Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 88, 1960-61
This text is also available in PDF
(475 KB) Opens in new window
– 631 –

On the Definition, Date, and Character of the Ross Glaciation,
Early Pleistocene, New Zealand

[Received by the Editor, June 3, 1960.]


Evidence for the Ross Glaciation is contained in a folded conformable sequence of Pliocene and Lower Pleistocene beds including till and laminated lake silts at Ross, Westland. Originally described as Pliocene, the glacial beds were later correlated on paleobotanical evidence with the Lower Nukumaruan Stage, now Lower Pleistocene. The sequence records a complete cycle of glacier advance and shrinkage superimposed upon the record of rapid alpine uplift and of the final emergence of North Westland from the sea. It is recommended that the term “Ross Glacial Stage” be dropped, and that instead the Ross Glaciation be recognized as a climatic event based upon the lithologic evidence of climatic change recorded in certain Lower Pleistocene beds at Ross, for which formal stratigraphic names are now proposed. Folding of the Lower Pleistocene beds accompanied differential tectonic elevation of the Mount Greenland mass and warping was still continuing after the early late Pleistocene Waimaungan Glaciation. It is not known whether the Ross Glaciation was due to ice-cap glacierization or piedmont coalescence of valley glaciers, but the former is perhaps the more probable.


The glacial character of some members of a folded and steeply-dipping sequence of beds was noted by the writer in 1945 when describing Tertiary and Quaternary deposits near Ross (170° 50′ E; 42° 54′ S) on the West Coast of South Island, New Zealand. The glacial members were separately described and letter-symbols on a small scale text-figure map gave some idea of their distribution, but they were neither mapped separately nor dissociated in age from the conformably underlying Waitotaran (Pliocene) marine beds. At the time there seemed to be no alternative to recognizing an early glaciation of the area in the Pliocene, a rather unpalatable suggestion which had nevertheless been advanced long ago by Hutton (1872). Apart from sampling of carbonaceous layers, there has been little further investigation of the critical outcrops near Ross, no doubt partly because they have not since been as well exposed as in 1945, when the beds of Jones Creek and some small tributaries were temporarily clear as a result of ground-sluicings for gold that have long since ceased.

Later, through advances in paleontology, the Ross beds were correlated more precisely with Pliocene and Pleistocene stages of the New Zealand standard succession, which in turn were becoming more accurately correlated with the European succession. A decision by an international commission regarding the position of the world Pliocene-Pleistocene boundary indirectly resulted in the Ross glacial beds becoming early Pleistocene in age. Wellman introduced the terms: “Ross Glaciation” and “Ross Glacial Stage” in 1951, since when they have come into general use. Brief, single paragraph outlines of the rather involved histories of these terms have recently appeared (Gage and Suggate, 1958, p. 593; Gage, 1959, p. 354), but there is a need for an explicit account of the evidence from Ross for Lower Pleistocene glaciation in New Zealand . The main objects of this paper are to present that evidence, to define the Ross Glaciation in stratigraphic terms, and to discuss the type of glacierization and the tectonic deformation suffered by the Lower Pleistocene beds.

– 632 –

Evidence for Glaciation

The beds under consideration unconformably underlie the extensive piedmont moraines of Westland, which show morphologic as well as textural evidence for their glacial origin, but with which we are not concerned in this paper. The evidence for the earlier glaciation at Ross comes from the two members of the succession formerly designated “R6” and “R7” (Gage, 1945, pp. 144–7). Glacial conditions of deposition are inferred for R7 for the following reasons: (i) it includes a typical till made up of boulders and pebbles unsorted as to size or shape and embedded in tough fine blue silt; (ii) the boulders and pebbles (many of them since found to be striated) include angular fragments of schist and hornfels, for which there is no known or likely source within 9 or 10 miles; (iii) the till is overlain by laminated, graded, fine silts indistinguishable from a type of sediment common in modern ice-margin or proglacial lakes. Large angular blocks of schist also occur in the upper part of the underlying R6 conglomerate, equally remote from the nearest source. At least the upper part of the conglomerate may therefore be interpreted as proglacial outwash from advancing ice, which eventually reached the Ross area and deposited R7.

It has been tacitly assumed that the ice advanced towards Ross from an elevated tract roughly coincident in position with the present South Island mountain axis. There is no evidence from striated rock surfaces, but a westward or north-westward direction of ice flow is demanded by the lithologic composition of the glacial deposits. New Zealand has been totally isolated from lands to the west since long before the Pleistocene, and so no suggestion of ice invasion from the west need be entertained.

Evidence for Age

Younger glacial deposits at Ross were described as “R9: High-level Terrace Gravels” (now regarded as glacial; Gage and Suggate 1958, p. 594) and “R10: Piedmont moraines and associated gravels and silts”. The age of these deposits, which rest unconformably upon the Lower Pleistocene beds, is accepted as late Pleistocene and will not be discussed further here.

The older glacial beds at Ross were inferred by Gage (1945, p. 155) to be Pliocene. From studies of migrations of marine molluscan faunas Fleming (1944) concluded that the seas in the New Zealand region cooled during the Lower Nukumaruan Stage (then regarded as mid-Pliocene) and subsequently became warmer. Finlay (1947, p. 352) correlated the Opoitian and Waitotaran stages with the Plaisancian stage of Europe. During the next few years the use of plant fossils, especially pollen and spores, developed rapidly in New Zealand, so that Fleming, in a paper presented in 1953 (published 1956), was able to quote Couper as supporting the correlation of a carbonaceous band at the base of R6 with the Lower Nukumaruan. While recognizing that direct faunal evidence was still lacking for correlation of the New Zealand stages above the Waitotaran with the Calabrian or Villafranchian, Fleming (1953) recalled that the Commission of the XVIIth International Geological Congress (London, 1948) when recommending that the base of the Pleistocene be fixed at the base of the Calabrian stage, had noted that the first indication of climatic deterioration in the Italian Neogene succession begins to appear at this horizon. In New Zealand, climatic deterioration setting in at the end of Waitotaran time thus provides reasonable grounds for drawing the Pliocene-Pleistocene boundary between the Waitotaran and Nukumaruan stages. Couper and McQueen (1954, table, p. 416) confirmed the Lower Nukumaruan age of the cool-flora lignitic bed at the base of R6 at Ross, and thus the older Ross glacial beds found themselves in the Lower Pleistocene.

– 633 –

Stratigraphic Terminology

Investigations by Gage (1945) and Wellman (1945) at Ross, which lies within the Mikonui Subdivision of the New Zealand Geological Survey (Morgan, 1908), showed the necessity for revising Morgan's stratigraphic classification in the light of recent paleontological advances, but in accordance with views then held with regard to overloading of the nomenclature with provisional terms (Gage and Wellman, 1944, pp. 352–3), Gage designated his divisions of the Ross succession merely by mapping symbols R1, R2, etc. The terms “Ross Glaciation” and “Ross Advance” came into use later, without precise formal definition, but by inference they were based upon R7 by Wellman (1951, p. 30) and upon R6 to R8 inclusive by Fleming (1956, p. 926). “Ross Glacial Stage” was introduced nominally and with local application by Wellman (1951, fig. 2, following p. 24), but was later adopted into a proposed scheme for a glacial chronology of the New Zealand Pleistocene by Gage and Suggate (1958, p. 593). By implication, R8 was excluded on the grounds that it lacks distinctively glacial characters.

Report 6 of the American Commission on Stratigraphic Nomenclature (Richmond, et al., 1959) advises against using evidence of climate change for the defining of time stratigraphic units. The writer, on the other hand, agrees with Suggate (1960) that important reversals of trend in climate change are probably world-wide and virtually synchronous and therefore are acceptable as criteria for time-stratigraphic definitions. A “Ross Stage” of the New Zealand succession might therefore be typified by the beds between the first indications of the Ross cooling and the first indications of post-Ross warming, but no such unit is required since the time-interval concerned appears to be covered by the existing time stratigraphic classification, with which the Ross sequence can be correlated reasonably by paleontological means. It would therefore be advisable to discontinue using the term “Stage” in connection with the Ross glacial beds, and instead to re-define “Ross Glaciation” as a climatic event (not as a stratigraphic unit, the recommendation of Richmond et al., 1959, Table I, p. 666) recorded by the character and sequence of certain beds at Ross. The beds themselves should now be properly defined as rock-stratigraphic units, especially in view of the greater significance which they have assumed since they were first described.

It is now proposed to establish the Jones Formation1to include that part of the Lower Pleistocene succession at Ross described by Gage (1945, pp. 144–6) under the headings: “R6: Lower Decomposed Conglomerate” and “R7: Laminated Fine Silts and Boulder Clay”. Henceforth, the Ross Glaciation may be considered to be based upon the evidence for climate-deterioration leading to glacierization that is shown by the Jones Formation.

The sections undoubtedly should eventually be re-traversed to fix the formation boundaries precisely, but until this can be done, the sequence as described in Jones Creek upstream from the site of Ross United Claim, in the north-western branch of Jones Creek and along the former Mount Greenland water-race are offered as a provisional composite type locality.

It is intended, for the present at least, to exclude the R.8 beds from the Jones Formation, but the possibility is recognized that less well-sorted conglomerate in the upper part may represent outwash from a glacier which failed to reach the Ross area, or for some other reason failed to leave a positive record. Meanwhile, for convenience, the name Mont d'Or Formation (from an old sluicing-claim of that name) is proposed for the beds previously designated “R8: Upper Decomposed

[Footnote] 1 The term “Ross Beds” was used in tables without definition by Hector on four occasions from as early as 1877 (Fleming, 1959, p. 354), and therefore unfortunately is not available. For the same reason the name “Ross Glacial Member” proposed by Suggate (1959, p. 276) is not favoured.

– 634 –
Picture icon

Fig. 1. —Geologic sketch map of northern end of Mount Greenland Range. (Modified after Morgan, 1908; Gage, 1945; Wellman, 1945.)

Conglomerate” (Gage, 1945, p. 147). No complete, continuous section is known, but the lower part conformably succeeds Jones Formation silt in lower Jones Creek immediately south of Ross township, while the upper part is exposed in the Ross Borough water-supply race and the Mont d'Or sluicing claim. It is not known whether the Mont d'Or Formation is appreciably younger than Lower Nukumaruan.

These formations are embraced by the Old Man Gravels, which is here assigned the status of a Group. Bowen (1957) confirmed and extended earlier tentative correlations between R7 at Ross and glacial beds elsewhere in Westland.

Lower Pleistocene members of the conformable late Tertiary-early Pleistocene sequence, which were mapped together by Gage (1945, Text-fig. 1, p. 139, after Morgan), are separated in the map accompanying this paper. It may be doubted whether the separation should be attempted before the additional boundaries are traced and mapped on the ground. This task will be lengthy and difficult owing to dense vegetation and widespread cover of younger glacial deposits, and it appears unlikely to be undertaken in the near future. As it is, the additional boundaries

– 635 –

may be interpolated with reasonable approximation on the existing small-scale map (Fig. 1). Wellman's interpretation of the Coal Creek section (1945, fig. 1) has been adopted.

Structure and Deformation of the Lower Pleistocene Strata

Mount Greenland consists of an elevated mass of ancient greywacke and slate (Greenland Group) of Lower Palaeozoic or pre-Cambrian age surrounded by Tertiary and Quaternary deposits. Thin outliers and a basal fringe on three sides are composed of Pliocene and Lower Pleistocene beds which are deformed into sharp north-east-plunging folds and faulted in at least one place, and which show a general periclinal arrangement. The outliers are probably more extensive and more numerous than the forest cover has allowed us to discover, but the whole Mount Greenland range may be regarded as an irregularly elevated and almost completely denuded mass of ancient rocks, its relief being due directly to comparatively recent tectonic movement on a large scale. Greenland Spur and Malfroy Spur thus appear to be denuded cores of anticlines or differentially-elevated sub-blocks of the Greenland mass.

In these circumstances it is held justifiable to extend the Greenland and Malfroy anticlines (Fig. 1) beyond the known limits of surviving Pliocene-Lower Pleistocene cover in order to demonstrate the character of the later deformation. The trends of these folds conform generally with the locally very uniform north-westerly strike of the steeply-dipping greywacke basement rocks. The undermass may therefore have been deformed by the mechanism of shear-folding or closely-spaced bedding-faults, while the incompetent covering strata accommodated themselves to the displacements largely by folding. But this cannot be the whole story, for Wellman (1945, pp. 10, 12) showed that the Lower Tertiary rocks in Coal Creek were sharply folded in mid-Tertiary times about an axis transverse to the trend of Greenland Group rocks.

The timing of the later deformation at Ross may be given approximately in terms of the glacial chronology. Our knowledge of the regional paleogeography is insufficient to confirm the possibility of earlier localized emergence which is suggested by overlap of Waitotaran beds upon greywacke at Ross, but there is no doubt that the climax of the disturbance occurred after the Ross Glaciation (Lower Nukumaruan) and was, therefore, part of the major paroxysm of the Kaikoura movements in early Pleistocene times. Deposits from the Otiran Glaciation (Gage and Suggate, 1958: correlated with “Last Glacial” of the northern hemisphere) are warped and faulted at several places along the Alpine Fault (e.g., see Wellman, 1955, p. 38, “Paringa Formation”; Bowen, 1954), but are not perceptibly deformed near Ross. Gravel remnants on benches on the valley sides high above the lower gorge of the Mikonui River, and ascribed to glacial deposition during the Waimaungan Glaciation (Gage and Suggate, 1958, p. 594; probably equivalent to “Penultimate Glacial”), are more than 2,000 feet above sea-level at a distance of six miles from the coast. Moreover, a distant view of the benches gives an impression of upwarping about a north-easterly axis. Additional altitude data are required to verify this impression which, if confirmed, would mean that the uplift of Mount Greenland continued into the earlier part of the late Pleistocene.

Nature of the Ross Glaciation

The Pliocene-Lower Pleistocene succession at Ross embodies the record of a complete cycle of glacier advance and shrinkage complicating the record of final expulsion of the sea from the North Westland region. The coastline migrated north-wetswards partly as a direct result of emergence, and partly as an indirect result through progradation promoted by the abundant transfer of waste from the rapidly

– 636 –

rising alpine region. In some places at least the material was contributed in the form of pro-glacial outwash gravels. The coincidence cannot be ignored, and one must count the growing elevation of the land as a direct factor promoting the first Pleistocene glacierization of New Zealand, but not the sole cause, for the Ross glacial cycle begins with the floral indications of cooler conditions in the basal R6 lignite which was deposited near sea-level. Allochthonous angular bounders near the top of R6 have been taken to indicate proximity of glacier ice, while the till and sub-glacial or ice-contact water-laid deposits at the base of R7 mark the actual invasion by the glacier. Laminated silts at the top of R7 indicate recession of the ice, leaving a pro-glacial lake impounded probably by terminal moraine. The cycle is completed with the infilling of the lake, and sufficient amelioration of climate to encourage recolonization of the region by plants, as indicated by a peaty layer at the base of the Mont d'Or Formation.

Correlatives of the Jones Formation are recognized elsewhere in the region, but so much has been eroded or covered by younger deposits that we cannot picture either the details of the early Pleistocene landscape, or the character of the Ross Glaciation. It is not known, for example, whether the ice responsible for the Jones glacial beds was at any stage continuous with that responsible for similar Lower Pleistocene glacial beds farther north at Humphreys Gully and Findlay Creek (Bowen, 1957) and if so, whether continuity would imply a Ross ice-cap or merely piedmont coalescence of valley glacier tongues such as developed in Westland later in the Pleistocene. One can, however, visualize an early stage in the mountain-building when dissection was incomplete, the relief less severe, and the scenery less strongly alpine in aspect than in the later Quaternary,* so that there may have been extensive plateaus upon which ice-caps or plateau-glaciers could have been generated.


Bowen, F.E., 1954. Late Pleistocene and Recent Vertical Movement on the Alpine Fault. N.Z. J. Sci. Tech., B 35, pp. 390–7.

Bowen, F. E., 1957. Early Pleistocene Glaciation in Westland. Austr. N. Z. Assoc. Adv. Sc., 32nd Meeting, Dunedin, Sec. C, Abstr. C 26.

Couper, R. A., and McQueen, D. R., 1954. Pliocene and Pleistocene Plant Fossils of New Zealand and their Climatic Interpretation. N.Z. J. Sci. Tech., B 35, pp. 398–420.

Finlay, H. J., 1947. The Foraminiferal Evidence for Tertiary Trans-Tasman Correlation. Trans. roy. Soc. N.Z., 76, pp. 327–52.

Fleming, C. A., 1944. Molluscan Evidence of Pliocene Climatic Changes in New Zealand . Trans. roy. Soc. N.Z., 74, pp. 207–220.

— 1953. New Evidence for World Correlation of the Marine Pliocene. Austr. J. Sci., 15, pp. 135–6.

— 1956. Quaternary Geochronology in New Zealand . Actes IV Congr. Internat. Quaternaire, Rome-Pisa, 1953, pp. 925–30.

Gage, M., 1945. The Tertiary and Quaternary Geology of Ross, Westland. Trans. roy Soc. N.Z., 75, pp. 138–59.

— 1959. Ross Glacial Stage, in Lexique Stratigr. Internat., 6 (Océanie), (4) New Zealand: Paris.

— and Suggate, R. P., 1958. Glacial Chronology of the New Zealand Pleistocene. Geol. Soc. America Bull., 69, pp. 589–98.

— and Wellman, H. W., 1944. The Geology of Koiterangi Hill, Westland. Trans. roy Soc. N.Z., 73, pp. 351–64.

Hutton, F. W., 1872. On the Date of the Last Glacier Period in New Zealand, and the Formation of Lake Wakatipu. Trans. N. Z. Inst., 5, p. 384.

[Footnote] * This suggestion was advanced by Dr. H. W. Wellman during conversation some years ago*.

– 637 –

Morgan, P. G., 1908. The Geology of the Mikonui Subdivision, North Westland. N.Z. geol. Surv. Bull. n.s., 6.

Richmond, G. M., et al., 1959. Application of Stratigraphic Classification and Nomenclature to the Quaternary (Report 6—American Commission on Stratigraphic Nomenclature). American Ass. Petrol. Geol. Bull., 43, pp. 663–75.

Suggate, R. P., 1959. Old Man Gravels, in Lexique Stratigraphique International, 6 (Océanie), (4) (New Zealand): Paris.

— 1960. Time-stratigraphic Subdivision of the Quaternary, as Viewed from New New Zealand. Quaternaria, V.

Wellman, H. W., 1945. The Geology of Coal Creek, Ross. N.Z. J. Sci. Tech., B 27, pp. 8–14.

— 1951. The Geology of Bruce Bay-Haast River, South Westland, N.Z. geol. Surv. Bull., 48.

— 1955. The Geology between Bruce Bay and Haast River, South Westland. N.Z. geol. Surv. Bull., 48 (2nd edition).

Dr. Maxwell Gage,

Department of Geology,
University of Canterbury.