Hornblende Andesite Rocks from Solander Island
[Read before the Wellington Branch, May 11, 1950; received by Editor, May 16, 1950.]
Examination of three rock specimens from Solander Island collected by Dr. R. A. Falla has confirmed that the island is composed of hornblende andesite. The rocks are petrographically described and a chemical analysis by Mr. F. T. Seelye is included. A comparison is made with certain rocks in the New Plymouth area. No data are available to assign a definite date to the volcanism but a late Tertiary age is suggested.
In 1909 R. Speight examined two small rock specimens collected from Solander Island, by Captain Bollons, of the G.S.S. Hinemoa, and identified the rocks as hornblende andesite. The presence of volcanic rocks was unexpected in view of the geological setting of the island and accordingly some doubt has existed as to whether the rocks are present in situ. Due to the isolation of the island and the difficulty of lauding except in the calmest weather, no additional specimens have been collected for many years but in December 1947, Dr. R. A. Falla landed and collected the rock specimens described in this paper.
Location and Topography
Solander Island, with the accompanying rocks and islets, lies to the south-west of New Zealand, about 21½ nautical miles southward from the South Island at Price's Point and about 35 nautical miles westward from the north-west point of Stewart Island (Fig. 1). No soundings near the island are available on the Admiralty chart (No. 2553) but soundings in Foveaux Strait, about 10 miles to the north, indicate depths of 60–70 fathoms, while between the northwest point of Stewart Island and Solander Island depths over 70 fathoms have been recorded. It is probable, therefore, that the island lies in comparatively deep water.
Information obtained from the N.Z. Pilot (1946, p. 355) and from the photographs accompanying this paper, reveals that Solander Island itself is about one mile in width, rises almost perpendicularly from the sea to a height of 1,095 feet (Figs. 2, 4 and 7) and that it is wooded except for its north-eastern extremity (Fig. 6). According to the N.Z. Pilot (loc. cit.), there is also a deep cave on the eastern side of the island and a large arch at the southern extremity. Three rocks lie within about 1/2 mile of the northern side of the island, the easternmost of which is about 25 feet high and the other two about 12 feet high. Southward of the island several large rocks about 100–150 feet high lie within a distance of 1/2 mile. An islet generally known as “Little Solander” lies about 1 1/4 miles westward of Solander Island and attains an elevation of about 400 feet (Figs. 3 and 4).
From the above description it will be clear that Solander Island does not now present the appearance of a volcanic cone. Its present topography is due to erosion and probably to depression or elevation resulting from fluctuations of sea-level. It is not certain, however, whether the island represents a remnant of a volcanic cone or a cone sheet.
The geological setting of the island is illustrated in Fig. 1. The Fiordland crystalline complex consists mainly of gneisses and associated granite, diorites and norites (Turner, 1938a, 1938b, 1939), with granodiorite the predominant plutonic rock in the area around Preservation Inlet investigated by W. N. Benson and his co-workers (Benson, 1933; Benson and Bartrum, 1935). These investigators make no reference to the presence of loose blocks of hornblende andesite in Cuttle Cove, Preservation Inlet as reported by P. Marshall (Speight, 1909a, p. 52). Near Riverton and Orepuki, diorite, gabbro, norite and granite are intruded into the Late Palaeozoic (Permian?) Maitai–Te Anau formation of argillites, greywackes, breccias and intercalated basic lavas including some displaying pillow structure (Park, 1921; Benson and Holloway, 1940, p. 11). At Bluff the injection of noritic rocks into tuffaceous rocks of the Te Anau formation has been described by H. Service (1937). Little information is available as to the character of the rocks on the northern part of Stewart Island but rocks comparable with those at Bluff are known to occur (Hector, 1870; McKay, 1889; Skey and McKay, 1889) and granite migmatites are common. The southern part of Stewart Island consists almost entirely of biotite granite with roof pendants of high grade metamorphic rocks (Williams, 1934). The Snares Islands, 62 miles south-west of Stewart Island, are composed of muscovite granite (Marshall, 1909, pp. 703–704).
It will thus be seen that there are no known Tertiary to Recent volcanic rocks in the neighbourhood of Solander Island. The closest areas are in Eastern Otago, near Dunedin (Benson. 1941, 42–43) and in the Auckland Islands which lie almost 190 miles north by west from the south cape of Stewart Island (Speight, 1909b).
Three slightly weathered specimens (8602–4*) were collected by Dr. R. A. Falla from rock outcrops on the eastern side of Solander Island. One sample (8603) is from near the shore line and the other two at heights of about 300 feet (8603) and 600 feet (8604) respectively, along the ridge to the summit. All the specimens are hornblende andesite with specific gravity ranging from 2·41 (∓0·01) in the two more massive specimens (8603–4) to 2·05 in the highly vesicular sample (8602). Macroscopically the rocks are light grey in colour with conspicuous large tabular crystals of plagioclase up to 1.5 cm. in length, and less abundant smaller crystals of hornblende and biotite set in a fine grained light grey matrix. In thin section the porphyritic structure is obvious, phenocrystic plagioclase, hornblende
[Footnote] * Numbers refer to specimens and thin sections in the rock and mineral collections of the New Zealand Geological Survey.
and biotite being set in a hyalopilitic ground mass composed of approximately equal amounts of colourless to light brown glass and microlites, or small crystals of plagioclase, hornblende and rarer biotite. The refractive index of the glass in the ground mass is 1·510 (∓0.001) and this indicates a composition of 67–68 per cent. of SiO2 according to the curves of W. O. George (1924).
The dominant phenocrysts are idiomorphic to subidiomorphic tabulae of plagioclase, occasionally 1·5 cm. in length, but usually less than 0·75 cm. In the ground mass the plagioclase occurs as microlites or narrow laths less than 0·25 cm. in length. Zoning of the phenocrystic plagioclase is common. The composition of the plagioclase was determined using the standard universal stage methods as described by W. Nitikin (1933) and F. J. Turner (1947). In all cases the plagioclase is andesine, varying in the non-zoned crystals from An39 to An44. The zoned crystals display rhythmic oscillatory zoning similar to that shown in the andesites of the San Juan area (Larsen and Irving, 1938, p. 229). In most crystals the variation in composition is small varying from An44–An40 but greater variations are occasionally encountered, e.g., one phenocryst varied from An33 to An48. In the plagioclase crystals examined, twinning on the albite and carlsbad laws dominated with less abundant pericline and rare mane-bach-ala types. Although some of the phenocrystic plagioclases are practically free of inclusions, the majority contain abundant globular or irregularly shaped inclusions of glass, portions of ground mass or, rarely, hornblende. These inclusions are sometimes haphazardly arranged, but more commonly are zonal, in which case a narrow marginal zone is entirely free of inclusions.
The most abundant mafic constituent is amphibole which occurs as prismatic crystals up to 7 mm. in length in the phenocrysts to less than 0·05 mm. in the small prisms, irregular fragments and flakes in the ground mass. The amphibole is a strongly pleochroic type of hornblende, the dominant variation being from light yellow (x) to dark brown or reddish brown (z), but a greenish brown type is present in one specimen (8602). The optical properties of a hornblende prism in specimen 8604 were determined as follows:—
|Z||=||dark rusty brown|
|Absorption Z > Y > X|
These properties are comparable with the lamprobolite type of A. Rogers (1940), equivalent to “oxyhornblende” or “basaltic hornblende” of A. N. Winchell (1933, p. 252). All the lamprobolite type possess a low extinction angle (Z ∧ c < 12°), a contrast with the greenish brown hornblende in 8602 which has the properties of normal hornblende. Very little absorption of the hornblende has occurred.
Biotite, though less abundant than hornblende, nevertheless is an important constituent in the specimens examined. It occurs in flakes up to 4 mm. in length and commonly these are curved or contorted. The biotite is strongly pleochroic, the pleochroic scheme being—Z—pale yellow, X—deep brown.
Accessory constituents are apatite, iron ore and zircon. Apatite is invariably present as colourless idiomorphic hexagonal or lath shaped prisms less than 1 mm. in length. Iron ore (titaniferous magnetite?) occurs both as irregularly shaped grains 0·4 mm. in width and as more abundant scattered granules less than 0·05 mm. in length in the ground mass. A zircon crystal was identified in one section (8603A).
Through the courtesy of the Geology Department of Canterbury University College, the writer has been able to compare the rocks collected by Dr. Falla with the thin sections described by R. Speight.* In both cases the hornblende andesites are identical in general composition and texture. Biotite is less abundant in Speight's sections and a reddish tinge is more evident in some of the amphibole (lamprobolite).
A chemical analysis by Mr. F. T. Seelye of the hornblende andesite from Solander Island is given in Table 1 (Anal. A). When this is compared with Daly's average hornblende andesite (Table 1, Anal. B), a close similarity is revealed. The analysis indicates that the andesite is a silicic variety and approaches a dacite in composition. No modal quartz or cristobalite, however, were detected in the thin section.
The hornblende andesites from Solander Island appear to have no known counterparts in the South Island. In the North Island, the writer believes the closest comparison can be made with the hornblende andesites and dacites from Paritutu and the Sugar Loaf Islands as described by C. O. Hutton (1944). Macroscopically and microscopically the rocks are closely similar in composition and texture. There is also marked agreement when the chemical analyses are compared (Table I, Anal. A, C and D). It may be noted here that the dacite analysed (9341), contains no modal quartz or cristobalite (cf. Hutton, op. cit., p. 148). No specimen or thin section is now available of the other rock analysed from Paritutu. The New Plymouth rocks differ from those on Solander Island in the following:—
Hydrothermal solutions have been active and have resulted in the deposition of colourless pools of cristobalite in most of the rocks.
Resorption of amphibole is important in the Paritutu group and has led to the formation of diopsidic augite. Resorption is less evident in rocks from Ngataierua Point.
[Footnote] * Fourteen thin sections are in this collection, thirteen of which were prepared by Vogt and Hochgesang. Of the latter, twelve (pairs numbered 1–3, 5–7) are hornblende andesites and presumably are made from the two specimens collected by Captain Bollons. Section No. 4 is a muscovite granite, but no reference to this section is made by Speight in his paper. No details are now available as to its origin except that it is labelled “granite, Solander Island”.
The rocks appear to be more massive as reflected in their higher specific gravity (2·65 compared with 2·41).
The New Plymouth rocks are believed to be remnants of former steeply dipping cone sheets injected into Upper Tertiary (early Pliocene or latest Miocene?) sediments. There is, of course, no evidence to indicate that Solander Island represents a remnant of a cone sheet, but it is of interest to note the presence, on the mainland at Te Wae Wae Bay and Preservation Inlet (Fig. 1), of Tertiary sediments including some of Upper Miocene or Lower Pliocene age (Finlay and Marwick, 1948, p. 37).
No data are available to assign a definite age to the hornblende andesite on Solander Island. If the rocks are comparable with those at New Plymouth a Late Tertiary age is probable. A comparatively recent date for the volcanism is also suggested by the unaltered character of the rocks.
Benson, W. N., 1933. The Geology of the Region about Preservation and Chalky Inlets, South-west Fiordland, New Zealand—Pt. 1. Trans. N.Z. Inst., vol. 63, pp. 393–432.
——, W. N., 1942–43. The Basic Igneous Rocks of Eastern Otago and their Tectonic Environment. Trans. Roy. Soc. N.Z., vol. 72, pp. 85–118, vol. 73, pp. 116–38.
——, W. N., and Bartrum, J. A., 1935. The Geology of the Region about Preservation and Chalky Inlets, South-west Fiordland, New Zealand—Pt. 3. Ibid., vol. 65, pp. 108–52.
——, W. N., and Holloway, J. T., 1940. Notes on the Geography and Rocks of the Ranges between the Pyke and Matukituki River, North-west Otago. Ibid., vol. 70, pp. 1–24.
Finlay, H. J., and Marwick, J., 1948. “Tertiary,” in Outline of the Geology of New Zealand. H. H. Tombs, Wellington.
George, W. O., 1924. The Relation of the Physical Properties of Natural Glasses to their Chemical Composition. Jour. Geol., vol. 32, pp. 353–72.
Hector, J., 1870. Notes on the Geology of the Outlying Islands of New Zealand with extracts from official reports. Trans. N.Z. Inst., vol. 2, pp. 176–86.
Hutton, C. O., 1944. Some Igneous Rocks from the New Plymouth Area. Trans. Roy. Soc. N.Z., vol. 74, pp. 125–53.
Larsen, E. S., Irving, J., Gonyer, F. A., and Larsen, E. S., 3rd, 1938. Petrologic Results of a Study of the Minerals from the Tertiary Volcanic Rocks of the San Juan Region, Colorado. Amer. Mineral., vol. 23, pp. 227–57.
McKay, A., 1889. On the Geology of Stewart Island and the Tin Deposits of Port Pegasus District. Rept. Geol. Explorations for 1888–89, no. 20, pp. 75–85.
Marshall, P., 1909. The Geology of the Sub-Antarctic Islands of New Zealand in Sub-Antarctic Islands of New Zealand, vol. 2, pp. 680–704. Philosophical Institute of Canterbury.
New Zealand Pilot, 1946. Admiralty, London, 11th ed.
Nikitin, W., 1936. Die Fedorow Methode. Borntraeger, Berlin.
Park, J., 1921. The Geology and Mineral Resources of Western Southland. Bull. N.Z. Geol. Surv., no. 23.
Rogers, A. F., 1940. Lamprobolite, a New Name for Basaltic Hornblende. Amer. Mineral., vol. 25, pp. 826–28.
Service, H., 1937. An Intrusion of Norite and its Accompanying Contact Metamorphism at Bluff, New Zealand. Trans. Roy. Soc. N.Z., vol. 67, pp. 185–217.
Skey, W., and McKay, A., 1889. On certain Rare Minerals associated with the Tin Ore of Stewart Island. Trans. N.Z. Inst., vol. 22, pp. 415–22.
Speight, R., 1909a. On a Hornblende-andesite from the Solander Islands. Ibid., vol. 41, pp. 52–53.
——, R., 1909b. The Physiography and Geology of the Auckland, Bounty and Antipodes Islands. Sub-Antarctic Island of New Zealand, vol. 2, pp. 705–44. Philosophical Institute of Canterbury.
Turner, F. J., 1938a. Progressive Regional Metamorphism in Southern New Zealand. Geol. Mag., vol. lxxv, pp. 160–74.
——, F. J., 1938b. The Metamorphic and Plutonic Rocks of Lake Manapouri, Fiordland, New Zealand. Trans. Roy. Soc. N.Z., vol. 68, pp. 122–40.
——, F. J., 1939. Hornblende Gneisses, Marbles and Associated Rocks from Doubtful Sound, Fiordland, New Zealand. Ibid., vol. 68, pp. 570–98.
——, F. J., 1947. Determination of Plagioclase with the Four-Axis Universal Stage. Amer. Mineral., vol. 32, pp. 389–410.
Williams, G. J., 1934. A Granite-Schist Contact in Stewart Island, New Zealand. Q.J.G.S., vol. xc, pp. 322–53.
Winchell, A. N., 1933. Elements of Optical Mineralogy, Pt. 2. Wiley, New York. Third edition.
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
Symbol and Classification
A = (1) 4. 4″ ″3. 4. Tonalose
B = (1) 4. 3. 4. Tonalose
C = 1. 4″ 3. 4. Yellowstonose
D = 1. 4. (5) ″3. 4. Yellowstonose.
A. Hornblende andesite (P.8603), Solander Island. Analyst, F. T. Seelye. G. = 2.42.
B. Average of 24 hornblende andesites, Daly 1933, p. 16, No. 52.
C. Dacite (P.9341) 2 chains south of Paritutu, Paritutu S.D. Analyst, F. T. Seelye. G. = 2.65. (Hutton, 1944, p. 150, analysis B.)
D. Dacite, seaward face of Paritutu. Analyst, J. S. Maclaurin. (Bull. N.Z. Geol. Surv., No. 14, p. 23, analysis 11.)