I HAVE already explained that all rocks are divided into two great classes, those of aqueous and those of igneous origin; and I have shown how aqueous rocks may generally be known by their stratification and by the circumstance that a great many of them contain relics of marine and freshwater life, in the shape of fossil shells, fish-bones, and other kinds of organic remains. The materials also of which these beds are composed generally show signs of having been in water, being rounded by the action of the waves of the sea, or by the running waters of rivers.

The other kinds of rocks, termed igneous, occasionally are associated in different localities with the formations named in the foregoing table. For instance, there are no volcanic rocks in Wales associated with the Carboniferous and Old Red Sandstone strata, while there are in Scotland, and true contemporaneous volcanic rocks are intercalated with the Lower Silurian rocks of Wales and Cumberland, while there are none associated with the equivalent strata in Scotland. Some of these contemporaneous igneous rocks consist of beds of volcanic ashes, others of old lavas, others of masses of matter which were intruded among the strata from below. Rocks that have been melted are known

[Igneous Rocks. 39]

to be igneous by their crystalline, slaggy, scoriaceous, vesicular, or columnar structures, and also by the effects

FIG. 9.

1. Dyke with veins.
2. Overflow of basaltic lava.
3. Altered strata at junction.
4. Unaltered sandstone and shale

they have produced on the strata  with which they are associated. Shales, sandstones, &c., are often hardened, bleached, and even vitrified at the points of junction with greenstone, basaltic,
and felspathic dykes, or old lava beds (fig. 9), and the same kind  of alteration takes place on a greater scale when large masses of igneous rocks have been intruded among the strata.

Then by comparing volcanic rocks of old date with those of modern origin, we are able to decide with perfect truth, that rocks which were melted long before the human race appeared upon the world are yet of truly igneous origin.

Changes of a more general character are especially marked in cases where granite, syenite, felspar and other porphyries and their allies, are associated with stratified deposits. Their igneous affinities are known by their crystalline structure, their modes of occurrence, and the effects they produce on the strata. Granite is composed of crystals of quartz, felspar, and mica; and syenite, according to old nomenclatures, of quartz, felspar, and hornblende. They often send veins or dykes into stratified rocks with which they are in contact, as in figs. 10 and 11, and frequently all along the line of junction, and often at great distances from it, alterations of the strata of an extreme character (metamorphism) are common. One marked distinction between granitic and volcanic and ordinary trap rocks is, that though injected veins of granite are common, granitic

[40 Granite.]

rocks never rose to the surface in a melted state, and overflowed like lava streams. This and their frequently

FIG. 10.

A, vein of granite; B, gneissic contorted mica-schist. The ramifying white spaces are white quartz. Milldam Goatfell, Brodick, Arran.

largely crystalline structure, together with peculiarities of crystallisation showing the presence of moisture, and

FIG. 11.

1. Granitic mass with injected veins among gneissic rocks.
2. Gneiss, metamorphosed strata.

also the transformations effected on the adjoining strata, prove the granitic rocks to have cooled and consolidated deep beneath the surface.

A third division, or sub-class, is known as metamorphic rocks. All strata as they assume a solid form become to a certain extent altered; for originally they were loose sediments of mud, sand, gravel, carbonate of lime, or mixtures of these. When these were accumulated, bed upon bed, till thousands of feet were piled one upon

[Metamorphic Rocks. 41]

the other, then, by intense and long-continued pressure, heat, and chemical changes that took place in consequence of infiltrations among the strata themselves, by degrees they became changed into hard masses, consisting of shale, sandstone, conglomerate, or limestone, as the case may be. But these have not always remained in the condition in which they were originally consolidated, for it has often happened that disturbances of a powerful kind took place, and strata originally flat have been bent into every possible curve.

For long it was the fashion to attribute most of the disturbances that the outer part of the earth has undergone to the intrusion of igneous rocks. The inclined positions of beds, the contortions of stratified formations in mountain chains, and even the existence of important faults—in fact, disturbance of strata generally—were apt to be referred to direct igneous action operating from below. Granite and its allies, from the time of Hutton, were always, without exception, included in the ordinary list of igneous rocks, and some writers of deserved reputation still do so. In connection with this subject, gneiss, and other kinds of metamorphic rocks were, and by some are still, supposed to be exclusively the effect of the direct intrusion of granite among previously unaltered strata.

As a general rule highly metamorphosed rocks occur in regions where the strata have been greatly disturbed. Such rocks, when the metamorphism is extreme, consist of gneiss, which may be micaceous, hornblendic, or chloritic; and of mica-schist, chlorite-slate, talc-slate, hornblende-rock, crystalline limestone, quartz-rock, and a number of others, which it is not necessary for my present purpose to name. In Scotland, Ireland, Norway, Canada, &c., limestones, calcareous

[42 Granite and]

sandstones, and sandstones, as they approach granites, lose their (sometimes fossiliferous) characters, and become changed into crystalline limestones, serpentine, &c., and quartz rock. In other cases gradual changes of a different kind are observed in slaty and schistose rocks as they approach granites. Clay-slates are simply clays consolidated by pressure, often affected by cleavage, and sometimes chemically altered. Approaching granites ordinary slates often assume a foliated structure by the development of distinct mineral layers of quartz, felspar, and mica. This is gneiss. Analyse some kinds of mica-slate, gneiss, and common sandy clay, and their average composition will not differ more than three clays, three pieces of gneiss, and three bits of granite often do from each other.

Granite is sometimes merely gneiss still further metamorphosed by heat in the presence of moisture; and, though this is not the popular notion, I have long held it, and some other geologists share this opinion. When slate is changed to gneiss, there is no development of materials which were previously absent, but simply a re-arrangement of its constituents, according to their chemical affinities, in rudely crystalline layers, which seem in gneiss to have found facilities for their development in pre-existing planes, whether of bedding or of cleavage; or, in other words, if the rocks be uncleaved when metamorphism occurs, the foliated planes show a tendency to coincide with those of bedding; but if intense cleavage has preceded, the foliation will generally tend to follow the planes of cleavage. Furthermore, in gneissic rocks, garnets, schorl, staurolite and staurotide, hornblende, and other minerals are frequent in some localities, especially near and in contact with granite. All the chief materials of these are

[Gneiss. 43]

such as occur in the unaltered rock, and the chemical action (brought into activity by heat and moisture) which induced their development, may perhaps in some cases have been assisted by sublimations proceeding from melted matter below. The intensity in many countries of these metamorphisms, extending over many thousands of square miles (as in Scotland, Norway and Sweden, and Canada), and through rocks thousands of feet in thickness, proves that it was the result of a long-continued process, taking place probably in all cases at considerable depths. The whole has then been upheaved and disturbed, often many times, and after denudation the gneissic and. the more thoroughly metamorphosed and sometimes intrusive granitic rocks were at length exposed at the surface.

Some of the metamorphic rocks, which I have to explain, have been highly disturbed, and in the north occupy about one-half of Scotland. Most of this area includes, and lies north-west of, the Grampian mountains; and I must endeavour to explain by what processes metamorphism of rocks has taken place, not in detail, but simply in such a manner as to give a general idea of the subject.

I have already said that typical gneiss consists of irregular lamine of mica, quartz, and felspar, and it frequently happens that they are bent, or rather minutely folded, in a great number of convolutions, so small, that in a few yards of gneiss they may sometimes be counted by the hundred. All these metamorphic rocks and granite, were by the old geologists called Primary or Primitive strata, and were considered to have been formed in the earliest stages of the world's history, because in those countries that were first geologically described, they were supposed to lie always

[44 Gneiss, Old Theory.]

at the base of all the ordinary strata. From the peculiarity of the minute contortions in the gneissic rocks, a theory now known to be erroneous was developed, which was this:

It is frequently found that granite and granitic rocks are intimately associated with gneiss. Thus we often find masses and veins of granite, with gneiss upon their flanks bent in a number of small wavy folds or contortions. Granite is a crystalline rock, composed of felspar, quartz, and mica, and the old theory (so far true) was that the world at one time was in a state of perfect igneous fusion; but by and by, when it began to cool, the materials arranged themselves as distinct minerals, according to their different chemical affinities, and consolidated as granite. The great globe was thus composed entirely of granite at the surface ; and by and by, as cooling still progressed, and water, by condensation, attempted to settle on the surface which still remained intensely heated, the water could not lie upon it, for it was constantly being evaporated into the atmosphere; but when the cooling became more decided, and consolidation had fairly been established, then water was able to settle on the surface of the heated granite. But as yet it could not settle quietly like the present sea: for by reason of strong radiating heat, all the sea was supposed to be kept in a boiling state, playing upon the granite hills that rose above its surface. The detritus thus worn from the granite by the waves of this primitive sea was spread over its bottom; and because the sea was boiling, the sediment did not settle down in the form of regular layers, but became twisted and contorted in the manner common in gneiss. All gneiss, therefore, was conceived to be the original primitive stratified rock of the world.

[contortion and Metamorphism. 45]

Subsequent research has shown that this theory will not hold; for this, among other reasons, that we now know gneissic rocks of almost all ages in the geological scale. Thus in Scotland the gneissic rocks are of Laurentian and Silurian age; in Devon and Cornwall we have gneiss both of so-called Devonian and Carboniferous ages. In the Andes there are gneissic rocks of the age of the Chalk, and in the Alps of the New Red, Liassic, Oolitic, and Cretaceous series; and in 1862 I saw in the Alps an imperfect gneiss of Eocene date pierced by granite veins, these strata being of the age of some of the soft and often almost horizontal strata of the London and Hampshire basins. It is therefore now perfectly well known to geologists that the term Primitive, as applied to gneiss, is no longer tenable; and the old theory has been abandoned.

I have stated that regions occupied by metamorphic rocks are apt to be much contorted. There seems, in fact, to be an intimate connection between excessive disturbance of strata and metamorphism. But by what means were masses of strata many thousands of feet thick bent and contorted, and often raised high into the air, so as to produce existing scenic results by affording matter for air and water to work upon? Not by igneous pressure from below raising the rocks, for that would stretch instead of crumpling strata, in the manner in which we find them in the Alps, Norway and the Highlands, or in less degree in Wales and Cumberland; but rather because of the radiation from the earth of heat into space, gradually producing a shrinkage of the earth's crust, which, here and there giving way, became crumpled along lines more or less irregular, producing partial upheavals, even though the absolute bulk of the globe was diminishing by cooling

[46 Shrinkage and]

(figs. 3, 12, and 57). This, according to the theory long ago proposed by Elie de Beaumont, and adopted by De la Beche in his 'Researches in Theoretical Geology,' is the origin of mountain chains. After water took its place on the earth, by such processes land was again and again raised within the influence of atmospheric disintegration, and rain, rivers, and the sea, acting on it, were enabled to distribute the materials of sedimentary strata. Such disturbances of strata have been going on through all known geological time, and I firmly believe are still in progress.

Such shrinkage and crumpling, where it has been most intense and on the greatest scale, is often (where I know it) accompanied by the appearance of gneissic or other metamorphic rocks, and often of granite or its allies.

The oldest rock in the British Islands is gneiss, but that originally was doubtless a common stratified formation of some kind or other. In fact, as far as the history told by the rocks themselves informs us, we cannot get at their beginning, for all strata have been made from the waste of rocks that existed before; and therefore the oldest stratified rocks, whether metamorphosed or not, have a derivative origin.

I must now briefly endeavour to give an idea of the theory of metamorphism. The simplest kind is of that nature mentioned in Chapter I. namely, when melted matter has been forced through or overflows a stratified rock, and remaining for a time in a melted state, an alteration of the stratified rock in immediate contact with it takes place. Thus sandstone may, by that process, become converted into quartz-rock, which is no longer hewable, like ordinary sandstone, but breaks with a hard and splintery fracture. Here then rocks

[Metamorphism. 47]

have been changed in character for a short distance from the agent that has been employed in effecting that minor kind of metamorphism (figs. 4 and 9).

On a much larger scale, the phenomena we meet with in a truly metamorphic region are as follows. In the midst of a tract of mica-schist, gneiss, or other altered rocks, a boss of granite (or one of its allies) rises, like those for instance of Dartmoor and Cornwall or of the north end of the Island of Arran. At a distance from the granite the beds may consist, perhaps, of unaltered shale, or of slate, sandstone, and limestone. As we approach the granite, the limestones become crystalline, and often lose all traces of their fossils; the sandstones harden and pass into quartz-rocks, and the shales or slates, or sandy beds and shales, lose their ordinary bedded texture, and pass by degrees into mica-schist, or perhaps gneiss, in which we find rudely alternating laminæ of quartz, felspar, and mica, often arranged in gnarled or wavy lines (foliation, figs. 10 and 11). As we approach the granite still more closely, we find possibly that, in addition to the layers of mica, quartz, and felspar, distinct crystals, such as garnets, staurolites, schorl, &c, are developed near the points of contact, both in the gneissic rock and in the granite itself.

It is not necessary for my argument that I should describe these minerals. It is sufficient at present to state the fact that such minerals are developed under these circumstances, and this is due to the influence of metamorphism.

Furthermore in some cases, as in the Lauretian rocks of Canada, great thicknesses of  interstratified gneiss are so crystalline that, when a hand specimen or even a small part of the country is examined, they

[48 Analyses of Rocks.]

might seem to be truly granitic; but when the detailed geology of the country has been worked out, they are found to follow all the great anticlinal and synclinal folds of metamorphosed strata that have also in a minor way been intensely contorted. The same is the case in parts of the Alps.

I have already stated that if we chemically analyse a series of specimens of clays, shales, and slates, often more or less sandy, together with various gneissic rocks and granites, it is remarkable how closely the quantities of their ultimate constituents, in many cases, approach to each other. They are never identical, while yet the resemblance is close, as close indeed as it may be in two specimens of the same kind of sandy shale or slate. In all of them silica would form by far the largest proportion, say from 60 to 70 per cent.; alumina would come next, and then other substances, such as lime, soda, potash, iron, &c., would be found in smaller varying proportions; and what I now wish to express is, that the distinct minerals developed in the gneiss, such as felspar, mica, garnets, &c., were not new substances introduced into the rock, by contact with granite, or by any other process, but were all developed under the influence of metamorphism from materials that previously existed in the strata before their metamorphism began, aided by hydrothermal action due to the presence of heated alkaline waters deep beneath the surface of the earth. Through some process, in which heat played a large part, the rock having been softened, and water—present in most rocks underground—having been diffused throughout the mass and heated, chemical action was set up, and the substances that composed the shale or slate, often mingled with silicious sandy material, were enabled more or less to re-arrange

[Mountain Chains. 49]

themselves according to their chemical affinities, and distinct mineral materials were developed in layers from elements that were in the original rock.

I have stated that to produce this kind of metamorphism, heat aided by water is necessary, so as to allow of internal movements in the rocks by the softening of their materials, without which I do not see how complete re-arrangement of matter accompanied by crystallisation could take place; and though it has always been easy to form theories on the subject, yet so little is known with precision about the interior of the earth beyond a few thousand feet in depth, that how to obtain the required heat is a difficulty.

From astronomical considerations it is believed by many persons that the earth has been condensed from a nebulous fluid, and passing into an intensely heated melted condition, by radiation into space at length cooled so far, that consolidation commenced at the surface, and by degrees that surface has gradually been thickening and overlies a melted nucleus within.

As the earth cooled and consequently gradually shrunk in size, the hardened crust, in its efforts to accommodate itself to the diminishing bulk of the cooling mass within, became in places crumpled again and again. Hence the upheaval of mountain chains and disturbances of different dates, which have affected strata of almost all geological ages.1

Reasoning on these disturbances, we know that strata which were originally deposited horizontally have often

1 This theory is not universally received, and has been variously developed by different authors, but it would be quite beyond my present purpose to discuss the subject in detail, and, as far as I know, the hypothesis proposed by Elie de Beaumont seems best to explain the phenomena exhibited by the outside of the earth.

[50 Internal Heat of the Earth]

descended thousands of feet towards the centre of the earth, by gradual sinking of the seabottom, and the simultaneous piling up of newer strata upon them. The layer that is formed today beneath the water forms the actual sea-bottom; but neither the land nor the seabottom are steady. The land is in places slowly descending beneath the sea, and sea-bottoms are themselves descending also. It has frequently happened, therefore, that for a long period a steady descent over a given area has taken place, and simultaneously with this many thousands of feet of strata have by degrees accumulated bed upon bed, as for example in the Pacific Ocean in the region of modern atolls and barrier coral reefs.

As we descend into the earth the temperature rises, whence, in the main, the theory of central heat has been derived. In our latitude heat increases about 1° for every sixty feet, and the temperature therefore, at so great a depth as 30,000 feet, to which it could be shown some strata have sunk, may at present be about 500°. Furthermore, strata that were deposited horizontally have been frequently disturbed and thrown into rapid contortions, or into great sweeping curves ; and by this means especially, strata which once were at the surface have often been thrown twenty, thirty, or forty thousand feet downwards, and therefore more within the influence of internal heat, as, for instance, in the bed marked * fig. 12, which may be supposed to represent a large tract of country. I do not wish it to be understood that the globe is entirely filled with melted matter-that is a question still in doubt; but were this book specially devoted to general questions of theoretical geology, I think I could prove, that the heat in the interior of the globe in places sometimes

[and Metamorphism. 51]

apparently capriciously eats its way towards the surface by the hydrothermal fusion or alteration of parts of the earth's crust, in a manner not immediately connected with the more superficial phenomena of volcanic action

FIG. 12.

—and for this, among other reasons, it may happen that strata which are contorted, have in places been brought within the direct and powerful influence of great internal heat. Under some such circumstances, we can easily understand how stratified rocks may have been so highly heated that they were actually softened; and most rocks being moist (because water that falls upon the surface often percolates to unknown depths), chemical actions were set going, resulting in a rearrangement of the substances which composed the sedimentary rock. Thus certain strata, essentially composed of silica and silicates of alumina, and alkalies such as soda and potash, may have become changed into crystalline gneiss.

This theory of re-arrangement leads me to another question—connected with, but not quite essential to my argument, as far as relates to physical geography—viz., What is the origin of granite, which in most manuals is only classed as an igneous rock? For my part, with some other geologists, I believe that in one sense it is an igneous rock—that is to say, much of it has often been completely fused. But in another sense

[52 Origin of Granite.]

it is often a metamorphic rock, because it is sometimes impossible to draw any definite line between gneiss and granite, for they pass into each other by insensible gradations. About halfway up the Matterhorn in the Alps, among the largely-contorted beds, a thick stratum occurs, one end of which is true gneiss, on the western side of the mountain, which striking towards the eastern cliff, gradually gets more and more crystalline till at length it passes into true granite. On the largest scale, both in Canada and in the Alps, I have frequently seen varieties of gneissic rocks regularly interbedded with less altered strata, the gneiss being so crystalline, that in a hand specimen it is impossible to distinguish it from some granitic rocks, and even on a large scale the uneducated eye will constantly mistake them for granites. Another very important circumstance is that granite and its allies frequently occupy the spaces that ought to be filled with gneiss or other rocks, were it not that they have been entirely fused and changed into granite. I therefore believe that many of the granite rocks I have seen, are simply the result of the extreme of metamorphism brought about by great heat with presence of water.

One reason why it has been inferred that granite is not a common igneous rock is that, enveloping the crystals of felspar and mica, there is generally a quantity of free silica, not always crystallised in definite forms like the two other materials. Silica being far less easily fusible than felspar, it seems clear that had all the substances that form granite been merely fused like common lavas, the silica ought on partial cooling to have crystallised first, whereas the felspar and mica have crystallised first, and the silica not used in the formation of these minerals wraps them round often in

[Fluid Cavities in Quartz. 53]

an amorphous form. Therefore it is said that it was probably held in partial solution in hot water, even after crystallisation by segregation of the other minerals had begun. This theory, now held by several distinguished physical and chemical geologists, seems to me to be sound, especially as it agrees exceedingly well with the metamorphic theory as applied to gneiss—granite being sometimes simply the result of the extreme of metamorphism. In other words, when the metamorphism has been so great that all traces of the semi-crystalline laminated structure has disappeared, a more perfect crystallisation has taken place, and the result is a granitic mass without any minor lamination in it. Even then, however, certain planes often remain, strongly suggestive of original stratification, and even of planes of oblique stratification or false-bedding.

These general results are not founded on mere conjectures. In a memoir by Mr. H. C. Sorby, 'On the Microscopical Structure of Crystals, indicating the Origin of Minerals and Rocks,' among other important points, he describes numerous small cavities in the quartz of granites, which are partly filled with water 'holding in solution the chlorides of potassium and sodium, the sulphates of potash, soda, and lime, sometimes one and sometimes the other salt predominating.' These 'fluid cavities' sometimes make up about five per cent. of the volume of the quartz, and he concludes that 'the fluid was not an accidental ingredient due to the percolation of water to a fused mass naturally containing none, but a genuine constituent of the rock when melted.' Reasoning on the underground temperatures necessary to expand the liquid so as to fill the cavities, by an elaborate process of argument he arrives at the approximate result, that 'the pressures under which granites were most probably

[54 Granite.]

formed' indicate depths from the surface varying from 15,100 to 65,500 feet. From certain passages it is evident that Mr. Sorby considers that gneiss and granite were formed approximately under similar circumstances. I quote this thoroughly philosophical memoir, that the reader may be less startled with the statement, that gneiss and some granites were formed by the metamorphosis of strata at depths counted by many thousands of feet, and also to give strength to the assertion, that under such circumstances water was present.1

If the above views be correct, though many granites having been completely fused have been injected among strata, and are thus to be classed as intrusive rocks, yet in the main, so far from the intrusion of granite having produced many mountains by mere upheaval, both gneiss and granite would rather seem to be often the results of the forces that formed certain mountain chains. Possibly this result was connected with the contraction of the earth's crust and the heat produced by the intense lateral pressure that, with much movement of parts, produced the contortion of vast masses of strata, parts of which, now exposed by denudation, were then deep underground, and already acted on by the internal heat of the earth in a degree proportionate to their depth.2

1 See 'Journal of the Geological Society,' vol. xiv., 1858. Sorby.
2 See Report, Brit. Assoc. 1866, p. 47: 'Address to the Geological Section,' Ramsay. Also an elaborate memoir by Mr. Robert Mallet, 'On Volcanic Energy, &c.,1 Trans. Royal Soc., vol. clxiii. p. 147.