Art. XXIII.—The Stratigraphy of the Tertiary Beds of the Trelissick or Castle Hill Basin.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 49, 1916, Page 321

Art. XXIII.—The Stratigraphy of the Tertiary Beds of the Trelissick or Castle Hill Basin. By R. Speight, M.Sc, F.G.S., Curator of Canterbury Museum. [Read before the Philosophical Institute of Canterbury, 6th December, 1916; received by Editors, 30th December, 1916, issued separatety, 30th October, 1917.] Plates XXI, XXII. Table of Contents A. Introductory B. General Physical Features. C Stratigraphy       1 Statements of Previous Workers; McKay and Hutton       2 Sequence, Thickness, and Arrangement of Beds in Particular Areas:—          (a) Lower Members of the Tertiary Sequence occurring in the Broken and Porter Rivers, Whitewater and Coleridge Creeks, &c; Volcanic Action.          (b) Occurrences of Limestone with Interstratified Tuff in the Porter and Broken Rivers, at Castle Hill, in Whitewater, Volcanic, Coleridge, Waterfall, and. Murderer's Creeks; Nature and Stratigraphical Relationship of the Limestones          (c) Pareora Beds occurring in the Thomas River, Home Creek, Porter River, Whitewater Creek, &c; no Evidence of Unconformity. D Tectonic Features E Volcanic Rocks F Palaeontology:—       1 Lists of Fossils from Particular Localities       2 Considerations of these Lists       3 General Table showing Occurrence and Range of Species. A. Introductory. The Trelissick or Castle Hill Basin, which is situated in the heart of the mountain region of Canterbury, has attracted considerable attention on account of the interesting outlier of Tertiary sediments located therein. It is somewhat surprising that no adequate reference is made to it in the writings of Haast, but its leading physical and geological features, with lists of fossils, have been given by McKay* A Mckay, Reports of Geological Explorations during 1879–80, 1881, p. 53 (with map by Hector) and by Hutton,† F. W. Hutton, Geology of the Trelissick or Broken River Basin. Trans N Z Inst, vol 19, 1887, p 392 and in connection with the former Hector gave a summary of results in the Progress Report introducing the results of the geological explorations for the year. A brief reference is also made to the locality by Marshall, Speight, and Cotton,‡ P. Marshall, R. Speight, and C. A. Cotton, The Younger Rock-series of New Zealand, Trans N Z Inst, vol 43, 1911, p 390 and to its features as an intermontane basin by the present author.§ R Speight, The Intermontane Basins of Canterbury, Trans N Z Inst, vol 47, 1915, p 341 There is a reference to the occurrence of certain fossils in this locality, with descriptions, in various lists by Hutton.∥ F. W. Hutton, Catalogue of the Tertiary Mollusca of New Zealand, 1873; Mollusca of the Pareora and Oamaru Systems, Proc Linn Soc N S W., vol 1, 2nd ser., 1887, p. 205

This comprises all that has been written on the locality, and the present paper is an attempt to add to the record of fact in regard thereto, especially in its bearing on the conformity of our Tertiary series. The situation of the basin almost midway between the classic districts of Amuri Bluff and Waipara on the one hand, and Pareora and Oamaru on the other, adds to its interest and importance and value as a critical locality, even though its stratigraphy is complicated by volcanic action and by structural movements. It may be urged that if the resolution of our problems of conformable or unconformable succession are a matter of such serious difficulty in areas where these disturbing factors do not exist little help can be obtained from such a disturbed area. There are some features, however, which render it peculiarly useful for study in this connection. The position taken by the author in this paper is that the locality has not furnished up to the present any positive evidence of uncontrovertible value of the physical unconformity in the Tertiary sequence; but this statement is not to be interpreted as indicating that none will be subsequently demonstrated. There are difficulties which can be explained on a basis of unconformity, though other explanations are equally applicable; but the position of the breaks will be chiefly determined on palaeontological grounds, and in order to do this it is necessary that systematic and complete collections be made from well-defined horizons. Great as were the services rendered to New Zealand science by Mr J. D Enys during his residence in this remote locality, he would have materially helped geological investigation had he realized the importance of this point. It is extremely unfortunate that his collections were inadequately labelled as regards place of origin, and the result is that the fine collection of New Zealand fossils in the Canterbury Museum identified and described by Hutton—many of them types—cannot be assigned with certainty to a particular bed, such indefinite place-names as “Porter River” or “Broken River” conveying little information of value from the standpoint of stratigraphical geology This deficiency has been remedied to some extent by the kindness of Dr. Allan Thomson, who has furnished me with a list of the fossils collected by McKay and at present in the Geological Survey collection, as well as a list of those collected by Enys, also of the fossils collected by himself in conjunction with the author in November, 1914—all recently identified by Mr. H Suter. The last-named has most kindly identified for me a number of specimens collected in the summer of 1915–16 B General Physical Features The general physical features of the district have been described by both McKay and Hutton in their accounts of its geology, and a reference to the origin of the basin has been made by myself in the paper on the intermontane basins of Canterbury referred to above The general sequence of events indicated therein is as follows: The Trias-Jura sedimentaries, of which the encircling mountains are composed, were folded by lateral pressure into a mountain area probably in early Cretaceous times. This was reduced to a peneplain by the end of the period, or perhaps a little later, and the land was then depressed below sea-level, the submergence continuing during the greater part of the Tertiary era, and on the eroded surface of Trias-Jura rocks a series of sedimentaries, including sands, green-sands, limestones, calcareous sands, and conglomerates, were laid down in

conformable sequence. The whole area was then elevated with differential movements, either of faulting or folding, or both combined, which resulted in the formation of a basin-shaped hollow, some five miles by three, in which the Tertiary sedimentaries were preserved from the active eroding agents which removed the weaker beds from exposed positions either on the elevated country in the neighbourhood or in the valleys more directly subject to glacier and river erosion. In all probability the area was a snow-field at the height of the glaciation; but there is little sign of glacier action in the area itself, except in the conchoidal hollows at high elevations in the mountains, where corrie glaciers probably nestled, and at the northern and southern end of the basin, where the truncation and faceting of spurs does suggest that glaciers invaded the area from the Waimakariri Valley Fig. 1.—Geological sketch-map of the Castle Hill Basin. (Scale, 1 mile to ½ in approx.) 1, greywacke; 2, sands, greensands. marls, &c; 3, tuff beds; 4, limestone; 5, sands, sandy clays, and shales—Pareora beds on the north and the Rakaia on the south (Plate XXI, fig. 1). But this evidence is not supported by the other signs of glaciation, such as lateral and terminal moraines, rounded and scratched surfaces, roches moutonnées, &c, and it is therefore possible that the truncation of the spurs may be due to faulting and the facets may be fault scarps of recent origin. These would be preserved at the ends of the spurs, while the course of the faults across the intervening valleys would be masked by the rapid and enormous accumulations of waste.

The general surface of the floor of the middle of the basin consists of a series of fiat benches, sometimes half a mile in breadth, planed by stream action from the less resistant sedimentaries, and covered with a thin veneer of gravel either washed out of the upper beds, which are composed of conglomerate, or derived from the encircling mountain ring and deposited by the present streams when they ran at a higher level. This is bordered on the eastern and western side, and to a certain extent on the northern side, by limestone hills, which project through the floor to a height of several hundred feet. The whole area is deeply dissected by streams which flow in beds incised at times hundreds of feet below the general floor, and bordered by narrow terraces cut in the Tertiary sediments and covered with thin veneers of gravel (Plate XXI, fig. 2). When cutting through the limestones the river-channels are narrow and gorge-like, and the sections thus furnished enable a clear insight to be obtained into the structure of the borders of the basin; but the mantle of river-gravels renders the exact interpretation of the central part a matter of extreme difficulty and a subject for speculation; this especially applies to the farm land immediately to the east of Castle Hill and to the country in the angle between the Thomas River and Hog's Back Creek, one of the largest feeders of the Broken River in its higher reaches. C Stratigraphy. 1 Statements of Previous Workers. The most interesting problems suggested by a consideration of the stratigraphy of the area are those connected with the question of conformity and the correlation of the beds. In order to have a basis of discussion it will be best to state the respective opinions of McKay and Hutton, and indicate the points on which they differed. McKay's classification, no doubt influenced largely by the ideas of Hector, is as follows — Table of Formations Age Formation Character of the Beds Lower Miocene 1 Pareora formation Principally loose sandy beds, with shell beds and beds of shelly sandstone Upper Eocene 2 Mount Brown limestone Shelly and coralline limestone 3 Hutchinson Quarry beds (fan-coral bed) Volcanic tufa and conglomerate with abundance of marine shells Cretaceo-Tertiary 4 Ototara limestone Close-grained calcareous sandstone 5 Lower tufa Intensely dark fine-grained and rubbly tufas 6 Chalk marls Chalk marls representing the Amuri limestone 7 Concretionary greensands Bright-coloured greensands with concretions 8 Boulder sands. Sands overlying the coal 9 Black grit and coal Black oyster bed, sandstone, grey sands, shale, coal, &c The substantial discrepancies between this table of formations and that of Hutton are—-(1) the latter did not recognize the existence of a volcanic-tuff bed below the lower limestone — that is, the Ototara limestone of McKay; (2) McKay correlated the chalk marls with the Amuri limestone,

whereas Hutton correlated what McKay calls the Ototara limestone with the Amuri limestone; (3) following on from this, McKay correlated the upper limestone of Hutton with the Mount Brown limestone, whereas the latter associated it with the Weka Pass stone; (4) they both put unconformities above the lower limestone, but on different evidence. The present author considers that there is strong evidence of physical conformity in the whole sequence, and thàt there are not two limestones but only one, which is divided apparently in some parts of the district into two parts by ash beds from contemporaneous volcanoes. 2. Sequence, Thickness, And Arrangement of Beds In Particular Areas. (a.) Lower Members of the Tertiary Sequence. The part of the district where the lower members of the sequence can be studied best lies on the east side of the basin. Here the Broken River and its large tributary on the south, the Porter River, cut across the strike Fig 2 —Section along Broken River Distance, about 4 miles Direction, W N.W. E S E 1, greywacke; 2a, sands, greensands; 2b, marls, &c.; 3, tuff beds; 4, limestone; 5, sands. sandy clays, shales, and conglomerates—Pareora beds of the beds, and good sections are to be observed on the high banks of the rivers. We get in the Porter River and in the Broken River below the junction the following sequence, in descending order, underneath the limestone (fig 2) — Volcanic tuffs and calcareous tuffs 70 ft. Marl 150 ft. White quartz sand 8 ft. Concretionary calcareous sand 2 ft. White quartz sand 20 ft. Concretionary band 6 in. White quartz sand 12 ft. Marl 4 ft. White quartz sand 3 ft. Marl 28 ft. White quartz sand 50 ft. Marl 100 ft. Concretionary greensand. 500 ft. Sands, &c., including the oyster and shell bed 800 ft. Sulphur sands, white quartz sands, sands with impure lignite, light greensands, sandy clay, and white sands, following in that order 300 ft. Greywacke. The beds strike N. 30° W., and dip S. 60° W. at angles approximately 15° above the junction of the two rivers, but which become slightly steeper on tracing the beds east Thoughout the series of beds enumerated above there is no sign of unconformity, as has been admitted by all who have

examined the section. A similar sequence also occurs in the Broken River between its junction with the Porter and the Broken River limestone gorge, but the clarity of the section is somewhat obscured by slips from the dominating limestone scarp above it: there is, however, no evidence of unconformity. Hector and McKay have correlated the marl with the Amuri limestone, whereas Hutton placed it on a lower horizon, since he correlated the lower Trelissick limestone with the Amuri limestone. There is strong evidence, which will be referred to later, that Hutton's contention is incorrect. The stratigraphical position of the marl, the fact that in places it is lithologically indistinguishable from typical Amuri stone, and its close resemblance in parts to a rock formed from Globigerina ooze, owing to the numerous Foraminifera that it contains,* P. Marshall, The Younger Limestones of New Zealand, Trans N Z. Inst, vol 48, 1916, pp 94, 95 render it extremely probable that Hector and McKay were right in placing it as they did The characters of the deposit show that it was laid down in fairly deep sea, but the interstratification with sands in its lower portions shows clearly that some land was adjacent. It is the higher parts, however, which have a true deep-sea facies The palaeontological evidence furnished by other localities in the neighbourhood indicates that the marl and the beds immediately associated with it are of Lower Tertiary age, which is not inconsistent with its correlation with the Amuri limestone. The beds exposed near the junction of Broken River and the Porter are continued round the east and south of Prebble Hill under the tussock-clad slopes, and there is no clear-cut section till the neighbourhood of the upper limestone gorge of the Porter is reached. In swinging round the hill the beds are bent up into a bowl-shaped form, which is tilted towards the west; on the edge of the southern rim a magnificent section is exposed. Near the river the beds strike N. 80° E., with a northerly dip at an angle of 50°, the river cutting directly across them The section exposed at the river consists of the following, in descending order:— Limestone Volcanic tuff. Marl, whiter in the upper portion and greyer in the lower. White quartz sand, with two concretionary bands similar to those in the section lower down the river. Greensands. These last apparently form the solid base of the western terrace of the Porter River for some distance up above the gorge, but a fairly heavy covering of shingle renders a proper examination impossible. The beds just above the gorge are apparently involved in a small fault with a down-throw to the north; but the section is somewhat obscure, and the appearance may be really due to a surface slip, the greensands having moved forward. If a fault really occurs, the throw is small. Owing to the close covering of well-grassed soil, and talus from the overlying limestone, no good exposure of these beds can be obtained on the east side of Castle Hill; the lower bed of limestone, however, outcrops in a small slip north-west of Mr. Milliken's house Following the basin round, the next important section where the lower members of the formation are exposed is in Whitewater Creek, a tributary

of the Porter coming in from the west and junctioning with the main stream close to the road-crossing. This tributary rises in Mount Enys and Fig 3 —Section along Whitewater Creek, Distance. 2½ miles. Direction, W N W E S E 1, greywacke; 2a, sands, greensands; 2b, marls, &c; 3, tuff beds; 4, limestone; 5, sands, sandy clays, shales. and conglomerates—Pareora beds Mount Olympus, peaks of the Craigieburn Range, and flows in a south-easterly direction across the whole series. The section exposed below the limestone is as follows (see fig. 3):— Limestone Volcanic tuff, of uncertain thickness but probably over 1,000 ft. Calcareous greensand, more calcareous in its lower portions—in fact, it is in its lower portions an argillaceous limestone, flaky and hard, and showing some resemblance to the lower bed of limestone, indeed, it is extremely probable that it actually represents that bed in this section. It is 12 ft. thick, strikes north-east, and has a dip to the north-west of 45°. Volcanic tuff, containing numerous fragments of rock, and glauconitic in character owing to decomposition, and passing down into a true tuff, the total thickness being 60 ft. Marl, fine-grained and flaky at times, containing a proportion of sandy material, sometimes decidedly sandy. The dip is here steeper, reaching 65°, and the thickness of the bed 500 ft. Greensands and sands, the latter sometimes concretionary, with thin beds of coal in their lower levels, striking N. 80° E., and dipping in a northerly direction at an angle of 45°, the total thickness being 950 ft. Although there is thus a variation in dip between this and the overlying marl, there is no evidence of a break. These beds rest unconformably on greywacke exposed in the channel of the creek The series here enumerated is somewhat different from that recorded by Hutton, the discrepancy arising, no doubt, through the obscuring of parts of the section by covering gravels, different observers having an opportunity of examining only portions of the sequence. I can see, however, no sign of an unconformity, any variations of the dip and strike being what might be expected from the circumstances of the deformation of the beds by folding It is natural under those circumstances that both will slightly vary from place to place. In Coleridge Creek, the westerly extension of Porter River, the volcanic-tuff beds become more important. The following sequence occurs in the lower part of the creek:— Limestone Marl and ash beds interstratified, the former being both grey and white, and the latter in places calcareous and with fossils. White sands, interstratified with volcanic beds, the sands passing down into greensands.

The special interest in this section is that it shows clearly that volcanic action was going on while the sands and marls were being laid down. The beds strike N. 30° W., and dip S. 60° W. at an angle of 60°, the beds striking in the direction of those in Whitewater Creek. In the upper part of Coleridge Creek, lying on the greywacke, are the greensands; above these again the marl, running almost along the creek; and this is succeeded by the volcanic-tuff bed, which has here no great thickness. This is followed conformably by the lower limestone. These beds strike north-west, and dip to the north-east at high angles The pronounced effect of volcanic action on the continuity of the limestone beds is seen clearly where this limestone crosses the creek near the old sheepyards, for there it ends abruptly and the bed passes on as a calcareous tuff with the same stratification. The junction has all the appearance of a fault, but no sign of dislocation can be seen in the over lying well-stratified beds. On the western side of the basin these beds are not seen, though they probably underlie the isolated blocks of limestone which occur there, the surface being masked by later deposits of Tertiary age and river-gravels and talus from the mountains lying immediately west. In the upper part of Broken River basin, especially in Waterfall Creek, in the main stream, and in Blackball or Murderer's Creek (a tributary coming in from the north), exposures of the beds can be distinctly seen, those in Broken River itself being most instructive. About a mile above the road-crossing the river runs through a narrow gap of limestone dipping almost vertically, the upper portion being forced over the lower, this being a subsidiary fault to a main one which runs north-east, the beds on the south-east being thrown down, the fault hades with the dip, and in consequence there is a suppression of the outcrops on the surface, the marls, and perhaps the overlying tuff if it really exists, being cut out. The beds in contact with the limestone are white sand with occasional white calcareous bands. These are succeeded by greensands, and then by a hard concretionary calcareous sandstone bed 20 ft. thick, and below this lie again greensands with large calcareous concretions containing fragments of saurian bones. The dip of these beds is downstream—i e, to the south-east, but the direction immediately changes to one up-stream, and the beds are repeated in reverse order. They have thus been folded up into an anticline with the eastern limb almost vertical or even slightly overturned, and the limestones which originally formed the roof of the anticline have been removed, but the structure is indicated by remnants to the immediate south-west of the locality. The lateral movement of the limestone at the gorge and the slickensided surfaces of the limestone are secondary phenomena resulting from the structural movements which have produced the folding These beds dip up-stream above the anticlinal axis till just past the junction with Waterfall Creek, when they take the form of a syncline, and the dip of the western wing of this continues for some distance up-stream—in fact, to the actual base of the series, where clay and occasional beds of lignite and brown coal occur. The banks of Broken River are here very high, but are thickly wooded and in places subject to slip, so that the actual sequence cannot be made out. Up Waterfall Creek, however, there are excellent sections of the beds between the greensands and the limestone. Hutton puts in a fault as necessary to explain the structure here, but this seems to be incorrect. The sequence is perfectly normal and without signs of dislocation.

The following beds occur in descending order (see fig. 2):— Limestone. Marl, of uncertain thickness. White sand, with occasional layers of greensand 3 ft. to 4 ft. thick, the total thickness of the whole beds being uncertain owing to the cover of bush and gravel, but certainly between 150 ft. and 200 ft. The beds strike north-west and dip south-west 35°. Greensands, dark in colour, over 100 ft. thick; these pass into the greensands seen at the mouth of Waterfall Creek. A noteworthy feature of this section is the absence of any volcanic tuff below the limestone In the upper portions of Waterfall Creek the limestones are most in evidence, and are subject to a series of anticlinal and synclinal folds. Where the anticlines have been denuded greensands are exposed; this especially is the case on the eastern side of the Hog's Back, a high ridge of limestone near the westerly margin of the basin. In Waterfall Creek and immediately south of it the beds are exposed on a well-marked scarp slope facing east, and the following section is visible:— Limestone. Limestone with volcanic-tuff pebbles, 6 ft. thick, passing into Tuffs and greensand, 8 ft. thick. Greensands. Light-coloured sands. These beds exhibit a conformable succession, and dip west at an angle of 55°. Immediately west of the section just referred to, above the spot where the river breaks through the limestone, there is apparently a reversal of dip, and coal-beds are exposed under the limestone; but the country is here bush-covered and the relations are obscure. In all probability the limestone and the underlying beds have been slightly upturned, perhaps owing to a fault movement which has produced a line of dislocation passing along the base of the Craigieburn Range; but no clear section can be obtained, owing to the covering of bush, soil, and loose debris. In that part of the area lying to the north of Broken River the sections are not clear, but exposures in various places show that the beds are existent in their normal stratigraphical relations, with the exception of one locality—viz., that in the immediate vicinity of Parapet Rock. The fault which affects the limestone in Broken River continues to the north-east just past this point, when it passes into a fold without any apparent disruption of the beds. At Parapet Rock the limestone is slightly overturned, but it rapidly flattens out and takes a basin-like form as it follows round Flock Hill. At Parapet Rock and where it crosses Murderer's Creek the limestone is underlain by calcareous tuff, but there is no exposure of the marls and white sands, and their outcrops may be suppressed, as they are in Broken River, as a result of a fault following the strike; but these beds appear in their normal position when traced round the eastern side of Flock Hill, and finally junction with those in the Broken River below the limestone gorge. To the north and east of Flock Hill the greensand beds and associated coal-measures are well developed, especially on the saddle between Murderer's Creek and the creek which runs east of Flock Hill. At the point where these beds cross Murderer's Creek Haast obtained numerous leaf-imprints, which were subsequently identified by von Ettingshausen,

and the same beds cover a fair tract of country to the eastward, and finally junction with the coal-measures which were described previously as occurring in Broken River. This completes the reference to occurrences round the whole circuit of the basin. In this account of the various localities where clear sections of the lower members of the sequence of Tertiaries are given it will be noticed that there is no indication of an erosion surface, and that the conformity is complete. This has been admitted by both Hutton and McKay. This is an important point, as will be further emphasized when the fossil content of the tuff is considered. The presence of the volcanic-ash beds at different horizons has been a matter of considerable difficulty, but in my opinion the following is a fair statement of the conditions:— 1. Volcanic action commenced at the time when the white sandy beds were being laid down—that is, between the deposition of the greensands and the marl. 2. Volcanic activity was more pronounced in the neighbourhood of Coleridge Creek, the ash beds being there much thicker than elsewhere in the area. The earliest signs of volcanic action are also furnished by that locality. 3. Elsewhere in the district the beds are thinner—in fact, they do not appear at all in the sections in Waterfall Creek, where they should be easily seen were they present. This remark applies to higher occurrences of the ash interstratified in the limestone. 4. This interstratification is not indicative of an unconformity, but that the deposit of ash went on contemporaneously with the formation of limestone and other marine beds in the adjacent sea. (b.) Occurrences of Limestone with Interstratified Tuff The limestones, which are in places separated by the uppermost layer of the volcanic-tuff beds, were considered by both McKay and Hutton as quite distinct, and belonging to different geological periods. The present author, however, regards them as belonging to the same period, with a difference in character which any limestone might exhibit as the conditions of deposit slightly changed during its deposition, the tuff bed interstratified in the limestone having just the same geological importance as similar beds interstratified in the sands and marls underneath the limestone. Before considering this question in more detail it will be best to give some account of the occurrences that the locality affords The most typical section, and one which is free from disturbing elements, is that obtained in the lower part of Home Creek, just above its junction with the Porter River. The creek cuts across the strike directly, and on the high walls of its gorge-like bed the relations can be easily seen. (Plate XXII, fig. 1.) The following is a description of this section:— Limestone, 60 ft. thick, the upper 10 ft. or 12 ft. of white stone, the lower 50 ft yellowish and weathering greenish-grey where run over by the stream; the rock is almost entirely composed of coralline fragments Calcareous volcanic tuff: This grades down from the overlying bed, and the proportion of volcanic matter increases in the lower part; the approximate thickness is 10 ft. Volcanic tuff, the upper part consisting of volcanic matter occasionally weathering green, and passing down into volcanic tuff weathering dark green, and finally into volcanic ash, the total thickness being about 75 ft.

Limestone, the “lower limestone” of Hutton and McKay, a somewhat flaky argillaceous stone, the lower part greenish, with small fragments of volcanic ash; thickness about 80 ft. All these beds strike N. 15° W., and dip westerly at an angle of 15°. When followed north towards Broken River the ash bed thins out and disappears, the only indication of the presence of volcanic matter being small particles scattered in places through the stone. The relations between the upper and lower layers of the limestone are obscured between the two rivers, but in the gorge of Broken River a clear-cut section shows no decided line of demarcation between them, and they apparently grade into one another insensibly. The spot, however, is somewhat difficult of access except at times when the river is quite low. From Home Creek the beds continue south across the Porter, but an excellent section of the upper layer of limestone and the subjacent tuff beds is seen just at the mouth of the Thomas River, the junction of the two being quite conformable. Followed south, the limestones form the highest point of the steep escarpments of Prebble Hill, facing east and south, and the strike swings round as described in the section dealing with the lower beds, till the upper limestone gorge of the Porter is reached. (Plate XXI, fig 2) Here occurs the following sequence:— Limestone. full of coralline fragments like the upper layer in Home Creek, 7 ft. thick This bed can be traced round on the slopes of Prebble Hill, so that it is continuous with the upper layer at the mouth of the Thomas, but it has thinned out very much. Volcanic tuff: Just as in Home Creek, the overlying limestone grades down into this bed, but its lower portions are a true volcanic tuff; the total thickness is about 60 ft. Limestone: The rock is flaky in its lower portion, but in the upper parts it loses this character and takes on the nature of the so-called upper limestone. Hutton says (loc. cit., p. 397), “On the south side of Prebble Hill the limestone is divided into two parts, the lower of which is composed of comminuted fragments of Bryozoa, Hydrocorallinae, &c., forming what is called a coralline limestone, thus differing altogether from its normal character, and resembling the upper limestone presently to be described.” This statement is quoted as showing that lithological character will not distinguish the two limestones, and that they are probably one member of the series, and that they grade into one another. At this point the beds strike almost due east, with a northerly dip of 50°. From this locality the beds apparently run west, but they are hid by the gravels of the plain till the north-east corner of Castle Hill is reached. Hutton puts in a fault at this place with a downthrow to the north, whereas McKay explains the arrangement by the presence of a fold. The mass of Castle Hill which dominates the interior of the basin, and whose picturesquely eroded rocks are responsible for the popular name for the locality (Plate XXI, fig. 1), presents a well-defined escarpment to the east, but unfortunately it furnishes no clear-cut sections showing relations of the two limestones (Plate XXI, fig. 2). The exposure along the eastern face of the hill shows that the beds have been folded into an anticline and syncline on going from north to south, the anticline having been eroded at the north-eastern corner of the hill (fig. 4). On a line about 300 yards to the west of this and parallel with it the hill has a dome structure, and the syncline to the south apparently flattens out. In a

section from east to west passing through the highest point of the hill it again exhibits an anticlinal structure, flanked on the east by a syncline (fig. 5), which probably passes into a fault in a north-and-south direction, so that the extreme north-eastern end of the hill is separated somewhat Fig 4 —Section along eastern face of Castle Hill from Whitewater Creek to Thomas River Distance, about 2 miles Direction, S -N 1, greywacke; 2a, sands, greensands; 2b, marls, &c; 3, tuff beds, 4, limestone; 5, sands, sandy clays, shales, and conglomerates—Pareora beds from the rocks on the northern flank. This faulting is apparently attended with a thinning-out of the beds, perhaps as a result of folding movements which have resulted in either a pronounced distortion of the limestones as suggested by McKay, or as an actual fault running almost parallel to the line of the Thomas River towards the upper limestone gorge of the Porter as suggested by Hutton. It is probable that both suggestions may be partly correct, the undoubted fold grading into a fault Fig 5 —Section across Castle Hill. Distance, about 1 mile. Direction, W N W -E S E 1, greywacke; 2a, sands, greensands; 2b, marls, &c, 3, tuff beds, 4, limestone; 5, sands, sandy clays, shales, and conglomerates—Pareora beds In a small slip about half-way up the north-eastern face there is an exposure of typical lower limestone exactly similar in lithological features to that in the Porter River, and this is succeeded by tuff beds (25 ft.), which pass up into a fossiliferous calcareous tuff with brachiopods, corals, and fragments of molluscs, and overlying this is the upper bed of limestone 150 ft. thick. This is typically developed in the splendid monolithic blocks which crown the hill, and in the fine eastern face. The stone was quarried at one time for building purposes, and was in good repute throughout the whole of Canterbury; in fact, it was looked on as the best building-stone that the province could produce. At the quarry itself it has a free, even-grained texture, nearly white, and admirably suited for general building purposes, its place being now taken by Oamaru and Mount Somers stone, which are inferior in quality; but the remoteness of the locality and the difficulty of transport will always militate against its general use On tracing the limestone south to the vicinity of the Whitewater Creek it thins out considerably, so that the total thickness diminishes to about 40 ft. In the bed of the stream there occurs a mass of limestone, but it is

difficult to say whether or not this is a slip from solid outcrop in the vicinity. It resembles in character the lower layers of the limestone in position above it, but this is not conclusive proof that it is a slip. In Whitewater Creek the sequence is as follows, in descending order (fig. 3):— Limestone, a hard flaky rock, breaking in the upper part into quadrangular blocks; thickness, 40 ft. Calcareous tuff, composed of limestone and fragments of volcanic rock, exactly similar to that in the analogous position in Home Creek and Broken River: from 2 ft. to 3 ft. thick; passing down into Volcanic tuff, 100 ft. thick. Limestone, thickness about 25 ft: this may be a part of the upper portion of the limestone beds. These beds strike N. 60° E., and dip in a westerly direction at an angle of 20° In Volcanic Creek the beds have the same dip and strike, and are as follows.— Limestone, the lower portion having a definite jointing into quadrangular blocks; the upper portion like the ordinary upper limestone in general appearance, but corals were absent in the specimens examined. There was a distinct separation into two facies, but no unconformable junction, and no tuff beds between them. Total thickness, 80 ft. Calcareous volcanic tuff: This is well bedded and clearly exposed under the limestone, and similar in lithological character to those beds in the same position elsewhere. Volcanic tuff, of uncertain thickness. There is no appearance of a lower band of limestone on the banks of this creek, but it may be covered up by surface accumulations; some indication of its presence should, however, be apparent. This absence increases the probability that the lower band of limestone in Whitewater Creek is only slip from the one bed which is in position, and that there are not two limestones. Traced towards Coleridge Creek the limestones appear to thin out and lose their distinctive character, especially in the lower part of the creek, where the volcanic tuffs are so well developed. The band of limestone dips to the west, but it is bent round in the form of a syncline and, passing close to the sheepyards, meets the creek again higher up and crosses it to the northern side. The succession here is as follows:— Limestone. Volcanic calcareous tuff, full of fossils. Volcanic tuff. Limestone, 100 ft to 150 ft. thick These beds are bent round on leaving the creek lower down, but at the upper crossing they have a north-easterly strike and a dip to the northwest at very high angles. On the western side of the basin there are numerous isolated blocks of limestone, detached from the main mass probably by a fault running north near the base of the Craigieburn Range, which is probably distinct from the main fault movement which bounded the area to the west. These blocks

occur in the small creek immediately below the limestone gorge of Coleridge Creek, at a height of 400 ft above its floor (fig. 6). They occur again higher up the valley-side in the next tributary on the north side of Coleridge Creek, a massive detached block occurs in the upper part of Whitewater Creek, and isolated fragments occur on the slopes of Leith Hill, on the western side of the basin, between the Whitewater and the Thomas Rivers. I have been informed that similar blocks occur in the bush immediately south of the Thomas River, and in the Thomas River itself, though I could not locate them. The occurrences near Coleridge Creek are especially Fig. 6.—Section across Coleridge Creek, just below limestone gorge. Distance, ½ mile Direction, N W -S E 1, greywacke; 2a, sands, greensands, 2b, marls, &c, 3, tuff beds; 4, limestone; 5, sands, sandy clays, shales, and conglomerates—Pareora beds. interesting as they are pinched in between the greywacke and the higher sandy beds of the Tertiary series, and the greywacke has apparently been forced over them along a fault-line. The blocks in Whitewater Creek consist of two beds of limestone, the upper 50 ft. thick; and underlying this white calcareous sands 40 ft.; greensands 10 ft.; and white sands of uncertain thickness. The lower limestone then follows, 40 ft. thick or more (fig. 4). The beds strike N. 60° E., and dip N. 30° W. at an angle of 35°. There is no appearance of overlying beds. The limestones are no doubt a part of the ordinary limestone series, and have been separated from the beds lower down the creek by a fault with a north-and-south direction. Only the occurrences in the northern portions of the basin are now left to be considered, and these present most interesting structural features. At the lower limestone gorge of the Broken River the following beds occur, all with a N.N.W.-S.S.E. strike and a W.S.W. dip of approximately 20°:— Upper limestone, 150 ft. thick, of the ordinary type but carrying fragments of volcanic tufa throughout, and at times segregated into well-defined layers. Calcareous tuff, 40 ft. to 50 ft. thick, well bedded, and containing a high proportion of calcareous material. Lower limestone, about 200 ft thick, flaky, and well jointed into quadrangular blocks. Both of these limestones are coralline in origin. In the neighbourhood of the gorge on the north side of the river there are several small faults with a north-and-south direction and a downthrow to the east. When followed round the eastern face of Flock Hill the strike gradually changes, and the dip becomes southerly at the northern end of the ridge, and farther on, when Murderer's Creek is reached, it is to the

south-east. The beds at this point have been subjected to a notable twist, as has been mentioned previously, and the angle of dip rapidly increases from 40° on the eastern side of Murderer's Creek to vertical with a slight overturning at Parapet Rock, and continues so till the bed again crosses Broken River at the upper limestone gorge. From just east of Murderer's Creek to beyond Broken River the limestones are involved in the fault, and some of the underlying members have had their outcrops suppressed (fig 1) Throughout this portion of the line of outcrop the rock is of a hard flaky nature, extensively jointed, a result due no doubt to the pressures and dislocations to which it has been subjected. In places it is crystalline, but traces of its coralline origin are visible throughout. It contains particles of volcanic tuff, and at Parapet Rock itself there appears to be a distinct layer of tuff, 4 ft. thick, dividing the stone into two layers, the lower 60 ft. thick and the upper about 80 ft., the latter being less flaky, but breaking into quadrangular blocks. It would therefore appear that there is the division into two layers even in this part of the area, remote from the centres of pronounced volcanic activity. In the area between Waterfall and Hog's Back Creeks the line of outcrops forms a series of loops, resulting from the erosion of a succession of anticlines and synclines. Continuing from Broken River crossing, it follows the same direction as it has north of the river till it reaches Trout Creek, when it swings round with a change of dip to the west, and forms the mass of Sugarloaf Hill, whose bold escarpment fronts to the north-east. This direction is continued across Waterfall Creek, the limestone itself forming the lip of the fall from which the creek takes its name. After crossing this creek the outcrop extends some distance up the steep slopes on the north and then swings round, forming a syncline, the edges of the bed cropping out on a series of low mounds in the direction of the Hog's Back Creek. When this is reached another reversal occurs and the outcrop follows back to the north, reaching Waterfall Creek once more and just crossing it. At the southern end of this stretch the limestone is almost vertical, and the upper edge of the beds forms a bold cliff along the base of which the Hog's Back Creek flows—in fact, the limestone is probably slightly overturned; but to the north it slowly flattens out, preserving a westerly dip, till in Waterfall Creek it is inclined at an angle of only 55°. Immediately north of this creek the line of outcrop turns back with a reversal of dip, crosses the creek again, and the bed apparently peters out, or is covered by surface accumulations and cannot be traced more than a few chains south of the creek A small occurrence of limestone is found on the south side of the Hog's Back Creek, in the vicinity of the greywacke slopes of the Craigieburn Range, on the spur dividing that creek from the Thomas River; but the exposure is so small that it is impossible to determine its relations. It may be a small block separated from the main limestone mass by faulting, or be a part of a fold connected with the Hog's Back Ridge under the creek, and overlain by more recent Tertiary beds. This limestone is generally hard, and somewhat jointed into flaky quadrangular blocks in its lower portions, but it varies much in character in its various parts. In many places the upper layers are distinctly crystalline. Marshall gives the following microscopical description of this rock from a specimen collected by the author: “Polyzoa are again the most frequent organisms in this rock, though echinoderm plates are common. Foraminifera

are less common, and only Cristellaria could be distinguished. There is some Lithothamnium.”*P Marshall, The Younger Limestones of New Zealand, Trans. N Z. Inst., vol 48, 1916, p 92. Hutton would discriminate between the two beds by insisting on the coralline nature of the upper as distinct from the non-coralline nature of the lower, but a careful examination shows that the coralline element may occur in both; in fact, Hutton himself has noted the coralline nature of the lower limestone at the upper Porter limestone gorge. Even where, through pressure as a result of intense folding or faulting, the rock has taken on a crystalline facies traces of its coralline origin are apparent. This can be well seen at Parapet Rock and in Coleridge Creek. The lower portions apparently contain a considerable mixture of volcanic-tuff fragments, and are more distinctly bedded in places; especially is this the case in the limestones exposed in the gorge of the Porter. In the upper portions the planes of bedding are less pronounced and the coralline fragments are more distinct, so that a decidedly different appearance is given to the rock. This led Hutton and McKay to class them as two distinct stones, and the absence of the decidedly coralline facies from certain localities led them to attribute its non-existence to erosion. They therefore placed an unconformity between the higher limestone bed and the overlying calcareous sands and shell beds, and the divergence in character between the upper and lower beds led them to place an unconformity between them, thinking also that the volcanic tuff marked an erosion surface. In my opinion they are members of one limestone, for the following reasons: (1.) There is no stratigraphical break, wherever they occur the beds are conformable, and when there is no tuff it is impossible to recognize a distinct break, but one bed passes into another insensibly. (2) The characters of the two rocks are not invariable. As Hutton pointed out, the lower bed is in places distinctly coralline, and the upper bed, even when there can be no possible doubt as to its identity, may show an entire absence of this feature. (3.) The palaeontological evidence will show that there is no great divergence in fossil-content between the tuff bed under the lower limestone, classified as Cretaceous by Hutton, and the tuff above it, which he assigned to the Oligocene period, the conformity of the upper limestone to the upper tuff being generally admitted. For these reasons I consider that there is only one limestone, that it is thinnest where volcanic action was most marked—that is, in Coleridge Creek, that in the deep water removed from the volcano sedimentation went on continuously, with a gradual change in the nature and condition of the organisms contributing to the formation of the rock, but that in closer proximity to the volcano interstratification of the limestones and the tuffs took place. Whereas in the lower horizons the limestone had more or less the character of a deep-sea deposit, as time progressed it was laid down in shallower water in preparation for the next suite of calcareous beds, which have the nature of shore material. (c) Pareora Beds. The next succeeding suite of beds, called “Pareora” by both McKay and Hutton, is considered by them to rest unconformably on the underlying series, a conclusion based chiefly on the supposition that the upper bed of limestone has been removed by erosion from certain parts of the area. If, however, there is only one limestone, then this argument falls

to the ground; but before dealing with it more fully it will be best to describe the relations between the various members of the overlying series as it is developed in different parts of the area. I have retained the name “Pareora,” since it is convenient to describe a particular facies of Tertiary deposits occurring in numerous places in Canterbury besides the name locality—that is, shore deposits of sandy layers with broken shells, occasionally with concretionary bands, which form the top of the Tertiary fossiliferous beds; but I do not intend to imply that they should be classified as distinct from the Oamaru series as developed in its typical district. The relations of these beds to the underlying strata are best seen in the neighbourhood of the junctions of the Thomas River and Home Creek with the Porter River. In the former, from the site of the old dipping-yards down-stream, occur the following:— Sandy shales; thickness exposed, 10 ft. Lignite, about 2 ft. thick, dipping S.W. 30°. Greenish-grey sands Yellowish sands, dipping S.W. 20°. Dark greensands. Yellowish sands. Dark greensands, extending past the junction of the creek on the south. Greenish-grey sands. Broken-shell beds, with sands, 12 ft. thick. Greenish-grey sands. Broken-shell beds, 2 ft. Concretionary sands, with broken-shell bands. Broken-shell beds, 3 ft. Sands, weathering rusty grey, sometimes a light-coloured greenish-grey, 100 ft. thick. Calcareous sandstone with many shells; thickness 15 ft.(?), strike N.N W, with a dip to the W S.W. 20°. Limestone. The relations between the lower bed and the limestone are well seen at the present time owing to a great fall of rock just above the junction of the Thomas with the Porter. (Plate XXII, fig 2.) Immense blocks lie here with thick layers of shells in excellent state of preservation. These were not available in the time of Hutton or McKay, and the overhang which the rocks no doubt had at that time would prevent their proper examination, the dangerous nature of the locality being remarked by Hutton. In none of the contacts, however, is there any sign of an erosion surface. In the neighbouring Home Creek the following sequence is exposed:— Calcareous concretionary sands with broken-shell bands, the shell-remains being concentrated into a narrow layer 6 in. to 12 in. thick; total thickness uncertain. Sandy beds, greenish in colour, with broken-shell layers, 15 ft. Calcareous concretionary sands, 2 ft. Greenish-grey sands, 3 ft. Sandy beds with broken shells, very much current-bedded, 5 ft. exposed. Broken-shell beds with sands, sometimes with hard bands, sometimes looser, 20 ft. thick. Sands weathering a rusty brown, with brown concretionary layers and shell-fragments, 80 ft. Shelly beds with loose irregular light-greenish-grey sands, 25 ft. thick.

There is no apparent unconformity here with the underlying limestone. Between the two limestone gorges of the Porter these beds are bent up into a syncline in sympathy with the underlying limestones. Starting from the fall of rocks and continuing up-stream, we have the following sequence, the first part in ascending order, and when the other limb of the syncline is reached with the beds in descending order:— Limestone Coarse-shell bed, about 2 ft. thick. Finer-shell and sandy bed, about 5 ft. Sands, rusty brown, inclined at times to grey, 100 ft. thick, striking N.N.W Sands, with bands of calcareous concretionary sands and shells, 40 ft thick Sands, 35 ft Struthiolaria bed, 4 ft thick, strike N.N W., dipping S.S.W 20°. Other shells are included besides the Struthiolaria, but this genus is dominant Yellow ochreous sand, 80 ft. Sandy shale, 12 ft Lignite, 2 ft Sandy shale, 120 ft thick, bent into a syncline. White sand, 2 ft. Sandy shale, 6 ft., white sands, 5 ft. Sandy shale, 150 ft Ochreous brown sand, 20 ft Greenish sand, 15 ft White sand, greenish where it weathers, but stained ochreous brown in places, with hard concretionary sandstone band in the middle; total thickness, 25 ft. Calcareous sandstone and concretionary bands with shells (Struthiolaria beds?), 1 ft, striking N E, dipping N W 45°. Greenish-grey sands, 30 ft (?) Broken-shell beds. Sands, thickness 130 ft. (?) There is some doubt about these beds, owing to the covering of soil. Limestone, 7 ft thick; strike, E by N.; dip, N by W 55° In all this section there is no evidence of unconformity The next important section is that occurring in the upper part of Whitewater Creek above the limestone. In a small tributary to the north an excellent show of sections is displayed. Working down this, the following beds are exposed — Sandy shales. Sandy beds Shell beds containing Struthiolaria. Sandy beds with layers of shells in their lower portions. Sandy beds with, concretionary layers containing shell-remains, largely Macrocallista, the same as in the Thomas and Broken Rivers at this horizon Sandy shell beds Limestone. These beds have a strike to the north, and dip west at an angle of 20° In Moth Creek, a tributary of the Thomas coming in on the west side of Castle Hill, there occur sandy shales, blue greensands, and shales with

coal. These strike along the creek to the north-east, and dip north-west at high angles—over 60°. In the Thomas River, below the junction with Moth Creek, occur sandy clays and shales with bluish greensands striking almost east and dipping north 65°, practically along the line of the Thomas at that part of its course, the strike of the beds swinging round in conformity with the direction of the limestones on the northern part of Castle Hill. Below the road-crossing the direction is also easterly, with northerly dip; but as the stream is followed down it gradually changes till the exposures near the mouth of the river are reached. Owing to the cover of slip accumulations from the high unstable terraces and river-banks, and the covering of grasses and soil, the exact sequence cannot be determined. In the upper part of the Thomas River, close to the roadman's cottage, the higher members of this series are developed. They consist of shales, sandy clays, greensands, sulphur sands with occasional beds of impure coal in the lower part but with occasional interstratified layers of conglomerate in the higher portions, and thin bands of lignite. They dip up-stream—that is, to the west—at angles varying from 20° to 30°; some of the variation may, however, be due to slip, and their total thickness must exceed 500 ft. It is important to note the occurrence of the pebble beds at this stage, as they are found not only in other parts of the Castle Hill Basin but widely distributed throughout North Canterbury towards the close of the Tertiary series, and in perfect conformity with the underlying fossiliferous beds. They are specially well seen at the mouth of the Waipara, in the Mount Grey and Mairaki Downs, and up the Okuku River, and indicate the presence of land of considerable extent in close proximity to this region at the close of the Tertiary era. In the Hog's Back Creek itself the higher members of this series of beds are exposed. At the mouth green sandy shales occur, and above them white sand; but the country is much slipped, and no good sections are exposed for some distance up the creek, where the directions of the beds are involved with the folds which have affected the limestones to the north in the vicinity of Waterfall Creek. However, they seem to conform to their direction. In a small tributary on the south side of the Hog's Back Creek the beds strike north, and dip to the west at an angle of 35°; but at the mouth of this small creek they apparently have a strike which runs west of north: this may, however, be the result of slip. Half-way up the main creek there is exposed an impure lignitic sandy shale with a N. 70° E. strike and a dip of 30° to the south; this overlies a blue sand weathering brown. Farther up still, the beds strike east by north and dip south at 60° approximately, the beds consisting of white sand and sandy clays, the creek following along the strike, These beds are succeeded again by conglomerates with pebbles of greywacke, representing the highest beds exposed in the series. At the point where the creek flows past the end of the Hog's Back it is impossible with the limited exposures to make out the relations, but in all probability the beds are pinched in, taking the form of a completely closed syncline, with the small outcrop of limestone noted previously as the western limb of the fold. On the other hand, they may be faulted down. On the west of the Hog's Back Ridge the fragmentary shell beds occur resting quite conformably against the limestone where it dips at high angles, but there is no exposure of the higher beds. In Waterfall Creek, too, there is a small exposure of the greenish sands in the syncline which terminates the limestone ridge.

In the northern part of the area the beds are best exposed in the main stream of Broken River above the gorge, and in its tributary, Murderer's Creek. In Broken River itself the beds are involved in a syncline, its axis being about half-way between the mouth of Trout Creek and the upper limestone gorge of the river (fig. 2). From this point the sequence of beds following down-stream consists of green sandy beds with shell fragments, greenish sands opposite the mouth of Trout Stream, sandy clays at the mouth of the Hog's Back Creek, succeeded by light-coloured sands and sandy clays, till the road crossing is reached, where a bed of lignite 2 ft thick crosses the road Lower still, greenish-grey sands with broken-shell beds in its lower parts, then 50 ft of sand and irregular broken-shell beds, followed by 140 ft. of sands, about 80 ft of sands and shell beds, and 25 ft. of shelly bands more or less concretionary, the last-named lying over the limestone without any sign of physical break. The beds here strike N N W -S S.E and have a westerly dip of about 20°, but farther up the stream the dip is flatter and the strike more east and west. Between the axis of the syncline and the upper limestone gorge the following sequence occurs greenish sands, sandy shales with lignite, greenish sands; the angle of dip rising from about 15° to nearly vertical as the fault is approached and the beds abut against the faulted limestone A similar sequence of beds is seen in Murderer's Creek to the east, between the Natural Tunnel and Parapet Rock, but the creek runs for most of its length on or near the line of strike, so that few good sections are seen. The coal bed which crosses Broken River and appears in the Thomas, and between the two gorges of the Porter, is present, and forms a part of the bank of the stream for some distance. At the two ends of this section the beds are bent round so that they cross the stream nearly at right angles, and at the upper crossing numbers of fragments of Struthiolaria tuberculata were found, showing that this particular bed persists to the north end of the basin Beds of this horizon also occur in Coleridge Creek near its limestone gorge. In a small tributary creek just below the gorge the following sequence in descending order is found (fig 5).— Limestone, entirely out of place, as it resembles in lithological character the lower bed of stone in Coleridge Creek—that is, it is a fine-grained somewhat crystalline polyzoal stone. It is apparently overlain by greywacke, and owes its position to being pushed up from below along a fault-plane, so that it now overlies the beds which are really higher in the sequence, and is itself apparently overlain by Trias-Jura rocks Greensands and dark sandy shales interstratified, the beds being repeated, and followed by sands, which pass down into sands with concretionary bands and layers of shells, mostly in a fragmentary condition. These are succeeded in descending order by brownish sands with shells, and bluish-green sands with shell layers, lying on the upper layer of limestone conformably. These beds are about 300 ft thick; they strike north-east and south-west, and dip north-west 60°. The lower members of the sequence are repeated in reverse order on going down Coleridge Creek, and the strike and dip gradually change till the strike is approximately north-and-south and the dip to the west at an angle of 30°. These beds lie conformably on limestone. Both McKay and Hutton have insisted on the existence of an unconformity above the lower limestone, but the evidence appears to me to be unsatisfactory. The former nowhere states definitely the reason why he

considers an unconformity necessary, but his sections, notably those on pages 64 and 65 of his report, leave no doubt as to his position. His representation of the block of limestone in the upper part of Whitewater Creek as lying across the denuded edges of the tuffs and greensand beds is decidedly open to question, and is not borne out by my observations. I could find no exposures which warrant the arrangement of beds as indicated, and the position of the limestone block is easily and more satisfactorily explained by the occurrence of a fault, of which there is undoubtedly other good evidence, bounding the western side of the basin. In his section, too, on page 61 the representation of the beds does not appear to be altogether satisfactory, however, he shows a perfectly conformable sequence near the gorge of the Porter; but his interpretation of the section up the Whitewater has no doubt influenced his ideas of the country to the west of Castle Hill. On page 406 of his article Hutton emphasizes the existence of the unconformity. His main line of evidence is that in the Coleridge Creek area the lower limestone does not now exist, but has been removed by denudation before the upper limestone was deposited. The absence of the lower limestone in this area can be explained satisfactorily on other grounds—viz., that it was never laid down, owing to the interference with deposition of calcareous beds by volcanic action. In the upper part of Coleridge Creek, more removed from the centre of activity, both beds of limestone, with their intervening tuffs, are quite normally developed and are conformable throughout, the whole sequence being represented. If, therefore, my interpretation of the structure is satisfactory, there is no reason for the unconformity. This opinion is further strengthened by the fact that in those places where the ash beds are not interstratified the lower limestone passes up without a break into the upper stone, and explains why it is that the lower stone does undoubtedly in general have the characteristic coralline character of the upper. They are in fact normally the same bed, and the ash beds do not indicate the presence of a break. The unconformity which is placed higher in the sequence—that is, over the upper limestone—has also, in my opinion, no foundation palaeontologically or stratigraphically. As far as the last is concerned, the main evidence of Hutton rests upon the disagreement in the strike of the beds in the tributary of Whitewater Creek. After a careful examination of the locality I can see no evidence of this; any change in dip or strike is perfectly gradual, and entirely explicable on the grounds that the beds of the area have been subjected to gradual deformation. The fact that the Pareora beds rest in the northern portion of the basin on the Waipara series is not entirely correct, even assuming that the lower limestone is of Waipara age. In the lower part of Broken River, below the road and near the gorge, the sequence is perfectly conformable and normal; the same is entirely true near the limestone gorge in the Thomas and Porter Rivers. The only apparent discrepancy occurs near the fault which crosses the Broken River, and then it is perfectly possible that outcrops may be suppressed, as so frequently occurs in connection with strike faults. I do not think that this affects the case as far as the contact of the Pareora beds and the limestone is concerned, since to the west of the Hog's Back the Pareora beds lie perfectly conformably on the limestone, in that part of the area there being no tuff bed interstratified with the limestone, although it occurs underneath it. Showers of ashes had evidently not fallen in that area while the limestone was being laid down. It must be noted in this connection that Hutton admits fairly the absolute conformity of the beds exposed between the two-limestone gorges of the Porter.

D. Tectonic Features. The main tectonic features are those indicated in my paper on the intermontane basins (p. 341), the conditions of faulting and folding being those which have resulted from the settling of a block of country with a consequent readjustment of the beds to a somewhat smaller area. An arrangement of inward-dipping beds towards the centre of the basin is materially departed from, although a cursory examination of the locality might encourage this belief. The main faults (see map) which are now clearly visible and affect the Tertiary beds are,— (1.) A fault running north and south to the west of Castle Hill along the base of the Craigieburn Mountains, with a downthrow to the east, as a result of which isolated blocks of limestone are left stranded at the base of the range. It is difficult to arrive at a precise estimate of the amount of throw, but it is certainly some hundreds of feet. The situation of the stranded blocks is somewhat hard to account for, but they may be blocks which were separated from the main mass when the whole of the area was faulted down, their contacts with the adjacent greywackes not being visible, though the occurrence of limestone so close to the latter without any intervening shore-line beds does certainly suggest that this is not the position in which they were laid down. In the section figured in my paper just cited the contact of the limestones with the greywacke to the west of the Hog's Back represents a contact without indication of a fault. At this spot the junction is extremely obscure, and it may be a shore-line or a fault contact—probably the latter, as this would be on the line of the undoubted fault continued north from the west of Castle Hill. (2.) There is another pronounced fault, running north-east and southwest through Parapet Rock on the main road in the direction of Broken River. This is evidently closely connected with the folding of the beds, and grades at both ends into folds. The downthrow is to the south-east, and there is considerable lateral movement of the blocks of limestone, as is evident from the brecciated and slickensided surfaces (3.) Hutton considered that a fault ran from the northern end of Castle Hill eastwards toward the upper limestone gorge of the Porter. This is apparently of the same nature as the last, the folding at the two ends being undoubted; but the middle portion, where faulting should be visible if it really existed, is covered with river-gravels and is completely masked. (4.) A well-defined line of fault runs parallel to Coleridge Creek in a north-east and south-east direction, as is evident from the stranded blocks of limestone left high and dry on the north-west side of the creek in fact, the whole valley of this stream appears to be determined by an earth-fracture, the later sedimentaries being dropped down and to some extent pinched in between the walls of greywacke. The folding which the beds exhibit is in all probability due to the folding which attended this dislocation of the strata. This line of fault may be continuous with the fracture which no doubt determined the eastern boundary of the basin, as it will conform thereto with but slight alteration in direction. When this line is continued to the south-east it reaches Coleridge Pass, which forms a slight depression in the lofty ridge of the Craigieburn Range; but there is no positive evidence of its continuance into the basin of the Rakaia on the western side of the range, although its extension in this direction is extremely probable.

E Volcanic Rocks The volcanic rocks of the area have been described by Hutton. They consist almost entirely of fragmentary matter, which is very thick between Coleridge Creek and Whitewater Creek; in the latter this reaches 1,000 ft. approximately. The beds are also strongly developed near the mouth of the Thomas and in Home Creek. The fragments are usually of small size, pieces over 3 in. in diameter being rare, and are extremely well bedded, indicating submarine deposition, from their association with limestones, and from the marked amount of calcareous material which they frequently show, and from the numerous marine fossil remains which they contain. The fragments are of glassy character, at times with the yellow colour characteristic of palagonite; at others the glass is clouded with iron-ore and with microlites of feldspar, and somewhat frequent crystals of olivine. Deposits of the same character occur just outside the Trelissick Basin near the junction of the Esk River with the Waimakariri, and near the junction of Sloven's Creek with Broken River. In both these cases the beds are well stratified and associated with beds of calcareous material, pointing to a wide extension of the sea over the region in which the volcanoes were situated. Hutton has noted the occurrence of dykes round the base of Prebble Hill. They do not, however, appear to radiate from any particular point, and are perhaps only remotely connected with the volcanic outbursts. Similar dykes occur on the northern flanks of Mount Torlesse, somewhat distant from the centres of explosive action, the most noteworthy being that on the northern side of the bridge across Broken River, and in Iron Creek, on the western boundary of the Mount Torlesse coal-mining lease. Other intrusions have not been found in position, but numerous boulders and pebbles of olivine gabbro are found in the bed of the Porter River. In all probability they have been shed from minor intrusions such as are known to occur in the greywackes farther south in the Acheron Valley, from which the large masses found in the gravels of the Rakaia River may probably be traced. F Palaeontology. Although the correlation of the beds in different parts of the area can be determined with reasonable certainty from their stratigraphical relations alone, the palaeontological evidence on which their correlation with beds outside the particular area is based must be considered in detail. The fossil-content of the various beds will be dealt with in order, the authority for the collection of the specimens being given in each case. The oldest bed in the area which has yielded plant fossils up to the present is the plant-fossil bed in Murderer's Creek, the collection from which was submitted to von Ettingshausen,* C. Von Ettingshausen, Contributions to the Knowledge of the Fossil Flora of New Zealand, Trans. N.Z. Inst., vol. 23, 1891, p. 250. of Vienna. There is apparently some doubt about the localities, but these were revised by Hector. In the list given there the following species are recorded:— Quercus lonchitoides Ett. Planera australis Ett. Dryandra camptoniaefolia Ett. Cassia pseudophaseolites Ett.

All these fossil plants are spoken of by the author of the article as having a distinct Tertiary facies, and to be related closely to plants of undoubted Tertiary age from other parts of the world, notably Australia. The characteristic Tertiary character of this flora is very important when the nature of the fauna of the beds immediately overlying is considered. The presumption is that the beds containing it would be undoubtedly admitted as Tertiary were there not beds in another part of the area in a slightly higher position containing animal remains which point distinctly to a Cretaceous age for the containing beds. These occur at the base of the series in Broken River itself, near the eastern margin of the basin, in beds overlying the coal-measures. From this horizon McKay collected Conchothyra parasitica Hutt, as well as species of Perna, Cerithium, and other fossils; and the present author has also collected Inoceramus fragments and casts, as well as Perna, and Conchothyra included among a mass of fragments of black oyster In beds of approximately the same horizon above the upper limestone gorge of Broken River fragments of saunan bones were also found. But the lowest horizon from which a representative suite of fossils has been obtained is that of the tuff bed in Coleridge Creek, which is interstratified with the sands and marls above the greensands. From this bed Thomson and the author obtained a number of forms, a list of which has already been published by Thomson* J. A. Thomson, The Flint-beds associated with the Amuri Limestone of Marl-borough, Trans. N.Z. Inst., vol. 48, 1916, p. 51. I have included this list, with additional finds of my own, which will serve to strengthen the position taken by Thomson in his paper as to the age of the beds. The following is the complete list.— *Admete trailli (Hutt.). Ampullina miocenica Sut. Ampullina suturalis (Hutt). Ancilla papillata (Tate). Ancilla subgradata (Tate) *Callistoma aucklandicum Smith *Cantharidus sp. very similar C proruninus perobtusus Pils. Chione chiloensis truncata Sut Clio (Styliola) sp like Clio tatei Sut. but larger. Cominella intermedia Sut. Coptochetus sp. Crassatellites obesus (A. Ad.) Epitonium marginatum (Hutt.). Epitonium rugulosum lyratum (Zitt.). Epitonium zelebori Dkr. var. *Fulgoraria arabica elongata (Swams) *Fulgoraria gracilis (Swams). Fusinus bicarinatus Sut. Hemifusus gonoides Sut. Lapparia corrugata (Hutt.). Leucosyrinx alta (Harris) = Turris altus (Hams). Limopsis catenata Sut. Marginella harrisi Cossm Paphia curta (Hutt) Pecten sp? Polinices gibbosus (Hutt) Polinices huttoni Iher Polinices ovatus (Hutt.). Seila huttoni Sut. Siphonalia costata (Hutt) *Siphonalia nodosa (Mart.). Siphonalia turrita Sut. Siphonalia sp? *Siphonium planatum Sut. Surcula seminuda Sut. Terebra costata Hutt. Turritella concava Hutt Of these thirty-seven species, seven (including a doubtful species of Cantharidus)—viz, those marked with an asterisk—are Recent, the percentage

being 19—exactly the same figure as Thomson arrived at from a restricted list. This clearly indicated a Lower Tertiary age. It is noteworthy, also, that this bed, one of the lowest in the series, does certainly contain the lowest percentage of Recent forms yielded by beds which are admittedly of Tertiary age. The next higher fossiliferous horizon in the tuffs is that underneath the lower layer of limestone, the special localities being (1) above the upper limestone gorge of the Porter, and (2) the gorge of the Broken River below the road-crossing. In the former the following were collected by Thomson:— *Callistoma aucklandicum Smith. *Calyptraea maculata (Q. & G.). *Cantharidus tenebrosus A. Ad. *Emarginula wannonnensis Harris. *Natica australis (Hutt.). *Natica zelandica Q. & G. Pecten hutchinsoni Hutt. Pecten yahlensis T.-Woods. *Siphonium planatum Sut. Of these, all but two species are Recent. In addition to the above McKay collected from this locality in the greensands a species of Tellina and also Ostrea subdentata (Hutt.), although on the authority of Hutton this specimen really came from the marls. In any case, this Tertiary species comes from beds underlying the tuffs conformably, according to all authorities, and serves to emphasize further the Tertiary age of the beds in this locality at this particular horizon. In the tuff bed below the limestone in Broken River the following were collected, principally by Thomson — Ampullina suturalis (Hutt.). Ancilla papillata (Tate). *Callistoma aucklandicum Smith *Calyptraea maculata (Q & G.) Cardium spatiosum Hutt Clio tatei Sut Crassatellites cordiformis Sut *Crepidula crepidula (L.) Cypraea ovulatella Tate. *Diplodonta zelandica (Gray) *Emarginula wannonnensis Harris Flabellum laticostatum T.-Woods *Fulgorana gracilis (Swains) Lima huttoni Sut. Modiolaria elongata (Hutt.). *Natica zelandica Q. & G. Panopea orbita Hutt. Pecten williamsoni Zitt. Pholadidea concentrica Sut. Phos cingulatus (Hutt.). *Psammobia lineolata Gray. *Siphonalia mandarina (Duclos). *Siphonium planatum Sut. *Tellina eugonia Sut. Trochus nodosus Hutt. Of these, eleven are Recent species, five of the seven from the upper gorge of the Porter being identical. From this tuff bed McKay also records (loc. cit., p. 74) numerous specimens of brachiopods of the genera Terebratula, Waldheimia, Terebratella, and Rhynchonella McKay also records the occurrence of spines and plates of Cidaris, and specimens of Echinus enysi Hutt., as well as Turbinolia and various other corals The upper tuff bed at the junction of the Thomas River and the Porter is perhaps the most prolific locality in the whole district. The following is the list of fossils collected from there, the name of the collector being

indicated by initials: E. = Enys; M. = McKay; T. & S = Thomson and Speight, 1914; and S. = Speight, 1915. *Ancilla australis (Sow.) E *Ancilla bicolor (Gray). M. *Ancilla mucronata (Sow.). T & S Ancilla pseudaustralis (Tate) E. Callistoma acutangulum Sut. T & S Callistoma filiferum Sut E Callistoma oryctum Sut. T & S, E *Calyptraea maculata inflata (Hutt) E, T & S. S. *Calyptraea scutum Less T & S *Capulus australis (Lamk). M Cardium spatiosum Hutt S Cardium waitakiense Sut T & S Chione subroborata Tate E *Chione yatei (Gray) Corbula humerosa Hutt S *Corbula zelandica Q & G E, M, T. & S. Crepidula striata (Hutt) E Cucullaea ponderosa Hutt. S Cylichnella enysi (Hutt) M, E. Cypraea ovulatella Tate E, T & S Daphnella neozelanica Sut M Dentalium solidum Hutt. M *Diplodonta striata Hutt E *Diplodonta zelandica (Gray) T & S *Divaricella cumingi (Ad & Ang) M. & E *Dosinia caerulea (Reeve) Epitonium cylindrellum Sut E *Euthria striata (Hutt) E *Fulgoraria arabica elongata (Swains) E Glycymeris cordata Hutt S Glycymeris globosa (Hutt) M *Lima angulata Sow M, S *Lima bullata (Born) S Lima colorata Hutt. T & S *Lima lima (L) M, E. *Limopsis aurita (Brocchi). S Limopsis catenata Sut T & S, S Limopsis zittelli Iher. E. *Loripes concinna Hutt M Macrocallista assimilis (Hutt) S *Macrocallista multistriata (Sow) T. & S Mactra attenuata Hutt. S *Mactra elongata Q & G M, E Marginella dubia Hutt S. Modiolaria elongata (Hutt) T. & S., E., S. *Modiolus australis (Grav) T & S, M., S. Modiolus dolichus Sut. M Monilea praetextilus Sut M. Monilea sulcatina Sut M Mya n. sp S *Myodora subrostrata Smith M. *Mytilus canaliculus Mart S Olivella neozelanica (Hutt) S. Panopea orbita Hutt T & S Panopea worthingtoni Hutt M. *Panopea zelandica Q & G T & S. Paphia curta (Hutt) S *Paphia intermedia (Q & G) M. Pecten athleta Zitt S Pecten beethami Hutt. M Pecten chathamensis Hutt T & S., S *Pecten convexus (Q & G.) M Pecten delicatulus Hutt E., S Pecten hutchinsoni Hutt S Pecten palmipes Tate E Pecten triphooki Zitt M Pecten williamsoni Zitt. E Pecten yahlensis T - Woods S Pholadidea tridens (Gray) M. Pholadomya neozelanica Hutt S. Polinices gibbosus (Hutt) M Polinices ovatus (Hutt) E *Protocardia pulchella (Grav) E. Protocardia sera Hutt M *Psammobia lineolata Grav E *Siphonium planatum Sut M *Tellina glabrella Desh S Teredo heaphyi Zitt T & S Trochus avarus Sut E Trochus nodosus Hutt E Turbo etheridgei T -Woods E Turritella bicincta Hutt E Turritella concava Hutt M., S Turritella murrayana (Tate) E., T & S *Venericardia difficilis (Desh) T & S Venericardia difficilis benhami (Thomson). E Venericardia purpurata (Desh) (australis Lamk) S

It is generally admitted that the tuffs underneath the limestone in the Whitewater Creek section are in the same stratigraphical position as those between the limestones in the Porter. From the former the following were collected — *Ancilla australis (Sow.). M. *Ancilla depressa (Sow.). T. & S. *Callistoma aucklandicum Smith. T. & S. *Calyptraea maculata (Q. & G.). M. *Cantharidus tenebrosus A Ad. M. *Capulus australis (Lamk.) M Cardium facetum Sut. T. & S Cardium patulum Hutt. M. *Cochlodesma angasi (C. & F.)? M. Cymatium minimum (Hutt.). M. Cypraea ovulatella Tate M. Cypraea trelissickensis Sut. M. Cytherea chariessa Sut M *Emarginula wannonensis Hars M Epitonium cylindrellum Sut M. Fissuridea annulata Sut. M. Lapparia corrugata (Hutt.). M. *Lima bullata (Born). M., T. & S. Lima colorata Hutt. M. *Limopsis aurita (Brocchi). M. *Mactra elongata Q. & G. M. Marginella dubia Hutt. M. Pecten palmipes Tate. M. Polinices huttoni Iher. M., T. & S. Protocardia sera Hutt. M. Siphonalia orbita Hutt. T. & S. *Siphonium planatum Sut. M. Teredo heaphyi Zitt. M., T. & S. Trivia zealandica T. W. Kirk. M. Trochus nodosus Hutt. M. Turritella bicincta Hutt. M., T. & S. In addition to the species enumerated, Hutton records the occurrence of the following in the tuffs and greensands, presumably from the upper tuffs, since he does not mention the lower tuffs (loc. cit., p. 405):— Ancilla hebera (Hutt). *Arca decussata (Sow.) Brissus eximius Zitt *Calliostoma spectabile (A. Ad.). Crassatellites attenuatus (Hutt) Echinus woodsn Laube Flabellum laticostatum T.-Woods Flabellum sphenodeum T.-Woods. Fulgoraria attenuata (Hutt.). Leicidaris australis Duncan. Lima jeffreysiana Tate. Mitra enysi (Hutt.). Mytilus striatus Hutt. *Natica australis (Hutt.). Paphia attenuata Hutt. Pecten polymorphoides Zitt. Pericosmus compressus McCoy. Turritella ambulacrum Sow. Venericardia pseutes Sut. Hutton remarks that Flabellum laticostatum is not found elsewhere, but it certainly occurs in the lower tuff bed at Broken River. He also places the percentage of Recent Mollusca at about 10 per cent. This, however, is much too low in point of fact, and the discrepancy emphasizes the danger of trusting to such percentages for the purposes of determining the age when the number of forms collected is comparatively small. He also records the occurrence of Pecten hochstetteri Zitt. and Waldheimia triangularis Hutt. in the overlying limestone. From the upper tuff beds McKay records obtaining Echinus enysi Hutt., Cidaris (plates and spines), Meoma tuberculata Hutt., Turbinolia, and numerous branching and leaf corals which I do not think have been definitely identified. The following is a list of the fossils collected from the shell bed above the upper limestone at the spot where the great fall of rock into the Porter River has taken place. The greater number occur in a layer which it is

evident lies conformably on the limestone in the layer next above, being composed of finely commmuted shells. Ancilla papillata (Tate) S. Ancilla pseudaustralis (Tate) T. & S. *Anomia walteri Hect. T & S., S *Arca decussata (Sow) T & S Astraea (Cyclocantha) bicarinata Sut S. Astraea transenna Sut T & S Astraea tuberosa Sut. T & S Brechites sp.? S. *Calyptraea alta (Hutt) S *Calyptraea maculata (Q. & G) Cardium spatiosum Hutt T & S., S Cardium patulum Hutt T & S Cardium subcordatum Sut T & S Cerithium nodosum Hutt S Chione speighti Sut. S *Cominella carinata (Hutt) S Crassatellites amplus (Zitt) S Crassatellites attenuatus (Hutt) T & S, S *Crassatellites obesus (A Ad) T & S, S Crepidula gregaria Sow. S *Crepidula monoxyla (Less) S Crepidula stata (Hutt) T & S. S Cucullaea alta Sow S Cucullaea alta vai B Hutt S Cypraea ovulatella Tate S Cytherea enysi Hutt S Cytherea sulcata (Hutt.) S *Dosinia greyi Zitt S *Dosinia inagna Hutt T & S, S Dosinia subosea (Grav) S *Fulgoraria arabica (Mart) T & S, S Glycymeris cordata (Hutt) S Glycymeris globosa (Hutt). T & S Hinnites trailli Hutt T & S, S. *Lima angulata Sow S *Lima bullata (Born). T & S Lima colorata Hutt T & S, S. Lima jeffreysiana Tate S Macrocallista assimilis (Hutt) S Modiolana elongata (Hutt) T. & S Modiolus dolichus Sut S *Mytilus canaliculus Mart T & S. Mytilus huttoni Cossm S *Mytilus magellanicus Lamk S Mytilus striatus Hutt T & S Paphia curta (Hutt.) S *Paphia intermedia (Q & G) T & S Pecten burnetti Zitt S Pecten chathamensis Hutt S Pecten hutchinsoni Hutt S Polnces gibbosus (Hutt) T & S, S *Psammobia stangeri Grav S Pecten semiplicatus Hutt. S *Siphonalia dilatata (Q & G) T & S. S *Siphonalia nodosa (Mart) S Siphonalia turrita Sut T & S *Stephopoma nucleogranosum Verco T & S Struthiolaria tuberculata Hutt T & S, S Trochus nodosus Hutt S Turbo superbus Zitt T & S Turritella concava Hutt T & S, S. Turritella patagonica Sow T & S *Venericardia difficilis (Desh) S Venericardia pseutes Sut (patagonica Hutt) S *Zenatia acinaces (Q & G) S The locality from which the above list of fossils comes is such a striking one that its absence of record by McKay and Hutton probably indicates that the fall of rock had not taken place in then time. The latter gives a list of fossils obtained from the beds immediately overlying the limestone in the Whitewater Creek and the Thomas River (loc cit, p. 409), which will correspond closely in stratigraphical position with these beds, and twenty-nine species are recorded as occurring in them of these, fourteen species occur in the above list from the “Shell bed,” and of the remaining

fifteen species five occur in lower beds of the series, while the following are unrecorded.— *Chione yatei (Gray) *Crepidula costata (Sow.). Cucullaea worthingtoni Hutt. *Cytherea oblonga (Hanley) *Glycymeris laticostata (Q & G.). Lima crassa Hutt *Serpulorbis sipho (Lamk.). Turritella cavershamensis Harris. *Turritella rosea Q. & G. Volutospina (Athleta) huttoni Sut. var. pseudorarispina (McCoy). Of these ten species, five are Recent and five extinct forms. From the upper beds of the Whitewater Creek and the Thomas River Hutton collected the following Mollusca (loc. cit., p. 409):— Ancilla australis (Sow.) *Anomia undata Hutt Bathytoma sulcata (Hutt) *Calyptrea maculata (Q & G.). *Cantharidus tenebrosus A. Ad. Cerithium nodosum Hutt *Chamostrea albida (Lamk) Cominella carinata (Hutt.). *Cominella maculata (Mart) Crassatellites amplus (Zitt) *Crepidula monoxyla Less. Cytherea enysi Hutt. Dentalium solidum Hutt *Fulgoraria arabica Mart. *Fulgoraria gracilis (Swains). *Glycymeris laticostata Q & G Hemiconus trailli (Hutt) Macrocallista assimilis (Hutt). *Mactra discors Gray. *Mytilus canaliculus Mart. *Natica australis Tate. Olivella neozelanica (Hutt.). Perna sp ind. Polinices gibbosus (Hutt.). Polinices huttoni Iher. Polinices ovatus (Hutt.). *Siphonalia mandarina (Duclos). Struthiolaria cingulata Zitt. Struthiolaria obesa Hutt. Struthiolaria tuberculata var. B Hutt. Surcula hamiltoni (Hutt.). *Thais succincta (Mart) Turbo superbus Zitt. *Turritella symmetrica Hutt. Venericardia pseutes Sut. Volutospina (Athleta) huttoni Sut. The following is a list of the fossils obtained in the small tributary of Whitewater Creek coming in from the north, in what may be called the Struthiolaria bed from the number of remains of this genus occurring. The same bed is met with in a similar stratigraphical position in the Porter River between the gorges and in the Thomas River. Ancilla papillata (Tate). *Crassatellites obesus (A Ad). Crepidula gregaria Sow Cylichnella enysi (Hutt.). Polinices gibbosus (Hutt.). *Psammobia stangeri Gray. Struthiolaria tuberculata Hutt. In Murderer's Creek, just below Parapet Rock, from beds which are in the same stratigraphical position as the foregoing, the following were obtained:— Crassatellites amplus (Zitt). Mesodesma australe (Gmel). Polinices gibbosus (Hutt.). Siphonalia cf. nodosa (Mart.). (An imperfect specimen.) The salient features brought out by these lists are as follows:— (1.) The lowest beds contain a Tertiary land flora. (2.) Over these lie conformably beds containing a Cretaceous marine fauna, a fauna which in other parts of New Zealand, such as at Malvern

Hills, Waipara, and Amuri Bluff, is associated with marine reptiles and belemnites and ammonites—ie., free-swimming forms. (3.) All the remaining horizons indicate a Tertiary fauna, the oldest—i e, the tuffs interstratified in the marls at Coleridge Creek—containing only 19 per cent. of Recent species, while the higher beds contain as follows: The tuff under the limestone at the Porter River, 7 species out of 9; the tuff bed below the limestone in Broken River, 11 out of 24; the tuff bed between the limestones near the mouth of the Thomas River, 33 out of 87, or 38 per cent.; the tuff bed under the limestone at Whitewater Creek contains 11 out of 29 species; the shell bed at the fall of rock yields 21 out of 66—ie., 32 per cent; while from the higher beds in the Thomas River and Whitewater Creek, out of 42 species 13, or 31 per cent, are Recent. The low percentage of Recent forms in the tuffs interstratified in the marls at Coleridge Creek is very significant, as it indicates a lower Tertiary age for these beds—granting that we can rely on percentages of Recent forms as an accurate basis for age-determination—whereas the higher beds contain a percentage that would lead one to think that they were mid-Tertiary There as, however, no regular increase in percentages as higher beds in the series are considered, a fact perhaps due to the conditions not favouring the entombment of a representative fauna, or, if it has been preserved, it has not been thoroughly examined or collected from Several well-marked and distinctive forms are, however, persistent right through the series. There is no indication of a distinct faunal break from the tuffs under the limestone up to the highest beds from which fossils have been recorded. There is, on the other hand, this low percentage of Recent forms in the tuffs interstratified in the marls, and a clear faunal break between these beds and those containing a Cretaceous fauna, although the flora associated with this is undoubtedly Tertiary There is, however, no evidence of the existence of a physical break between these marls and the oyster bed with its Cretaceous fauna, and this opinion has been held by every authority who has examined the section—even Hutton agrees with this, and, further, the last-named authority even agrees that there is no physical break up to the top of the lower limestone: nevertheless the tuffs underneath it undoubtedly contain a Tertiary fauna. This is one of the most important points brought out by a consideration of the locality, and it strongly supports the contention urged by Marshall, Speight, and Cotton, 1910 (loc. cit), and later by Marshall alone, as to the physical conformity of our Cretaceous and Tertiary series* P Marshall, The Younger Rock Series of New Zealand, Geol. Mag (n.s), dec. v, vol. 9, 1912, p. 314, The “Cretaceo-Tertiary' of New Zealand, Geol Mag. (n.s.), dec v, vol. 10, 1913, p. 286; New Zealand and Adjacent Islands, Handbuch der regionalen Geologie, Band 7, Abt. 1, Heidelberg, 1911, p. 28. The two chief explanations put forward to account for the association of Cretaceous and Tertiary forms in a conformable series are—(1) That the Cretaceous forms survived into the Tertiary era in this region after they had disappeared from other parts of the world, perhaps owing to their having been cut off by land barriers from competition with other forms, just as sea barriers enabled the different archaic forms of land-animals to persist on the land in this part of the world long after they had become extinct elsewhere; (2) that owing to the slow deposition of the marine beds in late Cretaceous

and early Tertiary times a comparatively thin series of deposits represents an enormous period, and that ample space of time was thereby afforded for great changes in the fauna. The main difficulty in accepting the former explanation is that in other localities in fairly close proximity the molluscan fauna of more or less fixed habitat is associated with free-moving forms and with marine reptiles, both of which are likely to have a wide range in space; and if they peopled other parts of the ocean they must, except under most peculiar circumstances, have penetrated the then New Zealand seas, seeing that it is very unlikely that a portion of the sea surface of the world was completely shut off from the general ocean. It is, of course, possible, but improbable The other explanation therefore appears to me to be more satisfactory. In the area under consideration the thickness of the beds between those containing the Cretaceous fauna and those with a Tertiary fauna certainly amounts to 1,800 ft—no mean thickness—and the conditions of deposition were in all probability very slow. In discussing the circumstances of deposition it is important to note that the form of the land surface was entirely different from that now existing,* R. Speight, The Intermontane Basins of Canterbury, Trans. N.Z. Inst., vol. 47, 1915, p 345. and that the land in the neighbourhood was probably of low relief and did not furnish any large amount of sediment. The greensands and marls were also laid down in fairly deep water, where deposition would be slow It seems probable that after the deposition of the coals and estuarine beds at the base of the series the area was depressed, and deposition was slow in the relatively deep water in which the marl was laid down, the unconformity, if any, being due to depression beyond the limit of deposition—that is, a per saltum unconformity, and not one due to elevation of the land, erosion, and subsequent depression. That a shallowing of the sea took place towards the close of this period is evident from the presence of the interstratified sands and conglomerates, which become increasingly important towards the top of the series. Numerous instances of slow deposition and small thickness of beds in one area associated with great thickness in an adjoining area can be cited from many parts of the world, but perhaps the most striking one is that afforded by the Silurian beds of Scandinavia, with their compressed though complete sequence, as compared with the beds of the same age in Wales, which are distinguished by their thickness, the former owing their relative thinness to slow deposition in deep water off the shore-line of a continent. Volcanic activity became manifest in certain parts of the Trelissick area while the marls and sands were being laid down, especially in the neighbourhood of Whitewater and Coleridge Creeks, the eruptions being almost entirely of fragmentary material, and submarine. In close proximity to the centres of volcanic action the limestones thin out and disappear, whereas they are thick in those parts of the area remote from volcanic activity, especially towards the east of the basin. It is probable that while volcanic activity was fairly continuous at one point deposition of limestones was continuous a little distance away, and that in the intermediate localities the limestones and volcanic products are interstratified

List of Tertiary Mollusca Obtained In The Trelissick Basin Arranged For Purposes of Comparison The numbers at the heads of the various columns correspond with the following localities:— 1. Tuff beds interstratified with marls, Coleridge Creek 2. Lower tuff beds, Porter River, just above the limestone gorge 3. Lower tuff beds, Broken River, just below the limestone gorge 4. Upper tuff beds, junction of Thomas River with Porter River 5. Tuff beds, below limestone, Whitewater Creek 6. Shell bed, at rock fall in Porter River 7. Sandy beds between the two gorges of the Porter River and in the Thomas River near its junction with the Porter (according to Hutton), also specimens collected by the author in beds of similar position in a tributary on north side of the Whitewater River (The suffix h indicates that Hutton is responsible for the record of the occurrence of the species in the locality) — 1 2 3 4 5 6 7 *Admete trailli (Hutt) × Ampullina miocenica Sut × — suturalis (Hutt) × × *Ancilla australis (Sow) × × ×h *— bicolor (Grav) × *— depressa (Sow) × — hebera (Hutt) ×h *— mucronata (Sow) × — papillata (Tate) × × × × — pseudaustralis (Tate) × × — subgradata (Tate) × * Anomia undata Hutt ×h *— walteri Hect × *Arca decussata (Sow) ×h × Astraea bwamnata Sut × — transenna Sut × — tuberosa Sut × Bathytoma sulcata (Hutt) ×h Brechites sp × *Calliostoma acutangulum Sut × — aucklandum Sm × × × × — filiferum Sut × — oryctum Sut × * — punctulatum (Mart) *— spectabile (A Ad) ×h *Calyptraea alta (Hutt) × *— maculata (Q & G) × × × ×h *— maculata inflata (Hutt) × × *— scutum Less × *Canthandus pruninus perobtusus Pils × *— tenebrosus A Ad × × ×h *Capulus australis (Lamk) × × Cardium facetum Sut × — patulum Hutt × — spatiosum Hutt × × × — waitakiense Sut × Cerithium nodosum Hutt × ×h *Chamostraea albida (Lamk) ×h Chione chiloensis truncata Sut ×

Chione speighti Sut × — subroborata Tate × *— yatei (Gray) × ×h Clio tatei Sut ×? × *Cochlodesma angasi (C & F) Cominella carinata (Hutt) × ×h — intermedia Sut × *— maculata (Mart) ×h Coptochetus sp? × Corbula humerosa Hutt × *— zelandica Q & G × Crassatellites amplus (Zitt) × ×h — attenuatus (Hutt) ×h × — cordiformis Sut × *— obesus (A Ad) × × × *Crepidula costata (Sow) ×h *— crepidula (L) × — gregaria Sow × × — monoxyla Less × — striata (Hutt) × × Cucullaea alta Sow × — alta var B Hutt × — ponderosa Hutt × — worthingtoni Hutt ×h Cylichnella enysi (Hutt) × × *— striata (Hutt) Cymatium minimum (Hutt) × Cypraea ovulatella Tate × × × × — trelissickensis Sut × Cytherea chariessa Sut × — enysi Hutt × ×h *— oblonga (Hanley) ×h Daphnella neozelanica Sut × Dentalium solidum Hutt × ×h *Diplodonta striata Hutt × *— zelandica (Gray) × × *Divaricella cumingi (Ad & Ang) × *Dosinia caerulea (Reeve) × *— greyi Zitt × — magna Hutt × *— subrosea (Gray) × *ginula striatula Q & G *wannonensis Harris × × × Epitonium cylindrellum Sut × × — marginatum (Hutt) × — rugulosum lyratum (Zitt) × — zelebori (Dkr) var × *Euthria striata (Hutt) × Fissuridea annulata Sut × *— arabica elongata (Swains) × × — attenuata Hutt ×h *— gracilis (Swains) × × ×h Fusinus bicarinatus Sut × Glycymeris cordata (Hutt) × × — globosa (Hutt) × × *— laticostata (Q & G) ×h ×h Hemiconus trailli (Hutt) ×h * Hemifusus gonoides Sut × Hinnites trailli Hutt × Lapparia corrugata (Hutt) × × Leucosyrinx alta (Harris) ×

*Lima angulata Sow × × *— bullata (Born) × × × — colorata Hutt × × × — crassa Hutt ×h — huttoni Sut × — jeffreysiana Tate ×h × — lima (L) × *Limopsis aurita (Brocchi) × × — catenata Sut × × — zitteli Iher × *Loripes concinna Hutt × Macrocallista assimilis (Hutt) × × *— multistriata (Sow) × Mactra attenuata Hutt × *— discors Gray ×h *— elongata Q & G × × Marginella dubia Hutt × × — harrisi Cossm × Mesodesma australe (Gmel) ×h Mitra enysi (Hutt) × Modiolaria elongata (Hutt) × × × *Modiolus australis (Gray) × — dolichus Sut × Monilea praetextilus Sut × — sulcatina Sut × *Myodora subrostrata Smith × Mytilus canaliculus Mart × ×h — huttoni Cossm × — magellanicus Lamk × — striatus Hutt ×h × *Natica australis (Hutt) × ×h ×h *— zelandica Q & G × × Olivella neozelanica (Hutt) × ×h Panopea orbita Hutt × × — worthingtoni Hutt × *— zelandica Q & G × Paphia attenuata Hutt ×h — curta (Hutt) × × × *— intermedia (Q & G) × × Pecten athleta Zitt × — beethami Hutt × — burnetti Zitt × — chathamensis Hutt × × *— convexus Q & G × — delicatulus Hutt × — hutchinsoni Hutt × × × — palmpipes Tate × — polymorphoides Zitt ×h — semiplicatus Hutt × — triphooki Zitt × — williamsoni Zitt × × — yahlensis T Woods × × — sp? × Pholadidea concentrica Sut × *— tridens (Gray) × Pholadomya neozelanica Hutt × Phos cingulatus (Hutt) × × × ×h — huttoni Iher × × ×h — ovatus (Hutt) × × ×h *Protocardia pulchella (Gray) × — sera Hutt × ×

*Psammobia lineolata Gray × × *— stangeri Gray × × Seila huttoni Sut × *Serpulorbis sipho (Lamk) ×h Siphonalia costata (Hutt.) × *— dilatata Q & G × *— mandarina (Duclos) × ×h *— nodosa (Mart) × × × — orbita Hutt × — turrita Sut × × — sp × *Siphonium plantaum Sut × × × × × *Stephopoma nucleogranosum Verco × Struthiolaria cincta Hutt × — cingulata Zitt ×h — obesa Hutt ×h — spinosa Hect ×h — tuberculata Hutt × × Surcula antegypsata Sut ×? — hamiltoni (Hutt) ×h — seminuda Sut × *Tellina eugonia Sut × *— glabrella Desh × Terebra costata Hutt × Teredo heaphyi Zitt × × *Thais succincta (Mart) ×h Trivia zealandica T W Kirk × Trochus avarus Sut × — cucinatus Hutt × × × × — nodosus Hutt × × × × — tiaratus Q & G Turbo etheridgei T -Woods × — granosus (Mart) — superbus Zitt × ×h Turritella ambulacrum Sow ×h — bicincta Hutt × × — cavershamensis Harris ×h — concava Hutt × × × — murrayana (Tate) × — patagonica Sow × *— rosea Q & G ×h *— symmetrica (Hutt) ×h *Venericardia difficilis (Desh) × × — difficilis benhami (Thomson) — pseutes Sut ×h × ×h — purpurata (Desh) × Volutospina huttoni pseudorarispina Sut ×h *Zenatia acinaces (Q & G) × I have compared this list with that published by Marshall in connection with the Target Gully beds, near Oamaru,† P. Marshall, Cainozoic Fossils from Oamaru, Trans, N.Z. Inst., vol. 47, 1915, p. 378. and find that there is a general similarity in fossil-content to the beds in the Trelissick Basin; but, of course, some forms are missing from each list. This discrepancy may be due to insufficient collecting, or to absence of the fossils, or to a difference in time allowing for some forms to become extinct. The

Oamaru locality is rich in small species, which perhaps points to insufficient collecting in the Canterbury locality; but certain larger forms present in one list are wanting in the other. For example, we do not find, in the former, species of Astraea, Calliostoma, Cardium, Cucullaea, Lima, Mactra, Mytilus, Pecten, Struthiolaria, Trochus, and Turbo, or find them comparatively rarely; while the latter are relatively poor in species of Bathytoma, Dentalium, Drillia, Latirus, Mangilia, and Turris By the kindness of Mr. Morgan, Under-Secretary for Mines, I have been allowed to see the lists of fossils to be published in the forthcoming bulletin on the Oamaru District, by Professor Park, and I find that this discrepancy is not sensibly removed even with longer lists There does seem, however, to be a slight preponderance of extinct forms in the species predominant at Castle Hill over those common at Target Gully, which indicates a rather older set of beds at the former place On comparing the list of fossils with that obtained by the author at the Lower Waipara Gorge* R. Speight, A Preliminary Account of the Lower Waipara Gorge, Trans. N.Z Inst., vol 44, 1912, p. 231. it is found that the resemblance is closer, the only marked difference being the presence of the numerous oysters of various species at Waipara, whereas they are practically absent from the mid-Canterbury district There is a remarkable similarity in the percentage of Recent forms from both these localities, which suggests an approximately identical age I have not been able to compare these lists with those from the beds in similar position in the Weka Pass section, as they have not been published up to the present date; and in face of the fact that these, and others of equal value for the purpose of correlation of our Tertiary series, are likely to appear at an early date, it seems unwise to comment further, or to attempt to draw conclusions which may be entirely upset after a consideration of fuller evidence