Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 41, 1908
This text is also available in PDF
(1 MB) Opens in new window
– 111 –

Art. XXI.—Geology of Signal Hill, Dunedin.

[Read before the Otago Institute, 10th November, 1908.]

General Geology.

The rocks of Signal Hill, being all volcanic, contain no internal evidence of their geological age. They form, however, a part of the Dunedin group of volcanics, which are known to overlie unconformably at Caversham and at Sea View a calcareous sandstone of the Oamaru system.* The Oamaru system was placed by Captain Hutton in 1875 in the Lower Miocene, but in his later works he classed it as Oligocene. Hector includes the Oamaru rocks in his Cretaceo-tertiary system; Park places them in the Miocene.§

The volcanic rocks are not in any case involved in the disturbances of the Oamaru rocks, nor have faults of any magnitude been observed in them. A considerable amount of elevation and depression has, however, since occurred, for the main valleys have been eroded far below the present base-level, and at Sandy Mount and elsewhere marine terraces are found at a considerable elevation.

Most observers have supposed the first volcanic outbursts at least to have occurred during the same period in which the Caversham sandstone was deposited. Professor Park, however, taking into account the erosion of the Caversham sandstone, suggests the beginning of the Pliocene as the first period of volcanic activity. There was either a prolonged period of volcanic activity, or there were two shorter periods separated by a period of quiescence, for conglomerate to a thickness of 100 ft., containing pebbles derived from the earlier volcanic rocks, lies in places upon the sedimentary rocks, and is covered by the later volcanic lavas. In places, also, fine shales with leaf-impressions occur above the sandstone and below the conglomerate. If, as suggested by Professor Park, these are of Pliocene age, the later eruptions which followed cannot be older than Pliocene. The occurrence of the conglomerate and the evidence of two periods of volcanic activity explains the unconformity between the earlier and later lavas of Signal Hill.

Physiography.

Signal Hill is a flat-topped elevation, extending as a ridge from Logan's Point in a north-easterly direction for a distance of three miles. The ridge then bends to the north and north-west, towards Mount Cargill. The greatest elevation of Signal Hill is 1,214 ft. Less than half a mile to the

[Footnote] * Marshall, “Geology of Dunedin,” Quart. Journ. Geol. Soc., lxii, 1906.

[Footnote] † Hutton, “Geology of Otago,” 1875; “Sketch of the Geology of New Zealand,” Quart. Journ. Geol. Soc., 1885; “Geological History of New Zealand,” Trans. N.Z. Inst., 1899.

[Footnote] ‡ Hector, “Outline of New Zealand Geology,” 1886.

[Footnote] § Park, Trans. N.Z. Inst., 1904.

[Footnote] ∥ Park, “On the Geology of North Head, Waikouaiti,” Trans. N.Z. Inst., 1903.

[Footnote] ∥ Park, “On the Geology of North Head, Waikouaiti,” Trans. N.Z. Inst., 1903.

– 112 –

south-east of the main peak is a steep rocky peak of about 1,000 ft.; and to the south-west of the main peak are several smaller peaks, all, however, with gentle slopes. Three-quarters of a mile along the ridge to the northeast of the main peak is a small elevation known as McGregor's Hill. On the south and east the hill slopes down to the shallow waters of Otago Harbour, and at the south-western end to an enclosed bay known as Lake Logan. The gullies on the eastern and southern sides are of slight depth, and are occupied by small streams.

A valley of considerable breadth and depth runs down to Lake Logan, and is occupied by the Opoho Stream, which rises in a swampy area near the summit of Signal Hill. On the north-west the ridge is bounded by the North-east Valley Stream. This stream always carries a considerable volume of water, and has eroded for itself a deep valley. The sequence of the rocks appears to indicate a great antiquity for this stream-valley. The stream has two branches, one of which rises on Mount Cargill, while the other has worked its head back in an easterly direction, and is cutting down the saddle to the north of McGregor's Hill. The upper part of this branch is still eroding its bed, but the lower part, and also the main stream, have reached base-level, and have filled the floor of the valley with alluvium. The North-east Valley Stream is fed by numerous small tributaries, the middle courses of which are generally through steep-sided gullies.

Occurrence of the Rocks.

The rocks of most widespread occurrence within the Signal Hill area are the phonolite of the Logan's Point type, the trachytoid phonolite of the Signal Hill type, and a basalt-flow which I describe as basalt No. 1. The basalt covers the top of Signal Hill, and extends across the upper part of the depression of the North-east Valley. The Logan's Point phonolite extends from near the summit of the hill to Logan's Point, forms a steep peak on the side of Signal Hill overlooking the harbour, and reaches to a point near Burke's. The Signal Hill phonolite has rather a wide occurrence about McGregor's Hill, and is found also in the North-east Valley.

Besides these there is a distinct variety of nephelinitoid phonolite, occurring probably as an intrusion, on the saddle to the north of McGregor's Hill; and there are several other basalts. The latter vary in composition from a very basic variety, occurring as a dyke on the Main North Road, which appears to be the youngest rock described, to a variety approaching andesite, which appears to be the oldest rock exposed. The least basic of the basalts, which I have called No. 5, clearly underlies the Logan's Point phonolite at a point on the Main North Road a quarter of a mile from Normanby. In the valley below this point it is not to be found, being obscured by slips of phonolite from above. Between the phonolite and the underlying basalt in the section exposed there is a stream-deposit, with boulders derived from the basalt, proving an intermediate period of erosion.

The Logan's Point trachytoid phonolite mass is of great thickness on the southern slope of Signal Hill, being continuous from below sea-level at Logan's Point to an elevation of 1,100 ft. near the summit. Towards the west, however, it thins out, until on the Main North Road there is a thickness of only 50 ft. between basalt No. 5 below and the main basaltflow (No. 1) above.

In the valley below, the Logan's Point phonolite gives place to the Signal Hill phonolite, but the junction is obscure. There is difficulty also in tracing the boundary between the two rocks on the eastern side. It is

– 113 –

however, practically certain, from sections exposed elsewhere—notably that at North Head, figured by Dr. Marshall*—that the oldest rock is the Logan's Point type, and that the Signal Hill type, which is closely related to the andesitic type of the North Head, rests upon it. The intervening basalt and basanite flows are not represented at Signal Hill, although a first inspection of the sections exposed along the Main North Road might lead to the belief that basalt intervenes between the flows of the two types of phonolite.

Going up the road from Normanby one passes successively Logan's Point phonolite, basalt, and Signal Hill trachytoid phonolite. A closer inspection shows that, while the basalt overlies the Logan's Point rock, it also overlies the Signal Hill type, and is continuous with the main basaltflow which covers the summit of the hill, and has the peculiar distribution to be described later. This mass of basalt evidently filled a depression which existed near the line of junction of the two types of phonolite, and which may have been the result of erosion, or perhaps was caused by the rapid solidification of the relatively acid phonolite lava. Section No. 1 gives the distribution of these rocks.

In connection with both types of trachytoid phonolite it may be stated that, in places where the rocks are exposed in situ, the cleavage is approximately horizontal.

Above the phonolite of the Logan's Point type there occurs, on the Signal Hill Road, about the 800 ft. contour-line, a bed of scoria 30 ft. in thickness. From an undecomposed core in this a specimen of basalt was obtained of a much more acid type than the main basalt-flow which followed it. It is described as basalt No. 4.

The main basalt-flow, described as basalt No. 1, covers the summit of the hill, and on the south-eastern side extends down only a short distance. It continues, however, much farther in a north-westerly direction, and to the north extends in a narrow belt across the head of the eastern arm of the North-east Valley towards Mount Cargill. This distribution is very peculiar, and it might be suggested that the basalt covering this area consists of a number of lava-flows, some of which alternate with flows of trachytoid phonolite. This explanation, however, is not upheld by field evidence.

The sections exposed on the Main North Road, and numbered consecutively 1 to 10 on the accompanying map, are as follows:—

(1.) Direction, N.E.-S.W. Height above sea-level, 330 ft. Basalt lies horizontally upon Logan's Point phonolite.

(2.) Direction, N.W.-S.E. Height above sea-level, 480 ft. Basalt lies upon trachytoid phonolite (Signal Hill type), dipping N.W. 12°.

(3.) The section is not clear, but basalt appears to overlie Signal Hill phonolite nearly horizontally. Height above sea, 520 ft.

(4.) Direction, N.W.-S.E. Height above sea-level, 550 ft. Basalt lies upon Signal Hill phonolite, dipping 12 ½° N.W.

(5.) Direction, N.E.-S.W. Height above sea-level, 580 ft. Basalt lies upon Signal Hill phonolite, dipping 4 ½° N.E. The basalt is then continuous for about half a mile.

(6.) Direction, E.N.E.-W.S.W. Height above sea-level, 750 ft. Basalt lies upon Signal Hill phonolite, dipping W.S.W. 8°.

At (7) and (8) the sections are not clear, but between these points basalt is continuous.

[Footnote] * Marshall, “Geology of Dunedin,” Quart. Journ. Geol. Soc, lxii, 1906.

– 114 –

(9.) Direction, N.W.-S.E. Height above sea-level, 860 ft. Basalt overlies Signal Hill phonolite horizontally.

(10.) At Junction School. Height, 920 ft. Basalt lies upon a convex surface of Signal Hill phonolite, dipping both N.W. and S.E. at an increasing angle.

The diagrammatic section along AA shows the relations of the rocks.

It is thus seen that the basalt-flow is everywhere later than the phonolite. Moreover, when all the above-mentioned basalt-outcrops are followed up they are found to be continuous with the flow covering the top of Signal Hill, and the flow is continuous across the deep valley to the north. At various levels, also, isolated patches of the same basalt occur on the spur leading down to Opoho.

It thus appears that the basalt flowed over a surface quite as uneven as the present surface of the hill, and filled up the depressions. That the basalt did not follow the phonolite in quick succession is proved by the occurrence of a considerable thickness of baked soil on the surface of the phonolite and below the basalt. No direct evidence of erosion, such as the occurrence of phonolite conglomerate beneath the basalt, has been observed, but it seems necessary to postulate a long period of erosion before the extrusion of the basalt which then filled up the depressions, and perhaps solidified on some of the slopes without filling up the valleys below. If, however, the basalt be supposed to have flowed from Mount Cargill, the valley which then occupied the site of the present North-east Valley must have been filled to a depth of 1,000 ft. with basaltic lava. From the lower part of the valley the basalt, if ever it occurred there, has been completely removed by erosion.

The ridges running down towards the North-east Valley are composed of basalt, while the gullies are eroded in the phonolite, which at the heads of the gullies is at a considerably higher level than the basalt at the lower parts of the ridges. This, together with the fact that the tongues of basalt occur in trough-like depressions in the phonolite, shows that the present ridges occupy the sites of ancient gullies.

The basalt mass is made up of numerous successive flows. At many points there are interbedded layers of scoria, and also red bands, which appear to be baked surface clays. If this is correct, considerable intervals must have elapsed between the successive extrusions of basaltic lava. These beds or layers cannot, however, be correlated on account of their resemblance to one another and of the similarity of the basalt of the different flows. The beds do not appear in any case to lie horizontally, nor to have a uniform dip and strike.

The strip of basalt filling the upper part of the North-east Valley is surrounded by older rocks at a higher level on all sides except in the direction of the mouth of the valley. Here the older rocks (phonolites) are at a lower level. Hence it appears that the ancient valley into which the basalt flowed had the same general direction as the North-east Valley. The ancient valley was nearly as deep as the present North-east Valley, and, if it be granted that there has been no subsequent tilting, had a gradient not quite so steep.

A small area of basanite occurring on the Lower North Road scarcely comes into the area discussed. Its relation to the basalt is not apparent. It occurs at a considerably higher level than the neighbouring Signal Hill trachytoid phonolite, and so is considered younger than that rock. It is distant about half a mile from the clear sections on the Main North Road

– 115 –

Several intrusions of both phonolitic and basaltic rock are found. The most striking of these occurs on the saddle north of McGregor's Hill. The rock, which is nephelinitoid phonolite, is not found in situ, but fragments are found on the surface over an area 40 yards by 150 yards, with its greatest elongation in the direction W.N.W.-E.S.E. Great variations occurring in the mineral characters of this rock in specimens collected within a few feet of each other, and its isolated occurrence, lead to the conclusion that it is intrusive. It is surrounded by a ring of basalt, an isolated remnant of the basalt-flow, which is here cut down by the action of streams approaching each other from opposite sides of the saddle. The intrusion, therefore, is later than all the lava-flows.

A group of dykes of a very basic basalt occur on the Main North Road. One is vertical, and 5 ft. in thickness; two others are irregularly inclined, and vary from 2 ft. to 3 ft. in thickness. Lower down the hill is a vertical dyke of the same character, 15 ft. in thickness. The strike of these dykes is about east and west.

In the upper part of the North-east Valley, about 500 yards south of the Junction School, there appears traversing the basalt a dyke of decomposed rock. On account of its decomposed nature, no detailed examination has been made of this rock. It has a fine lamellar structure, developed to such an extent that the lamellæ may be pulled apart by the hand. Its colour is a creamy grey, and it shows numerous white crystals of feldspar not entirely decomposed. Some of these are very large, and some of medium size. Their occurrence gives the decomposed rock a micaceous sheen. On account of the abundance of feldspar crystals, the rock has been tentatively classed as a trachytoid phonolite.

Sequence of the Rocks.

From the above description the following appears to be the sequence:—

1. Basalt of an acid type.

2. Logan's Point phonolite.

3. Signal Hill trachytoid phonolite.

4. Eruptions of basaltic scoria of an intermediate type.

5. Main basalt-flows.

6. Dykes of extremely basic basalt. Intrusion of nephelinitoid phonolite. Dyke of rock of doubtful composition, perhaps trachytoid phonolite.

Source of the Lavas.

Nothing definite can be stated as to the position of the vents from which the various lavas issued.

In the case of the Logan's Point phonolite the greatest thickness of rock occurs in the southern portion of Signal Hill. On the western side it thins out, and to the north dips below younger rocks. In the precipitous peak on the south-eastern side of Signal Hill the cleavage-planes of the rock are highly inclined, while elsewhere they are horizontal. Possibly this peak may be the neck or plug which solidified in the ancient crater; but the surrounding rock does not show signs of having flowed from this point. Similar rocks also are described from Mount Cargill, from the opposite shore of Otago Harbour, and from Otago North Head, where a bedded flow occurs; but such a great thickness does not elsewhere occur as that in the main Signal Hill mass.

*

[Footnote] * Marshall, “Geology of Dunedin,” Quart. Journ. Geol. Soo., lxii, 1906.

– 116 –

The trachytoid phonolite of the Signal Hill type does not occur over so great an area, nor, so far as is known, to so great a depth in the area described, but a similar rock has a wide occurrence to the north, as described by Dr. Marshall, and occurs also as a bedded flow at North Head. It may, then, be supposed to originate at some point north of Signal Hill.

The acid basalt underlying the phonolites does not appear to have any close allies in the neighbourhood, although I have compared it with a large number of sections kindly lent for the purpose by Dr. Marshall. It cannot, therefore, be traced to its origin.

The main basalt-flow closely resembles a basalt occurring as a neck on Mount Cargill. This may perhaps be its source. If so, the ancient valley intervening must have been filled to a great depth. The occurrence of the basalt at various levels on the inclined surface of the phonolites on Signal Hill is more easily explained on the assumption that basalt overflow occurred there, and the lava flowed down the slopes.

It is difficult to draw conclusions as to the nature of the magmas from which the rocks were derived, but they are best explained by supposing the rocks to be derived from two magmas, a basaltic and a phonolitic. The basalts became progressively more basic as vulcanism died out. The interval between the first basaltic flow and the next must, however, as before stated, have been great.

The phonolitic magma which supplied the lava of the Logan's Point type and the very much later nephelinitoid intrusion were similar in composition, and probably identical. The Signal Hill phonolite is chronologically intermediate between these two, but not intermediate in composition. It is much poorer in alkalies and richer in iron, and appears to be related to the trachydolerite of Mount Cargill.* It may have originated from a mixture of the basaltic and phonolitic magmas.

No explanation based on the theory of magmatic differentiation is applicable to the alternation of types, but the basalts, taken separately, may be conceived to be derived from a differentiating magma in its final stages.

Petrography.

Trachytoid Phonolite (Logan's Point Type).

This rock is mentioned under the above name by Dr. Marshall. At the typical locality (Logan's Point Quarry) the rock is distinctly trachytoid, and near the summit of Signal Hill even more so; but in the south-east peak feldspar is almost absent, and the rock should be classed as nephelinitoid.

Macroscopic Characters.—In hand-specimens the rock is of a greenishgrey colour, flecked with green streaks. It cleaves readily into flakes.

Microscopic Characters.—The structure is uniformly fine-grained in typical specimens, and holocrystalline. Flow structure is common wherever there is a considerable development of feldspar. In the typical rock from Logan's Point there are no phenocrysts. The chief constituent is ortho-clase or sanidine, but in some places feldspar becomes scarce. It occurs as bundles of fine needle-shaped crystals, some of which are of considerable

[Footnote] * Marshall, “Geology of Dunedin,” Quart. Journ. Geol. Soc., lxii, 1906.

[Footnote] † Marshall, “Geology of Dunedin,” Quart. Journ. Geol. Soc., lxii, 1906.

– 117 –

length—up to 0.5 mm.—but all are very narrow. Some show Carlsbad twinning. The crystals are often bent, and generally interwoven.

The constituents next in importance are nepheline and ægirine-augite. The nepheline is not distinguished without staining, but in a stained section fully one-third of the rock is seen to consist of nepheline, chiefly in short hexagonal prisms, and partly as fine allotriomorphous grains. The maximum size of the prisms is 0.1 mm. in length and breadth. Very small isotropic spots with low refractive index are put down as sodalite. The ægirine-augite occurs as flakes and needles without definite crystal outlines at Logan's Point. Near the summit better - defined elongated crystals occur, which are also larger, being as much as 0.5 mm. in length.

The maximum extinction-angle is 38°.

The colour is bright green, and the pleochroism as follows: (a) grass-green; (b) lighter green; (c) apple-green. With this may be compared the pleochroism of ægirine from Låven, given by Rosenbusch:* (a) pure green to blue-green; (b) olivine-green; (c) yellowish grass-green.

Associated with the ægirine-augite are numerous minute crystals of deep-brown, almost opaque cossyrite, varying in size from minute grains or needles to crystals 0.1 mm. in thickness. The form of the crystals is rather indefinite, but as nearly as can be made out the prismatic angle is that of typical cossyrite (66°). The mineral exhibits a deep-brown pleochroism. The grains are scattered evenly throughout the rock, being often associated with crystals of the green ægirine-augite.

MagAnetite is fairly abundant, in fine grains, while apatite occurs very rarely, in fine needles. No glass is present.

The high percentage of both ferrous and ferric oxides may be noted as peculiar to the rocks of this district. Treatment of the finely ground powder with hydrochloric acid dissolves a considerable amount. This points to the presence of a large proportion of nepheline; and the unusual amount of ferric iron obtained in solution by this method points perhaps to the solubility of cossyrite.

Variations of the Logan's Point Phonolite.—On the peak above Burke's mossy ægirine-augite is very abundant, and phenocrysts also occur up to 0.3 mm. in length. Feldspar is very rare, and the rock should be classed as nephelinitoid. Cossyrite and magnetite are abundant.

Near the top of Signal Hill the feldspar is relatively abundant, showing flow structure. At Normanby nepheline is abundant, as also is ægirineaugite, both as mossy growths and as phenocrysts of considerable size. These phenocrysts attain a length of 3 mm. or more, but are very narrow. They have sometimes a core of pinkish-grey augite. The extinction-angle of the ægirine-augite is 36°, and of the augite 40°. Orthoclase needles are moderately abundant. Cossyrite and magnetite occur.

The Logan's Point rock has been compared by Rosenbusch, in a private letter to Dr. Marshall, with the apachite of Osann. The chief resemblance, however, appears to be the presence of cossyrite. According to Rosenbusch, the distinctiveness of the apachite type is stated by Osann to be the richness in the younger amphibole minerals, the microperthitic nature of the feldspars, and the presence of ænigmatite. The last appears to be the only point of similarity.

[Footnote] * Rosenbusch, Mikr. Phys., 1905.

[Footnote] † T.M.P.M., 1896, xv, 394.

[Footnote] ‡ Mikr. Phys., vol. ii, p. 823.

– 118 –

In chemical composition the Logan's Point rock is similar to a tinguaite from Edda Gyorgis, Abyssinia,* the analysis of which is quoted for comparison:—

Trachytoid Phonolite, Logan's Point. Tmguaite, Edda Gyorgis, Abyssinia. Phonolite, Losuguta Type, British East Africa.
SiO2 57.00 57.81 58.37
Al2O3 16.06 18.74 16.65
Fe2O3 5.53 5.76 4.09
FeO 3.22 0.42 3.03
MgO 0.64 Trace 0.37
CaO 1.51 1.28 1.66
Na2O 8.00 9.35 7.28
K2O 6.18 4.52 5.46
H2O 2.10 1.50 2.36
TiO2 0.39 0.21
Cl 0.45

The analysis is also quoted of a rock from British East Africa, called by Prior the Losuguta type. This rock contains cossyrite, and its groundmass in particular appears closely to resemble the Logan's Point rock. Rocks with abundant cossyrite, which occurs in a manner analogous to its occurrence in the Logan's Point rock, are described by H. I. Jensen from Queensland.§ The Logan's Point phonohte was first described by Captain Hutton as an augite-andesite. He considered the groundmass to be chiefly glass, and mentions chlorite and a purple pleochroic augite. What he describes as chlorite is probably ægirine-augite, and the purple augite is not to be found. He evidently also mistook the cossyrite for magnetite.

Trachytoid Phonolite (Signal Hill Type).

This rock has been so named by Dr. Marshall. It is certainly a trachy-toid phonolite, with only a small amount of nepheline, but it contains amphibole as its chief ferro-magnesian constituent. Corresponding to the large amount of ferro-magnesian minerals and iron-ores, the content of both ferric and ferrous oxides is unusually high, as shown by the chemical analysis, given later. The rock is thus an exceptional one.

Macroscopic Characters.—The colour is greyish-green, and the texture rather coarse. Numerous cleavage-faces of feldspar are to be seen, and also crystals of amphibole of various dimensions, some being as much as 1 cm. in length and breadth. Sometimes a horizontal platy structure is noticeable, but the weathering is often spheroidal, and isolated boulders have the appearance of basalt.

Microscopic Characters.—The structure is porphyritic and holocrystalline, and the feldspars of the groundmass in places show flow structure.

[Footnote] * Prior, Mm. Mag., xii, p. 269, 1900.

[Footnote] † Contains also MnO, 0.43; P2O5, 0.08.

[Footnote] ‡ Prior, Min. Mag. xiii, 61, Feb., 1903, p. 238, and plate v, fig. 2.

[Footnote] § Jensen, “Geology of East Moreton and Wide Bay Districts, Queensland,” Proc. Linn. Soc. N.S.W., 1906, pt. i.

[Footnote] ∥ Hutton, Proc. Royal Soc. N.S.W., 1889, p. 134.

[Footnote] ¶ Marshall, “Geology of Dunedin,” Quart. Journ. Geol. Soc., lxii, 1906.

– 119 –

The largest phenocrysts are feldspar of different varieties in slightly corroded crystals. Anorthoclase is common in large crystals. It shows fine lamellar twinning, crossed at right angles by another set of very fine and rather indefinite lamellæ. As will be seen from the accompanying photograph (Plate X, fig. 2), these are too narrow and not regular enough to be attributed to pericline twinning of a feldspar of the albite-anorthite series. In thicker parts of the section extinction between crossed nicols is irregular. For these reasons this feldspar is regarded as anorthoclase.

As sections parallel to the cleavage are not obtainable, the extinction-angle cannot be verified.

Clear crystals of monoclinic feldspar with Carlsbad twinning are also common. They are referred to sanidine.

Oligoclase phenocrysts are also common, with polysynthetic albite twinning. Sections perpendicular to 010, having equal extinction-angles measured from the twin line, on either side of it, extinguish at an angle of 5° from the twin line.

Perthitic intergrowths are rather common.

The feldspars all occur in broad prisms, up to 3 mm. or 4 mm. in length.

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

Numerous phenocrysts of a deep-brown amphibole occur. They have definite crystal outlines when not entirely resorbed. Prisms and clinopinacoids are developed, giving hexagonal cross-sections. The crystals are elongated parallel to the c axis. The optical characters observed are as follows: a=a, b=b, cc=15°. The axis c lies in the plane of symmetry. Pleochroism: (a) Pale yellowish-brown; (b) deep brown; (c) deep brown. These characters show it to be intermediate between barkevikite and basaltic hornblende. The crystals are largely resorbed, being bordered, and in some cases entirely replaced, by fringes of magnetite grains associated with augite and a little calcite.

This amphibole has evidently been one of the first minerals to crystallize, having probably an “intratelluric” origin. It has afterwards become unstable in the magma owing to altered conditions, perhaps of pressure. The feldspar needles surrounding the amphibole phenocrysts have a fluxional arrangement.

An interesting comparison may be made between this mineral and the amphibole of a somewhat similar rock described from the “Beagle” collection.* In that rock the amphibole is converted to ægirine-augite, and no mention is made of magnetite. This is considered to be the source of all the ægirine-augite in the rock.

The size of the amphibole phenocrysts in the Signal Hill rock varies from 0.3 mm. in length and breadth up to 1 cm. or more.

Pyroxene phenocrysts are rare, and of small size. They are augite of a slightly sodic variety. The extinction-angle is 42°. The mineral is rather pale in colour, with very slight pleochroism, as follows: (a) pale green; (b) greyish, almost colourless; (c) slightly yellowish-green.

A few stout prisms of grey-coloured apatite occur, up to 0.2 mm. in length.

Olivine occurs rarely, and is largely altered to serpentine. A glomeroporphyritic inclusion has been observed. In the centre is a large cross-

[Footnote] * Geol. Mag., March, 1907, p. 100.

– 120 –

section of grey augite, surrounded by a mass of allotriomorphic crystals of lime-soda feldspar and grains of magnetite.

Groundmass.—The groundmass consists of lath-shaped feldspars, pale-green augite, numerous magnetite grains, and interstitial nepheline.

The feldspar is chiefly orthoclase or sanidine, but oligoclase also occurs, with the same characters as in the phenocrysts. The feldspar laths vary in size up to 0.5 mm. in length and 0.05 mm. in breadth.

The pale-green ægirine-augite occurs in irregular crystals. It has the same characters as in the phenocrysts. The pleochroism is very slight. The maximum size of the crystals is 0.5 mm. by 0.1 mm.

Nepheline does not appear without staining, but a stained section shows a moderate amount of interstitial nepheline. Associated with the nepheline is a little sodalite. There is no glass, and there is no cossyrite. The absence of the latter is rather remarkable, considering its abundance in the Logan's Point phonolite.

Chemical Composition.—The proportions of ferric and ferrous oxides are both very high, corresponding to the large amounts of ferro-magnesian minerals and magnetite present. The nepheline-syenite of Red Hill, New Hampshire, U.S.A.,* has a somewhat similar composition, excepting that its content of ferric oxide is much lower. An analysis of it is quoted for comparison:—

Trachytoid Phonolite, Signal Hill. Nepheline-syenite, Red Hill, New Hampshire.
SiO2 54.15 59.01
Al2O3 16.09 18.18
Fe2O3 7.35 1.63
FeO 4.90 3.65
MgO 1.61 1.05
CaO 3.86 2.40
Na2O 5.94 7.03
K2O 4.41 5.34
H2O 1.40 0.65
TiO2 0.41 0.81
P2O5 0.42 Trace
Cl 0.40 0.12

Order of Crystallisation. — (1) Some magnetite and all the apatite; (2) amphibole; (3) feldspars (after this corrosion of the feldspars and resorption of the amphibole took place, with liberation of magnetite); (4) more magnetite and pyroxene; (5) feldspar of the groundmass; (6) nepheline.

Nephelinitoid Phonolite.

Macroscopic Characters.—The rock varies in colour from lead-grey to green. It is dense, and fine-grained. Specks of dark pyroxene and large phenocrysts of feldspar, as much as ½ in. in breadth, occur. The green kind in particular has a platy structure.

Microscopic Characters.—The structure is holocrystalline and porphyritic, the phenocrysts, however, being rare. The groundmass in some cases appears cellular, and in other cases, where the nepheline is not so idiomorphic, there are alternate streaks of clear, fine-grained nepheline and mossy green ægirine. When phenocrysts occur they are large.

[Footnote] * Bayley, B.G.S.A., iii, p. 250.

– 121 –

Phenocrysts.—Large clear crystals of sanidine occur with Carlsbad twinning and cross-parting. In some cases they are elongated, and vary in length from 1 mm. or less to 1 cm. or more.

Nepheline occurs in sharply idiomorphic short prisms, but only rarely. They are as much as 1.5 mm. in breadth.

Groundmass.—Feldspar is very rare. Orthoclase occurs rarely in fine needles, showing Carlsbad twinning.

Nepheline is the chief constituent. It occurs abundantly in good idiomorphic crystals, in size about 0.06 mm., and also interstitially. It gives the section a honeycombed appearance.

Associated with the nepheline is abundant mossy ægirine or ægirine-augite of a deep-green colour. It occurs in very small crystals without definite outlines. The extinction - angle measured from the direction of greatest elongation is in many cases low, although it is impossible to determine it accurately. The pleochroism is from green to yellowish-green. At least some of the mineral is, therefore, probably ægirine; but ægirine-augite may also be present, though to a small extent.

Magnetite grains are very rare.

No glass is present, and no cossyrite.

Order of Crystallisation.—(1) Magnetite; (2) sanidine and nepheline phenocrysts; (3) mossy ægirine and the nepheline of the groundmass.

Chemical Composition.—The analysis of a tinguaite from Alnö, Sweden,* is quoted for comparison:—

Nephelinitoid Phonolite, Signal Hill. Tinguaite, Alnö, Sweden.
SiO2 53.80 50.26
Al2O3 18.72 20.15.
Fe2O3 4.99 3.67
FeO 3.59 2.62
MgO 0.86 1.43
CaO 2.80 3.28
Na2O 8.82 8.09
K2O 5.20 4.67
H2O 1.90 3.85
TiO2 0.30 0.24
Cl 0.14

Basalt No. 1.

Numerous flows of this rock occur, making up the chief basalt mass. All are very similar. Only one, which differs somewhat in texture, is briefly noticed separately as basalt No. 2.

Macroscopic Characters.—The rock is black and basaltic-looking, with numerous crystals of olivine and feldspar. No true columnar structure is to be seen, but some weathered boulders have the appearance of square prisms. Spheroidal weathering is general, and everywhere the solid rock is covered by a mantle of clay with spheroidal cores.

Microscopic Characters.—The structure is holocrystalline and porphyritic. Large phenocrysts of feldspar, augite, and olivine occur in a groundmass

[Footnote] * Rosenbusch, Elements, p. 215, 1898.

– 122 –

consisting of feldspar, magnetite, and augite, without glass. Flow structure is apparent in some localities and absent in others.

Phenocrysts.—Plagioclase phenocrysts vary in size up to 1.5 mm., and are abundant. Polysynthetic albite twinning is universal. The extinction-angle of adjacent lamellæ in sections perpendicular to 010 is 30°. The feldspar is therefore labradorite.

Augite phenocrysts are large, being as much as 7 mm. or 8 mm. in length and 2.8 mm. in thickness. The colour is grey, and the maximum extinction-angle 41°. The crystals are idiomorphic, and some show twinning, with the orthopinacoid as composition plane.

Magnetite occurs in crystals and rounded grains up to 0.4 mm. in diameter.

Olivine occurs in rounded crystals up to 0.5 mm., and, more rarely, larger. It is pale in colour. The olivine crystals are often replaced by serpentine pseudomorphs.

Groundmass.—The groundmass consists chiefly of irregularly arranged small needles and laths of labradorite and numerous irregular grains of grey augite and magnetite. At some places feldspar is very abundant, with flow structure, and at others no flow structure is to be seen, while magnetite and augite are relatively more abundant.

Chemical Composition.—For comparison the analysis of a dolerite* from Dyer's Pass, Canterbury, is quoted:—

Basalt No. 1, Signal Hill. Dolerite, Dyer's Pass, Canterbury.
SiO2 45.80 48.60
Al2O3 17.91 17.87
Fe2O3 6.14 6.20
FeO 8.69 5.76
MgO 3.92 4.32
CaO 8.10 9.11
Na2O 4.71 4.66
K2O 1.77 2.06
H2O 2.10 1.78
TiO2 0.35
Cl2 0.11

Basalt No. 2.

This basalt is not found in situ, but only as spheroidal cores in a deep clay. It has a cubical fracture. The minerals are similar to those in No. 1, and occur in the same proportions. The groundmass is similar to that of No. 1, the only difference being in the phenocrysts. The feldspar phenocrysts are much larger than in No. 1, and the olivine is nearly all decomposed.

Basalt No. 3.

This rock, which occurs in a dyke, is very hard and dense in appearance. Olivine crystals are seen in abundance.

Microscopic Characters.—In sections the rock is holocrystalline and porphyritic. The groundmass is much more coarsely crystalline than that of the lava basalts.

[Footnote] * R. Speight, Trans. N.Z. Inst, vol. xxvi, p. 409.

[Footnote] † MgO appears low to be associated with such a low amount of SiO2.

– 123 –

Phenocrysts.—The phenocrysts are augite, olivine, feldspar, and magnetite. Of these, olivine and augite are by far the most abundant. The olivine crystals are large, and are partly altered to serpentine. They are almost colourless. The augite crystals are grey in colour and up to 1 mm. in breadth. They have no pleochroism, and the extinction-angle is 41°.

Feldspar phenocrysts are rare. They are of basic labradorite, chiefly untwinned, and are about 1 mm. in length.

Magnetite grains occur up to 0.1 mm. in diameter.

Groundmass.—The groundmass is holocrystalline, and is coarser-grained than in the other basalts described. It consists chiefly of allotrimorphous grey augite, with labradorite laths averaging 0.3 mm. in length. The extinction-angle on either side of the albite twin line in sections perpendicular to 010 is 32 ½°. There are also abundant grains of magnetite. No glass is present.

This rock, owing to the scarcity of feldspar and abundance of augite and olivine, approaches a magma basalt or augitite.

Chemical Composition.—The chemical composition proves that the rock is derived from an extremely basic magma. For comparison the analysis of a basic basalt from Punta Delgrada, San Miguel, Azores,* is quoted:—

Dyke Basalt, No. 3, Signal Hill. Basalt, Punta Del-grada, S. Miguel, Azores.
SiO2 44.00 44.06
Al2O3 14.07 15.10
Fe2O3 5.16 5.23
FeO 10.87 7.93
MgO 11.18 9.84
CaO 10.28 12.56
Na2O 1.74 2.20
K2O 1.98 0.93
H2O 1.40 0.30
TiO2 0.47 1.80
P2O5 0.53
Cl 0.11
MnO 0.36

Basalt No. 4.

This basalt occurs in fragments in a scoria-bed on the Signal Hill Road, lying above the Logan's Point phonolite and below the main basalt-flow.

Macroscopic Characters.—The structure is porphyritic, the matrix being grey and fine-grained, with numerous white phenocrysts of feldspar and occasional black specks of augite.

Microscopic Characters.—In sections the porphyritic structure is strongly marked, augite and feldspar occurring in two generations. No flow structure is discernible.

[Footnote] * C. v. John, Jb. G. R.-A., Wien, xlii, p. 291, 1896.

– 124 –

Phenocrysts.—The most numerous and conspicuous phenocrysts are feldspar. These vary in size from 0.3 mm. to 3 mm. They show poly-synthetic albite twinning. The extinction-angle of adjacent lamellæ in sections perpendicular to 010 is 28°. The feldspar is therefore labra-dorite.

Less numerous are phenocrysts of grey augite and small pseudomorphs of serpentine with the form of olivine. The augite crystals vary in size from 0.2 mm. to 0.7 mm., and the olivine has been present in crystals 1 mm. or 2 mm. in length, and regular grains of magnetite up to 0.5 mm. occur.

Groundmass.—The groundmass consists of small laths of feldspar with a considerable amount of fine granular augite and magnetite. The feldspar is labradorite, and the augite is similar to that occurring as phenocrysts.

Basalt No. 5.

This basalt underlies the Logan's Point phonolite on the Main North Road, near Normanby, where fresh spheroidal cores are obtained, imbedded in a red clay.

Macroscopic Characters. — The rock appears dense and black, with occasional large phenocrysts of feldspar. It has a very irregular fracture.

Microscopic Characters.—The structure is holocrystalline and porphyritic. The groundmass is very dense, and the feldspar needles appear to have flowed around the phenocrysts.

Phenocrysts.—Feldspar is abundant, in irregular and lath-shaped crystals, showing polysynthetic albite twinning. The extinction-angle of adjacent lamellæ in sections perpendicular to 010 is 27°. The feldspar is therefore an acid labradorite bordering on andesine. Some of the crystals, especially the lath-shaped ones, are small; others are as much as 5 mm. in length.

Grey augite occurs very sparingly, in rounded grains, and there is a small amount of olivine in crystals up to 1 mm. in length.

Magnetite occurs in grains and octahedrons.

Groundmass.—In the very dense groundmass feldspar is moderately abundant, in fine needles. The remainder consists of grains of magnetite and augite.

Order of Crystallization for the Basalts.

(1) Magnetite; (2) olivine; (3) augite; (4) feldspar; (5) magnetite, augite, and feldspar of the groundmass.

Quantitative Classification of the Rocks.

The rocks have been classified on a quantitative chemico-mineralogical basis according to the system of Cross, Iddings, Pirsson, and Washington,* and the chemical results have also been plotted as diagrams according to the system devised by Iddings, and these diagrams have been combined in a multiple diagram (Plate IX).

[Footnote] * “Quantitative Classification of Igneous Rocks.” Chicago, 1903.

[Footnote] † Chem. Comp. of Igneous Rocks expressed by Means of Diagrams, U.S. Geol. Surv., 1903.

– 125 –
The chemical classification is as follows:—
Trachytold Phonolite, Logan's Point. Trachytold Phonolite Signal Hill. Nephelinitoid Phonolite. Basalt No 1, Main Flow Basalt No. 3, Dyke.
Norm. Norm. Norm. Norm. Norm.
or 36.70 or 27.80 or 30.58 or 10.56 or 11.68
ab 23.06 ab 36.15 ab 22.01 ab 19.91 ab 4.72
ne 8.52 an 4.73 ne 24.42 an 22.24 an 25.02
so 5.81 ne 2.84 ac 5.54 ne 10.79 ne 5.11
ac 14.32 so 3.88 di 11.23 di 14.97 di 21.16
di 6.44 di 11.96 ma 4.41 ol 9.44 ol 23.41
ol 2.50 ol 0.31 il 0.61 ma 8.82 ma 7.66
ma 0.70 ma 10.67 il 0.61 il 0.91
il 0.76 il 0.76
ap 1.01
Magmatic name. Class II, Dosalane Order 6, Norgare Rang 1, LaurdalaseSubrang 3, Judithose Magmatic name. Class II, Dosalane Order 5, Germanare Rang 1, Umptekase Subr 4, Umptekose Magmatic name, Class II, Dosalane Order 6, Norgare Bang 1, LaurdalaseSubr. 4, Laurdalose Magmatic name. Class II, Dosalane Order 6, Norgare Rang 3, Salemase Subr. 4, Salemose Magmatic name. Class III, Salfemane Order 5, Gallare Rang 4, Auvergnase Subrang 2,—–

The “norm” is seen to correspond roughly to the “mode,” or actual mineral composition.

It is interesting to compare the actual proportions of feldspar to feld-apathoid in the three types of phonolite: (1) In the Logan's Point phonolite, where feldspar and nepheline each constitute about one-third of the rock, silica is relatively high and alumina low, while alkalies are high; (2) in the nephelinitoid phonolite, where nepheline constitutes the greater portion of the rock, silica is lower, alumina is higher, and alkalies are about the same;

– 126 –

(3) in the Signal Hill phonolite, where nepheline is rare and feldspar of various types abundant, silica is nearly the same as in the nephelinitoid type, alumina is low, and alkalies are also low.

Explanation Of Plates VIII-XI.
Plate VIII.

Geological map of Signal Hill, with sections along lines AA, BB, CC, on map.

Plate IX.

Fig. 1. Map showing locality.

Fig. 2. Graphic representation of the chemical composition of the rocks (after Iddings).

Plate X.

Fig. 1. Logan's Point phonolite; × 200. Shows association of cossyrite and ægirineatugite.

Fig. 2. Signal Hill trachytoid phonolite; × 45. Crossed nicols. Shows large crystal of anorthoclase.

Fig. 3. Signal Hill trachytoid phonolite; × 45. Shows a large crystal of amphibole partly resorbed.

Fig. 4. Nephelinitoid phonolite; × 45. Shows the groundmass consisting chiefly of nepheline with mossy ægirie.

Plate XI.

Fig. a. Logan's Point phonolite overlying basalt No. 5 on Main North Road, near Normanby.

Fig. b. Dyke of basalt No. 3, intruded in basalt No. 1.