The Tertiary and Quaternary Geology of Ross, Westland.
[Read before Wellington Branch, June 14, 1945; received by the Editor, April 14, 1945; issued separately, September, 1945.]
In the Ross area, Early Tertiary, Late Tertiary and Quaternary groups of sediments rest upon probably early Palaeozoic greywaekes intruded by granite. At least two major orogenic disturbances have affected the region since Oligocene times. The date of the earlier can be placed only approximately—about the end of the Miocene—but the latter movements occurred after the first episode of glaciation. There is evidence of two and possibly three distinct periods of ice advance, the first two being separated by a period of intense deformation, and very probably by an amelioration of climate amounting to an interglacial state. The earliest glacial beds are probably Upper Pliocene (Nukumaruan) in age.
The area dealt with in this paper lies within the north-western quarter of the Totara Survey District, largely within the boundaries of Ross Borough, and entirely within the drainage areas of the Totara and Mikonui rivers. The township of Ross is situated at the foot of Mount Greenland, one of several ranges of moderate height which, in this part of Westland, lie between the coastal plain and the Southern Alps, separated from the latter by a lineament of low saddles marking the course of the Alpine Fault.
Apart from low-lying areas of farm land and swamp surrounding the township, on the coastal plain, and in the Totara and Mikonui valleys, the district is mostly under heavy forest, with particularly dense undergrowth. The clearings made in early days in connection with gold-mining are now overgrown with an almost impenetrable jungle of gorse and blackberry and regenerating native forest. Although outcrops are fairly numerous in stream beds, water-races, abandoned sluicing claims, and on tracks, the dense vegetation renders travelling slow and difficult, so that a thorough investigation of the area would take many months. In this paper are presented the observations made during a few short visits.
The Ross area received attention from members of the old Geological Survey, but mainly in connection with the gold-bearing drifts, and was later included in the regional survey of the Mikonui Subdivision. The Bulletin dealing with this subdivision (Morgan, 1908) is the most recent published account of the region, and contains numerous references to the earlier literature. A recent report by Wellman (1945) describes quartz sands in the early Tertiary beds near Ross.
In a recent paper of which the writer was co-author (Gage and Wellman, 1944) describing the stratigraphy of a nearby area, attention was drawn to the advantages of fine stratigraphic subdivision in a geological report of this type. The recording of numerous thin lithologic units, however, involves the provision of local names for a large number of formations, for in the present stage of our knowledge, it is seldom possible to make use of the established New Zealand Tertiary stage-names otherwise than in a general sense, while the recognised regional formation names (e.g., Cobden Limestone) tend to have more and more restricted application with the increase of stratigraphic knowledge. Pending the establishment of a standard Westland section, related to the New Zealand Tertiary stages, a system of numbering of formations has been adopted as a temporary expedient to avoid overloading the nomenclature with a multiplicity of names, many of which may prove never to have more than a local provisional significance.
The post-Cretaceous beds of the Ross district may conveniently be subdivided on the basis of two important unconformities, thus:
R.11. Recent Floodplain Deposits; Low-level Terraces, etc.
R.10. Piedmont Moraines, Associated Gravels and Silts.
R. 9. High-level Terrace Gravels.
Late Tertiary Group.
R. 8. Upper Decomposed Conglomerate.
R. 7. Laminated Fine Silts ana Boulder Clay.
R. 6. Lower Decomposed Conglomerate.
R. 5. Marine Breccia, Sandstone, Limestone, and Conglomerate.
Early Tertiary Group.
R. 4. Limestone.
R. 3. Calcareous and Glauconitic Sandstone.
R. 2. Coal Measures.
(Unconformably overlying R.1. pre-Triassic greywacke, argillite.)
In the following pages, the content, distribution and correlations of each formation will be described. Earlier classifications are discussed later. Correlation is based on the superposition of lithologically distinct groups of strata, with a few checks from palaeontology.
Complexly folded and faulted greywacke and argillite of uncertain, but probably early Palaeozoic age, intruded by pre-Cretaceous granite and metalliferous quartz veins, and by probably pre-Tertiary doleritic dykes, form the basement rocks, as fully described by Morgan (1908, pp. 96–101).
Early Tertiary Group.
At Ross, members of this group are, so far, known to outcrop only within a small triangular area east of Donnelly Creek. Elsewhere, they have previously been mapped on Ford Ridge and Doctor's Hill, six miles to the south-east, and at Koiterangi, nine miles to the east.
R. 2. Coal Measures.
At approximately half a mile up the main southern branch of Coal Creek, a slip on the south side exposes a contact of coal measures against basement greywacke. Local inversion of the beds brings about a steep south-easterly dip. The following section was observed:—
|(Stratigraphic upper contact obscured.)|
|Mostly soft white massive quartz sandstone:
micaceous and carbonaceous bands
|Mostly angular quartz pebbles up to 2in diameter||1|
|(Unconformable on leached, fractured greywacke.)|
These beds appear again in normal succession in the most northerly small branch of Coal Creek. The basal breccia here consists of greywacke pebbles. An account of the quartz sands, in which their possible use for glass-making is considered, has been given by Wellman (1945).
The following Proximate Analysis of coal from Coal Creek, Ross, done by Mr. W. Doherty, was made available by the Superintendent of the State Coal Mines, Greymouth:—
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
Morgan recorded the presence of coaly matter in these beds, which he included with his Upper Miocene, believing them to be younger than limestone that he correlated with the Koiterangi Beds (pp. 106–108). There is, however, very little doubt that the limestone overlies the coal measures, and no reason is known for dissociating this coal from the coal measures at Koiterangi. (Morgan, 1908, p. 105; Gage and Wellman, 1944, p. 357.) They may also be equivalent to the Brunner Beds at Greymouth, of early Eocene age.
R. 3. Calcareous and Glauconitic Sandstone.
Coarse calcareous sandstone or fine conglomerate, glauconitic in its upper part, outcrops in Coal Creek upstream from the junction of the northern branch. A soft, variegated red and greyish-green sandy bed downstream from the junction is interpreted as a weathered greensand. The attitude of these beds is obscure, but it is reasonably certain that they lie between R.2. and the limestone R.4. The thickness is not known, but is certainly more than 20 feet. Morgan did not differentiate the calcareous portion from the coal measures, believing them to overlie the limestone which then was known only from a solitary occurrence in Hodson Brook. Further, he considered the calcareous beds to pass upwards into brown fossiliferous sands, now known to be very much younger, and separated from the limestone by an unconformity.
Fossils recently collected from these beds by Mr. H. W. Wellman were examined by Dr. J. Marwick, who gave the age as probably Whaingaroan or Duntroonian.
No exact equivalents of these beds are known from other areas. The section at Koiterangi shows little resemblance, no comparable coarse calcareous sandstone or greensand being present, although the lower part of the limestone there contains fine sandy bands and some glauconite.
R. 4. Limestone.
The bulk of the prominent knob north of Coal Creek is composed of limestone described by Morgan (1919, p. 228), who correlated it with the Cobden Limestone at Greymouth. The rock is a hard, crystalline limestone, and Morgan's petrological description reads as follows.
“The Hodson Creek limestone is a nearly white close grained rock. Under the microscope it is seen to be composed largely of calcite crystals, together with Foraminifera, Polyzoa, and other organisms. A few grains of quartz and feldspar with some cloudy matter complete the section examined.”
The limestone overlies R.3. sandstones, etc., but the actual contact was not seen. An upper contact of the limestone was examined by Wellman (1945) and interpreted by him as an unconformity separating it from R.5. beds, and possessing an unusual feature in that older beds are now less steeply tilted than younger. On the spur running north from the knob on which is situated the Ross lime quarry, Pleistocene deposits overlap upon the surface of a steep dip-slope of limestone, while on the western side of the quarry, limestone is separated from beds of the younger groups by a north-south fault traversing the eastern valley-slopes of Donnelly Creek.
Morgan gave a minimum of 140 feet for the thickness of the limestone, and although an exact estimate is impossible, the full thickness exposed may be estimated as about 300 feet.
From the quarry, in which the highest exposed part of the limestone may be seen, the following fossils were collected by Mr. D. E. Morgan, of the Shell Oil Company:—
(G.S.LOC. 3051) Lentipeoten cf. hochstetteri (Zitt.) Chlamys compitum (Marw.) Mesopeplum sp.
In addition, Serripecten sp. cf. polemicus Marw. was collected in Coal Creek. Mr. C. A. Fleming, Assistant Palaeontologist, N.Z. Geological Survey, reported that these fossils broadly indicate a Whaingaroan age. A recent collection from Coal Creek by Mr. Wellman, included a new species of Serripecten and fragments of a coral, Graphularia, that Dr. Marwick considers to indicate Whaingaroan-Duntroonian age.
The upper part of the Koiterangi limestone (K.9., Gage and Wellman, 1944, p. 353) is classed by the palaeontologists as probably Duntroonian, so that correlation between the limestones of the two
localities is fairly certain. The Cobden Limestone at Greymouth includes Whaingaroan, probably Duntroonian, and definitely Waitakian portions, so that the fossiliferous part of R.4. is equivalent to the lower part of the Cobden section.
Late Tertiary Group.
Residuals of a once probably continuous covering of beds of this group remain on the slopes and the crest of the northern spur of Mount Greenland, while a more or less continuous strip fringes the base of the range on the northern and western sides. The beds unconformably overlie either basement or members of the Early Tertiary group.
R.5. Marine Breccia, Limestone, Sandstone and Conglomerate.
In a small north-western tributary of Jones Creek, the contact of this formation upon basement and the following almost continuous succession is exposed:—
|(Conformably overlain by lignitic clays of R.6.)|
|Fine, poorly-bedded quartz sandstone||15|
|(iii) Fresh, clean, well-sorted littoral conglomerate, pebbles of schist, gneiss, granite, some greywacke up to 6 inches diameter, average size 2 inch, mostly flattened||20–25|
|Moderately hard laminated brown sandstone||20|
|(ii) Layer of fresh, well-rounded granite, schist and greywacke boulders up to 2 feet size||3|
|Soft brown, fairly coarse sands, few lenses of fine, well-rounded conglomerate||100|
|Well-rounded conglomerate, granite, quartz, greywacke up to 3 inches diameter||6|
|Coarse, brown sandstone, thin layers of grit, flakes of mica up to ¼ inch, poor shell casts||50|
|Massive, fairly fine, brown sandstone||20|
|Finely laminated current-bedded grey sandstone, grading down to laminated siltstone||30|
|(i) Blue-grey, hard shell-limestone (G.S.Loc. 3156)||6|
|Sandstone, with scattered sub angular greywacke pebbles||3|
|Massive fine brown sandstone||20|
|Soft grey sandstone, scattered shell fragments||10|
|Breccia, angular greywacke pebbles and fragments of Ostrea in calcareous matrix||2|
|(Rests on R.1. greywacke.)|
Material from the shell limestone band (i) was examined by Dr. J. Marwrick, who identified the following mollusca, indicating a Waitotaran age:—
(G.S.Loc. 3156) Ostrea.
Lima cf. waipipiensis Marsh. & Murd.
Maorimactra chrydaea (Suter).
Bassina cf. yatei (Gray).
Marama sp. (lunule shorter than usual).
Gari cf. stangeri (Gray).
Antisolarium, aff. stoliczkai (Zitt.). (But low spire as at G.S.Loc. 2874, Stillwater.)
A very similar section of these beds is exposed in Donnelly Creek, east of the fault referred to earlier. The lower contact was not seen by the author, but is represented by Morgan as a sedimentary contact upon greywacke. A low cliff of soft sands on the eastern bank upstream from Bayley Gully contains limonitic bands rich in shell
casts, representing leached shell-limestone. Dr Marwick examined a collection from this locality, and reports as follows:—
(G.S.Loc. 3187) Glycymeris (Manaia) cf. manaiacnsis or rapanuiensis.
Mactra, cf. discors (Gray).
Maorimactra. chrydaea (Suter).
Dosinia cf. subrosea, (Gray).
Bassina cf. yatei (Gray).
“A Waitotaran age is indicated by crowds of Maorimactra chrydaea plus a large proportion of Recent species; Manaia is no later than Waitotaran and no earlier than Taranakian.”
The Donnelly Creek section includes two coarse-boulder layers, compared with one in Jones Creek (ii), separated by 20 feet of micaceous sandstone. The lower part of the littoral conglomerate (iii) contains in Donnelly Creek a layer of rounded pebbles up to 4 inches diameter. It is succeeded upwards by 3 feet of fine clean sand directly underlying a decomposed conglomerate (R.6.).
The section in Clear Creek and Macdonald Creek is less complete. Massive grey sandstone in Clear Creek probably represents the lower grey sands of R.5. A greywacke breccia containing Ostrea fragments is exposed about five chains up Macdonald Creek. Ten or twelve chains further upstream, a slip on the north side shows almost the same sequence as in Jones Creek. The boulder bed (ii) is again present as two separate, closely-overlapping lenses, and the littoral conglomerate (iii) is again succeeded upwards by fine quartz conglomerate.
R.5. beds are met with at intervals along the Mount Greenland track, and fossiliferous blue sands and marly clays were described by McKay (1893, p. 170) from the ridge between Totara River and Donnelly Creek at an elevation of 800 feet or 1,000 feet.
Fossiliferous sands, silts, and shell-limestone of Waitotaran (Lower Pliocene) age are known from several localities in Westland, including Westbrook (near Kumara), Maori Gully (near Stillwater), and Blackball. In the unpublished work of the geological staffs of the oil exploration companies recently operating on the West Coast, the Waitotaran has been mapped as an extensive formation conformably or disconformably overlying Opoitian (Lowest Pliocene) and older Tertiary beds.
R.6. Lower Decomposed Conglomerate.
In the north-western branch of Jones Creek and along the Mount Greenland water-race the following section appears conformable upon the R.5. sequence already given:—
|(Capped by R.7. laminated silts.)|
|(iii) Sub-rounded to sub-angular conglomerate of deeply weathered grey-wacke, hornfels, granite and schist, up to 1 foot diameter, average 4 inches, in fine sandy matrix||300–400|
|(ii) Finely laminated soft silt and sand, grey mudstone, lignite and carbonaceous clay||3|
|Coarse arkositic sandstone, with scattered pebbles of schist, quartz, granite, up to 2 inches diameter||12|
|(i) Soft yellowish clay, including 2 inches lignite||2|
The lignites (i, ii) indicate that marine conditions had come to an end. The first suggestion of glacial transportation appears in the form of occasional angular boulders of schist or hornfels up to 3 feet in diameter near the top of the formation. The nearest source of schist is east of the Alpine Fault, and of hornfels immediately to the west of the fault. No agent other than ice-streams could conceivably have transported debris a distance of nine or ten miles without producing some degree of rounding, whereas the large schist and hornfels boulders are decidedly angular. They can have suffered little water transportation from the time when they were liberated from ice until redeposited amongst the finer water-worn gravels of R.6. The upper part of the main conglomerate (iii) is generally more distinctly bedded than the base. The highest beds showing in the water race are laminated fine sands and silt overlying quartz sands containing scattered pebbles, and are thought to be the lowest beds of the next formation. R.6. conglomerate underlies R.7. silts on the Mount Greenland Track at the commencement of the ascent from the floor of Jones Creek, although the actual contact was not seen.
Fairly deeply weathered conglomerate possibly belonging to this formation overlaps upon greywacke basement towards the head of Cameron Creek, on the eastern slopes of Mount Greenland. The beds have a moderate easterly dip, and upwards contain an increasing proportion of angular and unsorted material, the upper part having almost the appearance of a moraine. These beds appear to have formed the surface from which overlying, auriferous, fresh greywacke gravels, apparently very locally derived, were sluiced in a group of very old claims.
Deeply weathered conglomerates forming a strip on the crest of the main north spur of Mount Greenland towards the summit may be either R.6. or R.8. They were reported by McKay to be richly auriferous.
This formation has not previously been distinguished from R.8., and its possible correlatives in other parts of Westland have been mapped as “Old Man Bottom” or “Moutere Gravels.” In the Arahura Valley, a poorly exposed, decomposed conglomerate, probably equivalent to R.6., underlies blue, laminated silts in the Humphrey's Gully sluicing claim, to be referred to later. Other possible equivalents are the conglomerates overlying fossiliferous marine Tertiary at Greenstone referred to by McKay as “Brighton Bottom,” and old gravels showing in the Hokitika River near Rimu (Bell and Frazer, 1906, p. 86). R.6. beds at Ross contain no internal evidence of age, and there is no evidence that a long period of time unrepresented by strata intervened between the deposition of R.5. and R.6.
At least 400 feet of beds was deposited above R.6. before the deformations occurred which produced the marked angular unconformity above R.8., to be described later. There is every reason to suppose that much thicker sections of these old gravels and
silts exist in other parts of Westland, notably in the Nelson Creek area. If the date of the last major earth movements in this region is taken as early as Pleistocene, then the age of R.6. is almost certainly Pliocene, and it could be as old as Waitotaran (Lower Pliocene).
R.7. Laminated Fine Silts and Boulder Clay.
The following section is exposed in the bed of Jones Creek upstream from the Ross United Claim.
|(Capped by peaty layer at base of R.S.)|
|Oxidised evenly and finely laminated fine micaceous silt or clay||10|
|Fresh, light blue-grey, finely laminated silt||10|
|Angular blocks hornfels up to 6 feet size, with finer grades of the same material showing all degrees of rounding from angular to sub-rounded, in blue clayey matrix; entirely unsorted||20|
|Fresh, light blue-grey fine silt, scattered angular and sub-angular boulders of hornfelsic greywacke, few lenses of well-rounded hornfels, schist and granite pebbles up to 3 inches diameter||30|
|Similar to bed above, but, oxidised||20+|
|(Contact with lower formation not exposed.)|
This formation is considered to be a glacial deposit because it contains coarse angular blocks of hornfels, a rock which could only have been derived from localities no nearer than nine miles away. The angularity of the material cannot therefore be due to an origin as a locally derived talus deposit. Furthermore, the formation closely resembles the moraines and varved silts produced by the Pleistocene piedmont ice (Plate 13, figs. 1 and 2). Evidence for superposition on R.6. comes from the Mount Greenland Track and Jones Creek sections. The sequence is continuous, with accordant dips, and although the exact contact was not seen, there is no doubt that the succession is as given. Previously mentioned laminated silts on the Mount Greenland Water-race almost certainly represent the base of R.7. The bedding of the underlying conglomerate can only be determined approximately, but there is no possibility of marked angular unconformity such as exists higher in the sequence.
Laminated blue clays are exposed in an extensive slip on the eastern bank of Donnelly Creek, downstream from Coal Creek, where they contain lenses and bands of fresh hornfels and granite pebbles, mostly fairly well rounded, and again on the western side of the Ross Lime Company's quarry, where they are faulted against R.5. limestone.
Downstream from the thickly overgrown sluicing claim at McLeod Terrace in the Mikonui Valley, R.7. beds form the western valley slopes for a distance of about 30 or 40 chains, and appear thicker than in Jones or Donnelly Creeks. Steeply dipping alternating fine blue silt and finely laminated blue clays pass upwards into a laminated brown weathered silt underlying R.8. conglomerate. The top of the silt has been deformed by contemporaneous ice movements. Morgan (1908, p. 110) refers to blue clays now obscured by tailings behind the Ross United open-pit workings, which probably are R.7.
This formation lias been confused with the considerably older middle-Tertiary blue clays of Westland commonly known as “Blue Bottom.” Finely laminated blue clays exposed in the bottom of the Humphrey's Gully sluicing claim are bedded between upper and lower conglomerates similar respectively to R.8. and R.6. They were included with the Blue Bottom by McKay, Bell and Morgan, but are very probably equivalent to R.7. In the Greymouth district, similar silts occur on the track from Notown to Molloy's Lookout, according to information received from Mr. H. J. Evans; also a folded series of laminated silts with moraine and lignites has been seen by the author in Nelson Creek, seven miles upstream from the village of the same name.
Tilted finely varved silts, with morainic material and sands containing marine fossils were lately described from Pug Creek, a tributary of the Omoeroa River, approximately 80 miles south of Ross, by Wellman and Willett (1942B, p. 212), and similar beds have since been discovered still farther south at the Paringa River by Mr. Wellman. The recent age indicated by the fossils renders correlation with R.7. improbable. A fuller discussion of the age of R.7. is reserved for a later page.
R.8. Upper Decomposed Conglomerate.
Good exposures of the upper part of this formation may be seen in Jones Creek immediately south of Ross township, and in old alluvial workings near the commencement of the Ross-Mount Greenland Track. The upper portions are exposed on the Ross Borough water-supply race and in the Mont d'Or sluicing claim. The following is a composite section:—
|(Overlain with strong unconformity by R.10. morainic gravels, etc.)|
|In part deeply weathered conglomerate of greywacke, schist, granite up to 6 inch size; bedding and sorting moderately good at base, poorer in upper parts; bulk of the deposit well-rounded||300|
|Fine conglomerate, mostly well-rounded quartz peas, with small amount of granite and greywacke||1–3|
|(Conformably overlies R.7. laminated silts.)|
Glacial characteristics are absent from this formation. It forms a sinuous belt skirting the foot of Mount Greenland on the northern and western sides while outliers occur on the higher parts of the range. In the Mikonui Valley the lower contact is well seen on the western bank downstream from the McLeod's Terrace sluicing claim. The peaty layer is absent here, and the upper part includes coarse sands and lenses of fine pebbles. No single complete section through the formation has so far been found. Mention has already been made of deeply weathered conglomerates forming a residual capping on the crest of the main northern spur of Mount Greenland from the junction of the Alpine and Cedar Creek tracks almost to the summit at 2900 feet, which may be either R.6. or R.8.
These beds are known elsewhere as “Old Man Bottom” (Greymouth district) and “Moutere Gravels” (Nelson). In the Arahura Valley, beds considered by the writer to be equivalent to R.8. were worked for gold, and were described as “Humphrey's Gully Beds” by Bell and Fraser (1906, p. 88). As at Ross, they are unconformably overlain by moraines, and rest upon fine varved silts similar to R.7.
R.8. is so far the youngest known formation of the group preceding the early Pleistocene movements. Its age is probably uppermost Pliocene (Nukumaruan or Castlecliffian), but could be earliest Pleistocene.
Most of the strata of this group occur at lower levels. They were deposited during the later stages of, or after the Pleistocene orogeny, and in contrast to the earlier groups, their distribution is closely related to the present physiography.
R.9. High-level Terrace Gravels.
On the south-western slopes of the Greenland Range, residuals of an alluvial deposit occur on the flat crests of a series of spurs of uniform height (1900 to 2100 feet) between small tributaries of the Mikonui River from Sandstone Creek south-eastwards to the Cedar Creek huts. (See Text fig. 2.) They consist of well-rounded, fairly deeply weathered greywacke cobbles up to 3 inches diameter, with some granite and schist and layers of sand, and show no appreciable dip apart from current-bedding. A similar series of spurs is present at about the same height on the opposing slopes of Mount Rangitoto, south-west of the Mikonui River, but it is not known whether these are gravel-capped.
Several exposures of the gravels occur on the Cedar Creek track. Indications of glacial origin are lacking, and they are best interpreted as normal stream deposits on the floor of a wide ancient valley ancestral to the Mikonui
The only likely equivalents of R.9. known to the writer are well-compacted gravels unconformably overlying Miocene (lower Ihungia) beds, and perched on the summit of Koiterangi Hill (Gage and Wellman, 1944, p. 360). Probably over most of the Westland piedmont region glacial erosion has destroyed similar evidence of earlier erosion cycles.
R.10. Piedmont Moraines and Associated Gravels and Silts.
Deposits of the extended glaciers of the piedmont phase of the Pleistocene glaciation fringe the base of Mount Greenland and Mount Rangitoto. Although locally deformed by contemporaneous ice pressure, they are for the most part gently dipping, and consist of fresh or moderately weathered moraine and fairly well-rounded and sorted greywacke and granite gravels. Some of the moraine is very coarse and some is interbedded with sands and laminated silts containing fragments of little altered wood.
Low Saddle at
Mikonui R. Tuke R. Alpinc Fault. Mt.Rangitoto
South Flank Greenland Range
2,200 Bench → Gravel capped (R9.)
Bench (? Gravel capped.)
Drawn by K. S. Murray from Photo.
Text Fig. 2.—View from prominent bend on Cedar Creek track (altitude approx. 2,220 feet), one mile west of Trig H.B. Angle of view-approx. S.E.-S. illustrates valley-in-valley form of lower Mikonui Valley. Cross-profile of old valley and remnants of its valley-floor deposits capping 2,200 feet benches lack glacial characters.
Morainic beds overlie indifferently R.8. and older formations with pronounced angular unconformity. They form a fringe overlapping on older beds on the lower slopes of the hills west of Ross Township. Contacts are to be seen in all the large sluicing claims from Jones Creek westwards to Donoghue's, and in the Mikonui Valley about one mile upstream from the main road bridge. Since both the glacial and R.8. gravels were worked for gold at Ross (McKay, 1893, p. 153), and steeply-dipping older morainic beds were also present, it is not surprising that a good deal of confusion is evident in early descriptions of the goldfield.
Varved silts may be seen on the roadside at Cemetery Hill, and on the saddle between the Totara and Mikonui Valleys. Brown sands with interbedded morainic gravels on the west side of the Totara Valley from Stony Creek to Cameron Creek were mapped by Morgan as marine Miocene. They have a general very slight northward dip, and are more probably the deposits of streams associated with an ice tongue which descended from the Upper Mikonui region via the Totara Saddle and Totara Valley.
The deposits of the main ice advance have been described from almost all parts of Nelson and Westland by numerous authors. From Abut Head southwards to Milford Sound they have lately been described by Wellman and Willett (1942B, p. 206 et seq.), while accounts of these in the Hokitika, Greymouth and Reefton districts appear in N.Z. Geological Survey Bulletins (Bell, 1906, p. 89; Morgan, 1911, p. 75; Henderson, 1917, p. 97a). The extended moraines of Nelson are described by Henderson (1931, pp. 154–160).
R.11. Recent Flood-plain Deposits, Low-level Terraces.
The lower valley-sides of the Mikonui and Totara Rivers, Donnelly and Jones Creeks and most other large streams carry remnants of terraces and the valley floors are covered by normal flood-plain deposits. No deposits attributable to the final valley-glacier phase of Westland's glacial history are known in the area described.
As in most other alluvial gold-mining areas, recent stream deposits include a large amount of material derived from older gravels, and supplied at an accelerated rate in the form of tailings. Aggradation of the bed of Donnelly Creek and of the lower part of Clear Creek has resulted from this process.
The classifications of formations employed by earlier writers dealing with the geology of Ross and the adjoining Hokitika area are summarised in Table I. The grouping adopted in this paper differs chiefly in the distinction of the R.7. morainic silts from the main mass of “Blue Bottom” marine fossiliferous sediments. The confusion of these beds by earlier writers led to the erroneous description of morainic Blue Bottom by Bell and Fraser (1906, p. 87). In his 1893 report, McKay avoided correlating the older auriferous conglomerate (R.8.) with the Moutere Gravels, but in another report the following year he uses this unfortunate term,
in accordance with the then official Geological Survey classification, to replace the suitably vague title of “Older Pliocene and Upper Miocene.” The R.9. gravels of this paper are aptly described by the heading of McKay's “Ia—Pleistocene” group: “High-level old river channels and terraces,” although he appears to include the Cedar Track gravels with the Moutere gravels. In the 1894 report, McKay, under the general heading of Moutere Gravels, describes the R.8. conglomerate of this paper as “Brighton Bottom,” correlating it with the Humphrey's Gully beds of Arahura and the Old Man Bottom beds of the Greymouth district. Bell and Fraser (1906) considered the Humphrey's Gully beds a distinct formation. They make mention of the Brighton Bottom as sands with pebble layers overlying the “Miocene” fossiliferous marine beds (p. 88), but suggest the possibility that they could be lateral equivalents of the Humphrey's Gully beds. These authors were on the right track in suspecting the existence of two distinct sets of older conglomerates, but were led astray when they confused the silts underlying the higher decomposed auriferous conglomerate with the true marine mid-Tertiary Blue Bottom sediments. Morgan generally followed the earlier writers and adopted the term Moutere Gravels, thereby inferring a long-distance correlation that has no direct evidence in its support.
The Tertiary and Quaternary geological history of the West Coast was summarised recently by Wellman and Willett in the first of their papers on the geology of South Westland (1942A, p. 304). When their historical scheme is reviewed in the light of the evidence from the Ross district, a general agreement is found as far as the lower Pliocene, but the later Tertiary and Quaternary beds at Ross suggest a somewhat different sequence of events.
Early Tertiary Sedimentation.
As in other parts of Westland, a marine transgression occurred after the deposition of the coal measures, but at Ross the sediments are thin. The Kaiatan stage has not yet been detected at Ross, and at Koiterangi, nine miles to the east, it is only doubtfully present (Gage and Wellman, 1944, p. 358), in strong contrast with the thick geosynclinal Kaiatan of the Greymouth area. The fossiliferous calcareous and glauconitic sands suggest clear shallow water with an absence of terrigenous waste. Limestones having general lithologic similarity occur at Greymouth (Cobden), Koiterangi, Ross, Abbey Rocks and thence at intervals southwards as far as Madagascar Beach (Healy, 1938, p. 87B; Wellman and Willett, 1942A, Fig. 1). Their ages are known with varying degrees of exactness, but are probably all between Whaingaroan and Waitakian. Although probably deposited from fairly shallow, clear seas, they are certainly not littoral. It is therefore practically certain that a continuous sheet of limestone once extended over the Ross area at least. The lithology of the limestones gives no indication of the direction in which the nearest land lay at the time, but suggests that no high land was near, Faunal breaks
|Approximate Age.||This Paper.||Bell, 1906.||Morgan, 1908.||McKay, 1893.|
|R.11. Recent floodplain deposits; low-level terraces, etc.||Fluviatile and marine gravels.||Recent morainic, fluvio-glacial, fluviatile and marine gravels.||I. Recent glacier, river alluvia, littoral.|
|Quaternary Group Pleistocene and Recent.||R.10. Piedmont moraines, associated gravels and silts.||Morainic material.||Pleistocene and Pliocene fluvio-glacial, fluviatile and marine gravels.||II. Pleistocene and Younger Pliocene extended glacier deposits, pre-glacial river deposits; marine grave- ls with black-sand leads.|
|R.9. High-level terrace gravels.||Not recognised.||Not recognised.||(?) Ia Pleistocene high-level old river-channels and terraces.|
|R.8. Upper Decomposed conglomerate.||Humphreys Gully Beds.||Moutere gravels.||III. Older Pliocene and Upper Miocene; Humph-reys Gully Beds, Old Man Bottom, brown sands.|
|Late Tertiary Group. Lower Pliocene (Waitotaran) and younger.||R.7. Laminated fine silts and boulder clay.||Not distinguished from Blue Bottom.||Not distinguished from Upper Miocene.||Not distinguished from IV Lower Miocene.|
|R.6. Lower Decomposed Conglomerate.||Not distinguished from Humphreys Gully Beds.||Not distinguished from Moutere gravels.||Not distinguished from III Humphreys Gully beds.|
|R.5. Marine breccia, limestone, sandstone and conglomerate.||Blue Bottom.||Upper Miocene; younger conglomerates, sandstones and clays.||IV. Lower Miocene; marine Tertiary beds; blue fossiliferous sands and marly clays.|
|Early Tertiary Group. Oligocene (Duntroonian) and older.||R.4. Limestone.|
|R.3. Calcareous and glau-conitic sandstone.||Koiterangi Series.||Koiterangi Series.||VI. Cretaceo-Tertiary; Upper, Middle and Lower.|
|R.2. Coal measures.|
|Pre-Cretaceous. Probably pre-Triassic.||R.1. Greywacke, argillite.||Kanieri Series.||Late Palaeozoic (? Carboniferous); Greenland Series.||X Triassic (?). Maitai Series.|
occur, as for example at Koiterangi, where the Waitakian stage is missing. The Lower Miocene sediments that succeed the limestones in Westland are of calcareous silty facies, but the first evidence of important emergence comes at approximately 50 feet above the base of the Lower Ihungia beds at Koiterangi, in the form of a 3–foot band of conglomerate composed of granite, schist, coal, etc. The appearance of schist indicates that elevation of the alpine axis had commenced. Marine sediments of generally increasingly coarse facies continued to be deposited during the Miocene stages, making up the lower part of the sequence generally known as the “Blue Bottom” in Westland. The non-discovery so far of Miocene beds at Ross is attributed to the overlap of considerably younger beds upon the limestone obscuring its upper contact.
Work as yet unpublished by the writer and others indicates that several breaks in deposition occurred in Westland and Nelson about the end of the Miocene. Urenuian, Opoitian or Waitotaran beds rest unconformably on beds ranging in age from Kaiatan to Tutamoe (Awamoan). Wellman (1945) has described a probable angular unconformity between beds of the upper and lower Tertiary groups in Coal Creek, Ross. The limestone which was once clearly continuous over the Ross area, was completely eroded at Jones Creek before the Waitotaran sedimentation. Deformation, elevation and erosion therefore occurred between the Waitakian and Waitotaran stages, but the date of the movements cannot be fixed more closely on the evidence available from the Ross district.
Late Tertiary Sedimentation.
The Waitotaran (Lower Pliocene) shell-limestone and associated beds (R.5.) are near-shore, shallow-water deposits. They are missing and probably were never deposited more than about a mile east of Ross, thus fixing fairly closely the most easterly position of the shoreline at this stage. A thin section of only Waitotaran marine strata is present at Ross, in contrast to the thick Urenuian, Opoitian and Waitotaran beds in other parts of Westland. Whether this sedimentation was initiated by slight depression is unknown, for the shoreline may have reached Ross merely through wave-attack upon land of low relief produced by sub-aerial erosive agencies operating after the Mid-Tertiary orogeny. The alpine region suffered continual uplifts to maintain the supply of schist-bearing conglomerate layers which characterise the marine Pliocene of Westland. Relief was not entirely destroyed west of the Alpine Fault, for R.5. conglomerates include granite and hornfelsic greywacke which do not occur to the east.
Late Pliocene Withdrawal.
Littoral and deltaic deposits at the top of R.5. mark the close of the later Tertiary marine deposition, being succeeded by terrestrial sediments. No measurable tilting accompanied this change of facies. Marine erosion, even if aided by subsidence of the depositional area, was nevertheless unable to cope with the abundant supply of waste transported by fast-flowing rivers from the ever
more rapidly rising alpine region. Deltas advanced, and the shoreline migrated an unknown distance westward of Ross. Elevation was still not confined to the region east of the Alpine Fault, for granite persists as a common constituent of R.6. conglomerates.
Isolated, large, angular blocks of schist at the top of R.6. for reasons given earlier, are interpreted as of glacial origin. Ice was already transporting material from the alpine region to within a short distance of Ross. R.7. silts are interpreted as varves, and their morainic inclusions as the deposits of floating ice. Nevertheless, it is not easy to visualise the conditions of deposition. Assuming as correct the correlation of R.7. with silts at Humphrey's Gully, the formations are so similar in lithology and relations as to suggest deposition in a continuous body of water. Unless subsidence occurred during and after the formation of R.6., the floor of the lake in which deposition of R 7. is assumed to have taken place would be at an altitude of over 400 feet above sea-level (this being the approximate thickness of R.6. non-marine deposits). If subsidence did occur, the absence of marine faunas from both R.7. and Humphrey's Gully silts makes direct connection with the sea improbable. What, then, was the nature of the barrier which retained the inland lake waters, or, in the alternative subsidence theory, prevented the introduction of marine faunas to the silts?
The answer to this question may be found when the distribution of the silts in other areas is better known. At this stage, two possibilities are worthy of mention. The first, and in the author's opinion the more likely, is that a barrier composed of terminal moraine marking the maximum advance of an early piedmont ice-sheet may have caused a large lake to form during the early stages of ice retreat. If formed at a low altitude, the barrier may have effectively prevented the intermingling of lake and sea waters. The second possibility is that the origin of the lake was deformational. A down-warping contemporaneous with the onset of glaciation could well have been too broad and gentle to produce appreciable angular unconformity at the base of R.7.
The writer feels that the available evidence is too meagre to justify more than the above tentative suggestion regarding the mode of origin of R.7.
Morgan (1926, p. 274) gives a resumé of the views of writers up to that time concerning the date of commencement of the glacial epoch. Although Hutton, in New Zealand, and Boule and others abroad considered this to be in the Pliocene, Morgan records a tendency among stratigraphers to define the Pleistocene as beginning with the onset of glaciation. Fleming (1944, pp. 207–220) concludes from a study of migrating molluscan faunas that a lowering of sea temperatures in the New Zealand region reached its climax in the mid-Pliocene (Lower Nukumaruan), whereas in the upper Pliocene, a warm sub-tropical current began to affect the region, and persisted throughout the Pleistocene, notwithstanding the glaciation of the land. At Ross, the lowest glacially derived beds are stratigraphically about 400 feet above the highest Waitotaran fossils, and no observ-
ableunconformity intervenes. On the other hand, decided unconformity exists between these beds and the Pleistocene piedmont glacial deposits. It is not unreasonable, therefore, to associate R.7. glacial beds with the Waitotaran-Nukumaruan cooling, and revive Hutton's Pliocene glaciation.
Tilted fossiliferous glacial silts and morainic gravels were described by Wellman and Willett (1942B, p. 212) from Pug Creek, a tributary of the Omoeroa River. Finlay, quoted in that paper (pp. 213–214), states that the age of the silts as indicated by foraminifera is certainly not older than Pliocene, and almost certainly Waitotaran or younger. Three molluscan species were all Recent. Since then, Mr. Wellman has collected similar material from the Paringa River containing a more satisfactory micro-fauna. Dr. Finlay is able to give the age as definitely post-Castlecliffian. Correlation of the Omoeroa and Paringa beds with R.7, once considered by the writer, is now scarcely tenable.
Possible Interglacial Stage.
Well-rounded, deeply decomposed conglomerates with peaty layers succeed R.7. at Ross. If any of the material was glacially derived, it suffered long enough transportation by normal streams to remove all sign of this. Concentrations of well-rounded quartz pebbles have been noted at the base of R.8. at two localities, suggesting that transportation was slow, and that in part at least, the characteristic deep decomposition of the material had occurred before it was re-deposited. This slowing down of the process of erosion and transportation is attributed to a lowering of relief of the alpine region, which was related, either as cause or effect, to the discontinuance of dominantly glacial processes of denudation after the formation of R.7.
Sharp folding, elevation and stripping of a large part of the cover of Late Tertiary beds from the structural “highs” had time to occur before glacial action is again in evidence. Glacial features are absent from R.9. gravels, which from their situation and attitude are intermediate in age between R.8. and the piedmont moraines. The interval therefore is interpreted as an interglacial stage, with which may also be connected the return of warm sea currents deduced by Fleming. But if sea currents have a controlling influence on glaciations, it becomes more difficult to explain the anomaly recognised by that author in that Pleistocene glaciation was apparently unaccompanied by a re-invasion of cold water faunas.
Early Pleistocene Orogeny.
The subdued land surface that prevailed during the formation of R.8. in late Pliocene or early Pleistocene was dissected following strong elevation east of the Alpine Fault, accompanied by folding, which resulted in strong differential elevations and depressions on the western side. Later erosion by streams and glaciers has greatly modified but not entirely obliterated the initial features of the cycle succeeding this orogeny. The Greenland and Rangitoto ranges and other highlands west of the Alpine Fault were elevated
at this stage, and retain residual cappings of younger covering strata, the attitudes of which show that the axis of up-folding is still approximately the axis of topographic elevation.
The term “Kaikoura Orogeny” has long been applied to strong deformations in the late Pliocene or early Pleistocene. Cotton (1938, p. 4B) recognises a general tendency to regard the movements as spread out over a considerable period. The evidence from the Westland region as a whole suggests continual movements from mid-Tertiary onwards, with no prolonged periods of rest. Indeed, maintenance of the supply of sediments for the thick Westland Tertiary sections requires continual elevation of the source areas, presumably in this case the Alpine region. The movements that brought about faulting and sharp folding of Pliocene beds were in Westland more severe than any preceding since perhaps the post-Hokonui Orogeny. Cotton (p. 2B) records that some areas affected by mild warpings in the mid-Tertiary movements were strongly affected by the Kaikoura Orogeny, and vice versa. If the term is permissibly used to denote an orogenic paroxysm affecting certain areas in late Pliocene or early Pleistocene times, then it is correct to say that the deformation of R.5., R.6., R.7. and R.8. beds at Ross is an expression of the Kaikoura Orogeny.
Strongly differential movement ceased after the Mount Greenland block had been raised at least 1000 feet, and probably not more than 1800 feet. A halt then occurred long enough for the sculpture by ordinary steam erosion of the wide, flaring valley in the floor of which the Mikonui River is now entrenched in a deep gorge. As mentioned earlier, there is no evidence that glacial processes were active at this period.
A series of comparatively uniform uplifts then occurred, continuing until the late Pleistocene. Intervening periods of rest have left traces in the form of elevated coastal platforms. In the Greymouth-Hokitika coastal strip, two prominent benches capped by gravel deposits are respectively at 50–80 feet and 400–600 feet above present sea-level. The absence of benches from areas where the piedmont moraines reached the sea has been considered by Wellman and Willett (1942A, p. 304) to imply that sea-level had attained approximately its present position before the main Pleistocene ice advance. Minor movements of the strand line, nevertheless, have occurred in fairly recent times. The total amount of elevation suffered by the Mount Greenland block as a result of all the Kaikoura movements is from three to four thousand feet. Wellman and Willett estimate that the axis of uplift of the Alps was raised 7,000 feet to 12,000 feet (p. 305).
High mountains near the coast combined with the warm Pleistocene sea postulated by Fleming may have been sufficient in themselves to initiate a great ice advance. Heavy precipitation would maintain abundantly supplied snowfields, from which thick, actively-eroding glaciers flowed, coalescing beyond the mountains to deposit the extensive moraine sheets of Westland.
From a source east of the Alpine Fault, inland from Ross, an ice stream passed to the lowlands by way of what is now the valley of the Totara, moraines and varves being deposited at the present Totara Saddle. Ice evidently did not completely over-ride Mount Greenland, and no ice stream could have passed through the elevated valley between Greenland and Rangitoto. The maximum thickness of ice was about 1600 feet around Mount Greenland. Farther south, opposite the region of greatest alpine elevation, glaciation was more intense.
The piedmont ice overflowed the foot of the Greenland and Rangitoto ranges, its deposits unconformably overlying steeply-dipping older beds. Lateral lobes of main ice streams descending along the present courses of the Hokitika and Waitaha Rivers, augmented by a lesser stream from the Totara-Upper Mikonui region skirted the base of the coastal mountains, and united to form a piedmont sheet in the vicinity of the Mikonui Mouth.
Much has already been written on the question of whether a period of normal stream erosion, during which all but the finer detail of the present physiography was carved out, intervened between melting of the piedmont ice and a temporarily renewed advance of glaciers down the valleys. McKay (1893, 1894) separates recent valley-glaciers deposits from those of the piedmont sheets, and in particular refers to a late advance of the Kanieri Glacier. Bell and Fraser (1906) held similar views. Morgan (1908) makes no suggestion of a second ice advance, and his classification of deposits implies that after the melting of the piedmont ice, glaciers gradually retreated up the present valleys. Later, however (1926, p. 279) he admits the possibility that two or more glacial periods might later be demonstrated in Westland. Henderson, dealing mainly with the glaciers of Nelson province, advances arguments for at least two post-Tertiary glaciations (1931, pp. 159–160).*
If it is true that a great intensification of glacial action can be initiated merely by increased precipitation due to uplift of coastal mountains, then it follows that unless continual re-elevation occurs, the glaciation will fairly soon bring about its own end through the lowering of relief due to rapid denudation. Likewise, a renewal of elevation would cause fresh advances of the previously shrinking glaciers. The historical scheme of Wellman and Willett, although making no subdivision of the glacial epoch, postulates a mid-Pleistocene re-elevation of the alpine region, accompanied by a change in the petrological composition of the moraines.
The Ross district has little to contribute to this issue, although it is noticeable that in its upper and middle reaches the Mikonui
[Footnote] * It is to be emphasised that these discussions do not concern the Pliocene glacial deposits (R.7.) of this paper; Morgan and McKay failed to record the glacial character of these beds, but Bell and Fraser, who mapped them as part of the Blue Bottom, speak of rock-flour and striated stones in the Hokitika district, which they considered were derived from an elevated glaciated inland region.
Valley exhibits glacial forms, eroded partly from earlier glacial deposits. The final retreat of ice into the mountains, and the modification of glacial landforms by normal erosion and deposition in the modern cycle bring to a close the geological history of the area.
The “Late Tertiary Peneplain.”
Wellman and Willett (1942A, p. 303) demonstrate the extension of Benson's Late Tertiary Peneplain to the Mount Cook region by means of a study of summit-height concordance, and consider that its formation took place “after the folding of the Tertiary beds, and before or during the deposition of the pre-Pleistocene Moutere Gravels…” They were aware that the surface truncated mid-Tertiary beds, but at that time it was not known that these beds are separated from Waitotaran by clear-cut unconformity. The erosion surface underlying the Waitotaran (R.5.) most probably represents the Late Tertiary Peneplain in the Ross district. Although the overlap of younger beds upon it at Mount Greenland shows the surface to have retained appreciable relief, there is no record from this district of any later erosion cycle having reached an equally advanced phase. The surface was severely deformed during the later Kaikoura movements.
Alluvial gold is present in almost all the members of the Late Tertiary and Quaternary groups, and has been mined from several of them. There is no record of the basal Tertiary conglomerates at Ross being explored for gold.
Large scale sluicing operations in the older claims at Ross were conducted in either (a) R.8. conglomerates, as in Ross United and other famous claims of Jones Flat, part of the ground worked in the Mont d'Or claim, according to McKay, and the deep lead workings; or (b) the R.10. morainic gravels, as in Mont d'Or, Mount Greenland and other claims between Ross Township and Donoghue's, as well as possibly the McLeod's Terrace claim in the Mikonui. The decomposed gravels of the Alpine claim near the summit of Mount Greenland also probably are R.8. The auriferous beds at Humphrey's Gully, Arahura, are believed to be equivalents of R.8.
Very old claims on the eastern slopes of Mount Greenland near Cameron Creek appear to have worked locally derived grey-wacke detritus, while on the floor of the Totara Valley gold has been worked from glacial gravels and recent re-concentrations of glacial and R.8. gravels. Recent flood-plain deposits and perched remnants in older formations have been worked in many places. The small claims still working are mainly operating in small patches of glacial and R.8. gravels left between the old claims, while tailings in Jones Creek are being successfully reworked for the nth time! The gold in R.8. is reported by present day miners to be very fine. It is also stated that traces are present in R.6. conglomerates and R.5. sands, but not in R.7. silts. Some of the younger deposits
contain very coarse, rough gold that cannot have suffered long transportation by water.
It is evident that although some of the Ross gold deposits have been enriched by successive concentrations, at the same time workable deposits in Recent and Pleistocene beds have originated as a result merely of a single cycle of erosion and deposition of gold derived from quartz veins near at hand.
The writer tenders his thanks to Drs. J. Marwick and H. J. Finlay and Mr. C. A. Fleming for palæontological determinations and advice in connection with this paper, and to Mr. M. Ongley and Mr. H. W. Wellman for helpful criticism. Mr. A. W. Hampton kindly re-drew the map for publication, and Miss K. S. Murray made the fine drawing of the Mikonui Valley.
Bell, J. M. and Fraser, C., 1906. The Geology of the Hokitika Sheet, North Westland Quadrangle. N.Z.G.S. Bull. 1 (New Series).
Cotton, C. A., 1938. Some Peneplanations in Otago, Canterbury, and the North Island of New Zealand. N.Z. Journ. Sci. & Tech., vol. 20, pp. 1B–8B.
Fleming, C. A., 1944. Molluscan Evidence of Pliocene Climatic Changes in New Zealand. Trans. Roy. Soc. N.Z., vol. 74, pp. 207–220.
Gage, M. and Wellman, H. W., 1944. The Geology of Koiterangi Hill, Westland. Trans. Roy. Soc. N.Z., vol. 73, pp. 351–364.
Healy, J., 1938. The Geology of the Coastal Strip from Big Bay to Professor Creek, North-west Otago. N.Z. Journ. Sci. & Tech., vol. 20, no. 2, pp. 80B–94B.
Henderson, J., 1917. The Geology and Mineral Resources of the Reefton Subdivision, Westport and North Westland Divisions. N.Z.G.S., Bull. 18.
—–1931. The Ancient Glaciers of Nelson. N.Z. Journ. Sci. & Tech., vol. 13, no. 3, pp. 154–160.
McKay, A., 1893. Geological Explorations in the Northern Part of Westland. Mines Rep., pp. 132–186.
—–1894. On the Geology of the Northern Part of Westland and the Gold-bearing Drifts Between the Teremakau and the Mikonui. Rep. Geol. Explor., vol. 22, pp. 11–50.
Morgan, P. G., 1908. The Geology of the Mikonui Subdivision, North Westland. N.Z.G.S. Bull. 6.
—–1911. The Geology of the Greymouth Subdivision, North Westland. N.Z.G.S. Bull. 13.
—–1919. The Limestone and Phosphate Resources of New Zealand. N.Z.G.S. Bull. 22.
—–1926. The Definition, Classification and Nomenclature of the Quaternary Periods, with Special Reference to New Zealand. N.Z. Journ. Sci. & Tech., vol. 8, pp. 273–282.
Wellman, H. W. and Willett, R. W., 1924A. The Geology of the West Coast from Abut Head to Milford Sound. Pt. 1. Trans. Roy. Soc. N.Z., vol. 71, pp. 282–306.
—–1942B. The Geology of the West Coast from Abut Head to Milford Sound. Pt. 2, Glaciation. Trans. Roy. Soc. N.Z., vol. 72, pp. 199–219.
Wellman, H. W., 1945. The Geology of Coal Creek, Ross. N.Z. Journ. Sci. & Tech, vol. 27, Sect. B, No. 1.