Rocks and Their Origins

THE ORIGIN OF SANDS

The essential characteristic of Sandstone is that it consists mainly of detrital grains of quartz, or occasionally of grains of chalcedonic silica (flint); these are found to scratch the steel blade of a knife, and are not affected by boiling in ordinary acids. The grains usually become cleaner in the boiling process, since the cement that has bound them together is liable to be destroyed. This cement may cause effervescence, being often formed of chemically deposited calcium carbonate.

When we consider the distribution of quartz in nature, we look to igneous and metamorphic rocks for the origin of the grains in sandstone. Quartz is one of the commonest minerals; but in granite and quartz-diorite it rarely forms more than half the bulk of the rock, felspar and mica and hornblende being its associates. Veins of quartz (quartz-rock) traverse many rocks, and become broken up into granular forms on weathering; but they are inconsiderable in comparison with the bulk of the slates or schists in which they lie. Mica-schists contribute a good deal of quartz-sand when they decay; but this is mixed with ferruginous clayey matter, and the soils produced are yellow loams.

We are easily impressed, then, by the enormous amount of denudation that was requisite to produce our existing sandstones. Though nowadays sandstones can be built up by the decay of older rocks of the same kind, the quartz must have come originally from igneous or metamorphic sources. Even in the metamorphic rocks, a large part of the quartz is probably detrital.

The microscopic characters of the quartz in sandstone commonly attest its origin. The minute liquid inclusions, with moving bubbles, that arise in the quartz of igneous and metamorphosed rocks, are easily seen in sections of sandstone. In some quartzites, these inclusions run in continuous bands from grain to grain, and have clearly arisen since the detritus was cemented. But in ordinary sandstones the inclusions in one grain have no relation to those in its neighbours. The felspars, moreover, of igneous rocks are commonly found, as rolled fragments, in sandstone. Their grains are usually whiter and duller than those of quartz, and may easily be distinguished by the naked eye.

Small gleaming plates of mica from the parent rock may accumulate with the quartz grains. The dark micas of decaying rocks, rich in iron and magnesium, together with mineral silicates of calcium, magnesium, and iron, such as the amphiboles and pyroxenes, form on hydration soft green chlorite. This mineral, in films and easily deformed flakes, at times occurs as a sort of groundwork to the coarser grains in sandstone, and colours the rock a delicate grey-green. Fine-grained sandstones of this type are difficult to distinguish from altered "greenstones," such as basaltic andesites. When the quartz grains, however, are large, as in the grits quaintly styled in old days "greywacke," they form a ready clue to the origin of the rock.

Nature sifts the products of decay so thoroughly, on any slope exposed to wind or rain, that the finest materials are carried far away, and the undecomposable quartz remains predominant. The alluvium in the upper reaches of streams is thus far more sandy than the mixed material supplied at the outset from the surrounding rocks. The more rapid flow of the water on the steeper upland slopes naturally removes the mud into the lowland.

When the detritus, still somewhat mixed, reaches a sea-shore, wave-action is rapidly effective. Before the continual wash and pounding of the water, any residual clay, and the finely comminuted portion of the quartz, are carried down the coastal slope. The colour of the sea after storms is sufficient evidence of the work that it performs. Beaches, then, arrive at a great similarity of type. The inviting yellow sands, formed of comparatively coarse material, occur alike off shores formed of chalk, slate, granite, or boulder-clay.

From the beginning of sedimentation, sands have thus tended to accumulate, and to become cemented into sandstones. These rocks, in turn uplifted and exposed, have yielded other sandstones. Since coarse sand does not travel far from the region where it is washed out of the parent rock, a thick mass of sandstone extending over many square miles may waste away, and yet become perpetuated in the district. Sandiness thus begets sandiness, and the physical conditions due to the presence of sandstone may prevail through long geological epochs (Fig. 5).

Of course, a submergence beneath the sea may change all this in a brief time; but wrinklings of the crust, raising the sandstones into severer atmospheric levels, may only accelerate their decay and render the surrounding lands more sandy.

THE CEMENTING OF SANDS

The cement of sandstones is very varied. On our modern coasts, springs draining from a limestone land, or even running through banks of broken shells, will deposit calcite in the interstices of the beach, until slabs and shelves of conglomerate and sandstone arise in defiance of the waves. On coasts where calcium bicarbonate is abundant, it may be precipitated by any cause that diminishes its solvent. Mere evaporation, and the escape of carbon dioxide from the water as it is scattered into spray, lead to the deposition of a cement between the grains of sand. As Linck[6] shows, calcite is thus laid down in temperate waters, while aragonite forms fibrous crystals between the detrital fragments on the flanks of tropic isles. Aragonite may also arise from the action of ammonium carbonate or sodium carbonate on calcium sulphate or calcium chloride in sea-water. Sands thus become cemented by one or other form of calcium carbonate. They include, moreover, calcareous algæ, foraminifera, and fragments of coral and sea-shells.

Fossil shells are usually represented in older sandstones by mere external and internal moulds. The texture of the rock allows of their being dissolved in percolating waters, while in clays belonging to the same geological series they may be exquisitely preserved.

In shallows, and especially in lakes, where soluble salts of iron become readily oxidised, brown iron rust, the mineral limonite, is continually forming at the surface and sinking to the bottom, where it firmly cements the sand. A group of bacteria[29] extracts iron in this form from the water of freshwater lakes and swamps, and greatly aids in its accumulation. Though a red colour may appear also in marine deposits, masses of red and purple conglomerates and sandstones may reasonably be assigned a freshwater origin. Such rocks are usually found to be devoid of marine fossils, and they often contain traces of land plants.

Barytes (barium sulphate), which sometimes occurs in veins simulating those of calcite, is an occasional cement of sandstone, evidently arising from subterranean waters.

Bands of flint (chert) occur in certain sandstones, such as the Hythe Beds of the English Lower Greensand Series. These are due to the cementing of certain layers by chalcedonic silica, and the source of this silica is seen in the hollow moulds of sponge-spicules, and the glauconitic casts of their canals, that commonly remain. G. J. Hinde[30] shows that in the Cretaceous examples, as in so many other flints, the majority of the spicules are of the tetractinellid type.

Under arid conditions, as in parts of Africa, loose superficial sands may become cemented by calcium carbonate, or even by silica, brought up in water rising by capillary action from below.

The sand-dunes of the coast of our own islands, which cannot remain wet for long, become in places toughened by a deposit of calcite derived from the abundant shells of land-snails. In the Cape of Good Hope[31] the dunes, as A. W. Rogers states, are converted by invasions of calcium carbonate, "into hard rock through a distance of many feet from the surface, and where repeatedly wetted and dried, as happens where the sea has encroached upon old dunes, the rock becomes intensely hard and weathers with a peculiarly jagged surface." The General Post Office and the South African Museum in Cape Town are mainly constructed of this recently consolidated rock.

The modern sandstones cemented by silica are still more interesting. In the Cape of Good Hope, and notably in the Kalahari desert, they form the intensely hard rock known as Quartzite[32]. The cementing material is true quartz, which sometimes deposits itself in bipyramidal crystals about the grains of sand. The molecules of such crystals are arranged in continuity with the grouping of those in the original detrital grain, as is proved in thin sections under the microscope by the optical continuity of the quartz of the grain and of its coating. As silica continues to be deposited, the coatings interlock, and the rock passes into true quartzite. It is now often difficult to detect the outline of the original grains. Such superficial quartzites may be ten feet thick at most, with uncemented sand below. Rogers suggests that the cementing process may have originated in shallow pools; but it has obvious analogies with that which forms iron-pans and superficial masses of calcium carbonate in regions where capillary waters are subject to prolonged evaporation. H. G. Lyons[33] has attributed the cementing of parts of the Nubian Sandstone in the desert of Lower Egypt to the silica set free by the alteration of the felspars in the rock. This change, he suggests, was accelerated by the infiltration of sodium carbonate of local origin. Fossil trees in these strata have been replaced by silica. A further example is recorded by Armitage[34] from Victoria, where friable ferruginous Cainozoic sands have been converted into quartzite. This type of rock, the hardest known, and associated in our minds with high antiquity and metamorphic action, proves, then, to be in process of construction at the surface at the present day.

The observations of Rogers show that quartz and not mere chalcedony is deposited on the grains of sand. The "crystalline sandstones" of Permian and Triassic age in England may, then, have acquired their remarkable characters at the actual epoch of their accumulation. This is rendered the more probable by the recognised occurrence of arid conditions, at any rate seasonally, when the strata in question were laid down.

These English "crystalline sandstones" were described by H. C. Sorby[35], who showed that the quartz deposited on the detrital grains was in optical continuity with that of the grains themselves. J. A. Phillips[36] regarded this quartz as crystallised out during the kaolinisation of felspars. The phenomena of laterisation, however, give us a further suggestion as to the origin of the secondary silica. It is now well known that tropical processes of weathering, with alternations of wet and dry seasons, allow alumina to be set free from combination with silica, "lateritic" crusts thus arising on a great variety of rocks. The felspars of a sandstone may, under such conditions, become laterised rather than kaolinised, aluminium hydrate being left, and the silica passing into solution and appearing again in certain layers as cementing quartz. The almost complete disappearance of silica from the more advanced laterites shows that it has been carried away elsewhere, and the cement of quartzite may thus be derived from rocks at a considerable distance. Just, however, as the destruction of siliceous sponge-spicules implies the formation of flint, so laterisation implies silicification as a complementary process.

The fact that secondary quartz in quartzite often arises in the rock itself is shown by the frequency of quartz-veins in quartzites, while they are almost absent from associated slates or schists. Hence it appears that a removal of silica goes on at some points, leading to an infilling of all the cracks and interstices at another.

It is clear, then, that sandstones, according to the mode in which they have been affected by percolating waters, may vary from the crumbling uncemented condition, known as Sand-rock, to that hardest and most resisting of rocks, quartzite. The permeability of sandstone is responsible for a wide variety of types.

THE SAND-GRAINS OF SANDSTONE

Sandstones are originally permeable by water, not because they possess a high percentage of pore-space, or "porosity," but because the pores between the grains are large. Water can thus move easily by gravitation through the mass. The capillary rise or spread of water is greatest in materials of very fine grain, though in these it may be extremely slow. For the most effective rise of water against gravity by capillary pull, a large proportion of particles about ·02 mm. in diameter should be present. Sand-grains, however, often measure ·5 mm. in diameter, and the fine mud or highly comminuted sand between the coarser matter is the cause of the spread of water through the mass when the supply comes from a subterranean water-table. Rain, however, is of course readily absorbed. It disappears so rapidly on some barren sandstone areas, coated as they are by loose sandy soils, that vegetation cannot make a start, even where water is supplied.

Daubrée, Sorby, and others have studied the characters of sand-grains, and it has been pointed out, that agitated water buoys apart and carries forward by flotation grains with a diameter of ·1 mm. or less. Hence coarser grains may become rounded like pebbles, by friction on the bottom of a stream; but small ones remain angular throughout geological periods, and even when transferred from one sandstone to another. When their surfaces have been cleaned by boiling in hydrochloric acid, the sharpness and irregularity of the quartz grains is strikingly apparent.

Mingled with these grains, in addition to the minerals previously mentioned, many interesting crystals appear that have become concentrated in the natural washing processes. Minute colourless zircons and brown rutiles, derived from granite, have collected, owing to their high specific gravity, in certain sands. Magnetite and ilmenite may darken the mass; monazite and thorite, which are sought after for their constituents cerium and thorium, become similarly selected in alluvial hollows, owing to their density of 5. Whatever gathers thus in sands may become preserved in sandstones, and the study of thin sections of the latter under the microscope is fruitful in suggestions as to their origin.

Some sandstones are remarkable for their highly rounded and almost spherical grains. J. A. Phillips[38] compared these with the wind-worn grains of deserts, which assume similar forms and a considerable polish. Large quantities of sand are carried from arid lands into rivers, into lakes, or into the sea, and hence well rounded grains, in bedded rocks, and even in marine sandstones, may have had a desert origin. J. W. Judd, when examining the deposits of Lower Egypt for the Royal Society, commented on the extreme freshness of the felspathic particles in sands accumulating in rainless areas, and recent observations on the soils of semi-arid districts show their comparative poverty in clay. Enough has been said to indicate the variety of geographical considerations that may arise from the examination of beds of sandstone. The grains often prove, especially in the coarser types, to be fragments of rocks rather than isolated minerals, and thus furnish a picture of the materials that formed the surface exposed to denudation.

The sandstones of finest grain may be found in beds deposited almost on the limits of sedimentation from the land, where they are interlocked with material of truly pelagic origin. Marine muds often contain a high percentage of comminuted quartz, and the study of shales and slates of ancient days shows how this almost indestructible mineral finds its way into beds that might easily be classified as clays[41].

SOME CHARACTERS OF SANDSTONE

Earth-stresses and shrinkage give rise to joints in sandstone, which may not be so clean and sheer as those in limestone, but which affect even the softer forms. Cemented sand-dunes of modern date tend to break away along vertical planes. Firmer sandstones give rise to stepped table-lands and "edges," and the resistance of many types to atmospheric decay renders their stratified structure strongly apparent. Small intervals in the process of deposition, or slight changes in the coarseness of the sand brought down by currents, give rise to laminated and flaggy types. Where a broad shore has been exposed between tide-marks, the drying and compacting of the surface before the next layer is laid down enables the latter to take a mould of the inequalities of that below. Ripple-marks, sun-cracks, rain-prints, and the footmarks of animals, are often preserved in this manner. Where the shore is subsiding, they may persist through hundreds of feet of strata.

Naturally, the best examples of these casts, and of the original structure in the underlying bed, occur where a little mud has been laid down over the sandy flat. Clay by itself, if damp, does not retain the impressions sufficiently long, and, when once thoroughly dried, it crumbles when the next water overflows it. But a foundation of firm sand with a thin mud-layer on its surface, as may be recognised in some Triassic deposits, furnishes excellent records of local weather or of the movements of errant animals. On the flat shores of lakes in a semi-arid climate, the water may retreat for miles, and return, perhaps months afterwards, when rains in the hills have given it a new burden of detritus. Under such conditions, broad sun-cracked flats may be preserved, with perhaps some plant-remains between successive layers[938].

The castings and tracks of worms, and the tubes of boring species, which are sometimes infilled by sand of a different colour, are common in sandstones of all ages.

SILICEOUS CONGLOMERATES

The deposits of wave-swept beaches leave us Conglomerates formed of various types of pebbles, among which quartz-rock and quartzite naturally predominate. In some cases the pebbles are ready formed when they reach their resting-place. They come rolling out from lateral torrents into the quieter waters of a main valley, as may be seen in summer in the broad pebble-banks of the north Italian streams. Thence they are washed by occasional floods into the great confluent deltas that constitute the upper part of an alluvial plain, or into lake-basins, where they promptly settle along the shore. But few such pebbles, except from pre-existing conglomerates or gravels on the shore-line, actually reach the sea. The rolled stones upon sea-beaches are mostly the products of marine action on the spot. While the fine sand-grains go seaward almost unharmed, the detrital stones, offering far less surface in proportion to their mass, strike on their neighbours as every wave shifts them on the beach, and soon assume a rounded form.

The conglomerates ultimately consolidated may reveal stratification only by the general arrangement of their pebbles. These can rarely be spheres, since they are not as a rule turned over, but are pushed this way and that until they acquire a flat ellipsoidal shape. They lie with their flatter sides in planes parallel to one another. Generally, however, alternations of coarser and finer beds mark out the stratification even in conglomerates.

The sands of deserts include abundant stones and blocks of rock, and the loose material becomes, moreover, sifted by the wind. True desert sands may accumulate at one point, the very finest loamy material may be carried away still farther to form fields of fertile löss, and a rock-desert, formed of stones resting on bare surfaces, may remain in large areas of the arid region. The loose stones here assume a characteristic shape, and have been known under the German name of Dreikanter. They are fairly flat below, and are cut away above by the drifting sand into a form resembling a gable roof dipping at both ends. Their surfaces are characteristically etched.

Dreikanter have been found in beds that were formerly ascribed to deposition on the shores of lakes, and it must now be borne in mind that continued attrition by drifting sand affects mixed detritus on a land surface much as the wash of waters does upon a beach. Certain materials are cut away more rapidly than others, and the residue assumes a more and more quartzose type. In this way, sandstones, and conglomerates in which fragments of quartzite and vein-quartz predominate over other constituents, may arise as æolian beaches on dry land.

SANDSTONE AND THE LAND-SURFACE

The permeability of sandstone has already been referred to. The surface offered by it is typically dry, and the soil, consisting mainly of grains of siliceous sand, can neither retain the rain that falls nor draw up water from below. The idea that trees can flourish on sandstone soils because they require nothing from the soil itself is of course erroneous. They depend to a large extent upon the materials set free by the decay of certain grains, or of the cement of the underlying sandstone. In proportion as the sandstone is impure, that is, the more its constituents deviate from pure quartz, the more chance there is that it will provide a fertile soil.

On the whole, however, areas of siliceous conglomerate and sandstone are given over, even in temperate climates, to forest and heather. Where the sandstone is still in the sand-rock state, bare patches are likely to appear even in the heath that has grown across it, and from these the wind carries away shifting sands.

Everyone familiar with the Carboniferous areas of the English midlands will realise the influence of hard grit and sandstone in forming "edges" across the country. The contrast between these escarpments and the slopes of crumbling shale that often underlie them gives diversity to the scenery of Yoredale and the Peak. The more yielding sandstones of Cretaceous age round about the Weald, or at the foot of the Chiltern Hills near Woburn, form rounded hills, mostly clad with woods of coniferous trees. In Surrey, unpaved cart-tracks, used for centuries, have cut gullies in the unconsolidated Folkestone Sands.

The underlying Hythe Beds, however, stand out between Reigate and Guildford as a bold escarpment, and it is interesting to reflect that this fine feature of south-eastern England is probably due to the chert which the beds contain (see p. 62). The local growth of siliceous sponges in a Lower Cretaceous sea enables Leith Hill in our days to dominate even the arch of Ashdown Forest, where another unfilled sandstone area rises in the centre of the Weald.

The sands of Bagshot Heath, and numerous similar areas in the Paris Basin, show how impossible it is to cultivate such strata, even near the best of markets. The flint gravels that cover much of the upland in the New Forest may also be borne in mind, as presenting the worst features of highly siliceous lands.

In a semi-arid climate, or one with only seasonal rains, the processes by which sandstone begets sandstone tend to develop desert wastes. The soils produced by weathering do not cake together, and are carried away by wind during the drier months. The bare rock appears over broad surfaces, just as it does in storm-swept limestone areas, and any hollow where shelter is afforded tends to become filled with sand (see Fig. 5).

The hummocky and extremely irregular surface of some of our Silurian areas, such as parts of the Southern Uplands of Scotland and the hard-won farmlands of Down and eastern Monaghan, is due to the presence of resisting sandstones among the shales. These sandstones, passing into true grits, are repeatedly folded, and their upturned edges have resisted even the passage of glacier-ice. They jut out along the crests of ridges, and even the smaller beds furnish angular fragments to the soils.

Far wilder scenery is formed by the more continuous sandstone masses of the Harlech Beds in western Wales, which are grits so firmly cemented that the rock breaks across the quartz-grains. Much of the Old Red Sandstone is of equally hard quality (Fig. 6). Its purple or grey conglomerates, the pebbles of which are quartzite in a quartz cement, form bare and rugged masses in the Great Glen south-west of Inverness, and are responsible in Kerry for some of the wildest rock-scenery in the British Isles. Variations in coarseness allow of the development of a marked stratification on the weathered mountain sides, and differential erosion of the beds has taken place where ice has pressed against them. Even on precipices, grassy ledges may occur, marking bands of sandstone or shale in the conglomeratic mass.

The red sandstones and conglomerates that form huge outstanding bluffs from Applecross to the north of Sutherland represent the denudation of a pre-Cambrian mountain region. These Torridon Sandstones cover a very irregular surface of old gneiss, with which their almost level strata are in striking contrast. P. Lake[39] has compared them with the deposits styled dasht in Baluchistan and Afghanistan, which similarly fill up valleys and cover hills, as products of extensive and rapid denudation. There is much, indeed, to suggest that the Torridon Sandstone, some 10,000 feet in thickness, was accumulated in a dry country on a continental surface, with the aid of floods during occasional rainy seasons.

Quartzite, which fractures into small angular blocks under earth stresses, yields an intractable surface of bare rock and taluses of shifting stones. The latter sometimes crumble down into white sand, which provides some basis for the growth of heather. The numerous joints, independent of the bedding-planes, cause the rock to break up almost equally on any exposed slope, and the crests of quartzite hills become typically converted into cones (Fig. 7). Viewed from a distance, the white taluses, streaming down evenly from the crests, resemble caps of snow.

The absence of soil and the smoothness of weathered surfaces render quartzite mountains hard to climb. The uniform cementing of the rock leaves the bedding with little influence on the surface-features, and rock-ledges and shelves are rare. The traveller ascends over taluses of angular and obstinate blocks towards slippery and inhospitable domes. But the wildness of the scenery will be his sure reward. It is of interest to reflect that the material of these bold outstanding mountains may in certain cases have originated, in all its hardness, in the levels of a sun-parched plain.