocene times, and it is probable that many of the bauxites of Europe and America have a similar origin.

T. H. HOLLAND

BIBLIOGRAPHY

For all papers published on the geology of India before 1893, references will be found in the second edition of the official manual (Manual of the Geology of India) published in that year. Results of importance which have been obtained since, and which are noticed in this chapter, will be found more fully discussed, with references to previous literature, in the papers named below. In addition to these papers, certain conclusions and changes in nomenclature are adopted in this chapter, which have not yet been made public. These, however, have been discussed by all the officers of the Geological Survey of India, and have been accepted as representative of their views.

1. F. Kossmatt.—On the importance of the Cretaceous rocks in Southern India in estimating the geographical conditions during later Cretaceous times. Records, Geol. Surv. Ind., vol. xxviii, p. 39 (1895). The Cretaceous deposits of Pondicherry. Ibid., vol. xxx, p. 51 (1897).

2. T. H. Holland.—The Mica deposits of India. Mem. Geol. Surv. Ind., vol. xxxiv, p. 11 (1902).

3. T. H. Holland.—The Sivamalai series of Elaeolite-Syenites and Corundum-Syenites. Mem. Geol. Surv. Ind., vol. xxx, p. 169 (1901).

4. T. H. Holland.—The Charmockite series, a group of Archaean hypersthenic rocks in Peninsular India. Mem. Geol. Surv. Ind., vol. xxviii, p. 119 (1900).

5. R. Bruce Foote.—The Geology of the Bellary District, Madras Presidency. Mem. Geol. Surv. Ind., vol. xxv (1896).

6. R. D. Oldham.—Geology of the Son valley in the Rewah State and of parts of the adjoining Districts of Jabalpur and Mizapur. Mem. Geol. Surv. Ind., vol. xxxi, pt. 1 (1900).

7. K. Redlich.—The Cambrian fauna of the Eastern Salt Range. Palaeontologia Indica, New Series, vol. i, pt. 1 (1899).

8. H. H. Hayden.—The Geology of Spiti. Mem. Geol. Surv. Ind., vol. xxxvi, pt. 1 (1904).

9. C. A. McMahon and W. H. Hudleston.—Fossils from the Hindu Khoosh. Geol. Mag., decade IV, vol. ix, p. 49 (1902).

10. A. Von Krafft.—Notes on the Exotic blocks of Malla Johar in the Bhoot Mahals of Kumaon. Mem. Geol. Surv. Ind., vol. xxxii, pt. 3 (1902).

11. T. Tschernyschew.—The Upper Palaeozoic deposits of Eurasia. Rec. Geol. Surv. Ind., vol. xxxi, pt. 3 (1904).

12. F. Noetling.—Beiträge Zur Geologie der Salt Range. Neues Jahr. für Min., etc., band xiv, p. 368 (1901).

13. H. H. Hayden.—On the Geology of Tirah and the Bazar valley. Mem. Geol. Surv. Ind., vol. xxviii, pt. 1 (1898).

14. C. S. Middlemiss.—Geology of Hazara. Mem. Geol. Surv. Ind., vol. xxvi (1896).

15. J. M. Maclaren.—The Geology of North-East Assam. Rec. Geol. Surv. Ind., vol. xxxi, pt. 4 (1904).

16. A. Seward and A. S. Woodward.—On some Permo-Carboniferous (Gondwana) plants and vertebrates from Kashmir. Pal. Indica, New Series, vol. ii, pt. 1, p. 40 (1904).

17. W. T. Blanford.—On the ancient geography of Gondwana-land. Records, Geol. Surv. Ind., vol. xxix, p. 52 (1896).

18. W. T. Blanford.—The distribution of Vertebrate animals in India, Ceylon, and Burma. Phil. Trans., vol. exciv, p. 335 (1901).

19. M. Bauer.—On the Jadeite and other rocks from Tammaw in Upper Burma. Rec. Geol. Surv. Ind., vol. xxviii, p. 91 (1895).

20. E. Vredenburg.—A geological sketch of the Baluchistan desert and part of Eastern Persia. Mem. Geol. Surv. Ind., vol. xxxi, pt. 2 (1901).

21. F. Noetling.—Fauna of the Miocene beds of Burma. Pal. Ind., New Series, vol. i, pt. 3 (1901).

22. R. D. Oldham.—Report on the Great Earthquake of June 12, 1897. Mem. Geol. Surv. Ind., vol. xxix (1900). After-shocks of the Great Earthquake. Ibid., vol. xxx, pt. 1 (1900), and vol. xxxv, pt. 2 (1903).

23. J. W. Evans.—Mechanically formed Limestones from Junagarh (Kathiawar) and other localities. Quart. Journ. Geol. Soc., vol. lvi, p. 559 (1900).

24. T. H. Holland.—On the constitution, origin, and dehydration of Laterite. Geol. Mag., decade IV, vol. x, p. 59 (1903).

25. A. C. Seward.—On the association of Sigillaria and Glossopteris in South Africa. Quart. Journ. Geol. Soc., vol. liii, p. 315 (1897).