Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 85, 1957-58
This text is also available in PDF
(4 MB) Opens in new window
– 247 –

The Intrusive Rocks of the Kaikoura Mountains, Marlborough, New Zealand

[Received by the Editor, April 26, 1957.]


The intrusive rocks of the Kaikoura Mountains are mainly gabbros, consisting of varying amounts of plagioclase, olivine, augite, and brown hornblende. Associated with these gabbro intrusions are numerous sills of diverse composition, ranging from extremely mafic to extremely felsic types. The mineralogy and petrography of these rocks is described, and chemical analyses of a hornblende gabbro, a camptonite, a nepheline syenite aplite, and a basalt are given. Analyses are also given of coexisting hornblende and biotite from the gabbro. The rock series is alkalic in nature, with an alkali-lime index of 50.4. The distribution of the different areas of these rocks indicates that they have been separated by transcurrent faulting since intrusion, which is believed to have occurred during the Hokonui Orogeny (Upper Jurassic?).


The presence of intrusive rocks in the Kaikoura Mountains has been known for many years. McKay (1886, 1890), in his pioneer reports on the geology of this extensive region, gave a fairly detailed and at times graphic description of the occurrence of these igneous rocks, but did not concern himself with their petrology except to tag them with names based on macroscopic appearance. Thomson, thirty years later, was obviously impressed by their abundance and variety and wrote a brief paper on their microscopic petrology (1913); this was apparently intended as preliminary to a detailed account, but later he devoted himself to the stratigraphy of the region and his monographic paper (1919) does not contribute further to the earlier description of the igneous rocks except to discuss their distribution and age relations. Since Thomson's work nothing has been added to our knowledge of these rocks save for the description of an analcime-tinguaite pebble from the Wairau Bar, probably derived from this region (Bartrum, 1936), and some details regarding the occurrence of igneous rocks in the northern foothills (King, 1934).

My interest in these rocks was aroused while I was working on the igneous rocks of the Mandamus-Pahau area in North Canterbury in 1944 and 1945. It was evident that the intrusive rocks in this area showed marked resemblances to those described by Thomson from the Kaikouras, seventy miles away to the north-east. During the following years a number of trips were made to investigate the distribution of intrusive rocks throughout this region. In November, 1944, I journeyed from Hanmer to the Clarence-Acheron junction, up the Acheron River to its junction with the Guide, thence through Bullock Head Gully to the Dillon River, up the Dillon, across the upper reaches of the Tweed River, past Lake McRae, and down the Elliot River; return was made by the same route, except that the Dillon was followed to its junction with the Clarence, thence up the Clarence to Jollies Pass. In February, 1946, the lower part of the Middle Clarence Valley was examined, by following the pack trail from Kekerangu to Coverham, thence up the river to Quail Flat, and out to Kaikoura by way of the Seymour and Kahautara Rivers. In February,

– 248 –
Picture icon

Fig. 1.-The Kaikoura Mountains, showing areas of intrusive rocks (lined). Major faults are indicated by heavy solid lines; the heavy dashed line in the Inland Kaikouras indicates the approximate limits of the sills associated with the intrusive rocks.

– 249 –

1954, the Clarence River was traversed by canoe from its junction with the Acheron to the mouth, which enabled the examination of the streams rising from the ranges on either side of the valley; at the same time a detailed traverse was made of the Muzzle River, which was clearly a major source of the igneous rocks in the river gravels. In April, 1954, a brief visit was again paid to the lower part of the Middle Clarence Valley to examine the igneous rocks in the Swale and Mead tributaries, and a rapid trip was made up the Awatere Valley as far as Upcot to determine the distribution of intrusive rocks in this area.


On all the journeys in the field I was accompanied by Mr. A. J. White, whose contribution in organization, companionship, and hard physical labour, cannot be adequately acknowledged in words. I am also greatly indebted to the few residents of this isolated region, particularly Mr. and Mrs. Arthur Roberts of the Bluff Station, for their hospitality and kindness Dr S. R. Taylor has made a valuable contribution to this research by determining the- minor and trace elements in the analysed rocks. My thanks are also due to Mr. Julius Weber, who made the photomicrographs. The costs of the field work and of thin sections and chemical analyses have been largely defrayed by grants from the Hutton Fund of the Royal Society of New Zealand and from the John Simon Guggenheim Memorial Foundation of New York.

Distribution of the Intrusive Rocks

The map (Fig. 1) summarizes the distribution of the intrusive rocks throughout this region. They occur in three separate areas: The Seaward Kaikoura Mountains, the Inland Kaikoura Mountains, and north of the Awatere River in Black Birch Stream. The area in the Seaward Kaikoura Mountains lies in the headwaters of the Dubious and Fidget Rivers, and evidently extends across the watershed into the Hapuku River, since igneous pebbles have been reported from that stream. In the Inland Kaikouras the intrusives are centred on the Tapuaenuku massif; igneous pebbles are abundant in the gravels of the Muzzle, Dart, Branch, and Dee Rivers on the eastern slopes, and in the Hodder on the western slopes: igneous pebbles are also found in the Mead and Swale Rivers but are much less abundant and seem to be derived entirely from small dykes and sills, or from the Great Marlborough Conglomerate. On the northern side of the Awatere Valley intrusive rocks seem to be confined to the headwaters of Black Birch Stream. The intrusions thus show a noteworthy limitation of occurrence, the significance of which will be discussed later.

McKay (1890, p. 130) mentions that igneous pebbles are abundant in the George Creek, draining the north-east flank of the Seaward Kaikouras, and the map accompanying his paper shows an area of intrusive rocks in this region. McKay's report on this was based, however, not on his own observations but on information he had received. This seems to be erroneous; I found no igneous pebbles in the George Creek, nor in any of the streams draining from the Seaward Kaikouras to the Clarence River below the Fidget River.


The minerals in the intrusive rocks were investigated by the standard techniques. Refractive indices were measured by the immersion method, and axial angles and optic orientations were determined in thin sections on a universal stage. Staining with sodium cobaltinitrite (Chayes, 1952) was used to discriminate between potash feldspar and plagioclase, and with methylene blue (Shand, 1939) to detect nepheline. X-ray powder photographs were used to confirm some identifications of fine-grained and isotropic material, Modes were determined by the point-counting technique (Chayes, 1956).

– 250 –

Since the characteristic mineral in many of these rocks is a brown hornblende it was clearly desirable to obtain a precise knowledge of its composition. A pure sample was isolated from a pegmatitic variant of the hornblende gabbro of the Muzzle River, using a combination of magnetic separation and density fractionation with methylene iodide-acetone mixtures. At the same time a pure sample of the biotite which is intergrown with the hornblende was also prepared and analysed chemically, thus providing useful data on the fractionation of certain elements between the hornblende and the biotite structures.


The only feldspar in most of these rocks is a plagioclase, generally a calcic andesine (An40 to An50). The comparatively sodic character of the plagioclase in these gabbroic rocks reflects their generally alkaline character and the presence of much of the calcium in hornblende and augite. The plagioclase is sometimes quite fresh, but is more commonly grey and turbid from beginning alteration. The alteration products were not studied in detail, but generally appeared to be kaolinite; in some specimens zeolites were noted.

Since the gabbros contain notable amounts of potassium, many sections were etched and stained, but potash feldspar was generally absent; a few sections showed up to 5 per cent. in interstitial grains. Evidently the potassium is usually contained in the ferromagnesian minerals, in the biotite and to a considerable extent in the hornblende. Potash feldspar is present inconsiderable amount in the less common felsic differentiates, which include aplites and some monzonites and syenites.


Fresh unaltered nepheline was found in an aplite cutting the hornblende gabbro in the Muzzle River. Here it occurred as irregular grains interstitial to the predominant feldspar, and was evidently the last mineral to crystallize. Although the hornblende gabbro itself shows considerable nepheline in the norm, none is present in the rock, evidently because the excess alkali is contained in the hornblende.

Large phenocrysts, up to 10 mm in diameter, showing the characteristic hexagonal and prismatic forms of nepheline are common and abundant in tinguaites from the Hodder River (Plate 22, Fig. 2). These phenocrysts are, however, completely altered to a fine-grained intergrowth of sericite and analcime. From the formulas of these minerals it is clear that this alteration cannot be interpreted as a simple hydration of a nepheline with the normal Na : K ratio of 3 : 1; it probably represents a late magmatic reaction of hydration and concomitant removal of part of the alkalis.


Besides the analcime in the nepheline pseudomorphs, this mineral also forms the groundmass in these tinguaites. It is greyish or brownish in colour from the presence of numerous tiny inclusions, but is otherwise quite fresh and is almost certainly a primary mineral in these rocks.


Olivine occurs only in the more mafic gabbros and camptonites, and is never abundant, being generally less than 10 per cent and seldom more than 20 per cent of the rock. In the gabbros it is generally fresh or shows only a beginning alteration, but in the camptonites it is often completely altered to chlorite or serpentine. Olivine from the gabbros has α = 1.678, indicating a magnesium-rich variety (Fo80).


Augite is a common and abundant mineral in most of the gabbros and related dyke rocks. It is a pale violet non-pleochroic variety, commonly twinned on (100). Zonal structure is not marked, except in phenocrysts in some of the camptonites. The

– 251 –

augite from one of the gabbros had β = 1 685, 2V = 51°, corresponding to a composition Ca42Mg48Fe12, according to Hess (1949). Brown hornblende is frequently intergrown with augite, with the b and c axes in common; the augite is then usually surrounded by a thick layer of hornblende, but occasionally the hornblende appears as irregular patches within the augite crystals.

Soda pyroxenes, aegirine or aegirine-augite, are present in minor amounts in a nepheline-syenite aplite from the Muzzle River and in the analcime groundmass of tinguaites from the Hodder River.


Brown hornblende is a common and characteristic mineral in most of the gabroic rocks. In some of the rocks it has formed in crystallographic continuity with earlier formed augite, but it also forms independently of this mineral. In an intrusive mass in the Muzzle River it is the sole ferromagnesian mineral, save for accessory amounts of biotite.

The hornblende in a pegmatitic phase of the Muzzle River intrusive was separated and analysed chemically. The results are given in Table I.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table I.—Analysis of Hornblende from Gabbro, Muzzle River.
Analyst: M. E. Coller.
Wt. per cent. Unit cell Content. Theoretical.
SiO2 38.0 Si 11.52
16 16
TiO2 4.13 AlIV 4.48
Al2O* 15.0 AlVI 0.88
Fe2O2 1.69 Ti 0.94
FeO 11.70 Fe3 0.39
10.39 10
MgO 11.43 Fe2 2.96
MnO 0.25 Mn 0.06
CaO 11.74 Mg 5.16
Na2O 2.80 Ca 3.81
K2O 1.48 Na 1.64
6.02 6
H2O+ 1.40 K 0.57
H2O- 0.04 O 45.23
48.05 48
Total 99.66 OH 2.82

Ideal formula: NaCa2(MgFeAlTi)5(AlSi3O11)2 (O2OH)2

Recalculation into empirical unit cell contents, using the measured density (3.22) and the lattice dimensions (α = 9.85 kX, b = 18.17 kX, c = 5.39 kX, β = 105° 45′) of a hornblende of similar composition from Kaersut, Greenland (Gossner and Spielberger, 1929) shows that the composition agrees well with the ideal hornblende formula.

The optical properties of this hornblende are as follows: α = 1.676, β = 1.695, γ = 1.705; optically negative, 2V = 74°; Z ∧ c = 12°; strongly pleochroic, × = pale brown, Y = Z = dark reddish brown.

– 252 –

The nomenclature of the amphibole group is in a confused state and an appropriate subspecific name for this particular variety is not easily arrived at. In most of its chemical and physical properties it agrees with kaersutite, but this name is generally restricted to brown hornblendes containing more than 5 per cent TiO2. It also resembles barkevikite, but barkevikite from the type locality in Norway is definitely an alkaline amphibole, containing more than 5 per cent Na2O. On the whole it is closest to the variety syntagmatite, described originally from the basalts of Mt. Vesuvius, but the non-committal appellation brown hornblende is most satisfactory.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table II.—Analysis of Biotite from Gabbro, Muzzle River.
Analyst: M. E. Coller.
Wt. per cent. Unit cell Content. Theoretical.
SiO2 35.1 Si 10.77
16.00 16
TiO2 4.05 AlIV 5.23
Al2O3 17.8 AlVI 1.21
Fe2O3 0.24 Ti 0.94
FeO 17.94 Fe3 0.06
MgO 10.07 Fe2 4.61 11.96 12
MnO 0.29 Mg 4.61
CaO 2.35 Mn 0.08
Na2O 1.09 Ca 0.45
K2O 7.74 Ca 0.32
H2O+ 2.46 Na 0.65
4.00 4
H2O- 0.24 K 3.03
O 42.55
47.59 48
Total 99.37 OH 5.04

Ideal formula: (KNaCa)2(CaMgFeAl)6, (AlS1)8O20(O,OH)4


The hornblende in the Muzzle River gabbro is intergrown with minor amounts of biotite, and a pure fraction was separated and analysed (Table 2) . Recalculation into empirical unit cell contents, using the measured density (3.08) and the lattice dimensions (α = 5.30 kX, b = 9.21 kX, c = 20.32 kX, β = 99° 18′) for biotite given by Strunz (1949) shows that the composition agrees well with the accepted biotite formula; the calcium is divided approximately equally between the six- and eight- co-ordinate lattice positions, and some of the OH- positions may be filled by O2- ions.

The optical properties of this biotite are: α = 1.624, β = 1.673, γ = 1.674; optically negative, 2V = 10°; strongly pleochroic, × = pale yellow, Y = Z = brown-black, almost opaque.

In view of the fact that the biotite and hornblende are intergrown and have presumably crystallized at the same time in equilibrium with each other and the liquiid, it is interesting to compare their compositions in terms of the 48 (O, OH) ions in their unit cells (Table 3) . From the nature of the structures calcium is much.

– 253 –
Table III—Comparison of Hornblende and Biotite Compositions, in Terms of 48 (O,OH) Ions in the Unit cells.
Hornblende Biotite Hornblende Biotite
Si 11.52 10.77 Mg 5.16 4.61
AlIV 4.48 5.23 Mn 0.06 0.08
AlVI 0.88 1.21 Ca 3.81 0.77
Ti 0.94 094 Na 1.64 0.65
Fe3 0.39 0.06 K 0.57 3.03
Fe0 2.96 4.61

more abundant in the hornblende and potassium in the biotite. Apart from these differences the most marked variation is in the relative amounts of iron and magnesium; the Mg: Fe ratio is 1: 1 in the biotite and about 1.5: 1 in the hornblende, i.e.,-the iron content in the biotite is considerably greater than in the co-existing hornblende. This decrease in the Mg: Fe ratio seems to be a general feature in the discontinuous reaction series augite-hornblende-biotite in igneous rocks crystallized from the same initial magma; augite from the Muzzle River intrusives has an Mg: Fe ratio approximately 4: 1, judging from its optical properties. Another significant difference in the compositions of the biotite and hornblende is the some what greater aluminium content in the biotite, both in the four-co-ordinate and six-co-ordinate positions; this increase in aluminium content is also a general feature in going down the discontinuous reaction series.

Accessory Minerals

Apatite is a constant accessory in all the gabbroic rocks. It occurs in euhedral prismatic crystals, generally from 0.05 to 0.5 mm long, and is colourless and free from pigmental inclusions. Sphene is present in practically all the rocks except the most mafic types. Trifling amounts of zircon occur in highly felsic rocks such as the nepheline syenite aplite. The opaque material is largely magnetite, sometimes accompanied by pyrite; despite the considerable amount of titanium in the basic rocks little of it is present as ilmenite, most of it being in the ferromagnesian silicates.


A. The Inland Kaikouras

The intrusive rocks of the Inland Kaikoura Mountains comprise gabbros in the Mt. Tapuaenuku-Mt. Alarm massif and associated dykes and sills which occur along the range from the headwaters of the Bluff River in the south to the headwaters of the Mead and Medway Rivers in the north. Pebbles of dyke rocks are rare in the Bluff River, but are more abundant in Bluff Creek, the next stream to the north. The Muzzle River carries large amounts of gabbros and dyke rocks, which are most abundant in the main stream, but its southern tributaries bring down minor amounts of dyke rocks. The Dart, Branch, and Dee Streams carry large amounts of gabbros and dyke rocks. Pebbles of gabbros and dyke rocks are not uncommon in the Mead and the Swale Streams; however, only dyke rocks were observed in the greywackes along the Mead Stream, and since the Great Marlborough Conglomerate contains numerous pebbles of the igneous rocks, much of the igneous material in the Mead and the Swale is probably derived from this formation. On the west side of the mountains gabbros and dyke rocks make up a large part of the gravels of the Hodder River; the igneous rocks are considerably less varied and less abundant in the Shin River, and igneous pebbles are comparatively rare in the other streams entering the Awatere River from the Inland Kaikouras. Evidently the plutonic intrusives

– 254 –

are concentrated in the high peaks of Mt. Tapuaenuku and Mt. Alarm; in fact, the high altitude of these peaks (about 2,000ft greater than the rest of this range) is probably the result of the buttressing effect of the massive plutonic rocks. The dykes and sills extend beyond the plutonic rocks for some miles in each direction along the axis of the range, falling off in abundance as the distance from the plutonic centre increases.

The intrusive rocks of the Muzzle River were examined in detail. The presence of igneous rocks in the watershed of this river is manifested by the abundance of coarse hornblende-feldspar rocks in its gravels. These rocks were traced to their source, which proved to be an area about one square mile on the south-west flank of Mt. Alarm.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table IV—Analysis of Hornblende Gabbro, One Mile South-West of the Summit of Mt. Alarm.
(Provisional 1 Mile Sheet S42, 925385.) Analyst: H. B. Wiik.
Weight per cent. Mol prop Norm. Mode (Weight per cent)
SiO2 42.26 7008 Or 9.7 Plagioclase (An48) 46.5
TiO2 3.80 474 Ab 11.6 Hornblende 42.7
Al2O3 17.41 1703 An 27.2 Biotite 4.7
Fe2O3 2.76 173 Ne 13.4 Opaque 3.2
FeO 10.06 1400 Sphene 1.5
MnO 0.25 35 Σ salic 61.9 Apatite 1.4
MgO 5.26 1305
CaO 10.05 1792 Ap 1.2 Total 100.0
Na2O 3.65 589 II 5.5
K2O 1.58 168 Mt 3.0
P2O5 0.55 39 Di 16.5
CO2 0.00 OI 11.9
H2O+ 2.33
H2O- 0.05 Σ femic 38.1
D = 2.93

Minor and trace elements in p. p.m. (analyst S. R. Taylor): Ga 23, Cr 18, Li 10, Ni 80, Co 35, Cu 45, V 330, Zr 190, Sc. 22, Sr 500, Ba 210, Rb 57, Pb and Cs not detected.

In hand specimen the typical rock consists of approximately equal amounts of brown hornblende and plagioclase, the brown hornblende occurring as crystals up to 5 mm long in the plagioclase groundmass (Plate 20, Fig. 1). The chemical analysis, norm (calculated according to Barth, 1952, p. 80), and mode are given in Table 4.

Under the microscope the plagioclase is often grey and turbid from beginning alteration. The alteration products may be largely zeolites, since they are stained by methylene blue after the rock is etched with phosphoric acid. The refractive index α of the plagioclase is about 1 553, indicating a composition An48. Because of its comparatively high potassium content, the rock was tested for potash feldspar by staining with sodium cobaltinitrite, but none was found. Nepheline, although indicated by the norm, is not present in the rock. The relatively high content of sodium and potassium is the reflection of the presence of about 5 per cent biotite, and of the alkaline nature of the hornblende. Accessory minerals include sphene, apatite, and opaque material (magnetite and ilmenite).

This rock is essentially a gabbro, using this name in a wide sense. It is distinctly alkaline in character, and is similar in composition to many theralites and essexites; however, it contains no feldspathoid, the excess alkalis being taken up in the

– 255 –

ferromagnesian minerals. It is very close in chemical and minerological composition to mafraite from the Tifao de Mafra, Cintra, Portugal (Lacroix, 1920).

The gabbroic rock varies somewhat from place to place in the relative amounts of hornblende and plagioclase; occasionally hornblende makes up as much as 70 to 80 per cent, of the rock. The amount of biotite also varies, some specimens being considerably richer in this mineral than the analysed sample. The gabbro is cut by veins of coarse hornblende pegmatite, in which the individual hornblende crystals are up to 10 cm long and 1 cm across (Plate 20, Fig. 2). The mineralogical composition of these pegmatite veins is similar to that of the main intrusive. The gabbro is also cut by narrow feldspathic dykes practically free from dark minerals. One of these has been analysed chemically and mineralogically (Table 5; Plate 21, Fig. 1).

Table V—Analysis of Nepheline Syenite Aplite, One Mile South-West of the Summit of Mt. Alarm.
(Provisional 1 Mile Sheet S42,925385.) Analyst: H. B. Wiik.
Weight per cent. Mol. prop Norm. Mode(Weight per cent)
SiO2 58.19 9650 Or 32.4 Orthoclase 45.9
TiO3 0.46 57 Ab 39.3 Oligoclase 33.8
Al2O3 21.39 2093 An 10.1 Nepheline 15.1
Fe2O3 0.93 58 Ne 13.8 Hornblende, Sphene,
FeO 1.64 228 Opaque 5.2
MgO 0.27 67 Σ salic 95.6
CaO 2.60 463 Total 100.0
Na2O 7.03 1134 II 0.6
K2O 5.57 591 Mt 1.0
MnO 0.01 Di 2.1
P2O3 0.01 OI 0.7
CO2 0.09
H2O+ 1.18 Σ femic 4.4
H2O- 0.16

Minor and trace elements, in p.p.m. (analyst S. R. Taylor): Ga 18, Li 31, Ni 25, Co 9, Cu 12, V 10, Zr 980, Sr 530, Ba 670, Rb 220. Cs<5, Cr, Sc, and Pb, not detected.

It consists essentially of orthoclase, oligoclase, and nepheline, with accessory amounts of sphene, hornblende (sometimes with cores of greenish augite), and opaque material; rare zircon crystals were noted. The oligoclase occurs as lath-like crystals up to 1 mm long. Staining with sodium cobaltinitrite shows that the orthoclase occurs partly as individual crystals similar in form to the oligoclase and partly in perthitic admixture with the oligoclase. The nepheline is present as small irregular grains between the feldspar crystals, and probably was the last mineral to form. The rock can best be termed a nepheline-syenite aplite.

The contacts between the gabbro and the greywacke which forms the country rock are sharp, but numerous apophyses of the gabbro penetrate the greywacke. At the contact the greywacke is recrystallized to a tough fine-grained hornfels with a distinct red-brown colour; this colour is due to the presence of a considerable amount of brown hornblende and some biotite (Plate 21, Fig. 2). These minerals represent in part the original ferromagnesian material of the greywacke, but some is probably the result of introduced iron, magnesium, and alkali from the magma.

– 256 –
Table VI.—Analysis of Camptonite Dike, at Junction of Two Branches of Muzzle River, 3½ Miles South of the Summit of Mt. Alarm.
(Provisional 1 Mile Sheet S42, 921333) Analyst: H. B. Wiik.
Weight per cent. Mol. prop. Norm. Mode (Weight per cent.)
SiO2 43.90 7280 Or 3.9 Augite 33.0
TiO2 3.13 391 Ab 12.6 Plagioclase 22.8
Al2O3 11.15 1091 An 24.1 Chlorite 21.3
Fe2O3 2.44 153 (after olivine)
FeO 9.42 1311 Σ salic 40.6 Hornblende 12.9
MnO 0.22 31 Opaque 10.0
MgO 9.86 2445 AP 0.5
CaO 13.24 2360 11 4.6 Total 100.0
Na2O 1.32 213 Mt 2.4
K2O 0.62 66 Di 35.3
P2O3 0.26 18 Hy 3.3
CO2 0.17 Ol 13.3
H2O+ 3.87
H2O- 0.35 Σ femic 59.4
D = 3.04

Minor and trace elements, in p.p.m. (analyst S. R. Taylor): Ga 17, Cr 1150, Li 12, Ni 430, Co 85, Cu 115, V 850, Zr 320, Sc 34, Sr 225, Ba 100, Rb 24, Pb and Cs, not detected.

The greywackes surrounding the intrusion are seamed with dykes and sills, most of which are fine-grained grey rocks with prominent augite phenocrysts up to 10 mm in diameter. They are particularly well exposed at the junction of two branches of the Muzzle River at the end of the spur named ” The Tongue ” on the Provisional One-Mile Sheet. The dykes at this place were commented on by McKay (1886, p. 51). A typical dyke from this point has been analysed chemically and mineral-ogically (Table 6; Plate 22, Fig. 1). The rock consists of phenocrysts of augite (up to 5 mm across) and altered olivine (up to 3 mm across) in a panidiomorphic groundmass of augite, brown hornblende, plagioclase, and opaque material. The olivine is completely replaced by chlorite, or a mixture of chlorite and serpentine. The plagioclase is greatly decomposed and its composition is not directly determin-able; judging from the chemical analysis it is probably a labradorite. The rock can be described as a porphyritic camptonite.

The hornblende gabbro evidently extends across the watershed into the upper valley of the Hodder River. This extension was not traced on the ground, but boulders of identical gabbro were found in the Hodder River gravels. Many of the gabbros in the Hodder River contain augite as well as brown hornblende, and in some it is the dominant mineral, making up 50 per cent or more of the rock. These highly augitic rocks usually contain some olivine. Some boulders of highly feld-spathic plutonic rocks were also collected; the feldspar in these is much altered, but they appear to be either syenites or monzonites.

The Hodder River also carries down a variety of dyke rocks, including types not seen elsewhere. These are dense fine-grained rocks with prominent nepheline phenocrysts up to 10 mm in diameter; the nepheline is, however, completely altered to a mixture of sericite and analcime (Plate 22, Fig. 2). The groundmass of the nepheline-bearing rocks is turbid and isotropic, and X-ray powder photographs show that it

– 257 –

consists of analcime. The groundmass often encloses crystals of brown hornblende, pale green augite, sphene, and apatite. In one specimen the analcime groundmass was crowded with tiny acicular crystals of aegirine or aegirine-augite.

The analcime-bearing rocks found as pebbles on the beach at Cloudy Bay by Dr. C. R. Laws and described by Bartrum (1936) are identical with these rocks and must have been carried down from the Hodder River. Bartrum classified them as analcine tinguaites; they might also be considered as monchiquites.

Plutonic intrusions evidently make up a large part of the Tapuaenuku massif, boulders of gabbro being common in the Dart, Branch, and Dee Streams. In most of these augite is the most abundant ferro-magnesian mineral; some brown hornblende is generally present, and accessory amounts of biotite. Olivine occurs in those variants richest in augite (Plate 23, Fig. 1). Modal analysis of a number of specimens showed the following ranges in abundance of the different minerals (in volume per cent): plagioclase, 19 per cent-68 per cent; augite, 7 per cent-53 per cent; hornblende 0–30 per cent. The average composition was plagioclase 42 per cent, augite 25 per cent, hornblende 20 per cent. From the modal analyses of these specimens it was obvious that the amounts of augite and hornblende were related, their sum being fairly constant at about 40 per cent to 45 per cent. In a few specimens augite was present to the exclusion of hornblende, but the reverse relationship, typified by the hornblende gabbros of the Muzzle and Hodder Rivers, was not observed. A few specimens of gabbro from the Dee River contain accessory amounts (about 5 per cent.) of potash feldspar. These augite-hornblende gabbros are very similar to the gabbro of the Mandamus River area (Mason, 1951) and like it can be classed as kauaiite (Cross, 1915, p. 16).

Some boulders of syenite also were collected in these streams. These rocks consist essentially of microperthite, with accessory amounts of biotite, brown hornblende, and occasionally a green uralitic amphibole, probably pseudomorphous after augite Plagioclase in minor amounts sometimes accompanies the microperthite. Some specimens contain a little quartz (up to about 10 per cent). These rocks show distinct mineralogical affinities with the gabbros, and probably represent minor differentiates of the original magma.

Dykes and sills extend far beyond the region of plutonic intrusions, as indicated on the map. These minor intrusions are mainly sills, generally two to four feet in thickness, but occasionally up to twenty feet. The general strike is about NE-SW, and their area of occurrence is also clongated, in this direction. These rocks are usually camptonite, with dominant brown hornblende and augite; often the augite is present mainly as large phenocrysts. The more basic varieties sometimes contain a small amount of serpentinized olivine. A few dyke rocks consisting largely of feldspar, either trachytes or andesites, were also collected.

B. Black Birch Stream

Black Birch Stream enters the Awatere River from the north-west about nine miles upstream from the bridge at Seddon. Its gravels contain large amounts of coarsely crystalline igneous rocks. Boulders of similar rocks are found in the Blairich River, the next stream to the north, but not in the same variety and abundance. This suggests that the outcrops from which these rocks have been derived are located in the area between Black Birch Stream and the Blairich River, mainly within the watershed of Black Birch Stream.

The igneous rocks are mostly gabbros identical in all respects with those of the Tapuaenuku massif. A few specimens of lamprophyres consisting of plagioclase, brown hornblende, and augite were also collected. Usually the gabbros consist essentially of plagioclase, augite, and brown hornblende, with accessory amounts of biotite; a little olivine is occasionally present. The relative amounts of augite and

– 258 –

brown hornblende vary considerably; in some specimens augite is reduced to accessory amounts or is missing entirely, and brown hornblende is very abundant; in others augite is the dominant ferromagnesian mineral, the amount of hornblende being quite small.

King (1934, p. 31) described the occurrence of a coarse-grained igneous rock ” in the vallev of Black Birch Creek, some miles from the mouth “. This rock was a coarse-grained aggregate of brown hornblende and feldspar. The feldspar was somewhat decomposed, and King described it as a mixture of albite, oligoclase, and orthoclase; he classified the rock as a syenite-pegmatite. From his description the rock is very similar to hornblende-bearing pegmatites of the Muzzle River intrusive, and it evidently represents a pegmatitic derivative of the gabbros.

C. The Seaward Kaikouras

Intrusive rocks in the Seaward Kaikouras are evidently limited to the area in the headwaters of the Dubious and Fidget Rivers, and extending across the watershed into the Hapuku River drainage, since only in these streams have pebbles of these rocks been found. In the Dubious River and the streams draining into the Clarence from the Seaward Kaikouras above the Dubious basalt pebbles occur, but these are derived from flows within the Clarentian sediments.

The intrusive rocks in the Dubious and Fidget Rivers are mainly gabbros, resembling closely those of the Tapuaenuku massif. They generally consist of varying proportions of plagioclase (about An50), augite, and brown hornblende, often with some olivine, and usually accessorv (< 5 per cent) biotitc. Sometimes plagioclase is practically absent, giving an ultra-basic variant consisting of approximately equal amounts of augite, olivine, and brown hornblende; the olivine is often partly or wholly replaced by chlorite and serpentine. In the Fidget River some pebbles of lamprophyres were collected, which proved to be camptonites, containing abundant brown hornblende, generally some augite (often as phenocrysts), and some olivine (altered to serpentine and chlorite).

D. The Clarence River Basalts

These rocks are distinct from the intrusive rocks with which this paper is concerned, in petrology, in mode of occurrence, and probably in age. However, during this investigation their occurrence was studied and samples collected for petrographic examination. They occur as thick flows within the Clarentian sediments on the eastern side of the Clarence Valley, from the Dubious River to the Gore River. In hand specimen they are dark grey to black rocks, generally aphanitic, sometimes with phenocrysts of augite up to 3 mm in diameter. They often contain amygdules filled with calcite. Under the microscope they are seen to be made up of tabular laths of plagioclase (generally an andesine about An40) in a groundmass of small crystals of augite and opaque material. In the analysed specimen (Table 7; Plate 23, Fig. 2) the interstitial ferromagnesian material is chloritic. The composition of this analysed specimen differs somewhat from that of a normal basalt; although SiO2 and Al2O3 are within the usual range, the alkalis are high and CaO and MgO low. In spite of the silica content being below 50 per cent, there is a little quartz in the norm. The measured composition of plagioclase is An38, close to that indicated by the norm; orthoclase is probably present also, although not certainly identified. This rock has distinct spilitic affinities. Study of thin sections of different specimens of the Clarence River basalts indicates that they can be generally classed as tholeiites.


The chemical and mineralogical affinities of the intrusive rocks of the Kaikoura Mountains are most simply and clearly expressed by the alkali-lime index (Peacock, 1931). By plotting the analyses of the igneous rocks of the Muzzle River area the alkali-lime index is found to be 50.4, which places these rocks in Peacock's alkalic.

Picture icon

Fig. 1.—Photomicrograph of analysed hornblende gabbro, Muzzle River Hornblende (grey and black), with numerous inclusions of small apatite crystals, in plagioclase (white). Length of scale is 5 mm.
Fig. 2.—Hand specimen of hornblende-bearing pegmatite, Muzzle River, showing contact with hornblende gabbro on lower light.

Picture icon

Fig. 1.—Photomicrograph of analysed nepheline syenite aplite, Muzzle River. Oligoclase and nepheline, white, orthoclase, grey (stained with sodium cobaltinitrite), hornblende, black. Length of scale is 1 mm
Fig. 2.—Photomicrograph of hornfels (contact-metamorphosed greywacke), Muzzle River. Hornblende and biotite (grey to black) in quartz and feldspar (white). Length of scale is 0.5 mm.

Picture icon

Fig. 1.—Photomicrograph of analysed camptonite, Muzzle River Augite (light grey), hornblende (dark grey), and magnetite and ilmenite (black), in a groundmass of chlorite, serpentine, and altered feldspar. Length of scale is 0.5 mm.
Fig. 2.—Photomicrograph of nepheline tinguaite, Hodder River Phenocryst of nepheline (altered to senate and analcime) in a groundmass of analcime which contains numerous small crystals of aegirme-augite and occasional sphere. Length of scale is 0.5 mm.

Picture icon

Fig. 1.—Photomicrograph of olivine gabbro, Dee River Olivine (light grey, many cracks), augite (light grey, darker borders), and hornblende (dark grey and black), with interstitial feldspar (white). Length of scale is 10 mm.
Fig. 2.—Photomicrograph of analysed basalt, Clarence River. Lath-like plagioclase crystals (white), with interstitial chlorite (grey), and magnetite crystals (black). Length of scale is 0.5 mm.

– 259 –
Table VII.—Analysis of Basalt, Clarence River, 1½ Miles Upstream from Junction With Dubious Stream.
(Provisional 1 Mile Sheet S42, 849220.) Analyst: H. B. Wiik.
Weight per cent. Mol prop Norm* Weight per cent.
SiO2 45.98 7625 Or 13.6 Feldspar 70.2
TiO2 3.80 474 Ab 37.8 Ghlorite 23.3
Al2O3 16.50 1614 An 22.3 Opaque 5.0
Fe2O3 5.41 3.39 C 0.5 Calcite 1.5
FeO 6.28 874 Q 0.3
MnO 0.16 23 Total 100.0
MgO 3.91 970 Σ salic 74.5
CaO 5.30 945
Na2O 3.84 619 Ap 0.6
K2O 2.10 223 11 5.8
P2O3 0.26 18 Mt 6.2
CO2 0.67 152 Hy 12.9
H2O+ 4.26
H2O- 1.63 Σ femic 25.5
D = 2.67 (probably low because of gas holes).

Minor and trace elements, in p.p.m. (analyst S. R. Taylor): Ga 18, Cr 14, Li 22, Ni 240, Co 70, Cu 60, V 55, Zr 410, Sc. 18, Sr 400, Ba 330, Rh 56. Pb and Cs, not detected.

group, close to the boundary with the alkali-calcic group, which is at an alkali-lime index of 51. This correlates well with the mineralogy of the rocks. In general terms the alkalic group and the alkali-calcic group are distinguished mineralogically by the occurrence of feldspathoids in the former and their absence in the latter. The Kaikoura intrusive rocks do not contain feldspathoids (except for extreme differentiates like the nepheline-syenite aplite), but are distinctly alkaline in nature. The nepheline in the norm is expressed mineralogically by the presence of biotite and alkali-rich hornblende.

The close similarity of the rocks from all three areas is strong evidence for their having formed from a common parent magma. The most abundant rock type is a gabbro with approximately equal amounts of augite and brown hornblende, and this probably represents fairly closely the composition of the parent magma. The degree of differentiation within the plutonic intrusives is not great, but the dykes and sills show that small amounts of highly alkaline and salic magma were ultimately developed Among the plutonic rocks separation of early-formed ferromagnesian minerals gave rise to ultrabasic variants practically free of feldspar, consisting in large part of augite, together with some olivine and brown hornblende. The various types of gabbro differ essentially in the relative amounts of plagioclase and ferro magnesian minerals, and in the proportions between olivine, augite, and brown hornblende. The trend is one of increasing amounts of plagioclase and brown hornblende coupled with the disappearance of olivine and decrease in augite content, resulting in the formation of hornblende gabbro such as that from the Muzzle River. The presence of monzonites and syenites in the Tapuaenuku massif indicates that in this area differentiation has proceeded even further, but along the same general lines.

The dyke rocks illustrate a greater degree of magmatic fractionation than the plutonic rocks. The commonest types are camptonites either close to the undifferenti-ated primary magma in composition, or enriched in early-formed ferromagnesian.

[Footnote] * After deducting CaCO3.

– 260 –

minerals. Some appear to have been intruded as a porridge of small crystals of augite, brown hornblende, and olivine suspended in an interstitial feldspathic liquid. These mafic differentiates are balanced by the felsic varieties typified by the nepheline syenite aplite of the Muzzle intrusive and the analcime tinguaites of the Hodder River. In general the intrusives of the Inland Kaikouras are more extensive areally and more diversified petrographically than those of the other two areas. This may reflect the position of the present erosion surface with respect to the igneous intrusions the surface in the Inland Kaikouras intersecting the intrusions at a relatively lower level than in the Seaward Kaikouras and the Black Birch Stream area.

The distribution of minor and trace elements in the analysed rocks follows a pattern that agrees on the whole with Goldschmidt's rules for the capture or admittance of ions by crystal lattices of the common rock-forming minerals. The amount of gallium varies little, although it is somewhat depleted with respect to aluminium in the aplite. Chromium in very low in all rocks except the camptonite, which contains over 0.1 per cent; this suggests it is present in the magnetite which is especially abundant in this rock. Nickel, cobalt, and copper show a similar pattern of distribution. The vanadium content is directly related to the amount of mafic minerals, the vanadium evidently proxying for ferric iron; the small amount of vanadium in the Clarence River basalt is noticeably out of line, and may be cited as additional evidence for a different magmatic source for this rock Zirconium shows the expected concentration in the aplite, but is also unusually abundant in the basic rocks, the usual average for a basalt being about 100 p.p.m. The amount of scandium shows a direct relation to the quantity of ferromagnesian minerals, and the amount of strontium with the quantity of plagioclase. Barium and rubidium evidently follow potassium closely.

Age and Correlation

The intrusion of these rocks was probably associated in time with the Hokonui orogeny. They are injected into greywackes and argillites whose age has not been directly established by fossil evidence, but which can be correlated on lithological and structural similarities with fossiliferous Triassic and Jurassic rocks in other parts of the South Island. Pebbles of these intrusive rocks are abundant in the Great Marlborough Conglomerate (Pliocene), showing that they were exposed to erosion at that time (Thomson's statement, 1919, p. 341, to the contrary seems to be erroneous.) Pebbles of igneous rocks occur in the conglomerates at the base of the Clarentian (McKay, 1886; Thomson, 1919) but it is not clear from their descriptions whether these rocks are similar to the intrusives described here. The fact that the intrusions are confined to the pre-Clarentian rocks indicates a pre-Clarentian age for them. However, it has been suggested that the Clarentian basalts are their extrusive equivalents. The question was discussed at some length by Thomson (1919), who concluded that the correlation was justified. It was hoped that chemical analysis of the basalt would settle this point, by showing clear-cut chemical differences, but the evidence is not unequivocal. The analysed basalt is slightly over-saturated, in contrast to the uniformly undersaturated nature of the intrusives; however, it shares with the intrusives a markedly alkaline character.

The intrusions thus either acompanied the Hokonui orogenic movements or occurred during the Clarentian sedimentation that followed. The fact that the intrusions show strong jointing like that of the enclosing greywacke suggests that they have been subjected to similar forces, and hence were emplaced and solidified before the orogenic movements were completely spent. On the whole I favour a pre-Clarentian rather than a Clarentian age for these rocks. A clinching argument would be the identification of pebbles of these rocks in the Clarentian conglomerates.

These intrusions can be correlated with those of Mandamus-Pahau area (Mason, 1951). The alkali-lime index of those rocks is 52.5, quite close to that of the Kai.

– 261 –

koura intrusives. The gabbro of the Mandamus-Pahau area is identical with many of the Kaikoura gabbros, and the dyke rocks in the two regions are very similar. The Mandamus-Pahau intrusions were dated on geological evidence as accompanying the Hokonui orogeny.

We thus have a series of four areas of comagmatic intrusions, distributed along a NE-SW strip of country about 100 miles in length. This distribution has interesting structural implications. Fig. 1 shows that each area of intrusions in the Kaikoura region is separated by one of the major faults from its nearest neighbour. The Black Birch Stream area lies north of the Awatere Fault, the Inland Kaikouras area between the Awatere Fault and the Clarence Fault, the Seaward Kaikouras area between the Clarence and the Kaikoura Fault. The Mandamus-Pahau area in turn lies in the block south of the Hope Fault, the continuation of the Kaikoura Fault. These intrusions thus show an en echelon relationship to each other, a relationship clearly associated with the major faulting. The transverse separation in a NE-SW direction between the intrusive centres is as follows:

Mandamus-Pahau area - Seaward Kaikoura area 60 miles
Seaward Kaikoura area - Inland Kaikoura area 12 miles
Inland Kaikoura area - Black Birch Stream area 18 miles

This relationship between these comagmatic intrusions is in striking agreement with Wellman's concept (1955) of the predominantly transcurrent nature of these major faults. The geographical distribution of these four relatively small areas of intrusive rocks, separated by regions completely free from such intrusions, is most intriguing It is difficult to conceive of a body of magma of uniform composition underlying a region 100 miles long and giving rise only to foui small widely separated intrusions. It is more reasonable to conclude that these intrusions were close together when formed, and that their present separation is largely the result of transcurrent movement along these major faults. If this is correct, then we have some indication of the amount of movement among these faults since the beginning of the Cretaceous. The movement on each fault is in the same sense, in that the block to the north west of each fault has been shifted in a north-easterly direction. The amount of movement has varied considerably from one fault to another; in particular, movement along the Hope-Kaikoura Fault has been of the order of 60 miles, while movement on the Clarence Fault and on the Awatere Fault has been between 10 and 20 miles. If the concept of transcurrent movement on these faults is correct, it should be possible to arrive at more precise estimates of the amount of movement over different times by a comparison of Cretaceous and Tertiary sections south-east of the Hope-Kaikoura Fault with those in the Middle Clarence Valley, and of the latter with those in the Awatere Valley.


Barth, T. F. W., 1952. Theoretical Petrology. New York. John Wiley & Sons.

Bartrum, J. A., 1936. Analcite Rock from Cloudy Bay, Marlborough, New Zealand. N.Z. Jour. Sci. Tech., 18, pp. 120–123.

Chayes, F., 1952. Notes on the stamning of potash feldsoai. with sodium cobaltmitute in thin section. Am. Mineral, 37, pp. 337–340.

—— 1956. Petrographic Modal Analysis. New York: John Wiley & Sons.

Cross, W., 1915. Lavas of Hawaii and their relations. U.S. Geol. Surv. Prof. Paper 88, 97 pp.

Gossner, B., and Spielberger, F., 1929. Chemische und roentgenographische Untersuchungen an Sihkaten. Ein Beitrag zur Kenntniss der Hornblendegruppe. Zeits. Krist., 72, pp. 111–142.

Hess, H. H., 1949. Chemical composition and optical properties of common clinopyroxenes. Am. Mineral., 34, pp. 621–666.

King, L. C., 1934. The geology of the Lower Awatere district, Marlborough. New Zealand. N. Z. Dept., Sci. Ind. Research, Geol. Mem. 2, 49 pp.

– 262 –

Lacroix, A., 1920. La systematique des roches grenues á plagioclases et feldspathoides. Comptes Rendus Acad. Sci. (Paris), 170, pp. 20–25.

Mason, B., 1951. The syenite and associated rocks of the Mandamus-Pahau area, North Canterbury, New Zealand. Trans. Roy Soc N.Z., 79, 260–275.

McKay, A., 1886. On the geology of the eastern part of Marlborough Provincial District. Rep. Geol. Explor. during 1885, pp. 27–136.

—— 1890. On the geology of Marlborough and the Amuri district of Nelson. Rep. Geol. Explor. during 1888–89, pp. 85–185.

Peacock, M. A., 1931. Classification of igneous rock series. Jour. Geol, 39, pp. 54–67.

Shand, S. J., 1939. On the staining of feldspathoids, and on zonal structure in nephelne. Am. Mineral., 24, 508–513.

Strunz, H., 1949. Mineralogische Tabellen. Leipzig: Akademische Verlagsgesellschaft.

Thomson, J. A., 1913. On the igneous intrusions of Mount Tapuaenuku, Marlborough. Trans. N. Z. Inst., 45, pp. 308–315.

—— 1919. The geology of the Middle Clarence and Ure Valleys, East Marlborough, New Zealand. Trans. N.Z. Inst., 51, pp. 289–349.

Wellman, H., 1955. New Zealand Quaternary tectonics. Geol. Rundschau, 43, 248–257.

Dr. Brian H. Mason

The American Museum of Natural History,
New York 24, N.Y, U. S.A.