Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 69, 1940
This text is also available in PDF
(3 MB) Opens in new window
– 283 –

Mineralogical Notes from the University of Otago, N.Z. No. 3—Kaersutite and Other Brown Amphiboles in the Cainozoic Igneous Rocks of the Dunedin District.

[Read before the Otago Branch, October 11, 1938; received by Editor, June 4, 1939; issued separately, December, 1939]

Kaersutite, the richly titaniferous end-member of the group of basaltic hornblendes or syntagmatites (Brögger, 1894), was first described by Lorenzen (1884-1886) from Kaersut, Nugsuaks Peninsula, West Greenland. It has since been found in several other regions, and recently Tomita (1934) expressed the view that the occurrence of basaltic hornblende and kaersutite may be regarded as a characteristic of the petrographic provinces of Cainozoic alkaline effusive rocks throughout the world. In his first announcement of the presence of nepheline-bearing rocks in the Dunedin District, Ulrich (1891) gave a careful description of the occurrence of brown amphibole in tinguaite porphyry (later termed “ulrichite”) at Portobello, and at Pine Hill, in a phonolite near the top of Mount Cargill, and in an unspecified rock, probably the kaiwekite at Long Beaeh. Marshall (1904, 1906) described other occurrences and in particular made a partial analysis of the kaersutite from the Leith Valley described below, but as the amount of alkalies and titanium contained therein was not determined, he was led to conclude that the mineral was a normal pargasite. Later, however (Marshall, 1914) he described the amphibole in the kaiwekite and certain other rocks as “probably barkevikite.” It is now possible to give further details of the development of brown amphiboles during the course of igneous activity in the Dunedin region, and two chemical analyses of kaersutitic examples. For these the writer must express his great indebtedness to Mr. F. T. Seelye, F.I.C., of the Dominion Laboratory, by whom the analyses were made, through the courtesy of the Directors of the Geological Survey and of the Dominion Laboratory. He is also indebted to Dr. F. J. Turner for the measurements of optic axial angles, and to Dr. G. D. Osborne, of Sydney University, for the determination of the refractive indices of the Leith Valley kaersutite. Many of the slides examined in this work (those in which an initial letter precedes the slide-number) were made by Dr. Marshall during his studies of the Dunedin District prior to 1914, which form the basis of much of the writer's later investigations. Thanks are due to Dr. Marshall also for the loan of additional material for this study and for personally communicated information.

– 284 –

It will be helpful, before describing the features of these minerals in the Dunedin District, to summarise the mode of occurrence of kaersutite at the type-locality, and the general properties of the brown monoclinic amphiboles. Kaersutite was reported by its discoverer, Steenstrup (1884), to occur in a thin vein or dyke cutting through a horizontal sill of peridotite, later described as picrite by Phalen (1904) and Rosenbusch (1908). The optical properties of kaersutite were determined and communicated to Rosenbusch by Ussing, who also provided the material for the more detailed chemical and optical studies of Washington and Wright (1908). Further field and petro-graphical work was done by Ravn (1911) and Heim (1911)*, by Krueger and Drescher (1928), and by Drescher (1932), whose studies indicate that a very interesting example of crystallisation-differentiation occurs at Kaersut, recalling some features of the Lugar sill in Ayrshire (Tyrrell, 1917). The Kaersut sill is 50 metres thick, and invades nearly horizontal Cretaceous sandstone, which is slightly fritted at the contact. The picrite-peridotite, though not homogeneous, is apparently not affected by gravitational differentiation. It contains sub-horizontal bands enriched in augite, and at least two well-marked augitic streaks inclined at 15° to the horizontal, intersecting the horizontal bands obliquely. These are 40 and 80 centimetres thick. Through 1he middle of the sill runs a horizontal sheet of ophitic dolerite 1.2 metres thick intersecting these oblique streaks. In addition here are numerous segregations of kaersutite-bearing pegmatite comprising a horizontal sheet 35–40 centimetres thick traversing the upper portion of the sill, and giving off below and above steeply inclined irregular veinlets, the latter occasionally rising into the overlying sandstones. There are also irregularly lenticular druse-filling masses of pegmatite, and rarely almost vertical thin veins which are bent against the dolerite sheet, or wedge out in the lower portion of the sill. Krueger's opinion that these veins are younger than the dolerite has been confirmed, but that they actually intersect it seems to be incorrect. Some of the larger veins contain in their middle portion finely granular aplite, either sharply distinct from the kaersutite pegmatite or passing gradually into it. Deuteric alterations have affected all these rocks in varying degree. Table I, compiled from Drescher's detailed descriptions, gives the approximate composition of the several types of rocks. Drescher calls attention to the concentration of kaersutite in the pegmatite, the material of which was derived by lateral secretion from the picrite with rising gas-pressure and decreasing temperature; strong resorption of the crystals occurred when the temperature fell below 350° C., and zeolites commenced to form.

Lorenzen's (1886) statement of the composition of the kaersutite was corrected by Washington's analysis of some of the original material, and very similar figures were obtained by Gossner (1929).

[Footnote] * Heim's figure illustrating the general occurrence is copied in Suess. La Face de la Terre, Vol. III-4, p. 1525.

– 285 –

Kunitz (1930) suggested that the rather different figures obtained by him might have resulted from an admixture of pyroxene in the powdered sample he had received, but the similarity of his results to those obtained by Schafer (in Drescher, 1932) from carefully purified material also provided by Drescher supports the latter's suggestion that there may be a real variation in the composition of the kaersutite even within the type locality. Other analyses of titaniferous amphiboles which may be classed as kaersutite (i.e. containing more than 5–6% of TiO2) have been made by Kawano, Teshima, Ushijima and Washington, and are given in Table II for comparison with the two analyses by Mr. F. T. Seelye of the Dunedin kaersutites. To provide a contrast, analyses of normal basaltic hornblende and barkevikite have also been added.

Table I.—Composition of the Kaebsut sill Complex.
Picrite-Peridotite Dolerite Pegmatite Aplite
Proportion of Total Mass. Normal 65.85% Augitic 30% 2.4% 1.65% 0.05%
SiO2 41.11% 38.48% 44.10% 48.50% 59.77%
TiO2 0.61 1.20 2.76 3.05 0.81
Na2O 0.72 0.74 2.69 4.39 5.68
K2O 0.12 0.13 1.17 2.15 6.33
Olivine and Serpentine 48 52
Angite 10 (a) 20 18 (b) 35 (c) 6
Kaersutite 4 5 44
Labradorite 6 3 33 10 81
Anorthoclase 4
Biotite 1
Iron-Ores 1 3 4 (d) 2
Apatite 2 1 1 1
Secondary 28 (e) 17 (f) 44 (g) 8 (h) 9 (h)

The chemical formulae assigned to kaersutite by different authors are in part as follows: Lorenzen (fide Doelter, 1914) stated the composition simply as 5R (Si, Ti)O3 Al2O3. Washington (1908) applied the formula of Penfield and Stanley (1907), which seems, however, less satisfactory than a type of ring-formula, which, while maintaining a general metasilicate composition, allows for the presence of a reibeckitic component of the type (R'R”)R”' Si4O12. Kunitz (1930) proposed the formula (SiO3)6(AlO2)2 H2 Ca2-Mg2 Mg2 with replacements of (CaAl) by (NaSi). This does not, however, agree with the results of X-ray structural analysis according to Berman and Larsen (1931), who proposed instead a modification of Warren's (1930) formula, of which a further modification was introduced by Machatschki (1932). Based on all three of these is the formula found by Kawano (1934) to correspond reasonably well with the structure of basaltic hornblendes:—

[Footnote] (a) Pale violet.

[Footnote] (b) Titaniferous.

[Footnote] (c) Titanaugite with some greenish diopside bordered by darker aegirine and kaersutite in varying proportions.

[Footnote] (d) Ilmenite.

[Footnote] (e) Chiefly chlorite with carbonates of Ca and Mg, biotite, analcite, chalcedony, aluminous minerals and limonite.

[Footnote] (f) Chiefly chlorite; also biotite, analcite, calcite and limonite.

[Footnote] (g) Chiefly nontronite and biotite, with chlorite, calcite and natrolite.

[Footnote] (h) Chiefly natrolite, with pectolite, mesolite and calcite.

– 286 –
Table II.—Cchemical Compositions of Brown Monoclinic Amphiboles.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
SiO2 41.38 39.52 39.50 39.52 39.70 40.85 39.78 38.30 39.20 38.79 39.01 36.12 42.48 41.12 40.17
TiO2 6.75 10.31 10.33 5.74 6.53 8.47 7.00 6.06 6.53 6.66 6.05 4.82 2.90 0.45 3.78
Al2O3 14.41 11.22 11.12 11.89 11.83 9.89 14.13 12.87 13.87 13.98 13.60 12.46 8.58 11.02 15.09
Fe2O3 abs. 1.22 0.06 1.50 4.92 8.85 4.61 7.98 4.08 3.15 5.25 9.60 6.81 6.54 5.49
FeO 11.28 8.81 9.44 8.08 5.98 3.9 7.31 6.96 7.33 8.44 7.42 10.43 15.62 17.73 5.99
MgO 13.51 13.31 12.90 14.19 14.72 12.47 11.01 11.79 11.98 11.88 11.73 9.09 2.78 6.14 12.48
CaO 12.97 10.93 10.91 13.72 12.52 12.16 10.75 10.47 12.37 12.20 12.05 12.01 13.45 10.42 11.21
Na2O n.d. 2.95 3.82 3.34 1.84 2.01 2.57 3.11 1.99 2.42 2.51 2.58 6.32 3.65 2.27
K2O n.d. 1.07 1.43 1.09 1.29 0.63 1.58 1.30 1.45 0.92 1.11 1.41 0.60 0.94 1.55
H2O- n.d. n.d. n.d. n.d. n.d. n.d. 0.22 0.62 0.27 0.40 0.19 n.d. n.d. n.d. 0.25
H2O+ n.d. 0.59 0.59 0.46 0.85 0.19 0.37 1.10 0.87 0.93 0.98 1.02 0.25 1.45 2.10
Co2 abs. abs.
P2O5 abs. abs. 0.03 0.03
F 0.28 0.05 abs. abs.
MnO 0.06 0.10 0.19 0.12 0.12 0.12 0.11 0.18 0.14 0.28 0.39 1.03 0.09
NiO 0.10 0.03 0.03
BaO 0.03 0.03
Etc. 0.26 0.02 tr.
Total 100.56 99.99 100.20 99.72 100.18 99.98 99.45 100.73 100.03 100.06 100.13 99.82 100.18 100.49 100.47
sp.G. 3.04 3.137 3.336 3.261 3.32 3.272 3.330 3.298 3.415 3.178

Kaersutite, Greenland, J. Lorenzen (1884) Anal (TiO2 low, Fe2O3) and alkalies are not reported. The stated presence of 0.26% of SnO2 was not confirmed by Washington (1908).


Kaersutite, Greenland, Washington (1908) Anal, (Corrected by excluding a small amount of Ca3P2O8).


Kaersutite, Greenland, Gossner (1929) Anal.


Kaersutite, Greenland, Schafer in Drescher (1932) Anal.


Kaersutite, Greenland, Kunltz (1930) Anal. Note.—Specimens 1–5 from kaersutite pegmatite.


Kaersutite (Linosite) Linosa. In basalt tuff, Washington (1908) Anal.


Kaersutite, Dogo, Okl Island. In basalt dyke, Teshima (In Tomita, 1934) Anal.


Kaersutite, Yododo, Korea. In andesine basalt. Kawana (1934) Anal.


Kaersutite, Uryon, Korea. In andesine basalt, Ushijima (In Kawano, 1934} Anal.


Kaersutite, Purakanui, Dunedin, N.Z. In kalwkite. F. T. Seelye Anal. (including S=0.02%).


Kaersutite, Leith Valley Quarry, Dunedin, N.Z. In trachybasalt. F.T. Seelye Anal. (including a trace of S).


Dark brown barkevikite, Fuerteventura, Canary Islands. In essexite. Kunitz (1930) Anal.


Red-brown barkevikite, Lugar, near Mauchline, Ayrshire. In lugarite. Scott (1914) Anal.


Green-brown barkevikite, Stavarnsjö, Norway. In nephelins syenite. Kunitz (1930) Anal.


Basaltic hornblende, Lake Balaton, Hungary. In basaltic tuff. Vendl. (1924) Anal.

[Footnote] * There are minor discrepancies between Kawano's and Tomita's citations of analyses Nos. 8 and 9.

– 287 –

(O, OH, F)2 (Ca, Na, K)2-3 (MgFe″Mn, Fe″',Ti)5((SiAl)4O11)12 Utilising this formula, the above analyses may be stated thus:—

Analysis (O, OH, F) (Ca, Na, K) (Mg, Fe″, Mn, Fe″', Ti) (SiAl)4O11) Note
No. 2 2 2.4 5 4 11 12
4 2 3.3 5 4 11 12
5 2 2.5 5 4 11 12 (a)
7 2 2.7 4.4 4 11 12 (b)
8 2 2.8 5 4 11 12
10 2 2.3 4.7 4 11 12
11 2 2.8 4.8 4 11 12 (c)
12 2 3.0 5 4 11 12
15 2 2.5 5 4 11 12 (d)

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

The usually small amount of Fe2O3 in kaersutite has caused Winchell (1933) to exclude it from the oxyhornblendes with which he groups most of the basaltic hornblendes. The kaersutite of Linosa is abnormal in this respect and may have been exposed to strongly oxidising high-temperature conditions, accounting for its high refractive indices and birefringence (cf. Kozu, Yoshiki and Kani, 1927, Barnes, 1930, Larsen, Irving and Gonyer, 1937). Murgoci's earlier suggestion (1922) was to define basaltic hornblende chemically as containing much (Fe,Al)2O3, with MgO/FeO = 6 and MgO/CaO = 3/2. The linosite differed in that MgO/FeO = 5, and kaersutite in containing more TiO2 and FeO with less (Al,Fe)2O3 than basaltic hornblende.

No means are here available to determine the lattice-dimension of the Dunedin kaersutite. The following figures are cited as being probably comparable therewith. The Anakie hornblende is strongly oxidised (Fe2O3 = 12.02%) and rich in TiO2 (5.40%).

Table IV.—Lattice-Dimensions OF Basaltic Hornblendes.
Mineral a b c Authority
Kaersutite, Greenland 9.85 Å 18.17 Å 5.37 Å Gossner and Speilberger (1929)
Kaersutite, Korea 9.77 17.85 5.32 Kozu and Tanake fide Kawano (1934)
Basaltic Hornblende, Anakie, Victoria 9.88 17.85 5.44 Parsons (1930), Greenwood and Parsons (1931)

Measurements have been made of the optical properties of amphiboles in well-known rocks that have been classed as barkevikite or barkevikitic hornblendes by authoritative petrologists. Thus, in a lugarite obtained by the writer from the type locality, though the extinction-angle of the amphibole was that cited above (11°), the optic axial angle determined by Dr. Turner was 68°-72°. In the teschenite of Neutitschein, Moravia, in which the amphibole was described by Rohrbach (1885), cited by Rosenbusch (1907) and Johannsen (1934), as barkevikitic hornblende, the extinction-angle is nearly 20°, but 2V is 74°-84°, as measured by Dr. Turner. The

[Footnote] (a) If a quarter of the TiO2 acts as SiO2

[Footnote] (b) If a third of the Al2O3 is included with R”’.

[Footnote] (c) If a thirtieth of the Al2O3 is included with R”’.

[Footnote] (d) If but little of the H2O+ is included in the molecule.

– 288 –

barkevikitic hornblende in sample of laurvikite from Quen and Tjolling, near Frederiksvarn, South Norway, originally described by Brögger, showed c∧Z = 12° with 2V = 60°-66°, in the former case, and 76°-78° in the latter (fide Turner).

The optical properties of the various amphiboles of which the analyses are cited here are recorded in the following table:—

Table V.—Optical Properties of Analysed Brown Amphiboles.
Anal. No. 110°110 c∧Z on 010 c∧Z′ on 110 α β γ γ-α 2V-Cal. 2V-Obs. Authority
1 55° 35′ 10° α Ussing
2 55° 35′ 8-9°β 1.676 1.694 1.708 .32 82° 81° Wright
4-5 12° 82° Drescher
6 1.4° 1.692 1.730 1.760 .068 80° 72° Wright
7 55° 43′ 4.4° 1-5° 1.687c 1.705 1.718 .031 79° 80°c Tomita
8 1.684 1.701 1.720 .036 Tomita
9 1.680 1.701 1.709 .029 75° Tomita
11 55° 25′d 1.6821 1.6975 1.717 .035 78°f (e)
12 10° 1.687 1.708 .021 (g)
13 55° 36′ 11° 1.680 1.701 .021 52° Scott
14 18° 1.687 1.701 .014 54° (g)
15 55° 44′ 8.7° 1.670 1.682 1.693 .023 84° 83° Vendl
X y Z
1 light brown dark reddish-brown darker reddish-brown Ussing
4-5 light straw-brown brown-red with-chocolate tinge brown-red Drescher
6 pale olive-brown brown very dark-brown Wright
7 pale yellow-brown, greenish tinge reddish-brown greenish-brown Tomita
11 pale straw-brown reddish-brown darker reddish-brown
12 light yellow-brown reddish-brown dark brown (g)
13 light yellow reddish-brown very dark brown Scott
15 pale yellow-brown dark green dark brownish-olive Vendl

Not stated whether c∧Z or c∧Z′ measured on 110.


c∧Z greater for red than for green light.


Measured in Na light. The complete range of refractive dispersion is determinable from Tomita's data.


Mean of eight measurements with range 4′.


Measurements of α and β and also of n1 = 1.6932 in a cleavage-flake (all ± .0002) were made in Na light on an Abbe refractometer by Dr. G. D. Osborne. Measurements of γ = 1 717 ± .003 and of n2 = 1.716 ± .003 were made in similar light on a Rayner refractometer by Dr. G. D. Osborne, of Sydney University.


Mean of 76°, 78° and 80° measured by Dr. F. J. Turner on different parts of one crystal slice cut perpendicularly to the vertical axis.


Measurements of refractive indices and extinction-angles made by Kunitz on analysed crystal. The optical axial angle is that given by Johannsen (1908) and Larsen and Berman (1934) as characteristic of barkevikite. Winchell (1933) gives 2V = 31°-52°.

– 289 –

From the above data it would appear that the distinction between barkevikite, barkevikitic hornblende and basaltic hornblende, which may be drawn chemically on the proportion of MgO to FeO (increasing in the above sequence), cannot be based satisfactorily on either the intensity of the reddish tinge for Y and Z or on the refractive indices. The range of the latter for barkevikites is α = 1.680-1.694 and γ = 1.701-1.708 with c∧Z = 10°-20°, and for basaltic hornblendes α = 1.667-1.681 (extended to 1.687 by Tomita) and γ = 1.688-1.701 (extended to 1.720 by Tomita) with c∧Z = 0-12°. The use of the extinction-angles and optic axial angles may, however, lead to a conventional separation, and it is proposed in the sequel to term “barkevikite” amphiboles with 2V less than 60°, and “barkevikitic hornblende” those with 2V greater than 60° and c∧Z more than 10°, while “basaltic hornblende” will denote those amphiboles with 2V greater than 70° and c∧Z less than 10°. There does not seem to be any optical grounds for the distinction of richly sodic barkevikite (Anal. No. 13) from varieties containing but little soda (Nos. 12 and 14). Nor are there obvious optical grounds for the distinction of richly titaniferous basaltic hornblendes (kaersutite) from those with a small content of TiO2. High refractive indices may result either from an increased content of Fe2O3 as in basaltic oxyhornblende or from increased TiO2 as in kaersutite. The highest figures occur in amphiboles in which both these oxides are very abundant, as in linosite (Anal. No. 6). The optic axial angle is high, the extinction-angle small and the pleochroic tints fairly uniform throughout the basaltic hornblende group, though greenish tints have been noted in some richly ferric examples (Kunitz, 1930; Parsons, 1930).

Occurrence OP Brown Amphiboles in the Dunedin District.

The igneous activity commenced in the late Cainozoic times after the peneplanation of warped Middle Cainozoic sediments (see Benson, 1934). The earliest products of eruption were anorthoclase trachytes as tuffs, breccias, flows and dykes, among which amphibole is occasionally present though usually too altered for investigation. What may be a late member of these intrusions occurs as a dyke (No. 2957 in the collection of the Geological Department of Otago University) crossing the roadway half a mile S.S.W. of Harbour Cone. It is a trachyandesite with oligoclase both as phenocrysts and in subordinate laths among the sanidine in the base. Though pale titan-augite occurs as phenocrysts and in the base, the more abundant dark mineral is a perfectly fresh unresorbed amphibole in prisms up to 2.2 × 0.4 mm. and small ground-mass prisms up to 0.1 × 0.03 mm. The optical characteristics of these are noted in Table V. The larger prisms are frequently twinned and permit the determination of the extinction angle on the universal stage by Nemoto's (1938) method.

Later than the trachytes, and possibly also later than this trachyandesite, are the hornblende basalts which are among the concluding members of the lavas of the first volcanic phase occurring in the Mornington portion of the Dunedin urban area. The amphibole in these is generally resorbed, but occasionally (e.g. in No. 26) is determinable. In this rock the phenocrysts are up to 3 × 1 mm. in size, though usually more slender. Sometimes they are entirely surrounded by

– 290 –

a mantle of titanaugite, the b and c crystallographic axes of the two minerals being common. In these rocks also are large (0.5 × 0.2 mm.) prisms of faintly dichroie apatite containing long fibres or sheets of minute greyish or brownish inclusions arranged parallel to the vertical axes or more rarely perpendicularly to the prism faces. Edwards (1938, pp. 303-6) considers that similar inclusions in the unusually large apatites in Victorian basalts may be remnants of hornblende, and remarks that “such apatites are associated with the breaking down of basaltic hornblende in trachytes from New Zealand and many rocks from Kerguelen Island.” Certainly the presence of such large apatites in the rocks of the Dunedin district is almost invariably in association with more or less resorbed hornblende, and usually with signs of greater alkalinity than that of normal olivine basalt. The amphiboles commonly display a narrow border-zone of densely aggregated minutely granular iron-ore (probably titaniferous magnetite) almost obscuring a background of titanaugite, the last often extended into a thin clear film enclosing the darkened interior. The pyroxene over considerable areas may be crystallographically in parallel growth with the amphibole (as noted above), but generally this condition does not extend throughout, and much of the pyroxene is quite haphazard in its orientation. Though the finely granular dark aggregate may often extend from the marginal zone throughout the whole of the space occupied originally by the amphibole, as in the “ghost crystals” of Tomita, the inner portion is as a rule less rich in iron-ore, and small secondary plagioclase laths may be seen to have developed among the new-formed pyroxene. Often this more open structure may extend to the margin of the paramorph. Concentration of the minute magnetite grains into vertical and rarely intersecting lines indicating that resorption changes proceeded along the cleavage-planes inwards from the margin (as Rosenbusch, 1908, noted) has frequently been observed in these rocks. It was not possible, however, to confirm Tomita's statement that in such replacements the new-formed pyroxene was never pigeonitic, though the large optic axial angle of normal titanaugites was observed in several instances. No hypersthene or olivine was noted among the resorption-products.

In other cases the material within the darkened margin is less finely granular. The distinctive feature is the occurrence of platy or “club-shaped” masses up to 0.20 × 0.02 mm. in length and breadth, commonly elongated parallel to the vertical axis of the amphibole, but also occurring either perpendicular or at 60° thereto, or forming tufts composed of a few such plates spreading towards the interior of the original crystal, or in shorter smaller and less regularly placed plates. The pleochroism of such new-formed plates is intense. Sections changing from greenish-brown to dark reddish-brown have their maxima markedly oblique to their elongation, which is nearer to the vibration plane giving reddish-brown. Sections changing from reddish-brown to an almost opaque red-brown show maxima more nearly parallel to the elongation, but because of the invariable presence of pyroxene within the thickness of the rock-slice the optical orientation cannot be stated definitely. This, doubtless the “brown material” described by Washington (1894), is probably

– 291 –

rhönite such as was formed in Tomita's (1934) kaersutites. Scattered magnetite and small secondary plagioclase laths are also developed in this connection (see Fig. 1, A). Rosenbusch's (1908) comment that the hornblendes in many basalts have xenocrystic characters or occur only where the basalts are rather abnormal or alkaline is generally applicable to the Dunedin district. The larger brown crystals, though often idiomorphic, are commonly more or less corroded or rounded. They may contain large apatite prisms, which, as noted above, may also occur separately in the ground mass, and

Picture icon

Fig. 1.—Transformation of kaersutite to rhönite and finely divided magnetite.
A. Commencement of the change in inclusion of pegmatoid gabbro-essexite in the breccia; Port Chalmers (D1).
B. Further stage with marginal concentration of the magnetite, and some separation of titanaugite. In hornblende basalt (26), Mornington, Dunedin.
C. Complete replacement of kaersutite, by rhönite, titanaugite, magnetite and plagioclase in gabbroid inclusion in basalt (522). Western side of Flagstaff Hill.

– 292 –

when either brown amphibole or large apatite prisms are present in a basaltic rock, a little sanidine is usually determinable between the laths of andesine or labradorite. Where this is more obvious the rock is classed as trachybasalt, a flow of which (697) occurs among the older first-phase lavas near Port Chalmers, another (387) among the later first-phase lavas near Purakanui, and others elsewhere, in which not only resorbed amphibole but also phenocrystic aegirine-augite is surrounded by a mantle of titanaugite.

The source of the presumably cognate amphibole xenocrysts may be inferred from the nature of the inclusions in the kaiwekite (Flow No. 13 of Marshall's North Otago Head sequence), a porphyritic olivine-bearing anorthoclase “trachyte,” one of the latest of the lavas of the first volcanic phase. Brown amphibole occurs in this in three ways:—


In inclusions of pegmatoid kaersutite-pyroxene-gabbro (or gabbro-essexite).


In inclusions of nepheline syenite pegmatite.


In large single crystals or crystal-clusters.

1. Specimens of these inclusions have been obtained both in the cliffs at the northern end of Long Beach (Purakanui), where they are common, and at North Otago Head. The presence of feldspar is not obvious in hand specimens, though a lustre-mottling of the hornblende cleavage-planes (up to 3 mm. long) shows the presence of inclusions therein. Microscopically (961, 4729, see. Fig. 2, A), kaersutite is seen to be the dominant mineral and has the optical properties noted on Table VI. It occurs in large poikilitic masses containing minute dark brown translucent schiller-plates elongated usually parallel, but also perpendicularly to the vertical cleavage. Irregular grains of magnetite 0.1-0.2 mm. in diameter are scattered sparsely through it, rare apatite prisms, more abundant colourless grains of olivine up to 3 mm. across (the optic axial angle of which 2V = 82° — indicates a composition of Fa 32%), and still more abundant masses of pyroxene occurring as subidiomorphic prisms 5 × 1 mm. in size fraying out into mutually poikilitic intergrowth with one or more of the large kaersutite individuals, or as separate but irregularly bounded grains scattered with or without optical parallelism through the amphibole or the plagioclase-background. Sometimes the parallel position of mutually intergrown pyroxene and amphiboles may be observed; at other times the latter mineral seems to be merely a thin film surrounding the former and giving it the appearance of possessing a faintly pleochroic brownish tint. The pyroxene is not homogenous. Though usually a pale greenish diopside, portions of a single optically continuous grain may pass into an irregularly not zonally bounded mass of titanaugite. Subidiomorphic grains of titanaugite are also scattered haphazardly or regularly oriented through the amphibole, or may occur embedded in plagioclase with or without an enclosing film of amphibole. The pleochroism of such titanaugite is very marked and follows the normal scheme. There is little difference of optic axial angle between the several types of pyroxene. The clear greenish diopside has 2V = 58°-62°, the schillerised colourless pyroxene 53°-62°, and the schillerised titaniferous

– 293 –
Picture icon

A, Kaersutite, titanaugite and diopside, olivine, labradorite, magnetite and apatite in pegmatoid gabbro xenolith (4729) in kaiwekite; North End of Long Beach, Purakanul.
B. Kaersutite, titanaugite, diopside and andesine with magnetite in pegmatoid gabbro-essexite inclusion (D3) in breccia; Port Chalmers.
C. Barkevikitic hornblende, oligoclase, orthoclase (?) and analcite with magnetite and abundant apatite in a turbid zeolitised base. Inclusion of zeolitised lugarite (C2) in breccia; Port Chalmers.
D. Pseudomorphs in finely divided magnetite after barkevikite (?) with orthoclase, oligoclase-andesine, and nepheline in nepheline syenite (C7). Probably inclusion in breccia; old shaft, Battery Creek, Harbour Cone.

– 294 –

variety 52°-60°. It is noteworthy that Drescher (1932) calls attention to the varying colour of the pyroxene in the dolerite invading the Kaersut picrite sill, and considers it to result from an uneven concentration of TiO2 such as is indicated by his detailed study of the associated iron-ores. But though diopside and titanaugite occur together in the kaersutite-augite-pegmatite of Kaersut (to which the rock under consideration has considerable resemblance), diversity of colour in a single grain has not been recorded therein by Drescher. The plagioclase grains in our pegmatoid gabbro are also poikilitic, up to 6 mm. in diameter, of labradoritic composition (Ab40) and faintly schillerised. It varies greatly in amount from 20% to 90% in the range of specimens examined.

Connecting these pegmatoid gabbro inclusions with those of nepheline syenite is a xenolith of a rather essexitic composition. It contains about 40% of andesine, which is not poikilitic, and is associated with a small amount of orthoclase and of nepheline with secondary analcite. Pyroxene in large poikilitic masses forms the dominant coloured mineral and is usually titaniferous, but with a faintly pleochroic bright green alkaline margin. Brown amphibole with c∧Z = 12° occurs intergrown with this pyroxene, and is marginally resorbed with the formation of finely granular magnetite. There is much idiomorphic apatite in both coloured and colourless minerals, but no divine.

2. A rock obtained by Dr. Marshall as an inclusion almost certainly from the kaiwekite of North Otago Head proves to be a fragment of nepheline syenite, in which a small amount of largely resorbed brown amphibole containing apatite and aegirine-augite occurs with more abundant separate crystals of this pyroxene and of anorthoclase, the last forming large poikilitic masses through which are scattered optically parallel grains of nepheline and of aegirine-augite. The coloured constituents together comprise less than a fifth of the rockmass. The cores of residual amphibole give tints of golden to deep red-brown. The optic axial angle was not determinable.

3. The scattered individual crystals of amphibole are up to 4 cms. in diameter. Several of these were crushed and a sample composed of clean cleavage-fragments free of any lustre-mottling selected under the microscope was analysed by Mr. F. T. Seelye (see Table II, No. 10). On such fragments c∧Z′ is nearly 10° in flakes parallel to (110), and the Y and Z brownish colouration has but little of the reddish tint of barkevikite. In smaller crystals scattered throughout a representative series of kaiwekites a range of optical properties was observed which are indicated in Table VI.

The characters of the coarsely crystalline inclusions of the group first described recall not only those of the kaersutite pegmatites of Greenland, but also those of certain pegmatoids described by Lacroix (1928) in volcanic rocks of basaltic facies. The mutual poikilitic intergrowth of the coloured constituents and their enclosure by plagioclase is in particular comparable with the features of pegmatoids, though there is not among the present group of inclusions any development of interstitial finely lathy alkaline feldspar (the association of which with more coarsely granular material is especially

– 295 –

charactersitic of pegmatoids), and but little primary zeolite, which, however, is seen among other coarse-grained inclusions occurring in rocks belonging to a later phase of the Dunedin volcanic series. It is probable, however, that in spite of their close similarity with the rocks of the Kaersut complex, the inclusions here described were developed as local segregations from magmas rather than as portions of individual rock-masses which have been included as fragments in the rocks in which they are now embedded. If that be the case, the diversity between the types of coarsely granular inclusions (1 and 2) in kaiwekite is in accord with the view to be discussed elsewhere, that kaiwekite is a hybrid rock produced by the mingling in varied proportions of diverse partially crystallised magmas.

The first volcanic phase in the Dunedin area closed with a varied series of intrusions, and explosive eruptions producing immense vent-filling breccias. Brown amphiboles of several types occur among the products of these. A “camptonitic” dyke (J.10, La. 16, La. 17*) described by Marshall (1906, p. 398) from near Portobello is noteworthy for the presence of much thompsonite and analcite, as indicated by chemical and optical tests including staining (Marshall, priv. com.). Phenocrysts of pale green or colourless diopside occur.

Picture icon

Fig. 3.— Barkevikitic hornblende associated with a little biotite mantling idiomorphic or corroded olivine in a camptonitic dyke (J10); Portobello.

The amphibole phenocrysts are free from any resorption and may contain a core of pyroxene. They are often twinned and faintly zoned, and form prisms tip to 2.0 × 0.8 mm. in size. Small (0.3 mm.)

[Footnote] * Microscope slides, of which the catalogue numbers are prefixed by a letter or letters, were used by Marshall, and in general no corresponding rock-specimens are now available.

– 296 –

phenocrysts of ilmenite are also present, and contain minute prisms of barkevikite and apatite. The groundmass includes many small thin amphibole prisms and shorter prisms of pyroxene. The optical properties of the amphibole are noted in Table VI, the smaller optic axial angle and more reddish colour of the central portions may betoken a more alkaline or barkevikitic composition than that possessed by the outer portion, though the birefringence and extinction-angle (indicated by Nemoto's method of determination) remain constant throughout. There is commonly a very thin greenish marginal film which is possibly arfvedsonitic (2V = 68°) fraying out into the zeolitic groundmass. The groundmass-amphibole occasionally appears wrapped about small sometimes corroded olivine phenocrysts, where it may be associated with flakes of biotite (see Fig. 3).

The “tinguaitic camptonite” or ulrichite of the Portobello Peninsula (Marshall, 1906, p. 397) is closely similar to the above. It (J.24) contains amphibole recognised as barkevikite by Marshall (see Table VI) associated with and sometimes containing idiomorphic

Picture icon

Fig. 4.—Irregularly shaped masses rich in barkevikite in a camptonitic groundmass containing relatively little amphibole (La2). Dyke, Portobello.

– 297 –

olivine (2V = 86° (—): Fa 24%), without regular orientation. Another allied dyke rock from Portobello (La.2) is noteworthy for the presence of irregular masses of a richly hornblendic camptonite in a leucocratic host (see Fig. 4).

A nepheline syenite porphyry forms a laccolitic intrusion occupying most of Varley's Hill on the northern side of Hooper's Inlet. It was first described as a tinguaite (Marshall, 1906, p. 396) and attention was drawn to the formation of groups of aegirine, magnetite and analcite derived from resorbed brown amphibole, of which some portions remain. The rock (J.3, J.15) consists chiefly of phenocrysts of anorthoclase, nepheline and more or less resorbed amphibole in a matrix of lathy sanidine, short prisms of nepheline aegirine-augite,

Picture icon

Fig. 5.—Basal section of a largely resorbed crystal of barkevikitic hornblende, replaced by green aegirine-augite partly in parallel intergrowth (stippled) magnetite, plagioclase and analcite. Abundant small prisms of apatite shown by dark margins. In nepheline syenite porphyry (103); Varley's Hill, Hooper's Inlet.

– 298 –

a little magnetite and interstitial analcite, the last mineral also occurring in goedes with natrolite and aegirine prisms. The remnants of amphibole, possibly as a result of the deuteric reactions, show a considerable range of optical characters. The value of 2V varies between 46°— (barkevikite) and 80° (—) (basaltic hornblende) though in ten other cases it lies within 8° of 60° (—). Where faintly-marked zoning is present, the darker outer zone has a lower value for 2V (62°— and 58°—), the same value for γ-β and the same or rather higher value for β-α. The extinction-angle c∧Z is 9°-12°, and the pleochroism shows × pale straw-yellow, Y deep red-brown or deep golden-brown, and Z deep brown or deep golden-brown with absorption X<Y<Z. Occasionally the amphibole contains diopside with 2V = 64°. The resorbed portions of the amphibole, though generally free of very finely divided magnetite, commonly contains rather coarsely granular magnetite, especially in the outer portions, probably as a result of the long duration of the conditions favourable to resorption. Aegirine-augite, either originally enclosed in the amphibole, or resulting from its resorption, is usually oriented so as to have the crystallographic axes b and c in common therewith, where the two minerals are in contact, but it often forms a series of short variously oriented prisms outlining or scattered through the space occupied originally by the amphibole. The inner portion is commonly a spongy aggregate of pale green aegirine-augite, nepheline, sanidine (?), analcite and sometimes natrolite with prisms of the originally enclosed apatite. There is a suggestion in the distribution of the pyroxene that it was to some extent strewed about during the resorptive reactions. Figure 5 herewith supplements Marshall's illustration (1906, Pl. XXXVII, Fig. 1) of these interesting pseudomorphs.

The explosive eruptions which preceded the effusions of the second volcanic phase produced a row of breccia-filled vents ranged along an anticline extending S.S.E. through Port Chalmers, where is the largest vent. Amphibole is present in the breccia either as separate crystals, or as a component mineral of various fragments of rocks included therein. Of these the most noteworthy are (1) the nepheline syenites or foyaites at Port Chalmers and the foot of Harbour Cone. In the latter locality Boult (1906) and Marshall (1906) thought these rocks to be portions of an intrusive mass ascending nearly to the present surface. The writer considers their immediate derivation from a breccia to be more probable. There are also (2) nepheline syenite porphyries as inclusions at Port Chalmers not unlike those in situ at Varley's Hill (3) a dark coloured rock formerly classed as diorite but closely similar to the above described pegmatoid gabbro in the kaiwekite, and (4) a probably analogous rock in the breccia at Harbour Cone described by Boult (1906), who suggested that the greenish-brown amphibole therein was allied to hastingsite, but gave no evidence to support this. None of this last rock has been available for the present study. Some details concerning these inclusions are as follows:—

2. In the Harbour Cone syenite the amphibole formed prisms in size about 2.0 × 1.0 × 0.8 mm. and often contained apatite and more rarely diopside or olivine, and was intergrown with irregularly oriented biotite (X = pale yellow, Y = Z a red-tan colour, 2V = 0°).

– 299 –

The mica remains absolutely fresh, but the amphibole is largely resorbed with the production of pale green diopside in the usual orientation, scattered magnetite and a little feldspar. A specimen in which the amphibole has been almost entirely resorbed is illustrated in Figure 2 D. The optical properties of such remnants of it as are determinable (see Table VI) are approximately marginal to those of barkevikite. Orthoclase, oligoclase-andesine and nepheline are the associated minerals. An originally comparable rock from Port Chalmers (C7) contains unresorbed barkevikitic amphibole with a brown rather than a reddish tint and a greenish margin indicating perhaps an approach to arfvedsonitic composition.

3. The rocks formerly termed diorite (Marshall, 1906, p. 413), but which resemble more the pegmatoid gabbro-essexite in the kaiwekite, find some analogy in the lugarites described by Tyrrell (1917). Characteristic examples (D1, D3, see Figure 2 B) are composed of large (more than 8 mm.) irregular or poikilitic basaltic hornblende (2V = 78° (—); 82° (—), c∧Z = 4°-6°) with occasional parallel intergrowth of pale green diopside, faint lilac titanaugite (2V = 48°), andesine (Ab60) and natrolite with a little analcite and large (1.3 × 1.0 mm.) corroded prisms of apatite. The less coarsely granular specimens differ from lugarite collected by the writer from the type locality, in that the Scottish rock contains rather more barkevikitic amphibole (2V = 68°-72° (—))* associated with some biotite and ilmenite, which are not present in the Port Chalmers inclusions. Linking the latter with syenites is a specimen (C2, see Fig. 2 C) in which perfectly idiomorphic fresh barkevikitic hornblende (2V = 62°-68° (—), c∧Z = 12°) unaccompanied by other coloured minerals, large prisms of oligoclase and of orthoclase (?) with some analcite lie in a rather turbid zeolitic matrix containing scattered grains of magnetite and minute prisms of apatite which are also present in the amphibole. These features are reproduced in the matrix of some lugarites. A further variation among the rocks under this head are certain other rocks at Port Chalmers (D4, D5, D6) containing masses of kaersutite, diopside (2V = 64°, c∧Z = 39°), labradorite and carbonate-pseudomorphs after olivine, with abundant apatite and magnetite and but little zeolite. They are very like the pegmatoid gabbros in the kaiwekite.

In some of the gabbro-essexite rocks the amphibole is marginally slightly resorbed with formation of magnetite, while within small narrow plates or prisms of rhönite have been formed elongated usually parallel to the vertical axis of the host. This change extends inwards along the cleavage-cracks in narrow bands exactly as illustrated by Tomita (1934, Plate xx, Fig. 2). The deep red ragged prisms of rhönite are up to 0.2 mm. long. (See Fig. 1 A.)

The conglomerates below the lavas of the second volcanic phase contain material derived from these breccias and various other products of the first volcanic phase. From the small exposure of these conglomerates near the head of Morrison's Creek was obtained a pebble of a peculiar trachyphonolite (735) in which phenocrysts of

[Footnote] * 2V = 52° in the richly sodic material examined by Scott (1914) and Tyrrell (1917).

– 300 –

almost fresh basaltic hornblende* (see Table VI) occur with zoned andesite, aegirine-augite and magnetite in a trachytoid matrix of oligoclase, sanidine, and minute (0.04 mm.) almost circular, weakly birefringent grains possibly leucite, together with small prisms of the coloured silicates. No such rock has been found in situ in the Dunedin district.

Elsewhere, particularly in Fraser's Gully, west of Dunedin, a coeval conglomerate contains pebbles of hornblendic trachybasalt (58,61) possibly derived from the above described Mornington flows, and of trachyandesite (P7, P8) derived from an unknown source. The amphibole in the latter is of two types, the yellowish variety (barkevikitic hornblende) being marginally absorbed, the reddish type (barkevikite) perfectly fresh. Their optical properties are noted on Table VI. The titaniferous character of the trachyandesite is emphasised by the presence of large (0.3 mm.) grains of perofskite. Pebbles of cossyritic phonolite (P26, P45) afford another variety of brown amphibole and the only known instance of the occurrence of cossyrite in products of the first phase of volcanic action in the region.

The earlier flows of the second volcanic phase are largely basaltic and rarely contain amphibole. The mineral reported to occur interstitially in the lowest of these basalts flow (No. 14) in the sequence at Otago North Head (Marshall, 1914) appears to the writer to be biotite rather than amphibole. Rocks tending to a trachybasaltic composition and containing more or less resorbed amphibole often accompanied by large prisms of dichroic apatite occur here and there among the older flows of this phase as e.g. on the south side of the Otago Peninsula opposite Ravensbourne (555) and on the western slopes of Flagstaff (522). The latter is of especial interest in that it contains a gabbroid xenolith of the type described above in which the rhönitic product of the resorption of amphibole is very well displayed (see Fig. 1 C). Near Harrington Point, by the entrance to the Harbour, a holocrystalline kulaite (691) forms a thin flow among this series of lavas, and contains abundant small resorbed phenocrysts of amphibole of which too little remains for optical study. This is part of a once rather widely extended flow of which the northern end six miles away forms the lowest of the flows at Omimi. At Brinn's Point, two miles north-east of the last-named locality, resorbed hornblende occurs either separately or enclosed in greenish diopside bordered by titanaugite forming phenocrysts in a remarkable hyalobasanite (1013, 1042).

At the base of the series of second phase flows a mile north of Port Chalmers there is a small development of trachyandesite (4834) containing xenocrystic (?) basaltic hornblende apparently derived from included pegmatoid fragments, aggregates of slender hornblende prisms both small and large varying in length up to 3 mm., the larger prisms being often prolonged into thin laminae which with the smaller prisms are more or less resorbed. In addition there are less idiomorphic stout prisms of augite and a matrix of more or less idiomorphic large and small tabulae of slightly zoned andesine containing small apatite needles, and rarely octahedra of magnetite. Sphaero—

[Footnote] * Optic axial angles determined by Mr. O. D. Paterson, M.Sc.

– 301 –

siderite, probably deuteric, occurs in considerable amount partly replacing the plagioclase.

Higher in the series of second phase flows is the very extensive Logan's Point phonolite, described in detail by Marshall (1906) and Cotton (1908), in which brown amphibole appears in the form of cossyrite, too irregular in its development for the exact determination of its optical characters. The member of the second phase series of flows which shows the greatest development of brown amphibole is, however, the rather younger and widespread Leith Valley trachybasalt first described as andesite by Marshall (1906), from which were derived the crystals of kaersutite which have been herein studied in greatest detail. (See Tables II, III, V and VI.) The kaersutite occurs in this either (1) in fragments of coarsely crystalline pegmatoid gabbro or (2) in isolated often deeply corroded crystals up to 5 cms. long which have been obtained in particular from the Leith Valley quarries. The gabbroid rocks (e.g. 4730a) consist of basic labradorite, kaersutite, red-brown by transmitted light, pale lilac titanaugite with a faintly pleochroic greenish core, large (up to 3.0 mm.) irregular grains of ilmenite, and grey markedly dichroic apatite in prisms up to 1.5 × 0.2 mm. in size, included in all the other crystals. The kaersutite when slightly resorbed shows the usual finely granular “opacitic” margin with or without rhönite. Natrolite has been formed in crevices extending out from the feldspar, and in irregular cracks between the xenoliths and the enclosing trachybasalt. Analcite may also be present, occurring as a rule in cracks or irregular spaces lined by natrolite, but occasionally forming large grains in kaersutite mantled by a thin reaction-zone of titanaugite and rhönite. The amphibole displays in different crystals dark yellow-brown or dark reddish-brown tints for Y and Z possibly depending on the degree of oxidation of the iron therein. The penetration of the enclosing trachybasalt into the coarsely granular xenoliths and the isolation of xenocrysts therefrom are visible in contact specimens. The scattered crystals of kaersutite yielded the material for analysis, the purity of which was determined by microscopic examination of partly-crushed fragments. The large crystals may be more or less idiomorphic, but are more often rather embayed by corrosion. Such absorption is not always accompanied by resorption-phenomena, for in places the amphibole is quite fresh up to the embayed margin. The optical properties of the analysed crystals and one other example have been tabulated. (See Tables V and VI.)

Resting on this trachybasalt is a phonolitic trachyandesite, the andesitic phonolite or Signal Hill phonolite of Marshall (1906) and Cotton (1908). It is nearly as widely distributed as the trachybasalt, from which it differs chiefly in the greater predominance of the colourless constituents. There are amphibole xenocrysts in it, the optical properties of which have been tabulated below (Table VI, p. 40, 445). These are as a rule distinguished by an unusually yellowish tint, and in the marginal resorption zone small grains of olivine may appear which seem to be primary rather than formed in association with the adjacent rhönite (cf. Tomita, 1934). With these may be a little calcite. Cotton (1908) concluded that, the amphibole was intermediate in character between barkevikite and

– 302 –

basaltic hornblende, and stated that the extinction angle c∧Z was 15°; Bartrum (1911) described a large (3 cms.) crystal with the same tint as barkevikite and gave c∧Z as 6°. The persistently large optic axial angle, however, betokens affinity with basaltic hornblende. It is worthy of note that Dr. Turner's universtal stage measurements have shown that plagioclase crystals in clots associated with the amphibole contain bytownite (Ab30) in their central portions.

The younger phonolitic conglomerates of Pine Hill and St. Clair rest on these and contain in the former locality among a variety of other rocks, boulders of the coarsely porphyritic tinguaite in one of which was the probably barkevikitic clear brown amphibole (a borderline type) described by Ulrich (1891). (Table VI.) These conglomerates are locally invaded, interstratified with, or covered by trachyandesite, a dyke of which (144) exposed at the junction of Pine Hill and Campbell's Roads shows beautifully fresh crystals of barkevikitic hornblende. (See Table VI.)

The basalts of the third volcanic phase cover these lavas and conglomerates, and are in turn succeeded by the widespread and varied phonolitic rocks which form the latest group of flows in the Dunedin district. Those of the summit and western slopes of Flagstaff and Mount Cargill are in a large measure the olivine-bearing “trachydolerite” of Marshall (1904, 1906) and Bartrum (1911), for which a more distinctive term is not affordeed by the standard nomenclature as the rock is apparently somewhat hybridised. Partially resorbed basaltic xenoliths, and clots of xenocrysts derived therefrom are scattered through the rock, and these include probably most of the very magnesian olivine present (cf. Benson and Turner, 1939). More normal cossyritic trachytoid or nephelinitoid ameletite phonolites extend further to the north-east. Amphiboles were not rare in the “trachydolerites,” though they have mostly been resorbed. Aegirine-augite and magnetite pseudomorphous after such amphiboles were described and figured by Marshall (1904, Fig. 2). More often the pseudomorph is an almost opaque aggregate of finely divided iron-ore. The original crystals are usually small and apparently phenocrystic rather than xenocrystic prisms, and rarely enclose relatively large prisms of apatite such as occur in xenocrystic kaersutite, though sometimes such apatite prisms are to be found in the ground mass of these rocks and seem foreign to it. The optical properties of one of these rarely remaining amphiboles were noted by Bartrum (44b) as those of a rather barkevikitic type, which is confirmed by the determination of the optic axial angle. There is, however, some variation in the size of the extinction angle c∧Z. (See Table VI.) In the ground-mass of these basified phonolites finely divided cossyrite is present.

In the more richly alkaline cossyritic ameletite phonolite, but little of the basaltic minerals appear save for the remnants of the largely resorbed magnesian olivine. Cossyrite is often very abundant in finely mossy aggregates notably in the rocks near Mopanui, but no developments of that mineral approach near enough to idiomorphism to permit a detailed determination of its optical properties. The strong bright red-brown to very dark brown pleochroism is, however, distinctive.

– 303 –
Table VI.—Optical Properties of Brown Amphiboles From Dunedin and Elsewhere.
Slide No. Rock and Crystals Locality, Amphibole, X Y Z c∧Z 2V(-) D.R.
Initial and First Volcanic Phase.
2957 Trachyandesite phenocryst. Groundmass. Portobello. Basaltic hornblende. Do. Light golden-brown. Light brown. Dark reddish-brown. Dark greenish-brown. Darker reddish-brown., Very dark greenish-brown. 5° 0°-2° - 72°-74° Med. Med.
26 Hornblende basalt. Inclusions in kaiwekite, Mornington. Do. Light brown. Dark yellowish-brown. Dark yellowish-brown.
4729 Pegmatoid kaersutite gabbro. Purakanui. Kaersutite, Light golden-brown. Reddish- brown. Darker reddish-brown. 8°-
D3 Pegmatoid gabbroessexite, Nepheline syenite. North Head. North Head. Barkevikitic hornblende (?) Do (?) Light yellow-brown. Dark red-brown. Darker reddish-brown, 12° - Large
La17 “Camptonite” phenocryst. Portobello, Barkevikitic hornblende, Straw-brown, Deep golden red-brown, Deep golden-red-brown, 11° 60°-74°
La17 Do. Inner Portion. Narrow margin, Groundmass crystals with dark greenish-brown border. Barkevikitic hornblende. Arfvedsonitic (?) hornblende. Barkevikitic hornblende. Straw-brown, Pale yellowish-green, Pale yellow-brown. Dark brown, Deep brown-green, Brown. Darker brown, Deep bluish, Deep brown. 11° 11° 11° 70°-82° 68° 72°
J24 Tinguaite porphyry (ulrichite). Portobello, Barkevikite, Pale yellow. Deep reddish-golden-brown, Deep reddish-golden-brown, 10°-11° 60° 62°
J3 and J15 Nepheline syenite porphyry. Hooper's Inlet, Inner portion barkevikite and barkevikitic hornblende, Outer portion barkevikite, Pale straw yellow, Straw. Deep reddish-golden-brown, Brown, Deep Golden-Brown to brown, Brown, 9°-12° 40°-52°-68° 80° 58° 62° Med
Inclusions in Breccia.
C7 Nepheline syenite, Harbour Cone, Barkevikite, Straw brown, Clove brown, Greenish-brown, 22° 56° 46° Low
– 304 –
C26 Nepheline syenite, Barkevikite, Straw. Red-brown. Red-brown. 10°
D3 Gabbroessexite. Port Chalmers. Basaltic hornblende. Pale brown. Dark yellowish-red brown. Darker red-brown. 4°-6° 78° 82° High
C2 Analcitised lugarite. Port Chalmers. Barkevikitic hornblende. Pale yellow, Deep brown with reddish tint. Deep golden brown. 12° 62° 64°
P7 Phonolitic trachyandesite. Large almost fresh phenocrysts. Smaller partly resorbed phenocrysts, Do. (fide Bartrum), Fraser's, Gully, Kaikorai, N.E. Valley. Barkevikitic hornblende, Barkevikite, Do. Basaltic hornblende. Pale yellow. Pale yellow, Do Pale golden-yellow, Deep golden-brown, Dark brown with reddish tinge, Do, Rich brown, Deep golden yellow-brown, Dark brown with reddish tint, Deep chocolate brown. Dark opaque brown, 8° 20° 6° 68° 68° 64°-60° 56° - Rather High
Inclusion in Second Phonolitic Conglomerate.
Tinguaite porphyry (fide Ulrich), Pine Hill Barkevikitic hornblende (?). Light yellow-brown. Deep rust brown, 10°
Third Volcanic Phase,
144 Trachyandesite. Pine Hill Barkevikitic hornblende, Pale yellow. Deep reddish-brown. Deep reddish-brown. 20° 74°
36C Basified phonolite or “Trachydolerite.” Mt. Cargill, Barkevikitic hornblende (?), Light straw brown. Reddish-brown. Reddish brown, 17°
44B Phonolite, Mt. Cargill. Barkevikitic hornblende, Light yellow-brown, Deep golden-red-brown, Deep red-brown, 74°
Brown Hornblendes in other well-known Rocks.
Lugarite, Ayrshire, Scotland, Barkevikitic hornblende Straw brown. Dark yellow-brown. Darker yellow-brown 11° 68°-72°
Teschenite, Neutitschenin, Moravia, Barkevikitic hornblende Pale yellow, Deep brown with reddish tint Deep golden-brown, 22° 74°-84°
Laurvikite. Quen Norway. Tjolling Norway, Barkevikite, Barkevikitic hornblende, Light golden, brown, Pale yellow-brown, Deep golden-brown, Reddish yellow-brown, Deep golden-brown Chestnut brown, 0°(?) 63°&3°
– 305 –


Collecting the results of the above discussion and of the data recorded in Table VI, it appears that in the Dunedin district brown monoclinic amphiboles occur very generally in the only slightly alkaline basalts and trachybasalts of the first and second volcanic phases, and prove to be basaltic hornblendes which in the two samples analysed have just sufficient content of TiO2 to permit their being classed as kaersutite. The amphibole occurs either in coarsely gabbroidal cognate xenoliths derived probably by segregation (Lacroix, 1893) at depth from such magmas, or in single grains separated from such segregations. Through processes of magmatic hybridisation such xenoliths or xenocrysts may be incorporated in a distinctly more alkaline milieu than that from which they originally crystallised. From magmas which have attained a more alkaline character, whether hybrid or derived by the ordinary processes of differentiation, increasingly alkaline sodic amphiboles may separate, namely types intermediate in character between basaltic hornblende and barkevikite, which are formed in magmas of trachyandesitic or camptonitic composition (and may subsequently appear as xenocrysts in more alkaline rocks). Definite barkevikites characterised by the possession of an optic axial angle approximating to 2V = 60° or less occurs in the richly alkaline nepheline syenites, tinguaites and some phonolites, and but rarely in any less sodic rock, e.g. phonolitic trachyandesite.

It should be stated that nearly all the universal stage measurements recorded herein were made by my colleague, Dr. F. J. Turner, who noted, moreover, that owing to the usual absence of minerals of which the birefringence is sufficiently definite to permit them to be utilised to determine the thickness of the rock-slices examined, it is rarely possible to give precise compensator-determinations of the birefringence of the amphiboles, and the terms high, medium and low are used in Table VI to indicate the approximate value within the known range for γ-α, i.e. 0-021 and 0.068.

The chemical compositions of a number of the rocks mentioned in the above table may be ascertained by reference to the works cited below. Two hitherto unpublished analyses of the rocks cited, for which the writer is indebted to Mr. F. T. Seelye, are appended.

La.16, 17. Camptonite, Portobello. Analysis by Marshall (1906), p. 398.

J.24. Camptonitic tinguaite (ulrichite). See Marshall, op. cit., p. 397.

J.3, J.15. Nepheline syenite porphyry. Varley's Hill, Hooper's Inlet. A rather incomplete analysis by Marshall, op. cit., p. 396, and a more detailed one by F. T. Seelye (704) is given below.

C7. Nepheline syenite in breccia (?) from old shaft, Harbour Cone.

For analysis by D. B. Waters, see Marshall, op. cit., p. 392.

– 306 –

4728. Trachybasalt containing permatoid kaersutite gabbro, Leith Valley. For a rather incomplete analysis of the host rock “andesite,” see Marshall, op. cit., p. 408. A more detailed analysis of part of the same flow has been made by F. T. Seelye. (See Benson and Turner, op. cit., p. 71.)

445. Phonolitic trachyandesite (“Signal Hill Phonolite”) For analysis, see Cotton, 1908, p. 120.

36C. Basified phonolite or “trachydolerite,” Mount Cargill. For analysis of a closely comparable or identical rock, see Bartrum, 1911, p. 177.

144. Phonolitic trachyandesite. Dyke, corner of Pine Hill and Campbell Roads, Pine Hill. Analysis by F. T. Seelye below.

Analyses Not Hitherto Published of Rocks Cited Above.
144 704 144 704 144 704
SiO2 52.27 51.82 H2O+ 3.43 3.80 NiO nt. fd nt. fd.
Al2O3 19.92 18.84 CO2 0.07 0.12 BaO 0.09 0.06
Fe2O3 3.08 3.71 TiO2 0.62 0.78 SrO 0.02 0.06
Feo 3.48 3.31 ZrO2 nt. fd nt. fd.
MgO 0.53 0.87 P2O5 0.27 0.34 99.92 100.00
CaO 3.71 3.20 S 0.02 0.03 O for Cl 0.03
Na2O 6.18 8.33 Cl 0.11 0.01
K2O 3.46 3.94 Cr2O3 nt. fd nt. fd.
H2O— 2.46 0.60 MnO 0.20 0.18 99.89 100.00


Papers marked with an asterisk have not been seen by the writer.

Barnes, V. E., 1930. Changes in hornblende about 800° C., Amer. Mineralogist, vol. 15, pp. 393-417.

Bartrum, J. A., 1911. Some Rocks of Mount Cargill, Trans. N. Z. Inst., vol. 44, pp. 163–179.

Benson, W. N., 1934. The Geology of the Dunedin District, N.Z., Abstract Proc. Geol. Soc. London, Dec. 14, pp. 10–13.

Benson, W. N., and Turner, F. J., 1939. Mineralogical Notes from the University of Otago, No. 2. Comparative Composition Variation Diagrams for the Cainozoic Igneous Hocks of New Zealand, with Determinations of the Optic Axial Angle of the Pyroxenes and Olivines therein, Trans-Royal Soc. of N.Z., vol. 69, pp. 56–72.

Berman, H., and Larsen, E. S., 1931. The Composition of Alkali Amphiboles, Amer. Mineralogist, vol. 15, pp. 140–144.

Boult, C. N., 1906. The Occurrence of Gold at Harbour Cone, Trans. N.Z. Inst., vol. 38, pp. 425–446.

*Brögger, W. C., 1890. Die Mineralien der Syenitpegmatitgange der Sudnorwegischen Augit und Nephelinsyenite, Leipzig.

—– 1894. Die Eruptivgesteine des Kristianiagebiets, Bd. I, Kristiania, pp. 34–38.

Cotton, C A., 1908. The Geology of Signal Hill, Trans. N.Z. Inst., vol. 41,. pp. 111–126.

Doelter, C., 1914. Handbuch der Mineralchemie, vol. 2, p. 630, Steinkepf, Leipzig and Dresden.

Drescher, F. K., and Krueger, H. K. E., 1928. Die Peridotit von Kaersut (Grönland) und sein Gangefolge als Beispiele einer Sekretions-differentiation Neu. Jahrb. Für. Min., Beil.-Band, 62, Abt. A, pp. 569- 616.

Drescher, F. K., 1932. Zur Kenntniss des Peridotits vom Grönland und seiner Gangefolges. Min. Petr. Mitt., Bd. 43, pp. 207–270.

– 307 –

Edwards, A. B., 1938. The Tertiary Volcanic Rocks of Central Victoria, Quart. Journ. Geol. Soc., vol. 94, pp. 243–320; esp. 333–6.

*Gossner, B., and Speilberger, F., 1929. Chemisette und rontgengraphische Untersuchungen an Silikate. Ein Beitrage zur Kenntniss der Hornblendegruppe, Zeits. Für Kristallographie, vol. 92, pp. 111–142. Min. Abstracts, vol. 4, no. 5, pp. 199–200.

Greenwood, G., and Parsons, A. L., 1931. The Lattice Dimensions of Certain Monoclinic Amphiboles, Univ. of Toronto Geol. Studies, no. 30, pp. 29–39.

*Heim, A., 1911. Ueber Petrographie und Geologie der Umgebungen von Karsuarsuk, Meddel. om Gronland, Bd. 47, pp. 1–175.

Johannsen, A., 1908. Determination of the Rock-forming Minerals, John Wiley and Sons, New York.

—– 1938. A Descriptive Petrography of the Igneous Rocks, vol. 4, Univ. of Chicago.

Kawano, T., 1934. Chemical Formula of Basaltic Hornblende, Proc. Imp. Acad. Japan., vol. 10, p. 349–352.

KozU, S., Yoshiki, B., and Kani, K., 1927. Note on the study of the transformation of common hornblende into basaltic hornblende at 750° C., Science Ret. Tohuku Imp. Univ., Sendai, ser. 3, vol. 3, pp. 143–159.

Kunitz, W., 1930. Die Isomorphverhältnisse in der Hornblendegruppe, Neu. Jahrb. fur Min., Beil.-Bd., 60 Abt. A, pp. 172–251.

Lacroix, A., 1893. Les Enclaves des Roches Volcaniques, Macon.

—– 1928. Les pegmatoides des roches volcaniques à facies basaltiques, Comptes Rendus Acad. Sci. Paris, vol. 172, pp. 321–327.

Larsen, E. S., and Berman, H., 1934. The Microscopical Determination of the Non-opaque Minerals, Bull. U.S. Geol. Survey, no. 848, p. 190.

Larsen, E. S., Irving, J., and Gonyer, F. A., 1937. Petrographical Results of a Study of the Minerals from the Tertiary Volcanic Rocks of the San Juan Region, Colorado, Amer. Mineralogist, vol. 22. p. 889–905.

*Lorenzen, J., 1884. Meddel. om Gronland, Bd. 7, p. 27.

*—– 1886. Zeits. für Kristallographie, Bd., p. 11, 318.

Machatschki, F., 1929. Uber die Formel des Monoklinen Amphibolen und Pyroxene, Zeits. für Kristallographie, Bd. 71, pp. 219–236. Min. Abst., vol. 4, no. 5, p. 202.

Marshall, P., 1904. Trachydolerites near Dunedin, Trans. Aust. Assoc. Adv. Science, vol. 10, pp. 183–188.

—– 1906. The Geology of Dunedin (New Zealand), Quart. Journ. Geol. Soc., vol. 62, pp. 381–424.

—– 1914. The Sequence of Lavas at the North Head, Otago Harbour, ibid., vol. 70, pp. 382–408.

Murgoci, G., 1922. Sur la Classification des Amphiboles bleues et de certaines Hornblendes, Comptes Rendus Acad. Sci. Paris, vol. 175, pp. 426–429.

Nemoto, T., 1938. A new Method of obtaining the Extinction Angle of Monoclinic Minerals especially of Pyroxenes and Amphiboles by means of Randon Sections, Journ. Fac. Sci. Hokkaido Imp. Univ., Series 4, vol. 4, pp. 107–111.

Parsons, A. L., 1930. A Chemical and Optical Study of Amphiboles, Univ. of Toronto Geol., Ser., no. 29, pp. 29–33. Min. Abst., vol. 4, no. 5, p. 390.

Penfield, S. L., and Stanley, F. C., 1907. The Chemical Composition of Amphibole, Amer. Journ. Sci., vol. 23, pp. 23–52.

*Phalen, W. C., 1903. Note on the Rocks of Nugsuak Peninsula and its Environs, Smithsonian Misc. Coll., vol. 45, p. 183.

*Rohrbach, C., 1885. Uber die Eruptivgesteine im Gebiet der schliesischmährischén Kreideformation, Min. Petr. Mitt., vol. 7, pp. 1–63.

Rosenbusch, H., 1904–8. Mikroskopische Physiographic der Mineral ien und Gestiene., Stuttgart.

*Steenstrup, J. K. V., 1883–4. Bidrag til kjendskap til de geognoatiske og geographiske Forholt i en Del af Nord Gronland, Meddel. om Gronland, Bd. 4 and 5.

– 308 –

Scott, A. 1914. Barkevikite from Lugar, Ayrshire, Min. Mag., vol. 17, pp. 138–142.

Tomita, T., 1934. On Kaersutite from Dogo, Oki Islands, Japan, and its Magmatic Alteration and Resorption, Journ. Shanghai Inst. Sci., Section 3, vol. 1, pp. 100–136.

Tyrrell, G. W., 1917. The Picrite-Teschenite Sill of Lagar, Quart. Journ. Geol. Soc., vol. 72, pp. 84–131.

Ulrich, G. H., 1891. On the Occurrence of Nepheline-bearing Rocks in New Zealand, Trans. Aust. Assoc. Adv. Sci., vol. 3, pp. 127–150.

Vendl, M., 1924. The Chemical Composition and Optical Properties of a Basaltic Hornblende from Hungary, Min. Mag., vol. 20, pp. 237–40.

Warren, B. E., 1930. The Crystal Structure and Chemical Composition of the Monoclinic Amphiboles, Zeits. für Kristallographie, vol. 72, pp. 493–517. Min. Abst., vol. 4, no. 5, p. 278.

Washington, H. S., 1894. The Volcanics of the Kula Basin in Lydia, Inaug. Dissertation, Leipzig.

Washington, H. S., and Wright, F. E., 1908. Kaersutite from Linosa and Greenland, Amer. Journ. Sci., vol. 26, pp. 187–211.

Winchell, A. N., 1933. Elements of Optical Mineralogy, Part 2, Third edition p. 254, Wiley, New York.


Professor Speight has kindly called attention to an error in “Mineralogical Notes, No. 2”; this volume, p. 57, last sentence in the second paragraph. He has already reported the presence of quartz in one Banks Peninsula trachyte, that of Cass Peak.