The Igneous Rocks of the Brocken Range-Ngahape Area, Eastern Wellington.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 72, 1942-43, Page 353

The Igneous Rocks of the Brocken Range-Ngahape Area, Eastern Wellington. By C. O. Hutton, Petrologist, New Zealand Geological Survey. [Read before Wellington Branch, October 9, 1941; received by the Editor, November 24, 1942; issued separately, March, 1943.] Introduction. In this paper are embodied the results of a mineralogical and petrological investigation of the igneous rocks of the Brocken Range—Ngahape area, Eastern Wellington. The geology of this area has recently been studied by Mr. D. A. Brown (1943) and the rocks herein described were collected by him. At his request this study was undertaken, and it is divided into three parts—viz., mineralogy, petrography, and petrogenesis. Mineralogy Two minerals, a pyroxene and an ore, have been separated in the pure state from the teschenite of Red Hill (P. 7616) by centrifuge and electro-magnetic methods. For the final stages of purification the two minerals were crushed finely in an agate mortar, and the powders thus prepared were centrifuged in order to free the products from inclusions. The two minerals were analysed by Mr. F. T. Seeyle, of the Dominion Laboratory. Diopsidic augite: The refractive indices were determined in sodium light and the accuracy is considered to be ± 0.002. The optic axial angle and extinction angle were determined on the Federov universal stage, and for the latter determination Nemoto's method (1938) was used. Chemically the clinopyroxene is a diopsidie type low in sesquioxides, and if the latter are disregarded as recommended by Hess (1941, pp. 516–517) the mineral may be considered as a member of the wollastonite, clinoenstatite, clinoferrosilite, ternary system. Calculated as such, the pyroxene has the approximate composition of Wo 49, En 37.7, Fs 13.3 (weight per cent.) and lies close to diopside, the mid-point of the CaSiO — MgSiO3 join. When plotted on Hess's diagram (1941, p. 518) the pyroxene falls into the field of salite. Plotting a pyroxene of this composition on the diagrams prepared by Deer and Wager (1938, p. 20) and Wager and Deer (1939, p. 80) a satisfactory comparison of the observed optical constants is noted.

Table I.—Analyses of Diopsidic Augites. A. B. C. D. E. SiO2 49.57 50.39 51.05 51.05 50.10 Al2O3 3.82 3.54 5.23 1.80 4.57 Fe2O4 2.00 1.95 0.00 2.03 2.34 FeO 6.59 8.47 7.35 6.56 7.14 MgO 13.75 15.82 14.18 13.82 14.20 CaO 21.44 18.41 19.10 22.06 20.18 Na2O 0.69 0.70 0.39 0.38 0.32 K2O + 0.08 0.14 0.07 0.08 tr. H2O + 0.10 0.11 0.50 0.17 0.24 H2O - 0.02 0.04 0.50 0.04 0.10 TiO2 2.05 0.87 0.50 0.36 0.70 P2O5 nt. fd. tr. 0.46 MnO 0.13 0.19 0.25 1.22 0.32 NiO 0.04 V2O3 0.04 SrO 0.006 100.33 100.63 100.02 100.03 100.21 α 1.690 1.684 1.684 1.684–1.700 β 1.697 1.692 1.691 1.692–1.708 γ 1.715 1.712 1.712 1.707–1.721 γ–α 0.025 0.028 0.028 0.020–0.021 Pleochroism very weak weak 2± 32–54° 46° 58° 60° 50° 41′–57° 55′ Z - e 42° 39° 40° 41° 40°–43° Dispersion r > v r > v Sp. Gr. 3.34 3.35 3.370 A. Diopsidic augite from teschenite, P.7616, Red Hill, Rewa S.D., Wellington Land District. Anal.: F. T. Seelye. B. Augite from Ivnarmiut, East Greenland, Wager and Deer (1939, Table 19, No. 6, p. 152). Anal.: W. A. Deer. C. Augite from Co. Antrim, Ireland, Harris (1937, P.102). Anal.: W. H. Herdsman. Harris's total is incorrect. D. Diopside from phenocrysts of Treasure Mountain quartz latite, 5 miles S.E. of Del Norte, Colorada, Larsen, Irving, Gonyer and Larsen (1936, Table 2, No. 1, p. 695). Anal.: F. A. Gonyer. E. Augite from Yoneyama, Etigo, Japan, Kuno and Sawatari (1934). Anal.: S. Tanaka. XY (Si,Al)2 (O,OH,F)6 has been suggested by Machatschki (1929) as the general formula for monoclinic pyroxenes, and a recalculation of analysis A in Table I on this basis is presented in Table II. As will be seen the requirements of this formula appear to be completely fulfilled. Most of the aluminium is required in the four co-ordinated position in order to satisfy the silicon chains of the pyroxene structure, while 0.0158 aluminium ions remain in six-fold co-ordination. The vexed question of the correct allocation of the titanium does not, therefore, arise, as this ion is included in the Y group of Machatschki's formula.

Table 2.—Calculation of the Formula of Clinopyroxene. Wt. per cent. Mol. ratios. No. of metal atoms on the basis of 6 (O,OH). SiO2 49.57 0.8253 1.8480 2.00 Al2O3 3.82 0.0375 0.1678 0.1520 0.0158 TiO2 2.05 0.0257 0.0575 Fe2O3 2.00 0.0125 0.0559 FeO 6.59 0.0917 0.2052 MnO 0.13 0.0018 0.0040 2.01 MgO 13.75 0.3410 0.7633 CaO 21.44 0.3823 0.8337 Na2O 0.69 0.0111 0.0497 K2O 0.08 0.0009 0.0040 The formula is thus:— (Mg,Fe”, Fe”, Ca,Mn,Na,K,Ti,Al)2 [(Si,Al)2 O6] Iron-ore: Table 3.—Analyses of Titaniferocs Iron-ores. A. B. FeO 32.30 34.9 Fe2O3 21.93 18.6 TiO2 38.33 46.1 Al2O3 0.70 MgO 3.03 MnO 0.49 NiO 0.03 V2O3 0.23 SiO2 2.63 0.7 CaO nt. fd. H2O at 103° C. 0.25 99.92 100.3 A. Titaniferous ore from teschenite, P.7616, Red Hill, Rewa S.D., Wellington Land District. Anal.: F. T. Seelye. B. Fine-grained ilmenite-haematite intergrowth, containing blebs of inter-grown rutile and titaniferous magnetite, Weld Range, Murchison, Western Australia, Edwards (1938, p. 41). Anal.: E. S. Simpson. The ore in the teschenite occurs in abundant idiomorphic grains, and because the analysis of the rock itself showed a high figure for TiO2 the ore was separated and analysed. An examination of a polished surface of teschenite revealed that the iron-ore is composite in nature, probably consisting of ilmenite set in a base of haematite. As pointed out by Edwards (1938, p. 43), this type of intergrowth would result from a gradual unmixing if cooling was slow enough. This unmixing would bring about the formation of two solid solutions, one rich in Fe2O3, and the other in FeTiO3. Actually the solid solution, rich in Fe2O3, would, according to Ramdohr (1926, p. 357) contain just under 10 per cent. of TiO2 if homogeneous. The MnO and MgO in the ore will presumably take the place of ferrous iron. Fibrous zeolites. Thomsonite, with very minor natrolite, is well developed in the teschenites (P. 3230, 3231, 7616) as sheaf-like

aggregates of sharply acicular crystals. Thomsonite is colourless, with straight extinction and elongation of the needles parallel to β; refractive indices have been determined as follows:— α = 1.513 ±0.002 β =1.516 γ=1.522 γ — α=0.009 2V = 68° ± 2°. Therefore, according to curves drawn up by Winchell (1933, p. 388), the mineral should contain approximately 37–38 per cent. of the Na3Ca9Al19O80.24H2O molecule; or should give an analysis somewhere intermediate between Hey's analyses Nos. 1 and 2 (1932, p. 54, Table I). Intimately associated with the radiating aggregates of thomsonite is a very minor amount of a zeolite with fibres positively elongated; birefringence is about 0.012 and γ = 1.485. The mineral is believed to be natrolite. Microchemical tests do not indicate any more than traces of K2O in these zeolites, hence the thomsonite differs from that investigated in a somewhat similar rock from Waihola, East Otago (Hutton, 1942, pp. 181–182). Amygdule mineral. In a group of variolitic pillow-lavas a green micaceous mineral forms amygdaloidal infillings and also some of the base or mesostasis. It is very minutely crystalline and occurs in aggregates of flakes, while the entire infillings occasionally measure as much as 2 mm. The colour varies from a greenish blue, through glaucous green to a rusty brown tint. Commonly a zonary structure is observed in the vesicles, usually with the olive-green material lining the walls of the amygdules, this material giving place in the centre to the greenish-blue phase. The brown material occurs in any position relative to the green varieties. Refractive index tests show that the composition must vary considerably, for values from 1.588–1.615 were determined for γ. Data determined on a single platelet are as follows:— α = 1.590 γ = 1.610 Elongation: positive α = pale green β=γ = greenish blue Qualitative microchemical tests show that (OH), Fe″, and Mg are the chief constituents; Al2O3, Fe″′, and K2O appear to be of minor importance only. Although the optical properties of the mineral are somewhat similar to those of bowlingite, the comparison ends there, for bowlingite appears to have much higher birefringence, and, furthermore, it is usually a product of alteration of olivine. It is believed that the amygdule mineral is a product of crystallisation of a ferruginous gel and that it does not represent direct decomposition of ferromagnesian minerals. Actually the mineral is comparable to Jolliffe's mineral X (1935, pp. 411–412), shown by analysis

(loc. cit. p. 416) to be a hydrated silicate of magnesium and iron, with most of the iron in the ferrous condition. On the other hand, the colour, pleochroism, and mode of occurrence are suggestive of celadonite, but the refractive indices are considerably lower than those recorded by Hendricks and Ross (1941, p. 687) and Winchell (1933, p. 436) for this mineral. Microchemical tests indicated some K2O, but whether it is present in sufficient amount to warrant classification as celadonite is by no means clear; certainly in some of the analyses quoted by Hendricks and Ross (loc. cit. p. 697) K2O is moderately low. In one case, however (P. 4888), greenish-brown biotite was observed filling the centre of one vesicle and merging into the green lining, but it is not thought that this indicates that biotite has developed from the green mineral. Therefore, until more work can be carried out on this problem, the material will be referred to as the amygdule mineral. Petrography. The igneous rocks of the Brocken Range-Ngahape area, as have been described by Brown (1943), occur in three distinct localities, and the rocks themselves are capable of subdivision into three petrological groups—viz.: A. Teschenites (Brocken Range). B. Altered olivine dolerites and dolerites with spilitic and teschenitic affinities (Kaiwhata River). C. Variolites (Ngahape). A. Teschenites (text fig. 1; plate 39, fig. 1). In the hand-specimen these rocks are melanocratic, fairly coarse, equigranular rocks, weathering to a reddish-brown soil. Microscopically the rock has a hypidiomorphic granular structure, though in parts where feldspar is completely replaced by zeolite a pseudo-porphyritic texture is developed. Plagioclase occurs in lath-like and tabular crystals up to 2.0 mm. in length. Twinning on Carlsbad, albite and pericline laws is general and zoning is most pronounced. The composition ranges from An67 in the interior to An45 in the most acid peripheral zones. The feldspar is heavily loaded with inclusions of apatite, biotite, and iron ore, and usually has a blotched appearance due to extensive replacement by thomsonite and minor natrolite. The pyroxene, a diopsidic - type moderately rich in TiO2, has crystallised in idiomorphic to sub-idiomorphic grains averaging 0.75–1.0 mm. in diameter. In thick grains the colour is pale pink, with the pleochroism faint but distinct. In the material separated for analysis the optical axial angle is 50°–54°, but in the large crystals in the thin slices strong zoning is developed, sometimes with poor hour-glass structure, consequently the angle 2V is not constant. In the narrow outer zones the angle was found to vary from 42°–46°, while the refractive index in one case rose from 1.715 to approximately 1.725. This general decrease in the optic axial angle and increase in refractive index in the outer zones of phenocrystic pyroxenes has been noted by Barth (1931), Kuno (1936)

and in New Zealand by Benson and Turner (1939, p. 68). These workers interpret these data as showing an increase in the clinoferrosilite molecule and concomitant decrease in the diopside molecule as crystallization advanced; that is, an enrichment of the later differentiates of the magma in iron. Figs. 1A, 1B, and 1C Teschenites. A. The normal mineral assemblage of altered feldspar, zeolites, pyroxene, iron-ore, and accessories (P. 7616). × 24. B. Analcite-rich portion showing replacement of feldspar and fibrous zeolites by analcite (P. 3230). × 24. C. Insertal areas of fibrous zeolites due to crystallization of late magmatic liquids (P.3231). × 24.

Irregularly shaped platy aggregates of bowlingite were noted. Although in some cases it appears to be partially rimming augite, more often the aggregates suggest the shape of former olivine crystals, after which it is possibly pseudomorphous. Occasionally it penetrates along the cleavage planes of the zeolite-flecked plagioclase crystals (P. 3231). The colour varies from a bluish green to bright green, and pleochroism is fairly strong according to the scheme: X = yellow, Y = Z = deep olive green, with absorption X < Y = Z. Some minor grains of ferriferous epidote are often associated with bowlingite. Sheaf-like radiating aggregates of thomsonite are abundantly developed throughout these rocks in the interspaces between the clinopyroxene and plagioclase, and as irregular patches within the feldspar itself (text Figs. 1A and B). Minor natrolite is closely associated with the thomsonite in the aggregates. Intensely pleochroic scraps of biotite occur, particularly in P. 3231, closely associated with grains of iron ore, although sometimes appearing to form a partial rim to clinopyroxene (text Fig. 1A). Much of the mica is altered to a pale green negative chlorite. The biotite appears to have crystallised at a period intermediate between the crystallisation of pyroxene and crystallisation of the zeolites. Some sphene, possibly a by-product of the biotite ± chlorite reaction, was noted. Analcime (n = 1.482) is most plentiful in some thin slices, and it is usually filled with the dust-like inclusions seen in the plagioclase; it has replaced much of the latter material. No thomsonite is visible in these “pools” of analcite, although the fibrous zeolite is abundant as flecks in the unaltered plagioclase. It is suggested that this points to the possible replacement of thomsonite as well as the feldspar by analcime. It might be contended, however, that on account of the absence of fibrous zeolite within areas of completely analcitized plagioclase, the crystallisation of the thomsonite and natrolite post-dated that of analcime. There is, though, no evidence in support of this contention. Apatite, in needles and stout hexagonal prisms up to 1.0 mm. in length, is a constant and abundant accessory, and it appears to be concentrated to some extent in the intersertal areas occupied mainly by zeolites and biotite (text Fig. 1C). The hexagonal prisms appear to be free from inclusions. Ilmenite-haematite solid solution is abundant in idiomorphic to sub-idiomorphic grains, usually octahedra, averaging 0.3 mm. in diameter, but occasional highly irregular grains, 2.0 mm. in diameter, are met with. Partial rimming with biotite is common, and ovoid areas of biotite, augite, bowlingite and zeolites occur as inclusions. These rocks are fairly typical of melanocratic teschenites, and their chemical composition is similar to that of the original teschenite from Silesia, although the New Zealand rocks lack barkevikite (compare analyses A and B, Table 4). The percentage of TiO2 is rather unusually high in the rock from Red Hill, and this is due to the important amount of iron ore present.

Table 4.—Analyses of Teschenites. A. B. C. D. SiO2 41.06 41.42 45.08 45.52 Al2O3 12.75 15.07 12.44 16.08 Fe2O3 2.24 7.93 3.25 4.18 FeO 11.48 6.95 6.37 MgO 6.53 4.82 8.56 4.85 CaO 9.55 10.16 10.43 8.34 Na2O 4.22 4.00 3.53 4.63 K2O 0.80 1.98 0.80 2.09 H2O + 3.78 2.73 2.97 4.92 H2O - 0.40 0.27 1.94 CO2 trace 0.15 TiO2 6.35 3.14 2.56 2.07 P2O5 0.39 1.57 0.52 ZrO2 nt.fd. nt.fd. S 0.09 0.37 0.10 MnO 0.12 0.20 0.29 NiO 0.05 0.03 Cr2O3 nt.fd. 0.04 V2O3 0.08 nt.dt. BaO 0.03 nt.fd. SrO 0.03 nt.fd. present Cl nt.fd. 0.05 nt.dt. F 0.04 0.10 nt.dt. 99.99 93.81*Total includes FeS 0.06, FeS2 0.04. (Johannsen, 1938, p. 229). Johannsen's total is incorrect; possibly FeO was omitted. 100.24 100.00†Total includes 0.95% of minor constituents. A. Teschenite, P.7616, from Red Hill, head of Upokongaruru Stream, west side of Brocken Range, Rewa S.D. Anal.: F. T. Seelye. B. Teschenite, Paskau, 25 km. west of Teschen, Mahren, Silesia. C. Teschenite, Warman's Point, P.6458, Mangapai, Ruakaka S.D. Anal.: F. T. Seelye. D. Teschenite, average of twelve analyses, quoted from R. A. Daly, 1933, p. 22. In many ways the Red Hill teschenites are comparable with rocks described by Benson (1942) from Eastern Otago, and especially the rock investigated by Bartrum (1925, p. 10) and Ferrar (1934, pp. 55–56) from Mangapai Estuary, Whangarei Harbour (Analysis C, Table 4). However, the Red Hill teschenites differ from these in that there is a complete absence of aegirine–augite mantling the titan–augite, or occurring as individual crystals in the mesostasis; a further distinction is the apparent absence of fibrous zeolites from the Mangapai rock. The Mangapai and Red Hill rocks appear to be remarkably low in K2O as compared with the average of twelve teschenites. (Analysis D, Table 4), and both rocks show development of leucocratic, possibly late magmatic, areas (c.f. Text Fig. 1C and Plate 39, Fig. 2). B. Dolerites (text Figs. 2A and B). Altered olivine dolerites: Macroscopically these rocks, represented by only two specimens, are dense, dark and fine-grained. In thin section they are seen to be somewhat altered, but holocrystalline with a granulitic structure. The plagioclase occurs in

lath-like and tabular crystals, often up to 2.5 mm. in length, showing single and multiple twinning. Zoning is quite pronounced, with the composition determined by universal stage methods, varying from An53–An58 centrally to An45 in some narrow peripheral zones. Inclusions of augite and fine-grained groundmass are common and are frequently concentrated at the centre of the plagioclase crystals. Figs. 2A, 2B, 2C. A. Quartz-albite dolerite with much bowlingite (P. 7601). × 24. B. Altered olivine dolerite with interstitial patches of finely crystalline, almost variolitic material (P.7602). × 24. C. Variolite (P.7608). × 24.

Clinopyroxene appears to be restricted to one generation, and occurs in granules averaging 0.3–0.4 mm. It is a pale purple or pink titan-augite with 2V = 51°–53° and Z±c = 43°. According to Wager and Deer's diagram (1939, p. 80), the composition of the pyroxene should be about CaSiO3 35, MgSiO3 45, and FeSiO3 20. Bowlingite occurs in fibrous aggregates that appear to simulate the outlines of olivine crystals, but whether all the bowlingite is pseudomorphous after original olivine is by no means certain. The colour is pale yellow for α and deep yellow or greenish-brown for γ. The borders of the bowlingite aggregates are often outlined with brown granular limonite, possibly due to decomposition of the mineral itself. This material is referred to bowlingite rather than to iddingsite on account of the pleochroism and refractive. index. Ross and Shannon (1926, pp. 13–14) show that the refractive index for α is 1.70 or higher and that a ruby-red tint is common for the γ vibration direction in iddingsite. However, the refractive index of the bowlingite is lower than 1.70 and the characteristic ruby-red tint is lacking. Ilmenite is an important accessory constituent occurring in rather ragged bars up to 0.6 mm. in length, in ragged grains and in dendritic forms. The bars or rods are very often parallel among themselves in a fashion similar to that described by Campbell, Day and Stenhouse (1932, p. 353) for iron-ore in segregation veins in some teschenites (text Fig. 2B). Needles of apatite and some calcite are present. Throughout the rock are numerous interstital patches of finely crystalline material that appears to have been formed from a late, more acid residual liquid (text Fig. 2B). They consist of augite, iron-ore, and bowlingite, set in a plexus of slender laths of plagioclase; the feldspar crystals are too fine for precise determination, but they usually show multiple twinning. In the feldspar laths inclusions of extremely narrow, longitudinal zones of cryptocrystalline material clouded with magnetite dust occur. Albite dolerites: In the hand-specimen albite dolerites are fairly coarse-grained, crumbling rocks, sometimes with long laths of feldspar up to 25 mm. in length, and sometimes large black crystals of pyroxene. In colour they vary from grey or buff to a glauconite-green. Microscopically they are all considerably altered. They are holocrystalline and often exhibit an intersertal or sub-ophitic texture with coarse idiomorphic crystals of plagioclase that are generally dusty in appearance, though the margins may be clear. The mesostasis between these crystals is varied and may consist of a mosaic of feldspar and quartz, or feldspar, quartz, and bowlingite; the quartz and feldspar do not show any micrographic relationship as has been described by Benson (1915, p. 142), although the association is an intimate one. The feldspar is in excess of pyroxene and occurs in sub-idiomorphic plate-like crystals and broad laths, universally twinned on Carlsbad, albite and pericline laws. The plagioclase crystals are slightly zoned and in some cases (P. 7615) have been affected by shearing with consequent cracking, bending or granulation. Inclusions

are often dense and are arranged centrally, leaving the periphery quite clear. The inclusions noted were clinozoisite, iron-poor epidote, and apatite, with a dusting of sericite and kaolin in most cases. Bowlingite has intimately penetrated the cracks and cleavage planes of plagioclase crystals. Augite occurs in two rocks (P. 7607, 7615) but is an important constituent of only one (P. 7607). It occurs in large, very pale pink grains up to 2.5 mm. in diameter, sometimes sharply idiomorphic, but with very faint pleochroism. The clinopyroxene is usually considerably altered to leucoxenized ilmenite and chloritic or bowlingitic alteration products. In places a tendency towards an ophitic relationship with feldspar is to be seen. Quartz occurs in small interstitial water-clear allotriomorphic grains, although a few large grains up to 1.0 mm. in diameter occur in one of the rocks (P. 7601, text Fig. 2A), while in P. 7606 rare clear zones of quartz were noticed surrounding dusty, somewhat altered acid oligoclase. One of the rocks shows transition to the micro-teschenites (Hatch and Wells, 1937, p. 226) with the development of plentiful analcime. This constituent extensively replaces the plagioclase, for fragments of feldspar occur floating about in analcime, while in other places the zeolite is crossed by lines of bowlingite, the latter mineral formerly occurring as a filling to cleavage cracks of the replaced feldspar. Bowlingite is a common constituent and has possibly been derived from olivine as well as from the clinopyroxene, the latter often occurs as relict granules in pools of bowlingite. It is usually interstitial and, sometimes microspherulitic in habit. It is uniaxial negative and distinctly pleochroic with X = colourless to pale yellow and Y = Z = green to greenish-brown. Acicular hexagonal prisms of apatite, up to 2.5 mm. in length in some cases (P. 7607), is a common accessory constituent. In one example (P. 7601) the prisms have a central, narrow zone of inclusions parallel to the C-axis. They are pale yellow in colour and have a refractive index slightly less than that of the apatite. It was not possible to identity the material, but it is suggested that it is an hydroxyapatite surrounded by normal apatite. Iron-ore occurs in bars and skeletal crystals and in subidiomorphic grains; knee-shaped twins have been noted (P. 7601). A positive titanium reaction was generally obtained on the separated ore, but the paramagnetic properties were stronger in some grains than in others; probably titaniferous magnetite and ilmenite are both present. Leucoxene was noted in some cases closely associated with the ore. Minor calcite was observed in the three rocks of this group, although no CO2 is recorded in the analysis of one rock (Table 5, analysis C.). C. Variolites (text Fig. 2C.). This group of rocks is represented by a number of specimens (P. 4886–4893, P. 7608). In the hand specimen the variolites are dark green to black massive fine-grained rocks with amygdules filled with calcite, and patches of black, shining micaceous minerals are very conspicuous in some specimens.

Fig. 3. Molecular percentages of normative feldspars of Brocken-Kaiwhata area rocks, compared with average spilite feldspar. 1 = anal. B, 2 = anal. E, 3 = anal. C, 4 = anal. A, 5 = anal. G (all of Table 5); 6 = Red Hill teschenite (normative nepheline calculated as albite); 7 = average spilite. Microscopically the rocks consist of radiating aggregates of acicular crystals of feldspar, bars and dendritic forms of leucoxenized ilmenite and slender blade-like crystals and granules of augite. There is an approach to the cervicorn structure so characteristically seen in certain of the cone-sheets of Mull, Thomas and Bailey, 1924, p. 303) and Ardnamurchan (Thomas, 1930, p. 171). These radiating aggregates do not exceed 0.75 mm. in diameter. Occasional minute tabular crystals of plagioclase occur with a central aggregate of inclusions and a clear peripheral zone. Owing to the very fine grain-size it is not possible to determine the exact composition of the plagioclase, except that the α refractive index direction is slightly greater than that of Canada balsam. The composition is therefore about acid to medium andesine, and this is supported by the feldspar of the norm calculation (Table 5, analysis G.). The rounded grains of pale purple augite average 0.75 mm. in diameter; occasional glomeroporphyritic groups of grains were noted. The amygdule mineral previously described in the section on mineralogy is fairly plentiful. Under low power it has a gel-like, glauconitic appearance, but with greater magnification, it is observed to be built up of aggregates of minute, positively elongated flakes. In one amygdule lined with the green mineral, a central zone is composed entirely of greenish-brown biotite in flakes averaging 0.15–

Fig. 1 (Upper). The normal association in Red Hill teschenite (P.3230). × 70. Fig. 2 (Lower). Leucocratic phase of the Mangapai teschenite (P.6458). showing partially analcitized plagioclase and plentiful biotite and apatite. × 70.

0.2 mm. in length. These flakes have a sheaf-like or radiating arrangement, and owing to the intense pleochroism good black crosses can be observed. Calcite is also plentiful in most of these rocks as a filling in vesicles, and it may occur with or without the green constituent. In some vesicles patches and ragged wisps of the green mineral occur “floating” in calcite, and this relationship suggests that the calcite is of later origin and that it has replaced the amygdule mineral. Occasionally a pale yellow glassy or cryptocrystalline material accompaies the amygdule mineral. It has not been possible to identify this, but under a high power magnification numerous radiating bundles of needle-like colourless crystals were noted; these needles were negatively elongated. This cryptocrystalline material might possibly be a zeolitic gel. Petrogenesis. On Table 5 the analysis of the dolerites and variolites are compared with similar rocks. The analysis of the albite-dolerites (B and C) are like that of a dolerite (D) described by Benson (1915) and associated in the field with spilites, while the albite dolerite (E) is very similar in composition to an albitic quartz dolerite (F) also described by Benson (1915). The variolite (G), on the other hand, can be compared chemically with the more normal basalt or diabase (H). However, the features that are of the greatest interest in all of these rocks from the Brocken-Kaiwhata area are: (1) poverty in K2O; (2) richness in Na2O; and (3) the fairly low figure for Al2O3. These characters at once point to a possible connection with the rocks of the so-called spilitic suite of spilites and keratophyres. In the case of the three albitic dolerites (Table 5, B, C, E) the normative plagioclase is Ab91, Ab76 and Ab88 respectively, but these figures do not completely support the evidence obtained from the feldspar determination by universal stage methods. This work showed that the plagioclase, though zoned, was never more basic than acid oligoclase (Ab88An12), thus the excess CaO must be present in the clinopyroxene. Also some clinozoisite or epidote is present, and the Al2O3 and CaO required for these minerals in the mode would be calculated as anorthite in the norm. The relative poverty in K2O pointed out by Sundius (1930, p. 9) as a special characteristic of the spilitic series is. for the Brocken-Kaiwhata rocks, clearly shown in the ternary diagram, Or–Ab–An (text fig. 3). With regard to the Na2O content, B. and E. are more sodic and C. very slightly less sodic than the averaged normative feldspar (point 7) of a number of spilites. The feldspars of the altered dolerites and variolites, A. and G. respectively, are on the other hand richer in CaO, though still poor in potash. They lie on the sodic side of the diabase group that Sundius (1930. p. 9) believes might possibly be separated from the field of spilitic feldspars, although Gilluly (1935A, p. 249) points out that there appears to be a complete gradation from spilitic to normal subalkaline rocks. From the Kaiwhata sill itself, altered olivine (?) dolerites, quartz-albite dolerites, albite dolerites; and analcime-bearing albite dolerites have been described, and this association suggests the possibility that after the magma was injected into the sill slight gravitative differentiation took place. Unfortunately, this was not realised until examination of the rocks was under way in the laboratory, and

then it was not possible to return to the field to verify the truth or otherwise of this hypothesis. The writer submits that all the dolerites, teschenites, and variolitic pillow lavas are closely associated magmatically and that they were emplaced about the same time. One cannot but be impressed with the chemical similarity of the teschenite, dolerite and variolite, and there would seem little doubt of their derivation from a single magma. It is suggested that the teschenite represents a differentiate of a fairly normal basaltic magma in place, somewhat similar to that described by Campbell, Day and Stenhouse (1934) in the case of the Braefoot Outer Sill in Fifeshire. With the intrusion of the teschenite at depth, a portion of the doleritic material was injected into the now westerly dipping Raukumara sediments to form the Kaiwhata sill. Slight differentiation of the doleritic magma in this sill may have occurred causing an enrichment of the residual magma in soda and water, resulting in crystallization of analcime and albite. At the same time as the Kaiwhata sill was emplaced, a portion of the magma was poured out on to the wet sediments of the Upper Cretaceous sea floor to form the variolitic pillow-lavas. In the teschenite itself the residual liquids after crystallization of the labradorite and diopsidic augite, appear to have been enriched in soda and water, resulting in the late crystallization of fibrous zeolites, followed by analcime, which has partially replaced plagioclase and earlier zeolites. This late crystallization of analcime after the other zeolites does not appear to be common, although Gilluly (1927, p. 204) states that thomsonite and analcime have crystallized simultaneously, followed by natrolite, in a diabase from Utah. Similarly in the Watchung basalts, Fenner (1911) has shown that one zeolite may crystallize at the expense of another, due to falling temperature of late-magmatic solutions. In the Red Hill teschenite both natrolite and thomsonite appear to have crystallized at the same time, and it is believed, chiefly on account of relative abundance and relationship to other minerals that they are primary magmatic constituents and are not the result of secondary alteration of plagioclase. In the variolites an interesting association of minerals has been observed filling the vesicles, viz., the green “celadonitic” mineral, calcite, biotite, and some cryptocrystalline material, referred to before as a zeolitic gel. It is the writer's opinion that the “celadonitic” amygdule mineral has crystallized from later magmatic ferruginous gels or solutions, and this supposition is in line with the general theory put forward by Fenner (1929, p. 245) that there appears to be a concentration of iron and alkalies in liquids given off by basic magmas; following Hendricks and Ross (1941, p. 705) the mineral may be termed a metacolloid. Furthermore, Jolliffe (1935, p. 422) comes to the conclusion that the green iron-rich mineral, greenalite, of the Biwabik formation of Minnesota, owes its origin to ferrous solutions of direct igneous origin; possibly a reducing environment would be necessary. As stated previously, brown rims to the green amygdule mineral are common, and these are possibly due to (1) oxidation of the ferrous iron in the green mineral,; or (2) the result of difference in composition of the gel, perhaps higher Fe2O3, at the time of deposition. The rare occurrence of greenish biotite in association with, but of later origin than, the green mineral, is interesting, and appears

to indicate that the very latest of the solutions to crystallize were rich in potash as well as in ferrous iron (cf. Fenner, 1929, p. 245). The question of soda-metasomatism versus primary magmatic crystallization of albite in spilites and allied rocks has been discussed by many writers and summarised by Gilluly (1935A and B), who contends that ophitic structure is no sure criterion of primary origin of the albite in the rocks concerned. This seems to be a very good clue indeed, for it is difficult to understand how albite could crystallize on ophitic relationship with augite when the physical chemistry of the problem is considered. However, attention is now drawn to the following characters shown by the Brocken rocks:— (1) No trace of the mottling of feldspars such as described by Gilluly (1935A, 229) was observed in the Brocken rocks, and in this respect they are similar to the sodic plagioclases in the rocks described by Bartrum (1936), although the feldspars in the Brocken rocks do not have an unaltered appearance. (2) No adinolization appears to have occurred, although it must be admitted that a specific search for this feature was not carried out. However, thin sections of near-by greywackes show no signs of enrichment in albite. (3) There is not any sign of the form that the expelled anorthite molecule would have taken if soda metasomatism has taken place. (4) The pyroxene, when present, in the sodic rocks of the Brocken-Kaiwhata area is considerably altered, and also occurs as remnants set in a bowlingitic matrix. But little or no epidote, clinozoisite, or calcite occurs associated with the bowlingitic pseudomorphs of the pyroxene. (5) Analcime is present in some of the dolerites, while analcime and fibrous zeolites are plentifully developed in the teschenites. No positive conclusion as to the origin of albite in the Brocken rocks is warranted on the above evidence. At this stage in the discussion it may be appropriate to mention the lack of spilitic characters in the majority of the post-Jurassic demonstrably submarine pillow-lavas of New Zealand. As instances may be cited the normal basaltic pillow lavas of the extreme northwest corner of the North Island of New Zealand (Bartrum and Turner, 1928), of the coastal area west of Auckland City (Bartrum, 1930) of the Oxford District of Canterbury (Speight, 1928), of the Oamaru District, north Otago (Park, 1918), and of Matira Creek in the Huntly-Kawhia area (Henderson and Grange, 1926). Finally it should be noted that other doleritic rocks are known in the Wairarapa area, but these await investigation. It will be sufficient to say that analcime-bearing types have been identified by the writer among specimens from the Eketahuna Subdivision at present being investigated by Mr. M. Ongley, while dykes occur in Cretaceous rocks near Flat Point, south of the mouth of the Kaiwhata River. These latter rocks have been quaintly described by Sollas (1906, pp. 153–154) as hypersthene andesites, but both are altered augite dolerites, some with variolitic structure. Similar igneous rocks from the Tinui area in Wellington Province remain to be described.

Table 5.—Analyses of Dolerites and Variolites. A. B. C. D. E. F. G. H. SiO2 44.79 47.20 49.18 48.35 53.37 54.88 44.45 44.87 Al2O3 11.45 13.24 14.46 14.12 15.22 12.62 12.78 13.60 Fe2O3 0.6+ 5.37 5.29 4.87 4.23 3.02 4.98 5.65 FeO 9.06 7.83 7.04 10.27 4.89 7.11 4.91 5.68 MgO 5.03 7.12 4.03 4.78 2.22 3.73 6.82 9.39 CaO 12.29 3.08 5.84 6.71 5.43 4.16 9.53 11.05 Na2O 2.93 3.93 4.96 4.63 6.88 6.01 3.13 3.41 K2O 0.75 0.78 1.29 0.38 0.91 1.10 0.44 0.45 H2O + 0.98 4.76 2.67 2.00 1.97 1.76 1.98 2.05 H2O - 1.58 2.05 0.90 0.30 0.90 0.23 5.36 1.48 CO2 6.61 0.34 nt.fd. nil. 1.38 trace. 1.40 TiO2 2.91 2.57 3.22 2.84 1.82 3.63 3.17 2.13 P2O5 0.73 1.49 0.88 0.35 0.70 0.44 0.72 ZrO2 nt.fd. nt.fd. nt.fd. nt.fd. nt.fd. S. 0.09 0.03 0.07 0.22(FeS) 0.02 0.71(FeS) 0.07 MnO 0.14 0.14 0.19 0.18 0.10 0.25 0.13 NiO 0.03 nt.fd. nt.fd. nt.fd. 0.05 0.03 Cr2O3 0.08 nt.fd. nt.fd. nt.fd. nt.fd. 0.04 V2O3 0.02 0.02 0.03 0.01 0.03 BaO 0.03 0.11 0.08 0.10 0.05 nt.fd. 0.02 SrO 0.05 0.01 0.06 0.025 0.10 Cl. nt.fd. trace? trace? trace? trace? F. 0.065 0.11 0.068 0.08 0.084 100.26 100.18 100.28 100.10 100.21 99.70 100.17 99.76 O. for F. 0.05 0.03 0.03 0.04 Total 100.26 100.13 100.25 100.10 100.18 99.70 100.13 99.76 Norms. A. B. C. D. G. H. Q. 5.82 6.58 0.68 2.85 Or. 4.45 4.62 7.63 5.39 2.64 2.78 Ab. 24.80 33.24 41.94 58.20 26.48 19.39 An. 14.46 3.40 13.41 7.96 19.53 20.29 Ne 5.11 C. 0.51 4.69 di CaSiO4 4.09 2.38 6.03 27.26 MgSiO3 2.77 1.51 5.21 FeSiO3 1.00 0.73 hy MgSiO3 12.53 17.73 6.81 4.03 11.78 FeSiO3 11.65 5.97 2.48 1.94 ol Mg2SiO4 0.36 9.08 Fe2SiO4 0.14 mt. 0.79 7.78 7.53 6.14 7.07 8.35 il. 5.52 4.89 6.12 3.46 6.03 4.10 hm. 0.02 Ap. 1.71 3.53 2.09 1.65 1.71 pr. 0.17 0.13 0.13 (cc) (15.02) (0.77) (3.14) (3.18) Normative plagioclase Ab03 Ab01 Ab70 Ab88 Ab58 A. Altered olivine dolerite, P. 7602. From sill crossing Kaiwhata Stream 20chs. upstream from Ngahape, Rewa S.D. Analyst: F. T. Seelye. B. Quartz-albite dolerite, P. 7601. Locality as for P. 7602. Analyst: F. T. Seelye. C. Albite dolerite, P. 7607. Locality as for P. 7602. Analyst: F. T. Seelye D. Dolerite, Munro's Creek, Bowling Alley Point, Nundle District, N.S.W. (Benson, 1915, p. 139, analysis 1002). E. Albite dolerite, P. 7615. Locality as for P. 7602. Analyst: F. T. Seelye. F. Quartz dolerite, Hanging Rock, Nundle District, N.S.W. (Benson, 1915, p. 139, analysis 145). G. Variolite, P. 7608, Ngahape “Bluff,” opposite junction of Waipapa Creek and Kaiwhata Stream. Analyst: F. T. Seelye. H. Basalt. Wiebel, near Steinbach, Giessen, Hesse. Analyst: A. Streng (Washington, 1917, p. 632, analysis 10). The total recorded in Washington is incorrect.

Acknowledgment. The author wishes to thank Mr. A. E. Leopard for his assistance in the laboratory. Literature Cited. Barth, T. F. W., 1931. Petrography of Pacific Lavas. Amer. Jour. Sci., vol. 21, pp. 377–405, 491–530. Bartrum, J. A., 1925. The Igneous Rocks of North Auckland, New Zealand. Gedenboek Verbeek. Verhandl, v.L. Geol. Minjb. Genootschap voor Ned. en Kolonien, Geol. Serie VIII, pp. 1–16, s'Gravenhage. —– 1930. Pillow-Lavas and Columnar Fan-Structures at Muriwai, Auckland New Zealand. Jour. Geol., vol. 38, pp. 447–455. —– 1936. Spilitic Rocks in New Zealand. Geol. Mag., vol. 73, No. 867, pp. 414–423. —– and Turner, F. J., 1928. Pillow-Lavas, Peridotites, and Associated Rocks of Northernmost New Zealand. Trans. N.Z. Inst., vol. 59, pp. 98–138. Benson, W. N., 1915. The Geology and Petrology of the Great Serpentinebelt of New South Wales, Pt. IV. The Dolerites, Spilites, and Keratophyres of the Nundle District. Proc. Linnean Soc. of N.S.W., vol. 40, Pt. 1, pp. 121–173. —– 1942. The Basic Igneous Rocks of Eastern Otago and their Tectonic Environment, Pt. 2. Trans. Roy. Soc. N.Z., vol. 72, pp. 85–118. —– and Turner, F. J., 1939. Mineralogical Notes from the University of Otago, New Zealand, No. 2. Trans. Roy. Soc. N.Z., vol. 69, Pt. 1, pp. 56–72. Brown, D. A., 1943. Notes on the Geology of the Brocken Range and the Kaiwhata Valley, East Wellington. Trans. Roy. Soc. N.Z., vol. 72, pp. 347–352. Campbell, R., Day, T. C., and Stenhouse, A. G., 1932. The Braefoot Outer Sill, Fife. Pt. 1. Trans. Edin. Geol. Soc., vol. 12, Pt. 4, pp. 342–375. —– 1934. The Baefoot Outer Sill, Fife. Pt. 2. Trans. Edin. Geol. Soc., vol. 13, Pt. 1, pp. 148–173. Daly, R. A., 1933. Igneous Rocks and the Depths of the Earth. New York. Deer, W. A., and Wager, L. R., 1938. Two New Pyroxenes included in the System Clinoenstatite, Clinoferrosilite, Diopside, and Hedenbergite. Mineral Mag., vol. 25, No. 160, pp. 15–22. Edwards, A. B., 1938. Some Ilmenite Microstructures and their Interpretation. Proc. Austr. Inst. Min. Metall., New Series, No. 110, pp. 39–58. Fenner, C. N., 1911. The Watchung Basalt and the Paragenesis of its Zeolites and Other Secondary Minerals. Ann. N.Y. Acad. Sci., vol. 20, No. 2. Pt. 2, pp. 93–187. —– 1929. The Crystallization of Basalts. Amer. Journ. Sci., Vol. 18, 5th Ser., pp. 223–253. Ferrar, H. T. The Geology of the Dargaville-Rodney Subdivision. N.Z. Geol. Surv. Bull., No. 34 (New Series). Gilluly, J., 1927. Analcite Diabase and Related Alkaline Syenite from Utah. Amer. Journ. Sci., vol. 14, pp. 199–211. —– 1935A. Keratophyres of Eastern Oregon and the Spilite Problem, Part I. Amer. Jour. Sci., vol. 29, 5th Series, pp. 225–252. —– 1935B. Keratophyres of Eastern Oregon and the Spilite Problem. Part II. Amer. Jour. Sci., vol. 29, 5th Series, pp. 336–352. Harris, N., 1937. A Petrological Study of the Portrush Sill and its Veins. Proc. Roy. Irish Acad., vol. 43, sect. B, No. 9, pp. 95–134. Hatch, F. H., and Wells, A. K., 1937. The Petrology of the Igneous Rocks. George Allen and Unwin Ltd., London.

Henderson, J., and Grange, L. I., 1926. The Geology of the Huntly-Kawhia Subdivision. N.Z. Geol. Surv. Bull., No. 28 (N.S.). Hendricks, S. B., and Ross, C. S., 1941. Chemical Composition and Genesis of Glauconite and Celadonite. Amer. Min., vol. 26, pp. 683–708. Hess, H. H., 1941. Pyroxenes of Common Mafic Magmas, Part I, Amer. Mineral., vol. 26, pp. 515–535. Hey, M. Studies on the Zeolites, Pt. II, Thomsonite (including faroelite) and gonnardite. Mineral. Mag., vol. 23, pp. 51–125. Hutton, C. O., 1942. In Benson, W. N., The Olivine Theralite of Waihola, East Otago, a Gravitationally-differentiated Sill. Trans. Roy. Soc. N.Z., vol. 72, pp. 181–183. Johansen, A., 1938. A Descriptive Petrography of the Igneous Rocks. Vol. 4, The University of Chicago Press, Chicago. Jolliffe, F., 1935. A Study of Greenalite. Amer. Min., vol. 20, No. 6, pp. 405–425. Kuno, H., 1936. On the Crystallization of Pyroxenes from Rock Magmas with Special Reference to the Formation of Pigeonite. Jap. Jour. Geol. and Geogr., vol. 13, Nos. 1 and 2, pp. 141–150. —– and Sawatari, M., 1934. On the Augites from Wadaki, Idu, and from Yoneyyama, Etigo, Japan. Jap. Jour. Geol. and Geogr., vol. 11, pp. 327–343. Larsen, E. S., Irving, J., Gonyer, F. A., and Larson, E. S., 3rd, 1936. Petrologic Results of a Study of the Minerals from the Tertiary Volcanic Rocks of the San Juan Region, Colorado. Amer. Min., vol. 21, No. 11, pp. 679–701. Nemoto, T., 1938. A New Method of Determining the Extinction Angle of Monoclinic Minerals, Especially of Pyroxenes and Amphiboles, by Means of Random Sections. Jour. Fac. Sci. Hokkaidô Imperial Univ., Ser. 4, vol. 4, Nos. 1–2, pp. 107–112. Park, J., 1918. The Geology of the Oamaru District, North Otago. N.Z. Geol. Surv. Bull., No. 20 (N.S.). Ramsdohr, P., 1926. Beobachtungen an Magnetit, Ilmenite, Eisenglanz und Ueberlegungen ueber das System FeO, Fe2O3, TiO2. Neues Jahrb. f. Min. Geol. Pal., 54, Beil. Bd. Abt. A, pp. 320–379. Ross, C. S., and Shannon, E. V., 1925. The Origin, Occurrence, Composition, and Physical Properties of the Mineral Iddingsite. Proc. U.S. Nat. Mus., vol. 67, pp. 1–19. Sollas, W. J., 1906. The Rocks of Cape Colville Peninsula, Auckland, New Zealand. Government Printer, Wellington. Speight, R., 1928. The Geology of View Hill and Neighbourhood. Trans. N.Z. Inst., vol. 58, pp. 408–431. Sundius, N., 1930. On the Spilitic Rocks. Geol. Mag., vol. 67, no. 787, pp. 1–17. Thomas, H. H., 1930. The Geology of Ardnamurchan, North-west Mull and Coll. Mem. Geol. Surv. Scotland. —– and Bailey, E. B., 1924. Tertiary and Post-Tertiary Geology of Mull, Loch Aline, and Oban. Mem. Geol. Surv. Scotland. Wager, L. R., and Deer, W. A., 1939. Geological Investigations in East Greenland, Pt. III. Meddel. om Gronland, vol. 105, No. 4. Washington, H. S., 1917. Chemical Analyses of Igneous Rocks. U.S. Geol. Surv., Prof. Paper 99. Winchell, A. N., 1933. Elements of Optical Mineralogy, Pt. 2, John Wiley and Sons Inc., New York.