terraces rising to hundreds of feet above sea-level, we must conclude that the former existence of such land in the present orogenic cycle is extremely improbable.

F. W. Hutton (1871, p. 393; 1899, p. 178) stated that: “An elevation of 500 feet would obliterate Cook Strait” and considered that it had been formed by more or less general subsidence. Hutton's work is less detailed than that of Crawford and adds but little to the Wellington geologist's views.

C. A. Cotton (1918, p. 324), discussing the origin of Cook Strait, has regarded the subsidence theories as inadequate because they require a downward movement for which there is no evidence, and are opposed to the clear evidence of uplift displayed along the present coast.

Class B.—Vertical Movement.

The three protagonists of this class of hypothesis are: Hochstetter (2nd hypothesis); E. Suess, who accepted Hochstetter's account; and Cotton.

F. von Hochstetter (1864B, p. xlvi) expressed his later views on the origin of Cook Strait as follows: “Indem aber das Land durch Hebung, durch Anschwemmung, und durch Hervorbrechen der Vulcane einen nicht unbedeutenden Zuwachs erhielt, versanken andere Theile gleichzeitig in die Tiefe. Einen solchen Ereignisse mag die Bildung der Cooks und Foveaux Strasse ihren Ursprung verdanken” (When the land had acquired a not inconsiderable increase in size through elevation, through alluvial deposit and through the ejectamenta of volcanoes, another part sank into the deep sea. Such an event, a sunken origin, caused the formation of Cook and Foxeaux Straits.). Later (p. 263) he wrote: “Der Senkung des Landes wahrend der Driftperiode schriebe ich die Bildung der Cooks und Foveaux Strasse zu.”

Hochstetter believed that the main mountain range of the North Island, which is continuous from East Cape to Wellington, was represented in the South Island by the Kaikoura Mountains between the east coast and the Awatere River.

This interpretation was accepted by Suess who, in his mighty work “The Face of the Earth” (vol. II, p. 144), wrote: “Hochstetter had already suspected that Cook Strait and Foveaux Strait, which separate the three islands, were due to the subsidence of mountain blocks; he was also aware that the mountain chain which follows the east coast of the North Island from East Cape to Wellington is continued on the other side of Cook Strait between the east coast of South Island and the River Awatere, and that this continuation lies to the east, outside the trend of the principal chain.”

Marshall (1911, p. 57) has put the same statement in another way: “One looks in vain for a mountain range in the North Island which might be regarded as a continuation of the dominant range of the South,” and Cotton (1916, p. 317): “In the North Island the absence of a continuation of the main mountain range—the

Southern Alps—of the South Island has often been remarked upon and the statements of Hutton and Suess on the subject are perhaps correctly interpreted as indicating their belief that the north-eastern continuation of the Alpine range has subsided independently. In view, however, of the late date of the movements to which the South Island ranges owe their present height, and in view also of the presence in the North Island opposite those ranges of Tertiary rocks of greater age than the orogenic movements, it would be more correct to say that this portion of the North Island has merely not been uplifted to the same extent as the South Island.”

Obviously the distribution of the various rock formations and mountain ranges was causing these theorists a difficulty which became more apparent as time went on. In the latest attempt to establish the hypothesis, Cotton (1918, p. 325) accepted the general idea as correct; but, realizing that the age of some of the rocks on the margins of the Strait required that its formation should be geologically very recent, he substituted for the “subsidence of earth-blocks” the conception of “earth-blocks which had failed to rise” with the neighbouring areas. A further difficulty, however, arose (noted by Cotton in the case of the Moas) in that the nature of the fauna and flora on the two islands near the strait requires that they should, at one time, have been joined. This he overcame by adopting the tentative hypothesis of connection after the major orogenic movements with subsequent separation as a result of the subsidence of blocks, “possibly contemporaneous with the partial subsidence of an adjacent portion of the South Island” (the Marlborough Sounds). This statement contradicts his original idea of earth-blocks which failed to rise, and tends to revert to the standpoint of Hochstetter, which he had already recognised as inadequate.

As will be seen later, these hypotheses fail to account satisfactorily for many of the facts of geologic structure of the neighbouring land areas which are now known. The recognition (King, 1937, p. 44) of the vertical displacements of Marlborough as expressions of a deep-seated compression of the New Zealand earth-ridge now gives ample scope for the concept of a lateral thrust of the island masses, permits their earlier connection, and thus fits the facts of biologic distribution more closely than will any combination of purely vertical movements besides bringing into harmony the present staggered eastern coasts, the offset mountain ranges and the geology of Marlborough and Wairarapa. The combined weight of the opinions of the authorities cited has, however, tended to discourage the discussion of the problem from other viewpoints.

Class C.—Lateral Displacement (Dislocation).

Though three geologists have already proposed theories of this type none has adequately discussed any aspect of the matter other than the obvious fact that the coastlines and mountain ranges of the islands are not continuous but are offset at Cook Strait. As the present writer presents an hypothesis of this type their pronouncements will here be quoted in full:

Fig. 1—Structure Map of North-Eastern Marlborough. Arrows indicate directions in which the earth-blocks are tilted.

and direction to that given above, i.e., the southern portion of the united islands moved to the west, the Marlborough Sounds block tending to stay with the North Island and a split developing between it and the mountainous country of North-west Nelson. When further movement of the South Island relative to the North occurred, the Marlborough Sounds block was also carried to the west, opening up a new rift which now forms Cook Strait.^*

Fig 3—Representation of the deformation undergone by two of the ancient strandlines of Cook Strait.

The terraces of Palliser Bay, moreover, exhibit a tilt to the North so that, if we may regard the cut marine bench and the covering deposits as we would a normal sequence of stratified rocks, we would define their formation as in the form of a broad asymmetrical syncline, one limb of which is partly transected by a fault, and which pitches gently to the north. That this movement has taken place progressively is shown by the fact that the older terraces are more strongly warped than the younger, the relation generally being similar to that of the older and younger terraces at Cape Terawhiti and Tongue Point (see King, 1930, pp. 500–2).

The combined warping and northward tilt are, no doubt, primarily responsible for the depressions occupied by Lakes Onoke and Wairarapa.

These features indicate the former presence of forces of compression in the Cook Strait region, though they represent probably only the final results of the application of such forces. The writer believes that one of the earlier results was the formation of Cook Strait itself.

Attention may also be drawn to the Marlborough Sounds. This rias-system has been produced by the partial drowning of a land surface of strong relief with the dismembering of its mature valley-systems. Cotton (1913, p. 318) has estimated the amount of drowning as 250–300ft. The fact that it is physiographically so distinct from any other area near Cook Strait and that it is bounded to east and west by major fractures indicates that it has had a history which differs considerably from that of neighbouring areas (Figs. 1, 4). This fundamental difference has not been sufficiently taken into consideration in previous theories bearing on Cook Strait.

Note on the Origin of the Cook Strait Margins.

There can be no doubt that the outlines of Cook Strait are determined by fractures. The north-west side of Wellington Harbour, the west side of Palliser Bay, the east and west coasts of the Marlborough Sounds Block and the western side of Tasman Bay are all governed by the occurrence of prodigious faults, which can be traced inland. Practically all the remaining coasts yield, on close scrutiny, evidence of having formerly belonged to this type. Several have been discussed by Cotton in a classic paper (1916, pp. 20–47) in which the criteria to be applied were set forth. The application of these criteria to the other steep coasts of Cook Strait^* reveals that many of them, too, originated as fault-coasts. The southern coast of the North Island, the western coast from Cape Terawhiti to

on regional grounds alone; but all that is at present required is a sufficient knowledge of the stratigraphy to permit matching of formations in the two islands. With this must go an understanding of the structure adequate for similar tectonic matching and explanation of the movements which have taken place.

The Moutere Gravels.

Much of the western rampart of the Southern Alps is bordered by outwash plains of gravel, which probably vary considerably in age from place to place. Their equivalents in the Nelson District are the Moutere Gravels, a deposit of coarse conglomerate certainly exceeding 1000 ft. in thickness and possibly in excess of 2000 ft. (Hector, 1879, p. 37). These occupy the triangular area at the head of Tasman Bay up to the convergence of the North-west Nelson Mountains and the Spencer and St. Arnaud Ranges. On either side, between the gravels and the faults bounding the mountain ranges, a narrow strip or series of lenticular outcrops of somewhat older Tertiary strata occurs, the Jenkin's Hill beds of Nelson and the Glenhope beds of the western side.^*

It is essential to the writer's hypothesis that the Moutere Gravel beds should be relatively young and occupy the rift formed by the westward migration of the North-west Nelson Complex of blocks in the first stage of the dislocation. Let us see what Henderson, the latest authority on the Moutere Gravels, says (1929, pp. 287–288): “The weathered Moutere Gravels occupy most of the floor of the Nelson graben from Tasman Bay to the St. Arnaud Range, mountains of quartzose greywacke. The gravels are mostly of the same rock and are in fact the piedmont deposits from the mountains of which this range forms a part; it should be noted that the highlands east and west of the graben are of quite different rocks and supplied but an insignificant amount to the gravels. These deposits, which, so far as known, are not folded or faulted, overlie the conglomerates, sands, clays, and lignites of the Glenhope series. The pebbles in the conglomerates of these latter beds are of the same igneous rocks that form the adjacent high country east and west of the graben. The Glenhope beds are upturned near the edges of the lowland and are there unconformable to the overlying Moutere gravels, but except near the bounding faults they lie flat and grade upward into the greywacke gravels.”

A further statement by Henderson (p. 288), made without any consideration of extensive rift-opening, may also be quoted in support of our thesis: “The sea does not seem to have covered the area of the Nelson graben until late in Ototaran time, perhaps even till the Hutchinsonian. The beds of this transgression were tilted and raised above the sea before the valley plain deposits of the Glenhope series accumulated. These movements were probably the prelude of the Kaikoura orogeny, a deformation that culminated after Glenhope time and raised the mountain mass of which the St. Arnaud

Range is a part. The Moutere Gravels, the piedmont deposits derived from the highlands then formed, are correlated with the Castlecliff beds, and the Glenhope beds with the Nukumaru beds.”

A slightly different viewpoint reveals here an example of sedimentation in an opening rift, the formation of which began in the middle Tertiary, but was continued into the Pleistocene.

The Marlborough Sounds District.

Since the time of McKay's (1879, 1879B, 1890, 1890B) rapidly executed investigations, no geological research has been undertaken other than Cotton's physiographic study of the Tuamarina Valley (1913), Morgan's investigation of the occurrence of coal at Picton (1921B, pp. 11–18) and Henderson's papers upon the gold deposits about Wakamarina (see 1935).

McKay showed that two main rock types are represented, each of which contains probably more than one series. The central mass, elongated north-east to south-west, is almost entirely schist. On this no petrological work whatever has been done and ignorance of the age, constitution and relations of the schist is profound. Certainly nothing short of a large-scale government geological survey with competent co-lateral petrological advice is likely to yield any adequate account of the Marlborough schist area. The writer, in the hurried period before his departure, endeavoured to ascertain some of the facts of structure; but the results are too disconnected for separate publication and did little more than lay bare the magnitude of the problems involved. The schist extends far to the south-westward as is indicated by Park's observations in the Spencer Mountains (1885, p. 357).

In the North Island, the only counterpart of the Marlborough schist is on Kapiti Island, some seven miles from the mainland (Ferrar, 1928, p. 314). Of this, only intensely weathered specimens are available which are useless for petrological examination and comparison with the Sounds material. This occurrence is of great importance to our argument.

Correlation.

In each island, the physiographic and stratigraphic backbone is composed of a system of alternating greywackes and argillites to which a Mesozoic age is usually assigned. West of the backbone, the correspondence of the Nelson rocks with those of Taranaki is, at first sight, poor; but closer inspection reveals evidence (Marshall and Murdoch, 1920, p. 118) that the shattered granite and Palaeozoics of North-west Nelson probably pass below Cook Strait and underlie the late Tertiary rocks of Taranaki.

Eastward of the backbone the match is surprisingly good. The Awatere beds correspond closely with the Tertiary rocks of Palliser Bay (King, 1934, pp. 12–13, 1936, p. 31; Thomson, 1919, pp. 281–282), and the Haurangi Mountain greywackes evidently belong to the Kaikoura Group. The latter do not reach Cook Strait owing to a torsional effect in the structure of the north-east corner of Marlborough, where, as the writer has already recorded (1934, p. 9), the

deep-seated movement of compression. Those to the west are bounded, and separated range from range, by immense reversed faults dipping east. Those to the east, the Kaikouras, are separated in the same manner by fault angle valleys with mighty reversed faults heading to the west. The two systems meet near the Wairau fault which, though truly a scissor fault, is mainly of the Kaikoura type. West of it is a scar of schist, and to the east, the Tertiary sediments of the Wairau-Awatere district. In the North Island the western type of structure is certainly present north from Wellington; but the Rimutaka fault is of the eastern type, and the composition and structure of the Haurangi Mountains, as far as these are known, permits them to be called the Kaikourides of the North Island. Thus, not only do geological formations have a similar distribution; but also similar or corresponding structures.^* As, however, the lateral dislocation is apparently later than the movements which gave rise to the present geological distribution, or was, perhaps, a final expression of them, it is to be anticipated that there will be distortions of major faults and trendlines in the immediate neighbourhood of the Strait.

These are shewn in North-eastern Marlborough where, as the writer has previously remarked (1934, p. 9), the faults bend from a north-north-east direction to the east (Fig. 1). Even the Great Clarence fault, transecting a previous overfold, is no exception. This may be interpreted as a drag effect, consequent on the dislocation, and caused more immediately by the fact that this segment of the South Island would be more closely in contact with the North Island than any other. Also, we have noted in our reconstruction (p. 548) that, during the relative movement of the South Island to the west, a segment of Nelson was left behind to form the stratigraphically and physiographically distinct block of the Marlborough Sounds. With farther straying of the South Island to the west, torsional structures have been developed in the north-east corner of Marlborough due to the resistance of the ancient rocks of the Marlborough Sounds block.

West of the main ranges is the relatively sunken area of Tasman Bay. With the triangular southern district infilled by Moutere gravels this is usually designated a graben, on the hypothesis here presented it is a split. The fault along the eastern margin is of the reversed type, that to the west has not yet been described. Observations by the writer indicate that only minor disturbance is shewn by the younger beds in the relatively downthrown area and this fracture may well be of the tension type.

A word of explanation may perhaps be offered regarding the absence of recent volcanic rocks along this assumed line of tension

of the two islands in the recent past. It would be interesting to know whether species of Paryphanta inhabit the Haurangi Mountain region and if so, to what group they belong.

On the whole, the marine molluscan evidence is disappointing in view of the concentration of effort on this group in recent years Moreover, not only are the relevant data scanty, but there seems small prospect of further work proving of value in a discussion of the relation between the two islands.

Insect Evidence.

No comprehensive account of the insect faunules has been written so the writer made a survey, from published lists of species, of the Lepidoptera found in one island only. From this survey were expunged those species which have been collected only in regions remote from Cook Strait, leaving a small residue of species which occur only in one island and near the shores of the strait. The writer then endeavoured to ascertain whether these were species which would have been definitely located in their present districts in any case, e.g., confined by altitude, or whether the presence of the Strait had exerted a definite restriction on their distribution. Not being a trained entomologist he soon found himself in difficulties, the mountainous nature of the terrain offering considerable opposition to insect dispersal and yet not wholly forbidding it, while the felling of bush and the destruction of natural habitats introduced factors which were difficult to evaluate. Entomologists consulted were unable to make any final statement on the matter. At the same time the distribution of the Coleoptera was similarly treated; but with a like result, namely: that there are definite species inhabiting regions adjacent to the shores of Cook Strait in one island, but not passing over to the other, is certain; but whether their restriction is due to the prior presence of the Strait or to other natural controls is at present unknown.

Myers (1926), however, has made observations on the Heteroptera which are significant in the light of the present discussion. He records (p. 455) nineteen species from both islands, ten from the North Island only, and three from the South; but the most outstanding result of his work is the discovery that those North Island species which are found in the South Island extend only to the northern portion of that island and follow the plant distribution. Although about half the species occur on trees, it is unlikely that the change of botanical province will govern their distribution so closely as to confine them north of the 41st. parallel of latitude.

This similarity of distribution between two groups of organisms, neither of which is absolutely dependent on the other, is worthy of further investigation and points to the conclusion that they conform to an older order of things and that the formation of Cook Strait is a recent feature.

Various other groups were examined by the writer and it was found that, though in individual groups uncertainty may exist, the trend of the biologic evidence as a whole admits of no mistake. The plant distribution alone renders former closer connection of the two

(Pliocene-Post-Pliocene). This movement was probably between 20 probably greywackes, must have been forced down and thrust below and 25 miles. When the movement took place, certain formations the Marlborough schist^† along the Wairau, and possibly other, fault-lines. At the same time torsional structures were developed in the north-east corner of Marlborough. I have noted elsewhere (1934, p. 9) how the north-eastward trending structures of the Kaikouras swing to the east in the Awatere-Ure River district. Possibly the eastward dipping limestones and marls of the Limestone Range block and the Benmore are are extreme cases of this (Fig. 1). These movements, obscure when considered only with respect to their own vicinity, become intelligible when interpreted as due to thrust against the rigid Sounds block.

On such an hypothesis, the shores of Cook Strait should be determined by fractures, and Cotton has demonstrated that this is so Moreover, the strong warping and tilting of ancient shorelines with respect to one another and to present sea-level is a clear indication that compression was an active agent in their elevation. The frequency of earthquake shocks and earth movements in the area must also be taken as an indication that final equilibrium has not yet been attained.

All movements—the remark is trite—are relative, and in a case such as this it may be argued that the result is the same which ever land mass underwent the major earth movement (South Island to the west or North Island to the east). This may be so between them; but with earth forces such as must have been in operation here, we should expect that the amount of deformation in the moving mass would be greater than in the stable one. If this is true, then undoubtedly the South Island moved, for the deformations of eastern Marlborough are much greater than those of the corresponding area in the North. Indeed, they display the only folding of considerable magnitude known from the Kaikoura orogeny and the greatest fault displacements recorded in New Zealand. We may say that the Island moved under the influence of thrust-forces operating at a maximum in Eastern Marlborough. Elsewhere, also, I have referred to under thrust from the south-south-east as responsible for the rise of the Kaikoura and Haurangi Ranges and classed these as the youngest

mountain-forming movements of the Kaikoura Orogenic Cycle. It may be noted that the highest mountains developed are the Inland Kaikouras in the Tapuaenuku-Mt. Alarm area, at or very near the point where the island-separating forces were at a maximum.

Furthermore, these mountains are all characterised by strong reversed faults with scarps facing to the east and their correlation with the Haurangi Mountains forms one of the most satisfactory features of the hypothesis. No longer is the Wellington-Tararua system of ranges with its westward-facing reversed faults to be correlated with wholly eastwardly directed Kaikoura system, as has been done since the time of Hochstetter.

The hypothesis requires also, if the Kaikoura Mountains are to be correlated with the Haurangis, that these elevations, together with the main ranges of the islands, should have been formed before Cook Strait. Moreover, the dislocation must have been sufficiently recent for the survival of topographical features which antedate it. In this connection we may instructively turn to published accounts of that classic area of deformation, the Swiss Alps.

In the eastern Alps many cases are known of mountain ranges transected almost at right angles to their trend by faults of horizontal displacement. For these the term “Blatt” was introduced by Suess (1885, pp. 115–126). “Flaws” and “shift-planes” are terms which have been used later. The continuity of the mountain ranges is broken, in some cases the ridges being so displaced that they block opposite valleys to form lakes. Some of these are, even yet, in existence, e.g., Fahlensee, Seealpsee (Heim, 1921, p. 369). “Flaws” are also well known in the Jura Mountains. In the European examples, the maximum movement on any particular “Flaw” does not exceed 800 metres, Cook Strait is on a much larger scale. It should be noted that these fractures are found only in the upper, folded formations, dying out in the material beneath the folds.

Suess has made the following pronouncement upon them: “In their normal form flaws are produced by a movement of two adjacent portions of the earth's crust in the same direction; but to an unequal extent. The parallelism of the movement of the two parts is often only present as regards direction, the dip of the beds being much steeper on one side than on the other.” Thus, according to Suess, both sides move in the same direction; but one to a greater extent than the other. They are clearly the result of shearing due to stretching along the arc of the folding, and as Heim (1919, p. 615) has pointed out, they occupy not the ends but the central portion of the arc.

Bucher (1932, p. 247) has succinctly summed up the essentials of the Jura Mountain flaws: “The relation of the flaws to the anticlines shows that they came into existence in the course of folding, after the main lines of folding had been established, but before they were completed. The major axes of folding can be recognised on both sides of the flaws yet the final pattern of the folds shows a good deal of independent movement on opposite sides of each fracture.” Substituting Cook Strait for the term “flaw” and (partly) block mountains for “anticlines and foldings” he has written a description of the central portion of New Zealand.