a rock with all the minerals characteristic of charnockite. In Malakanagiri dolerite dykes are equally unmetamorphosed in gneiss and in charnockite, for they always retain their ophitic texture and original augite.

The evidence from the texture of the Malakanagiri charnockites is also against the metamorphic theory. The occurrence in these rocks of large idiomorphic potash felspars is widespread. This is a feature only found in igneous rocks which have been little metamorphosed or not at all. Where igneous rocks have been subjected to high grade metamorphism, the potash felspar ceases to be idiomorphic and becomes xenoblastic with respect both to albite and quartz (see Harker, 1932, p. 38). It is noteworthy that true idiomorphic felspars have not been found in Uganda.

Groves makes a strong point in favour of the metamorphic origin of his rocks when he shows that the appearance of the successive ferromagnesian minerals, as seen in reaction rims, is in the reverse of the accepted order for plutonic rocks. The reaction rims round the ferromagnesians in the Malakanagiri charnockites appear to be always in the right order.

The field evidence shows the sharp and probably intrusive junction between the Malakanagiri gneisses and charnockites, and makes it clear that low grade metamorphism is the rule throughout this area. It also shows that charnockites of different chemical composition are surrounded by a uniform granitic gneiss.

The internal evidence shows idiomorphic felspars which could only occur in a normal igneous rock. This evidence is strengthened by the presence of reaction rims of a kind commonly found in cooling rocks, ones moreover which would be unlikely to survive any important rise in temperature.

It seems impossible, therefore, to avoid the conclusions that the metamorphism of the charnockite here is low, that the charnockite is not the product of metamorphism of older plutonic rocks but a normal igneous rock, and that the hypersthene in it is of pyromorphic origin.

As the Malakanagiri rocks are almost certainly apophyses from the main charnockite mass and as no evidence of high grade metamorphism has been found in the charnockites of that mass, it is probable that these conclusions apply equally to most of the charnockites in the Eastern Ghats.

X.-ALBITIZATION IN MALAKANAGIRI.

In previous Chapters I have vaguely referred the development of myrmekite, biotite, and hornblende to albitization of the original rocks, but have expressed no opinion as to the nature of this pheno

menon.

The first-formed minerals in the hypersthene-gneisses were felspars, pyroxenes, quartz, and perhaps iron-ore, none of which contain appreciable water. At a later date hornblende and biotite reaction rims have developed around the earlier pyroxenes, and albite has grown at the expense of potash felspar, usually with myrmekite as an intermediate stage. Moreover, it is clear from the comb-like structure of the biotite and hornblende, as well as from their relatively high water content, that they must have developed in an aqueous medium. It remains to discuss the myrmekite.

At first sight it would seem possible that the development of myrmekite and albite from potash felspars could be the result of uneven stress in the rock. It might easily be claimed, for instance, that the myrmekite was corroding the potash felspar at points of maximum stress.

If this were the case there should be little or no change in the general composition of the rock. If myrmekite and albite are to develop from potash-felspar, some soda must be introduced. This might come from soda-bearing minerals within the rock or from external sources. That the latter source is the right one is shown by the analyses of the rocks (p. 31) and by the fact that none of the soda-bearing minerals in the rock show any signs of decay.

This seems to rule out stress as the main reason for the development of myrmekite. This is borne out by field experience, for rocks, similar in composition to the Malakanagiri ones, which have been severely crushed, do not contain myrmekite. This is well seen in the acid charnockites of the main range. These have been severely stressed, but there is no myrmekite, and there are no large coronas of biotite or hornblende. Instead there is a considerable development of garnet.

If the myrmekite is not due to stress what is its origin?

The hypersthene-gneisses and soda-granites have been somewhat crushed, and contain numerous fine cracks through which water could readily pass. Water must have been present to enable biotite and hornblende to develop in the way they have done. That the formation of albite from potash-felspar also took place in the presence

of water is proved by the numerous inclusions of sericite in the albite. For albite and potash felspar contain no water but sericite does. It follows that the development of albite from perthite must have taken place in an aqueous medium, for otherwise there could have been no sericite; and as sericite has in many cases been deposited all the time during which the albite was developing from the potash felspar, it follows that there must have been a continuous supply of water at the contact. This means that there was a regular circulation of water in the rock while albitization was taking place.

I have already shown (p. 40) that the soda for the development of albite came from sources outside the rock. It seems almost certain that it was introduced into the rock in solution in the water circulating through it while albitization was taking place. This water probably corroded the potash felspars, the potash being leached out and the soda taking its place. Hence the great development of myrmekite and albite.

When the resulting alkali-rich waters came into contact with hypersthene an immediate reaction took place, followed by the development of coronas of hornblende and biotite.

Apatite occurs in the hypersthene-gneisses as minute idiomorphic crystals included in the first formed felspars, and as relatively large rounded masses associated with biotite and iron-ore.

Apatite.

In most charnockites the apatite occurs in the first form, and in very small amount. The rounded masses are therefore peculiar to the Malakanagiri ones, and probably account for their high phosphorus content (see p. 420).

It is reasonable to consider the idiomorphic apatites as primary and the others as secondary. The probability seems to be that the rounded masses have been introduced by the same solutions which have given rise to the myrmekite and albite. My chief reason for this view is that the rounded apatite is generally closely associated with biotite and iron-ore. These two minerals occur both as reaction products between the alkali liquids and the primary ferromagnesians, and as deposits along cracks. They must have been carried. along by the liquid, for otherwise they could not occur so frequently along cracks. The association of the apatite with these two suggests that it has also been transported by the same liquid.

The rounded shape of the apatite masses is possibly the result of their having filled cavities corroded in the original minerals of the rock by the alkali waters.

Iron-ore, like apatite, seems to have been frequently transported by the alkali waters. Whether it has been taken into solution by the alkali water from the rock itself and later redeposited, or whether it has been introduced

Iron-ore.

from external sources, is impossible to say.

Quartz, too, is probably partly criginal, partly a by-product of the corrosion of perthite, and partly introduced, but it is rarely possible to say what its origin is in particular cases.

Source of soda-bearing fluids.

The waters which have brought about these secondary alterations in the Malakanagiri gneisses cannot have been surface waters, for these would have brought about changes in which biotite coronas and myrmekite could have taken no part. They must therefore have been of deep-seated origin. They must also have been widespread, for rocks altered in the same way are believed to occur all along the western margin of the main charnockite range. The age of the alkali waters must be younger than that of the charnockite apophyses in southern Jeypore and older than that of the dolerites, for these have not been albitized.

The only source which would satisfy these conditions is the charnockite itself. This is deep-seated and widespread, and, like other igneous masses, probably extruded alkali waters in the final stages of its consolidation. This would probably have taken place after the consolidation of the Malakanagiri intrusions, and would certainly have been earlier than the dolerites.

It is most probable that the albitization was effected by waters driven off from the main charnockite intrusion in the final stages of its igneous activity. This would account for the widespread occurrence of these phenomena.

Albitization of the

Α very similar albitization seems to have affected the biotitegneisses and hornblende-schists along the charnockite border. Probably the same liquid from the charnockite older gneisses and was responsible for this also. The effects of the alkali waters on the gneisses and schists. are not so conspicuous as those described in the hypersthene-gneisses, but they are of the same kind. No doubt the total change in composition brought about in this way has been large.

schists.

Nature of the alkali solutions.

I have no clear ideas as to the nature of the alkali solutions. They must have been rich in alkalis, and probably carried much silica, iron, and phosphorus. The common occurrence of gold in the streams draining the gneisses and schists along the western margin of the Eastern Ghats suggests that it may have been introduced into these rocks in the same way, but there is no certainty on this point.

XI. THE KHONDALITES.

The main charnockite range is believed to be practically entirely igneous. Here and there patches of sillimanite-bearing rocks have been found in it, but only on a small scale, such as might be expected in the contact zone of any igneous intrusion.

Slides of these

It is not till the hills on the south-eastern side of the Machkund valley are reached that khondalites are extensively found. No effort was made in the field to map the boundary of the khondalites exactly, as this was outside the scope of my work. A few specimens were, however, collected in various localities. show that they are similar in essentials to other khondalites. In some cases (23246) they contain abundant orthoclase, in others (23248 and 23249) enstatite and rutile were noticed. The ferromagnesian minerals and also the sillimanite are much altered to serpentine.

The metamorphic grade of these rocks has never been questioned. They are essentially sillimanite-garnet-graphite-quartz schists, and must have been subjected to a high grade of metamorphism. The serpentine is probably a sign of later retrograde metamorphism.

A perusal of the literature on the charnockites and khondalites shows that those geologists who have studied them in the field all agree on the importance of stress in their development.

Thus Walker (1902, p. 10) holds that the khondalites are ancient sediments metamorphosed by great mountain-building forces, and by the numerous large igneous intrusions associated with them.

Smith (1900, p. 155) in discussing the khondalites of the Ganjam district says Since the consolidation of the gneisses much crushing has taken place. The folding that then occurred evidently enveloped large masses of the sedimentary rocks'.

Holland (1900, pp. 240-241) claims that the charnockites of the Madras Presidency have been subjected to dynamo-metamorphism,