Fourteen years ago I wrote an informal account of the geological history of New Zealand and its biota which was published in Tuatara (vol. 2). Since then, many advances have been made in New Zealand geology and in our knowledge of the fauna and flora and their predecessors in the geological past. In our age of specialisation ‘ the seamless coat of learning has been so rent and tattered ’ that it is difficult to synthesise data even from the several branches of Natural History. The attempt is worth while, however, to bring out points of agreement, misunderstanding or conflict between the separate disciplines.

The Biological Society, Victoria University of Wellington, is grateful to the Secretary, Department of Scientific and Industrial Research, for supplying the blocks for figures 6-11 and 13-15.

Continental and Oceanic Crust

A geologist's attitude to possible former land extensions depends a lot on his geophysical theory. The earth has two contrasting types of crust, continental and oceanic. Continental crust (or sial) consists of lighter rocks, roughly granitic in composition and density, averaging about 35 km. thick and generally standing above sea level or submerged to only moderate depths (less than 1,500 fathoms). The sial layer rests on a layer of denser rock, richer in iron and magnesia, known as sima. In oceanic crust, on the other hand, sial is very thin or lacking, and a thin layer of sediment lies directly on sima; oceanic crust is found in the ocean basins and under some of the mid-ocean ‘ rises ’. It does not come to the surface as land, but many volcanoes have erupted from simatic ocean bottoms and some rise above sea level as volcanic islands. Gravity studies suggest that continental crust blocks are floating in the denser sima, so that they ride high, and that the oceanic crust of sima could not ride high enough to form land areas. This is the geophysical basis of belief in the ‘ permanence of ocean basins ’ — deep-sea parts of the earth's crust have always been so, because there is no acceptable hypothesis whereby they could rise so high as to become land, nor whereby a former area of continental sial could have been changed to oceanic sima. Likewise, sialic crust, once formed, must ‘ ride high ’ in the crust, like a cork in water, as land or shallow sea. It is generally accepted that the amount of sialic crust has increased during earth history. If continental crust, once formed, is permanent, we cannot infer land connections across the oceans.

Geosynclines and Continental Accretion

Along linear belts, mainly along the borders of continental masses, the crust has sagged to receive from the adjacent land thick sediments which have later been deformed and welded on to the adjacent continent to increase its area. Several processes have been assumed gradually to concentrate the more siliceous parts of sima to form sial. The initial sag of many geosynclines seems to have been at the boundary between sial and sima at a line of weakness, which tends to migrate outward with the growing continental margin, but other geosynclines have formed between two areas of sial. America and Australia have grown during geological time by addition of marginal geosynclines. Island archipelagos, such as New Zealand and Japan, are narrow areas of sial with long geosynclinal histories, bordered on both sides by simatic ocean floor. Are they embryonic continents initially formed along lines of weakness within the sima with a linear rising ridge on at least one side, supplying the geosyncline with abundant sediment? Are they old continents in the process of changing to oceanic sima? Or are page 55 they dismembered fragments of other continents? If so, how could the segments of oceanic crust be formed between them and their continental neighbours? A Russian geologist, Belousov, who spoke at Victoria University in 1961, believes that small areas within the boundaries of the sial, such as parts of the Black Sea and Adriatic, have been ‘ oceanised ’ or converted to sima, but he has not suggested a mechanism. Geophysicists have found that Campbell Plateau, south of New Zealand, has crust about 20 km. thick; granite and schist suggest that the crust was once fully continental and has been partly oceanised.

Continental Drift

The most popular hypothesis for the dismemberment of continents and formation of separating oceanic segments is that of Alfred Wegener which has been modified in various ways by later writers. In the concept of continental drift, the sialic continental blocks were linked in a single Pangaea or two units in the Paleozoic and have subsequently drifted apart leaving secondary oceans in the gaps.

Paleomagnetism

Recently, paleomagnetic studies of dated rocks have convinced many physicists and geologists that drift has occurred. Results suggest considerable displacement of the magnetic poles. Data from different continents indicate different curves of polar wandering, and quite large amounts of drift are needed to make them coincident. When ancient climates (shown by fossil corals, or glacial beds) are compared with the ancient magnetic latitudes deduced from paleomagnetic studies, they show more reasonable patterns if continental drift has taken place (Blackett, 1961). The evidence is not unambiguous, however, much less proof of any particular pattern of drift.

Evidence for polar wandering and continental drift applies mainly to the Paleozoic, and to a lesser extent to the Mesozoic. In late Tertiary times, the earth's magnetic field was very close to that of the present, and few geologists advocate considerable continental drift during the Tertiary (when it would be important to biogeographers). Some paleomagnetic data suggest post-Cretaceous drift northwards of Australia and India, and some botanists invoke drift as late as Quaternary to explain plant distribution.

Transcurrent or Slip Faults

In mobile parts of the world geologists have recognised shears along which crust segments have slipped horizontally, and have inferred movements in the order of hundreds of miles since the Jurassic. One such shear-fault passes through New Zealand, others page 56 through California and the adjacent east Pacific floor. Demonstrated movement on such shear-faults is not proof of continental drift of Wegenerian type, as it does not entail the movement apart of sialic blocks, and it is on too small a scale in later geologic time to have major biogeographic significance. Professor S. W. Carey has reconstructed the former relationships of New Zealand and the submarine ridges by reversing the postulated movement of the fault, straightening out the curved arcs, and closing the intervening basins attributed to subsequent drift.

Geological Time Scale

To unravel the history of the biota, it is essential to have some knowledge of the geological time scale in relation to the dates of origin of different groups of plants and animals and to their probable dates of dispersal. Fig. 1 shows a geological time scale for New Zealand rocks, calibrated against an absolute time scale determined by radio-isotopes (Kulp, 1961).

Early Paleozoic

Older Paleozoic rocks, deformed and altered, are now preserved only in small areas of north-west Nelson and Fiordland, but they are important in showing that New Zealand's geosynclinal history goes back to these early times. Geologists now map some New Zealand rocks as Precambrian. Cambrian trilobites, ostracods, brachiopods, and sponges from Cobb River have diverse affinities in Queensland, Manchuria and the Baltic, which do not indicate any notable provincialism in the earliest known New Zealand fauna.¹ The more widespread Ordovician fossils (mainly graptolites, but a few brachiopods, trilobites, and other crustaceans) though most like those of Victoria, are also virtually cosmopolitan in their affinities.

By the Lower Devonian, the supply of sediment to the New Zealand region died down; restricted outcrops in Nelson include quartz sandstones and coral limestones implying that the border-lands were no longer rapidly rising, and were reduced to a low relief (peneplain). The Baton River and Reefton faunas include brachiopods and corals, with a few trilobites, molluscs, polyzoa, and a sponge, at first compared with well-known Devonian fossils of the Rhine and Bohemia (Allan; Shirley), but later studies have shown an expectable close relationship with East Australia. Gill (1952) therefore showed a shore-line between the Tasman and New Zealand geosynclines round a central Tasmantis land in the Tasman Sea. Land is unlikely to have occupied the south Tasman as it is floored

¹ Cambrian fossils shown in Fig. 2 are illustrated from textbook sources, as New Zealand specimens have not yet been described.

page 57

Fig. 1 : Geological Time Scale, showing important events in the fossil record, absolute time-scale (after Kulp, 1961), and main divisions of New Zealand rocks.

page 58 with oceanic crust. In Du Toit's reconstruction of Pangaea (before drifting) a continuous Samfrau Geosyncline is drawn to include Argentina, Falkland Islands, Cape of Good Hope, West Antarctica and East Australia, but Australasian Devonian faunas do not indicate such close relationship. A few brachiopoda, however, link the Tasman Sea Devonian with the African-Argentinian Devonian (‘Malvinocaffric Province’), and one of them has recently been found in the Horlick Mountains of Antarctica, near the base of the Beacon Sandstone sequence of non-marine rocks that contains much evidence of relationship between southern continents in the later Paleozoic. They are the first trace of an Austral fauna but what they mean in terms of paleogeography is not clear.

Later Paleozoic : The New Zealand Geosyncline

A fundamental reorganisation of New Zealand structure and the pattern of sedimentation followed the Devonian. The next sediments are extremely thick deposits accumulated east of the foreland of older rocks (of which Fiordland, coastal Westland and north-west Nelson are the emergent parts), in a rapidly subsiding geosyncline embracing the remaining parts of the South and all the North Island. The name New Zealand Geosyncline has generally been applied to this structure, but some geologists would extend the term to include the earlier rocks dealt with in the previous section.

The oldest rocks in the New Zealand Geosyncline are the unfossiliferous Haast Schists of Otago, Westland and Marlborough Sounds, and a narrow strip in the Kaimanawa Range, perhaps locally as old as Carboniferous but of uncertain age because of their great alteration. In Permian, Triassic and Jurassic periods Wellman distinguished a marginal facies of fossiliferous rocks deposited in shelf conditions on the west side of the geosyncline from an axial facies¹ of contemporary but poorly fossiliferous greywackes, more indurated, altered, and folded, deposited near the axis of most rapid sinking in the geosyncline (Fig. 3).

The belt of marginal facies runs from south-east Otago, through Southland to the Alpine Fault near the Holyford valley. It is absent for 300 miles along the Alpine Fault, but reappears from near Tophouse to Stephen Island and is presumably continuous northwards, as its Triassic and Jurassic representatives reappear again from Awakino Gorge north to Port Waikato. The long interruption along the Alpine Fault has been attributed by Wellman to trans-current movement of about 300 miles since the Jurassic, and although there are opponents of this view (e.g. Kingma) Wellman's hypothesis has received considerable support. To reconstruct the

¹ The terms marginal and axial facies are preferred to the terms Hokonui and Alpine facies used by Wellman.

page break

Fig. 2 : Geological Time Scale, with New Zealand fossils, modified from original drawing by Dr. J Marwick for National Broadcasting Service.

page break page 59 New Zealand Geosyncline, therefore, Carey moved back the parts that have been wrenched apart. Such a reconstruction is not attempted in this paper. As the rocks of the geosyncline have been folded steeply, the geosynclinal belt was probably a good deal wider than it now is, Possibly, too, its course was originally straight, like most modern ocean trenches, and its present pronounced swan's-neck curve is due to later movements. Carey's reconstruction brings in a further postulate — that triangular ocean basins (like the Fiji basin) have arisen by the drift apart of ridges on either side to form ‘sphenochasms’.

Beyond the main islands, the course of the geosyncline is speculative. The western foreland can be traced through Stewart Island and the Snares to the Auckland Islands. The geosynclinal sediments reappear at New Caledonia in the north, and at Chatham Island to the east.¹ East of New Zealand, granite at Bounty Island suggests that the foreland continues north-east to bound the sigmoid geosyncline to the south (Fig. 3).

The geographical framework sketched in Fig. 3 takes no account of the hypothesis that the New Zealand area of ‘sial’ has subsequently drifted away from the Australian continent. The geosyncline is extended north to New Caledonia but not south to West Antarctica and the Andes, although it is one of a series of more or less contemporary, if not continuous, down-welts that cicrle the Pacific and joins the Tethyan geosyncline that runs from Indonesia to the Himalayas. Ocean contours have been used to place the western and southern shore of the geanticline, but the depths of the New Caledonia and Bounty basins (presumably oceanic crust), are assumed to have been formed by later downfolding (apologies to the geophysicists!). Likewise, the Kermadec-Tonga ridge system (which has crust of continental thickness), is attributed to later movements.

Permian

The oldest dated rocks in the geosyncline, in the Takitimu Mountains and in Nelson-Marlborough, are volcanic sediments from a volcanic island arc parallel to its margins. Later deposits in Southland and Nelson are thick limestone, banded mudstone, conglomerate and sandstone in the marginal facies, and indurated greywackes of the axial facies in Canterbury and Northland.

Apart from fragments, the oldest New Zealand plants, from Southland (McQueen), are Permian Equisetites, Cladophlebis, Sphenopteris, ? Linguifolium, and Noeggerathiopsis mostly like those of Australia, reminding us that the Permian is pre-eminently the period when other southern lands showed the similarities that led to the concept of a continental Gondwanaland linking them together.

At this time, too, parts of southern continents were periodically glaciated. The two most characteristic Gondwana plants, Glossopteris and Gangamopteris, have not yet been found in New Zealand, and our marine sediments show no evidence of glaciation, but the lower Permian brachiopods, bivalves and corals include a strong southern cool-water element related to those of south-eastern Australia and to a lesser extent South Africa, South America and India, and contrasting with the Tethyan (tropical) faunas of the same period. New Zealand, however, received some elements from the tropics (J. B. Waterhouse, pers, comm.) and the Upper Permian reef corals and fusulinid Foraminifera of North Auckland are a Tethyan warm-water assemblage.

In the Permian, therefore, we see clearly for the first time two biogeographic elements that recur with different representatives and to varying extent in later times: (1) A cool, Southern Hemisphere, more or less circum-polar ‘Austral’ element, perhaps distributed by westerly winds and currents to the separate southern lands (or round the polar coasts of Gondwanaland if a drift hypothesis is favoured). (2) A tropical or subtropical Tethyan or Indo-Pacific element distributed from equatorial regions.

Much of later biogeographic history in New Zealand concerns the interplay of these two elements. The Glossopteris flora, though it reached to India, is a cool southern element.¹ Its presence in Antarctica when southern continents were glaciated, is argument for drift and polar wandering, but its distribution in the south, and the temperature gradients implied by the distribution of cool and Tethyan facies in New Zealand and Australia imply that the directions ‘polewards’ and ‘equatorwards’ were approximately the same as now.

R. Florin, a Swedish paleobotanist, concluded that since the Jurassic and perhaps since the Permian, southern conifers, particularly the Podocarps, had been distinguished from those of the northern hemisphere. Possibly, therefore, the ancestors of the modern New Zealand podocarps and kauri were already established here in the Permian, but their fossil record begins somewhat later. Some older groups of land invertebrates may be equally long established, but of this too there is no fossil evidence.

Triassic — Lower Jurassic

The New Zealand Geosyncline continued to sink and fill during the Triassic at an increasing rate, so that limestones (which depend on lack of detritus from land), are practically absent. The main sediments are the altered and strongly folded greywackes that now form the backbone of New Zealand, including the Southern Alps. Many geologists believe they were deposited near the centre of the

¹ Gangamopteris, however, is reported from the Pyrenées (Millot).

page 61

Fig. 3 : The New Zealand Geosyncline in the Triassic; a reconstruction based on Wellman (1952) and Grindley (1961) for New Zealand, extended to New Caledonia, Auckland and Chatham Islands.

page 62 geosynclinal trough (Fig. 3) by turbidity currents from its steep sides, an environment unfavourable for bottom organisms. The axial greywackes contain few traces of life: tube-worms (Terebellina), plant fragments, and odd fossils transported from marginal shelves or dropped from the pelagic zone. In the western marginal facies, however, bottom-dwelling brachiopods and molluscs abounded. The western foreland rose rapidly to supply abundant fresh sediment, and at times deltas advanced on to the shelves and even towards the axis so that plant remains are preserved at Waitaki Valley and Mt. Potts (Canterbury). Boulders in Triassic conglomerates near Marakopa are so large that they must have come from a steep rocky coast not far west of the present coast. Volcanoes were active on the western land, supplying ash to the geosyncline, Some of the younger granites of west Nelson were perhaps emplaced at this time, like the Bounty Island granite to the south-east.

The Triassic rocks and fossils of New Caledonia are so like those of New Zealand that the geosyncline and its border ridge probably extended at least to there, but geology cannot prove that the bridge was continuous land. In Eastern Australia, on the other hand, the Triassic sediments are very different: non-marine quartz sandstones from a stable continent which give little evidence of a Tasman Sea, but also speak against the close proximity of the actively rising ridge that supplied the New Zealand greywackes.

Many Triassic and Lower Jurassic bottom-dwelling brachiopod and mollusc genera are confined to New Zealand, or occur elsewhere only in New Caledonia, and the two countries formed a marine faunal province in Triassic time — the Maorian Province, especially characterised by its endemic archaic brachiopods. Apparently the Maorian Province was somewhat isolated from the Tethyan Province, and this is indirect evidence against Triassic land connection to Indonesia. On the other hand, ammonites, which probably lived in the surface waters, and pterioid bivalves¹ with pelagic larvae migrated freely. There are no distinct faunal provinces among Triassic and Lower Jurassic ammonites but local faunas vary in richness, our own being relatively poor in species and genera. The pterioids, Daonella, Halobia, Monotis, and Rhaetavicula contorta, spread dramatically at successive stages in the Mid and Upper Triassic so that similar forms occur in distant parts of the globe. Groups less easily dispersed suggest that the Pacific and Arctic oceans supported similar faunas (the Pacific-Arctic Realm, of which the Maorian Province is a part). The Tethys, from South-East Asia to the Mediterranean, had a somewhat different fauna, and the shallow seas that flooded northern Eurasia had an impoverished fauna related to the Tethyan.

The concept of a Pacific-Arctic Realm, contrasted with those of the Tethys and Europe, may be illustrated by the history of the Trigoniidae, a group of bivalves that flourished in the Jurassic and Cretaceous and are thought to have been derived from the Upper Paleozoic and Triassic family Myophoriidae. The first Trigoniidae occur in the middle Triassic of Chile and New Zealand, and they became diverse (at the generic level) in the Upper Triassic of curcum-Pacific lands: Peru, western North America, Alaska, the Arctic, Japan, Timor, New Caledonia and New Zealand. During this time Myophoriidae took their place in the Tethys and Eurasia, overlapping with Pacific Trigoniids in Japan and North America. Only near the end of the Triassic did an odd Trigoniid reach Europe, heralding their subsequent dramatic replacement of Myophoriids in the Jurassic. Under the Continental Drift hypothesis, the ‘Pacific-Arctic Realm’ would occupy the peripheral coasts of the restored Pangaea, which at first sight seems reasonable, but this realm would completely surround the Tethys and Muschelkalk realms that extended irregularly across Pangaea. A faunal province entirely surrounded by another faunal province seems an unlikely situation.

The only New Zealand Triassic vertebrate fossil is a large marine ichthyosaur, Mixosaurus hectori. No Triassic land fauna is known, but vegetation of Triassic and Lower to Middle Jurassic time is quite well represented and includes our earliest Araucarians and Podocarps. in addition to extinct, typically Mesozoic forms (Equisitales, Cycadophytes, Ginkgoales). Probably some archaic New Zealand animals were already established by the Triassic or Jurassic, in particular ancestors of the Tuatara (Sphenodon), so closely related to Jurassic Homoeosaurus, native frog Leiopelma (relations in Upper Jurassic) and certain archaic invertebrates. Climatically, New Zealand lay outside the presumably tropical zones of Triassic reef-corals and desert sandstones but there is no sign of cold nor of circum-Antarctic affinities.

Similar geological conditions and faunal affinities persisted into the Lower Jurassic (Lias) : the Maorian Province can still be detected in the endemic benthos in New Zealand and New Caledonia (e.g. Clavigera, Otapiria, Pseudaucella) whereas pelagic ammonites entered freely. Isolation of the benthos is further shown by the absence from New Zealand of Jurassic belemnites and trigoniids until after the Lias.

Middle and Upper Jurassic

The New Zealand Geosyncline's history continued through the Jurassic with some premonitions of change in the disposition of the basins of deposition due to the start of its evagination to form an elevated area (Wellman, Grindley). Another pterioid immigration (Meleagrinella) and the appearance of belemnites and trigoniids mark the mid-Jurassic. Upper Jurassic benthic faunas, in contrast page 64 with those of Triassic and Lias, have strong affinities with those of the Tethys (Marwick). Some species (Inoceramus galoi, Buchia malayomaorica) range to the East Indies (hence some have suggested a Malayo-Maorian Province), others have relationships as far away as India (other Buchia species, belemnites, Kutchithyris), and a few are even European species. A few endemic Triassic lineages linger as a minor faunal element but most have disappeared, their places taken by immigrants. For instance, instead of the endemic or Pacific Triassic trigoniids, we have representatives of six European Jurassic trigoniid genera. Some Jurassic elements were practically worldwide rather than strictly Tethyan, but the Tethys seems to have been the chief dispersal avenue. Some groups suggest that there was now a barrier in the Middle East separating the Mediterranean from the Tethys proper, which extended from Persia to the Pacific. Pelagic ammonites, previously world-wide, show three rather distinct faunal provinces (Arkell): Pacific, Boreal, and Tethyan (including New Zealand¹). Tropical reef-corals are not represented in the New Zealand Upper Jurassic, but nevertheless the Tethyan elements were strikingly dominant, in strong contrast with the endemism of the Triassic and Liassic, and the southern affinities of earlier Permian and later Cretaceous faunas.

Middle Jurassic vegetation is represented by the well-known Curio Bay fossil forest of Southland and Upper Jurassic vegetation by the Port Waikato plant fossils (once wrongly thought to include angiosperms). Poorly known conifers include Araucaria (sections Eutacta and Colymbea) and Podocarps (one perhaps a member of section Dacrycarpus which includes the kahikatea).

Rangitata or Post-Hokonui Orogeny

The New Zealand Geosyncline, persistent over most of New Zealand from pre-Permian to Jurassic times, became greatly restricted in the early Cretaceous, when most of its contained sediments, indurated by burial and deformed by folding, were elevated, and underwent erosion as land. During the ten million years or so of Upper Tithonian and Neocomian times, the sea may have retreated beyond our shores, for no fossils of these ages have been found. By Aptian time, when marine fossils take up the record once more, the area of geosynclinal deposition was restricted to the east coast from Raukumara Peninsula south to Marlborough (Fig. 4). The early Cretaceous has always been considered the period of maximum land extension, but all through Permian-Jurassic time the geosyncline was fed by a persistently rising ridge of land to the west, so that the Lower Cretaceous land area was not necessarily enlarged and there is no geological evidence to show how far the New Zealand ridge extended north or south. The land was at first mountainous, but was worn down to low relief before the end of the Cretaceous, when renewed transgressions indicate restriction of the New Zealand lands.

Angiosperms arose and spread throughout the world at the end of the Lower Cretaceous, and birds had developed from Jurassic beginnings sufficiently to take advantage of the more extensive lands formed by the Rangitata Orogeny. But mammals and snakes were also developing apace, and their failure to populate the area in this time of extended land seems to show that the New Zealand ridge was not continuous to New Guinea and Asia during the Cretaceous or later. Among Recent New Zealand land birds, only the ‘Ratites’ (Kiwi and Moas) are in fact ancient enough to have reached here in the Cretaceous. The period of extended land probably helped dispersal of terrestrial invertebrates.

Cretaceous

A non-marine deposit in the north-western South Island (Topfer Formation) classed as Lower Cretaceous contains the oldest known pollen resembling that of the Podocarp Microcachrys (now restricted to Tasmania), and spores of the ferns Gleichenia and Blechnum (Couper).

Marine lower Cretaceous (the Aptian Taitai Series) consists of rapidly deposited mudstones and conglomerates from the Raukumara Peninsula south to Marlborough (Fig. 4). Fossils include one bivalve (Maccoyella) found elsewhere only in Australia and others widely distributed, with Tethyan or cosmopolitan affinities (e.g. Spondylus, Dicranodonta).

The east-coast geosyncline persisted in the ‘middle’ Cretaceous (Clarence Series) and the sea reached the Chatham Islands, which may have formerly been part of the eastern land (Fig. 4), and Northland. Coarse conglomerates of varied rock types show that the land was hilly if not mountainous. ‘Middle’ Cretaceous marine faunas include ammonites and benthic molluscs representing northern species that presumably came by larval dispersal, via Tethys (e.g. the European Hyphantoceras, Inoceramus concentricus, and Exogyra), others with Austral relationships (lotrigonia spp.; Eselaevitrigonia), and some Australasian (belemnites of the Dimitobelidae).

The greatest biogeographic event in the mid Cretaceous was the dispersal of angiosperms, which apparently originated in the Lower Cretaceous. The oldest angiosperm fossils from New Zealand (Couper, 1960) are pollen grains in the Motuan and Ngaterian Stages (about Albian-Turonian), not attributable to any definite plant families, and forming a relatively minor element in a vegetation still dominated by gymnosperms. Their subsequent increase provided ecological niches for the specialisation of many invertebrates that are unlikely to have been dispersed earlier than the plants they were dependent on.

Marine transgression continued in the Upper Cretaceous when the sea reached southward to Dunedin and finally almost to Campbell page 66 Island (Fig. 4). The sea flooded over lowland swamps and gravel plains fringing a land of low relief in which long-continued chemical weathering produced quartz gravels and sands. Flints and other siliceous sediments were deposited in offshore waters to the east. Dr. J. T. Kingma believes that the whole of the southern North Island was submerged. This phase of peneplanation and quartzose sedimentation, which locally lasted into the Tertiary, may be the only really stable phase in New Zealand history since the Devonian. The early Senonian (Raukumara Series) marine faunas are similar in affinities to those of the mid Cretaceous. Late Senonian and Maestrichtian faunas include even stronger Austral elements — pelagic ammonites (e.g. Kossmaticeratidae) and benthonic lamelli-branchs (e.g. Lahillea, Pacitrigonia) and gastropods (Struthioptera) with relations in Chile and Seymour Island (Antarctica); other forms are more widespread, and several genera endemic. Some characteristic Tertiary genera appear for the first time. Vertebrates are represented by several sharks, a Chimaerid and an elephant fish (Callorhynchus) and by quite an impressive list of marine reptiles, including giant Mosasaurs and Plesiosaurs and a reputed Crocodile. Perhaps we may one day find ancestors of Tertiary Penguins in the sediments of this period, their likely time of origin.

In Upper Cretaceous coals, widespread in the South Island, are recorded the rapid rise to dominance of Angiosperms. Nothofagus pollen (of the brassi group) appears first in the lowest Paparoa Beds (about lower Senonian) and of the menziesii group in the Haumurian (Maestrichtian). Pollen of the Proteaceae, including grains attributed to Knightia (also recorded from Graham Land), appears in the Senonian. Several characteristic Podocarps (represented by pollen attributed to the groups of Dacrydium franklini, now confined to Tasmania, and of D. cupressinum, the rimu) appear for the first time (Couper, 1960). These elements (Nothofagus, Podocarps, Proteaceae) have definite South Temperate relationships, so that the Upper Cretaceous flora of New Zealand seems to have had much in common with that of Graham Land (Cranwell) and with early Tertiary floras of Chile, Australia, and Kerguelen (podocarps only). It would be an attractive hypothesis to attribute these plants to migration on a Lower Cretaceous land extension, with the possibility of land connections or approximations, but their fossil record suggests that they came after the Rangitata mountains were reduced, when the seas were spreading over a subdued land and reducing its extent once more.

Upper Cretaceous climate was probably moist warm-temperate, warmer than the present, judged by the presence in the South Island of the tropical family Olacaceae (represented by the pollen Anacolosidites) and marine reptiles, but the seas were not warm enough to support reef rudistids and corals. Increasing peneplanation probably caused deep podsolisation and slow replenishment page 67

Fig. 4 : Cretaceous Geography : an interpretation in part based on Wellman (1959) and Grindley (1961), extended to outlying islands.

page 68 of soil nutriments, and perhaps lowered rainfall through reduction of topographic relief. It is most unlikely that there were any subalpine or even montane biotopes in New Zealand during the period from Maestrichtian to Oligocene.

Cenozoic

In New Zealand, it is inconvenient to separate the Tertiary and Quaternary because deposition continued without interruption in many areas. Compared with many countries, the record of life in shallow seas is remarkably continuous throughout the Cenozoic and the record of terrestrial plant life is also very complete though not yet fully studied. New Zealand geologists have used a local system of Cenozoic Stages (Fig. 5) to classify the rocks and fossils of this era, because even in Europe the boundaries between the periods (Miocene, Pliocene, etc.) are uncertain and world correlation is doubtful. The mapping symbols are a convenient short-hand for the stage names and will be used in the following sections.

Structurally, and therefore geographically, Tertiary New Zealand became a good deal different from Mesozoic New Zealand. The pattern of folds, welts, and troughs that developed was on a finer scale than in the Mesozoic; instead of a broad trough some one or two hundred miles wide and thousands long, the land moved up and down as a series of narrow short interfingering or branching folds. The welts, which tended to be submarine ridges or land, were small, so we can think of Tertiary New Zealand as an archipelago. Changes in geography were frequent. Troughs sank rapidly but filled with sediment as they subsided so were seldom submerged to abyssal depths. Welts rose in complementary fashion, but owing to constant erosion were not mountainous. A kind of writhing of part of the mobile Pacific margin seems to have gone on in the later Tertiary. A changing archipelago would encourage speciation, i.e. the formation of two or more species by successive invasions, geographic isolation, re-invasions or back-invasions of populations from one island to another. This could account for genera with a multitude of species. Long ago F. W. Hutton, who had many advanced ideas on evolution, suggested that the diversification of the Moas (Dinornithiformes) was due to an archipelagic stage in New Zealand history. Other groups that show adaptive radiation (e.g. the Callaeidae among birds) may have developed on our changing archipelago just as the Galapagos finches (Geospizinae) and Hawaiian Drepaniidae did on their ancient volcanic island groups.

Paleocene — Eocene

The sea continued to lap further on to the peneplained land in the Paleocene and early Eocene. For instance, Paleocene seas lapped over the Proto-Chatham Island shown in Fig. 4, and are known to page 69

Figure 5
Table Of Divisions Of The New Zealand Cenozoic
The succession of series and stages has been established for New Zealand sediments and fossils. Mapping symbols (third column) are convenient abbreviations for the stages. Probable correlations with divisions of geological time in other parts of the world are shown. The Okehuan, Hautawan and Waipipian (Wanganui Series), formerly treated as substages, are here given stage rank. A few examples of sediments are given in the fourth column. For fuller descriptions see fascicule 4 (New Zealand) of Lexique Stratigraphique International, vol. 6 (Oceania).

page 70 have covered Campbell Island and at least parts of Chatham Rise. For the most part, however, the main outline of Upper Cretaceous geography persisted into the Paleocene and Lower Eocene, with increasing peneplanation on land and deposition either of globigerina ooze (now the Amuri Limestone) offshore, mainly in moderate depths (typically in North Canterbury but extending to Campbell Island and at times to east Wairarapa) or of fine siliceous silt and mud (Northland, East Coast) which includes turbidites and perhaps locally abyssal deposits in the down-sagging basins.

Marginal Paleocene deposits at Eyre River (Canterbury) and the Chathams contain orbitoid Foraminifera (Asterocyclina etc.) indicating at least subtropical seas. The benthic fauna (e.g. at Wangaloa) is highly endemic, containing descendants of Cretaceous forms, including members of Southern Ocean groups (Perissodonta, ancestor of Struthiolaria) but some of the newcomers in it suggest derivation from the north (e.g. Costacallista, Polinices, Sigaretotrema, Priscoficus). The earliest known fossil penguin (from Gore Bay, Canterbury) is Lower Eocene.

On land, Cretaceous podocarps still dominated the earliest Tertiary forest vegetation, but angiosperms become more important and incomers at this time include Nothofagus of the fusca group, additional Proteaceae, probably Casuarina¹ (now extinct here), the first Myrtaceae, and a member of the tropical-subtropical tribe Cupanieae, a mixed bunch, therefore, apparently from both the tropical north and from other parts of the Austral Realm.

In the Middle and Late Eocene, there is a change in geography. Hitherto, the position of the west coast of New Zealand has been unknown. In the Bortonian, the sea flooded on to coal-measure lowlands near Greymouth and in the Late Eocene pushed northwards to Nelson, leaving a large projection in north-west Nelson which I have called Karamea Peninsula (Fig. 6). In the eastern Bortonian sea (as from Paleocene onward) foraminiferal limestone, like globigerina ooze of the modern sea bottom, was deposited in East Marlborough, North Canterbury and Campbell Island, in fairly deep water away from coastal influence, and shallower limestones covered part at least of the Proto-Chatham Island. West and north of this calcareous sea bottom was a belt of more variable fine calcareous muds, glauconitic shales and sands in the subsiding East Coast and Northland basins, and south-eastward, on the more stable shallow shelf of Canterbury and Otago, a belt of glauconitic sand (green-sand) merging into glauconitic mudstone in deeper water at Hampden (North Otago). Glauconite formation required special biochemical conditions in seas where terrigenous detritus was minimal. On land, quartz gravels and sands formed wide swampy plains bordering sluggish meandering rivers draining a low land of deep leached podsols.

Fig. 6 : Middle and Upper Eocene Geography, showing main types of sediment of Bortonian Stage and transgression in Kaiatan and Runangan times.

Late Eocene seas locally transgressed further over this stable peneplain. Its stability was probably only relative, however. In the newly formed Kaiata Gulf inside ‘Karamea Peninsula’ dark carbonaceous silt was deposited above coal measures, bordered by land that rose quickly enough to form steep cliffed coasts locally in late Eocene time (Omotumotu). In the Waikato Basin, too, there was appreciable relief in the coal-measure landscape.

The Bortonian marine shelf fauna is one of the most distinctive in New Zealand's Tertiary history. It contains several restricted endemic genera and the first members of many common Tertiary genera.

The new gastropod arrivals include members of the Cypraeidae, Architectonica, Mitridae, Eocithara (a harp-shell), Conidae, Gemmula, suggesting dominance of Malayo-Pacific dispersal avenues and subtropical climates. Neilo and Speightia (aff. Andicula of Peru) are Austral immigrants at this time. A notable invasion of coral genera took place but its members do not indicate sea temperatures higher than 50° to 60° F. (Squires, 1958). The most important of many Foraminifera which entered in the Middle and Upper Eocene is Hantkenina, which had a world-wide distribution, and in the Upper Eocene a marked warming of our coasts almost to tropical conditions (Hornibrook, 1953) is suggested by the return of the Orbitoid Asterocyclina, the appearance of the cidarid Eucidaris (at the Chatham Islands; Fell, 1954), the regular echinoid Brochopleurus. and the lamellibranch Hinnites. Although the chief source of immigrants was the Indo-West Pacific, some Middle and Upper Eocene Foraminifera, corals (Discotrochus, Asterosmilia) and crabs (Laeviranina, Portunites; Glaessner, 1960) show rather notable central American relationships.

Despite evidence of warmth and tropical influence Eocene seas were probably not as warm as those of the Miocene. American paleontologists believe north-east Pacific seas gradually cooled after an Eocene maximum. Here is an unexplained anomaly.

On land, to judge by pollen, the Arnold Period saw the appearance of additional Nothofagus species of the fusca and brassi groups, of additional Proteaceous plants (including one with pollen indistinguishable from Banksia or Dryandra¹), and of Dracophyllum, Hoheria, Phormium (endemic), Elytranthe, Dysoxylum, and Rhopalostylis (Malayo-Pacific). and Laurelia (Austral fide Skotts-berg). Dominant trees included ? Casuarina, and increasing numbers of Nothofagus of the brassi group, now living in New Guinea and New Caledonia, which tended to replace the podocarps characteristic of older vegetation (Couper, 1960). No single direction of immigration can account for the Eocene newcomers in the fossil pollen record — Australian, Malayo-Pacific and Austral elements are all represented.

Oligocene

In the past decade there has been much uncertainty about the upper boundary of the Oligocene. Paleontologists have now transferred the Pareora Series to the Miocene. The distinction between Whaingaroan and Duntroonian Stages is ill-defined but they are kept distict here as their distribution makes reasonable sense to the paleogeographer.

The transgression that began in the Upper Cretaceous reached its climax in the Landon Epoch, except near Proto-Chatham Island, which remained emergent, Whaingaroan-Duntroonian seas flooded over the Late Eocene coal swamps of Northland, the Waikato, King Country and Taranaki, eastward at least to National Park, and persisted in the Eastern Geosyncline. Several districts not reached by the Whaingaroan were flooded by the Duntroonian sea. Karamea Peninsula was submerged, the South Island land either reduced to a narrow strip or obliterated in its northern parts. The lithologic and faunal facies of many Whaingaroan and Duntroonian marls, limestones and chalks (impure ‘globigerina ooze’) confirms the impression that land was distant or very low-lying. Geologists are quite uncertain about the whereabouts of land at the height of the Landon transgression, but glauconitic sandstones west of National Park and near the west coast, flanking the main development of limestone (Te Kuiti Group), suggest low islands on either side. In central New Zealand, Whaingaroan-Duntroonian calcareous mudstones and limestones are widely distributed (including remnants infaulted by later earth movements near Paraparaumu, Picton, the Upper Grey Valley, Brechin Burn and Smite River, Canterbury), so that the sea perhaps extended widely across the axis of New Zealand in this area (Fig. 7). In Otago, Duntroonian shallow seas and estuaries lapped inland to Pomahaka, Naseby, and Waikaia, but this maximum transgression of the Tertiary was shallow and temporary and probably did not submerge the whole land. In most districts, the same conditions persisted into the Waitakian, with local signs, however, of more rapid earth movements. Thus in Auckland Province, limestone gave way to Mahoenui mudstone and graded-bedded sandstone suggesting an increase of sediment from rising land, and sea retreated from Central Otago.

Many genera appear for the first time as fossils in the Landon Epoch, particularly in the Duntroonian and Waitakian, which yield faunas at Chatton (Ld), Wharekuri (Ld), and Otiake (Lw) in the Waitaki Valley. Newcomers include elements of Australian affinity (Bassina, Eucrassatella, Zenatia) and many endemic forms of unknown origin, but the majority are Indo-Pacific genera of warm-waters: Solecurtus, Maoricardium (living in the Indian Ocean), Astrea, Pyrazus, Xenophora, Ficus, Arca (s.str.), Isognomon, Bathytoma. The characteristic New Zealand Struthiolaria appears (Ld), a descendant of the middle Eocene Monalaria, but intermediate page 74

Fig. 7 : Oligocene Paleogeography, based on the distribution of sediments of the Whaingaroan and Duntroonian stages. Oligocene penguins, though commonest in North Otago, are known as far north as Kawhia in the North Island. Corrigendum : The panel marked ‘Limestones (shallow facies)’ should show coarse dots in a brickwork pattern.

stages apparently lived beyond our shores. The cidarids Phyllacanthus and Eucidaris, the regular echinoid Grammechinus, irregulars such as Echinolampas and Planilampas, abundant Alcyonarians (Graphularia, Isis, Moltkia) and Scleractinian solitary corals (species of Notocyathus, Stephanocyathus, Conocyathus, etc.), emphasise the Indo-Pacific relationships of the fauna, but absence of orbitoids. nummulites and reef-corals suggests that conditions were not fully tropical. Indeed the abundance of fossil penguins and whales (zeuglodonts, squalodonts) in the Oligocene gives an impression of coolness; the extinct subfamilies of Spheniscidae, undoubtedly an Austral group of late Cretaceous origin, seem to have been tolerant of at least warm water, and whales in general have wide tolerance. These were the basis of an earlier suggestion that the Oligocene was relatively cool.

Conditions on the reduced and low-lying lands of the Landon Period were similar to those of the late Eocene: subtropical forests with Nothofagus of the brassi groups as the dominant trees, ? Casuarina also common, but conifers, fusca-group beeches and ferns rare, suggest a savanna climate drier than at most times in New Zealand's history. Judged by pollen, the plants gained during this period included ex-tropical groups like Bombacaceae, Pseudo-nintera, and Freycinetia, additional Proteaceae (Australian?), Astelia and other dominantly southern forms (e.g. Restionaceae), as if both circum-Austral and Malayo-Pacific dispersal avenues were open for migrants. Other newcomers are Coprosma, Myrsine, Epilobium and Typha. The first Compositae pollen appeared throughout the world in the Oligocene, apparently as a result of the last spectacular evolutionary explosion in the flowering plants.

Lower Miocene

With revised correlations, the Otaian to Clifdenian are included in the Lower Miocene, and the well-known Waitemata and Pakaurangi Point beds of Northland are now thought to be Otaian.

The increased topographic relief heralded in the Waitakian of the King Country continued and extended to the rest of the country in the Miocene. Short gaps in the record suggest local withdrawals and readvances as the restless land began to stir. In many areas, sandstones and unweathered mudstones derived from rising land overlie Landon limestones. In the south, limestones and glauconitic sediments at first persisted. In Northland, the Waitemata Sandstone was deposited in shallow water, east of a rising ridge composed of Paleozoic granitic rocks (or perhaps of Triassic conglomerates) which was later the site of a chain of andesite volcanoes from Auckland to Hokianga, source of the Manukau Breccias. To the east, from Hunua to Kawau, sea transgressed over greywacke land that had been emergent long enough to lose nearly all trace of the early Tertiary peneplain and its covering Oligocene sediments.

Later in the Miocene, volcanoes were also active on this eastern ridge, producing the rocks now forming picturesque peaks at Coromandel, Great Barrier, Whangarei and Whangaroa. (Two ridges bounding Northland perhaps persisted from Cretaceous to Miocene as suggested by R. F. Hay, but the Oligocene sediments show that they were then very subdued if not submerged.) The Waitemata sea penetrated into the Waikato Basin from the west coast south of Port Waikato, perhaps linking southwards with the deeper Mahoenui sea, which occupied the southern King Country and Taranaki, and extended from north-west Nelson beyond Punakaike (Fig. 8), depositing calcareous and polyzoal sands, and silts, and occupied the Waiau Syncline whence shallow-water limestones extend east to Forest Hill. In eastern Otago and Canterbury, a shallow shelf of sand, locally glauconitic, bordered fossiliferous silts (Blue Cliffs). We cannot be certain whether the Otaian sea occupied all the Eastern Geosyncline, where Pareora microfossils have rarely been identified.

During the succeeding Hutchinsonian and Awamoan stages, rising lands, local volcanic outbursts and the consequent increase in sedimentation caused the sea to retreat from most of Northland and parts of the King Country (where the Mokau coals formed in swampy deltas), and perhaps also from parts of Southland. Apart from volcanoes, however, the land was not yet mountainous.

The revised correlation of the Pakaurangi Point beds from G. H. Scott's foraminiferal studies of the Haeuslerella bioseries gives us a picture of two strongly contrasting climatic provinces in the shallow-water Lower Miocene marine fauna. In the eastern South Island, the molluscan fauna does not greatly differ from that of the preceding late Oligocene; in Northland, however, a subtropical fauna is marked by the first appearance of many new genera, including Australian (e.g. Dimya) and especially Indo-Pacific elements, including Conidae, Murex (s.str.), Chicoreus, Pterynotus, Coralliophila, Oniscidia (gastropods), Ctenoides, Septifer, Cardita (bivalves), the only reef corals known in the Tertiary record (for instance Lobophyllia, Turbinaria and Alveopora; Squires) and orbitoid Foraminifera (Lepidocyclina, etc.). All told, the Otaian marks the high-point of mid Tertiary warmth and of the Indo-Pacific faunal influence that it brought. Additional genera of Australian and Malayo-Pacific origin had reached the eastern South Island by the Awamoan: Eumarcia, Cleidothaerus (Australian), Lutraria, Cymatiella, Aphera, Cronia, and others that first appeared in the Otaian of Northland. About 175 mollusc genera appeared for the first time in the New Zealand record during the three Pareora stages, most of Indo-Pacific affinity, but some distinctly Australian, and a few of Austral distribution, now characteristically Subantarctic (Hochstetteria, Kidderia). The marked Lower Miocene influx, spread over the Otaian to Altonian Stages, is also shown by corals, Foraminifera and Ostracoda.

Fig. 8 : Lower Miocene Paleogeography, based on the distribution of sediments of the Otaian Stage. When the map was drawn, the writer overlooked Wellman's evidence (1951, N.Z.G.S. Bull. 48) that the Tititira Formation of South Westland (? Lower Miocene) was derived from the west, suggesting a western ridge (near the words ‘no data’) and a marine basin over the western flank of the present Southern Alps.

Lower Miocene (Pareora) vegetation differs from that of the Oligocene in an increase of Nothofagus of the fusca group, podocarps, palms, and ferns, but Nothofagus of the brassi group persisted as dominant trees. Newcomers include Fuchsia (an Austral genus, otherwise mainly South American but with one montane species in Tahiti). Dacrydium of the kirkii-bidwillii group, and other species of Podocarps and Proteaceae (perhaps descendants of older local forms). The vegetational changes may reflect increased rainfall caused by higher topography.

The Altonian and Clifdenian stages (Southland Series) are also Early Miocene. Their marine sediments suggest some geographic changes — sea retreat in Northland, in the face of continued volcanic activity, but renewed transgression of seas which deposited Mokau Sandstone over the Mokau coals and over the western ridge, north of Taranaki, the thick offshore Ihungia Beds in the Eastern Geosyncline, and similar formations in Westland, north-west Nelson. and western Southland, after the minor emergences of the Pareora Series. Reinvigorated earth movements probably raised some of the anticlines of Westland and the Eastern Geosyncline as shoals and islands, separated by straits. Climate remained tropical to subtropical. On land, Nothofagus of the brassi group dominated the forests, and the jusca group, podocarps, and ferns became rare once more, suggesting a return to drier conditions as in the Oligocene.¹ The well-known fossil Cocos of Coopers Beach reinforces the marine faunal evidence for marginal-tropical climate in the Miocene.

Middle and Upper Miocene

In most districts, the Lillburnian is marked by coarser deposits than the preceding stages, reflecting movement of the sedimentary basins into shallower shelf zones as tectonic and geographic relief increased. The sea had left the central trough of Northland and most of Auckland, which henceforth were relatively stable. Sea remained south of a line from Awakino almost to National Park, and its western part received ash from volcanoes on the south end of the western ridge. Andesite volcanoes remained active on the east (Great Barrier-Coromandel). The Eastern Geosyncline received shallow off-shore sand with thin conglomerate beds (Tutamoe Formation), and graded beds in sinking troughs near Lake Waikaremoana.

Some Early Tertiary geosynclinal troughs had become ‘evaginated’ to form rising anticlines in the Late Tertiary (Gage). and Macpherson's insistence (1946) that such anticlines persisted during Tertiary deposition is most justified in the latter half of the epoch. On flanks of rising structures, an Upper Tertiary stage may

¹ We can expect some changes in ideas of the sequence of vegetation when Middle Tertiary plant fossils are better known than they are at present.

page 79

Fig. 9: Upper Miocene Geography, based on sediments of the Waiauan and Tongaporutuan Stages.

page 80 transgress unconformably over older rocks that had recently been emergent and eroded, making it difficult to draw shore-lines. Synclinal or fault-angle troughs between anticlinal shoals or islands subsided rapidly as tectonic relief increased, sometimes reaching even archibenthal depths, but generally they were filled as they subsided with thick sediments derived from the ‘highs’. deposited in only moderate shelf depths. Most Miocene sediments do not suggest that adjacent lands were truly mountainous but the Great Marlborough Conglomerate and the thick Longford Formation (Murchison) accumulated mainly in fresh water, near rising mountans. Peneplains of late Tertiary age — plain-like surfaces of low relief cut by erosion across older rocks — have been recognised in Otago. Fiordland and the King Country, so mountains were local.

South of the shallow Lillburnian seas in Marlborough, sediments of deeper facies were deposited both near the present coast at Kekerengu, Amuri Bluff, and Gore Bay (North Canterbury), and inland near Mandamus. About this time, the sea extended to within twenty miles of the present divide near the head of the Esk River (Wilson, 1956) but had retreated from the Weka Pass District, from Otago-South Canterbury, and from some of the West Coast basins. It persisted in the Waiau Syncline of Southland.

During the Waiauan (Fig. 9), the sea remained in most of the districts of Lillburnian sedimentation and transgressed over some areas of Lillburnian emergence — for instance at Weka Pass, Fox River, Cape Foulwind, and on the margin of the stable schist platform at Otago Peninsula. Tongaporutuan beds followed conformably in the main basins, but locally the sea transgressed further, on the west flank of the Eastern Geosyncline towards Dannevirke in the north and Awatere Valley in the south. In Hawke's Bay, rapidly deposited graded sandstones filled deep depressions attributed to active transcurrent faulting during phases of accelerated earth movement (Kingma). Some have suggested that rhyolitic tuffs near Mahia and Porangahau came from volcanoes off the east coast, but a source in the Taupo Zone is more likely. The Kapitean, youngest Miocene stage, is represented by marine beds in the same areas, but its seas spread further, towards Manawatu Gorge, and from Taranaki towards the west Taupo mountains, heralding the Pliocene climax of late Tertiary transgression. An acute spasm of mountain-building in Marlborough led to thick conglomerate deposits in Awatere district.

In the middle and late Miocene successive marine bottom faunas record a progressive cooling of the seas around New Zealand, by the extinction of subtropical genera that had marked the Mid Tertiary thermal maximum, e.g. the corals Platyhelia, Dendrophyllia, Flabellum pavoninum distinctum and Stephanocyathus ixine (still living in the Indo-Pacific) : the echinoid Schizaster and others; the mollusca Oniscidea, Conospirus, Hinnites, Solecurtus, Aturia and others of longer residence including endemic genera (such as Hedecardium, page 81 Serripecten, Magnatica, Neocola); the brachiopods Lingula, Magadina, Rhizothyris; and the commonest Tertiary crab (Tumidocarcinus, endemic to Australasia). Temperatures may have fluctuated but remained higher than now, as shown by the persistence of Cucullaea and the cidarid Phyllacanthus into the Kapitean. Despite cooling conditions, Malayo-Pacific and Australian immigrants continued to turn up: Olivella (SI), Pitar (SI), Trachy-eardium (Tk), Katelysia (SI), which had short-lived existence here, and others which remain as characteristic members of the Recent fauna: Striacallista (Tk), Zeacumantus (Tt), Atrina (SI), Mactra (SI). and the spider crab Leptomithrax (Tt). Immigrant Foramini-fera also came from the Indo-Pacific (e.g. Bolivinita quadrilatera, Tt) and from Australia, so that assemblages on both sides of the Tasman are more alike than previously (Hornibrook). Some mollusc incomers are southern or endemic forms which had perhaps begun to move north to New Zealand latitudes as the climate cooled: Cratis (SI), Aeneator (Tk), Buccinulum (Sw).

On land, the fusca group of Nothofagus, podocarps, and ferns (including tree-ferns) increased in the later Miocene suggesting increasing rainfall. The youngest New Zealand fossil coconut (Sw) was collected in Hawke's Bay by J. T. Kingma. Pollen of ? Casuarina remained common until the end of the period, but a number of other long-ranging Tertiary plants became extinct near the Miocene-Pliocene boundary, including members of the tropical groups Bombacaceae and Cupanieae. Incomers indicated by pollen include the last known Proteaceous addition to the flora but it did not survive the Pleistocene. The oldest moa bones recorded by Oliver (1949) are from the Upper Wanganui Valley (Maungapurua), apparently from Kapitean sediments.

Pliocene

Earth movements became even more intense during the Pliocene, approaching the climax of the Kaikoura Orogeny to which New Zealand owes its present geography. Differential movements were most acute in a mobile belt trending from Raukumara Peninsula obliquely south-westward through the Southern Alps to Fiordland, and decreased towards the more stable districts of Northland and Otago-Stewart Island on either side. The last major marine transgression, begun in upper Miocene times, reached its maximum in the Pliocene. In stable Northland the sea flooded the Manukau-Lower Waikato depression, as far south as Huntly and Meremere. Sea occupied Wharekahika Graben at East Cape. The southern third of the North Island was widely inundated, between the persistently subsiding Wanganui-Hawkes Bay basins, with extensions north to Wairoa and south to Palliser Bay. Deep-water Opoitian mudstones (continental slope according to P. V. Vella) suggest that coastal ranges east of Eketahuna were not elevated as continuous land till page 82 late Wanganuian time. Between these areas marine straits crossed sags separating the rising axial ranges, which were sometimes emergent, supplying coarse sediment, at other times reduced to islands and shoals. It is uncertain whether the Wellington area (flanked by sea in the Wairarapa and at Makara) formed another island or was linked to the emergent areas of Marlborough Sounds and Nelson as shown in Fig. 10. On the east side of the South Island, marine Pliocene beds are present in the Awatere and in Canterbury south to Timaru (here buried by later deposits) and, on the West Coast from Karamea to Resolution Island; in the south, a shallow sea still penetrated parts of the Waiau Syncline. The lands were not as mountainous as today but relief increased at the end of the epoch when schist from the alpine backbone was exposed to erosion and gravels from the rising South Island mountains drove back the coasts. Rhyolite volcanoes were active in the central North Island probably at first in the Coromandel-Thames district, but not as far south as Taupo until the Pleistocene. Volcanoes at Dunedin, Solander Island, the Auckland and Campbell Islands were probably built up largely in the Pliocene, Bank's Peninsula a little later. Chatham and Campbell Islands were emergent during at least part of the Miocene, were lapped by Pliocene seas and were the sites of Pliocene volcanoes.

Pliocene marine faunas (Opoitian-Waipipian¹) retained many long-lived Tertiary genera, some warm-water ones such as the corals Lophelia, Caryophyllia, the bivalves Arca, Eucrassatella, Miltha, Trachycardium, and the gastropods Polinices, Sconsia, Olivella, and even gained additional forms from the north such as the Foramini-fera Bulimina echinata, Globorotalia inflata, and Globigerinoides rubra and the Mollusca Glycymerula (Wo), Patro (Ww) and Zethalia (Ww). Pliocene sea temperatures in New Zealand were evidently nowhere cooler than those of modern Northland. But the most dramatic faunal event of Pliocene time was the progressive extinction of Miocene warm-water groups: the Haeuslerella lineage, the Globorotalia miozea lineage and other Foraminifera; Gemmula, Maoricardium, Mauicassis and many other mollusca (including some of those listed in the previous sentence); ostracods such as Loxoconcha australis and Cytherelloidea auricula. Other characteristic mid-Tertiary marine invertebrates, known or inferred to have been stenothermal in warm waters, had disappeared before the beginning of the Pliocene. Gaps in the fauna were only partly filled by newcomers to the fossil record. Ancestors of many groups still living appeared, some of them dominant organisms of our shores: for instance Flabellum rubrum (Sceleractinia); Amphidesma (Taria), Spisula, Cyclomactra, Notirus (Bivalves) : Nerita, Euthrena,

¹ To prevent ambiguity, the Waipipian. Hautawan, and Okehuan are treated as stages in this paper, and the Waitotaran, Nukumaruan and Castlecliffian stages are restricted to the beds in their type sections, formerly classed as the Mangapanian. Marahauan and Putikian substages.

page 83

Fig. 10 : Pliocene Geography, based on sediments of the Opoitian and Waipipian stages.

page 84 Cominella (s.str.), Penion of the adusta group, Paratrophon, Zeatrophon, Amphibola (Gastropoda); Coronula (whale-barnacle); Cancer novaezealandiae, Ovalipes, Hemiplax (Crabs); Fellaster (Echinoidea) became abundant. Some of the Pliocene immigrants are of Australian or northern origin but others are clearly Austral groups of cool water, distribution by the West Wind Drift (e.g. Lissarca, Aulacomya), and others ‘cryptogenetic’ — endemic genera of unknown ancestry, Some molluscan genera with a tranquil history of mid Tertiary stability showed a minor burst of speciation in the Pliocene (e.g. Bassina). At about this time the echinoid Pseudechinus arose in Australasia from Brochopleurus, and became distributed in a southern circumpolar zone by epi-planktonic dispersal in the West Wind Drift (Fell, 1962). A Pliocene seal, Arctocephalus caninus Berry, has been described from near Cape Kidnappers (Wo).

Terrestrial plant fossils (Couper) also give evidence of cooling climate, but even in latest Pliocene time the temperature was probably warmer than at present, judged by the dominance as far south as Dunedin of Nothofagus of the brassi group, now surviving in New Guinea and New Caledonia. Other dominant forest plants are the fusca group of beeches and podocarps. Locally the steno-thermal Dysoxylum (kohekohe), Knightia (rewarewa), and Agathis (kauri) were common, together with the last New Zealand species of Ephedra and of the tribe Cupanieae, neither of which survived the Pliocene. The pollen of ? Casuarina, common in middle Tertiary preparations, is so rare in the Pliocene that its parent plant must have been greatly reduced in numbers. Newcomers to the flora, judged by pollen, are lxerba,¹ Quintinia, Salicornia and Cyathea medullaris, thus mostly of Australian or northern origin. Couper and McQueen found that Pliocene samples show geographical differentiation of vegetation not evident in Miocene samples.

Early Pleistocene

Hautawan: Marine Hautawan deposits are largely confined to areas flanking the axial ranges of the central part of the mobile belt — Wanganui Basin on the west and the complex of synclines from Hawke's Bay to North Canterbury on the east. The Pliocene Manawatu Strait linking these areas was soon temporarily bridged by a gravel spit fed by coarse detritus from the rising Ruahine-Tararua Ranges. Extensive and thick aggradation gravels from young mountains were deposited in the Reikorangi and Kaitoke districts of Wellington, in Nelson (Moutere Gravels) and on both sides of the Southern Alps (Kowai and Old Man Gravels). Locally they contain cold climate pollen assemblages, the last-named moraine and varve-silts of the first or Ross Glaciation. Other non-marine Hautawan deposits have been recognised by cool-climate plant fossils in the Waikato, on the West Auckland coast, at Timaru, and Invercargill. The extent of cooling is shown by the presence of Nothofagus solandri var. cliffortioides (Mountain Beech) and N. menziesii (Silver Beech) near Port Waikato at the coast, and widespread subalpine grassland and scrub at low altitudes in the South Island. Geographic differentiation of ‘botanical districts’ is thus evident.

Marine invertebrate groups now confined to subantarctic waters came north to the southern North Island: the fan shell Chlamys delicatula, the cockle Tawera subsulcata, the gastropods Zeacolpus (Stiracolpus) symmetricus, Maurea couperi (aff. spectabile), Eucominia aff. nassoides, Trophon s.str., the crab Jacquinotia (pers. comm. D. Rodley and J. Yaldwyn). and perhaps the echinoid Notocidaris vellai Fell. Dr. J. A. Berry has identified a jaw bone from Napier as Ommatophoca, the genus containing the Ross Seal, now strictly Antarctic. Despite these indications of cold climate, a number of warmth-loving organisms survived the Ross Glaciation.

Nukumaruan: Seas of this first interglacial age occupied the flanking basins of Wanganui and Hawke's Bay-Wairarapa and were again linked through Manawatu Strait but did not leave identifiable traces in Auckland and South Island districts. Rhyolite volcanism, producing pumice and ignimbrite, began near Taupo. Several warm-water bivalve genera thrived for the last time: Patro, Isognomon, Glycymeris (s.str.), Lutraria and Eumarcia, and a few immigrants from Australia or the north trickled in: Pterynotus angasi, Ellatrivia, Leucotina ambigua.

Pollen assemblages attributed to the Nukumaruan from Waikato, Wanganui, Hawke's Bay and Invercargill contain few extinct species and indicate climates either similar to the present or cooler, those of Wanganui (Maxwell Formation) showing fluctuating climate evidently in early stages of the succeeding cooling. Several plants have their first records in this stage: Libertia, Plantago, Cardiomanes.

Okehuan: Evidence for glacial climate is weaker than in the Hautawan. Sea retreated permanently from most of the Hawke's Bay-Palliser Bay area and temporarly from Wanganui Basin as alluvial aprons spread out from rising ranges and from the central volcanic zone of the North Island, where ignimbrite eruptions produced vast quantities of pumiceous sediment. Okehuan is not known from the South Island. Periglacial vegetation is shown by a pollen sample collected by Mr. G. Neef from the east flank of the Tararua Range (W. F. Harris, pers. comm.), but otherwise evidence for glaciation rests mainly on the extinction of many of the warm-water mollusca that had persisted through the Nukumaruan. A retreat of the sea, assumed to be partly eustatic, intervened between the minor transgressions of the Nukumaruan and Castlecliffian at Wanganui. Okehuan marine faunas, known chiefly from Wanganui, are impoverished, locally even stunted, but include no distinctively cool-water forms. Dr. J. A. Berry has identified fur-seal vertebrae from Wanganui. The Okehuan was probably deposited during the advance of the sea in the period of increasing warmth after a eustatic withdrawal corresponding with the period of maximum cold.

Castlecliffian: The marine fauna suggests temperatures warmer than now. The sea lapped back on to the land in several regions, perhaps in the Manukau depression, certainly in the Bay of Plenty from Matata to Cape Runaway, at Gisborne, Mahia Peninsula and Cape Kidnappers, and occupied restricted parts of Wanganui Basin, where Castlecliffian fossils are best known. Otherwise, Castlecliffian geography did not greatly differ from the present. Earth movements were accelerated. Ignimbrite eruptions continued and the Tongariro volcanoes began to erupt. Immigrant mollusca, mainly from Australia but also from the north, included Xenogalea, Pecten (s.str.), Anadara trapezia, Bembicium melanostomum, Eunaticina, Agnewia, Pterynotus zelandicus (now persistent at Norfolk Island), Zelippistes (now northern Northland). On land, kauri lived near enough to deposit pollen at Wanganui, and the brassi group of beeches flourished for the last time.

Late Pleistocene (Hawera Series)

Owing to earth movements in central New Zealand late Pleistocene sediments are clearly separated from those of the early Pleistocene. Hawera Series deposits are either coastal terraces formed during interglacial stages of high sea level, or alluvial deposits of riverbeds aggraded during glacial stages of low sea level and of rapid erosion.

due to periodic deforestation of the mountains. At least four glaciations and three interglacial intervals contributed to Hawera stratigraphy. Late Pleistocene stratigraphy and chronology are at present very active fields of research.

Shore-line changes of the late Pleistocene were mainly (but not entirely) due to eustatic sea-level changes. In the Last Glaciation (for instance about 20,000 years ago), the sea withdrew to some 350 feet below present levels, perhaps more, so that much of the continental shelf was exposed, and Cook Strait was bridged (Fig. 11). In the Last Interglacial, perhaps 70,000 to 150,000 years ago, the sea reached about 60 feet above its present level, flowing into many lowland embayments and benching steep coasts, depositing marine and estuarine deposits, peats and dune-sands as it retreated again. Earlier interglacial sea levels (Penultimate, Antepenultimate) left their traces at still higher levels.

Each glacial stage contained two or more peaks of cooling (stadials) when ice advanced for some thousands of years, separated by interstadials, when temperatures rose a little. Rainfall was perhaps an independent variable. Each successive glacial and interglacial stage of Hawera time left its characteristic deposits, the subject of current studies (by M. Gage, R. P. Suggate, I. C. McKellar and others). The biogeography of the New Zealand Ice Ages is in its infancy, but their importance to the understanding of the composition, distribution and ecology of the living biota is already clear to geologists and paleontologists, and is being gradually appreciated by zoologists and botanists. The multiplicity of distributional changes caused by repeated climatic fluctuations of varying amplitude (stadial, interstadial, glacial, interglacial) probably had cumulative effects not to be judged by the effect of the most extreme single fluctuation. The extremes of Pleistocene climate, however, are a useful starting point for analysis of the changes caused in each fluctuation, and are as much as we can hope to discuss at the present stage of knowledge and in the scope of this essay.

Interglacial climate is hard to determine for several reasons, perhaps because each interglacial followed a cold period in which stenothermal warmth-loving organisms had been exterminated owing to the limited northward extent of land and shallow seas to provide a refugium. Many interglacial fossil assemblages seem to indicate present climatic conditions, rather than warmer conditions, which we would expect by analogy with interglacial climates of other lands and by inference from de-glaciation necessary to raise sea-level eustatically. Marine interglacial faunas of the Hawera Series contain a few mollusca (e.g. Leucotina, Ellatrivia, Marginella mustellina near Wanganui) south of their present ranges and thus suggestive of warmer sea temperatures. Interglacial Haweran lignites have not yet shown kauri or other Northland plants south of their present ranges. The strongest evidence for an interglacial warmer than the page 88 present is the widespread red-weathering of old soils and land surfaces in Wellington Peninsula, which occurred at an early post-Wanganuian date. Such deep weathering and ‘rubification’ are characteristic of warm regions with strongly seasonal rainfall and savanna vegetation. Even today, Northland is on the southern fringe of the Trade Wind belt, which probably reached south to Cook Strait in at least one interglacial. In each Interglacial, when the sea level was higher than now, many low areas were inundated, but the coasts later prograded and built dune fields as the sea retreated. At such times the main islands were separated from outlying islands, and the lower islets were submerged.

The extreme severity of glacial climate during the Hawera epoch is more easily judged than the mildness of interglacials. During the maximum phase of the Last Glaciation (Otiran Stage) (Fig. 11) permanent snow level was at least 3,500 feet lower than at present (Willett, 1950) and ice tongues carved glacial valleys on either side of the main divide from Fiordland to north-west Nelson and to a minor extent in northern mountains (Tararuas, Egmont, National Park). Glaciers extended below present sea level on the south-west coast of the South Island and in some eastern lake basins. In Westland ice tongues joined to form a continuous Piedmont Glacier. Moraine and varve silts were deposited in the South Island. Fell-field or tundra occupied vast areas in the South and in the central North Island. Periglacial conditions extended below present sea level at Wellington and solifluxion produced mantling deposits of breccia on the hills (Cotton and Te Punga, 1955), gravitating in steep fans to join the valley-plains of rivers which aggraded their beds with coarse debris, shed from the highlands even where there were no glaciers. From the gravel plains wind picked up dust to form a coating of loess on lowlands as far north as the middle North Island, there interfingering with ash from the active volcanoes. The coast was close to the continental shelf edge, linking the main islands and most outlying islets (Fig. 11). Extended rivers crossed the shelf across gravel plains. Banks now submerged were islands or shallow shoals and Antarctic icebergs may have stranded on the Chatham Rise (D. Cullen. 1962). The history of ice-age marine faunas, therefore, lies hidden below sea level, in the field of the submarine geologist. Pantin (1957) has described a shallow-water shellbed, now at about 70 fathoms (420 feet) off Cape Campbell, containing a southern fauna (with Chlamys delicatula) that lived about 19,500 years ago (by 14 C dating).

On land, deposits of glacial stadials commonly contain pollen of grasses and Compositae (probably subalpine daisy scrub) with few if any tree pollens. As many trees are anemophilous and pollens can be blown hundreds of miles from their source, absence from a deposit suggests that no trees lived nearby, but a flood of grass pollen makes it difficult to find the odd tree pollen in such samples. Refuges for forest of montane type on the emergent continental shelf page 89

Fig. 11: New Zealand during the Last Glaciation. The map is an attempt to show conditions during the peak of the Otiran Glaciation.

page 90 of southern New Zealand in glacial times cannot be ruled out but are hard to confirm. Cold climate pollen assemblages without any tree pollen at all are known as far north as Taranaki (pers. comm. W. F. Harris), and others as far north as Hauraki Plains contain so few that forest vegetation must have been scanty or distant. Subalpine Astelia linearis left abundant seeds near present sea level in cold climate deposits near Porirua (20,800 years old) and elsewhere, and in some places is accompanied by phylloclades of Alpine Toatoa (Phyllocladus alpinus). Plant fossils now available thus indicate that ice ages were even more frigid than Willett (1950) predicted on theoretical grounds and than biologists would expect from knowledge of the present biota.

Much collaboration between geologists, botanists and climatologists will be needed to determine the vegetation pattern during a glacial maximum. Were there forest refuges in the south? Was kauri pushed still further north, or is its present southern limit inherited from a glacial maximum? What were the changes in Northland's vegetation during an Ice Age? How rapidly did vegetation adjust to the rapid alterations of stadial and interstadial climate? Zoologists have corresponding problems to challenge speculation, which may one day be guided by research on Pleistocene insect fossils. For instance, did the endemic forest-floor fauna of Fiordland survive glaciation there in a coastal scrub refuge? How old (or young) are geographic races in Paryphanta and other land invertebrates?

Post-Glacial

In Europe pollen analysis established the zonal framework of post-glacial climatic history, a warming interrupted by temporary reversals including the final ‘Younger Dryas’ ice advance of about 10,000 years ago, and has permitted the dating of geographic changes due to the Flandrian transgression of the sea from its low glacial level to a level (or levels) somewhat higher than now (some 5,000 years ago; Thermal Maximum). In New Zealand a pollen zonation has been worked out, including a Thermal Maximum (Cranwell and von Post; Harris), and there is ample evidence (including some radio-carbon dates) of the post-glacial transgression that formed many inlets and harbours. A date from Foxton (Te Punga) suggests that Cook Strait was already a sea-way 9,900 years ago and is thus older than the English Channel. Post-glacial biogeographic changes include the southward and upward migration of vegetation zones (and animals) to reclothe the formerly glaciated and periglaciated areas, and adjustment of plant communities to the lesser fluctuations of later time, including the latest recognisable changes studied by Raeside (1948) and Holloway (1954).

Judged by their lack of fossil record, many Foraminifera, Echinoderms and Mollusca colonised New Zealand in post-glacial page 91 time. Most of the latter are of Australian origin (Saxostrea, Agnewia tritoniformis, Notocochlis migratoria, Tonna tetracotula) and it is likely that many Australian species in groups without fossil record are post-glacial immigrants. Unassisted immigrants among insects (e.g. the Monarch Butterfly) and birds (e.g. White-eye, White-faced Heron, Spur-winged Plover) have colonised in the past century, and many others became established earlier, but perhaps also in postglacial time (e.g. Pukeko, Crested Crebe, Pied Stilt, Grey Duck, Kingfisher). If, however, we subtract from our list of species of land and fresh-water birds the species of known and inferred late arrival from Australia, we are left with rather few species of probable Tertiary history in this country — a result of Pleistocene decimation.

The arrival of Polynesian man about a thousand years ago initiated many changes. At one time, extinction of the moas and associated birds (Aptornis, New Zealand swan, coot, goose, and eagle) was attributed to changes in climate and vegetation, with man merely delivering a coup de grace. Such a hypothesis was tenable (but opposed by the writer (1953) on theoretical grounds) when it seemed that the Moa-hunter knew only Euryapteryx, but radio-carbon dates have shown that other moa groups persisted well into the human period (Lockerbie, 1959), and their extinction can be attributed to the complex changes due to man's appearance. Several plants were also introduced, but on the whole, man's influence was destructive, through predation, agriculture and, above all, fire. Extinctions of large animals marked the spread of man in many parts of the world.

The Fossil Record in New Zealand

Fossil evidence for the history of life is very uneven because different groups or organisms vary widely in their chances of preservation and because the deficiency of the record is exaggerated by neglect of some groups by paleontologists. In the following section the main contributors to the study of New Zealand fossils (sources of data for this paper) are mentioned.

Marine and fresh-water algae have left no record, except for marine calcareous forms (‘Lithothamnion’) and marine and fresh-water diatoms, a neglected field. Abundant coals, lignites and peats, ranging from Jurassic to Recent, yield diagnostic pollen grains and spores of land plants (Couper, Cranwell, Harris, Moar, McIntyre), providing a good basis for study of the Tertiary history of the flora. Leaf, seed, and wood fossils, still not adequately studied by specialists, are a valuable supplement to the pollen record (Arber, Edwards, Oliver, McQueen, Bell, Florin).

Marine protistans are well represented in the fossil record, particularly Foraminifera (Finlay, Hornibrook, Vella), Coccolitho-phorida, Radiolaria, and the biologically obscure hystrichospheres.

Sponges occur sporadically, often as dissociated spicules, but have scarcely been studied as yet (Hinde and Holmes). Marine Coelenterates are well represented by the Tertiary corals (Squires): a few Hydrozoa and Anthozoa, and abundant Zoantheria (Scleractinia). Marine tube-forming Annelida are represented by tubes and shells. Echinoids are common as fossils (Fell), Asteroids and Crinoids rarer. Polyzoa (Brown, Uttley) and Brachiopods (Thomson, Allan) have a very good record.

Shelled Mollusca have the most complete record of all larger invertebrates (Hutton, Suter, Marshall, Murdoch, Finlay, Marwick, Laws, Powell, et al.) but land shells are unknown as fossils in New Zealand before the Pleistocene, and freshwater groups are little known (McMichael). Marine Crustacea are well represented by Ostracoda (Hornibrook) and Cirripedia (Withers): Decapoda are more patchy but of great interest (Glaessner). Other arthropod groups have extremely poor fossil records, and terrestrial groups are hardly represented at all, apart from insects in late Quaternary deposits (no publication).

Sharks and saw-sharks (Pristiophorus) are represented by teeth (Davis, Chapman). Elephant Fish and Chimaerids have left bone fragments in an Upper Cretaceous deposit (Chapman). Rays (Myliobatis, Trygon) have left rare grinding teeth and spines. Bones and scales of Teleost fish are occasionally preserved, but their commonest fossil remains are otoliths (Frost, Stinton) which should eventually give evidence for interpreting the history of the fish fauna. There is a single Pliocene record of the fresh-water Galaxias (Stokell).

Reptilia are known from Middle Triassic, Jurassic and Cretaceous beds. In the late Cretaceous, large marine Mosasaurs and Elasmosaurs were not uncommon. No terrestrial dinosaurs (known in Australia and in Patagonia) have been reported from New Zealand, and living Sphenodon has no fossil record before the Holocene.

Birds have a poor fossil record, except for the penguins. The oldest moa bones (Dinornithiformes) are Upper Miocene (Oliver) and foot-prints, perhaps Apteryx, about the same age. Extinct subfamilies of penguins (Sphenisciformes) are well represented from Lower Eocene to Oligocene and there is a single Pliocene skeleton (Marples), but the living Spheniscinae are unknown before the late Pleistocene. A possible albatross bone (Marples) is known from the Oligocene, but other flying birds have no record here before the Pleistocene.

Two groups of marine mammals are represented by scanty fossils: Cetacea by Archaeoceti (Kekenodon Hector, from the Oligocene), Odontoceti, and Cetotheres (Benham), and modern genera from the Pleistocene; Pinnipedia by a few bones of several genera from the Pliocene and Pleistocene (Dr. J. A. Berry).

Lack of Terrestrial Animal Fossils

Most non-marine deposits in this land of high rainfall and abundant vegetation are chemically acid, so that shells and bones are quickly dissolved. We should therefore be cautious in generalising from the absence of New Zealand fossil land animals, especially as the modern fauna contains ancient vertebrates that must have been here during the Tertiary. If no fossil Tuatara has been found, why should the absence of fossil land dinosaurs and early mammals mean that New Zealand never had a more extensive fauna of land vertebrates? On the other hand, it is difficult to suppose that mammals, once established (whether monotremes, marsupials or placentals) would have subsequently succumbed in a country that supported flightless rails and ratites until the dawn of the human period. Had dinosaurs (or early mammals) reached New Zealand by a land connection at the time Nothofagus came here, we would expect them to have survived as relics or to have radiated adaptively to fill some of the many empty niches in primitive New Zealand.

It is not strictly true to say New Zealand has no fossil land fauna. Pleistocene peats and silts yield remains of insects, and Tertiary coals prepared for pollen studies contain insect fragments too small to study. Future discoveries (in fresh-water shales, or fossil kauri gum), or new techniques, may eventually tell us something of the history of the land fauna.

Biogeography = Phytogeography + Zoogeography

I cannot accept the claim sometimes made that different rules govern distribution of plants and animals so that biogeographic conclusions drawn from (say) fossil shallow-water animals have no bearing on our interpretation of the history of the land fauna and flora. To be sure, dispersal mechanisms and vagility (ability to be dispersed) vary from group to group, and experience shows, for instance, that land plants have been dispersed across sea barriers that mammals could not cross, whereas mammals have readily crossed climatic barriers that limit plant dispersal. Granted this, however, the dispersal avenues for one group of organisms tend to be shared by others. Much dispersal, plant and animal, marine and terrestrial, is apparently due to currents of air and water, which tend to be almost parallel. Thus at times when New Zealand lay in an extended pan-tropical zone, the anti-clockwise circulation of Pacific air and water (a southward deflection of the summer Trade Winds and South Equatorial Current), irrespective of any change in the distribution of land and shallows, raised the chances for successful colonisation by organisms of tropical derivation, Foraminifera or ferns, swordfish or terns, screwpine or skink, crabs or bats. Likewise, at times when the cool West Wind Drift was the prevalent influence page 94 on the New Zealand Archipelago, down-wind and down-current movement favoured circum-polar dispersal of mussels and kowhai, seals and sea-stars, petrels and Petalurids, penguins and podocarps. In an oceanic region, too, marine and atmospheric ‘climates’ march hand in hand, so that a cold New Zealand washed by subtropic seas (or vice versa) is not easily conceivable. Emphasis on wind and sea currents does not mean that land extensions can be discounted, but the land was at its best to receive northern colonists when ex-tropical currents extended south, and likewise most receptive of southerners when the west wind belt moved north.

Ecological Fallacies?

In discussing the history of the New Zealand flora. Cockayne asserted that ‘it is not the species which move but the associations to which they belong’. Holloway's studies of beech invading podocarp forest (1954) do not support this statement in detail, and when applied to the colonisation of oceanic islands it is obviously untenable. The post-glacial Metrosideros forest of Auckland Island, for instance, contains only those few members of similar associations elsewhere that were able to cross 190 miles of sea in the 10,000 years or so since colonisation has been possible. Moreover, species in a community have diverse geographic origins and different geological history. Thus in Nothofagus forest, the beeches are an Austral element dating from the Upper Cretaceous, but their parasitic Elytranthe are of tropical derivation and date from the Upper Eocene here. Examples could be multiplied. Many formerly dominant plants became extinct at the end of the Tertiary, and some now important appeared late in the record (e.g. Cyathea medullaris, Dodonea). Obviously communities sort themselves out from the species available and are not stable in either space or time. Whether New Zealand has ever been anything but an island archipelago is still uncertain, but it will not solve the problem to assume that whole communities had to move as a unit if at all.

Lyel's principle of Uniformitarianism may be deceptive if it leads to the conclusion that organisms are bound to the environments where their successors now live. Kahikatea, often classed as a swamp-forest species, has survived through facultative adaptation to swamp conditions (Holloway) but is not an obligate swamp plant. Notornis and Kakapo are now mountain species, but formerly lived in lowland and coastal areas. Galeodea lived in shallow Miocene seas but persists hardly altered in the archibenthal zone (Dell).

The Alpine Biota

Geological history suggests that the present alpine and montane environment is very young, its topography not older than Miocene and its climate certainly post-Pliocene, yet it supports a diverse flora page 95 and fauna including endemic species and genera. Some of the high mountain species are probably facultative, not obligate, alpines, pre-adapted to the new environment, surviving there by a kind of ecological opportunism. This may apply to taxa that are light-demanding, and formerly lived in savanna conditions during the Tertiary; they could not compete in lowlands when post-glacial rain forests returned. Some ‘facultative’ alpines are in fact also found on coasts, but the problem cannot by any means be solved by such arguments.

The alpine-subalpine environment extended over the whole South Island and some of the North a dozen or so times during the Pleistocene, and offered colonising organisms a variety of empty niches. Into this ecological vacuum, we may assume, came colonists from several sources, perhaps at first including forms with a long history of life in similar conditions in Tertiary Antarctica, able to disperse northwards for the first (and last) time at the onset of the Pleistocene. If so they colonised across the sea, as land connections since the alpine biotope arose are out of the question. At about the same time, according to Couper, podocarps and Nothofagus were colonising New Guinea for the first time, presumably from Australia where they had lived throughout the Tertiary.

Despite these possibilities, I find it hard to reconcile the evidence of the youth of the alpine environment (from Tertiary climates judged by fossils) with the presence of a rich specialised biota, unless there was some rapid evolution and physiological adaptation, and the history of alpine communities remains an outstanding problem of New Zealand biogeography.

New Zealand Biogeographic Elements

Botanists and zoologists on the whole have agreed on the main elements in the biota, classified according to their geographic relationships and probable origin, but have used a variety of names for these elements. The names used here and in Fig. 12 are an attempt to reconcile different nomenclatures.

1. Malayo-Pacific (Malayan, Malayasian, Palaeotropic, Indo-Pacific etc.). This element includes plants and animals of tropical and subtropical derivation (other than those that came via Australia). Many organisms without an adequate fossil record are attributed to this element: Placostylus, plants like Meryta and Xeronema. seabirds such as Pterodroma hypoleuca. In many cases it is impossible to tell whether the colonists came direct from the north via the islands and shallows of the submarine ridges (perhaps more emergent in the past than now) or via Australia. In groups that were possibly once cosmopolitan but are now restricted to southern lands, it may be difficult to tell whether immigrants to New Zealand came by the Malayo-Pacific or Austral route and this recently led Forster to abandon the attempt to determine their route of origin and to group all older elements in an ‘Archaic’ category.

Fig. 12: New Zealand Biogeographic Elements, with some exam ples. Map modified from a diagram by J. A. Rattenbury (1962).

2. Australian Element. Australian relationships have always been strong in the fauna, and were much stronger in the flora before the rise of the xerophilous plants to dominance on the other side of the Tasman, a comparatively late development (Gill). Australia has been the source of the majority of New Zealand birds (Falla, 1953). In other groups, particularly of ancient dispersal, it is sometimes hard to distinguish Australian from Malayo-Pacific elements. After Fig. 12 had been prepared, Professor J. T. Salmon told me that the wetas (Hemideina, etc.), though related to the King Crickets of Australia, are probably Malayo-Pacific elements that reached both sides of the Tasman.

3. Austral (Antarctic, Subantarctic, Fuegian, Antarcto-Tertiary in part).

Many organisms have a circumpolar, or partially circumpolar distribution in cool-temperate latitudes but are absent from the warmer climatic zones to the north. Some have colonised, in postglacial time, islands severely glaciated in the Pleistocene (Dawson, Tuatara, vol. 7, p. 1), most have adaptations for dispersal, and some suggest by their limited fossil record that dispersal has occurred late in geologic time (Pliocene and later). This element, evidently distributed by the West Wind Drift of air and sea, under present geographic conditions, may be called Neoaustral. Some of its marine members can be derived from originally tropical or subtropical stocks that came to the Southern Ocean at any one of the three major points of entry (Australasia, South Africa, South America) as Fell (1962) has elegantly demonstrated for echinoderms. W. R. B. Oliver (1925) also pointed out that each southern land seems to have supplied colonists to the west wind zone, so that a decreasing number of species is found as a genus is traced down-wind (eastwards) from its headquarters in one of the southern entry points. Other Neoaustral species (the mussel Mytilus edulis, the King-crab Lithodes, the skua Catharacta) are bipolar (antitropical) in cold temperate seas of both hemispheres, and neither their hemisphere of origin nor their longitude of contact across the tropics is known.

The older and more puzzling division of the Austral Element is the Paleoaustral, consisting of plants and animals more distantly related to those of other southern lands, some with a fossil record going back to Tertiary or Mesozoic times, which are no longer being dispersed (judged by their systematic distinctness) and which have such poor dispersal abilities that many biologists have considered they were distributed along land connections linking southern lands with Antarctica. They include the podocarps. Proteaceae, Nothofagus and other plant genera (Aristotelia, Laurelia, Astelia, Fuchsia, Wein-mannia). The mollusc Struthiolaria, descended from Paleocene Perissodonta which ranged to South America and Graham Land (and now survives in South Georgia and Kerguelen) is an example from the littoral fauna, and there are many among the terrestrial page 98

Fig. 13: First appearance in the New Zealand fossil record of members of the marine fauna, illustrated by living species shown at the dates in the geological time scale when the taxa to which they belong are first recorded by fossil predecessors.
Time scale distorted.

page 99

Fig. 14: First appearance in the New Zealand fossil record of members of the land and fresh-water fauna and flora, illustrated by living species shown at the dates in the geological time scale when taxa to which they belong are first recorded by fossil predecessors. Time scale distorted.

page 100 invertebrates, including those of the forest floor litter. The fossil record suggests that the Paleoaustral elements entered New Zealand over a long period. Some of them, originally Malayo-Pacific, may have come south through New Zealand and subsequently spread in the Austral region.

4. Endemic Element (Aboriginal, Palaeozelandic in part, Archaic). Groups that have no close relations in other countries to indicate their place of origin, either because they developed here (primary endemics) or because they became extinct elsewhere (secondary endemics). The endemic element, shorn of forms that can be attributed to other elements, includes the moas, kiwis, short-tailed bat, and tuatara, among vertebrates, and many invertebrates (e.g. Peripatoides).

5. Cosmopolitan Element. Organisms so widespread that their particular route of colonisation cannot be determined on available evidence.

6. Holarctic Element. A few organisms related to north-temperate forms, absent in the intervening tropics, especially certain birds (Scaup, Merganser, Black-billed gull, South Island Pied Oystercatcher), plants such as Euphrasia¹ and Festuca novaezelandiae (aff. ovina), and others in the alpine flora.

Interpretation of modern distributions is subject to the same disadvantages as interpretation of fossils. Cockayne did not know that Nothofagus is living in New Guinea nor that a Fuchsia grows on the summit of Raiatea in Tahiti, but his conclusions on the history of the New Zealand Flora have stimulated a generation of botanists. The data of neontology cannot tell us the extent and duration of past distributions; but they supply the warp into which the woof of paleontology can be woven to give a pattern in time and space still full of imperfections but with hope of probability.

Interpretation of the Fossil Record

The fossil record includes remains of many living plant and animal taxa identifiable with varying degrees of confidence (Figs. 13, 14). The time-spans covered by the recognisable predecessors (some certainly ancestors) of living forms vary greatly. We obviously cannot accept the short fossil record of Sphenodon, of the snail Rhytida, or of the endemic fern genus Cardiomanes (for example) as significant of anything but the inadequacy of the record. How, then, can we dare to interpret such a fallible record? Interpretation must always be subject to revision as new information comes to hand.

Incomings in the fossil record (Figs. 13, 14) may be due to chance preservation or identification (as with Jasus and Notothenia, represented by single records). It may be due to evolution from an ancestral form, as with Struthiolaria, which appeared in the Oligocene but is a descendant of Monalaria (Eocene) and Perissodonta (Paleocene). In the diagram, living forms are shown at the point in the record where their presumptive ancestors first appeared, even if they differ generically or specifically from the first fossil record of the taxon they represent.

In many families and genera, the earliest fossil representatives are not the direct ancestors of the living species, which are the result of re-invasion of new stock from overseas. Thus the living Architectonica reevei and Mesopeplum convexum are probably not direct descendants of the first New Zealand members of these genera known in the Eocene and Oligocene but are due to re-invasion at much later dates. Immigration has been important throughout Tertiary time, and we no longer believe, as Marshall did, that the New Zealand marine fauna has been isolated here since the Cretaceous. In the marine fauna, therefore, the first appearance of a taxon that has no recorded ancestors but occurs regularly later may be taken (as a working hypothesis) to indicate an immigration. As such ‘first appearances’ are important to the stratigrapher, they are being constantly tested and amended.

Study of pollen and spores has yielded much data on the fossil history of living plant taxa. In some plants, pollen is so distinctive and abundant that its absence from older rocks gives confidence in the hypothesis that its appearance in the record is due to immigration. Other cases are ambiguous. Thus pollen of the miro-matai type (Podocarpus spicatus and ferrugineus) ranges back to Upper Miocene but the section Stachycarpus to which these species belong probably ranges back at least to Upper Cretaceous (Shag Point) judged by macrofossils. Similarly, kahikatea-type pollen first appears in the Upper Eocene, but macrofossils suggesting section Dacrycarpus range back into the Mesozoic, perhaps to Jurassic. Podocarps have a long history since the Triassic in the Southern Hemisphere, so the appearance of distinctive pollen in the Tertiary is probably due to evolution (at least of the pollen) in New Zealand, but the possibility of immigrations at late dates cannot be excluded.

Biogeographic Synthesis

The classification of New Zealand organisms into biotic elements from their geographic affinities (Fig. 12) has been combined with paleontological data on their probable age in the biota (Figs. 13, 14) to give a picture of changing biotic influences in later Mesozoic and Cenozoic time (Figs. 15A, 15B). Endemic, cosmopolitan, and Holarctic elements are not shown as they add nothing to the picture or have an inadequate fossil record. Extinct organisms of known page 102

Figs. 15a, 15b: Interpretation of the history of biogeographic elements. Immigrant taxa, living and extinct (with asterisks) classified under biogeographic elements from their geographic relationships, are placed near the geological date of their first appearance as fossils to give a picture of changing geographic influences through Mesozoic and Tertiary time.

page 103

Fig. 15b: For caption, see facing page.

page 104 geographic affinities are included as they contribute greatly to the understanding of past changes in the biota.

The alternation of Tethyan (tropical) and Austral (southern) influences in Paleozoic and Mesozoic faunal history was emphasised above. The Middle and Upper Jurassic period of Tethyan dominance left little trace in later marine faunas but it perhaps contributed to the origins of the Paleoaustral conifers for which, however, Florin claims an origin as early as Permian and a subsequent history entirely southern.

Whatever their ultimate origin, Paleoaustral conifers (Podocarps and Araucarians) were already prominent in New Zealand vegetation in the Jurassic and remained so in the Cretaceous, when the Paleoaustral influence was strengthened by the appearance of the southern angiosperms. Nothofagus and the Proteaceae, and of marine groups such as ancestors of the Struthiolariidae (Gastropoda), Lahilea and Pacitrigonia (Bivalvia). The Paleoaustral influence remained strong in the early Tertiary and contributed plant immigrants (Fuchsia, etc.) as late as Miocene. By this time, hoever, the Malayo-Pacific influence (successor to the Tethyan) had become dominant, and it seems possible that some Austral plant genera that first appeared in New Zealand in the mid Tertiary (Astelia, Aristotelia and Weinmannia, for instance) are in fact Malayo-Pacific elements that entered the Austral region at this time. It is most unlikely that land connections with Antarctica persisted through the long period of Paleoaustral invasion indicated by the paleobotanical record, and all we can be sure of is that dispersal was easier in the past than it is now. Antarctica was not heavily glaciated and was at least partly vegetated in the Tertiary with Paleoaustral plants, as Cranwell's pollen studies have shown, and this undoubtedly contributed (irrespective of land extensions) to the dispersal of the Paleoaustral elements, most of which probably entered the area from the north down one of the southern land masses. Some Paleoaustral organisms (e.g. Fuchsia and the extinct penguins) were probably distributed across the sea like their Neoaustral successors.

Australian elements can be detected as early as Lower Cretaceous (Maccoyella) and continued to arrive in a steady trickle, often hard to disentangle from the Malayo-Pacific, perhaps fluctuating in importance. Australia shared with New Zealand the Paleoaustral elements, so it is questionable whether the pollen attributed to? Banksia should be classed as Australian or Paleoaustral.

The dominant Malayo-Pacific influence (Fig. 15B) in the middle Tertiary is most obvious in the marine fauna, but it also brought plant immigrants (Metrosideros, Dysoxylum, Freycinetia, Rhopalostylis, Cocos, Elytranthe, Macropiper, Avicennia, Dodonaea). The Malayo-Pacific influence died away after a climax in the early Miocene and many of its immigrants became extinct. The most important of all Malayo-Pacific immigrants—Polynesian man— page 105 came in the Late Holocene to lay the foundation stones of New Zealand human culture.

In the Pliocene and Pleistocene an Austral influence again became important, marked by the appearance of marine animals classed as Neoaustral, distributed in the West Wind zone. The plants that (by inference) accompanied such animals are not well represented as fossils. I would judge that Ranunculus acaulis, Sophora and Hebe gained their wide southern distribution by easterly dispersal at this time. The Neoaustral is not a direct continuation of the Paleoaustral — it is much more restricted. There has been no further dispersal of beeches, Proteaceae and podocarps between southern lands, so that the appeal to land extension or continental drift to explain the difference is not unreasonable.

Land Connections

Geology cannot at present prove or disprove land connections but structural history can suggest the most probable dates and directions of land extension. At no time is there evidence for direct Trans-Tasman connection with Australia. The early Paleozoic is obscure but from Permian to Jurassic and especially during the Early Cretaceous orogeny the geanticline west of the New Zealand geosyncline could have extended north to New Caledonia, perhaps beyond. Its extension south of Campbell Plateau demands great changes in ocean depths, and probably in crustal thickness, for which at present there seems little evidence. If the circum-Pacific mobile belt is (or was) continuous from New Zealand to the Andean Province of Antarctica, and especially if continental crust can become more or less ‘oceanised’ as suggested by its thinness at Campbell Plateau, there may have been some land on the Macquarie-Antarctic ridges, but Campbell Island stratigraphy suggests that present crustal conditions date from some (considerable?) time before the end of the Cretaceous, and have been stable ever since. By the late Cretaceous when angiosperms were dispersing, the geological conditions for land extension had deteriorated, and remained poor till the Miocene. Failure of mammals and snakes to reach New Zealand at these times is evidence against any land connection beyond New Caledonia; nor could continental drift transport plants without bringing such animals. Late Tertiary faulting and folding movements probably affected the ridges to the north, and could have extended land. Conditions for dispersal in the south-west Pacific were formerly better than now, judged by the fossil record of immigrants to New Zealand throughout Mesozoic and Tertiary time. Antarctica was temperate and vegetated for an immense period prior to Quaternary glaciation, and provided a vast stepping stone in the Southern Ocean, doubtless supplemented by volcanic seamounts on ocean ridges. These concessions will not satisfy those who affirm the need for land to disperse Podocarps, Nothofagus, and many invertebrates, so the fascinating biogeographic enigma that intrigued page 106 Hooker and Darwin is still unsolved. It is not insoluble. Nothofagus and Podocarps apparently moved north into the mountains of New Guinea in Plio-Pleistocene time (Couper), presumably from Queensland. If beech-podocarp forest slowly moved down to the floor of Torres Strait during an early glaciation and was later forced up into the montane zones by warming climate, Pleistocene climate changed more than is generally claimed for the ‘tropics; the alternative is a leap of several hundred miles across strait and/or lowlands comparable with the trans-oceanic leaps postulated for Pre-Tertiary dispersal of these Austral groups. But whereas evidence for Mesozoic dispersal has been obscured by the immense passage of time and the destructive glaciation of Antarctica, evidence for this Plio-Pleistocene dispersal can be sought by studies of the paleo-geography, submarine geology, and paleobotany of Torres Strait and Arafura Sea and nearby regions.

Epilogue

I cannot end this paper without asking the reader to treat the tentative conclusions expressed as merely hypotheses to be critically tested. Between the raw geological data and a paleogeographic map yawns a gulf to be crossed by projection and extrapolation. Control for elements in past geography is seldom unambiguous and even the existing data have not yet been thoroughly analysed. Errors in identification and dating of fossils and in interpretation of the facies of sediments can easily lead to misconceptions about past history and geography. On the whole, the data are fuller and the degrees of freedom fewer in late geological time; ambiguities increase as we go back in time.

Two principles guide my interpretation of fossils. The principle of correlation of organs with organisms enables a specialist to make a mental reconstruction or identification of a fossil elephant from one tooth, of a gastropod from its operculum, or of a tree from its pollen grains. The principle of economy of hypothesis virtually requires us to conclude that identical things represent identical things until there is evidence that they represent different things. Pitfalls are obvious. Unrelated plants can in fact produce indistinguishable pollen grains, but in tracing back the history of our vegetation nothing is gained in assuming that pollen indistinguishable from that of a living plant taxon (whether it be species, genus, or family) was produced by an entirely different taxon. The accumulating data commonly force us to complicate our hypotheses but there is no justification for doing so in advance of the evidence.

The history of New Zealand organisms (Figs. 13-15) has been illustrated by drawings of existing species to emphasise that the dead fragments studied in the rocks represent whole plants and animals that are biological and not geological phenomena — green with chlorophyll, glowing with pigment, feeding, growing and reproducing like their modern representatives. But this illustrative device should page 107 not be misinterpreted. We do not, in fact, know that the Cretaceous Elephant Fish looked like the living one pictured, nor that Banksieaeidetes pollen came from a flower like that labelled ?Banksia in Fig. 15. The names of organisms shown on Figs. 13-15 are placed as close as possible to the positions on the geological time scale where they are first known as fossils, but not exactly, owing to lack of space, and original references should be consulted to determine their time ranges.

The biogeographic synthesis illustrated in Fig. 15 represents a subjective interpretation of data not yet fully analysed quantitatively, which, probably, no two workers would interpret in quite the same way. In the very inexact science of biogeography, such hypotheses cannot yet be directly proved and in Charles Darwin's words ‘the doctrine must sink or swim according as it groups and explains phenomena’.

Acknowledgments

Many of my colleagues, but particularly N. de B. Hornibrook, W. F. Harris, D. J. McIntyre, J. B. Waterhouse, G. L. Lensen, G. H. Scott, T. Grant-Taylor, and G. W. Grindley have patiently answered my questions. The Tertiary maps have benefited through comment from Geological Survey officers at Otahuhu, Wellington, Christchurch and Dunedin and also from M. Gage, P. Phizackerley, E. J. Searle, K. C. Short and P. Vella. I am especially grateful to Drs. M. Gage and R. P. Suggate for permission to use their unpublished may [sic] showing ice distribution in the Otiran Glacial Stage.

To save space few conventional literature references are appended, but I have tried to show the main sources of data, which the reader may trace in the Bibliography of New Zealand Geology (in press) and in the annual geological abstracts (N. Z. J. Geol. Geophys.).

The small figures were redrawn from a variety of sources, among which particular acknowledgment is due to the art of Dr. A. W. B. Powell in numerous publications. R. C. Brazier contributed a few drawings, S. N. Beatus photographically reduced them all, and I. W. Keyes helped with their assembly; C. T. T. Webb supervised the draughting of the maps and the preparation of diagrams for publication.

References

Allan, R. S., 1935: N.Z.G.S. pal. Bull. 14; 1947: J. Pal. 21: 436: 1960: Rec. Cant. Mus. 7: 233.

Arkell, J. W., 1956: Jurassic Geol. of the World, Edinburgh.

Benham, W. B., 1935-42: Trans, roy. Soc. N.Z. 65: 232, 239: 67: 1, 8, 15; 71: 260.

Blackett, P. M. S., 1961: Proc. roy. Soc. (A) 263: 1.

Brown, D. A., 1952: Tert. Cheilostomatous Polyzoa N.Z. (Brit. Mus.).

Carey, S. W., 1958: Continental Drift (Sympos. Univ. Tasmania).

Cockayne, L., 1928: Vegetation of N.Z., Leipzig.

Cotton, C. A., and Te Punga, M. T., 1955: Trans. roy. Soc. N.Z. 82: 1001.

Couper, R. A., 1960: N.Z.G.S. pal. Bull. 32; Proc. roy. Soc. (B) 152; and McQueen, D. R., 1954: N.Z. J. Sci. Tech. B35: 398.

Cranwell, L. M., 1959: Nature 184: 1782; and von Post, 1936: Geogr. Annal H.3-4.

Cullen, D., 1962: N. Z. J. Geol. Geophys. 5: 309.

Falla, R. A., 1953: Emu 53: 36.

Fell, H. B., 1954: N.Z.G.S. pal. Bull. 23; 1962: Evol. Living Organisms, Melbourne: 307; 1962: Nature 193: 759.

Fleming, C. A., 1953: Rep. 7 Sci. Congr. roy Soc. N.Z.: 114.

Florin, R., 1940: K. Svensk. Vet. Akad. Handl. 19 (2): 1.

Forster, R. R., 1961: Proc. roy. Soc. N.Z. 89: 51.

Frost, G. A., 1924-33: Trans. N.Z. Inst. 55: 605: 59: 91: 63: 133.

Gill, E. D., 1952: Trans. roy. Soc. N.Z. 80: 171.

Glaessner, M. F., 1960: N.Z.G.S. pal. Bull. 31.

Grindley, G. W., 1961: Proc. 9 Pacif. Sci. Congr. 12: 71.

Harris, W. F., 1949: N. Z. J. Sci. Tech. B30 (4); 1955: N.Z.D.S.I.R. Bull. 116.

Holloway, J. T., 1954: Trans. roy. Soc. N.Z. 82: 329.

Hornibrook, N. de B., 1951: N.Z.G.S. pal. Bull. 18: 1953: N.Z.J. Sci. Tech. B34: 436.

Kingma, J. T., 1958: N.Z.J. Geol. Geophys. 1:1: 1959: ibid., 2: 1.

Kulp, J. L., 1961: Science 133: 1105.

Lockerbie, L., 1959: Anthropology in the South Seas, New Plymouth: 75.

McMichael, D. F., 1957: P. Linn. Soc. N.S.W. 81: 222.

Macpherson, E. O., 1946: N.Z.D.S.I.R. geol. Mem. 6.

McQueen, D. R., 1954: Trans. roy. Soc. N.Z. 82: 231; 1956: ibid., 83: 673.

Marples, B. J., 1946: Trans. roy. Soc. N.Z. 76: 132; 1952: N.Z.G.S. pal. Bull. 20: Rec. Cant. Mus. 7: 185.

Marwick, J., 1953: N.Z.J. Sci. Tech. B34: 317.

Oliver, W. R. B., 1925: J. Linn. Soc. Bot. 47: 99: 1949: Dom. Mus. Bull. 15: 1950: Tuatara 3: 1.

Pantin, H. M., 1957: N. Z. J. Sci. Tech. B38: 781

Raeside, J. W., 1948: Trans. roy. Soc. N.Z. 77: 153.

Shirley, J., 1938: Q. J. geol. Soc. 94: 459.

Squires, D. F., 1958: N.Z.G.S. pal. Bull. 29.

Stinton, F. C., 1957: Trans. roy. Soc. N.Z. 84: 513.

Stokell, G., 1945: Trans. roy. Soc. N.Z. 75: 134.

Te Punga. M. T., 1958: N. Z. J. Geol. Geophys. 1: 92.

Wellman, H. W., 1956: N.Z.D.S.I.R. Bull. 121: 1959: Trans. roy. Soc. N.Z. 87: 99.

Willett, R. W., 1950: N. Z. J. Sci. Tech. B32: 18.

Wilson, D. D., 1956: N. Z. J. Sci. Tech. B37: 603: 610.

Withers, T. H., 1924: N.Z.G.S. pal. Bull. 10.