Early investigations of the Permo-Carboniferous glaciation of South Africa

A.G. Bain (1797–1864) was probably the first to describe the deposits of the Permo-Carboniferous glaciation of South Africa in 1844, but still attributed their formation to a volcanic origin. It was not until P.C. Sutherland (1822–1908) in 1868 and 1870 that the series was recognised as a glacial formation. J. E. Dunn (1844–1937) named this deposit the Dwyka Series or Dwyka Conglomerate after the Dwyka River near Prince Albert in South Africa in 1886. This series contains scratched boulders and varved sediments, and the basement is characterised by rounded boulders and striated surfaces as evidence of glaciation.


Introduction
The end of the Carboniferous and the beginning of the Permian time is the period of the Permo-Carboniferous Ice Age, which lasted between 310 and 285 million years.While the northern hemisphere was desert, the mainland of the large southern continent (Gondwana) was covered with ice.
The southern hemisphere continents have indications of extensive glaciations in common.Thus, in various horizons, solidified conglomerates, the tillites, occur which scratched the underlaying bedrock.The Permo-Carboniferous ice age exhibit alternations of warmer interglacial periods, as later in the Pleistocene.Stratigraphically, the glacial deposits of South Africa belong to the Karoo Supergroup, which ranges in time from the Upper Carboniferous to the Middle Jurassic (Catuneanu et al. 2002).Within this super group, the glaciogenic Dwyka group (Dunn 1886) represents the oldest deposits and chronologically comprises the Upper Carboniferous stages: Moscovian, Kasimovian and Gizelian, and the Lower Permian stages: Asselian and Sakmarian (Catuneanu et al. 2002;Gradstein and Ogg 2004).Almost all recent publications lack information on the early investigations of the Southern Hemisphere glaciation.Therefore, this will be briefly presented here and, in addition to original citations (in italics), the biography of the researchers themselves will be sketched.

Biographies
Andrew Geddes Bain 1797-1854 (Fig. 1) The exact date of his birth is unknown, but the date of his baptism is on 11th June 1797 in Thurso in Scotland.Nothing is known about his childhood and youth.
In 1820, Bain emigrated to the Cape Colony and first worked as a saddler in Graaf Reinet.He was recruited as commander of a battalion during the border war between the British colonial troops and the Xhosa warriors from 1833 to 1834.Here, he showed his engineering skills, which provided him a job as an assistant engineer within the Royal Engineers, who were engaged in the construction of military roads in the Cape Colony.During his first project in 1837, Bain discovered fossil remains of small reptiles and petrified wood during the construction of the Queens Road, between Grahamstown and Fort Beaufort.This prompted him to become a self-taught geologist and palaeontologist.In 1845, together with William Guybon Atherstone, he found the first dinosaur (Paranthodon africanus) in South Africa.
In the course of further road development, especially in the Western Cape, Bain investigated the stratigraphic 1 3 sequences of the younger Palaeozoic and finally produced the first geological map of South Africa (1845), in which the series known today as witnesses of the Ice Age are still described as claystone porphyry (Fig. 2).His work in the field of fossils earned him the reputation of being the founder of palaeontological research in South Africa.A handwritten list from 1846 of Bain's South African fossils, consisting of 31 plates, is preserved in the British Museum; part of his estate is kept in the archives of the Witwatersrand University of Johannesburg.Andrew G. Bain died in Cape Town on 20th of October 1864.(https:// en.wikip edia.org/ wiki/ Andrew_ Geddes_ Bain).
Peter Cormack Sutherland 1822-1900 (Fig. 3) He was born on April 7th 1822 in Latheran, County Caithness, in Scotland.His parents were peasants.In 1830, the family emigrated to Nova Scotia in Canada.There, he attended the school in Pictow, where coal was found.Disappointed staying in Pictow, he returned back to Scotland and attended now a Latheran parish school.In 1842, he enrolled as a medical student at King's College in Aberdeen.During these years at college, he travelled during teaching holidays first to West Africa as a chemist to study guano fertiliser deposits and later to Greenland as a military physician with whalers.Here, he met the polar explorer Sir John Franklin.In 1847, he graduated as a M. D. from Aberdeen University and obtained the Royal College diploma as medical officer and ship's surgeon.In 1853, he settled in Natal (South Africa) where he worked as a geodesist, military surgeon, and finally as state geologist for 33 years.Sutherland designed the first accurate map of Natal, which he successfully used in mineral exploration (Rosenthal 1966).
Although Sutherland worked primarily as geologist, he unofficially remained ship's doctor on the Sophia in the search for the missing polar explorer John Franklin (Kock and Krüger 1968-87).In 1857, he married Rebecca Urquart Leask and later Jane Garden Blaikie and they had two sons and two daughters.On November 20, 1900, Sutherland died in Pietermaritzburg in the Province of South Africa.

The first descriptions
A.G. Bain begins his description (1844, p. 53) with the words: "Having been employed, during the last seven years, under the Officers of the Corps of Royal Engineers, in superintending the construction of military roads in the colony of the Cape, more especially on its eastern frontier, and having also travelled far beyond that frontier in a northerly direction, I have had opportunities of observing the geological structure of that part of South Africa; and I venture therefore, although only a self-taught geologist, to submit the following observations to the Geological Society.
My principal field of research has been the tract of country extending north-wards from the sea-coast of the county of Albany to the heads of the rivers which enter the sea on that coast*.The sea-boundary of this county, commencing about 450 miles to the east of Cape Town, at the mouth of the Boschman's river, runs in a northeasterly direction about seventy miles, to the mouth of the Keiskamma river.In this length of coast are the mouths of the Great Fish and Gualana rivers, the former about 500 miles east of Cape Town, the latter about fifteen miles further to the north-east.The portion of the tract of country above described which I have examined with most attention, lies between the coast and the northern foot of the Winterberg mountain, whose summit is at the distance of nearly ninety miles from the sea.Respecting the country further in the interior I have also given some geological notices."This is followed by a brief description of the sequence of strata, which initially consists of a reddish quartzitic sandstone interbedded with talciferous shales.Bain was able to assign plant remains in these sediments with great probability to the genus Lepidodendron and thus attest a Carboniferous age to the series.He then continues on p. 54 as follows: "The next rock, in ascending order, is a claystone porphyry, consisting of a clear blue base, with numerous imbedded pebbles of quartzite, claystone, etc.It occupies the troughs in the undulating sandstone just described, which are bounded by parallel mountain-ranges of the same rock.In the hollows of some of these troughs it is from 300 to 500 feet thick, and, when seen in mass, has no appearance of bedding.It thins off towards the sides of the troughs.It often appears on the surface in small tongue-shaped portions, projecting from the ground at an angle of from 70° to 80°, and looking like gravestones in a churchyard.
I have traced one of the ridges of this rock through a distance of eighty miles, from the point where the ridge emerges from the sea, at the mouth of the Gualana river, to the point where the Great Fish and Little Fish rivers unite.The formation extends, I believe, much further both to the east and west.Its northern boundary, like that of the quartzose sandstone, passes eight miles to the north of Graham's Town." This series, described by Bain as claystone porphyry and not yet recognised by him as a glacial deposit, but attributed to a volcanic origin, is nevertheless marked in his summary writing (1845, pp.186), published a year later, with the note that the clasts it contains show no signs of high temperature action, as would be expected in volcanic deposits.The amorphous masses of the firmly lithified, thick marls between the components, however, let him stick to the volcanic origin.
"From the numerous imbedded pebbles of granite, sandstone, quartz, and clayslate, apparently not altered by heat, one might be led to believe that the whole was an aqueous deposit; but after inspecting the magnificent amorphous masses of hundreds of feet in height, as seen at Pluto's Vale in Albany, Toverberg in the Western Karoo, and at Klip Rug in Hantam, with many other splendid sections, where not the smallest sign of stratification appears, that idea also falls to the ground.
In this dilemma, I trust, I may not be considered visionary if I attribute the whole to the production of an immense volcano, which we may suppose to have existed somewhere near the junction of the Vaal and Orange Rivers, or perhaps about the site of the present Compass Berg, whose peak rises to the height of 10,000 feet above the sea-level; and thence deluged with fiery billows the Silurian (?) plains, and spread ruin and desolation over the carboniferous forests for tens of thousands of square miles.….Besides the imbedded pebbles, &c, above mentioned, I found at one place, near Zout Kloof in the Karoo, a great number of rounded calcareous nodules imbedded in this rock, from 3 inches to a foot in diameter, some of them being perfectly spherical, and resembling cannon-balls or bomb-shells; these balls I have never found in any other locality."

The first interpretations
Sutherland, in his short report Notes on the Geology of Natal (1855, p. 466), expresses doubts about the volcanic nature of the sequence, but did not yet draw any other conclusion: "At first I had considerable doubt about the true character of the trachytic rock, but they were removed by finding the sandstone beds scored and grooved where the superincumbent erupted matter had passed over them in a state of semi fusion.This condition of semi fusion must have conduced to the mechanical suspension of the granitic fragments.Where the grooves and scratches were observed, the inclination of the strata is 1U17r.Removal of the erupted matter by the denuding agency of the weather reveals the grooved surface of the stratified rock, which has been rendered the more enduring by contact with the former." First in the more comprehensive publication by Sutherland (1868, p. 17), we note a description of rocks with glacially abraded and polished surfaces: "It passes downwards into a very extraordinary brecciated formation which hitherto has proved the bête noir of South African geology.It is composed of a greyish blue argillaceous matrix containing in a state of dispersion of wondrous uniformity numerous fragments of granite, gneiss, graphite, and their other mineral associates, also quartzite, greenstone, and clay slate.Some of the included fragments, judging by measurement, weigh upwards of five tons; others are not the size of a pea, and the great mass is composed of indurated clay in a state of levigation of more or less fineness.It passes on one side into the Pietermaritzburg shale, and on the other it abuts upon the next succeeding formation underneath, upon which it appears to rest.A very good section may be seen in the road escarpment leading down to the bridge over Okkert's Spruit.The gradation of the one rock into the other is so extended that the exact point where the one ends and the other begins is impossible of definition.The included fragments are smoothed as if by the polishing attrition of muddy sediment, but not rounded as if they had been exposed to the friction pebbles usually sustain on an exposed sea beach." Following this description, he then concludes on p. 18: "In the present state of geological knowledge the deposition of this formation cannot be accounted for except by reference to glacial action.It is by the action of glaciers, coast ice, and icebergs, and by it alone, that fine silt and boulders of many tons weight can be deposited simultaneously on the same sea or lake bottom.The great Scandinavian drift is precisely the same in mechanical composition as the boulder clay of Natal, with which we are now dealing.Professor Ramsay [from Finland] has assigned certain breccias of Permian age to glacial action; there is, therefore, no reason why our quaestio vexata should any longer remain unsettled.Mr. Page does not hesitate to class all our coal-bearing rocks, and their included Saurian remains and plants, with the Permian system.This boulder formation passes conformably and imperceptibly into the Pietermaritzburg strata, which are continued with like imperception into the sandstones and shales containing these remains; we are, therefore, but following out a reasonable analogy." This interpretation of Sutherland is given in Dr. Mann's report in the Proceedings of the Geological Society of London (25 May 1870, p. 516) as follows: "Dr.Sutherland inclines to think that the transport of vast massive blocks of several tons' weight, the scoring of the subjacent surfaces of sandstone, and the simultaneous deposition of minute sandgrains and large boulders in the same matrix, all point to one agency as the only one which can be rationally admitted to account satisfactorily for the presence of this remarkable formation in the situations in which it is found.He believes that the boulder bearing clay of Natal is of analogous nature to the great Scandinavian drift, to which it is certainly intimately allied in intrinsic mineralogical character; that it is virtually a vast moraine of olden time; and that ice, in some form or other, has had to do with its formation, at least so far as the deposition of the imbedded fragments in the amorphous matrix are concerned.He dwells particularly upon the fact that Prof. Ramsay has already assigned certain breccias of Permian age to glacial periods and agency, and that there is good reason for referring the coal-bearing shale of Natal, into which this boulder-bearing clay passes almost imperceptibly, to the Permian system.
For these various reasons, Dr. Sutherland submits that the Boulder-clay formation of Natal should be classed with the Permian glacial breccias of Prof. Ramsay."This interpretation is confirmed by Dunn (1886, p. 4), who correlates the glacial deposits with the Dwyka Conglomerate: "Little by little further data have come to light, and lately quite another complexion has been given to the geology of the central portion of this country by the discovery that the Glacial Conglomerate of the writer occurring on the northern side of the Karroo is identical with and a continuation of the Dwyka Conglomerate of the south side of the Karroo." It is to Dunn's credit that he has recorded the distribution of these glacial deposits over a large area in South Africa.He reports (1886, p. 6): "Over the whole area occupied by the lake which would be 175,000 square miles, assuming a length of 700 miles by a mean breadth of 250 miles, the remarkable conglomerate appears to have been deposited.Vast as this tract is it must have been greatly exceeded by the lakes of the subsequent Karroo ages."He also comments on the possible age and geographical position of the glacial deposits (1886: p. 8): "The presence of a glacial conglomerate of such ancient date-probably carboniferous-is in itself an extraordinary fact, for it shows that the time of its formation icebergs of great thickness existed through a lengthened period of time in this inland sea and much nearer to the equator than it would be possible for them now to exist." Dunn is also the discoverer of a corresponding conglomerate in New South Wales in Australia, the Derrinal Conglomerate, which consistently parallels the Dwyka Series in time, providing the first evidence of an extensive southern hemispheric glaciation at the end of the Palaeozoic (1898, p. 208): "The conglomerate continues for some miles further northward along the railway towards Warwick.Similar deposits of the Derrinal Conglomerate should be found still further north, both E. and W. of the Dividing Range.lf the occurrence of the same conglomerate in South Africa (the Dwyka Conglomerate) is any guide then the Derrinal Conglomerate may be looked for at intervals right round Australia at the base of the Permo-Carboniferous, and later rocks that occupy so great an area of the interior of the continent.
Apart from the scientific interest which centres in the Derrinal Conglomerate as a geological benchmark connecting Australia with South Africa and possibly Asia and from the speculation it excites as to what conditions would have prevailed to produce a conglomerate so protean in its aspects, there is the very practical fact that the Derrinal Conglumerate forms the floor of all our valuable coal seams, and, what is almost as important in this droughty land, also the floor of the strata from which artesian water is obtainable." Hans Cloos, whose biography has been excellently presented by Seibold and Seibold (2000), reported in the Geologische Rundschau (1915) and already a year earlier in his habilitation lecture at the University of Marburg on the "Glacial Formations of the Cape County", with illustrations of the bed loads.His almost poetic words from his book "Gespräch mit der Erde" (1947, pp. 35-38) may conclude this short essay: "There are three points in South Africa that are famous for this [glaciation]: Prieska on the Orange, Riverton on the lower Vaal and Dwyka at the northern foot of the Cape Mountains.The ground moraine got its world-famous name from the river Dwyka (pronounced: Duaika).We drive to Riverton to see the Dwyka Conglomerate.
By car, by boat or on foot from the small station we reach the place where the Dwyka conglomerate and its old base are exposed under loose gravel on the shore from the small station across a table-level steppe with grass and bacopa.I knew pictures and descriptions of this geological pilgrimage site.Reality surpassed it: there is a low layer of rock, small and large boulders protrude from it like warts.You want to pick them up, but they're stuck.Otherwise everything is like in a gravel pit on the edge of the Alps or in the northern moraines of Germany.
The bed is of hard, black-green igneous rock and forms smooth undulations and shields; sharp, straight furrows run through it as if cut.When the moraine is lifted off, it has ridges on its underside that fit into the furrows.Grooves and waves run east to west and so did the ice." …… "It must have been a thick ice sheet that covered everything that is now called South Africa.South America, Australia, and western India were also under ice.Slowly, but with irresistible force, continents were eroded, mountains of rubble, sand and mud were carried off into foreign lands.Sheets of ice drifted into a southern polar sea and a rain or hail of sand and stones fell on its floor." "And the causes?How does such a huge glaciation of the entire southern crust of the earth come about when at the same time the northern earth is warm and dry?Does an ice sheet the size of today's Old World even have room on our small planet?Or must we, may we, gather the icy bits together, as Wegener's hypothesis does, around the Antarctic core, and bring them together in such a simple way, under the hat of ice?"

Fig. 1 Fig. 2
Fig. 1 Andrew Geddes Bain (1797-1864) adopted from Rogers(1937) Dunn 1844-1937 (Fig. 4) He was born on November 1st 1844 as son of Edward Herbert Dunn of Cheltenham, County Bristol in England.In 1849, he emigrated with his parents to Sydney in Australia.In 1856 the Dunns settled in Beechworth, where Edward John attended the "Church of England School", but was later taught only by a personal tutor.He then worked in an agricultural office doing breeding experiments.In 1864, he joined the Geological Survey under Professor G. H. F. Ulrich and studied geology until 1869.From 1871 to 1886, he compiled a geological map of South Africa and investigated the Cape Colony with the Stormberg coalfield and described the diamond deposits of Kimberley in 1871.At the same time, he came to the first findings of a glaciation in South Africa.For his geological research in South Africa and Australia, Dunn received the Murchison Medal from the Royal Geological Society in 1905.He was also Bessemer Medalist of the Royal College of Science in London (Johns and Roach 1934).From 1871 to 1886, Dunn was State Geologist of Cape Colony and from 1904 to 1912, he served as Director of the Geological Survey of Victoria (Johns and Roach 1934).In 1873, he went to London and studied at the Academy of Mines.Back in Victoria in 1886, he went into private practice in the Korumburra coalfield (Christholm, 1958).In 1929, he published a text on the Geology of Gold.He donated his collection of Victorian provincial rocks to the Mining Department at the Melbourne Museum.In 1875, he married Elisabeth Julie Perchard, and they had one son and two daughters.He died in Australia on April 4, 1937.