In an attempt to fulfil an oft-repeated request for a contemporary palaeontologist's views on the geological and biogeographical history of New Zealand I am writing something on this subject without looking up a single reference. In this way I may be able to give some idea of the feeling of at least one geologist-biologist of today (1948) on a subject so strewn with pitfalls that none of those who so freely criticise the attempts of previous generations have as yet come out into the open to give us their own version.

Few have tackled this subject since Frederick Wollaston Hutton, who was a pioneer biologist over a wide field, a geologist, and palaeontologist. His ideas are still given at the universities. Then we have contributions by the geologist Dr. P. Marshall; he tended, like everyone else, to interpret the story to fit his ideas of N.Z. geology. Such views as the mainly autochthonous nature of the N.Z. fauna, springing from Cretaceous ancestors were not based on critical palaeontoäogy, and were used to support geological theories as to the absence of breaks (unconformities) in the Cretaceous-Tertiary record. Speight, primarily a geologist, tended to lean heavily on biologists when he felt inclined to build land bridges. Benham, who would change the face of the globe for an earthworm, has built bridges on the plan of his master Beddard, and leaned in turn on the geologists, who vouched for their plausibility. Chilton's discussion (Subantarctic Islands of New Zealand, 1909) shows that the “Antarctic Continent theory” was constructed like an elaborate house of cards, an imposing edifice of unstable walls, each leaning for support on another equally weak.

Oliver's paper on the Biogeographical Relations of the New Zealand Region (1925), is a pretty sound assessment of the affinities of the flora and fauna, but it was not well correlated with geologic history. Prof. R. S. Allan prepared a paper for a symposium on this subject, given in Christchurch two years ago with Dr. R. A. Falla (animals) and Dr. W. R. B. Oliver (plants), but it is not published. Dr. J. Marwick published papers on the Origin of the Molluscan Fauna of N.Z., and on the Indo-Pacific element in the N.Z. Molluscan Fauna, and some of his conclusions have stood despite the advances of the last 20 years. Cockayne's history of the N.Z. flora was a noble attempt at correlation with geology. Some of his geological theory (such as the Peistocene elevation as a cause of glaciation) would not be held today. I need not itemise specialised attempts based on single genera or groups.

I have probably left someone out above, but it all boils down to this: the geologist, aware of the imperfection of his evidence and of

Fig. 1. Table of divisions of the New Zealand Cretaceous and Tertiary. The succession of series and stages has been established for actual sediments and fossil faunas represented in New Zealand; the probable correlation with the divisions of geological time in other parts of the world is indicated in the fourth column. Reproduced with permission from Finlay and Marwick, N.Z. Journ. Sci. Tech., vol. 28, p. 229, 1947.

past wild theories, is cautious of generalising, and the specialist biologist seldom knows the geological evidence with its uncertainties. Part of the trouble has been that there wasn't any “geology” for them to know. Most organisms underwent a lot of development, range changes, etc., during the Tertiary, and that is the part of N.Z. geology which has advanced tremendously in the last quarter century, through the application, first of molluscan palaeontology, by Dr. J. Marwick, and latterly of Dr. H. J. Finlay's foraminiferal work. This palaeontological work page 74 has had to go hand in hand with stratigraphical work, but it provided the time-scale upon which the deductions about the structural development of N.Z., were based. Much is unpublished. It is essential to have some knowledge of the geological time scale, particularly of the succession of New Zealand Cretaceous and Tertiary stages which are more important to us than the approximate correlations with the standard divisions, Miocene, Pliocene, etc. (See. fig. 1.)

There is no need to say much of the geological history prior to the Mesozoic as there is probably not a single living organism in N.Z., with fossil ancestors in our Palaeozoic. Palaeozoic marine faunas were fairly similar throughout the world; provinces are recognized, with some endemism, but, in general, the succession here, imperfect as it is, shows changes parallel with those elsewhere, suggesting, that, in general, the Palaeozoic world was almost Wendel Wilkie's dream … one world.

Much the same can be said of the earlier Mesozoic. Our Triassic and Jurassic fauna and flora are quite close to those of, say, New Caledonia, East Indies, the Himalayas, Mediterranean, and the Arctic. We have some endemics, but no typically “Neozelanic” living organisms have ancestors in our Trias and Jurassic, unless the Waikato Heads Jurassic Dicotyledon, Arber's Artocarpidium, is related to a living genus. In the Triassic and Jurassic, a deep, wide, linear trough (“geosyncline”) was present on the position of N.Z., running from the Nuggets, Otago, in a sinuous line across Canterbury and Marlborough, up the North Island to Kawhia and North Auckland. It was a gradually sinking basin, a down-fold, and from what we know of such structures, was almost certainly parallel to a welt, a rising mass of land, which is believed to have been west of the geosyncline. Possibly there were other welts; probably too, the structure was bigger than present N.Z., stretching N.W. and S. to form, perhaps, at times, land connections, or island arcs towards New Guinea … we don't know, but what I have said above makes it unlikely that Trias-Jura paleogeography has much to do with our modern biota; perhaps the tuatara (Sphenodon) could have arrived then. The Trias-Jura period of sedimentation closed with orogeny and uplift of the sediments in the trough, a frequent tendency for what has gone down to go up, like a stationary wave. This is the time (Early Cretaceous) when most bridge builders have tied us up in all directions, and with much excuse. But modern views on the mechanism and pattern of deformed geosynclinal areas would not allow us a “continent” such as many have asked for. N.Z. rocks have never included what are grouped as “continental” deposits; red beds indicating deserts, gypsum and salt deposits; this suggests that we have always had an insular climate of the west-wind belt, and never land of such dimensions as to allow rain-shadow deserts to form in the lee of mountain ranges. Still, we haven't much when it comes to arguing about what was happening out at sea; here in N.Z., uplift of the Trias-Jura geosyncline in the Cretaceous must have amounted to many thousands of page 75 feet, slowly of course, with erosion taking it off as it came up, but it doesn't take many thousands of feet to make a mess of the ocean floors. So you could have outside connections then if you wished, but keep them on lines of the known ocean-bottom ridges. What might have come? Sphenodon, as I've mentioned, the ancestors of the moas, kiwi, the N.Z. frogs, Leiopelma, some of the slow-evolving primitive invertebrates (e.g., Peripatus) and the ancestors of the present endemic, or slow evolving, plants. Re the latter we should eventually have some information from pollen, supplemented by leaf and wood studies.

When the geological record recovered from the blow of the post-Jurassic orogeny—which caused a big lacuna in the lower Cretaceous—we find quite a lot of changes, geological and biological. The pattern of the folds, welts and troughs, that later developed, was finer than before; instead of a broad trough some hundreds of miles wide and thousands long, the folds are narrower, short, interfingering, branching. The welts, which would tend to be land, were similarly small, so we are justified in thinking of post-Jurassic New Zealand as being, what it is now, archipelagic in its geography. Changes in the disposition of sea and land were relatively frequent, and were on quite a large scale, the troughs were locally relatively deep, although they became filled with sediment as they sank, the welts rose correspondingly great amounts, locally, but if composed of lately deposited sediment they would erode quickly. A kind of writhing of this part, presumably also other parts, of the Pacific margin, seems to have gone on. I may interpolate that this condition of a changing archipeligo is one that would encourage “speciation,” i.e., the formation of two or more species by successive invasions, re-invasions, or back-invasions of stock from one island to another, and this would account for genera with a multitude of species, and for species in which complete physiological isolation (in the sense of Dobzhansky and others), preventing interbreeding, had not been attained when the two daughter stocks came together again, so that they “hybridised.” It may be convenient to discuss the inferred palaeogeography step by step, before discussing the evidence of fossils.

The Upper Cretaceous sea washed the east coast of the South Island and the coastline, a fluctuating one, ran obliquely through Otago, up through Canterbury, separating areas of intermittent marine deposition to the east (perhaps progressive sea advance from this direction would be better) from an area of land, with intermontane basins, lakes and swamps, separated by mountains, to the west—over Westland and Nelson, and parts of Otago. In the North, sea occupied the strip from East Wellington to East Cape, and much of North Auckland, and we have little to reconstruct the land from, though it must have been on one side or the other of the strip of sea that received sediment. Most people would extend the Westland land, known definitely from its coal-measure record, to the north-west of the Cretaceous sea areas, page 76 but this problem of constructing land from our known areas of sea is one that constantly crops up. Certainly no one pattern applied to all the divisions of the Upper Cretaceous. The oldest beds (Taitai), are confined to the eastern strip; the Clarentian is present from Marlborough up to Bay of Plenty, and again in parts of North Auckland; the Piripauan is more extensive, from Otago (south of Dunedin), up to Bay of Plenty, and in many parts of North Auckland, and reaches further inland in Otago-Canterbury. The Teurian and Wangaloan, uppermost Cretaceous stages, are less widely recognised within the same area, and there is a chance that even in the area outlined shoals and islands may have existed, breaking up the area of ea. Our Paleocene and Early Eocene, only lately recognised, have not been fully enough mapped to base conclusions on, but they seem to have followed the Cretaceous basins of the eastern and northern areas. Quartz beds (sands and gravels) and leached light sediments at this time suggest that the land was of low relief, peneplained, and that orogenic movements were slowed down for a while (Macpherson, 1946).

The middle Eocene (Bortonian) marks a spread of seas over much of the area known or surmised to have been land previously; further inland in Otago, inland Canterbury, much of Westland and Marlborough, but still, apparently, only in eastern and northern North Island. Some believe, with good reason, that this mid-Eocene sea crossed the South Island to link the deposits now known at Broken River (inland Canterbury), and Westland, others, that a central axis persisted. The later Eocene, Kaiatan, and Runangan, saw the formation or growth of deep, somewhat localised troughs in some areas which were filled by fine sediment, carbonaceous in Westland and parts of North Auckland and Nelson, pale, perhaps deeper offshore sediments in Marlborough. The Kaiatan sea spread over coal measures in the Reefton-Murchison area which had apparently escaped the Bortonian transgression. There is no upper Eocene in the east coast, North Island, area, which apparently emerged temporarily.

The Oligocene Whaingaroan Stage saw a widespread transgression of the sea on to areas previously land. It advanced over Southland, persisted in Otago, Westland, Canterbury and Marlborough, Nelson and the Sounds, Wellington Province east coast to Hawke's Bay, if not further north, and advanced over the western North Island (we don't know from what direction), to cover at least the Taumarunui-King Country area, the Waikato basin, and much of North Auckland. In many of these areas coal below the Waingaroan beds testified to the land surfaces (probably upper Eocene) that were covered. You may ask, was any of New Zealand left? Well, the answer is that the Whaingaroan sediment must have come from somewhere, though it may have come from some distance, as it is generally fine in texture, and extensive limestones, lacking much detritus, were among its sedi- page 77 ments. We know that there were volcanic islands in the Oamaru district, and there were other areas, in or close to, the limits of New Zealand, that were emergent (central Otago, Ross, parts of Nelson).

In the succeeding Duntroonian and Waitakian stages, marine deposition continued with interruptions locally, temporary emergences, at least to shoals; limestones were deposited locally, and, it is believed, further transgression of the sea occurred in the Otago area, into “Central” Otago.

The upper Oligocene, Otaian to Awamoan stages, saw a recession of the sea from a number of areas. There were quite a few places where extensive Otaian beds were deposited, and Hutchinsonian seas still occupied parts of Westland and Otago, but Awamoan deposits are distinctly limited. The period (the Pareora Epoch) saw the withdrawal of the sea from the whole of the North Island, from Southland, Nelson and many parts of the South Island, even if the withdrawal was only temporary. In North Otago, where sea remained, coarser sediments bespoke increasing relief, inferred to be to the west, and the temporary withdrawals everywhere, or almost everywhere, attest a phase of mid Tertiary mountain-building. In most of Otago and parts of Canterbury, the sea withdrew after the Awamoan, and has not been back since. How large an area was exposed by this post-Awamoan or Awamoan withdrawal of the sea, perhaps better, elevation of the land, we don't know, but in many areas the sediments that would be exposed by such a movement were soft and would be readily washed away. I say this lest it be thought that this fairly widespread elevation, even if, as seems likely, it was of short duration, could have meant a land-bridge. It is certainly a good time to put one if there were evidence for it, and there were notable incomings among the mollusca and foraminifera at this time and a bit later. We know little of the Tertiary history of the New Caledonia ridge, except that that island has apparently been out of the sea since the Oligocene. Possibly, then, this late Oligocene pulsation of the New Zealand part of the Pacific margin could have opened fresh paths for immigration of certain organisms from the north. But not a continuous land bridge, mark you, or we would have a plague of snakes and mammals.

Very soon after the momentary shudder described above, the sea transgressed again over many parts of New Zealand, to deposit rocks of the Southland series. From now on, we feel, there is more justification in trying to draw shorelines; the structures that then determined land and sea are in some way related to the structures that now separate high range from lowland, land from sea, though increasing confidence is, of course, gained as we approach our own time.

The early Miocene Altonian sea lapped parts of Otago, but did not go far from the present coast; it flooded the Waiau depression of Southland; it was widespread in North Canterbury and Marlborough, in Westland, and in Nelson and the Murchison area. Possibly the page 78 main axis, the mass of Otago and the line of the Alps were by now delineated, as a pretty low, not highly mountainous, land. In the North, much of the East coast was flooded, certainly in an area east of a line projected through Masterton and Takapau; Altonian seas covered many parts, if not the whole, of the area west of Taumarunui north across North Auckland Peninsula, where the sediments are believed to be derived from a land to the west of the present coast. A western land, of gneissic and granite rocks, west of the North Island, has been inferred from the nature of sediments at various times between the Triassic and the Pliocene. In North Auckland, volcanoes broke out at the end of the Altonian, forming extensive deposits in the Waitakerei Range, at Whangarei, Great Barrier, and Coramandel, possibly also to the south; and it is almost certain that much of North Auckland, and further south perhaps to the latitude of about Raglan, emerged from the sea at this time, and was not again submerged for a long while, if at all. North Auckland thus has a fairly long history as land, but much of Otago is a bit older.

The Clifdenian sea followed the Altonian in parts of Southland (Waiau Valley), Westland, and elsewhere, and in the eastern part of the North Island and the central western part (Mokau River to King Country). The Lilburnian, however, is less extensively developed than the Altonian. The Waiauan sea advanced again locally, even on the margin of Otago, in the Waiau, in Canterbury (Oxford district) Marlborough, Westland and persisted through from the Altonian in the Eastern North Island. The Waiauan sea persisted in North Taranaki, inland to about Taumarunui at least, north to Mangapehi (south of Te Kuiti) and Taupo. It may have extended in from the Gisborne area to the Urewera … there is a blank in our knowledge there, but I think some Tertiary of about this age has been recorded from east of Waikaremoana. But nothing is known of any transgression to the north.

The Tongaporutuan stage saw a southward retreat of the north coast of the big King Country to North Taranaki “bight” which had had a long history as sea; the coast must have been out at sea north of about Raglan, and have swung in to the King Country, thence south inland of the headwaters of the Wanganui River, perhaps then obliquely south to the Wanganui, then S.W. to Cook Strait. On the east, the sea occupied much of the East Cape to Palliser Bay area. But there were probably at times small islets in both the eastern and western seas. Land between these seas is inferred to have crossed Cook Strait, and to have occupied much of the central axis. In the south, Tongaporutuan seas were persent in Marlborough and N. Canterbury but are not known elsewhere in Canterbury, nor in Otago. In Southland, the sea still occupied the Waiau depression, and, what is more, flooded eastern Fiordland, i.e., the area between Lake Hauroko and Port Craig (western Southland), which had previously been emergent to some extent. On page 79 the west, Tongaporutuan seas continued to occupy parts of the foothill-lowland area of Westland, and flowed well in to the Murchison basin, which had a rather violent period of geosynclinal sinking, as an inland basin in which thick estuarine beds, including coal, were deposited.

All this Miocene time, there were oscillations caused by orogenic movements everywhere; upfolds would emerge as island arcs, with fluctuating shorelines moving back and forth, straits and embayments would be formed and shortly be closed again, and we are far from being able to draw exact shorelines. Volcanoes are known to have been active in the Altonian of N. Auckland; in the Tongaporutuan of the Mokau area, probably from centres now out to sea; and at the same period, roughly, other volcanoes affected the east coast area near Hawke's Bay: these also may have been outside N.Z.'s present boundaries, or may have been early eruptions from the Taupo area.

The Kapitean (latest Miocene), was a period when the sea occupied much the same area as previously. There may have been further retreat in North Taranaki, but the main interest in Kapitean palaeogeography, however, is the transgression over the middle parts of New Zealand which introduced the great Pliocene depressions. Thus there is Kapitean over the East Cape peninsula as far as about Te Kaha in the Bay of Plenty; the sea came from the west right up to the shores of Taupo, perhaps further, and the sediments preserved show no sign of vulcanism so that all the volcanic rocks of that area are Pliocene or younger. Possibly a strait or straits formed across the Ruahine-Tararua axis at this time, or a little later. Certainly in the Opoitian there was further transgression over the North Island axis and for the first time since the Altonian, the sea bit into the Auckland Peninsula near Waikato Heads. The same areas were still flooded in the South Island, but there is evidence, from the estuarine nature of deposits in the Waiau, and in North Canterbury, and from coarser deposits locally, that the land areas were rising.

In the Waitotaran, I would say that North Auckland began to take its present shape. The sea flooded the Manukau lowlands, perhaps formed a strait through the peninsula, and possibly many other “lowlands,” such as the Kaipara, the Thames Graben, and Waikato Basins had been formed. By now, the Western Bight of Taranaki had dwindled, but the coast still went from about New Plymouth across to Raetahi, then flowed through a strait between islands into Hawke's Bay, and extended over the whole of that province, reaching up to Waikaremoana. The South boundary of this ancestral Cook Strait is hard to define; the Ruahines stood up as an island; so did the Tararuas which were probably linked to Marlborough. Possibly the coast was somewhere north of Nelson-Marlborough, swinging up to the Manawatu Strait, and then south through a maze of islands and shoals to Masterton and Palliser Bay. In Marlborough-North Canterbury the seas were becoming restricted between growing islands, and the Kai- page break

Geological Time Chart with New Zealand Fossils

page break page 82 kouras were rising rapidly. Nelson was emergent, but Westland was still lapped by seas which did not for a while show much evidence of the growth of the Alps. The area now occupied by Canterbury Plains was a shallow sea, receiving the debris from the rising Alps. In Southland, the fairly narrow trough that had lasted throughout the Miocene in the Waiau basin, was now filled with estuarine and fluviatile deposits, and the sea withdrew (see fig. 2).

The Nukumaruan of the North Island is confined to two basins, one on either side of the mountain axis (but it was not then mountainous) but connected by a straight at the Manawatu Gorge (which was not then a gorge). The western bay extended from Waitotara to Hunterville, to the Gorge, and thence S.W. to the enigmatical southern shore. On the east, the sea covered Hawke's Bay south of Mohaka, extending between eastern islands to near Pahiatua, and thence east to Castle Point. There was also sea in the lower Wairarapa, but most of Marlborough was emergent, the limited areas of sea still persisting receiving ever-increasing amounts of coarse debris from the growing mountains. The Waitotaran seas of Westland were now abruptly succeeded by advancing deltas of the gravel-bearing rivers coming off the rising Alps. In Canterbury, accumulating gravels from the same source estuarine at first, pushed the sea away from the range front. In the North Island, acid volcanoes became important.

In the Castlecliffian little of New Zealand was submerged; the sea occupied restricted parts of the two Nukumaruan basins, one bordering the coast from Wanganui to Marton and Palmerston, the other close to the southern shores of Hawke's Bay. We are now approaching the climax of the Kaikoura mountain-building movements which started well back in the Miocene … and when mountains rise, marginal areas frequently sink, or segments become down-warped or faulted. The Bay of Plenty, of which the earlier history is buried below later volcanics, had a marine phase near Whakatane at this time, and a “rift-valley” across the East Cape was also flooded. Other lowlands, similarly formed, were too close to the source of debris from the mountains and volcanos to let the sea make any headway against the piling up of land-derived debris. The great Ignimbrite eruptions of the Taupo and Bay of Plenty area date from the Castlecliffian, and the early pumice-dominated alluvium may be coeval. Thus old pumice deposits filling the Waikato basin, the Thames depression, the Bay of Plenty, and veneering much of the Hawke's Bay lowland, and such lowlands as the Waipaoa of Gisborne, date from this period, also the extensive older sands of North Auckland. In the south, without pumice and marine fossils to date deposits, we know no beds that are certainly Castlecliffian. Possibly the coasts were everywhere outside the present shores.

The Hawera series, the post-Castlecliffian rocks that are believed to be Pleistocene, are the deposits of coasts that everywhere lie roughly page 83

Fig. 2. An interpretation of New Zealand geography in the early Wanganui Epoch (? Early Pliocene). Marine fossiliferous rocks give positive evidence of the presence of sea in some areas, but the position of land and of coast lines must be deduced from indirect evidence of many kinds which is seldom conclusive.

page 84 parallel to the present ones, formed during the fluctuations of sea level. Such fluctuations in the Northern Hemisphere, and elsewhere, were related to glacial withdrawal and release of water from the sea. No precise correlation between glaciation and sea-levels has been made in N.Z., as yet. The advances of the sea became steadily less, and the intervening withdrawals were in the order of 300 ft., so that it can be deduced, roughly, what islands would be linked with the mainland at such times. Cook Strait, I should have said, seems to have been breached at the climax of the Kaikoura movement, i.e., before the Hawera. The bottom of Foveaux Strait could easily have been exposed at each glaciation, and high level seas of the interglacial periods carved the level tops of many small islands. It must be stressed that although the Kaikoura orogeny preceded the glaciation, no contemporary geologist (to my knowledge) considers that glaciation was caused by a general great elevation of the land.

Volcanoes were abundantly active in Hawera times. Egmont, Tongariro, Ruapehu, and other andesitic cones succeeded the rhyolites of the Castlecliffian, and there were further rhyolitic eruptions too, in the central area. It is not known what precise age the Banks Peninsula volcanoes are, other than post-Miocene, pre-Hawera, likewise the Otago volcanoes. In N. Auckland there is virtually no stratigraphic tie-up to date the older volcanoes, but the later ones are Hawera and post-Hawera.

You will have gathered from the above account that there is a lot we don't know, and that there were plenty of changes in the disposition of land and sea. The significance of these complex geographical changes to the faunal history is not always clear. Obviously, a changing and relatively isolated archipelago is favourable to race and species formation and to adaptive radiation, among the descendants of colonising stocks, especially if geographic isolation has the essential function in evolution attributed to it by Ernst Mayr and others. Actually there is a notable scarcity of groups that have radiated adaptively; although an “old” archipelago, New Zealand has no equivalent to the Drepaniidae (sickle bills) of Hawaii or the Geospizinae (Darwin's finches, Lack 1946) of the Galapagos in its bird fauna. The nearest approach is the “family” Callaeidae, containing the Huia, Saddleback, and Wattled Crow, which may be but the remnant of a group evolved in New Zealand from ancient stock derived from ancestors of the Australian Apostle Birds. The genera Nestor (Kaka and Kea), Xenicus (New Zealand Wrens), and the Mohoua-Finschia complex (Yellowhead-Creeper) are other small groups that may be the result of the same process. Geographic races of some littoral mollusc genera were more common in the past than they are in the Recent fauna (Pelicaria, Austrofusus) and it is a tempting analogy to suggest that groups in the terrestrial flora (e.g., species of Coprosma) and fauna (e.g., genera and species of Moa) underwent active geographic speciation in the Tertiary. The role of Pleistocene distribution changes has usually been page 85 underrated, e.g, the vigorous race-formation in Paryphanta dates from post-glacial recolonisation of the South Island mountains, and not from the Miocene as Powell originally suggested, and much of the local endemism among alpine plants must have developed since the Pleistocene. But before I pass on to the biota, I must give you some supplementary notes on the outlying islands, and on climatic changes.

Islands:

We know little about the Kermadecs, except that since they last arose as volcanic islands from the sea they have not been linked with New Zealand, or elsewhere. Three Kings: probably land-linked in late Tertiary, then isolated, and never again joined. Other “islands” at North Cape and Houhora have been joined by sand bars perhaps at the time (? Castlecliffian), when much rhyolitic sand was supplied to the Waikato. Some of the coastal islands were probably once landlinked; others have been submerged in Pleistocene times; others, in the Bay of Plenty, have arisen from the sea as volcanoes. Chatham Islands have had a lengthy history. Some land was probably there in the early Eocene, in the Oligocene, and in the Pliocene, at all of which periods there are littoral deposits. I do not know of any reason why they should have been joined to New Zealand: their probable great age as land makes the development of flightless rails there from flying ancestors quite plausible. Similarly with the Subantarctic Islands: there was Cretaceous land at Campbell Island, then late Cretaceous to Oligocene submergence, and probably no land very close: in the early Pliocene vulcanism broke out and the island as we know it began to develop from the erosion and Pleistocene glaciation of a volcano. The Auckland Islands have a less full history: there was Tertiary land, volcanoes and coastline, Pliocene vulcanism and then quite strong Pleistocene glaciation while the group was already insular. Stewart Island: virtually nothing is known of paleogeography before the Pleistocene, when it was insular at every interglacial sea-level rise; possibly linked at glacials.

Climate:

Generalising, one may say that the upper Cretaceous may have been cool, parts of the Eocene warm-temperate, the Oligocene a little cooler; the Miocene tropical, with cooling in the Taranaki series, continuing in Opoitian and Waitotaran stages to a climax in the lower Nukumaruan, when subantarctic sea temperatures reached north to about 40 degrees S. lat. At the same time glaciation affected the newborn Alps near Ross and elsewhere. This may mean that our Pliocene-Pleistocene boundary is wrong and should go before the Nukumaruan, but that it is too big a question to tackle here, though it is an important reason for using the local N.Z. divisions of time and not the more familiar Pliocene, Pleistocene, etc. The Upper Nukumaruan was warm, perhaps as warm as the seas of New South Wales. A lot of warm page 86 water forms became extinct in the early Castlecliffian and, though there are no cool-water immigrants, I think the lower Castlecliffian was fairly cool. This was the period of the big ignimbrite eruption, you will remember, and I like to correlate the older pumice deposits of the Manakau and Kaipara. Therefore it is possible that the cool flora lignites associated with pumice from Otahuhu and Kaipara (recently described by W. Harris), came from this period. The Upper Castlecliffian saw a return to warm conditions, as warm as North Cape, or warmer, in the Wanganui district. The climax of orogenic movement followed and if the climatic fluctuations mentioned above were accompanied by mountain glaciation, the evidence of it has been largely destroyed, or at least not recognised except for the Ross glaciation mentioned. At Wanganui, where the Castlecliffian sequence is preserved, it was followed by sea fluctuations representing two major sea advances which are probably interglacial, separated by inferred withdrawals which may have been glacial. This would correspond with the older Piedmont and the later Alpine Glaciations of the South Island, which are usually referred to by writers when they talk of the Pleistocene glaciation. Willett (unpublished), has plotted the glacial snowline for N.Z. Permanent snow, and glaciers, tipped the Tararuas, the N.W. Nelson mountains, and descended to about 3,000 feet at Stewart Island. At the Auckland Islands the snow-line was below 1,000 feet, and about 500 feet at Campbell Island. Under such conditions all vegetational life zones must have descended some 3,000 feet; or about one zone in Zotov's classification. Thus most woody vegetation must have left Stewart Island and none could persist in the Sub-antarctic: all that is there now must have come since. Big changes must have affected the plant communities of N.Z., and I would think that beech forest reached into North Auckland where it now persists as relict “islands” remaining after the post-glacial return of conditions favouring mixed forest, and kauri-taraire. I don't know whether the kauri could have maintained its present distribution, or whether it has advanced south since the glaciation, i.e., in the last 20,000 years. The pollen students will eventually elucidate the post-glacial vegetation changes.

Faunal Changes:

There is a great mass of information on the changes in the molluscan fauna, as yet scarcely adequately analysed. It is worth noting that molluscan distribution is fairly correlated with that of other marine organisms and that when we had, for instance, influxes of current-borne molluscan larvae from any direction, conditions would favour dispersal of drift-carried plants and animals from the same sources. Thus the molluscan changes may have paralleled those in other groups. Of course, if distribution was effected by migration along shorelines or more continuous shallows and islands than now, this would greatly expedite colonisation by land organisms.

Our Cretaceous fauna included marine, but not land, giant reptiles (Plesiosaurs, e.g.,Mauisaurus, and Ichthyosaurs); the molluscs include some Australian types (Maccoyella), and others of world-wide distribution (Inoceramus, Buchia, Trigonias, and Belemnites). There are a number of forms which might fairly be considered ancestral to the modern N.Z. mollusca. For instance, Conchothyra is a gerontic ancestor of the Eocene Monalaria and the Oligocene to Recent Struthiolaria. Among the ammonites affinities have been drawn with other circum-Pacific areas, including Grahamsland, and Marwick has noted that the forms common to the Tertiary of N.Z. and South America date back to the Cretaceous in most cases. The uppermost Cretaceous Wangaloan fauna includes some endemic lineages ancestral to our Tertiary faunas, but others that are in beds of similar age elsewhere, suggest that communication, in one way at least, was not cut off. The Eocene faunas have many typically N.Z. forms, including lineages not previously recorded, that must have arrived by migration. There is a handful of genera with affinities in the Eocene of other regions, e.g., Speightia (S. America) and Aporrhaids like Dicroloma. The Oligocene faunas are richer, and there are quite a few genera in the later Oligocene that have an Indo-Pacific aspect (e.g., Bathytoma). There is no doubt that much in-migration from the north, or perhaps from the west, occurred, but precise determination of the source of such immigrants is hampered by the scarcity of Oligocene beds in other S.W. Pacifice areas, e.g., most of the molluscan faunas of Australia are post-Oligocene. I mentioned the fairly general sea withdrawal at the end of the Awamoan or a little earlier. The next sea transgression (Altonian), brought substantial numbers of immigrants from tropical seas, of Indo-Pacific aspect (e.g., large foraminifera). This may have been assisted by the existence of shallow water migration paths, and must have required a southward advance of warm seas or the colonists could not have survived. But, with regard to possible land bridges, many, perhaps all, of the in-comers had larvae capable of long distance current transport so that there is no strong case for shore-line migration. They included Cypraea (cowries), Conus, Spondylus and Aturia (an extinct nautilus), which had been present earlier and re-invaded now, and many types now characteristic of tropical seas. There were no reef-building corals although temperatures must have approached those necessary for them; perhaps too much detritus entered N.Z. seas for the reef-builders.

The rich Miocene faunas did not persist to the Pliocene. Although the trends have not been fully analysed, they suggest progressive extinction of both autochthonous types, including as such those genera with early Tertiary ancestors, and of later immigrants. Nevertheless, there were continued immigrations from outside dribbling in through the late Miocene and early Pliocene suggesting the result of chance transport to an insular area. Some were, perhaps, of Australian deriva- page 88 tion; others came directly from the north, e.g., Waitara, a gasteropod of Japanese affinity, arrived in Tongaporutuan, and other N. Pacific genera in Opoitian; Zethalia (Indo-Pacific), in Waitotaran. I have elsewhere implied that southern cool seas crept up the east coast of the South Island in the lower Nukumaruan: this coincided with the extermination of a number of warm water types that had survived from the early Miocene or earlier (Olivella, Polinices, Ostrea ingens, etc.), and it brought a number of southern forms, but not many, perhaps because migration was difficult and because most of the shells we know as southern forms are now rock and kelp dwellers seldom preserved in the fossil record. Not all, by any means, of the warm water types were exterminated, and, when warm water conditions returned in the Upper Nukumaruan, they flourished (Eumarcia, Pedalion, etc., with perhaps a few re-immigrant warm water types, Pterochelus, etc.). When the cooling of the lower Castlecliffian exterminated so many of these forms, few took their places; but, with the upper Castlecliffian amelioration, quite a host of forms of Australian derivation (Xenophalium, Anadara, and others), and some from the north (Zelippistes, Pterochelus zelandicus), came in. The Eastern Australian current (Notonectian), is believed to be the dispersal mechanism for the west Tasman immigrants, and has continued to bring them till Recent times. Northern N.Z., perhaps, receives some drift direct from the north. Some of the Castlecliffian immigrants did not survive the later coolings (Anadara, Leucotina); others survived in North Auckland only (Pterochelus, Zelippistes), and there were doubtless extensive range changes during the later glacials although, perhaps because of the persistent Notonectian current, these have not left such a mark on the fauna as earlier changes.

Summarising, we may say that a nucleus of our molluscan fauna dates from the late Cretaceous; that there have been substantial additions and subtractions at various times through the Tertiary, particularly of Indo-Pacific types with a maximum influx about Altonian, and of some Australian forms, particularly in the late Pliocene; and that climatic fluctuations moulded and sieved the fauna in the latter part of the record, accompanying, it may be added, the development of the present faunal provinces.

What happened on land? Until the fossil botany has been unravelled, in close co-operation with stratigraphers to allow correlation with the marine record, it would be folly to guess. We don't know the precise age of the fossil coconuts which Berry thought to be Pliocene: probably they are older and thus among the oldest in the world. The pollen workers tell me that Nothofagus and Podocarp genera have a lengthy history in our coals which is not surprising. It will be interesting to find what Nothofagus did in the Altonian when the surrounding seas were warmed up. The absence of all record of a land fauna from our not inconsiderable coal measures, estuarine gravels and littoral deposits, surely cannot indicate that there was no land fauna. The page 89 earliest moas are Nukumaruan, and no early Tertiary bird faunas are known except the marine penguins of Eocene to Oligocene age. I have already noted how the archipelagic history of N.Z. may have influenced the moa speciation; that was also Hutton's idea. By the time we get extensive bird faunas, in the late Pleistocene or early Recent, Moas are diversified; there are endemic flighless rails such as Notornis and Aptornis derived from types that flew here before they lost their flight; and the goose, swan and other birds, the derivation of which from Australian types can be deduced. Later, continued sporadic colonisation from Australia has occurred and still occurs, a curious significant parallel to the mollusca where the record is better documented. Doubtless the endemic N.Z. birds are only remnants of the Tertiary fauna, and fairly ancient derivation is suggested by the affinities of some: the Blue Duck with a New Guinea relative; our wrens with the Indo-Australian Pittas; the thrush with a Javan form (R. A. Falla); the Crow with the Australian Apostle Bird; etc., etc. But many forms (Tits, Robin, Bellbird, Stitchbird, Fernbird), can be clearly derived from Australia quite lately; and others (Pukeko, Pied Stilt, Herons), might have arrived in human times. Some suggest derivation from the north rather than from the west, a further analogy with mollusca; and to complete the analogy there is a host of oceanic forms that are Subantarctic in derivation, comparable to the rock and kelp-dwelling molluscs that have come to N.Z. at various times, but chiefly in cool periods of the Pleistocene and late Tertiary.

In general, from consideration of the groups I know, I would put N.Z. organisms into several categories:

(1) A small number of extremely endemic forms that must have had a long history here since the Cretaceous; the mollusc Struthiolaria, the lamprey, tuatara, N.Z. frogs, kiwis, moas and the beeches (Nothofagus), to take examples from several groups.

(2) A larger number of subsequent migrants either from the north or from Australia, which have undergone varying degrees of differentiation according to the time when they arrived (many molluscs, some lizards, wekas, tomtits, the plants Xeronema and Meryta. Placostylus among land molluscs is a typical northern element).

(3) A substantial number of late immigrants from the same two general sources; many molluscs, perhaps some lizards, the sea-snake and turtle (which have not colonised permanently) many birds such as Pukeko, White and Reef Herons and the Silvereye, and Pisonia and other plants. Perhaps Meryta and Xeronema should come here rather than in (2).

(4) A number of more or less circumpolar types whose derivation is by the west wind, the west-wind drift, or by the birds of this zone; the marine Algae Durvillea and Macrocystis, kelp-dwelling molluscs, seals, penguins and many petrels: kowhai, korimikos, bidibids, page 90 and some grasses; many invertebrates. Many of these are of late derivation, but some show evidence of more ancient origin from this source which has doubtless been a repeated influence (endemic penguins, some petrels, plants). It is hard to say in some cases whether N.Z. was at the giving or the receiving end (Hebe elliptica, etc.). 16th April, 1948.