The fossil plant record of New Zealand is rather fragmentary. It does, however, show something of the kind of plants that succeeded one another in this region. Unfortunately the specimens from the oldest rocks in which plant remains have been found in New Zealand are too imperfect for identification but it is believed that the beds which contain them could be as old as the Carboniferous. From three localities in Canterbury and Otago a few species, moderately preserved, of a flora which can be classed as Rhaetic, are known. Except for three genera (Phyllotheca, Chiropteris, Baiera) characteristic of the Rhaetic, the plants are such as are found throughout the Triassic, Jurassic and early Cretaceous. With the much better known Upper Jurassic flora we have a group of Mesozoic plants that characterised a southern as opposed to a northern flora separated by an ocean bounding the north of Gondwana Land. In this flora members of the Gymnosperm families Araucariaceae and Podocarpaceae are characteristic and the presence of two species of Angiosperms has been proved. After an interval covering the first half of the Cretaceous we find bursting rather abruptly upon us a flora in which most of the characteristic Jurassic plants are absent, and, instead, Angiosperms are dominant. With them are the same families of Gymnosperms as were present in the Jurassic. This flora lasted until the Oligocene with important changes in that period. After another hiatus we find in the late Miocene a flora much like that of the present day and in the Pliocene floras still more modern in aspect.

Such, in brief, is the history of the succession of the floras that have existed in the past in the New Zealand region, a history that requires a continental connection to explain the presence of its main elements, more especially the group of gymnosperms. The degree of endemism in the present day flora, however, shows that New Zealand has been isolated for a very long time, probably for the whole of the Tertiary period.

The fossil flora of New Zealand, so far as it is known, must be taken into account when studying the origin of the present day flora. Comparisons of the early Tertiary floras of New Zealand and Australia show that the two were quite distinct, each having characteristics reminiscent of today's floras. But though the Tertiary flora of New Zealand developed into the present flora in isolation it received from time to time additions from Australia, presumably by some means of dispersal that carried plant disseminules across the intervening ocean. This seems a justifiable conclusion from the fact that there are some 300 species of plants in New Zealand indentical with Australian species. It is hardly believable that 300 species of flowering plants and ferns page 2 could cross from Australia to New Zealand during a period when land connected the two areas and remain without specific change during the whole of the Tertiary era.

Rhaetic Flora

A small flora consisting of 11 known species comes from Mount Potts, Clent Hills (Canterbury) and Owaka Creek. It is placed in the Rhaetic or uppermost Triassic because of the presence of three genera especially characteristic of the Triassic. These genera are Phyllotheca, an Equisetalian genus, Chiropteris, a fern-like plant, and Baiera, belonging to the Ginkgoales. Phyllotheca minuta, from Mount Potts and Clent Hills, is represented by detached whorls of leaves, about 10 in a whorl. Chiropteris lacerata, from Mount Potts, is a broad cuneate leaf with longitudinal parallel closely placed nerves frequently joining. Baiera robusta, from Mount Potts, is a fan-shaped leaf, slit longitudinally for half the length, with six or more of the basal parallel nerves extending into each segment.

Of the species found also in the Jurassic Linguifolium lillieanum deserves special mention. It is a spathulate leaf with a strong midrib from which arise, at a narrow angle, parallel nerves once or twice forked. Species of Linguifolium have been found in Rhaetic rocks in Chile and Australia. Cladophlebis is a genus of ferns with bipinnate leaves having oblong, acute and slightly falcate pinnae. C. australis has an almost world wide distribution from the Rhaetic to the end of the Jurassic. It is found in all the New Zealand Rhaetic localities. Cladophlebis is probably an Osmundaceous genus, at least the form of the petiolar bundles resembles those of the stems of Osmundites, so possibly the Cladophlebis leaves are the foliage of Osmundites though a direct connection has not been proved. Dictyophyllum is a fern having pinnatifid leaves with a pustular appearance over the whole surface. The sori are dispersed over the under side of the leaf. The New Zealand examples both in the Rhaetic and the Jurassic probably should be included in a single species, D. rugosum. Thinnfeldia is represented in the New Zealand Rhaetic by T. lancifolia, Mount Potts and Owaka Creek, and T. odontopteroides, Clent Hills and Owaka Creek. The forking of the veins of these two species is the basis of Gothan's genus Dicroidium but this classification was not accepted by Arber when he described the Mesozoic plants of New Zealand. In Thinnfeldia the leaves are pinnate, the pinnae being broadly ovate with a midrib and strong parallel secondary nerves arising at an acute angle. Sphenopteris is a genus of fern-like plants with compound leaves with free forking veins. Five species have been named from New Zealand, of which S. otagoensis and S. owakaensis from Owaka Creek are referred to the Rhaetic.

If the reference by H. H. Thomas of leaves of Taeniopteris to Williamconiella is justified then Taeniopteris, usually listed as a fern, page 3 belongs to the Cycadophyta. Taeniopteris has simple, lanceolate or spathulate leaves with strong midrib from which arise at right angles close parallel nerves single or once forked. Two species occur in the New Zealand Rhaetic, T. spatulata and T. thomsoniana. Of these T. spatulata is the most widely distributed. Elatocladus is the name given to coniferous twigs bearing small lanceolate leaves resembling those of Podocarps. E. conferta is found in the Rhaetic localities Mount Potts and Clent Hills, and in several Jurassic localities. Twigs of the Araucarian type with small leaves, named Brachyphyllum, have been found at Owaka Creek.

Jurassic Flora.

Plants of Upper Jurassic age have been collected in New Zealand principally in the following localities. Waikato Heads, Kawhia, Malvern Hills, Hokonui Hills, Mokoia (near Gore), Mataura Falls and Waikawa. More than 40 species are known and consist of Equisetales 2 species, Lycopodiales 1, fern-like plants 18, Cycads 6, Coniferales 15, Angiosperms 2. Several of these are not in sufficiently good state to be named specifically. No evidence has so far been obtained indicating that any of the fern-like plants are Pteridosperms.

Cladophlebis is represented in the New Zealand Jurassic by the widely distributed and common C. australis and by C. denticulata and C. antarctica. As stated above the foliage Cladophlebis possibly belongs to the stems Osmundites of which two species, O. gibbiana and O. dunlopi (including O. aucklandicus) have been described from New Zealand. The fern Microphyllopteris pectinata bears rounded pinnules, 5-6 mm. across. It resembles in shape Cladophlebis reversa and better specimens are needed to define the difference. A beautiful fern with tripinnately divided leaves is Coniopteris hymenophylloides. In the fertile leaves the lamina is reduced and the sori are borne at the ends of the veins. The sori are partly enclosed in a cup-shaped indusium and the sporangium appears to be of the Cyatheaceous type. A second New Zealand species is C. lobata. Dictyophyllum rugosum is found at Mataura Falls, and Linguifolium lillieanum at Malvern Hills, as well as at the Rhaetic localities already mentioned. A species of Thinnfeldia probably T. lancifolia, has been collected at Mokoia and Hokonui Hills. Sphenopteris gorensis and the allied Ruffordia goepperti occur at Mokoia.

Other Pteridophytes in the New Zealand Jurassic are two species of Equisetites and one of Lycopodites. These names denote species allied if not identical with the modern genera Equisetum and Lycopodium. E. hollowayi from Waikawa and Waikato Heads and E. nicoli from Mokoia are represented by whorls of acuminate leaves, those of E. nicoli being longer than those of E. hollowayi. Lycopodites arberi is a herbaceous species with undifferentiated ovate leaves. It was collected at Curio Bay, Waikawa.

page 4

There are about as many Gymnosperms as there are Pteridophytes in the Jurassic of New Zealand. Of the six species of Cycads Pterophyllum matauraensis has pinnate leaves, the pinnae being lanceolate, obtuse, and with conspicuous parallel nerves. Ptilophyllum acutilobum is also a pinnate form but the pinnae are narrow, acuminate and falcate. It has been found at Waikawa. In Nilssonia the leaf is narrow, broadest near the apex, with parallel veins arising from the midrib and with unequal-sized lobes. The two New Zealand species are. N. compta from Waikawa, deeply cut between the relatively narrow lobes, and N. elegans, from Mataura Falls and Waikato Heads, with broad lobes separated by shallow indentations. The general shape and nervation of the leaves of Nilssonia recall those of Taeniopteris. Taeniopteris, possibly a Cycad, is represented in the Jurassic of New Zealand by the large and broad-leaved species T. crassinervis from Mataura Falls; by T. spatulata, generally distributed; and by T. vittata, at Waikawa but doubtfully distinct from T. spatulata.

The family Araucariaceae, to which the present day Norfolk Island pine belongs, seems to have been well represented in Jurassic times. Leaves, cone scales and wood are present in the New Zealand rocks and are referred to five genera, but in life these might represent two or three species. This is because palaeobotanists have to give a generic name to each part of the plant as they are unable to prove which parts make up any particular species. Shoots with small leaves from Mataura Falls have been referred to Pagiophyllum. Large linear-lanceolate leaves are known as Podozamites. Two species, P. gracilis from Waikawa, and P. lanceolatus from Waikato Heads, are known. Cone scales are referred to Araucarites, those from Mokoia being named A. cutchensis and those from Waikawa, A. grandis. Dadoxylon australae from Waikawa is a wood of Araucarian type.

Jurassic conifers other than of Araucarian affinity include the cone Palissya bartrumi from Waikawa. This is an elongated cone with the scales loosely arranged when mature. It is possible that one of the species of Elatocladus represent the foliage belonging to these cones.

The last species belonging to the Jurassic of New Zealand to be mentioned is the Angiosperm Artocarpidium arberi. This was found at Waikato Heads on the same rock as the typically Mesozoic fern Cladophlebis australis. Because of the presence of this dicotyledonous plant Arber referred the beds to the Neocomian, that is, to the lowest Cretaceous, but Edwards thinks that the Waikato Heads beds would better be placed in the Upper Jurassic. Artocarpidium arberi is represented by two leaf impressions. These have the appearance of European beech leaves in which the marginal teeth are little developed. The straight parallel secondary nerves reaching the leaf margin especially recalls the beech family.

Concerning the relationships of the Jurassic flora of New Zealand Edwards says that about half the species have been recorded from page 5 Australia, about a third from India and about a quarter from Graham Land. Possibly this indicates that in the Jurassic there was a southern flora somewhat distinct from a northern one, the separation being the Tethys ocean which stretched from the Atlantic to the Pacific north of Gondwana Land. India accordingly would be included in the southern plant region. How this difference between northern and southern floras became more marked in the Cretaceous will be shown when dealing with that flora.

The Jurassic flora is essentially different from the Cretaceo- Tertiary floras though no doubt many of the ferns and gymnosperms of the Cretaceo-Tertiary are direct descendants of Jurassic genera. After the Jurassic there is a long gap in the New Zealand record and the next plant bearing beds, Upper Cretaceous, are dominated by true flowering plants.

Cretaceous Flora

The early Mesozoic floras are much more uniform throughout the world than are the floras from the Middle Cretaceous and onwards which are dominated by Angiosperms. In the Upper Cretaceous distinct plant provinces have come into existence, the flora of the New Zealand province, for instance, being almost as different from that of the Australian province as it is today. Surviving from Jurassic times there are in the Cretaceous flora ferns and gymnosperms but they are dominated by Angiosperms thus contrasting with the Jurassic flora where fern-like plants are dominant. The podocarps and araucarians are the characteristic gymnosperms in the southern floras both in the early Mesozoic and the Cretaceous as in fact they are at the present day.

The earliest Cretaceous plant bearing beds are found in the Paparoa coal measures in the Greymouth district. In these beds there occur such typically Mesozoic species as Coniopteris hymenophylloides and Cladophlebis denticulata together with a fern of more modern appearance, Pteris, a Gymnosperm like Podocarpus and several species of Angiosperms. This assemblage of plants indicates a middle or early Cretaceous date.

Quite different from the Paparoa flora and apparently of considerably later date is the Cretaceous flora at Shag Point. This flora has much in common with those at Pakawau and Wangapeka (Nelson) here classed as Eocene, several genera and species being common to two or more of these flora. A fern Cyclosorus tertiarozeelandicus, is found at Shag Point and is closely related to another species from Pakawau.

The gymnosperms of the New Zealand Cretaceous belong mainly to two families, Araucariaceae and Podocarpaceae. The Araucarians are referred to the genera Agathis and Araucaria. Three species from Shag Point were named and figured, but not described, by Hector in 1886. They are Araucaria buchanani, A. carinaria, and Agathis lanceo- page 6 latus and to these may be referred, five species described by Ettingshausen. in 1887. Hector's names are accepted as valid. Araucaria buchanani has small leaves somewhat like those of the Norfolk Island pine. A. carinaria has large leaves more or less triangular in section hence the specific name referring to the central ridge. Agathis lanceolata has large flat leaves similar to those of the kauri. A species with broadly ovate leaves, Agathis marshalli, has recently been described from the Kaipara district in beds of Upper Cretaceous age.

The Podocarps from Shag Point include Dacrydium praecupressinum (also from Pakawau), and two species of Podocarpus both found at Pakawau as well. Athrotaxis novazeelandiae, from Shag Point, has Cypress-like leaves and globose cones. Athrotaxis belongs to the family Taxodiaceae and at the present day is found only in Tasmania. The family includes the giant species of Sequoia of California.

Some Angiosperms from Shag Point which have received names and can with some degree of confidence be classified are mentioned below.

Cinnamomum is the name given to leaves having three principle nerves near the base, the next prominent secondaries arising some distance up the midrib. C. intermedium from Shag Point has rather large oblong leaves.

Nothofagus was in the Cretaceo-Eocene represented by several species with larger leaves than those belonging to living New Zealand species of the genus. It is impossible to say that the leaves definitely belong to Nothofagus, and not to Fagus, but as Fagus is not now found in New Zealand it will be convenient to group the New Zealand fossil forms under Nothofagus. N. lendenfeldi from Malvern Hills, an Upper Cretaceous locality, has leaves with the base truncate and the secondaries parallel and reaching the margin which is finely serrate. N. ulmifolia, from Shag Point, has close parallel secondaries and a finely denticulate margin. Leaves of N. ninnisiana, described by Unger from the Oligocene of the lower Waikato district, are hardly distinguishable from leaves found at Shag Point, Pakawau and Wangapeka.

To find an existing flora that, like the Shag Point flora, contains Podocarps, deciduous trees and evergreen trees one must go to the mid-mountain zone in New Guinea. In the Cretaceous there may have been a closer affinity between the floras of the New Guinea and New Zealand regions, and possibly since that time the deciduous element in the New Zealand flora has become extinct. It is difficult to see a cause for such a change. One would think that a general cooling of the climate in New Zealand, as has taken place since the mid-Tertiary, would have the opposite effect.

Acer tasmani from Shag Point was described by Ettingshausen as an Aralia but obviously he was dealing with a leaf impression broken page 7 away at the margin. On account of the entire base and three strong nerves it may be best to refer it to Acer. In any case it shows more resemblance to present day northern hemisphere deciduous trees than it does to any recent New Zealand species. Acer subtrilobatum, from the same locality, is a similar type of leaf. Unfortunately an impression of only a portion of a leaf is preserved.

Eocene Flora.

The plant beds at Pakawau, Wangapeka, Dunstan and Trelissick Basin may be classed as Eocene. These floras have much in common with each other and with that at Shag Point, and like it consists of ferns, conifers and Angiosperms, the latter including trees of both a deciduous and evergreen appearance.

The ferns described from Pakawau can all be referred to recent genera. They are Blechnum priscum, Cyclosorus cretaceo-zeelandicus, Sticherus obscurus, and Pteris pterioides. The Conifers are Agathis lanceolata, Dacrydium praecupressinum, Podocarpus cupressinum, P. maitai, P. parkeri and P. praedacrydioides.

In 1886 Hector named and figured, from Pakawau, a large, ovate, finely serrated leaf, Patete schefleri. This name can be accepted as no modern genus of New Zealand plants would include it. It has, however, no claim to be considered an araliad as both generic and specific names would seem to imply. More likely it belonged to the Tiliaceae. Ettingshausen's Grewiopsis pakawauica is apparently the same species.

Cinnamomum is represented at Pakawau by C. haastii, a species with small oval leaves. Haastia speciosa, from Pakawau, is a very large, broad leaf like that of a banana. Probably it indicated a warm climate at the beginning of the Tertiary period.

At Wangapeka in the Nelson district there occur some ferns, not yet described, Podocarpus trinervia, resembling the recent P. totara and two species of Phyllocladus. P. tanekaha recalls the mountain toatoa; but P. toatoa is quite different having veins occasionally joining together. There are leaves of Araucaria and several species of Angiosperms at Wangapeka, but no description of them has been published. There is a broad-leaved species of Cinnamomum, a narrow-leaved species of Nothofagus, a three-lobed leaf like the tropical genus Sterculia and leaves resembling those of Aristotelia and Melicytus.

Dryandra comptoniaefolia, from the Trelissick Basin, is an elongate, pinnatifid leaf with the segments having one to three nerves arising from the midrib. This is an interesting record, probably correct, as better preserved examples of this genus occur in the Australian Tertiary. Potamogeton ovatum, from the Trelissick Basin, has oval leaves with the characteristic nervation of the genus but it is not conspecific with any recent New Zealand species.

Oligocene Flora.

Plants of this age are known from Landslip Hill and Ohai in Southland and from the lower Waikato district. At Landslip Hill the leaf impressions, in several cases with the nervation preserved in great detail, are on quartzite. These beds are apparently older than those at Ohai and the Waikato. The most conspicuous species at Landslip Hill is Pittosporum elegans, a large-leaved form resembling the recent P. colensci. There are species of Carpodetus, Pomaderris, Araucaria and Dacrydium. Most remarkable, however, is a fern that recalls Linguifolium. It is a large leaf with midrib and parallel unbranched lateral nerves arising at acute angles.

At Ohai there are large leaves of the types of Acer, Castanea and Cinnamomum, relics, as it were, of the Cretaceo-Eocene flora, and there is an Araucaria or Agathis. But with these are leaves of more modern aspect recalling Rubus, Clematis and Hoheria.

The Waikato flora has been described by Unger and by Penseler Unger described an elongate, tongue-shaped leaf with a clasping bas under the name Myrtifolium lingua. It might be a Euphorbia similar to E. glauca. He also described Fagus ninnisiana, a large leaved species which would more conveniently be classed, under Nothofagus. From Pukemiro Penseler records species of Geniostoma, Beilschmiedia and Coprosma, thus showing the close affinity of the Oligocene flora to that of the present day.

Legends to Figures

page 9

Miocene-Pliocene Floras.

We have no knowledge of what kinds of plants existed in New Zealand during most of the Miocene period; for it probably was at the end of the Miocene that the plant bearing beds at Whangaroa and Wharekuri were laid down. At Whangaroa the few plants preserved are for the most part like those of the Pliocene and the present day. A well-preserved leaf is quite similar in appearance to Griselinia. Other impressions remind one of the leaves of Pittosporum, Brachyglottis and Cyathodes. Similarly at Wharekuri leaf impressions that could be classed as Pittosporum and Coprosma are found.

From the Pliocene two floras have been described, both preserved in fine volcanic tuffs. The flora of the Kaikorai Valley includes representatives of 19 genera of which possibly three are not now found living in New Zealand, namely, Parafagus, Patete (equals Ulmophylon) and Kaikoraia. Well knows recent genera like Laurelia, Coriaria, Metrosideros, Coprosma, Senecio and Knightia, occur. The Waipaoa flora from the Poverty Bay district consists almost entirely of recent New Zealand genera. Of 30 genera only two, Platycerium and Apocynophyllum?, are not now found in New Zealand. Characteristic New Zealand genera found at Waipaia are Carmichaelia, Rhopalostylis, Plagianthus, Hebe, Coprosma, Paratrophis and others.

Legends to Figures

The illustrations have been from the following sources: Nos. 14, 15, 29-30 original; Nos. 1-3, 8-10, 12 after Arber; Nos. 4-7, 11, 16-18, 25, after Buchanan (ined.); Nos. 19-24, 27 after Ettingshausen (Trans. N.Z. Inst. Vol. 23); Nos. 26, 28 after Unger; No. 13 after Edwards.

page 11