In order to discuss this a clear idea of the rocks in the border zone is necessary. The following pages are, therefore, devoted to a description of these.

I propose to make this description brief where the ancient gneisses and schists are concerned, as these belong to the Bastar-Jeypore rather than the charnockite province. They cannot be entirely neglected, as the grade of metamorphism shown by them must be compared with that observed in the main charnockite massif, and also with that found in the outlying intrusions of hypersthenegranite-gneiss so common in the border zone.

I give a somewhat more complete description of the charnockites, but here again no detail is given where such rocks have already been described by Sir T. H. Holland (1900) and others.

The commonest rock to the west of the charnockite massif is a very fine-grained biotite-gneiss. Typically (23197) this is made. up of fine angular grains of quartz and felspar Biotite-gneiss. with a small but variable quantity of green biotite. The commonest felspar is microcline in small angular grains lying in all directions. These grains are often crowded together, and seem to be the broken remains of larger crystals. Orthoclase also is almost always present. In some slides (23265, 23266, 23190, 23252) the potash felspars occur in large crystals surrounded by finer-grained material, and show excellent mortar structure. They are normally free from the microperthitic textures so characteristic of the potash felspars in charnockite.

Albite and oligoclase are always present. The former is frequently difficult to distinguish from quartz owing to the similarity of their birefringence, and the absence of twinning. Myrmekite optically continuous with the albite is often seen developing at the expense of the potash felspar, but the phenomena of this kind are not nearly so striking as those to be described in the sequel, in connection with the minor charnockite intrusions (see pp. 24-25).

Quartz, though always present, forms quite a small part of the rock compared with the felspars.

The biotite is a dark green variety quite distinct from the brown biotite seen in the aluminous schists further west in Bastar State, The crystals are usually arranged in irregular lines or streaks, but individual crystals are rarely oriented parallel to these lines. In some slides (23199, 23213) the biotite is arranged in felted masses around crystals of sphene which in turn encircle grains of iron-ore.

In such cases it is probably a secondary growth similar to the albite which develops at the expense of the potash felspar.

Hornblende is often present (23189, 23190). It has a pleochroism from pale greenish yellow to greenish blue and an extinction angle of about 24°. It is probably a soda-rich variety.

The

Iron-ore, epidote, zoisite, and sphene occur in many slides. sphene is in many cases a corona around iron-ore, and the epidote round zoisite. All these minerals are commonly associated with the biotite and are partly due to its alteration. In some cases the biotite is much altered to chlorite.

Muscovite is rather rare, but sericite is a common inclusion in the albite. Epidote and quartz occur included in this way also.

Garnet is rare, but is found in biotite-schists from the edge of the charnockite (23192, 23195). It does not appear to be due to any very high grade of metamorphism, for epidote, sphene, and green hornblende are associated with it. In slide 23213 the garnet seems to be developing from a felted mass of biotite.

Apatite is the commonest accessory.

seen.

Zircon is also frequently

On comparing the localities from which my specimens were collected with the slides, I find that the number of large potash felspars decreases gradually as the charnockite range is approached. With the disappearance of the large crystals, mortar structures are also lost. Nevertheless I consider the fine-grained biotite-gneisses of the charnockite margin to be granulated forms of the slightly porphyritic ones further west, and that these in their turn grade into the porphyritic gneisses lying just to the east of the JeyporeBastar border. All are probably metamorphosed porphyritic granites. The garnet which sometimes appears in the biotite-gneiss near the edge of the charnockite may be a contact effect, or may be due to increasing stress.

Hornblende-schists.

Associated with the biotite-gneisses are numerous hornblendeschists (23212, 23188, 23194). These are essentially composed of felspar and hornblende with abundant inclusions of quartz. The hornblende is similar to that which occurs in the biotite-gneiss. Albite, oligoclase, and labradorite are the usual felspars. These often show strain phenomena. Biotite, iron-ore, sphene, epidote, and zoisite are all very common. The last mentioned pair occur in most cases as small idiomorphic crystals included in all the varieties of felspar. Garnet

is present in many of the hornblende-schists (23188). It is rather sporadic in its occurrence, but becomes distinctly commoner as the main charnockite range is approached.

I consider the hornblende-schists to be metamorphosed igneous rocks. The prevalence of labradorite in them certainly supports this view.

Dr. T. L. Walker (1900, p. 168) considered both the biotitegneisses and the hornblende-schists north of Govindpalle to be paragneisses. He based his opinion on the relative abundance of quartz in them as compared with felspar. Examination of his slides, which were made in 1899, shows that they were too thick to distinguish between quartz and felspar, unless the latter was well twinned. As the rocks contain a large quantity of untwinned albite, he must have overestimated the quartz at the expense of the felspar. It is possible, however, that he is right and I am wrong. The point is not of great importance as the chief interest of these rocks lies in their grade of metamorphism.

Large masses of white crystalline quartzite occur in many places among the hornblende-schists and biotitegneisses. It is uncertain whether these are

Quartzites.

quartz veins or acid pegmatites.

On the margin of the charnockite near Gumiribadaguda (18° 28′ :82° 15′) is an occurrence of magnetite-garnet-grünerite-quartzite. Coarse potstones composed mainly of chlorite with a little ironore are common along the charnockite border. They are probably metamorphosed ultrabasic

rocks.

Potstones.

THE METAMORPHISM OF THE OLDER GNEISSES.

The characteristic minerals of the older gneisses and schists to the west of the Eastern Ghats are biotite, green hornblende, chlorite, albite, zoisite, epidote, sphene, and, in the vicinity of the main charnockite mass, garnet. Mortar structures, strain polarization, bending of crystals, and other indications of mechanical stress are all very common, though some rocks show them better than others. All these minerals are commonly found in rocks of low to medium. grade metamorphism. Minerals associated with rocks of higher metamorphic grades are not normally found in this area.

I conclude from the general mineral suite that the grade of metamorphism west of the Eastern Ghats is low to medium. The

absence of remains of minerals characteristic of the higher grades of metamorphism indicates that the mineral suite found throughout this area is not the result of retrograde metamorphism. The evidence of widespread mechanical stress among these rocks strongly favours my conclusions.

III. THE NEWER DOLERITES.

The newer dolerites occur locally as dykes and small intrusions in the gneisses to the west of the Eastern Ghats, and also in the charnockites of the main range. In hand specimens they are black rocks hardly distinguishable from the epidiorites so common in Bastar State.

Under the microscope they are seen to be composed of augite. labradorite, and iron-ore with a little mesostasis of quartz and undetermined felspar. In the less altered rocks such as 23222, the only trace of metamorphism is a greenish rim around the edge of the augite. In more altered varieties (23187, 23202) (Pl. 22, fig. 3) garnet generally displaces the green chlorite. In some cases the development of garnet has been so remarkable that only cores of augite and a mere fraction of the original felspar survive. Even in such cases the ophitic texture is not completely destroyed. Strangely enough the original constituents of the dolerite are often seen altered to garnet at one part of a slide, and to chlorite at another. Garnet, and less commonly biotite, form coronas round many of the original grains of iron-ore. Apatite in fibrous crystals is the only other mineral present.

The mesostasis of the dolerite is distinctly seen in many slides (23206). It is composed of an intergrowth of quartz and felspar, the former greatly predominating. These intergrowths differ strikingly from the myrmekite so commonly seen in the acid charnockites. In the first place they are definitely interstitial with respect to the older felspars, and never seem to have corroded them as the myrmekite has. Also their texture is graphic, the two minerals being separated sharply by straight lines, whereas in the myrmekite the quartz always occurs as vermicular streaks and rounded blebs. Finally quartz preponderates in the mesostasis of the dolerite, but is present in relatively small amounts in the myrmekite.

I am stressing the difference in the two types of intergrowth, as in the sequel I have attempted to fix the age of the myrmekite on the assumption that it does not occur in the newer dolerites.

Age.

The age of the newer dolerite is clearly younger than that of the charnockite, as dykes of it cut the charnockite in several places. It has never been found intruded into the Cuddapahs, though it has been found in the rocks surrounding the Sukma Cuddapah basin. In all probability it is older than the Cuddapah system.

The metamorphism of the newer dolerites is of considerable interest. They contain no epidote, zoisite, or sphene, although the epidiorite intrusions in the older gneisses Metamorphism. are crowded with these minerals. It is clear therefore that they have escaped the earlier metamorphism which converted the older dykes to epidiorites. On the other hand they show a quite remarkable development of garnet, and in a few cases (23231, 23232, 50/123) mortar structure, bent felspars, and other signs of stress. The garnet cannot, however, be correlated with the stress, because in some slides (23231) where there are marked signs of stress there is no development of garnet. The garnet does seem to be most common in rocks from the charnockite range and its immediate vicinity, but there are so many exceptions that it would be unsafe to draw any conclusions from this circumstance. In the slides where the augite has a narrow rim of chlorite and there is no other marked change, petrologists would agree that the metamorphism was low grade. When portions of a slide show rims of garnet and other parts rims of chlorite, the problem becomes a little more difficult. But I think that if chlorite rims be accepted as a sign of low grade metamorphism in one rock, they must be accepted as such in all rocks of the same kind; and if the chlorite rims are due to low grade metamorphism, then the garnet rims which occur along with them must be so also.

Exception to this view has been taken on the grounds that the garnet is the high-water mark of metamorphism in these rocks, while the chlorite is a later development due to retrograde metamorphism. This is a possibility, but an unlikely one, as the garnet is usually quite fresh and does not appear to be altering to chlorite.

Holland (1896, pp. 20-30) has described a similar development of garnet in rocks from Madras and Bihar. There are, however, important points of difference between Holland's rocks and the Malakanagiri dolerites.