A. THE ACID VOLCANIC SERIES.

Below the sedimentary rocks last described there is found another great series of igneous rocks, which, though much decomposed, appear to be on the whole of a distinctly acid character. The examination of numerous specimens shows certain points of resemblance in almost all cases, but still it is possible to distinguish two more or less welldefined types, which, as a matter of convenience, are here described as amygdaloid and quartz porphyry respectively. The differences are probably more apparent than real, and are more conspicuous in hand specimens than in thin slices. Speaking broadly, however, we may say that the amygdaloidal rocks contain 'few phenocrysts of quartz, while in the non-vesicular types this mineral is abundant in the porphyritic condition.

Our collection contains a good many specimens of these rocks from different depths, and from a consideration of these it is evident that the amygdaloid and the quartz porphyry occur in alternating beds or flows. The total thickness of the whole group is very great, amounting at the Kimberley Mine to about 1070 feet. At the De Beers Mine it is less, since the surface of the granite is here about 600 feet higher.

Although these rocks are here treated of under two separate headings, it must be clearly understood that there is every transition between them, and some of the specimens are so much decomposed that their original character is very doubtful.

(a) The Amygdaloid.

In a hand-specimen (208a, 208b), this is a pale, greyish green rock, fine in texture, and with a somewhat streaky appearance. It contains very abundant rounded or oval vesicles, filled with dark grey or white minerals. Examinations of slices show that the rock has undergone so much alteration as to make determination of its original character difficult, if not impossible.

It contains abundant porphyritic crystals, which have the characteristic form of feldspars, and are now represented by pseudomorphs of somewhat varying character. Many of them are more or less silicified, consisting of an aggregate of quartz, white mica, and some pale green chloritic substance, which shows very weak birefringence. The flakes of white mica usually show a distinct arrangement along two sets of planes, more or less at right angles, presumably the cleavage planes of the original feldspars. It is quite impossible to determine the original character of the feldspar, since no part of the original substance seems to remain.

The ground mass is of a peculiar character, consisting of a very fine-textured aggregate of quartz and feldspar it shows the peculiar patchy structure known as micropoecilitic, which is defined by Harker as follows:-"This consists of minute feldspar crystals with no orderly arrangement, enclosed in little ovoid or irregular areas of Petrology, 3rd Ed., 1902, p. 163.

*

quartz, the whole of the quartz in such a little area being in crystalline continuity." This structure is very characteristic of lavas which have undergone silicification, as has clearly happened in this case.

Some specimens show in the ground mass many well-formed rhombs of dolomite, which somethimes encroach on the feldspar phenocrysts, but do not appear to be derived from them; it would rather seem that they have originated in the ground mass, and have pushed aside the decomposing feldspar during their growth.

Vesicles are very numerous, but are rarely more than 3 or 4 mm. in diameter. They are filled with concentric layers of quartz, chlorite, calcite, and dolomite, and possibly some feldspar and zeolithic minerals.

Whatever the original character of this rock may have been, it is at any rate of volcanic origin, and in places the structure suggests that it may be something of the nature of an agglomerate, and not a true lava-flow. The specimens available are hardly of sufficient size to allow of the formation of a decided opinion on this point, but it is at least a possibility to be considered. A careful examination of the thin slices with a pocket lens yields no definite evidence. It should also be borne in mind that a simulation of clastic structure is often induced in viscous lavas by a kind of brecciation during flow. The question of the original character and origin of these amygdaloidal rocks must be left an open one.

(b) The Quartz-Porphyry.

The non vesicular type in a hand-specimen (209g, 209r, 209u, 211) is a rock of very crystalline appearance, and of the prevailing greenish grey tint which is so common in this series. It shows abundant phenocrysts of clear, glassy quartz and white, somewhat opaque, feldspar. Under the microscope the porphyritic character is conspicuous, and perhaps the most striking feature is the extraordinary manner in which the quartz phenocrysts have been corroded by the groundmass, so that they assume remarkable forms. This is a very common feature of the more acid hypabyssal and volcanic rocks, but is seldom so strikingly exhibited as in this case. This phenomenon is apparently due to the varying solubility of the quartz in the magma under different conditions of pressure.

In the intratelluric stage, where the pressure is high, quartz separates out in well-formed crystals, but in the effusive stage the pressure is decreased, and the quartz becomes more soluble, so that it is again partly dissolved by the magma. The peculiar forms are due to the varying solubility of the quartz in different crystallographic directions. *

The feldspars of this rock include both orthoclase and plagioclase in approximately equal proportions. Both occur in large, tabular crystals, which have often undergone a certain amount of magmatic' corrosion. This is perhaps most strongly marked in the case of the orthoclase.

* See Penfield, Trans. Conn. Acad., 8, 1889, p. 158.

The extinction angle of the plagioclase, measured on the albite twin lamellae, ranges up to about 22°. Its index of refraction does not differ much from that of orthoclase, so that it is probably albite.

The rock contains no phenocrysts of ferromagnesian minerals, but there are a very few small irregular crystals of a partiallydecomposed iron ore.

In some

The groundmass is microcrystalline, consisting of quartz, feldspar, and a good deal of pale green chlorite. Its general appearance is distinctly suggestive of silicification, as it is micropoecilitic, with a tendency in places to a microspherulitic structure. forms vesicles are absent, in other quartz-bearing specimens a few vesicles occur, and, as before stated, every transition can be traced to the highly-amygdaloidal type.

The foregoing description is principally taken from the freshest specimen, which comes from a depth of 1840 feet in the Kimberley Mine (208b), and probably approximates most closely to the original character of the rock. Other specimens are much more altered, and in particular many have undergone a high degree of silicification, so that both phenocrysts, vesicles and ground mass are converted into a mosaic of clear, granular quartz. A good deal of calcite is also often present.

The rocks described in the two preceding sections must be regarded as forming one continuous series of a volcanic nature, but it is not clear whether some of the non-vesicular types are true lava flows. It is possible that some of them may be of the nature of sills or other intrusive masses injected into a previously-existing series of lavaş. There is not sufficient evidence at our disposal to decide this point, which can only be settled by an examination of the sections on the spot.

B. BASIC DYKES.

At various depths in the acid volcanic series there are found rocks of a much more basic character, which are referred to in the field notes supplied as dykes. These specimens come from the Kimberley Mine, at depths of 1520 feet, 1840 feet, and 2160 feet respectively. They differ very markedly from the surrounding acid volcanic rocks, and present some special points of interest. The three specimens vary to a certain extent among themselves, but they have sufficient features in common to suggest a genetic connection.

The freshest example comes from a depth of 1520 feet (222), and is a rock of peculiar nature. Its macroscopic appearance is that of a slightly porphyritic, dark-coloured rock of distinctly basic appearance, with no visible feldspar. Under the miscroscope it is seen to consist almost entirely of colourless pyroxene and a green chloritic mineral, with only a little interstitial feldspar, and the merest trace of quartz.

The ferromagnesian mineral clearly occurs in two generations, both as phenocrysts, and constituting the bulk of the ground mass. The crystals of the first generation are idiomorphic, usually in very well-formed, eight-sided crystals, which often show polysynthetic and

interpenetrating twins. The mineral, when fresh, is colourless in thin slices, and its double refraction is low, giving interference colours not higher than yellow or orange of the first order in a slice of the normal thickness. The extinction is oblique, ranging up to 45°, so that the mineral belongs to the monoclinic system. It is therefore to be identified as one of the colourless members of the augite group, diopside or fassaite. Partly intergrown with the augite in parallel position, and partly occurring in independent prismatic crystals is a pale green chloritic mineral, possessing the characters of bastite pseudomorphs after a rhombic pyroxene. This is probably the mineral described by Bonney *, and by him referred rather hesitatingly to enstatite. Feldspar crystals of the first generation are entirely absent, but large crystals of ilmenite, showing the characteristic bar-structure, probably belonging to the intratelluric stage.

The ground mass consists chiefly of an aggregate of small prismatic crystals of the same pyroxenes, with a certain amount of interstitial feldspar, and, as before mentioned, a very small amount of quartz. The feldspar is much decomposed, and it is difficult to determine its original character. However, a careful examination, both by daylight and in artificial light, has failed to reveal any albitetwinning, except in one or two cases, and it is clear that the most of the feldspar must be referred to orthoclase.

It is somewhat difficult to assign this rock to its type. The very small proportion of feldspar, and the fact that this is orthoclase and not plagioclase, serve to distinguish it from the porphyrites and normal dolerites. Perhaps it can be most satisfactorily classified as a member of the lamprophyre group. According to Rosenbusch's definition, it must then be described as vogesite.

The other two examples of dyke-rocks (222c, 222d) differ from the preceding in showing a higher proportion of feldspar, especially plagioclase and quartz. They approach much more nearly to the normal dolerite type, and, as is common in this group, the last con stituent to crystallise in the ground mass is a eutectic of quartz and feldspar. They may be quite safely described as quartz-dolerites, and their macroscopic appearance confirms this view.

We have no evidence concerning the age-relations of these basic rocks to the acid series in which they occur, but they are probably subsequent to, and intrusive in the latter, and this supposition is confirmed by their greater degree of freshness.

C. LOWER SEDIMENTARY SERIES.

Below the acid volcanic series comes another small development of sedimentary rocks, which in the Kimberley Mine appears to be about 30 feet thick. It is represented in our collection by one specimen only, from a depth of 2470 feet (220). This is a heavy, black rock of crystalline appearance, which shows conspicuous and rather irregular bedding.

Under the microscope it is seen to consist of an alternation of layers of very fine-grained quartzose sediment and lenticular patches and streaks of crystalline calcite.

The quartzose bands consist essentially of very minute chips of quartz, and probably feldspar, with a good deal of mica, both colourless and brownish: there is also a considerable amount of black, carbonaceous matter along the bedding planes. Much calcite is

present in scattered crystals. The calcareous layers are usually more or less lenticular in shape, and are composed of crystals of calcite (or dolomite) showing secondary twinning. The whole rock shows distinct signs of crushing, and might almost be spoken of as foliated parallel to the original bedding planes.

So far as the origin of this rock is concerned, two explanations are possible :

(a) It may be a fine-grained sediment, containing pebbles of limestone, which have subsequently been squeezed out of shape and flattened by pressure.

(b) The calcareous matter may have been in the form of spheroidal or ellipsoidal concretion, due to segregation in the original rock, and afterwards recrystallised under pressure, perhaps accompanied by thermal metamorphism to a certain

extent.

All we can definitely say is that the rock is a sediment, containing both siliceous and calcareous matter, which has been subjected to a certain degree of metamorphism, so that it has been crushed and more or less recrystallised. The thermal metamorphism is probably due to the overlying volcanic rocks, while there is nothing to indicate the source of the crushing.

D. THE GRANITE.

Below this sedimentary rock, down to the greatest depth represented in the collection, viz., 2520 feet (215) is found a rather pale, grey granite, which in places shows a somewhat gneissose structure. It has already been pointed out that the granite is reached at a considerably less depth in the De Beers Mine, indicating either a very uneven surface of denudation, or else intrusion of a very irregular mass. This last supposition is unlikely, since the sedimentary rocks within a few feet of the granite show only a feeble degree of thermal metamorphism, so far as can be judged from the small amount of material at hand. It is much more probable that the upper surface of the granite represents an ancient buried landscape, as in the case of the Mount Sorrel granite and other pre-triassic rocks of the English Midlands. * However, it must be admitted that this is almost pure speculation, founded on very insufficient evidence.

It so happens that specimens of granite from the De Beers Mine are much less decomposed, and show somewhat greater variety than those from Kimberley, so that they will be treated of more fully later on.

* Watts, Geogr. Journal, June, 1903.