The trend of foliation-planes across a country is often, as Darwin pointed out, remarkably regular; in some cases, it follows that of the stratification, in others that of cleavage. The wrinkling of the foliation must be ascribed to subsequent compression, and all the features seen in the "strain-slip" structure of slate ([p. 92]) are repeated on a somewhat coarser scale in schists.
Some schists are undoubtedly produced by the contact-metamorphism of shales. On the flanks of mountain-chains, where argillaceous rocks have been arched into domes, and where granite has intruded as a core, the complete passage can be traced from sediment to schist. The clay-rocks lend themselves readily to the production of mica, usually of the pale type. Andalusite, and occasionally sillimanite and kyanite, arise. Andalusite often forms grey prisms of irregular outline, resembling slate-pencils, and standing out above the mica on any weathered surface. Almandine garnet is almost always present. Quartz occurs in streaks and patches, which resolve themselves into granular aggregates on microscopic examination. The mica imparts a distinct foliation to the mass; but the original stratification is very often preserved, and the minerals have developed along its planes. Small differences in the constitution of the original strata give rise to different types of schist, interbedded with one another. Andalusite, for instance, may occur only in certain argillaceous layers, while other layers are quartzose, through the presence of original sand. Mica-schist is the commonest type of metamorphic rock.
Where mineralisation has taken place over a wide area, it may be difficult to say if the foliation-planes in a schist are those of bedding, or of superinduced cleavage, or whether they indicate a sliding movement in the mass under pressure, whereby all preceding structures have become obliterated.
Amphibole-schist, often styled epidiorite, consists of foliated hornblende, or its greener ally actinolite, associated with granular felspar and sometimes with equally granular quartz. The amphibole being usually prismatic, the crystals are found with their longer axes arranged in parallel planes, and often streaked out parallel to one another. Minute wrinklings, due to subsequent yielding, are not so frequent as in mica-schists. Amphibole-schists occur commonly as knots and somewhat irregular masses among mica-schists, and represent basic igneous rocks that were interbedded or intrusive in the sedimentary series. The pyroxene of the original rock has become recrystallised as hornblende, and the felspathic constituent has rearranged itself in granular forms. J. J. H. Teall[95] has described in interesting detail an example from the older rocks of Sutherland, and his paper contains a useful discussion of problems of pressure-metamorphism.
AMPHIBOLITES
Hornblende-schists are often seen to pass into true diorites; but they also have relationships with the more puzzling rocks known as amphibolites. These, again, graduate into pyroxenites, or rocks rich in pyroxene, with granular quartz and triclinic felspar, and into eclogites, which may be defined as pyroxenites with garnet.
Pyroxene-eclogite, in South Africa, is associated with diamond[96], and fragments of exploded eclogite abound in the igneous vents from which the diamonds are extracted.
What has been called "pyroxene-granulite" is a dark granular eclogite, including rhombic pyroxene side by side with garnet, and associated, in Saxony and Skye, with igneous intrusions. In both localities it has been shown to result from the inclusion of basic rocks, such as dolerites and gabbros, in a bath of some invading magma. The lens-like form of the Saxon masses, and the occurrence also of sheets of pyroxene-granulite interlaminated with fine-grained granite, were till lately attributed to the rolling-out action of pressure-metamorphism. By what H. Credner calls a complete reversal of opinion, due mainly to the opening of new railway-sections, the granular eclogites of Saxony are now regarded as products of extreme contact-alteration, combined with igneous flow[97]. A. Harker[98] similarly points out that examples in Skye are derived from basaltic lavas, into which gabbro has intruded, producing a complete reconstruction of the rock.
Where a series of igneous rocks and sediments, in some cases already altered by pressure, has been attacked and partly melted up by granite, amphibolite-blocks are found as the common residue in the mingled mass. The quartzites and mica-schists of the mantle that overlies the granite dome may have disappeared by stoping and absorption (see [p. 126]). Rocks rich in amphibole remain, and they commonly contain pyroxene as well as hornblende. In some cases, as in Skye and Saxony, they may be traced to basic igneous rocks; but in others they may be referred with equal certainty to limestone. The interaction of the granite magma and the calcareous sediment has produced a silicate rock completely different from either.
Lévy[99] and Lacroix have shown how the amphibolites of France may sometimes represent dolerites, sometimes limestones. Their work has recently received striking support from the observations of the Geological Survey of Canada[100]. Streaky hornblende-gneisses over wide areas of Ontario are now attributed to the partial absorption of overlying limestone by what was once regarded as a "fundamental" granite. The amphibolite blocks have become drawn out into bands that follow all the flow-structure of the invading igneous mass. A small area of the same kind was studied in 1900 in north-west Ireland[101], where a remarkably pure granitoid rock, consisting of quartz and alkali felspar, has become enriched with dark mica at the expense of blocks of amphibolite included in it.