mass along certain narrow belts; and that recrystallization was set up along with the development of schistose structure. The operating cause cannot have been anything but pressure, especially as the foliated rocks occur not infrequently in lines of dislocation and shear; in other cases the foliated, types are at the margins of the dike, and the transition from massive igneous rock to metamorphic schist may take place within the space of one inch. The best examples of phenomena of this order are those described by J. J. H. Teall from Scourie in the north-west of Scotland.
Where rocks of any kind are traversed by powerful dislocations or thrusts they often present a schistose facies in the immediate vicinity of the planes of movement. In the Highlands of Scotland great thrusts occur, along which the rocks are displaced for distances which may be as much as ten miles; and immediately adjoining these thrust-planes very perfect foliation is induced in all kinds of rocks, sedimentary, igneous or metamorphic, which have been involved in the movements. The minute structure of these rocks is generally of the mylonitic, granulitic or finely crushed type. In the same way the serpentine of the Lizard in Cornwall passes into fine talcose and tremolitic schists along narrow zones of displacement. Many other examples of this might be cited from regions where folding and crushing have taken place on a large scale. As a rule, almost without exception, the foliation thus produced is parallel to the direction of movement in the rock masses.
In the mineral transformations which accompany metamorphism the operation of pressure is no less clearly indicated. There are, for example, three minerals which consist of silicate of alumina, viz. andalusite, sillimanite and kyanite. The last of these has the highest specific gravity. In andalusite-bearing rocks which have been sheared, with production of foliation, we sometimes find pseudomorphs of kyanite after andalusite, retaining the characteristic form of the original mineral. Compression, it seems reasonable to suppose, would produce that one of the three crystalline silicates of alumina which has its molecules most closely packed, and consequently the highest specific gravity. This explains the conversion of andalusite into kyanite. The principle that substances tend to assume that mineral form which has the least molecular volume is of wide application among metamorphic rocks. It has been calculated, for example, that when olivine and anorthite felspar are replaced by garnet (a change which takes place not infrequently when basic igneous rocks are metamorphosed) the molecular volume of the mineral aggregate diminishes from 145 to 121 or about 17%. On the other hand, when garnet is fused it recrystallizes as a mixture of olivine and anorthite. This has led to the generalization that all minerals formed by the crystallization of a fused magma at high temperatures have a large molecular volume, while those which are produced in rocks at temperatures below their fusion points and under great pressures have smaller molecular volumes. Loewinson Lessing pointed out that some minerals have a greater molecular volume than the oxides which enter into their composition; in other minerals the reverse holds good. The former group are, on the whole, characteristic of igneous rocks and products of contact alteration, both of which classes have been formed at high temperatures (e.g. wollastonite, spinel, nepheline, leucite and andalusite). The minerals of the second group are often of common occurrence in metamorphic schists and gneisses (e.g. staurolite, kyanite, hornblende, talc, epidote and garnet). Although there are exceptions to this rule, there can be no doubt that it expresses a generalization which is of great value in the study of mineral paragenesis.
The mineral changes are usually not of so simple a kind as those above enumerated. Mutual interaction takes place between adjacent components of the rocks. Titaniferous iron oxides, for example, obtain silica and lime from such minerals as augite or lime felspar and sphene results. Felspar often breaks up into epidote, quartz and albite; the epidote obtains its iron from adjacent crystals of augite or hornblende. Equations can be written to show the transformation of one rock to another; thus, diabase (labradorite, augite, ilmenite) may be converted into amphibolite (acid plagioclase, hornblende, garnet, sphene and quartz). In this case, the molecular volumes are for diabase 671 and for amphibolite 635·6, indicating a diminution on metamorphism. Many striking illustrations of this principle have been adduced. Caution, however, is required in applying it to concrete cases; if it was always strictly correct the metamorphic rocks should have higher specific gravities than their representatives among sediments and igneous rocks. Very frequently this is not the case, and there must be some counteracting process at work. We find this antagonistic principle in the tendency for the minerals of metamorphic rocks to contain water of combination, e.g. epidote, muscovite, chlorite, hornblende, talc. This indicates that they were formed at comparatively low temperatures.
We arrive then by many independent lines of reasoning (stratigraphical, microscopical, chemical and mineralogical evidence being abundantly available) at the conclusion that pressure acting on sedimentary and igneous rocks at temperatures below their fusion points has been able to change them into metamorphic rocks. This is the theory of dynamo-metamorphism, which has won acceptance from the majority of geologists who have made the petrology of metamorphic rocks their special study. It has still, however, many incisive critics, and in recent years dissent has on the whole gained strength.
One of the principal objections is that by these processes it is possible to destroy original structures and to break down the minerals of which a rock consists, but not to induce crystallization and build up rock structures of a new type. It is pointed out that in many regions the rocks though intensely folded are not highly metamorphic; in other places immense dislocations can be proved to exist, yet the rocks are only slightly altered or are converted into fine-grained mylonites and not into typical schists and gneisses. Conversely, it is argued, there are many districts where metamorphism is very intense, yet evidence of folding and pressure is only slight. It seems clear that another factor must be taken into account, and in all probability that factor is the action of water in rocks at a comparatively high temperature. All rock masses contain interstitial water, and many also consist of minerals in some of which water exists in combination. Hence all metamorphism must be regarded as taking place in presence of water. It is almost equally certain that metamorphism must be accompanied by a rise of temperature in nearly every case—in fact it is difficult to imagine such a process going on without considerable heat. Now heated water (or water vapour) is a most potent mineralizer. Crystals of quartz, for example, have been produced in glass tubes containing a little water, heated in a furnace to a temperature of about 300° C.
The heat required for the more intense stages of metamorphism may be derived from more than one source. Most regions of gneiss and schists contain igneous rocks in the form of great intrusive masses. These rocks themselves are frequently gneissose, and the possibility must not be overlooked that they were injected into the older rocks at a time when folding was going on. The metamorphism would then be partly of the contact type and partly the effect of pressure and movement, “pressure-contact-metamorphism.” The vapours already present would be augmented by those given out from the igneous rock, and intensely crystalline, foliated masses, often containing minerals found in contact zones (andalusite, cordierite, sillimanite, staurolite, &c.), would be produced. Cases are now known where it is in every way probable that the metamorphism is the result of a combination of causes of this order. Some of the Alpine schists which surround the central granite gneisses have been referred to this group.
Heat must also have been produced by the crushing of the rock components. In many metamorphic rocks we find hard minerals possessing little cleavage (such as quartz) reduced to an exceedingly fine state of division, and it is clear that the stresses which have acted on regions of metamorphic rocks are often so powerful that all the minerals may have been completely shattered. The interstitial movement of the particles must also have generated heat. There are no experimental data to enable us to say what rise of temperature may have been produced in this way, but we cannot doubt that it was considerable. If the crushing was slow the heat generated may have been conducted away to the surface almost as fast as it was produced. If the belt of crushing was narrow, heat would rapidly pass away into the colder rocks beyond. This may explain why in some rocks there has been much grinding down but little crystallization. The heat also may be absorbed in promoting chemical combinations of the endothermal type, but it is not likely that much was used up in this way. With rising temperature the rocks would become more plastic and fold more readily. Then if the crushing and folding ceased, a long period would follow in which the temperature gradually fell. The minerals would crystallize in larger grains after the well-known law that the larger particles tend to grow at the expense of the smaller ones, and finely granulitic aggregates would be replaced by mosaics of coarser structure. If there has been a considerable rise of temperature we might expect analogies in structure and constitution between the folded rocks and those which come from a contact aureole; this has in fact been noted by many geologists.
Another factor which must have been of importance is the depth below the surface at which the rocks lay at the time when they were folded. In the deeper zones the pressures must have been greater, and the escape of the heat generated must have been less rapid. The uppermost members of a complex which was undergoing folding are under the lowest pressures, are at the lowest temperatures and probably also contain most moisture. Hence minerals such as epidote, chlorite, albite, sericite and carbonates, which are often produced by weathering alone, might be expected to prevail. In the deepest zones the temperature and pressure are high from the first and are increased by folding; such minerals as biotite, augite, garnet, felspar, sillimanite, kyanite and staurolite might be produced under these conditions. The earth’s crust might in this way be divided into bathymetric zones, each of which was characterized by distinctive types of mineral paragenesis. Some geologists ascribe the greatest importance to this conception; they establish two or three types of metamorphism, each of which belongs, in their opinion, to a definite horizon. This is to some extent a resuscitation of the old idea, now discarded, that the Archean rocks are sediments of a peculiar kind formed only in the heated waters of the primal globe;
the first deposits were laid down under great heat and pressure and
