A Revision of the Type Clarentian Section at Coverham, Clarence Valley. S35.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 83, 1955-56, Page 93

A Revision of the Type Clarentian Section at Coverham, Clarence Valley. S35. H. W. Wellman Abstract The structure and stratigraphy of the type Clarentian section near Coverham, at the north-eastern end of the Clarence Valley is described and illustrated by a map. The section extends from greywacke of uncertain but pre-Albian age to lower Tertiary limestone. It was previously considered to be uniformly north-east dipping and substantially continuous, but is now known to be folded into two synclines and to be interrupted by several faults. The accepted stratigraphic order has been considerably changed, and this has made a reclassification of the New Zealand Cretaceous sediments necessary. A new classification is outlined and correlations made between the sediments at Coverham and those in other parts of New Zealand and in Europe. Introduction At the north-east end of Clarence Valley four streams—Wharfe Stream, Cover Creek, Nidd Stream, and Swale River—join to form Ouse River, which flows south-west for five miles to join Clarence River. Coverham is the site of an old homestead at Cover Creek and not far from the junction of the streams. Although in the valley of Clarence River, it is best reached from Kekerengu, a small railway settlement on the east coast road 35 miles from Blenheim, and at the mouth of Kekerengu Stream. From Kekerengu a formed road extends up the Kekerengu Valley for five miles to Remuera Homestead. From the homestead a track, which has recently been widened so that it can be used by tractors in fine weather, climbs 2,000ft. to Burnt Saddle and then drops 1,500ft. on the Clarence side to the homestead at Coverham. At Burnt Saddle an old track branches off to the north. It crosses the ridge between Wharfe and Cover streams and continues on the south side of Cover Stream Valley to rejoin the widened track about a mile east of Coverham (see Fig.1). Previous Geological Investigations The geology of Coverham has been described by Hector (1886) and by McKay (1886) and in greater detail by Thomson (1919), all of whom considered the sequence from the greywacke to the limestone to be essentially continuous and in order of progressively decreasing age. The geomorphology has been described by Cotton who, in numerous diagrams (1913, Fig. 2; 1945, Fig. 31; 1950, Figs. 5 and 6) has represented the structure as a simple fault angle depression. McKay thought that he recognized at Coverham Cretaceous strata of the same age as those at Amuri Bluff, later termed Piripauan by Thomson (1919). Piripauan does occur at Coverham in two synclines, but McKay does not mention them, and the rocks he considered to resemble those of Amuri Bluff are probably Clarentian. Thomson made the Coverham sequence the type for his Clarence Stage (now Series, see Finlay and Marwick, 1940). Within it he recognized six formations which are given below with his thicknesses and in the order in which he considered them to occur:—

Thomson's Clarentian Formations Feet. Sawpit Gully mudstones 3,200 Nidd sandstones and mudstones 550 Cover Creek mudstones 2,000 Wharf Gorge sandstones 450 Wharf mudstones 1,500 Basal conglomerates 250 Fig. 2.—Locality maps showing position of Sheet District S35 and position of that part of Coverham district illustrated by Figs. 2 and 6. Arrows show inferred direction of origin of redeposited Clarentian sediments as determined from attitude of sandstone tongues that filled scour channels in underlying mudstone. In the following account Cover Stream is used instead of Cover Creek, the singular is used instead of the plural for formation lithologies, and Wharfe is given its original Yorkshire spelling by the addition of a final “e”. Observations Made by Writer As the result of field work done by the writer on the east coast of the North Island during the summers of 1952–3 and 1953–4 it became apparent that the Cretaceous stratigraphy of New Zealand needed considerable revision. The type

section for the Clarence Series is at Coverham, and 16 days were spent in the district during the early part of April, 1954, about half being spent at Coverham itself and half in the middle Clarence Valley as far upstream as Bluff Stream. The Coverham area is compact and easily accessible, and the time available permitted a full examination of all the main stream sections; the eastern limit of the area examined being the headwaters of Kekerengu River, of Nidd Stream, and of Wharfe Stream. The first discovery was the recognition of the key Piripauan fossil Inoceramus australis Woods (Gs 6130)* Numbers prefixed “Gs” refer to collections at the Geological Survey, Wellington. in sandstone at Whernside Saddle and in similar sandstone at the headwaters of Nidd Stream. The sandstone is overlain at both places by flinty shale that weathers rusty and has a conspicuous sulphurous efflorescence that is probably mostly melanterite. Similar shale, four miles southeast, in the middle reaches of Kekerengu River was classed as Mangatu by Macpherson (1952) and by Finlay in the same paper shown to contain a Piripauan foraminiferal fauna with Rzehakina epigona. In the North Island similar shale with Rzehakina epigona was named the Whangai Formation by Quennell, in 1937. In the North Island, the Whangai “Series” was proposed for similar shale by Quennell and Brown (1937) and the name was modified to Whangai Formation by Finlay and Marwick (1948). It also contains Rzehakina epigona. Formations are best named from “a geographical name coupled with a lithological term which is descriptive of the rock” (Australian Code of Stratigraphic Nomenclature, 1950) so the name Whangai Shale is used in this paper for the Whangai Formation of previous writers. At Burnt Saddle and in Kekerengu River the shale is overlain by the well-known flint beds, and these in turn by the Amuri Limestone. The recognition of the Whangai Shale is an essential preliminary to the interpretation of the Coverham structure. From Whernside Saddle the shale was traced into the head of Cover Stream and into the north branch of Wharfe Stream where it is well exposed on the old track. It then strikes south-west into the Wharfe Stream downstream from the crossing of the widened track and extends as a progressively narrowing strip to within half-a-mile of the upper end of Wharfe Gorge. The synclinal structure is evident from dip and from the correspondence of the strata and fossils on each side of the Whangai strip. Another Whangai-filled syncline lies to the north within the valley of Nidd Stream, where the Whangai sediments extend to within a few hundred feet of the mouth of Sawpit Gully. The synclinal fold is visible in Nidd Stream, the axis pitching north-east at 15°, a direction of pitch that is in agreement with the distribution of the Whangai Shale, which becomes more extensive to the north-east and joins with the Whangai belt at the Wharfe Stream. Once the two synclines were recognized it became apparent that Thomson's type section line (from Wharfe Stream north to Sawpit Gully) is not the uniformly north-west dipping section he thought it to be. But by measuring sections on the flanks of the synclines the true sequence and the proper order of Thomson's formation were determined without difficulty. Thomson considered the Sawpit Gully Mudstone to be the youngest member of the Clarentian and to directly overlie the Nidd Sandstone It was found that the true sequence is the reverse of Thomson's.

The following table relates Thomson's formation to the formations and zones of this paper. It will be noticed that a new series—the Raukumara—is introduced, and that the Clarence Series is redefined. The reasons for re-defining the Clarence and the new Series boundaries will be given later. Table Relating Thomson's Formations to the Formations and Zones of this Paper. Tertiary. Wellman, this paper. Thomson, 1919. Landon, Arnold & Dannevnke 1.Amuri and Weka Pass limestones. Amuri and Weka Pass limestones. 2. Flint beds. Flint beds. Mata 3.Whangai Shale. Not recognized. 4. Inoceramus australis Zone. Not recognized. Upper Cretaceous. 5.Inoceramus A Zone. Not recognized. Raukumara 6. Inoceramus B Zone. Nidd Sandstone. 7.Inoceramus bicorrugatus Zone. Nidd Sandstone. 8. Inoceramus porectus Zone. Upper part Sawpit Gully Mudstone. Upper part Cover Stream Mudstone. Clarence Upper part Wharfe Mudstone. 9. Inoceramus concentrcus Zone. Lower part Cover Stream Mudstone. 10. Inoceramus anglicus? Zone. Lower part Wharfe Mudstone (p. 313). Lower Cretaceous. 11. Wharfe Gorge Sandstone. Wharfe Gorge Sandstone. Taitai? 12. Inocranms C. Zone Northern part Wharfe Mudstone. 13.“Basal Conglomerate”. Basal Conglomerate. Uncertain 14. “Greywacke”. Pre-Notocene sandstone and shale. Of the divisions used in this paper, 1 and 2 are as described by Thomson. Division 3—the Whangai Shale—was not recognized or else was confused with the overlying flint beds. Division 4 is not well developed along the type section line and was not included in Thomson's formation, but was mentioned by him (p. 316) as the green sandstones on the Kekerengu side of the pack-track that had been considered to directly overlie the “basal conglomerate” by Hector (1886: xxxii). Division 5 does not appear to have been recognized, although the key fossil Inoceramus sp. A. had been collected by Thomson from Wharfe Stream. Divisions 6 and 7 correspond with Thomson 's Nidd Sandstone Division 8 is repeated three times in the type section and corresponds to the upper part of the Sawpit Gully Mudstone, to the upper part of the Cover Stream Mudstone, and to the upper part of the Wharfe Mudstone. Division 9 corresponds with the lower part of the Cover Stream Mudstone. Division 10 is exposed in the Ouse

only and was considered as “lowest Wharfe” by Thomson. The Wharfe Gorge Sandstone—11—is as defined by Thomson. Division 12, the Inoceramus C Zone, corresponds to the northern part of Thomson's Wharfe Mudstone. The remaining two divisions are as described by Thomson. The true order of the fossiliferous parts of Thomson's formations is as follows, the formations missed by him being given in parentheses: Amuri Limestone Flint beds (Whangai Shale) (Inoceramus australis Sandstone) Nidd Sandstone Sawpit Gully, and part Wharfe Mudstone Cover Stream Mudstone Lowest part Wharfe Mudstone Wharfe Gorge Sandstone Northern part Wharfe Mudstone Proposed Cretaceous Divisions and Overseas Correlations The Clarentian as defined by Thomson in 1917 (p. 408) comprises all those Notocene rocks in the middle Clarence Valley lying below the flint beds at the base of the Amuri Limestone. The “basal conglomerate” was considered by Thomson to mark the base of the Notocene. The Clarentian as thus inclusively defined by him embraces not only the Clarence beds that he knew, but also younger beds—the Whangai Shale and the Australis Sandstone—that he had not recognized. The upper part of the Clarentian is thus synonymous with the Mata Series as defined by Finlay and Marwick in 1947. Thus on priority the Clarentian corresponds to the whole of the upper Cretaceous and to the upper part of the lower Cretaceous and is so large that it requires division. The writer restricts the Clarence Series to the fossiliferous beds best known to Thomson—that is to the Porrectus zone and the underlying fossiliferous beds down to and including the Anglicus? Mudstone above the Wharfe Gorge Sandstone. The intention is to define the Clarence Series so that it extends down to but does not include any rocks of the same age as the Taitai Series of the North Island. The available evidence, although far from conclusive, cannot be neglected. An Inoceramus like that immediately above the Wharfe Gorge Sandstone occurs a few hundred feet above Taitai-like sandstone of Mt. Arowhana in the North Island; and one like that below the Wharfe Gorge Sandstone occurs at several places below the Taitai Sandstone in Tapuwaeroa Valley. A Taitai age is thus the most probable one for the Wharfe Gorge Sandstone, and for this reason it is not included in the Clarence Series as defined here. The Mata Series has already been defined by Finlay and Marwick (1947) to extend down to include the Piripauan Stage. The type locality for this stage is at Amuri Bluff, the lowest fossiliferous beds being the calcareous sandstone with I. australis and I. pacificus. The Australis zone thus corresponds with the base of the Mata Series. Three Cretaceous zones—the A zone, B zone, and Bicorrugatus zone—he between the base of the Mata Series and the top of the Clarence Series. These are considered to comprise a new series, which is termed the Raukumara Series after the formation of that name in the North Island (Ongley and Macpherson, 1928) in which the three zones are well developed.

The key fossils and a tentative overseas correlation is set out briefly in the legend to the geological map (Fig. 5). The Whangai Shale contains Rzehakina epigona, and is thus uppermost Cretaceous or Paleocene (Glaessner, 1945: 207). The underlying Australis zone probably corresponds with the Calcareous Conglomerate of Amuri Bluff, which according to Wilckens (1922) and Woods (1917) is upper Senonian (presumably about Campanian) in age. The two Inoceramus species—A and B—that mark the underlying zones have not yet been described; they show some resemblance to the European species I. lobatus Gold. and I. lingua Gold. of lower and upper Campanian age, respectively. The underlying zone fossil I. bicorrugatus resembles varieties of I. inconstans that range in age from upper Turonian to lower Santonian (Woods, 1912). The next underlying zone contains several fossils that appear to be useful for overseas correlation. The zone fossil I. porrectus may be related to the lower Turonian fossil I. lamarki var. apicalis (Woods, 1912). At several places I. porrectus is associated with an elongated Inoceramus that resembles the widespread lower Turonian zone fossil I. labiatus (Schloth). In New Zealand Aucellina extends up to, but not above, the Porrectus zone. It is generally considered to have died out overseas at the end of the Cenomanian. This zone is probably upper Cenomanian and Turonian in age. The age of the next underlying zone is well defined. Turrilites is apparently restricted in Europe and North America to the lower Cenomanian. Inoceramus concentricus ranges slightly lower and extends down into the middle Albian. The flat Inoceramus in the dark mudstone resembles I. anglicus Woods from the middle Albian of England. The belemnite in the Wharfe Gorge Sandstone has not been described. It differs appreciably from that in the Concentricus mudstone and may indicate a significant age difference. Critical Sections The true order of superposition has been determined from the five sections described below, for which the approximate grid references are given at the end of each locality description. The five sections are correlated in Fig. 3. 1. Track and Stream South of Whernside Saddle (200460). Feet. Flinty, sulphurous siltstone (Whangai) c.1000 Light green massive medium sandstone in 5ft. bands 200 Massive medium-grained sandstone, Inoceramus cf. pacificus Gs 6103 10 Medium sandstone 10 Fine to medium-grained sandstone Inoceramus australis, Gs 6130 150 Band of maroon fine sandstone (cf. bands in Mangaotane Mudstone) 2 Dark blue grey micaceous siltstone, sulphurous, I. sp. A. c. 500 Siltstone as above with sandstone bands c. 500 Section 1 is clearly exposed on the track from Burnt Saddle to Remuera and in the tributary of Kekerengu Stream below the track. It is probably separated by a fault from the section below the flint beds on the south side of Whernside Hill and the full thickness of the Whangai sediments is uncertain. They are unlikely to be less than 1,000ft. thick. In its upper part the sandstone contains an Inoceramus similar to, but with more irregular ribbing than I. pacificus Woods and in its lower part an Inoceramus identical with I. australis Woods. I. cf. pacificus shows in the stream bed below the track (Gs 6105) and in the branch stream to the north-east. Well preserved specimens of I. australis

Fig. 3.—Chart showing variations in the thickness of four upper Cretaceous zones at five sections near Coverham. Numbers show vertical distribution of zone fossils. were collected on the track (Gs 6130) and in the stream below (Gs 6101) and it is difficult to understand why this key fossil was not collected earlier. The underlying beds, well exposed on the track and in the stream, consist of micaceous dark siltstone with a few sandstone bands, and a single band of basalt. Only one species of Inoceramus is represented, and it is common throughout. The contact with the unfossiliferous greywacke and conglomerate to the east is not exposed either in the stream or on the track, it is probably a fault. 2. Stream with Two Falls, 2,500ft. East of Sawpit Gully (165475). Feet. Flinty sulphurous siltstone (Whangai) c. 100 Not seen 10 Mudstone, dark and sulphurous 50 Hard glauconitic sandstone; waterfall, Inoceramus sp. B. 100 Soft siltstone with Inoceramus fragments 200 Medium-grained sandstone 200 Medium grey muddy sandstone, Inoceramus bicorrugatus 50 Medium blue-grey mudstone c.300 This section is well exposed in a branch of Nidd Stream flowing parallel to Sawpit Gully 2,500ft. to the east. The beds dip uniformly north-west from 70°

in the upper and to 60° in the lower part of the sequence. The lithologic and fossil sequence make the order of superposition certain and demonstrate that the beds are slightly overturned and that either a fault or an anticline (see Fig. 4) must intervene between this section and that at Sawpit Gully (Section 5). 3. Landslide Scar in North Branch of Wharfe Stream (188467). Feet. Flinty sulphurous siltstone (Whangai) 500 Interbedded 1ft. bands of Whangai and glauconitic medium sandstone 30 Well bedded alternating sandstone and siltstone, tabular concretions 15 Dark blue-grey micaceous siltstone I. sp. A. (Gs 6110) 20 Concretionary band I. sp. A. 2 Dark blue-grey siltstone I. aff. bicorrugatus (Gs 6111-2) 10 Dark blue-grey concretionary siltstone I. bicorrugatus (Gs 6113) 10 Medium blue-grey mudstone 20 Sequence not well exposed c. 10 Medium blue-grey mudstone I. porrectus (Gs 6109) 30 The landslide scar is 200ft. downstream from the junction of two sub-equal branches of the stream. The full thickness of Whangai sediments is probably exposed, but was not accurately measured. Flint beds overlie, and the flint is overlain in turn by the limestone of Whernside Hill (Thomson, 1919: 311). The beds below the Whangai Shale are well exposed, I. bicorrugatus showing near stream level and the other fossils in the scar above. I. porrectus was collected from the mudstone exposed at the stream junction. This mudstone is stratigraphically below the I. bicorrugatus beds, but the exact amount is uncertain. 4. West Side Syncline in Wharfe Stream (159458). Feet. Flinty sulphurous siltstone (Whangai) c. 50 Dark micaceous mudstone with Inoceramus sp. A. 20 Medium-grained sandstone, I. cf. bicorrugatus 30 Glauconitic gritty medium sandstone, I. bicorrugatus 20 Medium blue-grey mudstone, concretions with Aucellina euglypha (Gs 6123) 20 Lens of basalt 10 Alternating bands of mudstone and concretionary sandstone 30 Alternating beds as above with I. porrectus and Aucellina euglypha (Gs 6122) 30 Alternating beds as above with-Parapuzosia aff. haughtoni, etc. (Gs 6120) 30 The above section is composite, being based on the sections on each side of a narrow syncline in Wharfe Stream. The Whangai siltstone occupies the centre of the syncline and extends with progressively decreasing width to within 2,000ft. of the Wharfe Gorge. The siltstone is flanked by sandstone, and the sandstone in turn by mudstone. The three beds below the Whangai are exposed in a landslide scar immediately below the saddle crossed by the widened track as it passes from Wharfe Stream to Cover Valley. The remainder of the section is clearly exposed in steep bare hillside immediately downstream from the branch stream half a mile east of the north-flowing reach of Wharfe Stream. The determination of the Aucellina as A. euglypha Woods was made by Dr. C. A. Fleming. The ammonite was compared with one from Bushgrove Stream that has been examined and named Parapuzosia aff. haughtoni by Spath (1935). The Wharfe Stream ammonite was found in place at the point marked 6 on the map. When dug out it proved to be about 2ft. 6in. across (75 cm.) and weighed about 300lb. Only half was transported to Wellington, the remainder being placed on a rock ledge on the south side of Wharfe Stream 200ft. downstream from where found.

5. Sawpit Gully, Tributary of Nidd River (153480). Feet. Amuri Limestone c. 1000 Flint beds, calcareous c. 1000 Flint beds, carbonaceous bands and sulphur efflorescence 2 Dark sandstone with abundant sulphur efflorescence 4 Light blue-grey mudstone 20 Medium blue-grey concretionary mudstone. I. porrectus 80 Mudstone as above 200 Concretion in mudstone, with Gaudryceras subsacya 1 Medium to dark mudstone, fossils rare 2000 Thin bands of graded bedded sandstone 500 The critical part of this section—the junction of mudstone and flint beds—is clearly exposed. As mentioned by Thomson (p. 315) the junction appears to be conformable and unfaulted. Nevertheless several thousand feet of beds that are exposed nearby are absent (Fig. 3). The section at Sawpit Gully is thus one of the most misleading in New Zealand, and a striking example of how a strong unconformity* Throughout this paper “unconformity” is used for time unrepresented by strata. Irrespective of whether or not the time corresponds to a plane of angular discordance. or a fault may be impossible of detection even in most favourable exposures. In describing the junction it must first be made clear that the flint beds and Porrectus mudstone correspond closely with the flint beds and Porrectus mudstone in other parts of the district and that no reason exists for supposing that the missing strata have changed in character and disguised themselves either as flint beds or as Porrectus mudstone. The contact is either a fault or an unconformity, but it is not easy to decide which. Unconformity is favoured by the variations in the thickness of the Australis and the three Ran-kumara zones illustrated by Fig. 3. It will be noticed that the variations in thickness are due to thinning of all the zones and not to erosion at the base of one. It is consequently not unlikely that these four zones have thinned out to a few feet at Sawpit Gully. The Whangai Shale is also absent from Sawpit Gully. It is possible that the shale also varies in thickness together with the underlying zones, and that its apparent absence from Sawpit Gully is due to pronounced thinning. If the unconformity is due to thinning then the four feet of sandstone will represent the two thousand or so feet of missing strata. Evidence for an unconformity rather than a fault is provided by the sections exposed in the streams that cut through the flint beds south-west of Sawpit Gully. In the Swale, Mead, Limburne, and Dee rivers, the sequence is similar to that at Sawpit Gully, and it is only when the Branch is reached that the Whangai Shale appears below the flint beds. If the absence of the Whangai Shale is due to a fault, then the fault must have a nice and rather improbable parallelism to the beds for a total distance of ten miles. On the other hand, at several places between the Dee and Swale rivers, a small fault trace is noticeable at the base of the flint beds. In view of the conflicting evidence, the writer is uncertain if the Sawpit Gully section contains a hidden unconformity or an invisible fault, but he is certain that it contains either the one or the other. The stratigraphic position of the fossils mentioned in the text section above is taken from Thomson (p. 515). Two other fossils that may be important have been collected since, but it is difficult to give their exact positions relative to Thomson's collections. An Inoceramus (Gs 5824) with irregular detailed ornamentation like I. cuvieri Sowerby (Woods, 1911, Fig. 73) was collected by Messrs.

Couper and McQueen from about 50 feet below the flint beds. Slightly lower in the section a coral (Gs 5819) was found with I. porrectus. A similar coral (Gs 6094) also occurs with I. porrectus at the mouth of Sawpit Gully in mudstone that directly underlies the Raukumara Sandstone. Content and Distribution of Formations and Fossil Zones Limestones Within the mapped area limestone forms the escarpment of the Chalk Range and the ridge that extends north-east from Whernside Hill. The limestone has been described in detail by McKay and by Thomson. It is Tertiary in age and thus outside the scope of this paper. The limestone near Coverham is mostly too hard to provide foraminiferal faunas. It is less calcareous and softer in the Kekerengu valley where it was sampled by Macpherson (1952) and found to be Mangaorapan or Heretaungan (lower Eocene) near its base, upper Bortonian (middle Eocene) in its middle part, and Whaingaroan, Duntroonian, and Waitakian (Oligocene) in its upper part. Flint Beds Flint beds directly and conformably underlie the limestone in the Coverham district and at Kekerengu River. They resist erosion and together with the flinty part of the limestone form the cliffs of the Chalk Range, and the steep slopes of Whernside Hill. No fossils have yet been found in the harder part of the flint beds, and it has been assumed in the map and charts that the base of the flints corresponds with the base of the Dannevirke Series. This cannot be proved but is unlikely to be far in error. The origin of the flint beds has been discussed in considered detail by Thomson (1916) who concluded that the silica was chemically deposited under marine conditions. Samples from carbonaceous bands at the base of the flints were examined for spores and pollens by Mr. Couper, but only a few specimens found. These were sufficient to indicate a plant origin for the carbonaceous material but not sufficient to tell anything worth while about the nature of the plants. Whangai Shale Distinctive sediments recognized by Macpherson in the Kekerengu River, and named the Woolshed Shale, were described as “dark-brown and dark-grey siliceous carbonaceous shales, with splintery fracture …, rare nodules and scattered pyrites, and yellow efflorescence on some protected exposures”. Rare bands of greensand and glauconitic sandstone are interbedded with the shale at Coverham and Kekerengu River. Macpherson noted a resemblance to the lower part of the Mangatu Series of Poverty Bay and to the Waipawa Series of Hawke's Bay. The Mangatu and Waipawa Series both include sediments other than the distinctive shales, and for this reason Quennell (1937) proposed the name Whangai for the shale and its immediate associates. The name is taken from the Whangai Range in Hawke's Bay, and the writer uses it for the shale of Coverham. At places the shale is almost as resistant as the flint beds and forms bluffs and cliffs, but almost as often it slumps to form low knobs. It is conspicuous in the Coverham district but does not quite extend to the type section line proposed by Thomson. It underlies the flint beds of Whernside

Hill and extends north-east as a wide belt to Nidd River. To the west of the south-flowing reach of this river it extends as a narrow synclinal core almost to the mouth of Sawpit Gully. To the south it extends from the ridge between Cover and Wharfe streams, where it is well exposed on the old track, to the valley of Wharfe Stream, where it forms the core of a tight syncline and is well exposed on both sides of the stream. The shale is absent from the Sawpit Gully section. In the North Island ammonites have been collected from the Whangai shale by McKay (Marwick, 1950) and Macpherson (1951:264) but unfortunately they do no more than to make it certain that the shale is Cretaceous and not Tertiary. The best age evidence is from foraminifera, particularly from Rzehakina epigona which in overseas formations is considered to be restricted to the Maestrichtian, Danian and Paleocene (Glaessner, 1945) and in New Zealand to the Piripauan and Teurian (Finlay, 1939). At many North Island localities R. epigona first appears in the base of the Whangai Shale and extends up to or just beyond the top. At Kekerengu it was found in samples from the top and bottom of the shale, making correlation with the Whangai Shale of the North Island reasonably certain. The samples of Whangai Shale collected at Coverham have not yet been examined for foraminifera. As already mentioned, the shale is absent from Sawpit Gully. It is also absent to the south-west from Mead, Limburne, Dee, and probably from Swale valleys, but is well exposed in Branch River valley and in the valleys further to the south-east in which direction it progressively changes to a sulphurous sandstone with large concretions. A concretion with saurian remains (Gs 6145) was collected from a band of sandstone interbedded with the Whangai Shale at Branch River and McKay (1886, p. 75) described saurian concretions from the sulphurous sandstone of Muzzle River a few miles south-west. It is thus highly probable that the Whangai Shale represents the fine-grained (and more offshore ?) equivalent of the well-known sulphur (or saurian) sands of Amuri Bluff and Waipara. The distinctive nature of the Whangai Shale gives it a stratigraphic importance that can hardly be over-emphasized; yet as a formation it was completely missed by both McKay and Thomson. McKay's failure is unexplained. Thomson was too acute an observer to miss the distinctive sulphurous efflorescence. He collected it from Nidd River, had it analysed, and quoted the analysis in his description of the Saurian Sands of Waipara (1920, p. 343). Australis Zone Sandstone with Inoceramus australis and I. cf. pacificus forms a well defined ridge that can be traced for about half-a-mile immediately south of Burnt Saddle. The sandstone has been mapped as extending west to a fault in the south branch of Wharfe Stream, but no fossils were found in the western part. Similar sandstone 1,000ft. north of the middle reaches of Cover Stream contains I. cf. pacificus and is probably the source of boulders with I. australis in Cover Stream. To the north a band of sandstone with I. australis extends along Nidd River, mostly on the south side. The band appears to be overlain by Whangai Shale, but it is crushed and no clear contacts were seen. Still farther north similar sandstone with both I. australis and I. cf. pacificus crops out in the headwaters of Nidd Stream between two ridges of limestone.

The sandstone dips west and underlies Whangai Shale, which in turn underlies the western limestone ridge. It appears to be faulted against the eastern ridge of limestone. The Australis zone is either absent or thin and unfossiliferous in the western part of the Coverham district. It appears to be absent from the remainder of the Clarence Valley. It occurs as boulders in Kekerengu River, and is probably being washed down from Benmore Stream. At several places in the North Island the Australis zone with one or both of the two species of Inoceramus occurs immediately below sediments with Rzehakina epigona. No foraminifera have yet been obtained from the sandstone itself, and it either forms the base of the Rzehakina zone or more probably conformably underlies it. Its age is probably Campanian (upper Senonian). A Zone The Australis sandstone of Burnt Saddle is directly underlain to the southeast by a dark sulphurous micaceous siltstone with abundant bands crammed with a distinctive but undescribed Inoceramus, referred to here as Inoceramus sp. A. This fossil defines zone A and its distribution is marked on the map by the figure “3”. As well as on the track south-east of Burnt Saddle (Gs 5816-7) it occurs in the south branch of Wharfe Stream, immediately below the Whangai at the junction of the main tributary from the south and on the opposite side of the syncline in the landslide scar below the track to the north (Gs 6110). It also occurs in the upper part of Cover Stream (Gs 6097) and in the middle reaches of Nidd River. Outside the mapped area it is known from gritty greensand immediately below the sulphur sands of Bluff Stream (Gs 6149) and from dark siltstone at some distance below Whangai siltstone in the upper part of Kekerengu River just within the forest (Gs 6144). It is known from several localities in the North Island, and although not found in sequence there, it is closely associated with, and almost certainly not far below, the Whangai Shale. At most places the fossil occurs in dark micaceous siltstone, like that south of Burnt Saddle, but it occurs as well in concretionary conglomerate in Wharfe Stream (Gs 5821) and in Cover Stream (Gs 6097). A boulder of fossiliferous concretionary conglomerate with abundant well preserved specimens of this fossil with the locality label “Wharfe Stream above pack-track” in Thomson's handwriting was recently found in the Geological Survey collection. The thickness of the zone appears to be very great on the track near Burnt Saddle, but the stream exposures show numerous folds and the thickness may not be more than 500ft., but it can hardly be less. In the north branch of Wharfe Stream the full thickness of the zone cannot be more than 20ft. or 30ft. The thickness in Cover Stream is uncertain, but probably considerably greater than 50ft. B. Zone This zone is marked by an undescribed species of Inoceramus that is probably the most abundant and widespread in New Zealand. It is referred to here as Inoceramus sp. B, and its distribution marked on the map with the figure “4”. Its stratigraphic position is most clearly shown in the south-flowing tributary of Nidd River, 2,500ft. east of Sawpit Gully, where it underlies the Whangai Shale and directly overlies sandstone with I. bicorrugatus. It also occurs at several localities in Nidd River between this tributary and Sawpit Gully (Gs 6127) on

the old track 4,000ft. east of Coverham, in the main Wharfe Stream in the landslip scar below the track, and in the south branch of Wharfe Stream about 1,000ft. upstream from the junction of the two branches (Gs 5808-9). Its stratigraphic position in the North Island is clearly shown in Te Rata, Wairongamai, and Puketoro streams where it directly overlies the sandstone with I. bicorrugatus, and underlies the Australis Zone and the Whangai Shale. The sediments in the Clarence Valley are medium-grained sandstone with interbedded siltstone bands that are less sulphurous and less micaceous than those of Zone A. The thickness of the zone ranges from 30ft. in the landslip scar below the track to 200ft. in the tributary of Nidd River. Bicorrugatus Zone This zone is defined by Inoceramus bicorrugatus Marwick and by an allied species slightly higher in the sequence The distribution of the two species is represented on the map by the figures “5” and “(5)”, respectively. I. bicorrugatus was described by Marwick (1926) from a fragment from Mangaotane River collected by Ongley and Macpherson during their survey of the Waiapu Subdivision. It was shown by Marwick to match Thomson's large Inoceramus from the Nidd Sandstone. Marwick wrongly assumed from Thomson's stratigraphy and Woods' age determination that the Nidd Sandstone is Albian or, at its youngest, Cenomanian in age. The Nidd Sandstone is actually the youngest of the formations recognized by Thomson, and from stratigraphy alone the fossil is likely to be considerably younger than Albian. The species was originally described by Marwick from a fragment of the outer part of the shell. More complete material since collected from Mangaotane River and Puketoro Stream show that the shell at a particular stage in its life history suddenly changes its growth direction and ornamentation. The young shell had previously been considered as a distinct species, but had not been described. This sudden growth change and the shape of the young shell suggest that I. bicorrugatus is related to some shells of the group Inoceramus inconstans Woods (1912) that range in age from upper Turonian to lower Santonian. The growth change makes the correlation with I. andinus Wilkens suggested by Heinz (1928) unlikely. In addition to the locality at the mouth of Sawpit Gully (Gs 5818) from which it was recorded by Marwick, I. bicorrugatus also occurs at two other localities in Nidd River (Gs 6107). To the south it is well exposed in the north branch (Gs 6113), and in the south branch of Wharfe Stream (Gs 6104). An important locality is in the centre of the syncline at the junction of a tributary with the main branch of Wharfe Stream. The most distinctive sediment of the Bicorrugatus Zone is a slightly glauconitic gritty sandstone, but the zone also includes fine sandstone and dark siltstone. At Sawpit Gully the Bicorrugatus Zone is underlain by sandstone with belemnites and fragments of a huge Inoceramus. The fragments are up to half an inch thick and similar to fragments that occur immediately below I. bicorrugatus at several North Island localities and which are sufficiently widespread and distinctive to form a potential zone fossil. In all other sections the Bicorrugatus Zone appears to grade down into the Porrectus Zone, but the absence of the huge Inoceramus suggests that the gradation hides a minor erosion interval.

Porrectus Zone This zone is defined by an Inoceramus that was named I. concentricus var. porrectus by Woods (1917: 10). To the writer the fossil resembles Inoceramus lamarki var. apicalis Woods more closely than it does I. concentricus, and it is proposed to re-name the fossil I. porrectus. Its distribution in the Coverham district is shown by the figure “6” on the geological map. In addition to the upper part of the mudstone at Sawpit Gully, where it was first described, it occurs on the west side of Nidd River near the mouth of Sawpit Gully (Gs 6094), in the middle reaches of Cover Stream (Gs 6095 and Gs 6096), in lower Wharfe Stream on the north-east side of the syncline (Gs 6119), and in the main stream just below the junction on the south side of the syncline (Gs 6100). It also occurs in the south branch of Wharfe Stream. At all places, except at Sawpit Gully, it occurs immediately below the Bicorrugatus Zone and its distribution is thus an important part of the evidence for a major unconformity in the Sawpit Gully section. The variety of fossils in the Porrectus Zone is greater than that in any of the other zones near Coverham. An ammonite collected by Thomson from Sawpit Gully was described by Woods as Gaudryceras sacya (Forbes) and considered to indicate an Albian or Cenomanian age. In 1926 Marshall described an upper Senonian or Maestrichtian ammonite that he had collected from Northland as G. subsacya, and considered the Sawpit Gully ammonite to belong to the same species and to be appreciably different from the Cenomanian? G. sacya. The resemblance between the Sawpit Gully and Northland specimens is very close, and it seems evident that the species G. subsacya evolved slowly and is of no use for close dating. Aucellina euglypha Woods is probably confined to the Porrectus Zone. It has not been collected from Sawpit Gully but according to Thomson (p. 315) occurs in Swale River not far below the base of the flint beds. It is abundant in the lower part of Wharfe Stream from just below the base of the Bicorrugatus sandstone to several hundred feet below. Aucellina euglypha was also collected with I. porrectus at several places farther up Clarence Valley. The Porrectus beds appear to represent the upper limit of Aucellina in New Zealand. Four different kinds of belemnites were collected from the Coverham district, but the systematic importance of the differences are as yet uncertain. Woods (1917) described and illustrated Belemnites superstes Hector from Wharfe mudstone on the left bank to Wharfe Stream half a mile downstream from the packtrack. This locality is within the Porrectus Zone, but he mentions that the species also occurs in Cover Stream (Concentricus Zone) and in the upper Clarence and Awatere valleys. The origin of the type of B. superstes is probably from the Porrectus beds of Swale River. In 1886 (p. xxxii) Hector mentioned that B. superstes occurs in contorted sandstone with volcanic dykes immediately below flint beds. The reference to flint beds makes it likely that the locality is below the flint beds of the Chalk Range and the reference to dykes points to Swale River, where a dyke was reported by Thomson. The writer collected several belemnite fragments from the bare southern slopes of Wharfe Stream near the mouth of a north-flowing tributary half-a-mile from the upper end of the Wharfe Gorge. They are considerably flattened in a dorso-ventral direction, the average ratio of the two diameters being 1 to 1.4, which is slightly more than that of two of the specimens illustrated by Woods

(Pl. V, Figs. 5 and 6). The compression in the third specimen illustrated by Woods is considerably less. The belemnite in the Porrectus Zone is recorded on the stratigraphic column (Fig. 5) as “Belemnites sp. B.” The huge ammonite found on the south bank of Wharfe Stream 2,000ft. upstream from the gorge may be useful for correlation. It is similar to one from Bushgrove Stream in the Wairarapa district considered by Spath (1935) to be allied to Parapuzosia haughtoni and to be Senonian in age. The Bushgrove Stream ammonite is closely associated with an Inoceramus and with an Aucellina like that at Wharfe Stream. The Inoceramus is distinctive and close to I. labiatus (Schloth), a widely distributed lower Turonian species which, although absent from near Coverham, has been collected with Aucellina euglypha from several localities in the middle Clarence and upper Awatere valleys. In view of the stratigraphic and fossil evidence the huge ammonite is probably Turonian in age, and the name Parapuzosia aff. haughtoni may not be appropriate.* Mr. C. W. Wright, of London, after a preliminary examination, considers the huge ammonite from Bushgrove Stream as indeterminate, and the one from Wharfe Stream as Pachydesmoceras sp. possibly Turonian. Within the Coverham district the Porrectus beds consist of medium blue-grey mudstone with occasional bands of concretionary sandstone. Calcareous concretions are common, and up to 6ft. in diameter. Many are oblate spheroids with a reticulate pattern on the outside. McKay (1886: 90) shows a “dyke or boss of intrusive rock” on the south side of Wharfe Stream in his cross-section of the Coverham district. The rock contains large calcite-filled cavities and is probably a flow and hence the same age as the Porrectus mudstone. It is of interest that as far as can be judged from the associated fossils the volcanics that extend over a large area in the middle Clarence and middle Awatere valleys are either Porrectus Zone or basal Raukumara in age. The relation between the Porrectus and the underlying Concentricus Zone is somewhat uncertain. The two fossils are nowhere close to each other and are separated by a band of mudstone that contains only unidentifiable fragments of Inoceramus. Concentricus Zone This zone is defined by the presence of Inoceramus concentricus. Concentricus has a very restricted distribution in New Zealand and is known only from the lower part of Cover Stream near Coverham. Well preserved specimens are confined to calcareous concretions and are common only near the low saddle where the track crosses over into Nidd River (Gs 5815) at the point marked by the west end of the 60° dip symbol on Fig. 6. The other important fossil is the ammonite Turrilites circumtaeniatus (Gs 4707), collected by Thomson from Cover Stream “200 yards above the old sheep dip”. The ammonite appears to come from a horizon slightly higher than the typical specimens of I. concentricus. With the ammonite Thomson collected a belemnite that he described as B. superstes Thomson's specimen and one collected later are decidedly less compressed than the belemnites from the overlying Porrectus Zone and are recorded on the stratigraphic column (Fig. 5) as “Belemnites sp. C.”. From the same locality Thomson also collected the carapace of a crab, a small compound coral, the skeleton of a fish (Diplomystus coverhamensis Chapman), numerous fish scales and specimens of fossil wood. With the exception of the

fish, none of these fossils seem to have been further described, and the present whereabouts of some is uncertain. The writer collected the stem and leaves of what may be a cycad (B470) from a concretion in a gully around which the track passes to the south of Coverham. A sample (L 485) containing numerous spores and pollens was collected by Mr. Couper from Thomson's locality. The identification of these plant fossils is of great importance, for they make it possible to relate the comparatively well dated marine sequence of Cretaceous rocks with the non-marine coal-bearing sequence of the west coast of the South Island Mr. Couper has supplied the following list and comments:— Osmundacidites wellmanii Couper. Podocarpidites major Couper Cyathidites australis Couper. Podocarpidites cf. ellipticus Cookson Microcachryidites antarcticus Cookson Cyathidites minor Couper Araucariacites cf. australis Cookson Trilites verrucatus Couper Hymenophyllum sp. Lycopodium fastigioides Couper Tricolpites? sp. Cookson ex Couper “The first five fossils listed have a relatively restricted range and are useful for dating. Species 1, 2 and 3 first appear in the Bajocian (?) beds of Waimahaka in Southland and become extinct in the lower Paparoa beds of Greymouth, which are considered Senonian in age. Species 4 first appears in the Neocomian beds of Waikato Head and becomes extinct in the upper Eocene. Species 5 first appears in the Topfer Formation of Reefton, possibly Aptian, and becomes extinct in the upper Eocene Angiosperm pollen is extremely rare, of the 500 or so grams examined only one was from an angiosperm. It is dicotyledonous and is tentatively identified as Tricolpites sp.” The Concentricus mudstone is not noticeably different from the overlying Porrectus mudstone. It contains similar calcareous concretions and similar bands of concretionary sandstone. It was described by Thomson as being more generally fossiliferous than any of the other Coverham formations, but it did not seem particularly fossiliferous to the writer. Anglicus? Zone An Inoceramus with fairly regular concentric ribs separated by wide interspaces that resembles the middle Albian I. anglicus Woods marks this zone. The fossil is known only from dark mudstone immediately overlying the Wharfe Gorge Sandstone at Ouse River (Gs 5814) and was mentioned by Thomson (1919, p. 313) as the “large flat Inoceramus” in “the lowest Wharfe beds”. The thickness of this zone and its exact relation to the Concentricus Zone is uncertain. A similar Inoceramus (Gs 5995) was collected by the writer in the North Island from dark mudstone in the headwaters of Gorge Stream, a tributary of Mangaotane River, a few hundred feet above massive Taitai-like sandstone that forms the prominent mountain Arowhana.

Wharfe Gorge Sandstone The sandstone forms a narrow but not difficult gorge in the lower part of Wharfe Stream. It is about 1,000ft. thick at Wharfe Stream, but thins to the west, and is only a few hundred feet thick at Ouse River. Belemnites and fossil wood were recorded by Thomson and were the only fossils seen by the writer. The belemnites (Gs 6126) were collected from the base of a sandstone band near the lower end of the gorge. They are smaller and more elongated than those in the overlying Concentricus Zone and less compressed than those of the Porrectus Zone, and are recorded on the stratigraphic column (Fig. 5) as “Belemnites sp. D.” The fossil wood has been sectioned but is so poorly preserved as to be unidentifiable. C. Zone Mudstone that directly underlies the Wharfe Gorge Sandstone and forms the northern part of Thomson's Wharf Mudstone is exposed on the north side of Wharfe Stream for a quarter-of-a-mile upstream from the gorge. It contains many unidentifiable plant fragments and abundant carbonaceous material, and is much darker than the other Coverham mudstones. Its strong compaction and close jointing makes the extraction of whole fossils difficult. The only fossils are the unidentifiable plant fragments and an Inoceramus that was collected from two localities in Wharfe Stream (Gs 6125 and Gs 5812). The Inoceramus is recorded as I. sp. C. and its position shown on the map by the figure 9. A similar Inoceramus was collected by the writer from several localities at Tapuwaeroa Valley in the North Island from dark mudstone below the Taitai Sandstone. The Inoceramus is not well preserved and no overseas correlative is known, but as it lies conformably below the Taitai Sandstone (Marwick, 1939) it is probably either upper Aptian or slightly older. At Wharfe Stream the lower part of the mudstone is cut off by a fault, a total thickness of 500ft. being exposed. The details of a more complete section exposed in Ouse River have been supplied by Mr. Grindley. Section Exposed in Upper Part of Ouse River. Feet. Alternating bands of thick mudstone and thin sandstone 2,000 Siltstone with rare pebbles Inoceramus cf. anglicus (Gs 5814) 20 Graded bedded sandstone and mudstone (Wharfe Gorge Sandstone) 270 Dark mudstone with thin streaks of sandstone, Inoceramus frags. 1,000 Conglomerate 10 Thin bedded mudstone and sandstone (pre-Notocene of Thomson) 10,000 The C Zone is represented by 1,000ft. of dark mudstone below the Wharfe Gorge Sandstone. Thomson interpreted the rare pebbles at the base of the Anglicus? Zone as being the base of the Clarentian. Mr. Grindley says that the underlying sandstone is very like the Wharfe Gorge Sandstone and the dark mudstone below very like the C Zone mudstone of Wharfe Stream. The stratigraphic relation between the section at Ouse River and that in the branch stream which contains the “basal conglomerate” of Thomson is uncertain.

“Basal Conglomerate” The base of the Clarentian beds was not examined at Coverham, and the following description of the conglomerate in the largest north flowing tributary of Wharfe Stream is taken from Thomson (p. 312). The pebbles are set in a mudstone matrix, are well rounded, and up to a few inches in diameter. They are composed of sandstone and quartzite (greywacke), of white, bright red, and pink jasperoid quartz, and rare granite and porphyry. Dark crystalline rock, schist, and limestone are described as being absent. Thomson asserts that the conglomerate is basal, resting unconformably on significantly older greywacke and argillite. These observations are supported by members of the Geological Survey who visited the contact in this end in an adjoining stream in 1953. The distinction between Clarentian and pre-Clarentian is not everywhere easily made. Thomson admits that he is unable to tell whether the conglomerate bands in Ouse River are Clarentian or pre-Clarentian. Farther up the valley a clear section exposed at a bend in Clarence River between Dart River and Ravine Stream shows:— Section at Bank Clarence River (S 42. 984317). Feet. Dark mudstone, abundant concretions 500 Alternating graded bedded sandstone and siltstone 150 Bands of well sorted Clarentian-like sandstone … 70 Finely alternating beds …. 130 Hard gritty sandstone band 20 Gritty sandstone, coal frags, calcareous conglomerate 30 Gritty sandstone with many pebbles; congretions and jasper, etc. 100 Gritty sandstone with concretionary bands and pebbles 50 Gritty greywacke with bands of argillite and coal fragments 300 Dark argillite and greywacke, very irregular bands, concretions and rate fragments of wood 500 The top of the above section is certainly Clarentian and the base equally certainly pre-Clarentian in the sense used by Thomson, but the writer could find no trace of a break that is sufficiently clear-cut to convince him that the lower part of the section is pre-Cretaceous or even significantly older than the upper part. The base of the Clarentian is equally difficult to define at Kekerengu River. Macpherson (1952: 263) has mapped in descending order: Woolshed Shale, Good Creek Sandstone, and Whernside Group. The Woodshed Shale is Piripauan in age. The Good Creek Sandstone was considered by Macpherson on lithology to be Tapuwaeroan in age but it contains Inoceramus cf. porrectus and an elongated Inoceramus like labiatus, and is probably part of the Porrectus Zone. It dips steeply east, and this was accepted as the true dip by Macpherson, but numerous graded sandstone bands all indicate that the beds are slightly overturned, the top facing north-west (upstream). Except that it contains conglomerate bands and is apparently unfossiliferous, the Whernside Group did not appear to the writer to differ appreciably from the Good Creek Sandstone. It was considered by Macpherson to resemble and to be of the same age as the Raukumara Series. Graded bedding shows that the tops of individual beds face upstream. The formation was traced without noticeable change for a mile to Remuera Station, and then for two miles farther upstream. Finally, after crossing a crush zone that probably represents a major fault, Raukumara Sandstone with Inoceramus sp. A. was encountered just within the boundary of the forest. The structure is evidently much more complex than it was considered to be by Macpherson. The way in which the tops of the beds

face north-west has much in common with the Coverham section (Fig. 3) where the bulk of the beds have the same attitude, the continuity being interrupted by several faults. The apparent thickness of the Kekerengu Section is so great as to make it almost certain that it, too, is faulted. Conglomerate bands are conspicuous in the Whernside Formation and up to 100ft. thick. They strike from Remuera to Whernside Saddle and doubtless continue south-west to connect with the conglomerate bands described by Thomson in his pre-Clarentian greywacke. The bulk of the Whernside Formation is probably older than any of the Clarence Series of Coverham. To the writer it resembles the Taitai sandstone and conglomerate more than any other beds that he has seen. Although Thomson visited Coverham by pack-track from Kekerengu at least twice and must have had ample time to examine the rocks on the way, he makes no mention of them in his text and in his map of eastern Marlborough his formation boundaries discreetly fade out before reaching the valley of Kekerengu River. To the writer it seems likely that he had as much trouble in deciding if they are Clarentian or pre-Clarentian as other geologists have had since. Foraminifera in Clarence Section Foraminiferal samples were collected from the type Clarentian section near Coverham by Dr. B. H. Mason, in 1946. Reasonably abundant foraminifera were obtained from the upper part of the Cover Stream Mudstone, from the base of the Nidd Sandstone, and from the whole of the Sawpit Gully Mudstone. These samples were briefly examined by Dr. H. J. Finlay, and have been re-examined by Mr. N. de B. Hornibrook, who supplied the following lists and description. “The Sawpit Gully and Cover Creek mudstones contain one fauna and the Nidd Sandstone another. The two faunas are so different from the Whangai faunas that they have not been studied with the attention they deserve. “The Sawpit Gully fauna contains:— Lagenidae, small and distinctive. Anomalinidae, small and distinctive. Gyroidina, tiny and like G. nitida Reuss. “Bulimina” tiny. Globigerina cretacea d'Orb. Karrerulina clarentia Finlay. “A similar fauna with a large number of species, including Globigerina aspera (Ehrenberg) was found in mudstone below Whangai Shale in Tangaruhe Stream that was mapped by Dr. A. R. Lillie in the Dannevirke Subdivision as Raukumara and part of the Clarence Series. “The Nidd fauna contains: Ammobaculites sp., and other large coarsegrained arenaceous foraminifera: Lagenidae, moderately large Glomospira charoides (Jones and Parker) Ramulina sp. Gyroidina aff. globosa (Hagenow) “Anomalina” sp.

“The difference between this and the Sawpit Gully fauna may be due to facies but is more probably due to a difference in age. The Puketoro fauna, from mudstone of the Raukumara Series just below the Whangai Shale, is similar and contains:— Ammobaculites sp. (as above) Glomospira charoides Ramulina sp. (as above) “Anomalina” sp. (as above) Globigerina cretacea Gaudryinella cf. delrioensis “A similar fauna with Gaudryinella is also known from Northland (N15f584), and from the Dannevirke Subdivision (Al 88 = Gs 2410). The Sawpit Gully, Nidd, and similar faunas with Globigerina cretacea were considered by Finlay to be part of a single faunal unit for which he used the names Clarentian, Raukumara, and Puketoro, as best served the occasion. The re-examination of the Clarentian section has made it likely that the Nidd fauna is younger than the Sawpit Gully one. In terms of the series divisions proposed by Wellman in this paper the Sawpit Gully fauna represents the middle and upper Clarence Series, and the Nidd fauna the middle part of the Raukumara Series.” Each of the two micro-faunas is associated with a different species of Inoceramus, and the approximate age of each is known. The “Sawpit Gully” fauna of Coverham and Tangaruhe Stream is associated with I. porrectus and is upper Clarence Series in age. The Nidd fauna of Coverham, Puketoro Stream, Northland, and Dannevirke is associated with I. sp. B. and is middle Raukumara Series in age. Fig. 4.—Cross section at 330° along north west flowing tributary of Wharfe Stream and through mouth of Sawpit Gully to Chalk Range. Horizontal and vertical scales are the same. Intervals of 1,000ft. marked on right. C = Clarence; R = Raukumara; T? = probable Taitai.

Structure Although the strata are so strongly deformed that the cross section (Fig. 4) gives only a very generalized picture, the pattern of deformation appears to be reasonably consistent, and of the same general character throughout the Clarence Valley, at least upstream to Bluff River. The strata are strongly folded, with fold axes that dip north-west and only extend a few miles before being faulted off. In plan the faults and fold axes, although sinuous, are almost parallel. The two synclines near Coverham have already been described; both plunge 10°–20° north-east and have axial planes that dip north-west at about 60°. Visible anticlinal folds are conspicuously absent. It seems likely that they have been destroyed by faulting. The faults are less obvious than the folds. The strata are severely crushed near the major faults and no major fault planes were clearly seen. The faults are conveniently described under: (1) active faults expressed by surface traces, and (2) faults inferred from structure and stratigraphy. (1) The largest fault trace extends from near the mouth of Sawpit Gully east-north-east near the crest of the ridge between Nidd River and Cover Stream. For the first mile the trace is a conspicuous trench; it then becomes less definite and appears to die out after two miles. A fairly well-defined trace branches off to the north-east half-a-mile from Sawpit Gully, and is considered to follow through a strike valley to rejoin the major fault south of the Nidd landslide. In the opposite direction the trench cannot be traced across the steep banks of Nidd River, but it re-appears in the flattish uplands to the west of the river where it swings to the south and gives off two branches as it swings. A short but definite fault trace interrupts a Recent gravel fan at the lower end of Nidd Swamp. The strike of the fault continues along the line of Nidd River where crushed rocks have already been mentioned. It seems likely that active faulting is taking place along the Nidd River and that erosion, except at the gravel fan, is too rapid for a fault-trace to form. Although it cannot be proved at all places, it is probable that considerable displacement of the Cretaceous strata has taken place along the line of all the active faults. The main fault-trace east of Nidd River coincides with the boundary between sandstone to the north and mudstone to the south. Beyond the north-east branch fault the sandstone contains Inoceramus cf. pacificus and is part of the Australis zone. At the west end of the trench no fossils were found at the fault itself, but as I. bicorrugatus crops out a few hundred feet away in similar sandstone it has been assumed that the sandstone at the fault is part of the Raukumara Series. The fault traces to the east lie within poorly fossiliferous mudstone where a large displacement would not be evident. Although the changes in lithology at the trench suggest considerable displacement they do not make the direction of displacement certain. The relation of the Australis sandstone to the adjoining beds is particularly difficult to explain. To the writer it seems most likely that the strata have been cut up into slivers and shifted laterally by large clock-wise-transcurrent displacements. The only evidence near Coverham is the horizontal slickensides seen at several minor faults. The regional evidence is more conclusive. The fault traces at Coverham are parallel to fault traces at Awatere and Kekerengu valleys. From topographic evidence it is certain that the Awatere and Kekerengu faults are essentially transcurrent (Wellman, 1952) and as these faults lie one on each side of Coverham it is almost certain

that the Coverham faults, being parallel, are transcurrent also. In view of the rapid variations in thickness of the different zones and formations, definite evidence for transcurrent movement might be determined from detailed mapping. The possible transcurrent movement cannot be represented by displacement on the cross section, and the fault that has caused the main fault trace is shown there as a reverse fault with a dip parallel to the fold axes. The true dip is not known. (2) The three inferred faults are marked with roman numerals on the map. Inferred Fault I is shown crossing the lower part of Sawpit Gully and extending east-north-east parallel to the base of the flint beds. The evidence for this fault is purely stratigraphic. On the north side of the inferred fault graded-bedded sandstones interbedded with the Sawpit Gully Mudstone dip north-east at 70° and are not overturned. They are separated from the Porrectus beds at the base of the flint beds by about 3,000ft. of strata that dip north-west with about the same slope, hence the sandstone bands are probably 3,000ft. lower than the Porrectus horizon. Judged from the section exposed in the stream half-a-mile east of Sawpit Gully the strata on the south side of the inferred fault are less than 1,000ft. below the Porrectus horizon and are overturned. The hypothetical fault has thus been placed at an anticlinal axis, the north side of which is considered to have been upthrown about 2,000ft. The evidence for Inferred Fault II is both structural and stratigraphic. The fault is considered to cross a saddle between Ouse River and Wharfe Stream, to follow the lower part of Wharfe Stream, and to swing north-east through the saddle at the east end of the ridge formed by the Wharfe Gorge Sandstone, and to continue north-east across Cover Stream to join with the main active fault. Structurally the fault is needed to explain the abrupt ending of the Wharfe Gorge Sandstone to the east, and to cut off the anticline in the upper part of Cover Stream. The throw of the fault is probably greatest in Wharfe Stream. The dark mudstone on the north side of the fault is at least 5,000ft. below the base of the Bicorrugatus Zone, whereas the Porrectus mudstone on the south side of the fault can be no more than a thousand feet below that zone. The throw is thus about 4,000ft. It will be noticed that at the line of the cross section, Fault II also corresponds with an anticlinal axis. Inferred Fault III is less certain than the other two. It is considered to cross the south branch of Wharfe Stream not far from the junction and to follow the south side of the north branch to cross into the valley of Kekerengu River by the saddle near Whernside Hut. The direct or close contact between the Porrectus mudstone in the north branch of Wharfe Stream and the Whangai Shale on the track on the south provides the best evidence for the fault. The northern is the upthrown side, as in the other two hypothetical faults. The structure is complex and a full interpretation would require a detailed examination of the ridges as well as of the stream section. This was not attempted and the structural interpretation given must be considered as no more than a simplified version of the truth. The writer considers it extremely improbable that the structure is less complex than that shown. Thomson was fully aware of the presence of limestone at Whernside and must have realized the complexities of the regional structure when he drew his “Map of eastern Marlborough” (p. 293). He was also aware of complex minor

folds in the middle (opp. p. 316) and upper Swale River (p. 349), yet in a spirit of optimism that was followed by faint misgivings he wrote: “The thicknesses given above for the various divisions of the Clarentian at Coverham were estimated, except in the case of the conglomerates, by measurements from a section drawn to true scale (sic). An average dip of 55° was allowed, but it will be observed that the dip is often steeper and seldom less than the figure. The reversals due to folding are unimportant, but have been allowed for. An almost continuous section of beds has been observed, and no faults of any consequence were seen. Consequently unless there is a very strong unconformity between the Clarentian and the Amuri Limestone and a repetition of the beds by closely adpressed folds, there is no escape from the conclusion that the thickness given is approximately correct.” The thickness is. of course, much less than it was considered to be by Thomson. Mode and Direction of Origin of Clarentian Sediments Thomson (1919, p. 315) was of the opinion that the whole of the Cretaceous rocks of the Coverham district were “deposited as the topset beds of a continental shelf undergoing rapid depression and near the mouth of a large river, the sands and gravel of which were arrested nearer shore, or up the estuary by its drowning”. It is now known (Kuenen and Migliorini, 1950) that turbidity currents are capable of transporting sands and conglomerates beyond the continental shelf, and there is thus no reason to suppose with Thomson that the sand and conglomerate bands are evidence for near-shore deposition. It is generally accepted that many graded bedded sandstones were transported by turbidity currents. Almost all the sandstone bands within the Clarentian (as defined in this paper) show graded bedding. The sandstone bands in the overlying Raukumara beds do not show this feature. Graded bedding was used during the revision to find the true order of superposition, which is indicated on the dip symbol by an arrowhead pointing towards the younger sediments. Many sandstone bands filled grooves that had been cut in the underlying mudstone, presumably by the turbidity currents that transported the sandstone. Being less resistant than the sandstone the underlying mudstone has been eroded away to expose the counterpart of the grooves at the base of the sandstone bands as sandstone tongues. The tongues are an inch or so long, and they narrow to a point in one direction. Within each sandstone band the tongues point in the same direction to within 5°, and it seems clear that this direction is important for indicating the direction of turbidity current flow. All the directions measured within the Clarence Valley are given in the table below. The dips are given in column 1, the one overturned dip being given a value greater than 90°. The sandstone casts invariably lie at the base of the sandstone bands. When they are viewed in a direction at right angles to their strike, the observer is looking in the direction of decreasing age of the sediments. This direction, which can be referred to as that which the beds face, is given by column 2. The position angle, column 3, gives the direction that the sandstone tongues point when the sandstone band is viewed as the face of a clock, the direction being given in degrees measured in a counter-clockwise direction from the 6 o'clock position. If, for example, the tongues point downwards the position angle is 0°, if the tongues are horizontal and point to the right the angle is 90°, if they

point upwards the angle is 180°, and if horizontal and pointing to the left the angle is 270°. The advantage of this method of measuring the position angle is that the original direction in which the tongues pointed before the beds were folded can be obtained merely by adding the position angle to the facing direction, and subtracting 360° when necessary. The unfolding process and the reason for measuring in a counter-clockwise direction is best shown as follows:— If a book with one edge resting on a table be used to represent a sandstone band and the position angle of the sandstone tongues be marked on the back cover (to correspond to the base of the beds), the original direction is inferred to be the direction in which the tongues point when the book is swung around its lower edge so that the back cover with the position angle marked on it lies flat in contact with the table. It will be shown later that the tongues probably point in the direction of origin of the turbidity currents, and this direction as determined by the mode of unfolding outlined above is given in column 4 of the table above. Table Giving Direction of Casts at Base of Graded-Bedded Sandstones. Locality of Sandstone Band (1) Dip. (2) Facing Direction = Direction of Decreasing Age. (3) Position Angle. (4) Inferred Direction of Origin. Sawpit Gully, in mudstone 60 340 10 350 Wharfe Stream Gorge 70 340 150 130 Wharfe Stream Gorge 70 340 40 20 Wharfe Stream Gorge 70 340 0 340 Wharfe Stream, in dark mudstone 90 350 180 170 Wharfe Stream, near track crossing 80 320 180 140 Dee River, 300ft. below flint beds 100 50 350 40 Track north side Branch River 20 260 290 190 Ravine Stream 70 315 270 225 Muzzle River, near base Clarentian 80 320 240 200 From the table it will be seen that the sandstone tongues point in two quadrants only—from north-west to north-east and from south-east to south-west. From the map (Fig. 2) it will be seen that they show a reasonably geographic orientation: on the north side of a north-east line just north of Clarence River they point north; and on the south side of this line, south. From this it is inferred that the tongues point in the direction of origin of the turbidity currents and not in their direction of flow. References Cotton, C. A., 1913. The Physiography of the Middle Clarence Valley, New Zealand. Geogr. J., 42: 225–46. —— 1945. “Earth Beneath “. Whitcombe & Tombs, Wellington. —— 1950. Tectonic Scarps and Fault Valleys. Bull. Geol. Soc. Amer., 61:- 717–57. Finlay, H. J., 1939. New Zealand Foraminifera: The Occurrence of Rzehakina, Hantkenina, Rotaliatina, and Zeauvigerina. Trans. Roy. Soc. N.Z. 68: 534–43. Finlay, H. J. and Marwick, J., 1948. Cretaceous. In: “The Outline of the Geology of New land” (Nz. Geol. Surv.) Glaessner, M. F., 1945. “Principles of Micropalaeontology.” Melb. Univ. Press. Hector, J., 1886. Progress Report, 1885. N. Z. Geol. Surv. Rep. Geol. Explor. 1885, 17: ix-xl. Heinz, R., 1928 Uber die Oberkreide-Inoceramen Neu Seelands und Neu-Kaledoniens. Mitt. Min.-Geol. Staatsinst., 10: 111–30.

Kuenen, P. H., and Migliorini, C. L., 1950. Turbidity Currents as a Cause of Graded Bedding. J. Geol., 58: 91–127. McKay, A., 1886. On the Geology of the Eastern Part of Marlborough Provincial District. N.Z. Geol. Surv. Rep. Geol. Explor. 1885,17: 27–136. Macpherson, E. O., 1951. The Stratigraphy and Bentonitic Shale Deposits of Kekerengu and Blue Slip, Marlborough N.Z. J. Sci. Tech., B, 33 (4): 258–86. Marwick, J, 1926. Cretaceous Fossils from Waiapu Subdivision. N.Z. J. Sci Tech., 8: 379–82. —— 1939 Maccoyella and Aucella in the Taitai Series. Trans. Roy. Soc. 68: 462–5. —— 1950. The Type of the Ammonite Madrasites mckayi (Hector). Trans. Roy. Soc. N.Z. 78. 482–4. Ongley, M. and Macpherson. E. O., 1928. The Geology of the Waiapu Subdivision, Raukumara Division. N.Z. Geol. Surv. Bull. 30 (n.s.). Quennell, A. M., and Brown, D. A., 1937. Dannevirke Subdivision. N.Z. Geol. Surv. 31st Annu. Rep. (n s.) : 1–6. Spath, F., 1935. In: Marwick, J., Palaeontological Report, N.Z. Geol. Surv. 29th Annu. Rep. (n.s.): 11. Thomson, J. A., 1916. The Flint-beds associated with the Amuri Limestone of Marlborough. Trans. N.Z. Inst., 48: 48–58. —— 1917. Diastrophic and other Considerations in Classification and Correlation, and the Existence of Minor Diastrophic Districts in the Notocene. Trans. N.Z. Inst., 49: 397–413. —— 1919. The Geology of the Middle Clarence and Ure Valleys, East Marlborough, New Zealand. Trans. N.Z. Inst. 51: 289–349. —— 1920. Notocene Geology of the Middle Waipara and Weka Pass District, North Canterbury. Trans. N. Z. Inst. 52: 322–415. Wellman, H. W., 1952. Data for the Study of Recent and late Pleistocene Faulting in the South Island of New Zealand. N.Z. Journ. Sci. Tech., B, 34 (4): 270–88. Wilckens, O., 1922. The Upper Cretaceous Gastropods of New Zealand. N.Z. Geol. Surv. Pal. Bull. 9. Woods, H., 1911. “A Monograph of the Cretaceous Lamelhbranchia of England”, 2 (7). Palaeont. Soc., London. —— 1912. “A Monograph of the Cretaceous Lamellibranchia of England”, 2 (8). Palaeont. Soc., London. —— 1917. The Cretaceous Faunas of the North-Eastern Part of the South Island of New Zealand. N.Z. Geol. Surv. Pal. Bull. 4. H. W. Wellman Geological Survey 156 The Terrace Wellington

Series Formation or Zone Lithology and Thickness Important Fossils Age Arnold & Landon Weka Pass and Amuri limestones Sandy limestone glauconitic at base. Fine grained limestone, flinty at base. c 2,500 feet. Numerous foraminifera Oligocene. Tertiary Danne Virke Flint beds Flint beds, calcareous at top, sulphurous and carbonaceous at base 1000 to 2000 feet. No fossils found. Eocene Paleocene Mata Whangai shale Flinty sulphurous siltstone, dark when fresh rusty weathering. 500 to 1000 feet. Rzehakina epigona. Danian Upper Cretaceous Maestrichtian. Australis zone (solid black on map) Thick bedded medium sandstone slightly carbonaceous 0 to 500 feet. Inoceramus cf. pacificus 1* Campanion Senonian Inoceramus australis 2 Raukumara=R A zone Dark micaceous sulphurous siltstone, sandstone bands. 0 to 1000? feet Inoceramus sp. A 3 B. zone Siltstone and medium sandstone. 0 to 200 feet. Inoceramus sp. B 4 Santonian Bicorrugatus zone Medium to coarse sandstone, glauconitic 0 to 400 feet. Inoceramus bicorrugatus Belemnites. sp. A 5 Coniacian Clarence =C Porrectus zone Medium blue-grey mudstone with concretionary bands c 1000 feet. Inoceramus porrectus Gaudryceras subsacya Porapuzosia off. haughtoni Belemnites sp. B 6 Aucellina euglypha Turonian Barren zone Medium blue-grey mudstone with concretionary bands c. 1000 feet Inoceramus frags. only Cenomanian Concentricus zone Medium blue-grey mudstone, concretionary bands, plant fragments c. 1000 feet Turrilites circumtaeniatus Belemnites sp. C Inoceramus concentricus7 Albian Lower Cretaceous Anglicus? zone Dark mudstone Inoceramus cf. anglicus 8 Taitai? = T? Wharfe Gorge Sandstone Graded bedded sandstone Belemnites sp. D Aptian? C. zone Dark mudstone Inoceramus sp. C 9 Basal conglomerate' Conglomerate and mudstone bands Conglomerate with well rounded pebbles of greywocke, red chert, quartz, and igneous rocks. c. 50 feet. No fossils found Pre-Clarence Strongly compacted dark muddy sandstone and sandy siltstone. Many conglomerate bands. c. 5000 feet. No fossils found near Coverham. Neocomian or Jurassic. ? Fig. 5 (above).—Stratigraphic Column to accompany geological map of Coverham district showing lithology and important fossils. Numbers show distribution of zone fossils on map opposite. Fig. 6 (on right).—Geological map of Coverham district on about the same scale as the air mosaic Fig. 1. Thomson's formations are shown on left in sloping capitals. Numbers 1 to 8 show distribution of zone fossils as indicated in Fig. 5.

Text-Figure 6.