properly mica gneiss-a rock only differing from granite in possessing a foliated parallel structure: an attempt to give some idea of this structure is made in Pl. LXXIII., fig. 5, which is taken from a piece of ordinary gneiss. The origin of gneiss is still involved in much obscurity, and there can be no doubt but that some varieties of gneiss have a totally different origin from others. It is known that beds of true mica or hornblende schist, i.e. those composed only of hornblende or mica along with quartz, often become felspathic close to their junction with eruptive granites, so as to become lithologically gneiss, although petrologically they can only be termed felspathic schists, since by following up these beds they are soon found to be true schists.

Other and more characteristic gneiss, in which the felspathic element is inherent and much more prominent, has from old times, and with much show of reason, been regarded as formed from the débris of eruptive granites, due either to subærial disintegration or to their having been ejected into or under water, and thus converted into the condition of tuff, which, after having been arranged under water as sedimentary beds and consolidated, have become subsequently crystalline by metamorphic action, thereby causing them to assume the foliated appearance they now present.

Whatever may be the true origin of ordinary gneiss, there is, however, another variety of this rock, called granitic gneiss, which cannot have been other than eruptive granite originally, in which the parallel arrangement of foliation is a superinduced structure, developed in it subsequent to its solidification; in fact it is a metamorphic eruptive rock, just as the ordinary schists are but metamorphic sedimentary beds.

The proof that such granite gneiss was originally eruptive is seen in the disturbance which it has occasioned in the rocks through which it has broken, as also in the fragments of these rocks which it encloses; thus Pl. LXXIII., fig. 3, is taken from Darwin's admirable work on the Geology of South America, and represents a fragment, 7 yards long by 2 wide, of dark coloured rock with garnets in it, enclosed in the ordinary granitic gneiss of the country about Rio Janeiro, both being cut through by a still more recent small granite vein; fragments and patches of the other rocks are also seen in the granitic gneiss of Donegal and Galway in Ireland.

A magnificent section, several miles in length, of such granitic gneiss can be seen in the naked and almost perpendicular cliffs on the north side of Eidsvand, a lake in the south of Norway: a portion of this section, elsewhere published by the author in 1856, is represented in Pl. LXXIII., fig. 1, and shows the original dark horablendic schists of the country broken through and

dislocated, whilst immense fragments of them are detached and enclosed in the mass of light-coloured granitic gneiss; and, as further evidence of its eruptive nature, fig. 6 shows how, at Haukeraadalen, this granitic gneiss throws out a small vein breaking through the adjacent hornblende schist and possessing all the structural characters of true granite, and entirely wanting the parallel lines of foliation seen dipping invariably at a high angle to the east in the main mass of granite gneiss. Similar cases have been described by Keilhau in other parts of Norway, and by Scott and Haughton in Donegal.

Even in the most characteristic eruptive granites, it is not at all uncommon to find portions of the rock possessing a more or less defined parallel structure, due to the position of the plates of mica in them, in consequence of which the rock, to use the workmen's language, "has a grain," and splits more easily in this direction—a feature which-as, for example, in some parts of the Aberdeen white granite quarries-is made use of for the purpose of making curbstones, &c. The lithologist, on seeing such stones, would describe them, and correctly so, as granitic gneiss, although petrologically-that is, studied en masse in the field-they are true granite. In Germany the distinction gneiss granite has been employed for such undoubted granites as possess a foliated structure.

The direction of the lines of foliation was for a long time regarded as that of the original stratification, and in many instances, no doubt, this is the case; for example, in the section

shown in the annexed woodcut, which is taken near Christiansand, and which shows alternating beds of ordinary and augitic gneiss with foliated crystalline limestone, which are without doubt the representatives of sedimentary beds of silicious and calcareous nature, originally deposited by water, and broken

through along the planes of stratification by the two granite dykes shown in the section.

Many years ago, however, attention was directed by Keilhau and others to the fact that the planes of foliation did not always coincide with those bounding the larger beds or masses of rock, seen to be distinct from one another by their differing in mineral composition, and that they were sometimes even at high angles to them. The subsequent observations of Darwin and Sharpe proved, in very different parts of the world, that, so far from being identical with stratification, the lines of foliation over large areas were also those of cleavage; and Mr. Sharpe, who regarded foliation as the final result of the same cause which induced cleavage in rocks, generalised, from observations in Scotland and the Alps, that the lines of these two structures, taken together, were parts of great arches many miles across-an hypothesis, however, which has not received subsequent corroboration. In 1854 the author announced, as the results of experimental as well as field observations, that the foliation in rocks appeared to be a structure induced subsequent to the consolidation of the rock, following the direction of the planes of least resistance in the rock, whether such planes were those of original sedimentary stratification, subsequent cleavage, or (in eruptive rocks) the striæ of fusion; and this view explains why the lines of foliation and cleavage so often coincide, since, if a rock had once undergone cleavage and subsequently became foliated, the foliations would naturally follow the planes of cleavage in preference to those of stratification, since the former would be those of least resistance.

It is therefore of the utmost importance that geologists, when observing in the field, should-especially in districts consisting of metamorphic schists and gneiss-continually bear in mind that the planes of foliation may not necessarily be in any way connected with those of sedimentary deposition, and that, in such districts, the only means of arriving at any sound conclusion as to what the probable original bedding had been, is by carefully studying the difference in mineral character of the various rock masses superposed one on another.

What the cause of foliation may be, is a problem as yet but little investigated or understood. Heat (not necessarily intense) appears certainly to have played an important part, since foliated rocks are rarely or ever met with unassociated or at any great distance from rocks of eruptive, i.e. of igneous origin. Experiments made by the author between 1849 and 1853 showed, when blocks of massive or amorphous soapstone were exposed for some months to a temperature not exceeding redness (infinitely below what would be sufficient to fuse or even

soften the rocks), under a slight pressure of from about seven to twelve pounds per square inch, that they became totally changed in structure and converted into an aggregate of finely developed crystalline foliæ of a brilliant white or greenish colour, identical with tale; in fact, a tale schist. Under similar circumstances, if protected from oxidation-as they contain some iron in the state of protoxide-ordinary clayslates were converted into rocks possessing a beautiful parallel structure, resembling gneiss so closely, that some of the hand specimens could not be distinguished by the eye from parts of the same clayslate altered at its points of contact with eruptive rocks in nature. In the first of these cases, mere re-crystallisation or molecular re-arrangement in the solid mass will explain the change; but, in the second, chemical action also has evidently come into play in re-arranging the chemical elements of the clayslate into other mineral forms not pre-existing in the slate. The effect of the heat being to expand and render the pores of the rock more open, doubtless admits of the molecules rearranging themselves and crystallising in the still solid rock, whilst at the same time the superincumbent pressure tends to force the crystals to shoot out or develope themselves in one direction only, i.e. at right angles to the pressure.

These experiments, and the formation of the well-known Réaumur's porcelain, show how such crystalline structure can be developed in solid bodies after their perfect solidification without any return to a fluid or molten condition; and the

FIG. 4.

annexed woodcut, taken from Keilhau, which represents an appearance in the gneiss of Jomfrueland, an island on the southern coast of Norway, can only be explained on the assumption that the direction of the lines of foliation in the rock (which are represented by the dotted lines) had been determined subsequent to its complete consolidation, since it will be noticed that the subordinate bed of dark hornblendic character has been dislocated by the fault AB, without any corresponding disturbance in the lines of foliation.

In the present state of science it would be premature to say more as to the probable causes of foliated structure; the hope may, however, be expressed, now so much attention is devoted to geological research, that the study of this interesting although abstruse subject may no longer be neglected.

Striation, or that structure which is due to the presence of what are termed the striæ of fusion, has frequently been called

slag, glass, or lava structure, from its being so characteristic of these substances also.

Everyone has probably noticed the existence, especially in glass of inferior quality, of certain lines which more or less injure the transparency of the glass, although in themselves quite transparent. In the portions of glass remaining attached to the bottom and sides of old or broken glass pots these lines are still better seen, being commonly rendered distinctly visible by alternating lines or layers of glass possessing different tints of green with others white or colourless, and often presenting a parallel structure of great beauty, especially when these lines are seen to be contorted and crumpled up into all manner of shapes. The same structure is seen in the vitreous slags from iron furnaces, the colours being usually shades of blue, yellow, green, and grey; whilst the slags from the coppersmelting furnaces often show extremely beautiful alternating striations of a deep red and black colour.

In obsidian or the so-called volcanic glass, an exactly identical striped or banded structure is extremely common, and in Pl. LXXIII., fig. 7, which represents a piece of obsidian from the Lipari Islands, an attempt has been made to convey an idea of this appearance.

This structure appears to be caused by the movements of the different parts of a viscid molten mass, which flow over one another at different rates of progression, whilst its existence is usually denoted by the bands or stripes in the rock either differing in colour or in the shades of some one colour. Sometimes, especially in obsidians and traps, this structure is accompanied by an innumerable number of small air-bubbles (?gas or steam also), which, being drawn out or elongated from their original spherical shape by the progressive movement of the molten mass, develop a parallel structure of very peculiar appearance. In the older lavas these cavities often become, through infiltration, filled up with carbonate of lime or other minerals, and thus form what is termed amygdaloidal trap or lava.

When such lavas, glasses, or slags cool quickly, they retain their vitreous character and the striated structure already described; when, however, the cooling takes place very slowly, or when, by natural or artificial agencies, they have been again heated and kept so for a considerable time, devitrification commences to take place, the crystalline or stony structure developing itself first along the lines of striation. In Pl. LXXIII., fig. 2, which depicts a section of a fragment of greenish plate-glass from the St. Helen's works, an attempt has been made to illustrate the gradual development of crystallisation (or foliation, in other words) along the lines of striation, origi