Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 40, 1907
This text is also available in PDF
(602 KB) Opens in new window
– 176 –

Art. XIV.—On a Soda Amphibole Trachyte from Cass's Peak, Banks Peninsula.

[Read before the Canterbury Philosophical Institute, 6th November, 1907.]

The oldest rocks found on Banks Peninsula consist of slates, cherts, and greywackes of uncertain age; but the last show a marked lithological resemblance to Lower Mesozoic greywackes that occur at the Malvern Hills. The only exposure of these rocks on Banks Peninsula is near Gebbie's Pass, where they occupy a considerable portion of the main ridge, and extend down on both sides of it, but especially towards the head of Lyttelton Harbour. Here they form a large part of the solid floor of the valley in which Teddington lies. Over them lie solid flows of rhyolite and beds of agglomerate penetrated by dykes of rhyolite and pitchstone. The age of these beds is also uncertain, but they resemble very closely in lithological character the garnet-bearing rhyolites of Mount Somers, Rakaia Gorge, and the Malvern Hills, which are certainly of Cretaceous age, as rhyolite pebbles are found in conglomerates forming the lower members of the coal-bearing series, which, as well as the rhyolites, overlie unconformably Jurassic sedimentaries. At Mount Somers, too, rhyolite tuffs, according to S. H. Cox, are interstratified with coal-bearing beds. It is therefore highly likely that the Gebbie's Pass rhyolites are of Cretaceous age.

After a considerable lapse of time, during which the rhyolites were heavily eroded, the main mass of Banks Peninsula was formed, consisting chiefly of andesites of basic type and basalts. These were poured out as subaerial lava-flows, and thrown out as scoria and ashes from two craters which now form Lyttelton and Akaroa Harbours. Onawe Peninsula probably marks the centre of the latter volcano, as the extremity of the peninsula is composed of a syenite, and this is the only occurrence of a plutonic rock in the locality. The remaining part of this small peninsula near the narrow isthmus is principally formed of intercrossing dykes; it thus shows the structure which characterizes the neighbourhood of the pipe of an old volcano. Sir Julius von Haast suggests that a third centre of eruption, belonging to this period, occurs in the valley of Little River. He is not very definite about it, and says that the remains of the lavas that were poured out from it are not very extensive. I believe, however, that he modified his views somewhat at a

– 177 –

later date, and considered that the peninsula was built principally from the centres of Lyttelton and Akaroa. I cannot speak definitely from personal observation as regards this point, but from what I have seen I am inclined to think that it is unlikely that a crater occupied the valley of Little River, but that the lavas occurring there were poured out from both Lyttelton and Akaroa, and that the form of the valley can be well explained as the result of prolonged stream erosion. When all the lavas are basalts and basic andesites, and good sections showing their relations are practically absent, an accurate estimate is extremely difficult to make. Good sections showing contacts of the andesites with the earlier rhyolites are also rare, owing to the completeness with which the soil covers over everything. However, a section near the end of the spur which divides Gebbie's from McQueen's Valley affords convincing proof of their relative age. Here the actual contact is seen, and andesites undoubtedly overlie denuded rhyolites.

The andesites always contain augite, with a small amount of olivine generally added, and thus show close relations to the basalts; but the silica percentage of some varieties is too high (about 56) for them to be classified as such. There are gradations, however, from the less basic to the thoroughly basic types, which finally pass into undoubted basalts. It is highly likely that the Akaroa lavas are of a slightly later date than those from the Lyttelton volcano. They are generally of a more basic character, true basalts forming a large proportion of effusive mass. This evidence is perhaps very unreliable, but it is supported by the fact that the crater-ring of Akaroa is in a much more perfect condition than the Lyttelton ring, denudation not having exerted such a marked influence over its original form. However, this may be accounted for by the more resistent character of the rocks constituting it. In the subsequent section I have represented the Akaroa lavas as being slightly younger than the Lyttelton ones.

The andesitic eruptions from these two centres were succeeded by an outpouring of basalts and andesites from Mount Herbert, and probably from Mount Sinclair. The latter mountain forms the geographical centre of Banks Peninsula, being situated at the junction of the Port Levy, Pigeon Bay, and Little River Valleys, with outlying parts extending nearly to the edge of the crater-ring of Akaroa.

Sir Julius von Haast mentions a fourth centre of eruption at Quail Island, within Lyttelton Harbour; but this may be contemporaneous with that at Mount Herbert, and it is even possible that the Quail Island basalts came from Mount Herbert, and that the connecting rocks have been removed by denudation.

– 178 –

On examining a map of Banks Peninsula it will be seen that the centres of volcanic activity lie approximately on a line running east-south-east and west-north-west. It seems, therefore, a reasonable inference that the eruptions took place at different points of a fissure or line of weakness in the earth's crust running in that direction; that eruptions broke out first at the Lyttelton and of the fissure, and that afterwards the centre of maximum disturbance moved eastward to Akaroa, and then back to Mount Sinclair and Mount Herbert, and possibly to Quail Island.

As the Lyttelton volcano has thrown out rocks belonging to three different periods, and perhaps to four, I think it would be convenient to refer to the rhyolites as the Gebbie's Pass series, to the olivine-andesites as the Mount Pleasant series (named from one of the chief peaks on the northern side of the harbour immediately behind the Town of Lyttelton), and to refer to the lavas which come from Mount Herbert as the Mount Herbert series. All these are quite distinct in age: the Gebbie's Pass series being almost certainly Cretaceous, the Mount Pleasant series being early Tertiary, and the Mount Herbert series middle Tertiary; but the last two are extremely uncertain as regards their age, and may be much more recent. Although stream erosion has exerted a marked influence in forming valleys, yet the form of the crater-ring is fairly perfect, especially as regards Akaroa, so that a more recent date may very well be assigned to the two later series.

Picture icon

Diagrammatic Section of Banks Peninsula, from Tai Tapu to the Coast near East Head.


Slates and greywackes; Lower Mesozoic(?). 2. Rhyolites, Gebbie's Pass series; Cretaceous. 3. Augite-andesites and basalts, Mount Pleasant series; early Tertiary(?), perhaps later. 4. Basalts and andesites, Mount Herbert series; mid Tertiary(?), perhaps later. 5. Syenite, Onawe Peninsula; early Tertiary(?), perhaps later.

Note.—The line of this section is not straight, but altered in direction to show the relative position of the rocks of different age. A good deal of the section is problematical, particularly that portion between Mount Herbert and Akaroa Harbour.

– 179 –

The foregoing section shows the relativ position of the different outpourings of volcanic rocks. It will be noted that in it I have classified the lavas from Akaroa as belonging to the Mount Pleasant series; but this classification is merely tentative, and for reasons just stated they should probably be marked as belonging to a later date.

The Mount Pleasant series is penetrated by a remarkable series of dykes, well described by Sir Julius von Haast, who pointed out in the case of Lyttelton that they are oriented in a somewhat striking manner. They all, with few exceptions, converge on a point at the back of Quail Island, no matter in what parts of the crater-ring they are found. The dykes of the Gebbie's Pass series are not so arranged, while there are none visible in the Mount Herbert series. Some of these dykes have been previously described by Hutton*, Ulrich,† Marshall, Filhol,§ Kolenko, and the author. They consist, as far as is known at present, of dolerites, basalts, hornblende, and augite andesites, some containing olivine, trachytoid phonolites, and trachytes, the last being probably the most numerous, although basaltic dykes are also common. Some of the trachytes contain hornblende and others augite, and they are in general of whitish, pale-grey, and sometimes of a greenish colour, and very vesicular. Chemical analysis shows that some contain a high percentage of soda. It is to this class that the rock to be described belongs.

The Soda Amphibole Trachyte.

Its Occurrence.

The rock is found as a massive dyke on the northern side of Cass's Park, one of the highest points on the west side of the old crater-ring of Lyttelton. The dyke can be traced fully half a mile from near the crest of the ridge, through the Kennedy's Bush reserve, and down one of the valleys towards Lansdowne. At times it is fully 60ft. wide, but it thins out towards the top of the ridge, and also when followed down the valley. At one spot it was quarried as a building-stone, and several buildings in Christchurch were built of it, notably the present Tourist

[Footnote] * “Eruptive Rocks of New Zealand,” Trans. Roy. Soc. N.S.W., 1889.

[Footnote] † “Transactions of the Australasian Association for the Advancement of Science,” vol. 4, 1891.

[Footnote] ‡ “On a Tridymite Trachyte of Lyttelton, Trans. N.Z. Inst., vol. xxvi (1893).

[Footnote] § “Mission de l'Ile Campbell,” Paris, 1883.

[Footnote] ∥ “New Zealand Journal of Science,” vol. ii.

[Footnote] ¶ “On a Doleritic Dyke at Dyer's Pass,” Trans. N.Z. Inst., vol. xxvi (1893).

– 180 –

Office, and it has been used in the construction of others. Blocks of this stone occur in the archway over the entrance to Canterbury College. As a building-stone it is very easy to work, and stands the weather extremely well; but its appearance is somewhat spoiled by the presence of fragments of the country rock, which are irregularly distributed through it.

Macroscopic Appearance.

The rock is of a light greenish-grey colour, with phenocrysts of feldspar visible in a rather porous groundmass. A number of black specks also are to be seen, and these are either the soda amphibole or aggregates of iron-ore derived from it. No other porphyritic crystals are visible. The included andesitic fragments are of all sizes, up to 10cm. in length.

Specific Gravity.

The specific gravity determined immediately after immersion in water was 2.35; on leaving the rock to soak for twentyfour hours it was 2.48; and determined by a specific-gravity bottle it was 2.57. These figures afford some idea of the vesicular character of the rock.

Chemical Analysis.

A chemical analysis of the rock was made in the chemical laboratory, Canterbury College, by several students, under the direction of Dr. W. P. Evans, who has kindly furnished me with the following result:—

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

Per Cent.
SiO2 70.04
Al2O3 15.40
Fe2O3 4.65
CaO Slight trace only.
MgO 0.55
Na2O 4.35
K2O 4.65
H2O 0.57

The following points with regard to this result are specially noticeable: The high percentage of SiO2 (70.04), the low percentages of CaO and MgO, and the moderately high percentage of alkalies for a rock of its character. These peculiarities are explained by the microscopical examination, and will be dealt with more fully subsequently.

– 181 –

Microscopical Examination.

A microscopical examination of the rock shows it to belong to the trachytes, but with characteristics connecting it with the rhyolites. The phenocrysts are apparently all sanidine, clear, fissured, some in Carlsbad twins, their greatest length being about 4mm. Anorthoclase was specially looked for, in order to explain the fairly high percentage of soda, but no undoubted crystals were detected, although some of the crystals suggested the microscopic twinning of anorthoclase very faintly. In the absence of decided characters I have classified them all as sanidine. No phenoerysts of plagioclase were observed.

The only other porphyritic mineral is the soda amphibole. This mineral occurs in very irregular-shaped individuals of small size, re-entrant angles being extremely common. Pleochroism is very strong, the maximum absorption occuring when the cleavage is parallel to the short diagonal of the nicol. The colours are a deep-blue, greenish-blue, and brownish-yellow. The mineral is somewhat opaque, and only translucent in thin sections. The angle of extinction is therefore somewhat difficult to determine, but it ranges up to about 10°, measured from the cleavage traces in sections where they are parallel. These characters show the mineral to be, in all probability, the soda amphibole arfvedsonite, or a closely related variety. The rock also contains aggregates of iron-ores, which are apparently derived from this amphibole.


The groundmass is holocrystalline, but the size of the individual crystals varies considerably in different parts of the dyke. It is composed chiefly of rectangular and short lathshaped crystals of sanidine frequently twinned, with interstitial matter of smaller microlites of sanidine, and sparingly plagioclase; this last is almost certainly albite. The feldspars exhibit at times a rough fluxion structure, especially when the groundmass is somewhat coarse in texture. Small grains of quartz are commonly seen in the groundmass forming part of the interstitial matter between the larger individuals of feldspar. The most brilliant polarisation colours and the index of refraction (Becke's test) show clearly that the mineral is not tridymite—which might have been expected in a trachyte. The high percentage of SiO2 in the chemical analysis shows that a considerable quantity of free silica must be present, and this must occur in the groundmass, as the clear phenocrysts are apparently all sanidine. The soda amphibole also occurs very plentifully in the groundmass, in the form of irregular flakes. This exhibits the characteristic pleochroism of the larger individuals. In

– 182 –

many cases it is moulded on the larger crystals of sanidine in the base, and has evidently separated out at a late period in the consolidation. Small irregular flakes of a greenish augite also occur, but it is very difficult to differentiate them from the blue amphibole, their colour and faint pleochroism being the special criteria for discrimination.

This description of the rock shows that it belongs to the phonolitic variety of trachyte, using that term in its general acceptation—viz., a trachyte which contains sanidine (and anorthoclase), with alkali iron pyroxenes or alkali iron amphiboles.

The microscopical examination thus explains the peculiarities in the chemical analysis. The fairly large percentage of soda is due to the presence in large quantities of the soda amphibole and the green augite. The practical absence of lime shows the absence of all plagioclase feldspars except albite, and, taken in conjunction with the poorness in magnesia and the absence of any other mineral explaining the percentage of iron-oxide, it shows that amphibole is most probably an almost pure sodairon variety. This may contain a small proportion of magnesia, although the presence of this oxide may be due to the fragments of augite in the groundmass. The high percentage of silica, and its presence in the free state in the groundmass, though rare in trachytes, seems undoubtedly to occur in those of orthophyric character (vide Rosenbusch's “Elemente der Gesteinslehre”). The tridymite trachyte dyke from the Lyttelton-Summer Road described by Marshall also shows a high percentage of free silica; but he came to the conclusion that this was of secondary origin, whereas it appears to me to be a primary constituent in the groundmass of this rock. I include the analysis of the tridymite trachyte made by Marshall for the purposes of comparison, as well as all other analyses which I have come across of the trachytes and related rocks belonging to the Mount Pleasant series.

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

Analyses of Trachyte and Allied Dykes of the Mount Pleasant Series.
A. B. C. D. E. F. G. H.
SiO2 58.93 59.87 61.99 61.38 60.69 52.18 71.09 70.04
Al2O3 23.95 21.22 13.08 20.60 17.75 20.00 15.45 15.40
Fe2O3 5.43 4.42 8.65 2.57 3.83 5.00 1.50 4.65
FeO 1.16 0.34
MnO 0.14 4.42 1.19 1.21
CaO 1.75 2.58 2.21 2.18 1.20 4.92 3.25 Slight trace
MgO 0.96 0.91 Trace 0.40 1.43 1.03 0.89 0.55
K2O 4.32 4.06 1.61 Traces 2.30 2.35 4.35
Na2O 5.61 5.34 4.22 9.70 13.10 14.57 4.81 4.65
H2O 1.36 3.82 1.98 0.79 0.07 0.57
Total 102.31 99.70 100.00 100.00 100.00 100.00 99.75 100.21
– 183 –

A. Trachytoid phonolite, Lyttelton-Sumner Road; analysed by P. Marshall; Trans. N.Z. Inst., vol. xxvi (1893).

B. Trachytoid phonolite, Heathcote; analysis by T. Butement; quoted by H. F. Ulrich, Trans. Aus. Assoc. Adv. Sci., vol. iii (1891).

C. Vesicular trachyte from agglomerate bed; analysis made in laboratory of the Geological Survey; quoted in Haast's “Geology of Canterbury and Westland.”

D. Dyke (side) cut by tunnel, No. 29B, same dyke; analysis made in laboratory of the Geological Survey; quoted in Haast's “Geology of Canterbury and Westland.”

E. Dyke (centre) cut by tunnel, No. 29A, same dyke; analysis made in laboratory of the Geological Survey; quoted in Haast's “Geology of Canterbury and Westland.”

F. Dyke (centre) cut by tunnel; analysis made in Paris by Dr. H. Filhol; “Mission de l'Ile Campbell.”

G. Tridymite-trachyte, Lyttelton-Sumner Road; analysis by P. Marshall; Trans. N.Z. Inst., vol. xxvi (1893).

H. Soda amphibole trachyte, Cass's Peak; analysis made in chemical laboratory, Canterbury College; inserted again for convenience of comparison.

This list includes nearly all the published analyses of trachytic and allied rocks of this series. There seems to be one or two striking features about some of them. Assuming that they are tolerably correct, those marked D, E, F show an abnormal percentage of soda, and also a very small percentage of potash; also A, B, C show an excess of soda over potash. The high percentage of MnO in C is also remarkable; this apparently explains the presence of frequent thin coatings of a black mineral resembling pyrolusite, which occurs on the fracture surfaces of the rock. Analyses A and B undoubtedly show the characters of a trachytoid phonolite, and C, D, E, and F those of a soda trachyte. These last rocks have anorthoclase as a common phenocryst, but the practical absence of potash in the analysis is rather peculiar. The two analyses G and H afford interesting comparisons. The marked agreement of the silica, alumina, magnesia, and the soda are very noteworthy. The only differences appear to be the greater proportion of iron-oxides and the practical absence of lime in H. These peculiarities are explained by the microscopical analysis. There is a fair proportion of plagioclase (andesine) and a very small amount of iron-bearing mineral in the tridymite-trachyte. In his description of the rock Dr. Marshall noted the percentage of magnesia without being able to account for it. On looking over a section of it I found in the groundmass a considerable quantity

– 184 –

of greenish-blue pleochroic mineral in very minute fragments, which may be either the soda amphibole or the greenish augite. These would account for the small percentage of magnesia that does occur. This greenish mineral with slight pleochroism is found in other rocks occurring as dykes in this series. In some cases it is undoubtedly an augite of a soda-bearing variety; but in other cases where it has the bluish tinge of varying degrees of intensity it is, in all probability, a soda-iron amphibole.

Perhaps the most interesting occurrence of this mineral is in the syenite of Onawe Peninsula, Akaroa. In his description of this rock Captain Hutton says,* “The hornblende goes up to 0.50in. in length; when fresh it is greenish and pleochroic, changing from blue-green to yellow-green, the polarisation colours not brilliant.” On examining this rock further with the advantage of thinner sections I find the masses of iron-oxides which have in most cases replaced the hornblende show not merely a greenish-blue, but a deep-blue colour, and in other cases I noticed small pieces of hornblende exactly resembling the amphibole of Cass's Peak. This, therefore, seems to me a case of the occurrence of an arfvedsonite syenite. Just as in many dykes of the Mount Pleasant series, this rock is very light in colour, and shows a small proportion of iron-bearing mineral.

The fairly wide occurrence of the rocks of the phonolitic trachyte variety so closely connected with the trachytoid phonolites, as well as the occurrence of arfvedsonite syenite at Akaroa, is of special interest when we note the existence at Dunedin of the magnificent series of alkaline rocks discovered by Ulrich, and well described latterly by Marshall. The occurrence of the rocks previously mentioned in the Banks Peninsula area shows distinctly that the distribution of alkali rocks in New Zealand is wider than at first supposed.

[Footnote] * “The Eruptive Rocks of New Zealand,” by Professor F. W. Hutton. Read before the Royal Society of New South Wales, 7th August, 1889.