The Geology of Banks Peninsula—a Revision.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 74, 1944-45, Page 232

The Geology of Banks Peninsula—a Revision. By R. Speight. [Read before the Canterbury Branch, October 6, 1943; received by the Editor, June 26, 1944; issued separately December, 1944.] Part II.—The Akaroa Volcano. Table of Contents. 1. Introduction. 2. General Geological Features, Including:— I. The Basement Beds. II. Upper Slopes Facing the Harbour. 3. Original Extent, Form, and Height of the Cone. 4. Special Localities:— I. Scenery Nook. II. Dam Rogers Cliff. III. Little River Valley. 5. The Dyke System, With Special Reference To:— I. Panama Rock. II. View Hill. III. The Devils Gap. IV. Pulpit Rock. 6. Age of The Volcano. 7. Surface Deposits, Absence of Folding, etc. 8. Composition of Lava Flows and Comments Thereon. 1. Introduction. Although the Akaroa region has received a good deal of attention, chiefly on account of its historical associations and its landscape features, precise references to its structural geology are somewhat infrequent. Haast (1879) spends on it a little more than a page in his account of the geology of Banks Peninsula, and Hutton (1885) gives it very brief reference. The present author has contributed some account of the area (1917) as well as references to its petrology (1923 and 1924), while his last geological paper on the area (1940) gives an account of the basal beds exposed round the middle and upper parts of the harbour. Later references are by Cotton (1941), in his description of “Some Volcanic Land-forms in New Zealand,” and in his masterly account of the phenomena of vulcanity (1944) he states that the Lyttelton and Akaroa volcanoes have the form of lava domes. In view of this absence of extended description the present author submits the following observations which may serve to fill gaps in accounts already given. As the area under consideration is of considerable extent, with steep slopes, and requires detailed observation in many places, some difficult of access, this account is not meant to be complete, but only an addition to what has already been written in connection therewith. 2. General, Geological Features. (Plates 30 and 31.) Except near the upper edge of the crater-ring and near sea-level the slopes facing the harbour are almost completely masked by a covering of soil and slip-material, and therefore sequent observations from sea-level to the highest peak are virtually impossible. In this respect there is a contrast with the Lyttelton area, which is perhaps

explained by a greater readiness of the Akaroa rocks to respond to weathering under the influence of a heavier rainfall. A proportion of loose material in the surface soil adds to the masking effect. Thus, well-exposed faces at intermediate levels are rare, the Pulpit Rock in French Farm Valley and the bastion fronting French Peak affording striking exceptions to the general rule. On the external slopes exposures are more frequent. They occur frequently on the crests of the ridges, on the sides of the valleys near their seaward ends, and on the cliffs of the coastline. These latter cannot be closely examined since access to them is only by boat, and this is generally difficult and often impossible. As far as can be seen the upper slopes of the caldera and all the slopes of the periphery exhibit the structures normal to a composite volcanic cone—Cotton would call it a dome—with a subordinate number of those accidental features which modify its regularity. There is the usual alternation of lava-flows and ash-beds, but a precise estimate of the relative importance of the solid and fragmentary constituents cannot be arrived at owing to the completeness of the soil cover. Near the edge of the crater-ring massive breccias are at times strikingly developed. Exposures in the cliffs of the periphery indicate that the thickness of solid material is there greater than that of the fragmentary, if allowance is made for that constituent which is due to deposit under an advancing flow. The finer-grained fragmentary beds are usually rudely stratified, are reddish in colour, with a frequent purplish tint, not only near the shore but at all levels. These beds have rarely a wide extent laterally. (1) Basement Beds. The basement on which the volcano has been built consists of plutonics—syenite and gabbro, forming the distal end of Onawe Peninsula—and also trachytoid rocks, which form the major portion of the shelf stretching from Tikao Bay to the vicinity of French Farm (Speight, 1940, pp. 60–76). Although no contacts between these rocks, either plutonics or trachytes, and those of the cone itself are visible, it seems reasonable to consider them the older of the two, and that they formed an integral part of the foundation on which subsequent building took place (op. cit. pp. 74–75). There is also a possibility that an episode of basic extrusion antedated that to which the present cone belongs. It has been pointed out that the rocks exposed round the head of the harbour in the vicinity of Duvauchelles Bay show a degree of lecay that would hardly be expected had these rocks belonged to the stage during which the cone was actually built. This suggestion has been adopted by Cotton (1941, p. 303). Examples of such decomposed rock can be noted at the back of the Council Yard at Akaroa, on the lower slopes alongside the road leading from Akaroa towards Stony Bay, and along the shore between Barrys Bay and French Farm. Further, the irregularity in the direction of inclination of the flows round the foreshore of the harbour from Duvauchelles to Takamatua is not in agreement with the hypothesis of an outward quaquaversal dip from a volcanic centre located near Onawe, the position of such a centre being determined by a consideration

of the direction of dip of the beds forming the upper part of the cone. All due allowance must be made for the occurrence of reasonable irregularity due to the chance form of the surface on to which the lavas were outpoured. If the irregularity in direction is due to sagging or incipient fold movements, then it must be urged that the lava flows of the cone show no sign of similar variation in direction of inclination. If this phenomenon is attributed to structural causes then it strengthens the case for the existence of a previous episode of basic extrusion. In this connection special mention should be made of Otipua Hill, the semi-detached elevation lying between Takamatua and the bay on which the town of Akaroa is situated. The depression east of the hill over which the road passes, and the fact that some of the flows on the western front of the hill do really dip to the east, suggest that it has been constructed of material discharged from a centre beneath the floor of the harbour in close proximity, and that it forms the lower part of the sequence exposed in the neighbourhood of the road leading up the spur to the east in the direction of Long Bay Saddle. There is some doubt, however, concerning this apparently obvious interpretation. It is unfortunate that the summit and lower slopes of Otipua Hill are in general completely masked with soil, and that rock exposures are visible only in places on the western facing and along the shore; it is the latter which furnish most direct evidence. On the shoreline of the hill nearest to the town the beds consist of inter-stratified flows, agglomerate, and ash-beds with a general dip to the south-east. This is specially true of the fragmentaries, though all beds have an occasional easterly dip. They are penetrated by trachyte dykes. Over them lies a massive basalt with definite dip to the south-east and forming a considerable stretch of the shoreline with flanking shelves following along the hillside at higher levels. This basalt overlies a different basalt and also fragmentaries, which at their western exposure contain a small amount of trachytic material, some of the pebbles having been rounded in water; they also dip to the south-east. On the next point to the west a massive trachyte occurs and continues on the shore-platform for a chain and a-half. The contacts with the basalt at higher level are obscured so that one cannot give a definite opinion as to whether the trachyte is intrusive or not, but the large extent of the exposure as compared with that of the general run of the dykes suggests that it is not intrusive, so that it may be an outlying member of the trachytes just across the harbour between Tikao and Broughs Bays (Speight, 1940, pp. 68–71). Towards the western point of the bay basalt flows and fragmentaries are interstratified, the amount of the trachyte in the latter increasing and the basic component decreasing as the point is approached; it is probable that a little decomposed rhyolite is also present. The inter-stratification of the trachyte fragmentaries with basalt flows and the inclusion of a proportion of basic material in them—a proportion which is small in the lower levels but increases upward—suggests that the first basic eruptions broke through a trachyte cover, deposited fragments round the vent, and then as the volcano grew the proportion of trachyte declined and finally the deposited material became entirely basic.

On the western slope of the hill, basalt beds are seen dipping to the east, while along the shore beneath them the dip is variable, and the rocks are much decomposed and intersected by trachyte dykes, some of the latter so wide that they may not be intrusive at all. Fragmentaries, trachytic at the base and with an increasing basic component at higher levels, also occur. Considerable attention has been paid to this locality, since a correct interpretation of its features is most important if a true idea of the early history of the Akaroa volcano is to be obtained. While it must be admitted that the easterly dip of the beds on the western facing must be seriously considered, the prevailing southeasterly dip of the flows and fragmentaries throughout almost the whole length of the southern shore of the hill affords material for serious consideration of the hypothesis that a volcanic episode occurred antecedent to, and perhaps independent of, that responsible for the building up of what is usually regarded as the present Akaroa cone. Some reference should also be made to the beds exposed near sea-level on the south side of the bay on which the town of Akaroa lies. The first exposures of basalt that one encounters in this direction appear to have an inclination towards the middle of the bay, a direction which is contrary to what they should have if they were outpourings from a centre located somewhere near Onawe. Further along the shore these basic beds have an uncertain inclination; but when Green Point is reached, a solid basalt lying on a scoriaceous bed shows a definite inclination towards the mouth of the harbour, a direction in agreement with what one would expect were this bed a member of the Akaroa cone series. It seems reasonable to consider the partially submerged reef off Green Point to be a continuation of this basalt. When speaking of the dip of the beds towards the middle of the bay off Akaroa town as a proof that they did not come from a centre located near Onawe, one must not forget that the inclination of the flows and inter-stratified scoriaceous beds in the direction of this centre may be due to their having been poured out over an irregular surface or that some amount of sagging has taken place after the decline of volcanic activity; but the whole circumstances of the beds south of the town, taken in conjunction with those of the beds exposed on the shore of Otipua Hill, do strongly suggest that they belong to a distinct volcanic episode antedating that responsible for the formation of the cone itself. The possibility of the presence of basic volcanics belonging to an earlier period is to some extent supported by the fact that, near the gabbro on Onawe, basic volcanics show signs of metamorphism (Speight, 1940, pp. 62–63), that is, these rocks were in position before the gabbro was intruded, and it is also possible that they antedated the intrusion of the syenite. Otipua Hill and its neighbouring slopes seem to have some physiographic relation with the shelf, composed largely of trachyte flows, on the opposite side of the harbour. Also the break in the profile of the spurs reaching down into the upper harbour, where the decomposed rocks are specially developed, gives some support to

the hypothesis that all these rocks formed part of a basement on which the cone was subsequently constructed, and that this physiographic break marks the approximate position of the pre-existing surface. I am unfortunately unable at the present time to indicate any factor in the analyses or mineralogical composition of the rocks of these two series that may enable them to be differentiated, nor can I indicate the precise upward limits of the presumably older series owing to the obscurity arising from the mantle of soil, beyond the presence of a presumed physiographic break mentioned earlier. It might be stated here that there are reports of the presence of coal near Lushington Bay, on the western margin of Otipua Hill, but I have seen no evidence of its presence. Haast mentions that fragments of wood turned into anthracite, are to be found in some of the tuffs, and this may account for the report. However, the presence of waterworn pebbles in the fragmentaries on the shore near the south point of Otipua Hill cannot quite be neglected in this connection. (2) Upper Slopes Facing the Harbour. (Plate 30.) The upper slopes facing the harbour are formed by the truncation of outward dipping flows, and are frequently bold and precipitous. Notable among these on the western side are Mount Bossu (2386 feet), Saddle Hill (2758 feet), French Peak (2685 feet), Rocky Peak (2297 feet), and on the eastern side Duvauchelles Peak (2406 feet), Okains Peak (1880 feet), Lavericks Peak (2478 feet), and Purple Peak (2643 feet), while extending from the last-named towards the entrance to the harbour are the most striking faces occurring round the crater-ring. Their upper surfaces rise to 2668 feet in Flag Peak, and to 2643 feet in Mount Berard, and fronting them are the precipitous slopes known as Brazenose. From this locality thick flows form the crest and upper slopes of the ridge extending towards the entrance, and are probably continuous—or at all events they belong to the same horizon—with the massive flow, some 250 feet thick, which forms the major portion of the beds exposed in the face of Dam Rogers cliff. The sharp-pointed masses which dominate that part of the crater-ring between Duvauchelles Peak and Lavericks Peak at the head of Okains Valley, the most westerly one known as Okains Peak, appear at first sight to be dykes, but a closer examination shows this to be unfounded. The top of Okains Peak consists of a mass of basalt, the lower part fine-grained, laminated, and evidently chilled, resting on fragmentary beds under which lie interstratified basalt and fragmentaries. The upper basalt, forming the actual summit, is irregularly columnar, the columns being of large diameter. There is no evidence of intrusion except a small dyke, which does not penetrate the capping. While making an examination, one had always in mind the possibility that it might be a mass which had mushroomed out over the adjoining surface from even a small dyke, but this possibility was considered unlikely. A similar mass forms a slightly higher peak about half a mile to the south, and they both seem to be remnants of a widely extended flow, which may have formed part of the massive capping near the

summit of Lavericks Peak, a higher elevation about half a mile to the east of Okains Peak. The peaks round the harbour are almost invariably the terminations of the radiating ridges dividing the external valleys, while between the high points the general level of the edge of the crater-ring takes on beautiful catenary forms, which mark the heads of the external valleys. The internal valleys seem to have developed independently, and their heads rarely coalesce with those of the external slope. These circumstances seem to indicate a considerable erosion of the upper ends of the ridges and that they once extended a considerable distance over the fringe of the area occupied by the harbour. 3. Original Extent, form, and Height of the Cone. The boundary of the beds forming the Akaroa cone, as it stands at present, follows round the seaward margin of the peninsula from the western shore of Pigeon Bay on the north to the vicinity of Kaituna on the south, the area covered forming a rough circle 20 miles in diameter. A marginal fringe of uncertain width lies submerged beneath the sea on the eastern periphery and on the south is covered by the alluvial deposits marginal to Lake Ellesmere and by the marine gravels of the Ellesmere Spit. The western boundary presents serious uncertainties. As mentioned in my former account there is no reason to consider either Pigeon Bay or Little River valleys as centres of activity. Also I consider Mount Sinclair to belong to the Akaroa cone for reasons already given (Speight, 1917, p. 383). But the question of Mount Herbert appears uncertain still. The lavas of which its summit and a section of the northern flank are formed were credited by both Haast (1879, p. 346) and Hutton (1885, p. 216) to a centre located in the upper part of Kaituna Valley or even to subsidiary craters on the summit of the mountain and its westerly extension. In my first account (1917, pp. 382–3) I accepted these conclusions, but in a subsequent article (1933, pp. 41–51) the matter was fully discussed and the conclusion tentatively reached that the Mount Herbert lavas came from the direction of Akaroa; the flows on the eastern bastion of Mount Herbert are certainly inclined upward in that direction. However, the great thickness of fragmentaries, greater than in any other locality on the peninsula, seems to suggest for them a locus of origin in the immediate vicinity. If that is the case, the upper part of Kaituna Valley is probably the spot from which they came, and the valley itself, south of the basin in its upper reaches, has been eroded on the southern flank of a cone constructed in this area, although the valley can quite well be attributed to development along the junction of beds of undoubted Lyttelton and Akaroa origin. The presence of the thick beds of fragmentary material is certainly not conclusive, for deposits of considerable thickness are exposed in various other localities clearly associated with Lyttelton, and the Kaituna deposit may have come from the same source. The original form of the cone can be arrived at only by considering the evidence furnished by the existing remnant. There is no doubt that it is constructed of lava flows with a slightly larger volume of interstratified material. The relative importance of these

constituents cannot be precisely determined owing to the small number of clear-cut exposures at intermediate levels and the difficulties of approach to the marginal wave-cut cliffs, and estimates based on observations of the latter are largely impressional. However, on the sea-cliffs the amounts seem in general to be sub-equal, but on the shore-platforms inside the harbour and on the road-cutting near the edge of the crater-ring fragmentaries exceed the lava-flows in thickness by a considerable amount. The flows near the present summit have an inclination varying between 3° and 15°, but the higher values are rare and the beds are frequently almost flat. At intermediate situations the inclination is generally lower. For example, on the eastern side of Little River Valley (Plate 31) it is 5°, and it is the same at the Devils Gap—to be referred to later—while at the margin of the cone it is still less, and in places the flows are almost horizontal. This low inclination can be clearly seen in the cliffs near Dam Rogers (Plate 31) and at the Akaroa Heads, on the sides of the Little River Valley near Lake Forsyth, and on the periphery between that lake and Kaituna. These inclinations have the general character of those of a lava cone, but it is extremely probable, as Cotton suggests (1944, p. 91) that the roof over the caldera had the form of a flat dome, and in its completed form it resembled in shape the shield volcanoes of Iceland and specially those of Hawaii. The general uniform height of the dominant elevations right round the crater-ring gives support to this suggestion, since such uniformity might be expected to arise as a result of the erosion of a flat dome rather than of a cone. In the latter case irregularities would certainly occur had a long period of erosion followed on the cessation of volcanic construction. There is some difficulty in arriving at a tolerably accurate estimate of the former height of the volcano owing to the uncertainty of the angle of inclination of the flows that formed the actual summit. The general angle of those on the middle slopes, such as those shown on the flanks of the ridge east of Lake Forsyth and the Devils Gap, is taken to be about 5°, and if this value persisted as far as a centre located near Onawe, a distance of about seven miles, it would mean that the summit formerly reached a height of approximately 5,000 feet. If the steeper angles occurring near the edge of the crater-ring continued towards the probable centre, then the height would have been somewhat greater, and if, on the contrary, the angle flattened then the height would have been correspondingly less. In this connection it should be noted that the flows on the eastern bastion (2997 feet) of Mount Herbert are inclined at an angle suggesting an origin near Akaroa and distant from Onawe some eight miles. Even if they rose in that direction at an angle as low as 3° it would mean that the summit of the volcano was 2,300 feet higher than the summit of the bastion—that is, it exceeded 5,000 feet. This is as accurate an estimate as facts appear to warrant. The same method of calculation would give Lyttelton volcano a height of between 3,000 and 4,000 feet, so that Akaroa was the greater.

And there is a reason why the two volcanoes were probably higher, although their forms remained the same. Owing to a lowering of the land relative to the sea after volcanic activity declined or had actually ceased, the present height is less than it was formerly. The precise amount of lowering is not known, but earlier in this revision (Speight, 1943, p. 15) evidence is given that in the Lyttelton area, and inferentially in that of Akaroa as well, the land formerly stood at least 700 feet higher than at present, therefore it is probable that the Akaroa volcano formerly had a height approaching 6,000 feet. Thus it did not reach the height of the present cones of Ruapehu or Egmont, though, allowing for the wide area that it covers, its dome-shaped form might give it a greater volume. 4. Special Localities. (1) Scenery Nook. A specially interesting landscape feature, with striking colour effects, occurs in an indentation of the southern coast about a mile and a-half west of the harbour entrance; this is known as Scenery Nook. Exposed in the face of the cliff on the western side are well-stratified beds of ash, scoria and agglomerate, reddish in colour, above 100 feet in thickness, bent into a syncline, and capped by the ordinary basalt; an underlying basalt is exposed at the base of the beds. They are penetrated by dykes, presumably trachytic, some not oriented according to the usual arrangement of Akaroa dykes to be referred to later. The occurrence is possibly a parasitic cone or outburst on the flanks of the volcano during its construction, which has been covered up by subsequent outpourings from the main centres of eruption. (2) Dam Rogers Cliff. This is the most striking scenic feature on the shore of the harbour. It lies about a mile inside the north head. The cliff is vertical, 500 feet in height, and a solid flow of basalt occupies half the face; this is inclined at an angle of about 7°; narrow bands of scoria lie beneath it and divide the flows at lower levels. In these the sea has driven caves. Towards the entrance to the harbour the flows are numerous, only a few feet in thickness, separated by thin scoria and ash-beds which are inclined seaward at low angles. (3) Little River Valley. (Panorama and Plate 31.) The panorama has been included in order to give some idea of the general features of the valley. The two trailing spurs which converge towards the floor of the main reach give it a form entirely different from that of the caldera of Akaroa, a difference due primarily to the fact that, while the tributary valleys of the former have been eroded in beds dipping initially in the direction of the valley profile (Plate 31), those of the latter, except near the entrance to the caldera, have been eroded across the scarp of the beds. Nevertheless, the possibility that Haast and Hutton might be correct in attributing the major features of Little River Valley to explosive action was always kept in view when examining the area from a stratigraphical standpoint. If they were correct some of the flows should show a quaquaversal dip, including directions towards

Akaroa; and this does not occur. The outcrops of flow so strikingly developed on the eastern side of the valley indicate an origin in the direction of Akaroa, and the rocky knobs which stand up like the vertebrae of a backbone on the skyline are the remnants of the upper edges of flows oriented to that centre and not to Little River. The inclination of a well-defined outcrop on the northern flank of Saddle Hill, seen clearly from both Akaroa and Little River, has an appearance suggesting that it might have come from a centre located in the upper part of the latter valley; but a careful examination on the spot discloses this to be illusory and that it really came from a locality in the direction of Onawe. Of course the flows shown in Plate 31 could have come from either centre, but other circumstances rule out Little River Valley as a possible source. In one or two cases the outcrops of flows on the western side of the main valley at an intermediate level show little inclination, but in my opinion these are accidental features dependent on the form of the surface over which the lava flowed. There is a possibility that some of the lower beds were derived from the Lyttelton centre, and the level outcrops can be thus explained, since they would in that case be on the very periphery of the cone where flat beds might be expected. 5. The Dyke System. (Plates 33 and 34.) The dyke system of Akaroa is analogous to that of Lyttelton. There is the same radial arrangement in the outer parts of the area and the same criss-cross pattern near the centre, the peninsula of Onawe, in Akaroa Harbour, corresponding with Quail Island, in Lyttelton Harbour, as the locality on which the outlying dykes in general converge. A great many show on the shore-platforms round the upper reaches of the harbour and specially so in the vicinity of Onawe. Fewer are visible on the shore lower down the harbour, and there are short reaches where they are quite absent. For example, they do not show at all in the western part of the massive basalt that flanks Otipua Hill for a quarter of a mile on its south side, though a few, both trachytic and basic, occur near the eastern end. They are very rare beyond a line extending across the harbour from Green Point to Wainui Island, though they are reasonably common near both those places. This is no doubt partly due to the fact that the shoreline in the lower reaches of the harbour is parallel to the strike of the dykes in that sector, so that only those departing from the general direction have a chance to show, except where they cut irregularities in the line of the coast. Some show on the external cliffs, as at Scenery Nook, and a high promontory, some two miles east of the harbour entrance, is called Dyke Head. They are, however, practically absent from that stretch of the periphery extending westward from Birdlings Flat, where the ends of the spurs were once cut back by the sea. A very limited number appear on the exterior of the cone or in cuttings of the Summit Road and of the roads on the flanks of the volcano. There is nothing at all like the great succession round the Lyttelton tops, although there are certain sections where they occur sparingly. For example, they show on the road to the Lighthouse,

one more than 15 feet thick, and three trachyte dykes occur within a chain on the Summit Road at the head of Okains, also basic dykes appear occasionally on Cameron's Track, which passes round the north side of Lavericks Peak, and on the road to Peraki. Two basic dykes cut across the Saddle Hill ridge, one about a chain north of the trig, another some ten chains away to the south-east, while a large trachyte dyke, over 18 feet thick, and oriented to Onawe, crosses the south-eastern shoulder of the hill. There is, however, a marked deficiency in number as compared with Lyttelton. This may be due to an actual paucity, but other factors afford reasons for the small number. (1) The more complete covering due to weathering as distinct from erosion may partly explain the small number of Akaroa dykes exposed. (2) Owing to its more youthful age it has not been exposed so long to erosion, and therefore the ribs of the volcano, so to speak, have not been exposed to view. (3) The force behind the intrusions may not have been competent to raise a large number to the neighbourhood of the crest of the Akaroa crater-ring, the effort dying out at lower levels, and particularly so since the rocks of which the cone has been constructed, judging from available exposures, are more resistant to rupture than those of Lyttelton and contain a lower percentage of fragmentary material. Haast has pointed out that dykes of the Lyttelton area frequently do not reach the present surface (op. cit., p. 340 et seq.), and he mentions that five of the dykes which cut the floor of the Lyttelton Tunnel did not reach the roof. In his explanation he stresses the fact that many of the dykes not exposed on the present surface of the ground did actually see the light when they were injected, but that subsequent flows covered up their outcrops. This serves to explain the occurrence of a dyke on the western side of the entrance to Akaroa Harbour. Any or all of these explanations just cited may account for the small number of dykes seen on the exterior and interior slopes of the Akaroa cone above the shoreline. There is perhaps a tendency to regard the present general outline of the cone as representing its maximum development and the dykes now visible as having at the time of their intrusion reached the height they now show and not reached higher, whereas a considerable stripping of their projection upwards and of the overlying rocks has taken place. Though there is a dominance of trachyte in this area, basic dykes certainly occur. Analyses have already been published by the author (1923, p. 149, and 1940, pp. 71–2), and additional ones are given on page 243. There are four occurrences of trachytoid intrusions not already mentioned and worthy of special notice. Two of these lie on the eastern slopes of the volcano—viz., (i) Panama Rock and (ii) View Hill, and two on the western side—viz., (iii) the Devils Gap, near Peraki, and (iv) the Pulpit Rock in French Farm Valley, the last on a slope facing the harbour.

(1) The Panama Rock. (Plate 33.) This lies on the eastern side of the Panama Track, a little-used road connecting Okains Bay with Le Bons Valley across the upper levels of Lavericks Valley. When viewed from the Summit Road at the head of Le Bons (Plate 33), it presents a most striking appearance, and it rises from the crest of the ridge north of Le Bons like a typical volcanic plug, with a long trailing dyke stretching from it in a south-westerly direction. The general level of the ridge from which it rises is about 1,600 feet, and the mass itself reaches a height of 1,891 feet. It extends about 15 chains in a north-easterly direction, almost in a line with the dyke, and its slopes are precipitous on all sides but the north-east; its summit is fairly even and of varying width, with a maximum of about four chains. At its north-eastern end it slopes down rather rapidly, and does not cross the road along the Lavericks-Le Bons ridge to the east-north-east. On the north face it shows rudely columnar structure in addition to vertical partings. The exposed terminal mass appears to be an expansion of the end of a dyke which fed it, and the material discharged from the fissure apparently mushroomed out over the adjoining surface, thus accounting for the great volume exposed. Near it the dyke is over a chain in width, but half a mile away down the track to Le Bons it narrows to about 50 feet, and after another 10 chains it does not show on the surface at all, and it cannot be recognised on the Summit Road towards Onawe where it should occur if the alignment continued. The terminal mass may represent either (i) the summit of a subterranean dome of laccolitic character, or (ii) a volcanic neck, or (iii) a volcanic plug. In the first case, the present exposed position would require that the former cover had been removed by subsequent erosion; but I do not think that the existing form of the Lavericks-Le Bons ridge in the vicinity of the mass is substantially different from that obtaining when the intrusion took place. It must be admitted, however, that I have not examined the ground at Trig. Q (1,865 feet), a mile east-south-east of the Panama Rock, or Le Bons Peak (1,641 feet), a mile and three-quarters away to the east, both on the ridge just mentioned. In the second case there is no sign of the remnant of lava flows or fragmentaries usually associated with a neck, and the petrological character of these should be related to that of the central mass, and therefore easily recognisable. These two hypotheses are considered unsatisfactory, and so there remains the third—viz., that the rock is a volcanic plug. It will be found to answer the description of a plug given by Daly (1914, p. 131), which is as follows: “In exceptional cases the highly viscous lava of relatively cool vents has exuded in quantity sufficient to form distinct domes at the surface, and notably overlapping the limits of the vents. These domes have grown endogenously, as bodies of unbroken, massive lava.” This statement seems to agree with Stearns' (1942, pp. 21–22) description of the mode of formation of the bulbous domes of Maui, and it certainly fits in with the phenomena to be seen at the Panama Rock.

Another point to be considered is whether or not the mushrooming over the adjacent surface occurred all along the line of the dyke or was located only near its termination. Unfortunately this cannot be determined for certain, since the dyke is exposed at lower and lower levels as it continues to the south-west, and any expansion at the surface as it was originally may have been removed by subsequent denudation. But there is no remnant that I have seen which might suggest a former expansion. I therefore conclude that this bulbous dome, to use Stearns' term, was located on the present ridge at the end of the extension of the dyke towards the north-east. An analogous mushrooming out on the margin is given by the dyke called Dover Castle on the western side of Heathcote Valley in the Lyttelton area. This has a somewhat peculiar chemical and mineralogical composition with trachytoid affinities (Speight, 1923, pp. 136–37). The Panama Rock has a definite fissile texture with a glistening appearance on the fractured surface, and this is specially to be noted where the material has expanded over the adjacent rock, but it shows distinctly all over the summit. The surface is marked with definite indentations, with the cores of the projections formed of harder rock. Petrologically it is a soda trachyte, as is disclosed by the analysis given herewith. Note.—No attempt is being made to give a full petrographical description of the rocks mentioned here and subsequently. That is left to a more competent petrologist.—R. S. Analyses of Trachytes. Analyst: F. T. Seelye. No. 1. No. 2. SiO2 62.36 62.82 Al2O3 17.07 17.10 Fe2O3 2.97 2.49 FeO 2.09 2.54 TiO2 0.33 0.34 MgO 0.26 0.29 CaO 1.40 1.29 Na2O 6.48 6.75 K2O 5.40 5.16 H2O+ 0.72 0.56 H2O- 0.76 0.51 CO2 nt. fd. trace P2O5 0.07 0.11 V2O3 nt. fd. nt. fd. ZrO2 0.03 trace Cr2O3 nt. fd. nt. fd. MnO 0.07 0.08 NiO nt. fd. nt. fd. BaO 0.06 0.03 * Determined spectrographically directly on a sample of the rock. SrO < 0.01 0.002 S. 0.03 0.01 Cl. 0.01 trace 100.12 100.08

C.I.P.W. Norms and Symbols. No. 1. No. 2. Q. 1.47 0.94 or. 31.89 30.50 ab. 54.79 57.10 an. 1.59 1.14 di. 3.29 3.86 hy. 0.93 wo. 0.41 mt. 4.26 3.61 il. 0.63 0.66 ap. 0.17 0.27 pr. 0.05 No. 1. Soda Trachyte, Panama Rock, Le Bons Bay. I”. 5. 1. (3) 4. Nordmarkose. No. 2. Soda Trachyte, Devils Gap, Peraki. I (II). 5. 1. “4. Nordmarkose. Under the microscope the rock appears to be composed of short laths of alkalic felspar some characterised by denticulate margins, like the anorthoclase of a bostonite. Scattered through an even-grained mesh of this material are numerous short laths of greenish, pleochroic, aegerine-augite. Phenocrysts are rare and the few I have seen are of sanidine. Crossing the junction of the roads just north-west of the Panama Rock is a basaltic dyke, one of the few visible on the external surface of the cone, and its visibility results entirely from excavations in forming the road. (2) View Hill. View Hill forms an elevated ridge rising to 2,491 feet, oriented north-east and south-west, about half a mile in length, near the proximal end of the divide between Little Akaloa and Okains valleys. The summit of the ridge is fairly even except for a slight depression which cuts across about a third of the distance along it from the north-east. Passing along it from this direction the following sequence is exposed:— (a) Basalt, the remains of a solid flow dipping north-east, with approximately vertical columnar structure, and probably originating from Akaroa. (b) Fragmentaries, containing scoriaceous masses, exposed on the surface of the ridge for several chains, thickness uncertain They mark the depression in the crest of the ridge just referred to. Their development suggests the possibility that they mark the position of a parasitic cone. (c) Basalt, differing in texture from (a). Contacts with (b) and (d) are obscured, though it probably overlies (b) and underlies the margin of (d). (d) Trachyte, forming a long oval dome, about 15 chains in length and 10 in width, oriented in the direction of Onawe; the summit is slightly inclined in this direction, and the slopes at the margins are only moderately steep, and are covered with grass; exposures are rare and the contacts obscured.

(e) Basalt appears on the steep slopes at the western end of the ridge, having come apparently from the direction of the Akaroa centre. The remaining part of the ridge to the southwest is narrow, with interstratified flows of basalt and fragmentary material. It furnishes a typical example of the destruction of inter-valley ridges near their proximal ends and the semi-isolation of a planeze. The trachyte of section (d) is the interesting feature of this occurrence. At either end of the exposure the rock is brownish in colour and very flaky, being reminiscent of the fissile material of the Panama Rock, but almost half-way along there is a small exposure of more massive form, just rising above the grassy surface, very much lighter in colour, somewhat vesicular in texture, with some of the vesicles partially filled with an undetermined amygdaloidal mineral, softer than glass and not acted on by acid. Both facies of the rock are aegerine-augite trachyte, the brown variety containing more augite than the light-coloured one. The oval, dome-shaped form of the trachyte mass suggests an analogy with the bulbous domes of Stearns. It is certainly an intrusion or results from one, but the contacts are obscured and its relationships to the associated basalts cannot be determined for certain. The inclination of the lava flow at the eastern end of the ridge—presuming that it originated from Akaroa—suggests that it once continued over the trachyte mass, and that the cover has been subsequently removed by erosion, but how much was removed cannot be said There may be some analogy between this case and those described by Daly (1925) in his account of the geology of Ascension Island, where viscous trachyte masses are very intimately associated with basalt eruptives (pp. 23–38). Both the form of the mass and the probable existence of a former cover indicate a laccolitic origin, the longer axis of the dome corresponding possibly with the direction of a short feeding dyke oriented to Onawe. It does not continue to the south-west beyond the end of the ridge, though it may continue after some break to the north-east. It is possible, but not probable, that the tilt of the basalt at the eastern end of the ridge may be due to a warping of the overlying beds resulting from the intrusion, but there is no sign of such warping at the western end. I was not able to locate the dolerite previously referred to (1924, p. 262) as coming from this locality. It contained greenish augite, brown hornblende, a little mica, and much apatite. This may indicate the presence of another intrusion to the north-east, since the rock is entirely different from any known to occur in the Akaroa area or on the remaining portion of Banks Peninsula. (3) The Devils Gap. (Plates 33 and 34.) An analogous trachytoid intrusion of massive size forms the eastern wing of the Devils Gap, a striking landscape feature on the western side of Peraki Valley on the southern flank of the volcano. The western wing has been referred to earlier in connection with estimates of the former height of the cone. This massive capping of basalt, one hundred feet thick, the summit 2,412 feet above sea-

level, stretches down in a southerly direction along the middle part of the ridge dividing Peraki and Te Oka valleys. The intrusive mass lies east of the basalt, extends down the slope to the east (Plate 33), and after an interval obscured by soil and surface deposits, appears again as a massive buttress nearer the floor of the valley (Plate 34), the whole occurrence forming the most extensive example of this class of intrusion that I have come across on the peninsula. The upper part seems to be plastered, as it were, on the side of the valley, suggesting an orientation towards Onawe, but it cannot be traced uphill along the ridge more than a few chains beyond the end of the basalt, where a very limited exposure occurs in the tussocks, and, further, the great mass at lower levels on the side of the valley discounts this apparent orientation, and indicates an extension along an east and west axis. Had the Peraki Valley been eroded when the intrusion took place the lower part might have been regarded as a viscous tongue directed downhill from a dyke oriented to Onawe from near the position of the trachyte mass which forms the eastern wing of the gap, but it does not seem likely that the intrusion took place at such a late date in the history of the volcano. The total length of the exposure in an east-west direction is about half a mile, and its width over 15 chains, but I have seen no clear-cut contact between it and the associated flows and fragmentaries. The basalt mass and the beds beneath it about two chains away from the trachyte across the gap, show no disturbance as a result of the intrusion. No feeding dyke, like that at the Panama Rock, can be seen, and it is apparently more closely related to the occurrence at View Hill, though it does not exhibit the dome-like form of that intrusion. Perhaps this difference is due to modification of its upper surface when the Peraki Valley was eroded, for I cannot think that its intrusion occurred after that valley had taken on its present form. Also it cannot be said for certain that the intrusion reached the ground surface at the time of its injection, though the present situation implies that it did not do so. It was probably laccolitic in form and the erosion responsible for the excavation of Peraki Valley has been the reason for the exposure to view of this sub-surface intrusion. The rock is resistant in the mass though the surface exposed to the weather is soft. The lower part exhibits the vertical parting characteristic of the Panama Rock. Analysis No. 2, p. 243 shows it to be a soda trachyte closely related to this rock. Under the microscope occasional phenocrysts of sanidine (? anorthoclase) and aegerine-augite appear in a base of alkali felspars laths and stumpy forms of aegerine-augite. Thus its mineralogical composition is analogous to that of the Panama Rock, though it differs slightly in texture, notably in the form of the felspar of the base. (4) The Pulpit Rock. (Plate 34.) This is the last of the four large trachytoid intrusions to be mentioned. It stretches across the upper part of the French Farm Valley for over a quarter of a mile, and its summit lies at a height of about 1500 feet. The stream occupying the floor of the valley has cut deep into the rock dividing the exposed portion into two unequal

parts but not exposing the base. The southern part forms a bold landscape feature and rises precipitously for 250 feet above the ground fronting it; this part is known as the Pulpit Rock. The northern part is thinner and less striking; it exhibits a closely-spaced and parallel lamination, perhaps due to chilling by the underlying basalt when the mass was injected. The surface deposits have so obscured the contacts that a satisfactory conclusion cannot be arrived at concerning the form of the channel up which the magma came or what was the original shape and extent of the mass. This uncertainty as to its original form is accentuated by the erosion it has suffered during the formation of the caldera, and its exposure to the light of day may be due to that cause. It appears to be oriented on approximately N.N.W.-S.S.E. lines, and it probably extended in the former direction across the top of the ridge dividing French Farm Valley from the western branch of Okuti Valley; for, after an obscurity due to the covering of soil, trachyte in position appears for nearly a chain close to the old yards on the summit of the ridge (height 1,815 feet), and scattered fragments continue on the surface for more than five chains further into the upper part of Okuti Valley. This rock is a variant of that of the Pulpit Rock, and may be a marginal facies of it; it is a dark-coloured, very hard, fine-grained aegerine-augite trachyte. The basalt flows above the main mass in the direction of Saddle Hill do not appear to have suffered any dislocation or warping from the intrusion of such a thick body of rock, and this suggests that it had flowed out at the surface as a bulbous dome and been covered up subsequently by material discharged as the cone was built up further. All the same it is doubtful how far away from it such a mass would cause warping in pre-existent flows. But if the exposure on the crest of the ridge near the yards does link up with the mass in the valley, then this explanation becomes unsatisfactory, especially as the exposure near the yards is interstratified with basalt flows at a higher stratigraphical level, indicating an upward transgression through them of the intrusive mass. The variation in facies can then be easily explained as due to the smaller size of the intrusion at higher levels or to more effective chilling of the margins. However, the occurrence near the yards presents resemblances to another trachyte mass, which occurs on the crest of the ridge, height about 2,250 feet, about three-quarters of a mile to the south. This outcrops immediately above the Pulpit Rock, on the downhill side of the prominent escarpment of basalt and its underlying breccia which marks the northern flank of Saddle Hill; the dip of this basalt is between 5° and 7°. The trachyte outcrop extends along the crest of the ridge for about four chains. It reaches down the slope towards the Pulpit Rock for about a chain, when an undisturbed basalt in position cuts it off. The relation of the mass to the basalt of Saddle Hill is obscure; on the ridge to the north—that is, toward the yards, it is also cut off by basalt breccia and flows in position; while on the slope towards Okuti Valley the ground is soon covered by soil and debris, and the true extent of the mass in that direction cannot be determined, but basalt on both sides narrows the exposed width

to less than two chains. It does not appear to have any close connection with Pulpit Rock, though this absence of connection may be illusory. The form of the occurrence suggests that it is a small, independent intrusion extending down the slope of the Okuti Valley from just beyond the crest of the ridge, or perhaps an offshoot of the intrusion at Pulpit Rock, as suggested for the occurrence near the yards. It does look something like a flow, but if so, it is the solitary case known from the area. Although the connection with the Panama Rock is not obvious, I think it represents an off-shoot from that mass. The exposed surface is coloured brown owing to the oxidation of the ferrous constituent, and is marked by denticulations like those of the trachytes referred to earlier. The rock is very hard, dark-coloured, and is a fine-grained aegerine-augite trachyte, without phenocrysts or with very small ones, and with a base composed of much augite in stumpy laths and alkali felspars in short, rarely lath shaped forms. This description applies to the rock exposed near the yards further north on the ridge. The two occurrences thus differ to some extent from the Pulpit Rock, but such difference may be due to marginal chilling or to the relative smallness of the intrusions or extrusions, if indeed they are actual flows; the latter suggestion I do not endorse. It is possible that others of similar character may ultimately be located on the slopes of Saddle Hill. Since writing the above, I have observed from the upper slope of Saddle Hill a mass resembling the Pulpit Rock in appearance and with similar orientation, outcropping in the upper part of Wainui Valley, but was unable to examine it closely. The large trachyte dyke, referred to on page 241 as occurring in this locality, meets it at its south-eastern end, and may be genetically connected with it. On the northern side of Reynolds Gully, on the south-western slopes of Saddle Hill, there is a massive upstanding occurrence of basalt, only a few chains in length and oriented towards Onawe. At first sight it looks like an intrusion, but at the base of the northern face basaltic breccia shows underneath the exposed mass, so no doubt it is a remnant of a thick flow of basalt. Analyses of the smaller trachyte intrusions, belonging either to the basement beds or to those of the cone, as well as one of a basic dyke near the neck at Onawe, are given in two articles by myself (Speight, 1923, p. 149; and 1940, pp. 71–2). The composition of these trachytes dykes is closely related to that of the massive intrusions. The four main trachyte intrusions have much in common. They are all aegerine-augite trachytes with similar mineral and chemical composition, although they differ slightly in texture. All, except View Hill, have analogous minor surface features such as denticulations, and major features such as pronounced vertical partings. The flaky outcrops of the limited exposures of View Hill suggest that it might also have similar features if the surface had been similarly exposed to the weather. The intrusions are all short in length, of great relative width, and all but that at the Devils Gap dome-like in form. These resemblances point to a similar origin, and this has already been

Akaroa Caldera from the summit of Mount Bossu, view looking north; Wainui in the foreground; Duvauchelles Peak. View Hill, and Okains Peak in the distance. Akaroa Caldera from the slopes of Rocky Peak, view looking south-east; Barrys Bay and Onawe in the middle distance; the harbour entrance in the far distance on the right.

Eastern ridge of Little River Valley, view looking east across Lake Forsyth. Inclined lava flows. originating from Akaroa, show on the hillsides, and gullied slopes cut in loess, show just above the shore of the lake. Photo: J. A. Bartrum. Thin-bedded flat lava flows, on cliff near Dam Rogers.

Panorama of Little River Valley, taken from Te Oka Saddle, view ranging from west to north. Main valley in the middle distance; Western Valley beyond it in the distance; Puaha Valley behind the ridge in the middle; Okuti Valley in the foreground. The ridge in the left foreground is merely a bastion extending a short distance from the eastern wall of the valley. It will be seen that this landscape is entirely different from those shown in Plate 30.

Panama Rock, with dyke stretching south-west from it; view taken from the Summit Road at the head of Le Bones Valley. Photo: Miss Thelma Kent. Devils Gap, looking south-east. Trachyte intrusion on the left, and massive basalt flow on the right.

Pulpit Rock, view looking south-east from Okuti Ridge near the yards; Saddle Hill to the right; the extension of the rock to the north-west hes under the face covered with serub. Trachyte intrusion on the west side of Peraki Valley; Devils Gap at the top of the picture; trachyte bastion towards the floor of the valley below the Gap. View from the east side of Peraki Valley, looking south-west.

considered at some length in connection with Panama Rock, and perhaps the evidence of this well-exposed occurrence may furnish the clue to the origin of the others—viz., that they represent bulbous expansions, to use Stearns' term, of relatively short dykes, either extruded on the surface or they are laccolitic expansions at depth. In no case have I come across any evidence of alteration of rocks or faulting in their vicinity. The formation of bulbous domes composed of soda-trachyte on the Island of Maui, Hawaii, is attributed by Stearns and Macdonald (op. cit. pp. 310–11) to differentiation arising partly from crystal settling of the heavier elements such as augite and olivine in the sub-surface magma chambers and partly from the transfer of the volatile constituents, particularly the alkalies, to the upper chambers feeding these domes The general similarity in the character of the igneous rocks and of the forms of the Hawaiian volcanoes to those of Banks Peninsula is interesting in this connection, and suggests that these intrusions may be attributed to similar causes This explanation adopts and amplifies that given by Daly (1925, p. 79) when accounting for the numerous trachyte domes of Ascension Island and their association with its basalts. However, the forms of the Akaroa intrusions appear to resemble more closely those of the trachyte masses noted in the same author's account of Saint Helena (1927), as, for example, his drawing of Speery Islet (p. 53) and his account of Great Stone Top (p. 55) which he attributes (p. 91) to “emanation of the alkaline magma through a dyke fissure.” In the case of Panama Rock both the dome and the fissure are plainly visible, the latter indicated by the dyke. The association of trachyte with basalt is so common, especially in the islands of the Pacific, that mention of particular cases is unneccessary. However, the hypothesis just referred to may serve to explain the formation of trachyte dykes as well as bulbous domes at various stages in the history of a basalt volcano. It is inconceivable that all the dykes were injected at one stage in the development of the volcano. If it be granted that they have been discharged from magma chambers at varying intervals, some coming out on the ground-surface and suffering erosion before they were covered up by subsequent eruptions, their infrequency at higher levels and their great number at lower levels will easily be explained. In the case of Akaroa I have not come across types intermediate between trachytes and basalts, but in the Lyttelton area the whole suite which would result from differentiation in a magma chamber are well represented—trachytes, trachy-andesites, andesites, alkaline basalts, and ordinary basalts occurring. Perhaps a more complete examination will show that these intermediate types occur at Akaroa. Besides the intrusions just mentioned there are no doubt others awaiting location. It was not found possible to examine all localities, promising or otherwise, and I feel certain that other occurrences as interesting as the four major intrusions just referred to will be discovered eventually, and those already mentioned examined more thoroughly. They undoubtedly suggest complexities not hitherto suspected in what has up till now been regarded as a simple composite volcanic cone.

6. Age of the Volcano. Owing to the non-existence of sedimentaries in the basement beds of Akaroa one important criterion is not available for the determination of the age of the volcano, and so reliance must be placed very largely on comparisons with Lyttelton. At the head of Lyttelton Harbour there is the following sequence of beds in ascending order lying on an older series of volcanics which in turn rest on grey-wackes of Triassic age—(i) Sandstones and greensands, of Awamoan, that is, Mid.-Tertiary age; (ii) Rhyolites and pitchstones; and (iii) Basic andesites and basalts, of which the cone was built. Between each pair of these occurrences there was a period during which the lower set of beds was heavily eroded, and considering the intervals of time necessary to carry this out, it is reasonable to date the Lyttelton volcano from Late Tertiary times. Now, the basalt flows of Mount Herbert overlie the third series with marked erosional unconformity, and must therefore be of later date. It has been suggested earlier in this account that the Mount Herbert lavas come from the direction of Akaroa and possibly from Akaroa itself. If the latter supposition is correct then the Akaroa volcano must date from Latest Tertiary or even Early Pleistocene times. If the lavas did not come from Akaroa but from another centre in the same direction, such as Kaituna, then reliance must be placed on physiographic evidence as an age criterion, and this clearly indicates that Akaroa is much younger than Lyttelton, so that the age just suggested appears to be quite reasonable. It is unfortunate that no help in this difficulty is furnished by a comparison with the later volcanics of other regions of Canterbury, such as Timaru, Geraldine, the Malvern Hills, or View Hill near Oxford. The later rocks of these areas, though undoubtedly basic, differ in texture and composition from those of Akaroa, and must have been extruded under conditions and possibly at times differing from those of the Banks Peninsula volcano. 7. Surface Deposits, etc. From time to time references have been made to occurrences of diatomite in the area under consideration, and one near Wainui has recently been described by Willett (1944, p. 90). It is extremely probable that small deposits may be located in the future, since temporary ponds must have been formed occasionally in times gone by, especially behind rock falls and landslips, but these cannot be of large extent, and their discovery will no doubt be largely a matter of chance On the slopes facing both ways near the saddle between Wainui and Peraki valleys lie areas favouring such deposits. The soil covering of the western and northern slopes facing the harbour have been fully dealt with by C. S. and A. C. Harris (1939, pp. 1–10). A point of special geological interest is the distribution of loess. This forms deposits up to 50 feet thick near sea-level, the amount gradually declining at immediately higher levels, and then disappearing entirely, to be resumed on the flat ridges near the summit. This statement appears to be generally true not only on the internal but also on the external slopes of the cone. Plate 31 shows a thick deposit on the shore of Lake Forsyth. I can confirm the general truth of the statement that it may be thick on gentle summit slopes even up to heights of over 2,000 feet.

There is another matter which concerns surface deposits and the origin of the Akaroa caldera. If the basin has been formed by a great explosion then some remains of the material ejected should be found on the external slopes of the cone; but I have found no trace whatsoever. It is inconceivable that all the fragments necessarily produced by such an explosion and scattered in thick drifts could have been removed by denudation and leave not a trace behind. For this reason alone I consider the explosion hypothesis untenable. Some consideration should be given to the possibility of earth movements following on the decline and extinction of volcanic activity, but there is no clear evidence of such. The irregularity in the direction and amount of dip to be observed on the shore-line round the middle and upper part of the harbour can be explained as resulting from irregularities arising from conditions obtaining during an earlier volcanic episode or immediately following it. A survey of the dip of the beds in the upper part of the cone gives no definite indication of incipient warping or other crustal deformation. Some of the lava flows round the crater-ring do exhibit at times a moderately steep inclination, but this is a purely local feature. The drowning of the harbour may indicate a settling of the land after volcanic action had ceased, but it more probably indicates a rising of sea-level following on the recession of the ice after the Pleistocene glaciation; and there is no reason why both factors should not have been simultaneously operative. The shore-platforms round the harbour seem to be rather high to have resulted from marine erosion at present land-level, and it is possible that a slight rise has set in during recent times, but this amount cannot exceed a maximum of two feet. The physiography and the origin of the landscape forms typical of the area, such as the mode of formation of the great basin which now occupies the heart of the volcano, and the erosional forms of the exterior surface, have been dealt with previously. 8. Composition of Lava Flows and Comments Thereon. The Akaroa lavas—lying above the basement of plutonics, trachytes, and basic rocks—appear to be entirely basalts. Some criteria for differentiation of basalts from andesites would insist on some of them being assigned to the latter group, and these points will be mentioned later as individual cases are considered. From the mineralogical point of view there does not appear to be any feature occurring in the earlier rocks, such as the relative abundance of olivine, which is not matched in the later rocks, and the same applies to textures. While finer-grained types appear to be more common in the later flows, they occur quite freely in the earlier ones. In order to determine if there was any significant variation in chemical composition as extrusion proceeded, specimens for analysis were collected at various heights and in different localities, and these have been analysed by Mr. F. T. Seelye, of the Dominion Laboratory. Results already published include three from near sea-level (Speight, 1924, p. 260, No. 3; and 1940, p. 71, Nos. 4 and 5). The first of these, from just north of Duvauchelles Wharf, has a composition analogous to that of rocks at higher levels, although from stratigraphical

considerations it should perhaps be assigned to the basement rocks of the volcano; the second is entirely different and can certainly be assigned to the basement beds; while the third presents anomalies owing to the presence of a magnesian or iron carbonate which render it useless for comparative purposes. Nos. 1 and 2 on page 260 are of rocks from intermediate levels, the former from about 600 feet above sea-level, and the latter from about 1,600 feet, and they show close relationships in composition to those in the accompanying list. These two rocks were called andesites, following the criterion used by Washington for rocks on the andesite-basalt borderline. He calls andesites those basic rocks containing over 62.5 per cent. of normative felspar, and those with a less percentage he calls basalts. However, the Akaroa rocks which contain both labradorite and olivine, even if they have a higher percentage of normative felspar than 62.5 per cent. are called basalts in this account. Analyses of Akaroa Lavas. Analyst: F. T. Seelye. 1 2 3 4 5 6 SiO2 48.25 44.45 46.01 46.56 47.53 49.25 Al2O3 16.73 15.07 16.34 16.25 16.31 16.53 Fe2O3 3.86 3.37 3.63 3.37 4.10 4.16 FeO 8.00 9.76 8.81 8.73 8.10 8.11 TiO2 2.65 3.05 2.86 3.17 2.93 2.45 MgO 3.96 7.77 5.19 4.92 4.65 3.28 CaO 7.45 9.70 8.62 8.84 7.87 7.14 Na2O 4.46 2.77 4.07 3.85 4.02 5.05 K2O 1.83 1.01 1.53 1.43 1.35 1.72 H2O+ 1.28 1.56 1.26 1.50 0.94 0.62 H2O- 0.30 0.74 0.34 0.42 1.00 0.30 CO2 0.17 0.12 0.03 0.02 0.03 0.02 P2O5 0.94 0.50 0.92 0.81 0.89 1.07 V2O3 0.01 0.028 0.02 0.025 0.018 0.01 ZrO2 nt. fd. nt. fd. nt. fd. nt. fd. nt. fd. nt. fd. Cr2O3 nt. fd. 0.03 0.015 0.02 0.015 nt. fd. MnO 0.17 0.17 0.18 0.17 0.18 0.21 NiO nt. fd. trace 0.01 trace nt. fd. nt. fd. BaO 0.04 0.03 0.04 0.03 0.03 0.07 * Determined spectrographically directly on a sample of the rock. SrO < 0.01 0.07 0.07 0.06 0.06 0.05 S 0.05 0.02 0.03 0.03 0.02 0.02 Cl 0.01 0.01 0.03 0.02 trace 0.03 100.17 100.23 100.00 100.22 100.04 100.09 C.I.P.W. Norms and Symbols. 1 2 3 4 5 6 Q. or. 10.80 5.96 9.02 8.46 7.96 10.19 ab. 35.23 21.81 26.53 28.63 34.03 38.33 an. 20.25 25.70 21.80 22.84 22.47 17.35 ne. 1.34 0.88 4.29 2.13 2.38 di. 7.80 14.95 12.18 12.84 8.71 9.17 hy. 2.82 ol. 9.90 16.34 11.51 10.44 8.36 8.41 mt. 5.51 4.88 5.21 4.84 5.95 6.04 il. 5.04 5.80 5.43 6.03 5.57 4.66 ap. 2.22 1.18 2.19 1.92 2.12 2.52 pr. 0.09 0.95 0.05 (cc) (0.39) (0.27) No. 1. Basalt, Shoreline, Otipua Hill, opposite township, Akaroa. II. 5. 3. 4.—Andose. No. 2. Basalt, Shoreline, Green Point, Akaroa. III. 5. (3) 4. 4.—Auvergnose. No. 3. Basalt, Quarry near Hill Top, height about 1,600 feet. II (III). 5″. 3. 4.—Andose. No. 4. Basalt, Okains Peak, height 1,880 feet. II (III). 5. 3. 4.—Andose. No. 5. Basalt, Devils Gap, Peraki, height 2,400 feet. II”. 5. 3. 4″.—Andose. No. 6. Andesite, Devils Knob, Birdlings Flat, near sea-level. II, 5, (2) 3. 4″.—Andose.

The following comments may be made on these results. First of all, it must be noted that earlier in this account it was suggested on strategraphical grounds that No. 1 might belong to the basement beds of the cone and its slightly more acid character than its position demands can be satisfactorily explained. The basicity of No. 2, from near sea-level and the earliest of the cone rocks of this series, confirms the description already given (Speight, 1924, p. 262), and it accords with the fact that blebs of olivine up to half an inch in diameter show in the hand specimen. No. 3, from a quarry near the Hill Top (height 1600 feet) is an undoubted olivine basalt, containing 4.40 per cent. of normaive nepheline. Although the slide does not respond as a whole to staining, certain small anisotropic patches of irregular form do take the stain, and they may perhaps be nepheline. No. 4, from Okains Peak, is an olivine basalt containing both olivine and labradorite, as also is No. 5, from the Devils Gap, although it contains over 62.5 per cent. of normative felspar. No. 6, from Birdlings Flat, is the most interesting occurrence. In the hand specimen the rock is fine-grained, with definite flow structure, and with the peculiar grey tint that sometimes marks an alkaline basalt. Under the microscope it appears to be composed of a mesh of andesine laths, augite grains and laths, little olivine in small grains, but phenocrysts are practically absent, those present being of untwinned felspar (andesine) and little larger than the laths of the base. Some small rectangular forms that respond to staining may be nepheline. This rock is certainly the latest of those listed, and it was discharged towards the close of the Akaroa volcanic episode. Further it presents an interesting parallel as regards composition to the latest flow on the summit of Mount Herbert, classed as oliogoclase andesite (op. cit. pp. 252–3, No. 4, and p. 257). The similarity in composition, etc., serves to strengthen the suggestion made earlier that the flows of Mount Herbert may have come from the Akaroa centre. Of course this similarity may be only a coincidence, but taken in conjunction with other evidence it is certainly suggestive. Leaving out No. 1, since there is some doubt as to its true stratigraphical position, there is a steady increase in acidity from the earlier flows to the later. However, the two specimens from intermediate levels mentioned earlier present some anomalies. No. 2 from 600 feet on the Barrys Bay Road fits in quite well, although it would be called an andesite according to the Washington criterion. But No. 1, from 1,600 feet on the Stony Bay Road, is more acidic than the rock from the Hill Top at approximately the same height. All the same it fits in quite well before the rock from Okains Peak and that from the Devils Gap. So the statement appears to be fairly accurate. I wish in conclusion to express my indebtedness to various people. First of all, I should like to thank Dr. Henderson, Director of the N.Z. Geological Survey, for his kindly interest, and especially Mr. F. T. Seelye, of the Dominion Laboratory, for his excellent analyses, the last of many such which he has done at my suggestion in connection with various investigations I have undertaken during the past few years, and I wish to express my sincerest appreciation of his valuable help. I have also to thank Miss Thelma Kent, A.R.P.S., and Professor Bartrum for photographs; Professor Cotton for suggestions

and the loan of literature; and, finally, Mr. F. O. Waymouth, for much assistance in connection with transport to remote corners of the Peninsula. Ciiief References. Benson, W. N., 1941. Cainozoic Petrographic Provinces in New Zealand and their Residual Magmas. Am. Jour. Sci., vol. ccxxxix, pp. 537–552. Benson, W. N. and Turner, F. J., 1939. Mineralogical Notes from the University of Otago—No. 2. Trans. Roy. Soc. N.Z., vol. lxix, pp. 56–72. Cotton, C. A., 1941. Some Volcanic Land-Forms in New Zealand. Journ. of Geomorphology, vol. iv, no. 4, pp. 297–306. —– 1944. Volcanoes as Landscape Forms. Whitcombe and Tombs, Christchurch. Daly, R. A., 1914. Igneous Rocks and their Origin. —– 1925. The Geology of Ascension Island. Proc. Amer. Acad. Arts and Sciences, vol. lx, no. 1. —– 1927. The Geology of Saint Helena Island. Proc. Acad. Arts and Sciences, vol. lxii, no. 2. Haast, Julius von, 1879. Geology of Canterbury and Westland, pp. 324–354. Harris, C. S. and A. C., 1939. Soil Survey of Duvauchelle Bay-Wainui District, Banks Peninsula. Soil Survey Division, Dept. Sci. and Indust. Research, Publication no. 1. Hutton, F. W., 1885. Sketch of the Geology of New Zealand. Q.J.G.S., vol. xli, p. 216. Speight, R., 1917. The Geology of Banks Peninsula. Trans. N.Z. Inst., vol. xlix, pp. 365–92. —– 1923. The Intrusive Rocks of Banks Peninsula. Rec. Cant. Mus., vol. ii, part 3, pp. 121–50. —– 1924. The Basic Volcanic Rocks of Banks Peninsula. Rec. Cant. Mus., vol. ii, part 4, pp. 239–67. —– 1933. The Source of the Mount Herbert Lavas. Rec. Cant. Mus., vol. iv, pp. 41–51. —– 1940. The Basal Beds of the Akaroa Volcano. Trans. Roy. Soc. N.Z., vol. lxx, pp. 67–76. —– 1943. The Geology of Banks Peninsula—Part I. Physiography. Trans. Roy. Soc. N.Z., vol. lxxiii, part 1, pp. 13–26. Stearns, H. T. and Macdonald, G. A., 1942. Geology and Ground-water Resources of the Island of Maui, Hawaii, Terr. of Hawaii, Div. of Hydrography, vol. vii. Willett, R. W., 1944. Diatomaceous Earth, Wainui, Akaroa, N.Z. Journ. Sci. and Technology B, vol. xxv, no. 2, p. 90.