The Basic Igneous Rocks of Eastern Otago and Their Tectonic Environment, Part 5.
The Quartz Porphyry of the Shag Valley and Its Relation to the Mid-Tertiary Basaltic Magma of North-East Otago; A Petrogenetic Essay.
By W. N. Benson.
[Read before Otago Branch, Royal Society of New Zealand, November 13, 1945; received by the Editor, February 13, 1946; issued separately, June, 1946.]
|Introduction: Occurrence, Tectonic Environment, and Age of the Shag Valley Quartz Porphyry||1|
|Petrology of the Quartz Porphyry and of its Xenoliths||9|
|The Chemical Composition of the Quartz Porphyry||13|
|The Origin of the Quartz Porphyry of Shag Valley.|
|A Comparative Petrogenetic Discussion.|
|(A) Hypothesis of Derivation by Fractional Crystallisation-differentiation from Uncontaminated Basaltic Magma||14|
|(B) Hypotheses involving Assimilation, with Differentiation of the Syntectic Melt, and/or Transfusion||24|
|(C) Hypotheses involving Re-fusion, Anatexis or Palingenesis||29|
Introduction: Occurrence, Tectonic Environment, and Age of the Shag Valley Quartz Porphyry.
The general occurrence of the quartz porphyry was described by its discoverer, McKay (1887), who showed it as extending for a little more than a mile northwards from a point on the Old North Road near the foot of the Horse Range and two and a-half miles northeast of Palmerston. His work thereon was revised recently by
Paterson (1942). The general geographical relation between the quartz porphyry of Shag Valley and the coastal districts of Otago is shown in the writer's map (Benson, 1941, Plate 36), and in particular its geological relation to the immediately adjacent rock-formations is displayed on Figure 1 herewith, which is based on Paterson's (1942, Plate 7) map and section as modified by the present writer. McKay showed that two major faults run south-eastwards near the foot of the Kakanui Mountains or Horse Range, and considered that the quartz porphyry was injected into the more westerly of these two fissures. Paterson recognised that the quartz porphyry ran between the two faults mapped by McKay and himself, but did not notice that it occupied there a third fault-fissure, as is shown on Section YY in our Figure I, since no attempt was made by him to map separate units in the “Coal Measures” of his stratigraphical classification. Nor was this done by the present writer, the relations being readily described in general terms.
As will be seen from Figure I, the basement formations are schists to the south-west of the Cretaceous and Tertiary formations which occupy the lower portions of the Shag Valley, and less metamorphosed greywackes, semi-schists and phyllites, rising from beneath these covering formations to form the Horse Range. The sequence of Cretaceous and Tertiary formations in the lowland regions may be briefly described thus: A basal series of Upper Cretaceous (Upper Senonian) coal-measure conglomerates, etc., about 1200 ft. thick occurs on the north-eastern side of the valley, but is only about 100 ft. thick on the south-western side of the valley. Outliers of the basal portions of these coal-measures extend south-westward up the slopes of the schist, and north-eastward up those of the greywacke, etc., rocks forming the Horse Range. The stratigraphic subdivisions of these Upper Cretaceous rocks are distinctive and, being significant for our inquiry, are detailed below. Conformably on the Coal Measures there rest the “Katiki Beach” Sandstone and above the more or less glauconitic mudstones which together form the “Abbotsford” beds of Lower Eocene age and 600–900 feet thick. Upon them follow with apparent conformity the local equivalent of the “Green Island” Loose Sandstone about 120 feet thick, finer in grain-size and much more argillaceous than that of the type area near Dunedin, but containing the same characteristic assemblage of heavy minerals (Paterson, 1942, pp. 41, 44). Above this again is the Upper Eocene (Tahuian) “Burnside” Mudstone. All these formations occur within two miles of the porphyry intrusion. Elsewhere, chiefly south of the Shag River, the “Burnside” Mudstone is followed disconformably by the Lower Miocene “Caversham” Sandstone and overlying “Goodwood” Limestone, with which we are but little concerned. In compiling this generalised summary use was made of the observations of Brown (1938), Paterson (1942) and the recent age-determinations kindly communicated by Finlay, and by Finlay cited by Benson (1943).
Fig. 1.—Geological map (after Paterson, 1942, as modified by Benson) and section across the lower Shag Valley.
Greywacke and Argillite.
Senonian Coal Measure Conglomerate, etc.
Danian Concretionary Mudstones.
Lower (?) Eocene Mudstones.
Mid Eocene (?) “Loose” Sandstone.
Upper Eocene Mudstone.
Oligocene (?) Quartz Porphyry.
Late Tertiary Tuff and Basalt.
100–120 ft. terrace.
40–60 ft. terrace.
20–30 ft. terrace.
The Upper Cretaceous (Senonian) Coal-Measure conglomerate in the neighbourhood of the quartz porphyry is about 1200 feet thick. As shown by McKay and Paterson its lowest portion (A) is a very massive irregular aggregate of more or less angular fragments several inches long of greywacke, semi-schist and schist, of phyllite and quartz pebbles derived from fault-scarps of the pre-Cretaceous formations which rose into relief during the strong early Cretaceous (?) crust-movements, and were deposited more or less as semi-talus or fan-conglomerate. A vivid and unconventionally humorous description of this portion of the Coal-Measures was given by McKay (1892). It is several hundred feet thick and passes up into (B) more or less friable quartz sands containing thin beds of quartz pebbles. Following on these and forming the upper portion of the Coal-Measure Conglomerates are: (C) a limonitic coarse-grained pebbly sandstone with abundant plant-fragments, (D) haematitic quartz conglomerate. (E) carbonaceous shales and coal-seams followed locally by (F) limonitic sandstone with rare Upper Senonian marine mollusca. It is not here asserted that the order C, D, E is the true sequence in the upper portion of the coal-measures near the porphyry, the essential point for our present purpose being that, whatever be the actual inter-relationship, the beds (C–F) all lie above (A) and (B).
The narrow belt between the quartz porphyry and the greywackes, etc., of the Horse Range is a syncline bounded on each side by N.W.–S.E. faults and pitching south-eastwards. Its northern extremity consists wholly of the semi-talus (A). Where the belt is crossed by the Old North Road (Section YY, Figure 1), the centre of this belt is occupied by the quartz sands (B) which here lie nearly horizontally in the trough of the syncline, and further to the south-east are covered by the beds (C), (D), (E), and (F), and these by the Lower Eocene beds, according to the information available in the literature. The north-eastern boundary of the quartz porphyry intrusion abuts against formation A, which dips north-eastward at the contact, except at one point, namely, in the pine-covered ridge 250 yards north of the road, where is exposed, though badly, a narrow (10–20 yards?) strip of deeply-weathered high-grade schist [Chl 3 on Turner's (1938) classification]. This strip of schist is bounded on the east by a fault-breccia composed of schist-fragments, north-east again of which is the downthrown, steeply dipping semi-talus.
Though the schist in this small inlier adjacent to the porphyry is not of quite so high a grade of metamorphism as that (Chl 4) which characterises the Central Otago schist exposed south-west of the Shag Valley in the area illustrated by Figure 1, the distinction is not great. There is no evidence available to show whether there is a gradual transition from the Chl 4 to the Chl 3 grade beneath the Cretaceous-Tertiary sediments in the Shag Valley, or whether here, as in the Waipiata district thirty miles to the north-north-west, a block of Chl 3 schist was faulted down into the basement of Chl 4, before the Cretaceous peneplanation which truncated both (Turner, 1939, p. 40, and Figure 10). Since the quartz porphyry rising through the Chl 3 schist contains abundant xenoliths of rocks displaying lower grades of metamorphism (Chl 1–2) such as may occur in the deeper portions
of the Kakanui Series which forms the Horse Range, it is evident that the rising magma came into contact with such low-grade material, and hence probably followed the major pre-Cretaceous fault which brought the two very different grades [Otago (Chl 4–3) and Kakanui (Chl 2–1)] of material into apposition, though the intrusion of the magma must have occurred in post-Cretaceous times and was probably accompanied by a revival of differential crust-movement along this member of a frequently rejuvenated series of faults. That the porphyry should have diverged obliquely from the major fault, to traverse the higher grade schist only in the upper part of its course, may perhaps be the result of the opening of a branch fault-fissure during this rejuvenation of fault-movement. (See Figure 1.)
Immediately south-west of the porphyry is a second belt of Cretaceous rocks, only about a hundred yards wide where it crosses the roadway. It consists of nearly horizontal or gently dipping limonitic pebbly sandstone with plant fragments (C) by the roadway and haematitic quartz conglomerate (D) further north. The porphyry thus occupies a fault-fissure between Cretaceous formations. It is difficult to see any clear evidence of the inclination of this fault, but from various hints it appears probable that it is a very steep reversed fault, with a westerly downthrow of varying amount, generally several hundred feet.
A similar qualitative description might be given of the western member of the pair of faults mapped by McKay and Paterson. It runs parallel to the western boundary of the porphyry and about a hundred yards from it where it crosses the roadway. On the western side of this fault, fortunately exposed by a recent clearing of the roadside cutting, is a white, finely-granular, richly argillaceous sandstone spangled with muscovite, macroscopically identical with the Shag Valley facies of the Green Island Loose Sandstone which was recognised by Paterson, who did not, however, note the occurrence of this rock in the immediate vicinity of the porphyry. The boundary of an area of this formation which he recognised less than half a mile from the porphyry has therefore been extended in our Figure 1 to indicate its probable continuity with the newly discovered occurrence. The well-grassed slopes of the areas underlain by Tertiary sediments rarely afford the opportunity for the exact mapping of their boundaries. Though Paterson found this Shag Valley argillaceous equivalent of the Green Island sandstone contained, though but very sparsely, the heavy minerals, notably andalusite and kyanite, characteristic of the Green Island sandstone near Dunedin, Dr. Turner was unable to recognise their presence in the specimen collected by the present writer from near the porphyry. So far no fossils have been recorded from the Green Island Sandstone. It is therefore of interest to note that Dr. H. J. Finlay recognised in the writer's specimen numerous siliceous micro-organisms, especially radiolaria and several of an as yet undescribed species of the foraminifer Bolivinopsis which he has found in the Palaeocene-Lower Eocene Abbotsford beds from the Katiki formation upwards in the coastal region between the Shag Valley and Hampden, and in coeval formations and in the Te Uri district of the North Island, but which does not extend into the Mid
or Upper Eocene Bortonian and Tahuian formations. Unless the outcrop by the roadway is of a facies of sediment within the Abbotsford Mudstone elsewhere unknown, it seems appropriate for the present to regard it as the equivalent of the “Green Island” Loose Sandstone of the Dunedin district, into which the Abbotsford Mudstone passes upwards by gradual transition, and to be of latest Lower Eocene age.
If this be correct, it will have an interesting bearing on the geology of the Dunedin district in that it will indicate the absence there of the Middle Eocene Bortonian formations, except as may be represented by the thick richly or purely glauconitic beds at the base of the “Burnside” mudstone, which may have accumulated there during a period of partial cessation of deposition of clastic sediments in the Dunedin district while the Bortonian sediments were accumulating north of the Shag Valley. Further, the throw of the westernmost fault in the Shag Valley will be therefore of the order of 800 feet, also with a westerly downthrow.
McKay followed by Paterson, mapped the quartz porphyry as a strip of almost uniform width (50–60 yards) extending from about 250 yards south of the roadway to about 2000 yards north thereof. The present writer found, however, that the intrusion was more nearly lenticular, reaching its greatest thickness of about 130–140 yards in the gully and spur-ridge 400–500 yards north of the roadway, but thinning out northwards, and could not be seen more than about 1000 yards from the road in this direction, though the continuation of the fault-fissure which it occupied was traceable by the contact of the Lower Coal Measure semi-talus (A) with the Upper Coal Measure haematitic quartz conglomerate (D) for at least a mile beyond the northernmost exposure of the porphyry.
It has not been possible to observe the contact of the porphyry with the formations which it invades. In several places, however, and especially in the spur-ridge 500 yards north of the roadway, the pale green ground-mass of the porphyry contains, though sparsely, small joint-bounded flaky xenoliths of dark grey micaceous phyllite comparable with the finer-grained or argillitic bands in the greywackes and more especially semi-schists of the Kakanui rocks of the Horse Range. These xenoliths may be as much as 8 mm. long, but range down to those of microscopic dimensions, the minute xenoliths being very abundant. These xenolithic flakes are more or less parallel to one another, their common direction being a plane of laminar flow in the enclosing porphyry, which is clearly marked by the orientation of the phenocrystic plates of biotite, and to a less extent by the more or less linear parallelism of the longer axes of the phenocrystic plagioclase-prisms, and by the streaks in the more or less glassy groundmass. This plane of laminar flow strikes about N.W.–S.E., i.e., parallel to the elongation of the intrusive mass of porphyry, but the northward-directed dips vary from about 20°–80°, are difficult if not impossible to measure by ordinary means, and the data obtained therefrom are insufficient to form an adequate basis for discussion of their structural significance. Nevertheless, they afford a suggestion that the intrusion as a whole forms a sheet which dips rather steeply
to the north-east, which accords with the view that it was injected into a steeply-dipping reversed fault-fissure under the influence of dominant pressure from the north-east. Indication of the effect of such pressure deforming the nearly rigid cooling mass is described below.
In considering the age of the Shag Valley quartz porphyry it will be best to commence with an outline of the geological history of the Shag Valley. It is immaterial for our purpose to consider whether the Otago Schists (Chl 4–3) are the most strongly altered and deeper portions of the same huge complex of Palaeozoic to Early Mesozoic sediments as yielded the Kakanui semi-schists, etc. (Chl 2–1) in their less deeply buried members during one period of metamorphism, or alternatively whether two sedimentary series are involved, with an intense pre-Mesozoic orogeny causing the high grade of metamorphism of the Otago Schists derived from Palaeozoic sediments, and a much less powerful orogeny at the close of Jurassic times changing a largely Lower Mesozoic series of sediments into the Kakanui formations, and affecting to a minor degree the previously strongly metamorphosed Otago Schists. The point is discussed by Turner (1939, pp. 42–3), the balance of probability being thought by him to lie with the latter hypothesis. Be that as it may, it is clear that very strong block-faulting occurred along the Shag Valley fault-zone and its north-westward continuation into the northern portion of the Maniototo region, before the peneplanation of Early-Middle Cretaceous times which produced a surface of low relief, which was, as it were, a large-scale mosaic of planed fault-blocks wherein rocks of very varying grades of metamorphism were brought into apposition, and the dominant fault-fissures were ranged in a zone running N.W.–S.E. Rejuvenation of the movement of these fault-blocks in Middle Cretaceous times produced fault-scarps, during the rapid destruction of which the semi-talus of the Lower Coal Measures was laid down north of the present Shag River, and the more normal clastic sediments of the Middle and Upper portions of the Coal Measures transgressed beyond the fault-angle depression in which they formed to extend far over the Cretaceous peneplain on either side of it, chiefly as terrestrial deposits, but containing in their highest portions Upper Senonian marine sediments. Deposition of the Palaeocene-Lower Eocene “Abbotsford” marine formations, with the “Katiki” beds at their base, and the “Green Island” Loose Sandstone at their top and after a period of non-deposition, the formation of the Upper Eocene “Burnside” beds followed in the Shag Valley, and southwards thence to Dunedin, but there may be traced northwards from Moeraki the Middle Eocene “Bortonian” mudstones laid down during this period unrepresented by clastic sediments south of the Shag Valley. Again in Lower Oligocene times there was a marked difference between the geographic conditions on either side of the Shag Valley. To the south of it sediments of this age are practically absent, and the Upper Oligocene “Caversham” Sandstone is separated from the Upper Eocene “Burnside” Mudstone (on which it rests disconformably) by a thin bed of glauconite only. North of the valley, however, there is a very extensive record of Oligocene forma-
tions. Submarine eruptions breaking out in the later portion of Eocene times (Benson, 1943, p. 130) formed of the thick basic Waiarekan tuffs, with intercalated more normal marine sediments, and a diatomaceous band, were overlain (near Oamaru) by the Otataran limestone, through which the basalt magma rose to form the pillow lavas and breccias of the Deborah group, more or less simultaneously with an extensive development of sills and dykes of dolerite invading formations in and beneath the Ototaran limestone. On the cessation of eruptive activity the Lower Oligocee “Kakanui” Limestone was laid down near Oamaru, and, with minor disconformities and lacunae, late Oligocene sediments were laid thereon. None of these formations with an aggregate thickness of the order of 1000 feet, are represented south of the Shag River and the facies of the Upper Oligocene (“Hutchinsonian” greensand) and the overlying “Awamoan” beds near Oamaru are very different from those of the coeval “Caversham” Sandstone and “Goodwood” Limestones immediately south of the Shag River, affording palaeogeographic evidence of noteworthy crust-movements probably along the Shag Valley fault-zone between Eocene and Miocene times. It is in this fault-zone that the intrusion of quartz porphyry occurs. If this intrusion was genetically related to the Mid-Early Upper Oligoeene activity of basaltic magma beneath north-eastern Otago, crustal movement in the Shag Valley fault-zone must have been vigorous, possibly more so than would suffice to explain the palaeogeographic distinction between the regions north and south thereof in Late Oligocene-Early Miocene times deducible from stratigraphical considerations only.
Further crust-movements occurred in later Miocene-Early Pliocene times, both folding and faulting occurring throughout Eastern and North-Eastern Otago, and especially the Shag Valley district (see Benson, 1941, Figure 2). The resulting differential relief then produced was obliterated by peneplanation during the remainder of Pliocene times, and in particular the quartz porphyry, which is not now associated with any topographic features such as fault-scarps which may have formed at the time of its intrusion, was probably exposed by the removal of the enclosing Cretaceous formations during this second peneplanation. But if the quartz porphyry magma rose into its present position during the Late Miocene deformation it must have had but a slight relation if any with the basic magma which was active beneath North-Eastern Otago in Oligocene times, unless the activity of the latter were maintained or renewed (though without overt expression) until this later period, which is perhaps not beyond the bounds of possibility. The occurrence of rhyolites, etc., resting on the eroded surfaces of Mesozoic-Lower Miocene formations and overlain by the Pliocene-Pleistocene (?) basalts of Banks Peninsula may be recalled (Speight, 1935), though these rhyolites occur in an environment with a geological history very different from that of the Shag Valley. Nevertheless, on both sides of this valley the Pliocene peneplain was covered by effusions of basaltic lavas of Late Pliocene-Pleistocene (?) age, developed before the Pleistocene foldings and faulting which again rejuvenated the Shag Valley fault-zone
elevating the Kakanui formations to produce the basalt-covered Kakanui-Horse Range. No evidence known to the writer is available to form a critical distinction between the relative probabilities of these two estimates, Late Oligocene and Late Miocene respectively, of the age of the quartz-porphyry, though the former is deemed to be the greater. The suggestion that the quartz porphyry rose into position at the close of Cretaceous times before any overt expression of the basaltic magma beneath North-Eastern Otago, namely, during the crust-movements accompanying the subsidence which permitted the accumulation of abnormally thick Cretaceous sediments in the Shag Valley fault angle depression would mean that it rose into its present position before the formation of the fault cutting off the Cretaceous sediments immediately west of it, and is without any supporting evidence, and the suggestion that it may have risen during the Pleistocene deformation seems excluded by the absence of any topographic expression of differential relief adjacent to the porphyry, and would, like the immediately preceding suggestion, involve the occurrence of igneous activity at a period during which no other igneous activity is as yet known to have occurred within hundreds of miles of the Shag Valley. Hence, of these four suggested ages for the quartz-porphyry that involving an Oligocene age, and hence approximate contemporaneity with the basaltic activity beneath North-Eastern Otago seems the most probable, though it has not yet been proved to be correct.
Recognising the uncertainty, though probability, both as regards the age of the quartz porphyry and its genetic relation to the Oligocene basaltic activity, the concluding petrogenetic portion of this paper, assuming tentatively that such a genetic relation exists, takes the occasion to review the current hypothesis concerning the genetic relations of acidic and basic igneous rocks, and to consider, on the basis of a widely distributed series of comparisons, what may be the mechanism of such genetic relationship in the case of the Shag Valley quartz porphyry and the basic igneous rocks of North-Eastern Otago in terms of each of the principal current hypotheses.
Petrology of the Quartz Porphyry and of its Xenoliths.
Paterson's (1942, p. 467) petrographic description of the quartz porphyry, based on the study of three sections, 5090–2 in the collection of the Otago University Geological Department, has been supplemented by a re-examination of these sections, of P.8662 in the collection of the Geological Survey cut from McKay's specimen of the intrusive rock, and of five other slides 5923–6 cut from the freshest rocks obtainable from the gully a quarter of a mile north of the Old North Road, together with 5922 cut from a whitish rock showing a peculiar form of alteration occurring within a hundred yards north of this road. The bulk of the rocks have a pale greenish to grey-buff colour with a porcellanous groundmass, containing phenocrysts of quartz, feldspar and biotite, and numerous small vesicles. A laminar flow-structure is shown by the parallelism of the mica flakes and, in sections cut perpendicular to this (e.g., 5926), by the approximate
alignment of the long axes of the colourless phenocrysts and of the small xenoliths, and by the faint streaking of the vitreous groundmass.
The xenoliths in the quartz porphyry were not recognised by Paterson, and when seen in micro-section are rarely more than a millimetre long, and consist chiefly of fine-grained argillite probably derived from the pebbles coming from the higher members of the Kakanui rocks in the Horse Range, which were incorporated in the lowest portion of the Coal Measure conglomerates. Less often they are of fine-grained quartzose schists which have attained about the Chl 2 grade of metamorphism according to Turner's (1938) classification, and were probably derived from the more metamorphosed lower portions of the Kakanui semi-schists. No xenoliths of the Kakanui greywacke or of the typical Chl 4 or even Chl 3 schists of Central Otago, exposed south of the Shag Valley, and in the faulted inlier on the north-eastern side of the intrusion, have yet been recognised in the porphyry.
The argillite xenoliths may form nearly 5% of the bulk of the porphyry. P.8662 is the only slide free from them. They are quite inconspicuous in micro-sections, since their colour is almost that of the glassy matrix, though they are sometimes paler through lack of the faint limonitic staining of that matrix. Their grain-size (∠ 0.002 mm.) is so minute that mutual compensation of the birefringent granules renders them so nearly isotropic that a gypsum plate is necessary to show their granularity. Some are nearly rectangular (∠ 1.0 × 0.4 mm.) and bounded doubtless by close-spaced joint-fractures, though the majority are smaller, have rounded or less regular boundaries. Sections cut transverse to the bedding planes of the argillite show thin close-spaced lamellae containing flakes of pale yellow-green clinochlore, in a few cases as much as 0.15 mm. long. Parallel to these are rarely very thin segregation veinlets of quartz with optic axes perpendicular to the bedding planes. Sections cut parallel to the bedding plane show only a minute mosaic structure, though rarely (5923) there may be scattered rounded granules of quartz up to 0.1 mm. in diameter. In addition to such siliceous xenoliths there are small scanty turbid brown chips of more kaolinitic material.
Schistose xenoliths are more varied, and two groups of these may be distinguished. In some (5923, 5925) [see Fig. 2] the grain-size (0.02–0.03 mm.) is much smaller than all but the most finely granular or lowest grade Chl 2 of the quartz-albite-chlorite-muscovite schists and the structure more hornfelsic than schistose. The dominant minerals are quartz with albite (?) and biotite and little or no muscovite, suggesting that the rock has undergone some thermal metamorphism and that combination of chlorite and muscovite has occurred, as in the case of the more coarsely granular and relatively high-grade schists of the biotite zone (Turner, 1938), though the heat, not of plutonic magma, but of the quartz-porphyry magma may have been effective here. Small xenoliths of hornfelsic to semi-schistose quartz-rich rock free from biotite (e.g., Fig. 5922) may have a like origin.
5925. Quartz porphyry with glassy matrix (wide-spaced dots) containing irregular vesicles (V), phenocrysts of quartz (Q), plagioclase (P), and biotite. Small xenoliths (X) of argillite and semischist. Only a few of each are lettered.
5091. Quartz porphyry with large xenolith of schist.
5922. Quartz porphyry with abundant bauerite pseudomorphs (B1) after biotite, and a small hornfelsic xenolith.
Schistose xenoliths of the second group have more strongly marked schistosity (e.g., in 5091, Fig. 2), and cut transversely show an approach to augen- or mortar-structure with relatively large (∠ 0.05–0.40 mm.) residual, marginally granulated and optically strained grains of quartz and feldspar set in finely crushed but partly recrystallised minerals divided by impersistent, crumpled and irregular laminae composed of minute flakes of muscovite and very
little chlorite, and rendered turbid by carbonaceous and limonitic particles. Large (∠ 0.3 mm.) flakes of muscovite may lie transversely across the schistosity or the lamellation. In (5936), the dark grey angular xenoliths contain porphyroblasts (∠ 0.15 mm.) of biotite lying obliquely across the schistosity plane of a fine-grained matrix containing relatively large (∠ 0.05 mm.) angular grains of quartz and feldspar, and extremely abundant flakes (about 0.02 mm.) of weakly pleochroic biotite grouped in laminae alternating with laminae in which almost colourless to pale greenish-brown mica is abundant. The rock-structure approaches that of rocks in the Chl 2–3 grade of metamorphism, the highest exhibited by the older rocks in the Kakanui Range. The largest xenolith (5 × 5 mm.) seen in the quartz porphyries is of this nature (Fig. 2, 5091).
The quartz porphyry has an almost to completely istotropic matrix making over 80% of most rocks, faint buff in colour with refractive index 1.508–1.513 ± 0.001 as determined by Dr. Hutton, such as may distinguish rock-glasses with 70–67% of SiO2 according to Tilley's (1922) and George's (1924) observations. Difficulty in measuring this index was occasioned by the presence of crypto-crystalline material and minute indeterminate feldspar laths. The presence of kaolinitic material and limonitic staining also decreases the translucency. There may occur also indeterminate colourless almost isotropic sharply-bounded prismoids (∠ 0.03 × 0.01 mm.), or fragments with embayed margins sometimes recalling “bogen” structure in vitric tuffs, and others larger and less regularly shaped, the last possibly being what Paterson termed “cavities filled in by zeolites”. These have a lower refractive index than the glass and contain irregular platelets which rarely are almost as birefringent as quartz. The suggestion is tentatively advanced that they consist in the main of opal in part changed to chalcedony.
The most abundant phenocrysts are quartz, more or less corroded, and in some cases derived from bi-pyramidal crystals up to 2 mm. long. Others are sharply angular and of varying form, including many small splinters with elongation less often perpendicular than parallel to the optic axis. These, and the irregularly-bounded fragments occasionally show optical strain.
Much less abundant but varying in amount and smaller (∠ 0.8 mm.) are phenocrysts of oligoclase about An21 rarely showing a core of An29, sometimes in idiomorphic prismoids but nearly always with fractured boundaries, occasionally seeming to have been dragged out a little into sub-individuals by movements in the stiffening magma. A few irregular and fractured grains doubtfully referred to sanidine are present.
Biotite in hexagonal plates up to 20 or 30 mm. in diameter forms 1–3% of most rocks. It is generally fresh, strongly pleochroic, with slight marginal zoning or borders darkened by reaction with the glassy ground-mass. The highest colour is a deep yellowish or greenish brown. The optic axial angle is very small and slight chloritisation occurs in some cases. Many transverse sections of biotite display bending, slight corrugation or fraying out, especially of the smaller flakes. The commonest alteration is an incipient
bleaching accompanied by the separation of dusty limonite, and at first a retention of strong birefringence. This process commences around the margins and in the cleavage planes. In 5922, however, baueritisation* of the mica is almost complete. In this rock the mica was more abundant than in any other of the quartz porphyries, and has been almost completely replaced pseudomorphously by colourless material with R.I. about 1.545–1.550, D.R. about .008, optical orientation as in mica, 2V=40° (—) kindly determined by Dr. Turner, with simultaneous separation of dusty magnetite and limonite in the cleavage-planes and small cracks perpendicular thereto. The cleavage-laminae are commonly bent or crumpled, and may be arranged in subradiating fashion, and the pseudomorphs may show strongly marked strain-shadows, or be broken into sub-individuals with diverse orientation. Commonly such broken grains may be dragged out into the matrix as a veinlet or streak of colourless bauerite, with very undulose or irregular extinction-directions, suggesting that the stiffening rock may have been sheared, possibly by faulting movements prior to its complete solidification. The glassy matrix in this rock is abnormal, and has devitrified in thin wispy streaks of faintly birefringent material with extinction parallel to their extension suggesting the presence of minutely acicular alkaline feldspar and appearing (under crossed nicols) against the dark background like moon-illumined cirrus clouds or wispy nebulae. (Cf. Jeans, 1944, Plate VII.) In hand-specimen the rock shows faintly-marked slickensided-surfaces.
Apatite prisms (∠ 0.1 × 0.05 mm.) are occasionally included in the biotite flakes, and in one case an apatite prism was seen half embedded in biotite, the remainder being enclosed in the glassy ground-mass.
The Chemical Composition of Quartz Porphyry.
A specimen of the quartz porphyry collected by Mr. D. A. Brown adjacent to the Horse Range road was analysed by Mr. F. T. Seelye (Analysis No. 11 of Table II). No sample of the rock was retained for petrographical study, and as the high content of water (now believed to be dissolved in the glassy matrix) and excess of alumina (normative corundum) suggested that the rock was weathered and possibly contaminated—the frequent presence of xenoliths had become known by then—search was made for the specimen (5925) that was most fresh and contained the fewest xenoliths, so that a more representative analysis (No. 12) could be made by Mr. Seelye. Later, however, a sample of the rock first analysed (5936) was obtained from Mr. Seelye and found to be almost free from decomposition, though it contained many small xenoliths of dark grey richly mica-
[Footnote] * Bauerite (or “Silica metabiotite”) is a term applied by Rinne (1911, 1924, 1925) to the colourless material, chiefly silica holding water loosely and in variable amount, left when the majority of the bases in biotite have been removed in solution. It forms pseudomorphs after biotite and retains some of the optical characteristics of biotite and of its atomic structure as shown by X-ray spectrography, though with much lowered refractivity and birefringence. [Compare Gaskin's (1944) recent study of bleached biotite in a Victorian granodiorite.]
ccous schist. It was collected a few chains north of the roadway. The second rock analysed is also rather hydrous and shows a greater excess of alumina than No. 11, and since its xenoliths are few and highly siliccous rather than kaolinitic, it must be concluded that excess of alumina (normative corundum) as also content of water are characteristic of the glassy base, as of other acidic rocks the analyses of which are cited in Table II. Variation in silica-content of the glass is also suggested, for, though the second analysis shows higher SiO2 than the first, the rock contains less phenocrystic quartz than is usual in these Shag Valley rocks. Variation in the content of lime and of baryta in the quartz porphyry is also indicated.
The Origin of the Quartz Porphyry of Shag Valley.
A Comparative Petrogenetic Discussion.
(A) Hypothesis of derivation by fractional crystallisation-differentiation of uncontaminated basaltic magma.
Geological occurrence suggests that the quartz porphyry of Shag Valley was injected into its present position during the period (Oligocene) of activity of the basalt-tholeiite-dolerite magma in North-Eastern Otago. There are many regions throughout the world, e.g., at Lassen Peak (Harker, 1909, p. 125; Bowen, 1928, p. 95), where the co-magmatic relationships of a series of volcanic and/or intrusive rocks are made clear by the Cape Colville Peninsula (Henderson, 1913, p. 79) and North Auckland (Bartrum, 1925, p. 13), and fractional crystallisation-differentiation is held by most though not by all petrologists to be the chief genetic process concerned in the development of these varied rock associations. There are other regions such as North-Eastern Otago, where there is a general lack or extreme scarcity of rocks intermediate in composition between the basic and acid extremes of rock-variation. It will be helpful, therefore, to discuss briefly several of these discontinuous sequences, and to note the grounds for concluding that fractional crystallisation is for these also the chief process concerned in their differentiation.
We commence with the Carboniferous tholeiitic-quartz doleritic intrusions of Northern Britain. The petrology of the great Whin Sill of Northumberland and Durham has been very thoroughly described petrographically by Teall (1888), Holmes (1928), Tomkeieff (1929) and others, and the range of chemical composition has been exhaustively studied by Harwood (in Holmes, 1928), and especially by Smythe (1930). The average composition of the mass is given in Table I, No. 2. The points of the various oxides as plotted on Harker's (1909, p. 119) variation-diagram fall very close to continuous curves, and Smythe's (1930, p. 55) illustration is copied with additions in Fig. 3 herewith. Besides normal and quartz-bearing dolerites and more or less pegmatoid facies of both, there are locally small amounts of “felsitic” or aplitic differentiates, which form small “inclusions” or segregations and ramifying veinlets without sharp boundaries or spreading irregularly through coarsely granular rocks or filling vesicles. Analyses of three of these (Nos. 1, 2 and 3 of Table II) are
plotted on the diagram, together with two “micropegmatites” which are derived from Analyses No. 1 and 2 by recalculating these as free from ferromagnesian minerals and decomposition products. These Whin Sill analyses, however, leave the range SiO2 55%–73% almost unrepresented, and to fill this are plotted analyses of rocks derived from the coeval tholeiite-dolerite complex of the Scottish Midlands which have a closely analogous average composition (Table I, No. 3). It is to be noted that the composition of a Whin Sill aplite (Table II, No. 2) is almost identical with that (Table II, No. 4) of the residual glass (Walker, 1935, p. 151) in the tholeiite of Kirkintillock, Dumbartonshire (which contains 49.10% SiO2), which glass Shand (1943, p. 226) considers to afford the clearest available evidence of the derivation of an acidic residuum by crystallisation differentiation of a “basaltic magma”.* It also suggests that in spite of Fenner's (1926, 1937) caveat, there are good grounds for inferring a magmatic origin for the micropegmatite in the Carboniferous dolerites of North Britain, and also for that of the quartz-bearing dolerites of North-Eastern Otago, which were derived from a magma of very similar average composition (Table I, No. I), so far as may be inferred from the few analyses available. It may be noted that the pegmatoid rocks of intermediate composition in the Scottish quartz dolerite complex, shown by the few representative analyses plotted, are, as Kennedy (1933) notes, rather strongly sodic Their alkali variation-curves, plotted by broken lines in Fig. 3, differ in this respect from those of the Whin Sill complex.
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
All analyses calculated as free from H2O and CaCO3.
Average composition of the Oligocene basalts, tholeiites and dolerites of North-Eastern Otago. Benson (1944), p. 104, No. 6.
Average composition of the Whin Sill. Cf. Smythe (1930), p. 32.
Average tholeiite of Scotland. Cf. Walker (1935), p. 152, No. 17.
Average Palisade dolerite. Walker (1940), p. 1081 and Table 3.
Average Karroo dolerite based on representative analyses selected by Professor Walker (priv. comm.).
Average Keweenawan basalt calculated from 21 analyses cited by Broderick (1935). In many of these oxidation of FeO to Fe2O3 has occurred.
Felsitic vein rock in Whin Sill Cushat Steel, Dunstansburgh, Northumberland. Smythe (1930), p. 45, No. 7.
Pink veins in the Hampeth Dyke, Northumberland. Smythe (1930), p. 98, No. 2.
Felsitic vein in the Whin Sill, Alnwick Moor, Northumberland. Tomkeieff (1929), p. 110.
Glass isolated from an analysed tholeiite with 49.10% SiO2. Kirkintillock, Dumbartonshire, Scotland. Walker (1935), p. 150, No. 4.
White albitised vein in Palisade dolerite, New York. Walker (1940), Table 3, No. 20.
Albitised diabase aplite, Goose Creek, Virginia. Shannon (1924), p. 28.
Microgranitic dyke, Insizwa, Griqualand, S. Africa. Scholtz (1937), p. 143, No. 14.
Microgranite sheet in hornfels near contact with dolerite. Scholtz, loc. cit., No. 15.*
Granophyric dolerite. Thomas River, Transkei Gap, S. Africa. Mountain (1944), p. 61.
Rhyolite, Manwan Creek, Lebombo Range, Zululand. Prior (1910), p. 154.
Quartz Porphyry, Shag Valley, N.E. Otago.
Quartz Porphyry, Shag Valley, N.E. Otago.
Rhyolite, Tardree Mountain, Co. Antrim, Ireland. Holmes (1936), p. 94.
Hornblende granophyre, Beinn a Chairn, Skye. Harker (1904), p. 216.
Granophyre, Kastardal, Iceland. Hawkes (1928), p. 513.
Leucogranite, Upper Skeggil, Iceland. Hawkes (1928), p. 525.
Liparite, Berufjordskard, Iceland. Washington (1917), p. 65, No. 30.
Spherulitic Rhyolite, Cape Franklin, E. Greenland. Backlundand Malm-qvist (1935), p. 55.
Quartz Rhyolite, Cape Franklin, E. Greenland, op. cit., p. 32.
Glassy matrix of “Andesite obsidian” erupted May, 1902, Mont Pelee, Martinique. Lacroix (1904), p. 527.
“Andesite obsidian” containing the glassy base yielding analysis No. 18. Loc. cit. supra.
“Red Rock” (granite), Pigeon Point, Lake Superior, Minnesota. Grout (1928), p. 560, No. 4.
Rhyolite associated with diabase sill, Duluth, Lake Superior. Schwartz and Sandberg (1940), Table I, No. 7.
Ibid. Loc. cit., No. 22.
Granophyric Felsite, Bare Hill, Keweenaw Point, Lake Superior. Broderick (1935), p. 356, Table 3, No. 3156.
Felsite, Mount Houghton, Keweenaw Point. Loc. cit., No. 3159.
Granophyre associated with the Breven dolerite dyke, Saterstugan, Sweden. Krokstrom (1932), p. 305.
Granophyre vein in the Skaergaard gabbro intrusion, Kangerdlugssuak, East Greenland. Wager and Deer (1939), p. 208.
The Palisadan dolerite of Eastern U.S.A. has again an average composition (Table I, No. 4) similar to Nos. 1, 2, and 3 of the same table, though characterised by rather higher content of Al2O3 and MgO, and lower content of Na2O.
After making a detailed petrographical study of this dolerite and considering numerous chemical analyses, Walker (1940, p. 1097, 1101) concludes the evidence from the white aplitic veins which invade the chilled phase of the sill is such that they could not be “arkose mobilised by the igneous rock and then injected into it,”
[Footnote] * Nos. 7 and 8 may not be strictly of magmatic origin. Nos. 15a and 15b are represented by points plotted in Figure 4.
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
|Fe2O2||0.37||0.19||3.63||0.97||1.23||1.28||1.39||1.20||0.77||0.63||0.85||0.63||1.47||1.33||2.34||2.76||1.21||1.05||2.20||0.68||2.54||5.20||1.69||1.51||2.0 [ unclear: ]||1.56|
|MgO||0.55||0.71||1.01||0.50||1.48||2.12||0.58||0.72||1.19||0.09||0.56||0.34||0.08||0.56||0.44||0.24||0.11||0.63||3.68||0.58||Tr.||0.08||0.24||0.01||0.17||0.1 [ unclear: ]|
|Les [ unclear: ] :||0.12||0.02|
|Total||100.15||99.50||100.22||100.41||99.91*||99.29||99.95||99.70||100.58||100.49||100.21||100.11||99.81||100.18||99.88||100.49||99.99||99.69||100.56||100.31†||99.9 [ unclear: ]||99.74||99.96||100.09||99.17||100.02|
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
|Norms calculated free from [ unclear: ] alcr and calcite.|
|O [ unclear: ]||31.53||30.82||46.81||23.79||2.23||4.00||34.47||21.10||28.97||28.97||33.30||36.65||25.62||35.55||17.39||19.98||46.43||7.23||7.21||29.97||30.62||37.15||40.35||48.97||27.20||22.84|
|Fs||2.03||4.03||2.49||1.51||0.62||9.90||1.31||0.20||1.60||2.53||2.24||6.0 [ unclear: ]||3.21||0. [ unclear: ] 8|
|il||1.25||0.80||0.76||0.51||.77||0.74||1.42||3.19||0.02||0.69||0.48||0.77||0.30||0.15||0.15||0.1 [ unclear: ]||0.77||0.46||0.30||0.21||0.15||0.61|
|mt||1.72||0.24||5.30||1.54||1.90||2.13||1.16||0.9 [ unclear: ]||0.25||0.41||1.86||3.28||2.35||0.72||1.62||3.24||0.94||0.70||0.46||0.70||3.10||0.93|
[Footnote] * 99.04 in original.
[Footnote] † 100.37 in Washington (1917).
The Carboniferous dolerites, etc., of North Britain, especially the Whin Sill, chiefly after Smythe (1930, p. 55), with additional data.
The Oligocene dolerites, etc., of N.E. Otago and the Shag Valley quartz porphyries.
The Jurassic dolerites of the Karroo, South Africa, and the Lebombo rhyolites. Also the syntectic granophyres. Compiled from available data.
also, that “apart from the possibility of certain upward migration of volatiles and resultant hydro-thermal alteration, simple crystal fractionation accounts adequately for all the differentiation phenomena displayed by the (Palisade) sill, there is no evidence for introduction of silica by syntexis with floated arkose slabs.” Though no variation diagram could be constructed for the series of analyses of the rocks of the Palisade sill indicating clearly the effect of crystal-fractionation, this may be seen by contrasting the form of the curves for MgO in the variation-diagram (Fig. 3) for the comparatively little differentiated Whin Sill, and that for the few analyses of the Mid-Tertiary Basic Igneous rocks of North-Eastern Otago, in which crystal-fractionation and the gravitation separation and accumulation at depth of olivine crystals, which was so clearly indicated on microscopical grounds (Benson, 1944, p. 92).
Still more pronounced, however, is the effect of crystal-sorting and formation of picrites in the thickest gravitationally stratified portions of the Karroo series of basic intrusions as may be seen in Fig. 3. The average composition of the parent magma of these rocks (Table I, No. 5) is based chiefly on the compositions of the undifferentiated marginal facies of the dolerite, but the very wide extent, approximately 200,000 square miles, over which these rocks occur, and their huge volume, 50,000–100,000 cubic miles [the estimates are due to Du Toit (1920)], there may have been a considerable variation in the local character of the initial magma and there have certainly been a very wide range of conditions of cooling and differentiation and consequently of rock types produced. Numerous good analyses of these by Brink, Frankel, Herdsman, Hillebrand, Mountain, Murray, Nell, Preetz, Prior, Radley, Reinesch, Scholtz, de Swardt, and Weekes are given in petrological papers by Daly and Barth (1930), Frankel (1943, 1944), Henderson (1910), Mountain (1944), Nell and Brink (1944), Poldervaart (1944), Prior (1910), Scholtz (1937), de Swardt and Murray (1944), and Walker (1940, 1941, 1942). As the bulk of these analyses show between 50% and 55% of SiO2, only a representative selection of these were plotted to avoid overcrowding Fig. 3. The less abundant but more basic rocks are thus proportionately over-represented in the diagram, even though the analyses of the least siliceous rocks were not utilised. Only three analyses (one incomplete, see Table II, Nos. 7, 8, 9) of acidic magmatic (?) differentiates from this basic magma are available to the writer, those of micro-granites associated with the gravitationally differentiated Insizwa laccolite. Goodchild (1916) held that these “represent material squeezed out from the differentially melted hornfels floor at an early period after the consolidation of the intrusion,” and Scholtz (1937, p. 146) cited Daly's view that they “may be regarded as later differentiates of acid material previously absorbed by the gabbroid magma” though concluding that “assimilation could only have played a relatively unimportant role”. There are several analyses of syntectic and rheomorphic granophyres occurring in the dolerite which will be considered below.
Instead of wholly intrusive magmatic acid differentiates, however, the Karroo basic magma yielded the voluminous acidic lavas of
the Lebombo Range, which are of particular interest to our present discussion. The following information concerning them is derived chiefly from Du Toit (1929). They occur in Zululand near the north-eastern extremity of the huge Karroo magmatic province, and are especially associated with the Stormberg basalts, the effusive phase of the Karroo magma, two analyses of which (Stockley, 1940) are plotted on Fig. 3. The igneous rocks of the Lebombo Range can be triply divided into a lower basaltic group commencing with limburgite and augitite,* a middle rhyolitic group and an upper basaltic group, with an intercalation of rhyolitic rocks. The thickness of erupted material must be tremendous—several miles at least. The flows, tuffs and breccias form a series inclined eastward at 5°–20°, but near the axis of flexure their inclinations rise to 30°, 40°, or occasionally 80°. Within the lower basalts are large linear bodies of dolerite, gabbro and granophyre which are regarded as the feeders of the material erupted, the volcanism being of the fissure type. Some of the dykes have been introduced along slip-faults. Crustal tension in an east-west direction is specifically indicated throughout the eruptive period; the growth of the monocline can, with good reason, be dated to the middle portion of that period, and its development is considered responsible for the explosive and acidic character of the material emitted during the second volcanic period which occurred between the effusion of two “quiet” suites of plateau basalts. “Both extrusive and intrusive suites can be regarded as products of a single basic intercrustal reservoir within which differentiation had been brought about.” These acidic rocks have been described petro-graphically by Henderson (1910) and Prior (1910) and four analyses thereof are plotted on Fig. 3. If these four analyses, together with that of the Insizwa microgranites (if they be not syntectic), and those of the granophyric dolerite (Table II, No. 9) and more basic dolerites, be used for the construction of a variation diagram, a series of curves are obtained remarkably like those derived from the Whin Sill and its associates, and differing from the curves obtained from the plotted analyses of the igneous rocks of North-East Otago chiefly in the low content of soda in the Shag Valley quartz porphyries, which may perhaps result from the more advanced degree of fractional crystallisation. (See below.)
Indeed, there is yet another and very suggestive point of analogy between the developments of the African and Otago rocks. The acid rocks, apart from minor or even interstitial developments, rose in bulk only along lines of crustal deformation (unless the Lebombo rhyolites should extend far eastward beneath the Cretaceous and Tertiary sediments which cover their down-warped area), while, as noted above, the quartz porphyries are known in Otago only along the Shag Valley fault-zone, which was undergoing some deformation during the period of activity of tholeiitic-doleritic magma in the region bounded by this fault-zone. In both areas localised differential pressure may have been partly responsible for the expulsion of
residual melt from the crystallising basic magma in the reservoir at depth, affording examples of the process resembling that recently discussed by Emmons (1940) in a paper which may have other and varied applications to the rock-series discussed in the last few pages.
This discussion of the genetic relation of associated basic and acidic volcanic or hypabyssal rocks may be extended by brief references to a few of the many other instances of such associations. Thus: the eruption of the rhyolites of Keweenaw Point, Lake Superior, followed on voluminous effusions of basalt and was succeeded by eruptions of gabbro and basalt, folding with the formation of a great synclinorium being in progress throughout the period of eruptive activity (Iddings, 1913, p. 336). Broderick's (1935) detailed examination of the chemical petrology of the rocks shows that it may be expressed by a variation-diagram (deduced from a series of almost collinear points) very like those in Fig. 3 herewith, but distinguished therefrom by the rather more potassic nature of the American rocks. He concludes that the association resulted from crystallisation-differentiation together with some volatile transfer of alkalies. Schwartz and Sandberg (1940), discussing the association of rhyolites with diabases at Duluth, west of the last-named area, concluded that “while syntexis” (as suggested by Daly in 1917) “cannot be excluded as a process in the Duluth sills, it appears improbable that it was the dominant process. If any ‘material’ was incorporated ‘in the basic magma’, it has become changed so much during differentiation as to be unrecognisable. The differentiation is clear, but there is an entire lack of evidence of any earlier process.” They do not, however, regard crystal-sorting as the principal method of differentiation, Grout's (1918) and Fenner's (1926, 1937) objections thereto having, they believe, valid application to the Duluth area. The origin of the granophyres in this area they consider to have been closely connected with the amount of mineralisers present in the upper portions of the sills, certainly an igneous process, independent of the action of differential pressures, the significance of which was urged by Emmons (1940).
Williams (1935), discussing the Newberry (Oregon) volcanic complex of basaltic and rhyolitic rock with very subordinate andesite (a feature contrasting with the more fully represented composition range at Lassen Peak), noted that the repeated eruption of rhyolite and basalt more or less contemporaneously from a common fissure, suggested the co-existence of two contrasted magmas beneath the region studied, but concluded (p. 303) that these two magmas were probably derived from a common plateau basalt magma through a process of crystal-differentiation. It did not seem possible to account for the association by any method involving assimilation.
The Brito-Arctic Tertiary igneous complex affords classical instances of this association of basic and acidic rocks. Thus Harker (1904, p. 55) notes that rhyolites were poured out between the earlier and later periods of basaltic eruption in Antrim (Ireland), where there occurs the quartz porphyry or rhyolite of Tardree (Table II, No. 13). Rhyolites, of which no analyses are available, occur intercalated among the earlier basalts of Skye, and the presence of com-
posite sills formed at a later stage is evidence of the co-existence of acidic (Table II, No. 14) and basic magmas in the intercrustal magma-reservoir (Harker, 1904, p. 57, and chaps. xii–xiii). In Mull (Thomas and Bailey, 1924) and Ardnamurchan (Richey and Thomas, 1930), the association of rhyolite with basic lavas and the occurrence of composite intrusions has also been observed. Composite intrusions occur also in Arran (Týrrell, 1928). Iceland (Rosenbusch, 1908, pp. 836–8, Iddings, 1913, p. 372, Hawkes and his co-workers, 1925, 1928, 1933) consists almost entirely of basaltic rocks which have been erupted intermittently from Tertiary times to the present day. Intercalated with and invading these are rhyolites, rhyolite-porphyry, granophyres and granites, as well as differentiated gabbroid masses. Intrusive rocks of medium acidity are very subordinate. The most abundant of the acidic rocks with 70–73% of SiO2 are dominantly sodic or sodi-potassic (Table II, No. 15), while the more richly siliceous rocks are products of a later stage in differentiation, and are dominantly potassic (Nos. 15A and 15B in Figure 4). The structure of a composite dolerite-quartz porphyry dyke gives evidence of the co-existence of acid and basic magma (Hawkes, 1925).
At Cape Franklin, in North-Eastern Greenland, effusions of spherulitic rhyolite and a probably younger more potassic quartz rhyolite (Table II, Nos. 16, 17) occurred during the interval between the eruption of an earlier series of dolerites and a later series of basalts (Backlund and Malmqvist, 1935). At Kangerdlugssuak, the Tertiary basalts are invaded by the large differentiated gabbroid Skaergaard intrusion, which is cut by dykes of granophyre (Table II, No. 26) similar in composition to the spherulitic rhyolite of Cape Franklin, and forming the latest differentiate of the gabbro magma. (Wager and Deer, 1939, pp. 207–8, 324.) Except for a relatively small amount of hedenbergite granophyre containing 59% of SiO2, igneous rocks of medium acidity are lacking in this complex.
Attention may be called to the occurrence of “plagioliparites” with 73.4% SiO2, 3.7% Na2O and 3.7% K2O, as the latest extrusive products of the basaltic volcano of Mount Tanggamoes in Southern Sumatra. The liparites contain phenocrysts of quartz, oligoclase and biotite set in a glassy basis which contains a noteworthy amount of normative orthoclase, though no modal orthoclase is determinable. The point which indicates the ratio SiO2: NaAlSiO4: KAlSiO4 in these rocks is almost coincident with No. 21 in Figure 4. Bemmelen (1932) regards them as the extremely siliceous final differentiation product of the basaltic magma. No ultrabasic or intermediate rocks have been found in this complex.
Several points may be noted in concluding this section. Grout's (1926) calculations lead to the estimate that a normal plateau basalt could, through fractional crystallisation-differentiation, yield a volume of residual acidic or granitic magma equal to about a tenth of that of the original magma. Holmes (1936, p. 231), basing his calculations especially on the composition of the Whin Sill, found “that by the time the residual liquid has reached the composition of granite, it forms only 5 to 10 per cent. of the whole mass”, and concluded that its presence within a framework of augite and plagioclase
is fatal to the idea that such residual acid liquid could collect to form a discrete body of granite magma, except by squeezing with intense deformation of the crystal meshwork.
Collins (1917, p. 97), inspecting the veins and intergranular micrographic derivatives of basalt at Onaping Lake, put the fractional volume-relationship as about one-seventh. At Kangerdlugssuak the quantity of the late stage intermediate (hedenbergite granophyre with 58.81% SiO2) and acid rocks is small, probably not amounting to more than “one per cent.” (Wager and Deer, 1939, pp. 218–293) or “two or three per cent.” (op. cit., p. 236). In each of the other cases noted, similar volume-proportions between parent basic magma and acidic residuum may hold, though in regard to the very voluminous Lebombo rhyolites a quantitative estimate is not possible.
It does not seem certain, however, that the physical objections cited above raised by Holmes against the hypothesis of the derivation of felsitic veinlets as residual products of fractional crystallisation of a relatively small basic intrusion such as the Whin Sill, must necessarily be fatal to the hypothesis of the derivation of such a small amount of acidic melt as rose to form the Shag Valley porphyry as a discrete body during a late stage in the crystallisation of a large volume of basaltic magma at considerable depths. Grout's concept (1918, p. 491) of fractional crystallisation with gravitational separation is approvingly cited by Daly (1933, pp. 324–5). “During much of the magmatic stage of a thick igneous body, crystals are locally concentrated along wall and roof, rather than distributed uniformly throughout the mass. This crystal-sown phase is probably denser than the unmodified liquid. If the group of crystals settles down a short distance into the original liquid beneath, the mixture has a higher density than the original liquid where free of crystals. Hence the local mixture tends to sink. The downward shearing stress of the denser body of material increases with the diameter of the body, which, even if only a metre in diameter, may sink much more rapidly than an isolated crystal. The result is a powerful two-phase convection of a special kind”, which would aid in the accumulation of an acidic phase in the upper portion of the magma chamber. Since gravitative differentiation with the sinking of olivine and pyroxene crystals, and the upward concentration of the acidic residual magma a short distance beneath the chilled roof of sills which may be little more than 150 feet thick, and such acidic material, though comprising not more than one or two per cent. of the total volume of the sill, may amount to nearly twenty-five per cent. of the volume of the portion of the dolerite into the interstices of which it has been concentrated, as has been proved to have occurred in an Oligocene dolerite sill at Moeraki (Benson, 1944, pp. 91–9, and especially Fig. 6), it does not seem beyond the bounds of possibility, even if one accepts Holmes' quantitative considerations as generally true, that a large body of basic magma crystallising slowly at depth may yield, if subjected to moderate deformation before its complete solidification, an acidic residual liquid, which, breaking through the upper chilled margin of the magma-reservoir, might rise into the fissure
The composition-indicating points for the acidic rocks in Table II, derived (?) from basic magmas, plotted on Bowen's (1937) NaAlSiO4–KAlSiO4–SiO2 residual magma diagram. For significance of the points * * *, E. and S. see text pp. 27, 29, 31.
of an actively-moving fault, and there solidify as a discrete body. Such an intrusive mass would be largely glassy or wholly crystalline according to its size and rate of cooling, and would contain dissolved gases. If it rose quickly to a level not far below the earth's surface it might be expected to solidify as a dominantly vitreous rock containing some phenocrysts of quartz and feldspar, and a number of vesicles. That the slightly vesicular hypohyaline Shag Valley quartz porphyry consolidated within about 1,000 feet from the surface (the combined thicknesses of the “Abbotsford”, “Green Island” and “Burnside” sediments) must follow if its Oligocene age be acceptable.
Many, though not all, of the acidic rocks generally thought to be differentiates from basic magmas contain (as do the Shag Valley quartz porphyries) a small excess of alumina appearing as normative corundum (Table II). That such may be concentrated into a residual melt is shown by the contrast between the composition of a Peléan “andesite obsidian” containing phenocrysts of basic labradorite and hypersthene with a minor amount of olivine and ores, and that of the glass derived from the rock (Lacroix, 1904, Table II, Nos. 19 and 18), to which attention was called by Rittman (1936).* This pair of analyses affords as striking an example of the derivation of an acidic residuum from a parent basic magma as that of the Scottish tholeiite noted above (No. 4).
[Footnote] * Most of the analyses of Peléan andesitic lava and tuffs show up to 2% of normative corundum and occasionally cordierite is noted in such rocks. An extreme case is that of a cordieritised andesite with 8.06% of normative corundum (Washington, 1917, p. 333).
Bowen (1937) held that acidic magmas containing more than 80% of normative sialic minerals (excluding anorthite) are sufficiently near to the compositions of members of the system NaAlSiO4–KAlSiO4—SiO2 to have a cooling history comparable with that of this experimentally investigated system, and as a consequence of fractional crystallisation during cooling the compositions of their residual magmas should change towards those represented by points within the field containing mixtures with the lowest melting point, the “low temperature trough”. (See Fig. 4.) This relation was found by him to hold for Central African rocks, and was applied to averaged analyses of a wide range of late differentiates both volcanic and plutonic by Barth (1939, p. 82). It was also well exemplified by the phonolites and trachytes of New Zealand, but many of the points representing the sialic compositions of New Zealand rhyolites and quartz porphyries were found to be scattered outside of the sialic end of the low temperature trough (Benson, 1941, p. 542), a divergence from Bowen's generalisation concerning non-porphyritic rocks, which, it was suggested, might be due to the presence of a noteworthy amount of phenocrystic quartz in the aberrant rocks. The same explanation may hold for some of the “super-siliceous” analyses listed in Table II and plotted in Fig. 4. Thus the Antrim and the Shag Valley quartz porphyries (Nos. 11, 12 and 13) hold about 5% of phenocrystic quartz, so also may (cf. Rosenbusch, 1908, pp. 936–8) the Icelandic liparite, No. 15B, but not so the purely glassy Peléan material, No. 18. The development of richly potassic residual melts* may follow the course discussed by the Larsens (1938, pp. 423–5). They hold that while there is a eutectic relationship between sanidine and plagioclase with more than 5% anorthite, there is probably a reaction-relation if later the content of anorthite becomes smaller. Sanidine low in soda in the early stages of crystallisation becomes richer therein as crystallisation proceeds, finally exceeding the normative proportions, increasing the proportion of potash in the residual melt. “A few of the [San Juan] rocks at the extreme rhyolite end of the diagrams indicate a sharp increase in potash and decrease in soda without much other change. Numerous rock-analyses fall on the variation curves just before the change in slope of the curves of the alkalies, but very few on the steeper parts of curves that represent the potash-rich rhyolites. It seems therefore that only under exceptional circumstances does differentiation proceed to potash-rich rocks. Such rocks will be relatively high in silica and low in lime and the ferric constituents.”† This appears to accord satisfactorily with the features of the Shag Valley and E. Greenland potassic rocks (Nos. 11, 12 and 17), and to explain the apparent departure from the purely eutectic considerations on which Bowen's (1937) diagram is based. It affords
[Footnote] † Sundius (1926) discussion of the association of sodic and potassic varieties of aplites within single intrusive masses derived from Swedish granite magmas is of interest in this connecton. (For an alternative explanation by Holmes (1926) see below, p. —–
also an instance of the general principle by which the crystallisation of a compound of potassium with ionic radius (K=1–33 Å) occurs later than that of a similar compound of an element with smaller ionic radius (Na=0.98 Å), (Goldschmidt, 1938, p. 29). On the other hand, the richly sodic rocks in the acidic segregations from the Palisadan magma (Nos. 5 and 6. Fig. 4) have been deuterically altered with resulting albitisation of the originally more calcic plagioclase and almost complete removal of potassic compounds (Walker, 1940, p. 1093). This replacement at the transitional pneumatolytic-hydrothermal stage of rock-development may be associated with the greater volatility of potassic compounds than of sodic, which determines the albitisation of pegmatites (Bowen, 1933, p. 123; Barth, 1939, p. 99). Possibly it is as a result of concentrations of potassic compounds so liberated and moving up into the lower temperature environment existing simultaneously at levels higher than that where their volatilisation occurs that accounts for the replacement of plagioclase by valencianite and other potassic feldspars in propylitised rocks as at Waihi (Fenner, 1934; Morgan, 1924; Terzaghi, 1935). Hydrothermal processes of enrichment in potassium can, however, hardly be concerned in the development of such glasses as form the ground-mass of the Shag Valley porphyries.
(B) Hypotheses involving Assimilation with subsequent Fractional Crystallisation, Gaseous Transfer or Transfusion.
Though it might be concluded by many petrologists from considerations similar to those discussed above that it is probable that the Shag Valley quartz porphyries were the final residuum after prolonged fractional crystallisation of the basic magma that rose under North-Eastern Otago in Oligocene times, strong objections have been raised on physico-chemical grounds to the view that acidic magmas could so be derived in the absence of any antecedent development of intermediate rocks, and alternative hypotheses, such as are outlined by Daly (1933, Chap. xiv) call for consideration in so far as they may bear on the origin of the Shag Valley porphyries. Fenner (1926, 1931, 1937) has argued that whereas the Bowen (1937) concept of the development of acidic magmas as the residual products of fractional crystallisation would involve their concentration at temperatures below that of the consolidation of basalts, the acidic magmas in several regions, notably at Katmai, Alaska (Fenner, 1926) and the Gardiner River, Yellowstone Park (Fenner, 1938), had so high a degree of superheating that they were able to assimilate large amounts of the basalts which they came into contact with. “The rhyolite was able to do things that the theory of crystal-fractionation declares impossible” (Fenner, 1937, p. 167). He advanced the view that highly heated gases rising through the magma conveyed volatile compounds of various elements, not only from the lower to the upper portion of liquid magmas, but from a magma into the formations invaded by it, i.e., “gaseous transfer” and “transfusion”, e.g., of Reynolds (1936) and Holmes (1936a). The last-named (Holmes, 1931, 1932, 1936b), as noted above, based objections to the hypothesis
of the derivation of acid rocks from basic magmas on other grounds, including the “impossibility” of forming discrete bodies of acidic igneous rocks by removing by filtration under gravity an acidic residual melt amounting only to 5–10% of the original volume of the parent basic magma from a liquid-filled meshwork of closely inter-locking crystals comprising the 95–90% portion of that original magma which had previously crystallised, though it might be possible to remove the acidic residuum by squeezing through intense deformation of the meshwork by externally applied stresses. It is at least open to question whether the local stress adjacent to the Shag Valley fault zone, even if active during Oligocene times, would be sufficient to meet this requirement in a region where so little deformation affected the rest of the region underlain by basic magma during this period, and the same consideration might be raised in connection with the development of the Lebombo rhyolites along a narrow warped and faulted strip through the huge, elsewhere almost undeformed area of Karroo dolerites and Drakensberg basalts. The absence of any rocks of intermediate acidity in very many associations of basic and acid rocks (our own among them) Holmes holds to be a fatal objection to the application of the Bowen hypothesis of the derivation of the latter, since it would be “completely at variance with the results of work on silicate systems.” He states that “the only hypothesis that is genuinely worthy of the name of contrasted. differentiation is that [of Fenner, viz.] distillation of a gaseous phase and its consequent separation from a liquid phase,” but notes that “We do not yet sufficiently understand the nature of magmatic behaviour to describe … the manifestation of such distillation accurately.” (Holmes, 1936b, pp. 231, 236.)
This hypothesis offers no explanation of the marked concentration of the points indicating the ratio SiO2: NaAlSiO4: KAlSiO4 in rocks of rhyolitic, trachytic or phonolitic compositions, whether intrusive or effusive, into the “low temperature trough” of the triangular diagram indicating the consolidation phenomena of this experimentally investigated system, which is demonstrated by the great majority of chemical analyses of rocks within the above range of compositions (Bowen, 1937, p. 18, Fig. 9; Barth, 1939, p. 82, Figs. 54–5; Benson, 1941, p. 542, Fig. 3, and Fig. 4 of this paper) in so striking a manner that, on statistical grounds alone, it is almost impossible that the phenomena indicated could be fortuitous, but almost certainly result from some general physico-chemical evolutionary processes in crystallising magma, whether they be that set forth by Bowen, or processes as yet scarcely adumbrated involved in the alternative hypotheses qualitatively put forward by Fenner and Holmes. In the absence of any quantitative indication of such alternative evolutionary processes in magma no discussion of their possible application to the Shag Valley porphyry can here be attempted. We turn, therefore, to consider the possibility of producing acidic magma through assimilation of sialic formations by invading basic magma and the fractional crystallisation or other modes of differentiation of the resulting syntectic, reserving for the succeeding section of this paper the consideration of Holmes' (1931) hypothesis
of the formation of acidic magmas by the fusion or refusion of sialic formations with lower range of fusion temperatures than the invading basic magma, and without essential mingling of the basic and acidic melts.
There is a strong body of opinion that some assimilation of sialic material into a basic magma is a necessary preliminary to differentiation yielding an abundant richly siliceous residue. This may occur either at great depth, the familiar abyssal assimilation of Daly, or which carried out on a broad scale, according to the view of Barth (1936, p. 350, 1939, p. 80), permits the development of an acid magma where the acid rocks of the Tertiary Circum-Pacific folding zone have been kneaded into the original basaltic magma, or it may occur marginally in relatively small bodies of magma at higher levels within the earth's crust. An often cited example of the latter (following Daly, 1917) occurs at Pigeon Point, Minnesota. Bowen (1928, p. 214), while agreeing that some assimilation of quartzite into basic magma has taken place, holds that it could give rise only to increased amounts of late differentiates similar to those which might form from the original uncontaminated magma by crystallisation differentiation alone. “The rims about the xenoliths are not simply melted xenolith, but essentially normal igneous material of a late stage of the reaction series.” Grout's (1928) restudy of the area [not affected in this regard by Bastin's (1938) later work], lead independently to similar conclusions. “Though some assimilation of the quartzite is indicated, probably much less than a fourth of the granite (Table II, No. 20) is directly or indirectly due to assimilation of sedimentary rock.” “About the same amount of granite would have formed, and its composition would have been about the same had there been no assimilation.” Bowen's comments on the nature of the “rims about the xenoliths,” explains Grout's observation that the material of such rims contains twice as much alkalies as either the quartzite or the enclosing gabbro.
Drusy granophyric patches occurring in the Skaergaard gabbros, interpreted as being formed by recrystallisation of fused xenoliths of the grey gneiss (with 68% SiO2) invaded by the gabbro, are surrounded by more coarsely granular hybrid granophyres, and the same fused product of gneiss-xenoliths is considered to be the source of the material forming small veins of granophyre extending for a few metres at most through the gabbros, not chilled against but merging into them. But “it is unlikely that the source of the magma of the larger transgressive veins of acid granophyre” (see Table II, No. 26) which are unchilled at their contact with the gabbro, “was so derived, but was a late differentiate from the magma forming the highest unlaminated layered rocks” (basic hedenbergite granophyre with 59% of SiO2), which they also intersect. Fenner's (1931, p. 549) view that fractional crystallisation would tend to produce a residuum relatively enriched in iron is supported in some measure by the features of the rocks at Kangerdlugssuak formed up to the middle stage of differentiation, but thereafter the differentiation suddenly takes a new direction, and in the latest stage filter-press action has produced a final highly acid residuum now found in veins and sills
cutting all the earlier differentiates. An amount of acid gneiss estimated as being not more than one per cent. of that of the whole intrusive complex, was assimilated into the magma, but it is considered that the granophyre would have formed without such assimilation of acid gneiss, and that such assimilation did not radically affect the nature of the late differentiates but only their amount. “The complete incorporation of acid material into the Skaergaard magma, during the protracted period of cooling, complicates the interpretation of the late stages of differentiation, but it may be safely assumed that it did not significantly modify the early and middle stages” (Wager and Deer, 1939, pp. 185–212, 239, 306–9).
Mountain (1936, 1937, 1944), Polderwaart (1944), and Walker (1940, 1941, 1942) have described many occurrences of granophyric rims about inclusions of sediments in the Karroo dolerites, or at the margin of sills and of veinlets of granophyre derived therefrom injected into or beyond the dolerite originating from the reaction between the basic magma and the sediments—a process of mobilisation or rheomorphic change of sedimentary material. Analyses of these syntectic granophyres are given and are plotted on Fig. 3 and are stated in normative form in Table III.
Fine-grained granophyre, Rietkop, W. Cape Province. Walker (1942), p. 297.
Upper coarse-grained granophyre, Rietkop, W. Cape Province. Walker (1942), p. 297.
Granophyre, Hangnest, W. Cape Province. Walker (1942), p. 297.
Granophyre. Upper border of xenolith, Alewyn's Gap. Walker (1942), p. 297.
Second transitional zone around sandstone xenolith No. 50. Loc. cit.
Normal granophyre around sandstone xenolith E, No. 51. Loc. cit.
Grit in roof of dyke converted into granophyre, No. 33. Loc. cit.
A few of them containing nearly 80% normative sialic minerals are indicated by asterisks on Fig. 4 and lie close to the normal curves of Fig. 3 and in or near the “low temperature trough” of Fig. 4. Nevertheless, if Poldervaart's (1944, p. 110) view of the origin of at least one of these granophyres be true of all, they cannot be considered as differentiates from syntectic magma. Walker (1942) and
[Footnote] * Analyses Nos. 7 and 8 of Table II would perhaps be more correctly placed if they were transferred to Table III.
Poldervaart (1944) consider the differences between their compositions and those of the sediments from which they were severally derived are the result of “transfusion” of various elements from the invading or enclosing magma, following the work of Drs. Reynolds (1936, p. 403) and Holmes (1936, p. 417). Dr. Reynolds found the general order of diffusibility from a lamprophyre magma into quartzite to be: Al2O3, K2O, Na2O, with minor amounts of P2O5, S, NiO, BaO, and SrO. Holmes (loc. cit.) found the order of relative diffusivity from an alkali basic magma into quartzite to be: SO3, H2O, K2O, MnO, Al2O3, Na2O, TiO2, FeO, CaO, MgO, with minor variations in this sequence in the different areas and samples studied. What is produced is “neither fused silica nor a solution of silica in the material of the enclosing lava. It is a metasomatic replacement product of quartz due to the introduction into the latter of various constituents in proportions surprisingly different from those in which they could have been present in the magmatic part of the lava.” (Holmes). Dr. Reynolds (1936, p. 39) noted that “the circumstances which contributed to the present distribution and relative concentration of the elements in the transfused xenoliths include (a) sequence of introduction into xenoliths, (b) relative rate or power of diffusion through the xenolith, (c) sequence both in time and space of fixation by the xenoliths.” Variation in permeability adds a further cause for the extreme irregularity of the amount of concentration of the several elements transfused from the invading magma into the more or less granophyric, or less often glassy (cf. Holmes, 1936) rocks produced. Concerning South African rocks, Poldervaart (1944, pp. 109–110) comments: “the responsible magma was highly differentiated. Hence … the emanations derived from this magma would (probably) be of a constantly changing character. With all these variable factors it is scarcely surprising that the metasomatic granophyres are of such different compositions.” It is noteworthy, however, that the above list and relative abundance of the typical non-gaseous elements transfused from magma into sediments resembles but is not identical with the list, “Si, Na, K, Fe, Ti, Al, most prominent in the vapour phase, from boiling pegmatitic liquid, among which Ca and Mg should hardly be expected at all.” (Bowen, 1933, pp. 119–120.) Walker (1942) and Poldervaart (1944), however, found that the materials transfused into the sediments in order of relative abundance appeared to be (a) K, Fe, Mg, Al; (b) Ca, Fe, Mg; (c) Ca, Fe, Mg; (d) Fe, Ca, Na, Al, Mg, Ti; and (e) Fe, Ca, Ti, Na, Mg, P respectively in five cases studied, Mg and, in general, Ca being concerned in transfusion, Si and and K being the chief elements lost from the xenoliths. Hence, power to form volatile compounds may not be the chief factor in determining “transfusibility”. We may here recall the development of minute veinlets or sheets of (“transfused”)? alkaline feldspar extending from the margin of the enclosing igneous rock into the quartzose xenoliths in the dolerites of Moeraki (Benson, 1945, p. 297, fig. 3).
But all these “transfused” rocks are associated with relatively large developments of basic igneous masses, are themselves relatively small, and in general holocrystalline. Of one granophyre it is stated
(Poldervaart, 1944, p. 110). “the sandstone was rendered plastic by heat and by emanations from the magma … the crystals grew in a viscous medium. The granophyre veins in dolerite indicate that the mass eventually acquired rheomorphic properties, but signs of mingling or gradation between the two rock types are wholly lacking…. The granophyre was at no time completely liquid”. Since, from the data summarised above, the quantitative effects of transfusion-processes would seem to be at present unpredictable, no useful discussion of their possible rôle in the origin of the Shag Valley quartz porphyry can be given here. Possibly, however, such limitations imposed on the processes of transfusion by their occurrence at a relatively high level and therefore with rapid cooling within the crust need not completely exclude the hypothesis of transfusion of material from basic magma into a heated siliceous invaded formation from having any possible role in the origin at depth of the magma of the Shag Valley quartz porphyries, though they are several miles from the nearest surface exposure of the coeval (?) dolerites.
(C) Hypotheses involving Re-fusion, Anatexis or Palingenesis.
The origin of acidic magmas by the pure melting of old granitic and other siliceous masses has been put forward in a variety of forms by Holmes (1915, 1926, 1931, 1938), and has been considered appreciatively by Krokström (1932, 1937) as affording the most probable explanation of the origin of granophyre (Table II, No. 25) associated with the Breven dolerite dyke in Sweden. He comments (1937, pp. 21–2): “I think it is high time that we should try to break the spell by which the differentiation curves have for rather a long time held many petrologists bound. No reliable proofs of the consanguinity of a couple of rocks may be furnished by such curves … Even the most useful instrument when not properly handled may do more harm than good…. Now experience seems to have brought out that the projective points of a series of consanguineous magmatic rocks tend to fall along simple and approximately rectilinear curves, whereas a markedly aberrant analysis suggests that the rock in question does not belong to the main suite. The reverse conclusion is, however, not valid. If an analysis fits in with the differentiation curves we are entitled to conclude that the chemical relations do not speak against it belonging to this series, but no decisive proofs whatever may be gained from this fact.”
Recalling the experimental evidence that granites will melt and flow at much lower temperatures than basalt magmas, he adds: “If we accept Holmes' view of rising cupolas [and fault-invading wedges?] of basaltic magma, it is clear that the rising thermal surfaces must be able to effect refusion of acid rocks to a fairly large extent … The composition of the melt thus generated is practically independent of the rocks affected, provided they are mainly quartzofeldspathic, as it must necessarily approach that of the lowest melting mixture—the granite eutectic. It is very interesting to find that the Breven granophyre is almost identical with this eutectic [near E, Fig. 4] as deduced by Vogt. I became convinced that the association of dolerite and granophyre was due not to a differentiation of a
characteristic type, but to a rheomorphism of the country rocks.” Krokström's (1937) discussion deals with material in part injected along major fault planes into which basic magma may have ascended, mobilising the immediately adjacent sialic invaded rocks, thus producing a magma which in turn might rise higher in the crust, either in association with the basic igneous rocks as in the Swedish area, or unaccompanied thereby. The hypothesis does not, however, seem to afford an explanation of the feature upon which Fenner has laid such stress, namely, the marked super-heating of the acidic melt where it is seen to have been in contact with the basalt.
Backlund (1944, p. 73) adopts a similar view. While discussing the upward transfer of heat involved in the effusion of the vast thickness of plateau basalts in Eastern Greenland, he adds: “One may realise that such a transfer contributes to the mobilisation not only of minor (molecular, etc.) emanations, but also of major bodies, allied geologically with the basalts, and more or less contaminated in the course of events. It seems probable that the acid and felsic mobilisations of “various regions in Greenland including Kangerdlugssuak” may belong to this type of connection, and yet also some liparitic and granophyric representations in the older parts of Iceland (Hawkes, 1933) may have developed by some similar derivation.”
If the acidic rocks of the Lebombo Range and the vast extent of the Karroo basic intrusions and the Drakensberg basalts be reconsidered in this connection, the limitation of the acidic rocks to a narrow zone of crustal tension and warping may be viewed in the light of the three types of genetic hypotheses already considered, somewhat as follows: To Du Toit (1929) these acidic rocks appear as late magmatic differentiates merely, invading a region of tension (cf. Emmons, 1940). On Barth's (1939, p. 80) hypothesis, they might have resulted from assimilation of the sialic crust into basic magma rising into a dislocating zone. On Holmes' (1931), Krokström's (1937), and Backlund's (1944), they are chiefly the product of the sialic crust mobilised by initial and latent heat of crystallisation provided by the basaltic magma rising into and partially crystallising within the lower part of the tension fissure, the composition of the acid rocks being varied to some extent by mingling with the basaltic magma and by crystallisation-differentiation.
How far can the last hypothesis be applied to the relationship of the far smaller masses of basic and acidic rocks in North-Eastern Otago? Twelve analyses of schists, three of semi-schists, and two of greywacke (Seelye, in Williamson, 1939, p. 30), afford an indication of the composition of the lowest accessible material from or through which the quartz porphyry has come. In Table IV, line R gives the average composition of these seventeen rocks; S the average composition of the three most siliceous schists (quartz-albite-sericite schists with accessory epidote and chlorite or actinolite); 11 and 12, the compositions of the samples of the Shag Valley quartz porphyry, all expressed normatively.
It will be seen that the feldspathic portions of both R and S are dominantly sodic, and that the composition of Q, Or and Ab of S, represented by the point S in Fig. 4, is more sodic than that of the average granite eutectic E, while that of the quartz porphyry 11 and 12 is not only more siliceous, but much more strongly potassic than E. Clearly we have not here a case of magma-modification by assimilation of potassic sediments (cf. Brammall and Harwood, 1932, pp. 11, 228–9). If the material of the quartz porphyry were derived in part from the selective fusion of the schists, further modifying processes, including perhaps those discussed by the Larsens (1938, pp. 423–5), will have been needed to account for the high concentration of potassic feldspar in the quartz porphyry. Krokström's hypothesis alone does not seem adequate, therefore, to account fully for the derivation of the quartz porphyry.
A further suggestion may be that the material of this porphyry might in part be derived from the continued or repeated differential fusion (cf. Holmes, 1926, p. 317) of granite, presence of which invading the schists at depth, but nowhere exposed, has been inferred from, inter alia, the presence of authigenetic tourmaline in the schists (Turner, in Williamson, 1939, p. 40). But to deal with this would take our discussion far into the realm of speculation.
A final tectonic comment may be made. The imperfectly revealed geological structure of the rocks adjacent to the quartz porphyry suggests that the fissure into which they rose was not, near the surface at least, a tensional fissure. But any magma rising into it from depths would be moving into a region of decreasing pressure, and Emmons' (1940) discussion on magmas is not necessarily without bearing on our problem.
From the above discussion, it will be inferred that it is not appropriate to base definite statements regarding the origin of the Shag Valley quartz porphyry upon the limited data available. To do so would involve a premature solution of one of the most difficult and debated problems before petrologists at the present time, the problem of the association of the acid and basic igneous rocks. We may recall the concluding sentences of Shand's (1943, p. 226) stimulating discussion. “Evidence … compels one to admit that basalt and rhyolite may have a common origin, in which case it must follow that the rhyolite was derived from a basaltic parent, since converse is clearly impossible. Whether the process was identical with one of those we have discussed in connection with the gabbro-granophyre association, or whether an explanation can be found in terms of assimilation of siliceous crustal material as A. Holmes argues very cogently in 1931, we have no means of deciding.”
It has been the writer's purpose in the present petrogenetic essay to discuss a few of the salient features of current hypotheses offered in explanation of such associations, in so far as they may bear on the origin of the Shag Valley quartz porphyry, in the hope that a summary thereof, though admittedly very incomplete, may be of some service to fellow petrologists, and afford a fresh instance of the wide range of inquiry which may be involved in the study of comparatively small geological features.
It would appear probable that the basic magma which rose and became active beneath North-Eastern Otago in Oligocene times was concerned in the production of the quartz porphyry. It is known that, though its exposed products are dominantly olivine-bearing, it also gave rise to olivine-free dolerites containing micropegmatitic material. But the evidence is as yet not adequate to determine to what extent the Shag Valley quartz porphyry, occurring four miles from the nearest exposure of a basic product of that magma, could be considered as the product of such an acid differentiate removed by filter-press action at a late stage from the basic magma crystallising at depth, or may have incorporated material derived by assimilation and rheomorphism of invaded sialic rocks.
The writer's indebtedness to Mr Seelye for the analyses of the quartz-porphyry, to Drs. Hutton and Turner for certain optical determinations, to Dr. Turner for helpful suggestions, to Dr. H. J. Finlay for microfaunal observations, and to the Director of the Geological Survey for permission to publish this petrological discussion of a group of rocks in an area falling within the region to be described by Mr. D. A. Brown in a forthcoming bulletin, must be gratefully acknowledged.
—— and Mamlqvist, D., 1935. Zur Geologie und Petrographie der nordöstgrönlandischen Basalt-formations. II. Die Sauren Ergussgesteine von Kap Franklin. Med. om. Grönland, vol. 95, no. 3.
—— 1939. In Barth, Correns and Eskola. Die Enstehung der Gesteine. Julius Springer, Berlin.
Bartrum, J. A., 1925. The Igneous Rocks of North Auckland, New Zealand. Verbeek Gedenboek, Verhandl. Geol. Mijnb. Genootschap v. Ned. en Kolonien, vol. 8, p. 13.
Bastin, E. S., 1938. Hydrothermal Alteration in the Rocks of Pigeon Point, Minnesota. Journ. Geol., vol. 48, pp. 1058–1072.
Bowen, N. L., 1928. The Evolution of Igneous Rocks. Princeton Univ. Press.
Collins, W. H., 1917. Onaping Map Area. Geol. Survey Canada, Geol. Series 77.
—— 1934. Igneous Rocks and the Depths of the Earth. McGraw Hill Book Co., New York.
Dixey, F., 1937. (With Campbell Smith, W., and Bissett, C. B.) The Chilwa Series of Southern Nyassaland. Geol. Surv. Dept., Nyassaland. Bull. No. 5.
Fenner, C. N., 1926. The Katmai Magmatic Province. Journ. Geol., vol. 34, pp. 675–771.
—— 1938. Contact Relations between the Rhyolite and Basalt in Gardiner River, Yellowstone Park. Bull. Geol. Soc. Amer., vol. 49, pp. 1441–1483.
Harker, A., 1904. The Tertiary Igneous Rocks of Skye. Mem. Geol. Surv. U.K., Grt. Britain.
—— 1909. The Natural History of Igneous Rocks. Methuen, London.
Henderson, J., 1913. The Geology of the Aroha Subdivision. N.Z. Geol. Surv., Bull. No. 16.
Iddings, J. P., 1913. Igneous Rocks. Vol. II. John Wiley & Sons Inc., New York.
Jeans, J., 1944. The Universe Around Us. Camb. Univ. Press, 4th Edn.
Lacroix, A., 1904. La Montagne Pelée et ses Eruptions. Paris.
Morgan, P. G., 1924. The Geology of the Waihi District, Hauraki Goldfield. Geol. Surv. N.Z., Bull. No. 26, pp. 36 et seq.
Richey, J. E., and Thomas, H. H., 1930. The Geology of Ardnamurchan. Mem. Geol. Surv. Scot.
Rosenbusch, H., 1908. Mikroskophische Physiographie der Massigen Gesteine. Zweite Halfte. E. Schweizerbartsche Verlagshandlung, Stuttgart.
Schwartz, G. M., and Sandberg, A. E., 1940. Rock Series in Diabase Sills at Duluth, Minnesota. Bull. Geol. Soc. Amer., vol. 51, pp. 1135–1172.
Seelye F. T., 1939. In Williamson (1939), p. 30 q.v.
Thomas, H. H., and Bailey, E. B., 1924. Tertiary and Post Tertiary Geology of Mull, etc. (Petrology). Mem. Geol. Surv., Scotland.
—— 1939. In Williamson, 1939, pp. 40, 43 q.v.
Tyrrell, G. W., 1928. The Geology of Arran. Mem. Geol. Sur., Scotland.
Wager, L. R., and Deer, W. A., 1939. The Petrology of the Skaergaard Intrusion, Kangerdlugssusk, East Greenland. Med. om Grönland, vol. 105, no. 4.
—— 1940. Differentiation of the Palisade Diabase. Bull. Geol. Soc. Amer., vol. 51, pp. 1059–1106.
Washington, H. S., 1917. Chemical Analyses of Igneous Rocks. U.S. Geol. Sur., Prof. Paper 99.
Williamson, J. H., 1939. The Geology of the Naseby Subdivision. N.Z. Geol. Surv., Bull. No. 39.