Stream, while in the South Karori district points which must have been near the middle of the old valley-floor are 200 ft. to 300 ft. higher. On the other hand, the manner in which the branch valley joins the main one suggests a direction of flow from north to south.

Post-Kaukau Uplift. The sub-senile land-mass was elevated about 600 ft., and immediately active degradation was resumed during a fairly long stillstand until the streams, following courses inherited from the previous cycle, became graded in their lower reaches with respect to a base-level not greatly different from that of to-day. Cotton, however, postulated a change of level equal only to the vertical interval between the Kaukau surfaces and those of his Tongue Pont Cycle, represented by the major Makara terraces and the lower Tongue Point platform. The reasons for favouring the greater uplift involve consideration of the Makara fossils which will now be dealt with.

The Makara Fossils. Fragments of massive, marine, worn and broken oysters are contained in a fine conglomerate consisting of fairly well rounded greywacke and argillite debris in a sparse, muddy matrix which outcrops in a small branch of the Makara Stream. The deposit was first described by A. McKay (1877, p. 50), who considered it to be Miocene in age. F. K. Broadgate (1916, p. 78) thought that, on the physiographic evidence alone the beds could not be younger than Pliocene, and that they are a remnant of a once extensive Tertiary cover preserved from erosion locally at Makara by down-faulting; but he cited no facts in support of his contention.

Dr. Marwick informed the writer that merely from the massive type of oyster that the indeterminable shell fragments represent, and from the degree of induration of the matrix, he would infer a probable Pliocene age for the deposit. Against Broadgate's faulting hypothesis is the absence of signs of the implied post-Pliocene faulting in the vicinity, although zones of intense faulting show in the greywacke nearby. It is unlikely that a portion of a widespread Tertiary cover has been preserved in a depression of the undersurface as other relics would surely have been found. The deposit is confined between fairly steep greywacke walls, but the actual contact with the basement rock is not exposed. A fossiliferous bed about 2 ft. thick underlies 6 ft. of barren conglomerate, while at another locality, 200 ft. distant across the valley, poorly rounded fine conglomerate indistinguishable from the barren layer merges upwards without visible break into consolidated gravels such as make up the bulk of the main Makara Valley terraces. Stratification is poor, but neither folding nor tilt were detected.

Re-submergence. The writer considers that the post-Kaukau Cycle elevation was sufficient to enable the sculpturing of youthful valley systems below the previous cycles to a lower level than the floor of the present Makara Valley, but that before much widening was effected a reversal of movement permitted estuarine waters to invade the land, the Makara shell-bed being deposited at the head of an estuary. Submergence continued, but was either intermittent or slower than the initial plunge, so that bay-head progradation was able

not only to maintain the shoreline, but evidently also to cause it to advance down the valley. Delta deposits and then flood-plain alluvium were laid down in turn over the littoral material, deepening as depression of the land continued, until base-level was raised to the level of the highest set of marine erosion benches on the south coast. Then occurred a pause long enough to allow of the sculpture of these features.

The Tongue Point Cycle. At the time of the above-mentioned pause, valleys carved earlier from the Kaukau surface were deeply filled with alluvium. This is the interpretation placed upon the terrace gravel deposits of the middle reaches of the Makara Valley, of the “Long Valley,” and perhaps of Tinakori Valley, laid down in eroded depressions below the level of the Kaukau Cycle remnants, although the Tinakori example is obscured by the Wellington Fault. The nearby important valley system of the Hutt contains a great thickness of gravels which may have been deposited under similar circumstances. Following upon the stage of maximum submergence, a series of earth movements, probably with intervening halts, then brought sea-level to its present position. A major halt in uplift saw the formation of the wide, lower, wave-cut bench at Tongue Point, re-excavation of the Makara Valley, and its widening to a mature cross-profile with walls outside the ancient gravel-filled trough. If a Pliocene age may be accepted for the Makara material, it would follow from the above discussion that the Tongue Point period of erosion occurred since that time, but that the Kaukau Cycle is earlier.

Makara-“Long Valley” Capture. The high-level terrace remnants near Karori Park mark the most southerly traces of the floor of the “Long Valley” of Cotton. A south-westward flowing branch of the Makara River of Tongue Point times, as mentioned earlier, worked back northwards and captured the headwaters of the “Long Valley” which also then evidently possessed a widely flaring cross-profile, but was dominated over by its westerly neighbour, which probably had the advantage of a more direct course to the sea.

Cook Strait Foundering. A glance at the Admiralty chart of Cook Strait (N.Z. No. 695) shows extreme irregularity of the sea floor off the south Wellington coast, except where masked by sediments opposite the entrance to Port Nicholson and the Wairau River mouth, strongly suggested a drowned land surface (Cotton, 1918, p. 325). There is a rather abrupt change to shallower depths about the 50-fathom line, inshore from which is a more even submarine bench, free from sediments in many places, and ascribed to wave-planation acting at a depth somewhat greater than usual owing to the effect of strong tidal scour and the prevalent stormy conditions of the Strait preventing accumulations of protective debris. The line of sudden deepening is believed by the writer to mark approximately the position of a fracture separating the subsided from the upstanding-block. The drowning occurred after the Kaukau Cycle, but necessarily before the uplift of land preceding the Tongue Point Cycle to account for the raised wave-cut platforms. He favours a period co-eval with the sinking discussed in connection with the Makara

fossils, and considers that the subsidence of the southern block may have continued after the northern block halted, with dislocation along the fracture referred to and that the submerged area did not take part in the subsequent rise of the peninsula. During each pause in uplift, the coast has receded from the fracture under wave-attack. To-day recession is proceeding rapidly, and the smaller streams flowing to the south coast cannot cut down rapidly enough, reaching the shore in cascades.

Cotton considered (1912, p. 254) that the west coast of the peninsula also is derived from an initial fault coast. It is thought, from the relatively more advanced development of the streams flowing to the west coast compared with those flowing to the south, and from the more mature condition of the west coast, that the foundering of land to the west occurred at a more distant date.^*

Karori-Makara Capture. Following on the formation of a new south coast much farther north, rapid south-flowing torrents dissected the older valley-floors, steadily pushing the divides northwards. These streams had the benefit of a short steep course to the new coast, and more rapidly experienced the advantages of recurrent uplifts of the land; so it is not surprising that the largest and most successful of them, the Karori, should in time break through into the Makara system, and effect an important capture. One by one the headward tributaries of the Makara were diverted to the south coast, and finally the important branch that effected the “Long Valley” capture vas claimed, giving rise to the Karori Stream as we now know it, flowing in a deep valley far below the remaining traces of the earlier Makara southward extension.

A western tributary of the Karori became established in the main Makara Valley, causing the migration of the divide farther northwards, and the formation of the higher set of terraces south of the Makara Golf Course, the terraces along the South Karori road, and the change of slope of the walls of the narrow upper Karori Valley.

Post-Tongue Point Uplift and the Port Nicholson Subsidence. King (1930, p. 502) could not make out whether a uniform uplift terminated the Tongue Point Cycle, followed by a later deformation that tilted the coastal platforms and Makara terraces. It seems logical to associate these tiltings with the movement that depressed beneath the sea the Port Nicholson area (Cotton, 1912, pp. 250-4) as phases of the same disturbance. The effects would be produced by a rotation about a north-north-east axis a little east of Karori Stream, lowering to the east and elevating to the west.

The Karori Stream had no extensive floodplains such as would leave terraces to bear witness to the uplift in the lower reaches. The western tributary dissecting the Makara Valley had its gradient increased by the warping. The main stream would have been hindered as it flowed obliquely up the dip of the warp, but its steep gradient coupled with rejuvenation must have offset the adverse effects. Cotton