Chapter III

PETROGRAPHY OF THE GNEISSES AND SCHISTS AND ASSOCIATED IGNEOUS
INTRUSIONS.

Subdivision of the gneisses and schists; predominant meridional strike. Gneisses and schists of sedimentary origin; variation in degree of metamorphism; phyllites and slaty schists; green schists; quartzites; quartz schists and granulitic gneisses; crystalline limestones; micaceous and felspathic gneisses; calc silicate rocks; epidote hornblende rocks; magnetite and hæmatite schists; sheared trachytes of Bussa; altered dolerites of Ngashki; amphibolites and talc schists; granulites. Petrography of the banded Archæan gneisses: relationship of the two groups of gneisses. The gneisses and schists of the Sahara and of West Africa. Age of the sedimentary gneisses. Intrusive rocks: the older granites; variation in degree of reconstruction; brecciation and injection structures; the younger granites; aplitic and tourmaline- bearing granites and pegmatites; tourmaline quartz veins; stanniferous pegmatites; schorl rocks; soda granites; biotite and riebeckite granites; tinstone associated with soda granites; biotite and augite syenites; augite diorites; riebeckite quartz porphyries; ægirine granophyre; augite syenite porphyries; alkali syenite porphyries; diorite porphyrites; mica porphyrites; hornblende porphyrites; diabases and basalts. Granites and dyke rocks of the Sahara. Age of the soda granites; successive periods of intrusion.

THE GNEISSES AND SCHISTS.

For descriptive purposes it has been found convenient to separate the crystalline rocks of the Protectorate into two groups, (1) a group of quartzites, phyllites, schists, and soft felspathic and micaceous gneisses, and (2) a group of hard well-foliated, much granitised and frequently banded and contorted micaceous and hornblendic gneisses. The first group occupies by far the greater area within the Protectorate, and presents in the west a well-marked alternation of belts in which quartzites, phyllites and schists predominate, with belts of which the prevailing members are felspathic and micaceous gneisses. The second group has no definite distribution within the Protectorate, but appears to replace locally and irregularly the members of the first group throughout the whole crystalline area. The members of both groups appear to be irregularly folded along axial planes, which are usually vertical or highly inclined, but it is probable that detailed investigation would reveal a regular system of major folds, accompanied by much local and minor crumpling. The folding, which is typically meridional, has been affected by equatorial stresses in such a way as to produce a transverse twisting or corrugation of the predominant strike. The result is that individual exposures may vary in direction of strike from N.–S. to E.–W., while in places, as in Bauchi province, the strike may become regionally equatorial. A similar equatorial divergence of the prevalent meridional strike has been noted by M. Hubert[28] in eastern Dahomey, but his suggestion that the equatorial direction becomes predominant in Borgu and the Niger valley has been found to be unsupported by local evidence. It is impossible to believe, moreover, that there can be any connection, as suggested by M. Chudeau,[29] between this locally divergent strike of the gneisses and the distribution of the overlying Cretaceous and Eocene rocks.

The sedimentary origin of the first group is fairly certain. The pebbly quartzites, quartz schists, schistose grits, mica schists, slaty schists and phyllites represent corresponding varieties of sandstones, grits and shales, and frequently exhibit traces of their sedimentary origin in the deformed pebbles and clastic grains which they contain. The granular felspathic and micaceous gneisses on the other hand are more highly reconstructed, and afford in themselves little or no evidence of a clastic origin. Their constant association and alternation with the quartzites and schists make it probable, however, that they represent the arkoses and greywackes of the original sedimentary series. Many of the green hornblendic and epidotic beds probably also represent original gritty accumulations of chloritic and epidotic character, while the quartz magnetite schists of eastern Kabba correspond to ferruginous grits or gritty ironstones of sedimentary origin. The fine-grained crystalline limestones of Igbo, and the course granular marble between Jakura and Wa, similarly represent original clastic calcareous rocks within the sedimentary succession. To what extent the recurrence of belts of quartzites, phyllites and schists in the western part of the Protectorate represents the repetition of one or more belts is quite unknown. The phyllites and schists are moreover everywhere unfossiliferous and the age of the original sedimentary rocks is consequently undetermined. The foliation of the reconstructed rocks is, as a rule, parallel with the bedding, but the more micaceous schists frequently show a more or less well-defined and oblique strain-slip corrugation.

The alternation of belts of less and more highly metamorphosed rocks in the west of the Protectorate coincides to some extent with an alternation of belts in which shear structures predominate, with belts in which reconstruction is prevalent. The various belts appear to pass insensibly into each other, but it has been found impossible in the course of a rapid survey to decide either to what extent shearing may have everywhere preceded reconstruction, or to what extent the varying degree of reconstruction in alternate belts may represent a corresponding variation in the intensity of the accompanying folding. All that can be said with certainty is that the sedimentary series as a whole has been subjected to a highly complex folding accompanied by much shearing and reconstruction, and that while over the greater part of the Protectorate the stresses have been for the most part relieved by, or translated into, a molecular and mineralogical rearrangement, there are certain belts in which relief has been primarily obtained by mechanical readjustment or shearing. Moreover, as in the reconstructed areas there are certain beds which appear to be more coarsely crystalline and less distinctly foliated than others, so in the shear belts there are certain beds which have assumed the character of granulitic gneisses, while others in the immediate neighbourhood have retained to a very large extent their original clastic and gritty character. Most of the reconstructed gneisses exhibit under the microscope faint strain phenomena which, with the incipient strain-slip foliation of the more micaceous schists, are probably to be ascribed to regional stresses affecting the whole series after the period of maximum reconstruction.

Grey and purple sericitic phyllites with knots of magnetite and pyrites are extensively developed in the neighbourhood of Ngashki in the Niger valley, around Anka and Bukwium on the borders of Sokoto and Kontagora, and in the Zungeru belt in eastern Zaria from Kato to Birnin Gwari and Koriga. In Kabba and Nassarawa and southern Zaria their place is taken by fine-grained quartz muscovite schists with abundant magnetite. With the phyllites are associated in the Niger valley and Sokoto province green quartzose slaty schists (1195, 1325),[30] whose colouration is due to recrystallised chloritic and hornblendic material. Mr. Parkinson[31] has described similar slaty schists from Southern Nigeria. In the neighbourhood of Zungeru these green schists become softer and show under the microscope (1563–4) ragged porous or fibrous crystals and crystal aggregates of reconstructed hornblende in a matrix of granulitic quartz, felspar and magnetite, which encloses also scattered grains of hornblende, biotite, chlorite and epidote, as well as relics of original clastic grains of quartz and felspar. These rocks present a considerable resemblance to the “green beds” of the Scottish Highlands.[32] True quartzites are comparatively rare and only at Kazaure do pure siliceous rocks attain any extensive development. Elsewhere the rocks which present the characteristic appearance of quartzite in the field are usually more or less felspathic and micaceous, and include schistose grits, micaceous quartzites, quartz muscovite schists and granulitic gneisses which show all the familiar stages in the deformation of felspathic grits and sandstones. Some of the white and pink granulitic gneisses, such as those of Zungeru, may, however, be ultimately proved to have been originally of an eruptive character. A characteristic feature of these rocks is the development of tourmaline and sometimes of pyrites along the shear and foliation planes as the result of later pneumatolytic action. At Kuta a belt of coarse quartz muscovite schist contains abundant sillimanite in sheaf-like aggregates.

Mica schists in which biotite predominates are common throughout and frequently carry garnet, magnetite and staurolite in conspicuous crystals. In the Igbo limestones the original impurities have recrystallised as tremolite, diopside, scapolite (122), sphene, biotite, felspar, quartz and pyrites, while the streaks of silicates in the limestone between Jakura and Wa contain diopside, epidote, idocrase, felspar, and quartz. In the latter limestone also, scales of graphite are common throughout and pyrites on the margin. Amongst the reconstructed gneisses, which show a remarkable and rapid variation in texture, the paler-coloured felspathic rocks are usually rich in orthoclase and microcline, and occasionally muscovite, while the darker- coloured varieties, in which biotite predominates, frequently carry garnet, sillimanite, staurolite and in places also sphene, tremolite and carbonates. Andalusite-bearing rocks are known from certain localities, but kyanite gneisses and schists have not been definitely recognised in place. The local occurrence of kyanite on the river sands, however, points to its presence within the neighbouring rocks. Calc silicate types are represented by the tremolite schists of the Nassarawa tableland (668, 689), by a diopside tremolite schist from Borgu (1221), a diopside and wollastonite-bearing rock with sphene from Maibirro in the Niger valley (1192), a biotite tremolite schist with calcite from Olle in northern Kabba (879), and a diopside and scapolite-bearing rock with sphene and green spinel from Atuka, near Okuruku, in eastern Kabba (843). The epidote hornblende gneisses of Boi in southern Bauchi (976) contain much diopside and sphene, and are probably of sedimentary origin, while relics of clastic felspar are common in an epidote rock from Goshin Duchi a few miles east of Boi (983). A similar clastic origin is presumed for many of the epidosites and epidote hornblende rocks associated with the sedimentary gneisses in other parts of the Protectorate. The quartz magnetite and quartz haematite schists include coarse granular and fine-grained granulitic types (844, 1370), the latter presenting considerable resemblance to the itabirites of the United States. Their sedimentary origin is presumed throughout and only in the case of the massive quartz magnetite rock of Okuruku (148), which is associated with a magnetite-bearing epidiorite, is there any probability of an igneous origin.

Two stages of alteration can be recognised among the sheared trachytes of Bussa (p. 72). The least altered types (1164, 1167) show fracturing of the phenocrysts of felspar and the replacement of the original ferro- magnesian minerals by fibrous uralite, chlorite, epidote and calcite. The more altered types (1166) are characterised by the production of abundant white mica in the ground-mass and the development of biotite, magnetite and carbonates from the original ferro-magnesian material. Original vesicles of quartz and calcite in these rocks have been deformed and the quartz replaced by a mosaic of angular grains. A sheared quartz-porphyry from Kanikoko (1206) exhibits deformation and granulation of the phenocrysts of quartz and felspar and the abundant development of white mica in the ground-mass. In the altered dolerites of Ngashki (1175, 1176) the original structures are almost entirely obscured by the extensive production of granular epidote, calcite, uralite, chlorite, magnetite, felspar and quartz. Knots of tourmaline are common in these altered rocks. The types grouped as amphibolites range from epidiorites to hornblende schists and amongst them may be recognised all the familiar stages of saussuritisation and recrystallisation of original tonalitic, dioritic and diabasic rocks.

It is probable, however, that amongst the amphibolites are included rocks which were intruded at many different periods during the metamorphism of the gneisses and that to difference of age as well as to local intensification of the metamorphism is to be ascribed the variable degree of construction which the rocks have severally undergone. The chlorite and talc schists are to be regarded as basic intrusions which had suffered considerable decomposition before being subjected to metamorphism.

Amongst highly reconstructed rocks of doubtful derivation, associated with the sedimentary gneisses, are a garnetiferous quartz enstatite biotite granulite from Atuka near Okuruku in southern Kabba (841) and a garnet granulite from the adjoining village of Kuturu (168) composed of labradorite, quartz and garnet with a distinct centric structure. The garnet is in spongy masses associated with a small quantity of green pyroxene, sphene, apatite and iron ore. A pyroxene granulite, from Againyi (823), consists of a porous intergrowth of pale green pyroxene and green hornblende, the former predominating, in a matrix of clear labradorite. Sphene and ilmenite are also present in irregular intergrowths as well as a little accessory epidote and apatite. Another variety (832) contains no amphibole and consists of pale green pyroxene and garnet in a matrix of labradorite and quartz. A third type from the same locality (821) is an epidote granulite consisting of predominant epidote and garnet with subordinate green pyroxene, sphene and ilmenite in a matrix of labradorite. A pyroxene granulite from Nassarawa province (854) shows ragged crystals of diallage with quartz inclusions and rounded garnets with inclusions of both quartz and diallage in a matrix of coarsely granular quartz. A hornblende granulite from the same province (859) consists of rounded spongy garnets and radiate groups of actinolite fibres with a little accessory ilmenite in a typical fine- grained quartzo-felspathic mosaic. A remarkable biotite granulite from Eri in eastern Ilorin (1056) contains in addition to biotite and granular garnet as the predominate ingredients much sillimanite, staurolite and andalusite in a matrix of untwinned felspar and quartz.

The harder gneisses of the second group are subordinate in amount and irregularly distributed throughout the micaceous and felspathic gneisses of the first group. They consist of a series of much granitised medium and even-grained biotitic and hornblendic gneisses which usually show a lenticular or linear foliation but frequently also a ribboned or banded and contorted character. In some cases the banding is broad and only to be distinguished in the field, while hand specimens present quite a homogeneous character. All variations may be found between acid aplitic and basic hornblendic bands, and knots and lenticular masses of more basic material are common within the more acid types. As a rule the banding is much interrupted by pale coloured granitic material into which the lighter coloured bands seem to pass imperceptibly. A striking feature is the presence of a series of early pegmatite veins folded and foliated with the gneisses. There is in fact every indication that these banded rocks are of a primarily composite character and represent either an originally sedimentary or a differentiated igneous succession invaded and injected by acid granitic material and pegmatite veins in the deeper portions of the crust and subsequently subjected to a later folding, foliation and recrystallisation. The foliation is in general vertical and parallel with the banding. Under the microscope the striated and banded gneisses exhibit all the characteristic features of typical archæan rocks (801, 846). They are composed of a granular aggregate of quartz, felspar, hornblende or biotite with sphene, apatite and iron ores as accessories. The felspars are usually allotriomorphic and vary from labradorite in the more basic to orthoclase and microcline in the more acid rocks. The biotite occurs in tabular flakes singly or in aggregates with the basal plane well-developed but with ragged margins in the prism zone, while the hornblende forms spongy growths or irregular plates with embayed margins and inclusions of quartz. Strain shadows are common in the quartz and felspar but only traces of cataclastic granulisation or of secondary micropegmatite are as a rule recognisable. Van Hise[33] is of opinion that the strain phenomena and the vermicular quartz to be observed in these thoroughly reconstituted gneisses are due to the partial lagging of reconstruction behind deformation. It is equally probable, however, that the strain effects and micropegmatitic structures have been induced after the period of maximum reconstruction by later strains of lesser magnitude.

The mutual relations of the two groups of gneisses are difficult to determine. The actual contact between the two has nowhere been found clearly exposed and its character must therefore be to a certain extent conjectural. The apparent irregular replacement of the softer gneisses by members of the harder group when traversed across the strike and the apparent extension into the softer gneisses of tongue-like masses from the larger areas of harder rock is certainly suggestive of an intrusive character as advocated by Mr. Parkinson[34] in Southern Nigeria and Liberia. Upon this hypothesis the harder gneisses would appear to correspond to an intimate mixture of granitic and dioritic material, whose intrusion took place during the folding and reconstruction of the sedimentary gneisses. It is difficult, however, to understand why these rocks should present in their even-grained and closely foliated and banded character such an entirely different aspect in the field from that of even the earliest recognisable granitic intrusions within the sedimentary gneisses and why they alone should possess an early set of folded and foliated pegmatite veins in addition to the later pegmatites which are common to both groups of gneisses. All the phenomena indeed are more readily explicable on the assumption that the two groups differ in age, that the sedimentary gneisses were originally deposited unconformably upon an earlier and probably archæan gneissic or igneous complex and that the latter series was refoliated and recrystallised during the reconstruction of the former. The peculiar distribution of the harder gneisses and their intimate alternation with the softer series would then be due in part to the original unconformability between the two groups and in part to the irregular character and the varying amplitude of the later folding to which both were subjected. The abrupt replacement of the sheared but little altered quartzites of the Niger valley by the harder gneisses at Jebba, Wuru and Maibirro probably indicates also a certain amount of faulting and thrusting accompanying the folding. On account of the subordinate and ill-defined character of the distribution of the harder rocks no attempt has been made to distinguish upon the map between the archæan and the later sedimentary gneisses.

PLATE IX

The Parade Ground at Bukuru.]

In the Sahara to the north, M. Chudeau[35] has described the occurrence of archæan and sedimentary gneisses whose mutual relations are similar to those which prevail in Northern Nigeria. The archæan areas are of comparatively small extent, and are primarily distinguished, according to M. Chudeau, by the presence of numerous domes and turtlebacks of granite. This method of distinguishing between the two groups, however, cannot be applied in Nigeria, where intrusive granites are equally common amongst both the earlier and the later gneisses. The metamorphic rocks of sedimentary origin attain a great development in the Sahara[36] and include quartzites, phyllites, schists, gneisses and crystalline limestone similar in type to those of Nigeria, and folded in a similar complex manner along axes whose direction is predominantly meridional. The decrease northward of the intensity of the metamorphism, as advocated by M. Chudeau, requires much further investigation in view of the local occurrence of little altered rocks within the Protectorate. M. Chudeau believes in the original unconformability of the later gneisses upon the archæan and points out, with M. Haug,[37] the analogy between the system of folding of the gneisses of the Sahara and the Sudan, and the Caledonian folding of the crystalline rocks of Scotland and Scandinavia. He demonstrates the pre-Devonian age of the sedimentary gneisses, and assigns them provisionally to the Silurian on the strength of their supposed continuity with the graptolite-bearing shales of Tindesset[38] and Hassi el Kheneg.[39] The stratigraphical evidence available, however, seems rather to point to an unconformability between the shales and the schists, and M. Flamand[40] prefers to assign a pre- Cambrian age to the whole series of crystalline schists and gneisses.

In Dahomey, M. Hubert[41] has described a similar complex of archæan and sedimentary gneisses in which the latter clearly predominate. His separation as Silurian of the quartzites and schists of Atacora from the other metamorphic rocks appears, however, to be unjustifiable. Quartzites, gneisses and schists of sedimentary origin associated with granites and archæan rocks, are found also in West Africa, in French Guinea and the Fouta Djalon,[42] in Liberia,[43] the Ivory Coast,[44] the Gold Coast,[45] Togoland,[46] and Southern Nigeria,[47] and in Central Africa to the south and south-east of Chad,[48] in the Kameruns,[49] and in the French Congo.[50] Similar rocks attain also a remarkable development to the south of the equator.[51]

The comparison which M. Chudeau has drawn between the strike of the gneisses and schists of the Sudan on the one hand, and of the Scottish Highlands on the other, may be further extended not only to the petrographical character of the constituent groups of the gneissic complex in West Africa, but also to the respective relations of the various groups to each other, and to the intrusive rocks which they enclose.[52] Dr. Voit[53] has, moreover, arrived at similar conclusions in the course of his investigation into the structure of the crystalline complex of the northern Transvaal, the constitution and mutual relationships of the Swaziland system and the older granites and gneisses being closely paralleled in the rocks of Nigeria.

THE INTRUSIVE ROCKS.

Many of the thin bands of well-foliated augen and granitoid gneiss enclosed within the metamorphic series undoubtedly represent original granitic intrusions, but these are so intimately bound up with the adjoining gneisses and schists, that no discrimination can yet be attempted. Attention has been more particularly directed in the field to those granitoid rocks which form the prominent inselberge and isolated groups of hills, and amongst them it has been found possible to distinguish an older from a younger series of intrusions. The older intrusive rocks consist very largely of granite, while the younger intrusions include granitic, syenitic and diabasic types with numerous associated dyke rocks. To what extent the epidiorites, amphibolites, and granulites of the gneissic series represent more basic intrusions associated with the older granites is quite unknown.

(_a._) _The older Granites._

The older granites are never sharply marked off from the adjoining gneisses, and invariably show a certain amount of foliation in the field, and of cataclastic deformation and reconstruction under the microscope. They appear to have been intruded at different times within the period of the evolution of the gneisses, and the amount of reconstruction which they themselves have undergone may be taken as a general indication of their age. Vermicular micropegmatite is most extensively developed in those schistose granites which have stopped short of complete reconstruction and which still afford abundant evidence of cataclastic deformation. It may also occur, however, in small amount in thoroughly foliated and reconstructed granites, in which it is assumed to have been developed, as in the case of the banded and striated gneisses (p. 118), by minor crustal strains, after the period of maximum reconstruction.

The most highly altered types form extensive lenticular or phacolitic masses of well-foliated and fairly coarse-grained biotitic and hornblendic gneisses (849, 481), containing streaks and patches of darker-coloured and finer-grained basic material. Under the microscope they possess a coarsely granular and thoroughly reconstructed character, and in the mutual relations of their constituents present a considerable resemblance to the banded gneisses of similar composition. The quartz and felspar show much later straining with incipient granulisation and traces of micropegmatite. These rocks have participated in all the changes which the adjoining gneisses have undergone, and were apparently intruded before the period of maximum reconstruction of the latter.

The least altered of the older granites (806, 1560) are typically distinguished by the irregular batholitic character of the intrusions and the partial and frequently only marginal character of the foliation. They sometimes have a tendency to become finer-grained towards their junction with the gneisses, and they usually enclose portions of the adjoining rocks and vein and inject with granitic material the marginal gneisses and schists. The inclusions may be of either the harder or softer gneisses according to the character of the rocks which the granite has invaded. In places a parallel arrangement of the felspars appears to have been induced through streaming during crystallisation. The partial foliation is parallel in a general way with that of the surrounding gneisses and schists. The granites vary much in character from biotitic and hornblendic rocks with syenitic modifications to muscovite and microcline granites and aplites. The quartz and felspar are strained throughout and cataclastic granulitisation is common on the margins of the crystals. Vermicular quartz is most abundant in the marginal and more distinctly foliated portions. These rocks would appear to have been intruded after the period of maximum reconstruction of the gneisses, but before the final cessation of stress and strain. In places, as at Abuja, crush lines cross the granites, and in their immediate neighbourhood the rocks show a higher degree of reconstruction and a greater development of micropegmatite in the matrix.

These later members of the older granites appear to have produced little or no contact metamorphism in the surrounding rocks. Brecciation and injection structures on the other hand are very frequent. The hard banded gneisses on the margin of the Abuja granite, for example, are abundantly injected with quartz and quartzo-felspathic material. Similarly at Zwoll, in Bauchi province, a porphyritic biotite granite has brecciated a series of banded and contorted biotite and hornblende gneisses, and many of the enclosed fragments are partially absorbed or minutely injected with granitic material. The Okuruku granite in southern Kabba, on the other hand, has invaded and brecciated members of both the harder gneisses and the softer series of biotite schists, magnetite schists and amphibolites. In the neighbourhood of the inclusions the granite becomes knotted and richer in biotite, hornblende, sphene and other minerals, while the inclusions themselves are irregularly veined and minutely penetrated by quartzo-felspathic material. In every case it would appear that the rocks, which the granites have brecciated and invaded, possessed their gneissic or schistose character before the actual intrusion of the granites, and that they, as well as the larger inclusions, suffered little further change as the result of the stresses and strains which brought about the partial foliation of the granites. The gneisses as a whole show strain phenomena, but only traces of granulitisation and of secondary formation of micropegmatite, while the granites show extensive marginal formation of micropegmatite, together with intense cataclastic deformation. It would seem indeed that these granites were intruded after the period of maximum reconstruction of the gneisses, but before the final cessation of stress and strain, and that while the gneisses which had originated under conditions of maximum strain were comparatively stable under similar strains of diminishing intensity, the granites which had crystallised from igneous fusion were in a condition of unstable equilibrium with respect to the stresses and strains to which they were subjected after crystallisation. The gneisses, as it were, possessed a certain amount of elasticity and transmitted the later strains with little resultant deformation, while the rigid masses of granite arrested those strains and suffered in consequence a partial deformation. The production of a strain-slip foliation in some of the more micaceous beds is probably to be referred to the same period of differential deformation of the granites. These later strains became locally intensified along crush lines, and in the case of the Abuja granite produced granulitisation both of the granite and of the felspathised gneisses in the neighbourhood of the crush line.

In addition to the two extreme types of the older granites there are many elliptical masses of similar material, elongated in a direction parallel with the strike of the gneisses, which show a decided foliation throughout, sometimes more and sometimes less intense (48, 1223). As a rule they are indistinctly marked off from the adjoining rocks and preserve their normal texture up to the margin of the intrusion. The granites are predominantly biotite granites, more rarely muscovite biotite granites, and the felspars are frequently porphyritic. Under the microscope they show a higher degree of reconstruction than the least altered granites, but a less complete reconstitution than the most highly altered types. The felspars vary from oligoclase to microcline, and the plagioclases often show basic interiors and zoned margins. Both the felspar and the quartz show abundant strain phenomena, and the felspars are usually partially surrounded by a secondary fringe of vermiculate micropegmatite in optically discontinuous grains. Frequently also the felspars are corroded or irregularly penetrated by quartz or micropegmatite, while the matrix consists largely of granular and partially recrystallised felspar, muscovite, microcline and quartz. The darker patches within these granites may be taken to represent portions of the adjoining rocks caught up during intrusion and partially absorbed, recrystallised and foliated with the granites. These rocks may possibly be correlated in the period of their intrusion with the least altered granites and regarded as the smaller and therefore more readily reconstructed representatives of the larger partially foliated masses. On the other hand, in the absence of the inclusions of adjoining gneisses which form the characteristic feature of the least altered granites, it is possible to regard them as of somewhat earlier origin and as having participated to a considerable extent in the general reconstruction of the gneisses themselves.

It seems probable indeed that the crustal movements which brought about the metamorphism of the gneisses were of a somewhat spasmodic character and spread over a very considerable interval of time and that the intrusion of granitic and other igneous material took place at intervals throughout the whole period. No hard and fast line can therefore be drawn between the highly reconstructed and gneissose granites and the imperfectly or only partially foliated types, a complete series of intermediate varieties being apparently possible between the two extreme members of the series.

(_b_) _The Younger Intrusions._

1. Granites and pegmatites.

Amongst the later granites, which present for the most part a batholitic habit, two types may be distinguished, differing in mineralogical composition and probably also to some extent in age. The rocks of the first type never show chilled margins, and were evidently intruded while the adjoining gneisses still possessed a comparatively high temperature. They include non-foliated aplitic granites and pegmatites usually rich in muscovite and tourmaline. The second type includes riebeckite and biotite granites which are sharply marked off from the adjoining rocks by felsitic and fine-grained margins. The rocks of the first type usually exhibit under the microscope a faint straining of the quartz and felspar and were probably intruded before the final cessation of the crustal movements to which the reconstruction of the gneisses and the foliation of the earlier granites were due. The members of the second type, on the other hand, are practically free from cataclastic phenomena and evidently assumed their present position after the cessation of movement within the crust. There is no reason, however, to assume any great difference in age between the two types or any marked discontinuity in magmatic conditions. The same magma, of a somewhat acid and alkaline character, was probably present in the depths throughout the whole period of intrusion, and the varying character of the products may be taken as affording some clue to the nature of the differentiation within the magma itself.

The alkaline granites of the first type are pale-coloured granular rocks composed almost entirely of quartz, microcline, orthoclase, perthite, and acid plagioclase. Biotite and other ferro-magnesian minerals are comparatively rare, while muscovite and tourmaline are frequently abundant and in large primary crystals. The quartz and felspar have a tendency to pegmatitic intergrowth, and sometimes masses of graphic pegmatite occur porphyritically in a normal quartzo-felspathic matrix. The tourmaline and muscovite may occur also as at Igbo in two generations, in large crystals and in the ground mass. Frequently also these granites possess a tendency to a coarsely banded structure, the banding being due either to variation in size of grain of the quartz and felspar or to variation in the abundance of tourmaline and muscovite in successive bands. These granites rarely form intrusions of any size, the largest known being the tourmaline granites of Igbo and Darroro, but even they are insignificant in comparison with the alkaline granites of the second type.

The pegmatite dykes and veins which are so abundant throughout the gneissic series evidently belong to the same period of intrusion. Like the aplitic and tourmaline-bearing granites they never exhibit fine- grained margins at their junction with the gneisses. They vary much in composition, and include graphic granites, quartz-felspar pegmatites with biotite, magnetite or ilmenite, quartz-felspar pegmatites with muscovite or tourmaline or both, and quartz-muscovite pegmatites with or without tourmaline. At Eri there are stanniferous pegmatites carrying quartz, muscovite and cassiterite, but no tourmaline. The massive quartz-reefs which frequently enclose portions of the surrounding schists and carry tourmaline, pyrites and sometimes gold, are also to be referred to the same period of intrusion. Inseparable from them and from the pegmatites are the smaller veins of quartz and tourmaline quartz and the schorl rocks of a foliated or striated character which occur as a rule in association with the veins of tourmaline quartz.[54] The schorl rocks with a gneissic habit in which the quartz is minutely granulitic and the tourmaline is in small crystals evidently recrystallised and full of quartz inclusions, are probably altered gneisses and schistose granites in which the felspars have been replaced by quartz and tourmaline. Traces of decomposed felspathic material are to be found in every specimen, but no pseudomorphs in tourmaline of felspar or of other minerals have been definitely recognised. The striated schorl rocks in which the darker bands are formed of finely fibrous tourmaline, appear to arise also as modifications of the adjoining rocks in the vicinity of the tourmaline quartz veins. Sometimes a foliated rock may be traced through a striated type into a massive tourmaline rock on the margin of the vein in which the original structures have entirely disappeared, and in which the original minerals, including the quartz, have been entirely replaced by finely fibrous tourmaline. In the quartz veins, pegmatites and granites, on the other hand, the tourmaline is in well-formed crystals of primary origin. It would seem, therefore, that the pneumatolytic or metasomatic formation of the schorl rocks was brought about by emanations from the quartz veins and pegmatites before the final consolidation of the latter. The extensive development of tourmaline and pyrites in the quartzites and schists of certain areas is also to be ascribed to the circulation of mineralising vapours and solutions throughout the adjoining rocks during the same period. The remarkable absence of tinstone in the tourmaline pegmatites, together with its appearance in pegmatites at Eri which do not carry tourmaline, seems to point to the stanniferous pegmatites having originated at a somewhat later date than the normal tourmaline-bearing pegmatites. This suggestion is supported by the occurrence of tinstone in association with the soda granites of Bauchi province.

The alkaline granites of the second type are typically distinguished by possessing fine-grained porphyritic and felsitic margins which indicate a considerable difference in temperature between the granites and the surrounding gneisses at the period of intrusion. To this group are referred the biotite granite of the Anagoda hills in S. Nassarawa, the riebeckite granite of the Gurkawa Hills in northern Muri, the biotite granite of Bukuru and Ngell, and the riebeckite granites of the Kwandokaya Hills and of the Ningi-Burra massif in Bauchi province, the riebeckite granite of Fagam, Gadama and Galambi on the borders of Bauchi and Kano, the riebeckite granite of Shira in the sub-province of Katagum, and probably also the Kailema granite near Jebba and the biotite granite of Abara near Ngashki in western Kontagora. The alkaline character of the original magma is reflected in the nature both of the felspars and of the ferro-magnesians. The felspars are much kaolinised and consist chiefly of tabular crystals of untwinned orthoclase probably rich in soda, and of streaky perthitic intergrowths of orthoclase and albite. Where enclosed in quartz, the terminations of the crystals are frequently idiomorphic. Microlites of riebeckite are common inclusions and small plagioclases of an earlier growth are frequently enclosed in the later felspars and in the quartz. The quartz forms aggregates of irregular grains showing faint strain shadows and has a great tendency to form coarsely pegmatitic intergrowths with the felspar. Where biotite alone is present, it is of a normal greenish-brown colour. Riebeckite, however, is the most characteristic ferro-magnesian, and occurs for the most part in irregular plates moulded on the felspars and sometimes enclosing small crystals of felspar in poikilitic fashion. Occasionally, as in the Shira granite, the riebeckite builds elongated crystals more or less idiomorphic in the prism zone and frequently intergrown with ægirine. Crystals of riebeckite over an inch in length are common in knots in the Kila granite, while on the other hand the margin of the Gurkawa granite is stippled with minute porous crystals of the same mineral. Associated with the riebeckite in places are other blue and brown soda amphiboles[55] as well as green ægirine and a golden-coloured mica, while zircon, apatite and iron ores are constant but scanty accessories. Pyrites is sometimes abundant in the porphyritic and felsitic marginal rocks.

A remarkable feature of some of these later granites is the occurrence in them of tinstone and numerous sulphide ores. It would appear that in places these later granites have been broken and fissured, and that along these fissures vapours and solutions have risen which have chloritised and mineralised the adjoining rock and in part refilled the fissures with ore-bearing pegmatite and vein stuff. In the Bukuru granite the cassiterite is associated in the neighbourhood of Ngell with pyrites, chalcopyrite, tetrahedrite, blende and galena. Topaz and brown zircon are found accompanying the cassiterite in the tin-bearing alluvium at Naraguta and elsewhere, and are probably also of pneumatolytic origin. The granites of the Kwandokaya hills, of the Jengre and Limoro hills, of the Ningi-Burra massif, and of the Gadama and Fagam hills also yield in places a tin-bearing drift, and have probably suffered, like the granite of Bukuru, a local fissuring and mineralisation. Tourmaline has not been definitely recognised as a product of these later emanations.

Similar soda granites with microgranitic and felsitic margins have been described by MM. Foureau, Gentil and Lacroix[56] from Zinder, by M. Chudeau[57] from Mounio and Goure, by M. Garde[58] from Machina and by M. Hubert[59] from Dahomey, while Captain Freydenberg[60] has noted the occurrence of a soda syenite with granitic modifications at Melfi to the south-east of Chad.

PLATE X

2. _Syenites and Diorites_

These rocks, which are by no means abundant, occur as small bosses or intrusive masses within the gneisses and schists. As a rule their junction with the surrounding rocks is not exposed, and their relationship to the later granites has not been definitely ascertained. They should probably be regarded, however, as somewhat basic modifications or differentiation products of the later granitic magma. Slight strain effects are usually apparent in the quartz and felspar.

The low hills around the town of Bauchi are composed of a coarse-grained augite syenite, in which orthoclase felspar predominates in large twinned crystals with a peculiar brownish tint and a resinous lustre. Under the microscope (49) subordinate oligoclase is present in smaller crystals, while the knots of darker minerals are composed of intergrowths of violet-coloured augite, green hornblende, brown biotite and iron ores. Quartz is interstitial and occurs also in secondary vermicular growths on the margins of the felspars. The accessories are zircon and apatite, and the whole rock is much stained with secondary limonite. A rock of similar appearance from Kende to the east of Bauchi (69) is somewhat more basic and may be termed a quartz augite-diorite. The predominant felspar is a basic andesine, while the ferro-magnesians are pale-coloured augite and green hornblende in irregular plates and intergrowths. The quartz is interstitial or intergrown with the felspar in pegmatitic fashion. A similar rock from Kanna (1677) contains much biotite in addition to the augite and hornblende. The rock of Kogon Dutsi (1338) within the walls of Kano is a basic augite diorite without quartz, consisting predominantly of acid labradorite and pale-coloured augite. Both the augite and felspar have a tendency to schillerisation and the augite is accompanied by, and in part intergrown with, green hornblende, brown biotite and iron ores, while hypersthene is a scanty accessory. An augite-diorite from eastern Kontagora (1302) contains a little interstitial quartz and felspars varying from oligoclase to acid labradorite. The ferro-magnesian minerals are much intergrown and include pale-coloured augite, green hornblende and brown biotite.

3. _Porphyries, Porphyrites, Diabases and Basalts._

The later sills and dyke rocks are most abundant in the central provinces and especially in Bauchi, eastern Zaria and northern Nassarawa, but are somewhat sparingly distributed throughout the remainder of the Protectorate. They include quartz porphyries, orthoclase porphyries, mica and hornblende porphyrites, diabases and basalts. In the central area the high content in soda which characterises the more acid types undoubtedly indicates a genetic connection between the later soda granites and the dyke rocks, although investigation has not yet proceeded far enough to allow of the recognition of the major foci to which the dyke rocks may be referred. Amongst the quartz porphyries micro-granitic, granophyric and felsophyric types are equally common. In the central provinces riebeckite is frequent both among the phenocrysts and in the ground mass, while elsewhere biotite is the commonest coloured constituent. An ægirine granophyre occurs at Chikobo (1417) on the borders of Bauchi and Zaria. Orthoclase porphyries are commonest in the central area and possess typically a trachytic ground mass, of which riebeckite and ægirine are frequent constituents. Augite syenite porphyries are known from the Kudaru Hills in eastern Zaria (1397) and from the vicinity of Limoro on the borders of Bauchi and Zaria (1413). The latter occurrence contains hornblende and biotite in addition to the augite, as well as a small amount of quartz in the ground mass. An extensive mass of granophyric diorite porphyrite occurs in western Kontagora between Auna and Bussa (1162). The felspars are turbid and much epidotised and surrounded by a fringe of more acid material which passes outwards into the granophyric matrix, while smaller irregular crystals of green augite and ilmenite and abundant minute plates of green mica are also enclosed within the pegmatitic material. A peculiar rock from Toro in Bauchi province (958) may be termed a quartz diorite porphyrite, and presents a certain affinity with the basic members of the charnockitic series of the Ivory Coast.[61] The phenocrysts have a tendency to aggregate in groups and include rounded quartz grains, tabular crystals of andesine and labradorite with inclusions of granular augite, elongated crystals of hypersthene and irregular plates of biotite much corroded by the ground mass and enclosing grains of augite and apatite. The ground mass is composed of granular green augite, felspar and quartz, with much magnetite and apatite. The felspar of the ground mass is frequently untwinned and more acid than the phenocrysts, while the quartz forms in places micropoikilitic plates enclosing the other ingredients of the ground mass. As alkali-syenite porphyries may be grouped the bluish-grey trachytoid rocks of Kanikoko (1217–1219). The phenocrysts include tabular plates of sanidine, microcline, microperthite and microcline- microperthite and fairly well-formed crystals of diopside and enstatite, much intergrown with brown mica and bluish amphibole. The ground mass consists of microlites of alkali felspar and granules of the ferro- magnesian minerals. Typical mica porphyrites with phenocrysts of biotite and zoned andesines occur at Barazara (1274), and between Dan Gerumfa and Gwashi (1276) in eastern Kontagora. A hornblende porphyrite between Naraguta and Tilde (956) is inconspicuously porphyritic and composed of allotriomorphic green hornblende with subordinate biotite and augite in a matrix of acid plagioclase, orthoclase and quartz. A hornblende porphyrite from Borgu (1220) contains large poikilitic crystals of augite and hornblende in a granular matrix of labradorite, augite, hypersthene, hornblende and iron ore. A somewhat similar rock from Bauchi province (1602) contains large poikilitic hornblende crystals in a matrix of granular augite, magnetite and lath-shaped crystals of labradorite. A labradorite porphyrite from Toro in Bauchi province contains large tabular phenocrysts of acid labradorite in a ground mass of green augite, brown biotite, secondary green hornblende and lath- shaped crystals of acid plagioclase. The diabases and basalts of the central area are particularly numerous round the margin of the Bauchi plateau and the Nassarawa tableland, and probably include intrusions of many different ages. Some of the coarser-grained diabases present affinities with gabbros in the idomorphism of their augite crystals (47, 1695). Others exhibit a typical ophitic structure (962, 985), while others are predominantly granulitic (690). A hornblende diabase with pale green augite, greenish-brown hornblende and deep red biotite occurs in the Ruruma hills (1466) in eastern Zaria. Biotite also accompanies the augite in an ophitic diabase from the Kudaru hills (1390) in north- east Zaria. Uralite diabases are fairly common, but olivine diabases (747) are conspicuously rare. The basalts present few features of interest and are typically compact, non-porphyritic rocks with a holocrystalline, ophitic or granulitic structure and free from olivine.

MM. Gentil[62] and Chudeau[63] have described the occurrence of intrusive granites, pegmatites, porphyries, lamprophyres, diabases and gabbros, throughout the crystalline areas of the Central Sahara, and Dr. Passarge[64] has recorded similar intrusive rocks from the northern gneissic region of Adamawa. Ægirine rhyolites and alkaline trachytes of an intrusive character are known also from Zinder, Gabana, Hadj el Hamis,[65] and M’Burao.[66] M. Chudeau believes that the alkaline granites of Mounio with their associated hypabyssal types pierce the sandstones of the Tegama, and would therefore refer all the known occurrences of soda granite in Central Africa to a late Cretaceous or Tertiary age. In Nigeria, however, there is little or no evidence to support this view. The soda granites are nowhere found piercing sedimentary rocks within the Protectorate, and while it is possible that they may have been originally intruded into rocks of Cretaceous age which have now been removed, it is remarkable that, with the exception of rare sills of dolerite and the Tertiary pipes of basalt, no intrusive dykes or masses of hypabyssal or plutonic material are to be found within what remains of either Cretaceous or Tertiary rocks within the Protectorate. Such intrusive rocks, however, are particularly numerous within the crystalline areas, and especially in the neighbourhood of the larger masses of soda granite, which are themselves frequently pierced by marginal dykes of basalt and porphyry. In view of the abundance of soda types among the dyke rocks, moreover, it is hard to escape the conclusion that the later non-foliated alkaline granites of the second type (p. 132) with the associated syenites, diorites and dyke rocks of acid, intermediate and basic composition, all possess a certain genetic relationship, and should be properly ascribed to one and the same period of igneous activity. All the facts, indeed, are more readily explicable on the assumption of the pre-Cretaceous age of this period of intrusion within the Protectorate. The stratigraphy of Mounio and Koutous is somewhat doubtful, and M. Chudeau’s section[67] of the faulted margin of the Goure granite does not indicate an actual transition from the mica schists and quartzites into the sandstones and clays of the Tegama. M. Garde,[68] moreover, while indicating the occurrence of crystalline gneisses to the north of Goure does not appear to confirm M. Chudeau’s interpretation of this important section. It should be remembered, however, that even although further investigation may prove the truth of M. Chudeau’s correlation, there is no reason to believe that the intrusion and effusion of igneous rocks rich in alkalis has been confined, as M. Chudeau would suggest, to one particular period in the history of the Sudan.

It is interesting to trace in a general way, and subject to the limitations indicated above, the successive changes in the character of the igneous intrusions from the earlier to the later granites and dyke rocks on the hypothesis of a continuity in magmatic conditions, or at least of a periodic renewal of activity in the same igneous focus. The successive intrusions may be tabulated as follows:—

1. Earliest well-foliated granites of intermediate composition with
numerous associated amphibolites and epidiorites.

2. Partially foliated granites of both acid-alkaline and intermediate
character, with partially sheared dyke rocks of acid, intermediate and
basic composition.

3. Aplitic granites and pegmatites, rich in muscovite and tourmaline
with associated reefs and veins of quartz and tourmaline quartz.

4. Alkaline granites rich in soda with associated syenites, diorites,
and dyke rocks.

If the normal sequence of events within the period of activity of any igneous focus be taken to be the successive manifestation of volcanic, plutonic and hypabyssal activity, there may perhaps be recognised in the successive intrusions indicated above four periods of igneous activity, each of which is now represented by the products of the plutonic and hypabyssal phases only, while the products of the accompanying volcanic phases, if originally present, have been removed in the course of denudation. The tendency to the production of rocks rich in alkalis should be remarked as a notable feature of the various periods of intrusion; and in this connection it is interesting to note that Messrs. Horwood and Wade[69] have likewise emphasized the alkaline character of many of the earlier granitic intrusions to the south of the equator.

[Footnote 28: Hubert, _Mission Scientifique au Dahomey_, 1908, p. 448.]

[Footnote 29: Chudeau, _Sahara Soudanais_, 1909, p. 18.]

[Footnote 30: The numbers in brackets refer to microscopic preparations in the Imperial Institute Collection, NN. F.]

[Footnote 31: Parkinson, _Q. J. G. S._, Vol. LXIII, p. 315.]

[Footnote 32: _Mem. Geol. Sur. Scotland_, Expl. of Sheet 55, 1905, p. 13.]

[Footnote 33: Van Hise, _Treatise on Metamorphism_, 1904, p. 696.]

[Footnote 34: Parkinson, _Q. J. G. S._, Vol. LXIII, 1907, p. 308; Vol. LXIV, 1908, p. 313.]

[Footnote 35: Chudeau, _Sah. Soud._, p. 2.]

[Footnote 36: Gentil in _Doc. Sc. Miss. Sah._, 1905, p. 721.]

[Footnote 37: Haug, _C. Rd. Ac. Sc._, 7 Aug., 1905.]

[Footnote 38: Haug in _Doc. Sc. Miss. Sah._, 1905, p. 753.]

[Footnote 39: Flamand, _C. Rd. Ac. Sc._, 140, p. 954.]

[Footnote 40: Flamand in Voinot, _Bull. Com. Afr. Fr._, 1908, p. 218.]

[Footnote 41: Hubert, _C. Rd. Ac. Sc._, 146, 3 Feb., 1908; _Miss. Sc. au Dahomey_, 1908, p. 253.]

[Footnote 42: Chautard, _Le Fouta Djalon_, 1905.]

[Footnote 43: Parkinson, _Q. J. G. S._, 1908, LXIV, p. 313.]

[Footnote 44: Chevalier, _La G._, XVII, 1908, p. 201.]

[Footnote 45: Lenz, _G. M._, 1877, p. 27; Knox, _Geology of Africa_, 1905, p. 93.]

[Footnote 46: Von Ammon, _Geologie von Togo_, _M.H. Geog. Gessell. Mün._, 1903, I, p. 393.]

[Footnote 47: Parkinson, _Q. J. G. S._, 1907, LXIII, p. 308.]

[Footnote 48: Courtet in Chevalier, _L’Afr. Cent. Fr._, 1908, p. 646.]

[Footnote 49: Passarge, _Adamaua_, 1895, p. 312; Esch, _Geol. von Kameruns_, 1904, p. 23.]

[Footnote 50: Barratt, _Ann. des Mines_, VII, 1895.]

[Footnote 51: _Cf._ Horwood and Wade, _G. M._, V, 6, 1909, p. 455.]

[Footnote 52: “The Geology of the North-West Highlands of Scotland,” _Mem. Geol. Sur._, 1907, pp. 75, 214, 595, 607.]

[Footnote 53: Voit, _Trans. Geol. Soc. S. A._, 1905, VIII, pp. 106, 141.]

[Footnote 54: _Cf._ Flett, “Geology of Land’s End District,” _Mem. Geol. Sur. Eng._ 1907, p. 25.]

[Footnote 55: _Cf._ Gentil et Freydenberg, _Bull. Soc. Geol. Fr._, 4th Ser., VIII, 1908, p. 44.]

[Footnote 56: Gentil, _C. Rd. Ac. Sc._, 8 Aug., 1904; Foureau and Gentil, _C. Rd. Ac. Sc._, 2 Jan., 1905; Lacroix, _C. Rd. Ac. Sc._, 1 May, 1905; Gentil, _Doc. Sc. Miss. Sah._, 1905, p. 697.]

[Footnote 57: Chudeau, _Sah. Soud._, 1909, p. 265; _C. Rd. Ac. Sc._ 1 July, 1907.]

[Footnote 58: Garde, _C. Rd. Ac. Sc._, 149, 5 July, 1909, p. 43.]

[Footnote 59: Hubert, _C. Rd. Ac. Sc._, 145, 1907, p. 764.]

[Footnote 60: Freydenberg, _Chad et Shari_, 1908, p. 107; Gentil et Freydenberg, _Bull. Soc. Geol. Fr._ (4), VIII, 1908, p. 44.]

[Footnote 61: Lacroix, _C. Rd. Ac. Sc._, 150, 1910, p. 18.]

[Footnote 62: Gentil, in Foureau’s _Miss. Sahar._, 1905, p. 720.]

[Footnote 63: Chudeau, _Sah. Soud._, 1909, p. 256.]

[Footnote 64: Passarge, _Adamawa_, 1895, p. 382.]

[Footnote 65: Garde, _C. Rd. Ac. Sc._, 5 July, 1909 (149), p. 43; _cf._ Gentil and Freydenberg, _Bull. Soc. Geol. Fr._ (4), VIII, 1908, p. 44; _cf._ also Suess, _The Face of the Earth_, Vol. IV, _Oxford Trans._, 1909, p. 283.]

[Footnote 66: Hubert, _C. Rd. Ac. Sc._, 1904 (139), p. 378.]

[Footnote 67: Chudeau, _Sah. Soud._, 1909, p. 267.]

[Footnote 68: Garde, _C. Rd. Ac. Sc._, 149, 5 July, 1909, p. 43.]

[Footnote 69: Horwood and Wade, _Geol. Mag._ V, 6, 1909, p. 455.]