!53 Palaeont. afr., 23, 153-171 (1980) LAST INTERGLACIAL SHORELINES IN THE SOUTH CAPE by Oliver Davies Natal Museum, Pietermarit;:;burf! CONTENTS Page INTRODUCTION ................ ... ...... ........ ..... ... .. .. ........ ........ ...... .. ..... ..... ..... .... ...... .. ........... 153 THE SOUTH CAPE .................... .. .......... ........ .... ... ..... ........ ........ .... ... .. ..... ....... .......... .. .. .. 154 HIGH SEA LEVEL AFTER EEM 1.. ................. .... .. ...... .. .. ........ ......... .......... .. .. .. ... .. ......... 164 EEMIAN CURVE OF SEA LEVEL .... .. ... .. .. .. ............ .. .. .. ...... .. ................... ..... .. ......... .. ... 165 Camp Century ice core, Greenland .... .. .. .............. .. .... ..... .. .. .. ... .. ........... .. .. ...... .. ... .. ........ ,165 Barbados .. ....... . .. .. ...... . .... . ............. ... .. . .. . ... .... ........ ... .. .... ....... . ... . .... . .. ..... . .. .. ... . ...... .... .. .. 165 Huon Peninsula, N ew Guinea ...... . ... .. ..... ...... . ......................... .. .. .. ....... ... ........... ............ . 166 California .. ..... . ... .. .. ......... ... .. ... .. ... ...... .. ........ .... . .. .. .. .. .. . ... ... ......... ........... ... .. .. .. .. ........ ..... 166 General .. . ......... .. ... . .......... .. ...... .. .. ...... ... ... .. ... . ....... .................... .. .......... . ... . ........ .... ...... .. 166 ACKNOWLEDGEMENTS .... .... .......... .. .. .. .. ....... .. .... ... .... .... ..... .............. .. ..... .................. 167 REFERENCES .................. ..... .. .... ... .... .. ....... .. .. ... .. ... ................. .... ..... ... .. ........... .............. 167 INTRODUCTION My paper discusses the Last Interglacial shore­ lines of the South Cape with reference to findings from the rest of the world. I call them Eemian shorelines, and define Eem as an interglacial with three peaks of sea level separated by two stadials when the ocean dropped. I use the term Eem though it has been wrongly applied in north-west Europe (Kukla 1977). Other names have been given to this interglacial; in the deep-ocean record the stage is referred to as Stage X or Stage 5. It has been dated mostly by Uranium/Thorium as lasting from about 125 000 to 80 000 B.P. The dates and the fluctuations will be discussed below. The Eemian was preceded by a glacial stage known in the Alps as Riss and in North Europe as Saale (Warthe), during which there were one or two interstadials perhaps of interglacial rank; dates are uncertain. The Eemian was followed by the last glaciation, Wiirm or Weichsel, which con­ tained two well-defined interstadials less warm than today and several minor fluctuations, and it lasted from soon after 80 000 to 10 000 B.P. The Eemian interglacial began with very rapid warm­ ing (Shackleton 1969, Ruddiman 1977a) and ended with very rapid cooling (Lalou et al., 1971 , Mcintyre et al., 1972). I avoid the presumed equivalence of the North American Sangamon, since Cooke ( 1973) has suggested that the Sanga­ mon is much older and that the period 125 000- 80 000 B.P. was an interstadial of the last or Wis­ consin glaciation. I refer to the three warm peaks of Eem as Eem I, the oldest, Eem II and Eem III. Other names have been given: for example Barbados III, II and I; Stages 5 e-a for deep-ocean cores; for the pollen­ record Q ,R ,S in Macedonia (Wijmstra 1969); Eem, St. Germain I and II in the Vosges (Wail­ lard 1978). Far more confusing has been the pot-pourri of names and dates given to support varying theories of climatic sequence. It is generally agreed that the whole Eem lies beyond the range of radiocarbon, but extreme finite dates of about 68 000, 64 000 and 60 000 (Woillard 1978) have been obtained for the first Wiirm interstadial: three peaks existed, Amersfoort, Brorup - the warmest - and Odde­ rade. It is also agreed that molluscs are liable to contamination, so high finite dates on them may be invalid. Most thorium dates have statistical er­ rors of 10 000 years or more so cannot be used for assignation to any stage as short as an interstadial. Corals are used, but molluscs are chemically sus­ pect (Kaufmann 1971). Sequences of pollen and of stable isotopes in deep-sea cores indicate four stages of decreasing warmth separated by colder stages; this pattern suits the sequence Eem I-III and Brorup. But contrary to Shotton (Shackleton 1977a), French scholars have challenged the finite dates of the Brorup complex and have moved Bro rup back to what I call Eem II and/ or III; so Du­ plessy ( 1976, cp. Matthews 1973) equates his Eem with my Eem I, Amersfoort with my Eem II and Brorup-Odderade with my Eem III. Marner (1972a, b) starts the Wiirm at the end of my Eem II and equates my Eem III with Brorup. Confu­ sion has reigned also in the Mediterranean nomen­ clature. Three stages have been named (Bonifay and Mars 1959): Tyrrhenian I or Paleotyrrhenian which is nearly always pre-Eem; Tyrrhenian II or Eutyrrhenian, probably my Eem I; Tyrrhenian III or Neotyrrhenian which is almost certainly Eem II or III but frequently defined as a Wiirm intersta­ dial (e.g. De Lumley 1976: 311- 25). Keraudren !54 ( 1970-2) rejects the Paleotyrrhenian, and Perthui­ sot ( 1972) jumbles up the nomenclature. A further source of confusion is the transference of names from the emerging Moroccan Atlantic coast (An­ gelier et al., 1976) to the much less stable Mediter­ ranean coast of Morocco and Algeria. THE SOUTH CAPE In estuaries of the South Cape between Coega and Mossel Bay and apparently as far west as Ar­ niston beds occur with warm water fauna at peak altitudes of +4-+9 m. Some have probably been eroded, and all would have been laid down under water, so M.S.L. would have been over +9 m and high water (H.W.) about +10m. In addition to many modern species of mollusc these beds have yielded several species which today live in the tropics: Cerithium scabridum rufonodulosum (Gulf of Oman to Mascarenes) Pitar madecassinus (Natal) Paphia textilis (Natal) Rhinoclavis kochi (Natal) Anodonta edentula (Natal) Mactra ovalina (Natal) Tapes sulcarius (Indo-Pacific) Pseudostomatella orbiculata (Indo-Pacific) The beds have also yielded six species which today are found in West Africa but not in the South Cape (Tankard 1975; Kilburn and Tankard 1975) and several species now extinct: Cantharidus suarezensis jultoni (formerly Calliostoma mosselense) M onilea ponsonbyi Diala infrasulcata Pupa daviesi. I have named this fauna the Swartkops fauna from several sites on the Swartkops River near Port Elizabeth. Such beds have not been identified in all estu­ aries. Some such as Keurbooms River are very narrow; others, as perhaps Bietou River, have no exposures. In dredging the incised trench of the latter, none of the typical species were found (Schwarz 1899). Swartkops beds are unknown east of the Coega River where many estuaries are steep and narrow; only modern species occur in estu­ arine collections at +9 m from Richmond Strand and Kleinemonde and at +6 m at Port Alfred (Mountain 1962 ). Fauna of perhaps slightly warmer conditions than today's has been found at +9 min the Great Fish estuary. Those species that today live north of Natal have not been found in the poorly exposed and decalcified estuarine beds at +6-+9 min this province. West of Mossel Bay molluscs now extinct or no longer in the area have been reported from estu­ arine beds at Sandvlei near Cape Town (Davies 1972: 277) and at Wortel Gat on the Klein River (Barnard 1962); but there are no typical members of the Swartkops fauna. Indurated estuarine beds at the Bree mouth, probably Late Pleistocene, have yielded only modern species (Wybergh 1919). In the Western Cape a warm fauna occurs on open coast sites much older than Eemian; but the typ­ ical species are not the same, and their extinction is unconnected with the appearance of the Swart­ kops fauna. In estuaries between the Olifants River and Langebaan Lagoon at +5-+7 m there are several West African but no Indo-Pacific species. Warm water forms occur on ramparts at about + 7 m at the mouth of Saldanha Bay (Tan­ kard 1976). On open coast sites in the South Cape where the top of the beach is traceable at about +9 m (i.e. M.S.L. about +8 m) only modern fauna is found except for single specimens from King Neptune, Ferreiratown and perhaps Frog Rock at Mossel Bay. There is confusion about the exposures at the last site and two shorelines may be mixed. I col­ lected modern fauna and one archaic species from a beach up to + 17,4 m; Goodwin (1933) claimed Cantharidus suarezensis jultoni from a beach up to +8, 7 m. This site is not estuarine. Several other species in the South African Museum (Kilburn and Tankard 1975) may have come· from some­ where in the general district and not specifically from Frog Rock. Artefacts have been found on few of the estu­ arine sites with Swartkops fauna. The most signifi­ cant is Mouritzkraal at the Gamtoos mouth ex­ posed in a road cutting (fig. l) . I interpret the beds above 4,3 m (perhaps +6 m M.S.L.) as the tidal estuarine beach of a trans­ gressing sea on which men at low tide collected Donax and left their artefacts. As the exposed sur­ face of the water-laid beds may be eroded, maxi­ mum M.S.L. may have been +8 or +9 m . The ar­ tefacts are typically Middle Stone Age (fig. 2), so this industry must have been well established be­ fore the maximum of the transgression. At Coega Salnova I found artefacts on and just above the surface of the estuarine beds at + 7,2 m in red dune sand (fig. 3). These are suggestive of a late Acheulian biface industry rather than of the Middle Stone Age; but as in Europe, there must have been a transitional period when small bifaces were current along with Middle Stone Age flake and blade tools. There are two small bifaces from a gravel at +9 m at Gibraltar Rock on the Great Fish estuary (Davies 1971: fig. 6:2). At the mouth of the Klein Brak River a shelly indurated sand is exposed up to +2,5 m. I found only modern fauna; but Haughton (1937) reports, apparently from the same bed, one Swartkops species. The sand is probably conformable with the lagoon deposits 3 km inland which rise to +8 m and contain abundant Swartkops fauna. With the sand at the mouth I found a lightly rolled core, probably Middle Stone Age (fig. 4). Evidence from other sites is less definite. In the caves at Die Kelders and Klasies River lower, Middle Stone Age remains overlay an 8-metre ter­ race, not necessarily conformably. At Redhouse I above 5·1 flood- plain, m. 5·0 4·9 4·8 4·7 4·6 4·5 4·4 4·3 BP-K o<:;::)o~ 0 0~ 0 CJ D 0 0 0 0 \) 0 0 Clc? D D 0 0 0 \) 0 0 a 0 o a 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o~oQOO {fu rather gritty C6 \) Q very fine 0 0 sand White dune- sand, no shells ? Erosion- surface Harder sand many Assiminea 2 sharp flakes, a few Donax Many Assiminea 155 Fine water- laid sand, many very small Assimi nea and other mollusc fragments Soft sand, flakes, cores, set pebbles many Donax serra Varved water- laid si Its and sands Scattered shells, incl. Loripes liratula ( Swartkops- fauna ) BASE NOT EXPOSED MOURITZKRAAL, GAMTOOS MOUTH, SECTION Flood- plain estimated I· 5 - 2 m above m.s.l. Figure 1. !56 em 3. 78/113 +6,4 m m.s.l. 78/111, +6,0 m m.s.l. Figure 2. Mouritzkraal, Gamtoos Mouth. Fluctuations at top ofTransgression. found Middle Stone Age pieces in sand above but not on the estuarine clay. In a shingle-bar up to +9 m at the mouth of the Bietou River Macfarlane ( 1958) claimed to have found rolled Middle Stone Age flakes and an unrolled cleaver that should have been Acheulian; it may have been derived, and this author's reports are particularly unsys­ tematic. Vol man ( 1978) reports Middle Stone Age artefacts from a midden on a rock terrace at +8 m at Sea Harvest, Saldanha, but the terrace is prob­ ably Early-Middle Pleistocene (Kilburn and Tan­ kard 1975: 214). Numerous wave cut platforms at +8-+9 m are known on the open coasts of the South Cape; some carry pebbles, modern species of mollusc and sand. Artefacts are scarce. In general they are un­ rolled Middle Stone Age pieces subsequent, there­ fore, to the maximum transgression; for instance on storm ridges at + 7,5 m at Cape Hangklip (Mabbutt 1954). Rolled Middle Stone Age flakes occur on a beach at Nahoon Point which passes beneath a Late Pleistocene dune so that its crest is concealed; perhaps also on a storm ridge about +5 m at Kidds Beach (Macfarlane 1937: 900-1). !57 em Figure 3. Coegasmond Salnova Artefacts at base of red sand, resting at + 7,1 7 m . m .s.l. on top estuarine bed with Swartkops fauna. Similarly, in Natal rolled flakes were cemented to a beach ridge up to +6 m at Tiger Rocks (Beater 1967); small rolled bifaces comparable to those from Coega Salnova were found at +5,5 m in an estuarine gravel at Ngane Mouth (Davies 1970: figs. 41-2); near Durban there were lightly rolled and unrolled picks of the Tugela industry, which I regard as an early facies of the Middle Stone Age. Rolled bifaces have been found in similar gravels in the South Cape at Kleinemonde and Pletten- berg Bay (Davies 1972: figs. 11, 13); they are older than the marine beds and may have been derived from a continental surface. Unrolled bifaces are doubtfully reported from Plettenberg Bay and Knysna East (Mortelmans 1945), and better authenticated from a 9-metre terrace near the Hlu­ hluwe mouth in Zululand (Davies 1970: 433). I would hesitate to date the latter assemblage as late as the Tugela Industry. Considering the altitude, it is possible that this is an estuarine gravel of the 158 em Figure 4. SC 268, Klein Brak Mouth: Estuarine beach ex­ posed to +2,4 m m.s.l. and rising. N.M. 72/3, rolled core, ventral pebble-surface save one flake. + 18-metre sea. In South West Africa probably only a +2-4m beach (Davies 1973, Wieneke and Rust 1975) is Eemian. The beach above carries Acheulian tools. As the single handaxe from Knysna could have been carried by man to a 9-metre beach, it can be inferred that, apart from a few very small bifaces, all products of the Acheulian Industry had ceased before the maximum +8-9 m transgression. The Middle Stone Age and Tugela Industry were fully established on all the coasts of South Africa by this time. On the assumption that coastal South Africa, like Morocco (Biberson 1961) and Lebanon (Fleisch 1956), has undergone regular epeirogene­ sis during the Pleistocene and that therefore the crest of each transgression is older than that of the one below it, I have identified both estuarine and open shore exposures at + 15-21 m. There is no evidence for or against a theory by Butzer and Helgren that Nelson's Bay Cave at +18m was opened in the Early Eem (Klein 1972). Some re­ ports on artefacts from estuarine gravels at these levels seem to indicate that they were Acheulian, such as New Brighton (Breuil MS: 3) and the Bie­ tou estuary (Macfarlane 1938: 239, S.A. Mus. 6833). Elsewhere undoubted Acheulian pieces have been found: probably unrolled Late Acheu­ lian at Duineplaas (Davies 1972: 237), Gouritz Mouth, Goukamma Drift and perhaps the Harvey­ ton Terrace on Sundays River (Ruddock 1957); and more archaic looking rolled pieces from sev­ eral sites in the Coega estuary (figs. 5-7) and the Kariega estuary (Davies 1972: 227). The same pic­ ture emerges at the upper gravel at Cape Hangklip up to about +18m: numerous unrolled Acheulian pieces have been found, but apparently nothing rolled and nothing of Middle Stone Age technique. In Natal also unrolled, perhaps Late Acheulian, pieces have been found in the Mzeneni and Mvoti estuaries and not unexpectedly unrolled "Sang­ oan" at two or three sites at about +18m. A few sites at +18m have yielded an archaic fauna, but we do not know when certain species became extinct in South Africa. At about + 17 m at Frog Rock, Mossel Bay, I found one Isognomon gaudichaudi, a mullusc assigned to the Early Pleisto­ cene and probably also to the Mio-Pliocene. I found Pitar schwarzi beneath and perhaps earlier than the + 18-metre gravel on the National Road at Coega in a fine silt; this belongs to the early part of the Alexandria Formation, probably Late Miocene to early Middle Pleistocene. A vague and inaccurate report (Engelbrecht et al., 1962) of a mollusc collection from Fishwater Flats on the Swartkops River, said to have been at +9 m, lists several extinct species. I have not identified the site; it may have been a pit exploiting Cretaceous, Miocene and Quaternary beds near the old Gra­ hamstown road. On the other hand, sites far up estuaries near Port Elizabeth have yielded a few specimens of .:.;:_:·: ··. .. :::::' ': -~·~ ... ;:·~·~;." ' . . ~/·· . '%-,!(/~~.,.:· ,. -:·: . ·~:·. ·. : .... :.· }.: :! /.;: .. 1. em 2. ventral flake­ surface .·.·· · ....... :·· ·.:.: . :··.'.;. Figure 5. PE 535, Gravel at +18m m.s.l. exposed on N2 road, west side of R Coega bridge overlooking Salnova tanks. N.M. 68/40, Rolled artefacts in gravel. .· . . ·: .. .. ·.·· (.}> <0 :;tr~}~J~)' ~. unrolled ... · ~- ... ' f· ' .~-' .,, ~', ' ' ' . ,y""" ' - . -~{>'.!: .. : ~-~\ ' / ,..-. fl~ifl~ ... ·:· .. . ~ • ' '," • ,.,r:s, ' ' : N t.J f • :: ... /:'•'•~ ~~ • / ... /;. ~ol~ .- . .· .. . . . / . . , .. • . . .,..,- · · ·)!$. I;\ · i : . ..j;,> ,,. · -. ~~'?;·" .. · • .. · ;.~ :::;~ . / . ~·!loi,;l\; •( .... : . . .. .. . "' . ,. . . . ..... . . . . , .. '·· "' .. · .· "-""' . . ... , ... .. : ' . ". . ., . . . . ~" . . . . : .. . : ..... ·• ~···' · . ' . . . "' ; . . . ·.'. ··'·' ':}: .... : .,ifj·m1il ~~~:~!~·~!i.'t/1. · ,-: .- .. ~~,, ,·1{11 . · - . ·. ·.w, ·~·. ,; . ' .. ,J ,; . . . ....... ... . N•>·' . ... :,.,·. '·•i' . . •· I .. p.. . . . - ~ ·I·• . '"'I ,,:·.: ,~·· ,!,~:/ ·. · ,i ! . r.-. ::~· __ · · · · ~· .,....... · . · ·. ·. 1V·)'~z . . , ... " >): •: ,; ·' . , •. f""'· : : ~ ....... . . . . " ;.,,~'.';f. .. ll ~· _.,.., . . . . . . . ·- ~~~· .( .• ~ .S11hn: .,..-., :~1//;·~~.~.f~' ilt · ,'': . .' · , 1 , ·, .'· j ··" 'J"I · .. ·:i· ·&~\tf/lJI,.~ ·, f!! . i · · . - . . ··. ~~·A ·\ "'~•;. i ' ,, ., .. •!.lit. . ·;· . ., ' • . ~. ~ " M,.,.If .. >!''r . : . . . . . .. , ,~,'~-·~. :::.~_}:'~-. } ·: . ··.:::. ·-· .. _ ..... ·..~-. . . . .,.,.,,,,~: .. l .. .... ,d;;l~ '• .. ~ -.. ~~ .:_ ~ ---. slightly abraded .. ': ·::'-- . .. i .· ... \ ' I unrolled 4. l . ~ .. ::.:·:'· ·.:_:_. .. .. --\ ,. unrolled; ventral flake­ surface ·.··. · · .. .· . cleaver, heavily rolled /J~~t:r\ -~::-~· ·: :: :·.::-; : ·, . .. -_: . ···.· -~ - . ·· · .. .,, .. , 4ifi :1~ ~·,. .. ,·; 1~ :~~\ · ·,· t I~, ·&!\>· \\;:: '!t;..L.:· -,:.}.::: .. ::' .· ~: · ;· .. '·' ;;1,~f~~fP ·.··· ,~~:;r·;!;Jfit'f'+9 m. In the Kariega estuary at +18m is an entirely modern fauna save for one species today not found west of East London. I previously dated the + 18-metre level to Eem I and the widespread beds with Swartkops fauna at +9 m to Eem II. No suitable radiometric test was possible. However, I recently asked Dr. G.H. Miller of Boulder, Colorado, to make racemisa­ tion-tests on Donax serra from three sites in the Swartkops valley (table 1) . He regards the ratio of D-alloisoleucine to L-isoleucine in both free and combined fractions as the most reliable figures for dating: PE-557, Deal Party, bar at mouth of Swartkops River, contemporary with the +8-metre beds up­ stream and containing shells washed down the estuary, rising to about +6, 7 m. "The first two specimens may be 130 000 B.P. or perhaps in the range 160 000--220 000. The third specimen is rather older, probably 160 000." It may be derived from an older beach. PE-553, Cradock, estuarine bed up Papkuils River at + 12,7 m . "Though there are some discrepancies TABLE I Alti- D-alloisoleucine: tude % Free Amino-acids L-isoleucine Site m.s.l. valine isoleucine leucine combined free DEAL PARTY bar 6,7 66 42 67 0,64 1,09 96 65 95 0,81 1,26 70 43 71 0,67 1,18 CRADOCK estuarine silt 12,7 69 42 67 0,41 0,45 0,87 56 25 42 0,41 0,41 0,90 57 24 46 0,31 0,87 BLUEWATER BEACH 7,5 3 2 4 0,06 0,054 0,17 storm-beach 0,054 12 8 14 0,10 0,10 0,32 95 72 93 0,78 1,14 in the figures, it is most realistic to consider all the specimens as of the same age, and markedly younger than those from Deal Party. They may be Eemian, but it cannot be disproved that they date from a post-Eem high sea-level." PE-528, Bluewater Beach, shelly shingle up to +6, 1 m. Owing to the presence of a few Cerithium scabridum, it was assumed that this deposit is a con­ tinuation of the Deal Party bar interrupted by the present mouth of the Swartkops River and banked against its left bank which is sharply incised into older beds of the Alexandria Formation. Dr. Miller reports that the third specimen is of considerable antiquity, perhaps > 160 000, while the first two are much younger and could well be Holocene. In view of these figures, it appears that the bed is a Holocene storm beach which has incorporated molluscs eroded from older beds up the valley. It should be noticed that none of the analysed specimens were of typical Swartkops species. These were too rare from the Cradock site to be spared for dating. Donax serra is common today and probably has been for a long time past in the South Cape. Furthermore, racemisation dating is still on trial and is not yet as dependable as radio­ carbon. Generally, the above evidence permits the fol­ lowing conclusions: A. A transgression to +8-9 m with some estu­ arine beds rather lower contains abundant warm water Swartkops fauna in estuaries but hardly any specimens on the open coast. B. Previous to this transgression there had been a fluctuation; at Coega Salnova sea level dropped below +6 m to permit formation of thick terrestrial calcrete on underlying estu­ arine beds. This fluctuation need not have lasted long. Typical Middle Stone Age material was being used at this time. C. Before this fluctuation there was a transgres­ sion at Coega to at least +6,4 m and at least +5,5 m at the Gamtoos mouth. Warm water Swartkops fauna had already reached the South Cape and several species have been found in beds belonging to this transgression: at Coega Cerithium scabridum, Loripes liratula, Diala infrasulcata; at Gamtoos mouth Loripes li­ ratula . C.E. Stearns ( 1976) suggests two peaks of Eem I, the first <3 m below the second, but his dates are unacceptable; so again on Hawaii (H.T. Stearns 1961), New South Wales and probably elsewhere in the Pacific (Mar­ shall and Thorn 1976). D. Assuming that A-C are part of Eem I, world­ wide evidence (Shackleton 1969, Mangion et al., 1974, Ruddiman 1977a) indicates that this was preceded by a rapid transgression from a very low sea level. In Hawaii there was a peak of glaciation about 140 000 B.P. (Porter et al., 1977). E. At some earlier stage, probably before the fin­ est Late Acheulian pieces were being made, there was a transgression to +18m with ar- chaic looking tools and remnants of a Plio­ Pleistocene fauna. F. Probably considerably later than stages A-C there was a transgression to + 12-15 m with a few specimens of Swartkops fauna, not necess­ arily still living but perhaps eroded from the +8-metre beds and carried up the estuaries. There is evidence for this transgression from other South African sites: 1. At the open-shore site of Skoenmakerskop is a beach sand forming an encrusted terrace at +15m overlying dune. In the superficial beach sand were only fragments of shell, but some displaced slabs contained numerous shells, all living toclay i!'. the area, with fragments ofVer­ metid, a few bones and many echinoid spines. The whole is characteristic of an exposed rocky coast. At the west end of this terrace is another crust of beach rock at +18m. The lower beach rock could equate with the F-beds. 2. Confused accounts (Laidler 1933, Macfarlane 1937) of a former exposure at Blind River, East London, describe a boulder bed from +9,0 to + 11,58 m, perhaps overlain by white marine sand to + 13,7 m, and overlying rolled Middle Stone Age flakes between +8 and +9 m. No significant fauna was found. It would appear that Transgression A overran a surface with Middle Stone Age and was perhaps overrun by Transgression F. The boulder bed should cor­ relate with the cemented beach beneath Na­ hoon Point 1 km to the east. This beach is ex­ posed up to+ 1,5 m and passes beneath a dune dated by radiocarbon about 30 000 B.P. (Mountain 1966); it rests on hard aeolianite. 3. On the Atlantic coast Tankard ( 1975, 1976) found in the Berg River estuary a later trans­ gression lower than the earlier or peak, but claimed that at Churchhaven a lower trans­ gression to + 1,5 m was earlier than a higher to >+5 m after a rapid surge. 4. In Natal Hobday (1974) describes at Port Durnford a spit with lignite lenses up to +18m which he claims are washover deposits, perhaps from an ocean at +8 m. It is difficult to believe that waves 10 m high could be gen­ erated regularly on the Indian Ocean. It is eas­ ier to suppose that the sands and lignites are largely sub-aerial and freshwater deposits, per­ haps contemporary with or later than the 8- metre sea. Subsequently, sea level rose to an altitude from which waves could wash over a bar at+ 18m. 5. Hobday ( 1976) describes on· St. Lucia Promon­ tory three fluctuations of a sea level up to +5,3 m. This is probably Eem I because near Durban it appears that in a later stage of Eem a platform was incised slightly below modern M.S.L. 6. At Mbokadweni Mouth there is probably evi­ dence for fluctuation during Eem. The highest exposure is a potholed platform at +6,3 m with no cliff. Since all beach deposits on it 163 have been eroded, sea level was probably up to +8 m. In front of it at one point is a small de­ posit with modern fauna up to +2,3 m. The original altitude is unknown. It has been widely assumed that the South Afri­ can +8-metre stage A-C is Eem I because this is the approximate level of the peak of Eem I dated by thorium to about 125 000 B.P. at tropical sites round the world; for example in Jamaica (Moore et al., 1974), Florida (Osmond et al., 1965, Lalou 1971), Victoria (Gill and Amin 1975), some stable islands in the Pacific (Marshall and Launay 1978, Bernat et al. , 1976, Sterns H.T. 1974) and a few stable islands in the Indian Ocean (Thomson and Walton 1972, Battistini 1976b). Some authors have tried to put back the peak of Eem I to 140000 B.P. or earlier (Evans 1971; Dai Pra and Stearns 1977); a few dates suggest this also from Antilles (Schubert and Szabo 1978), Tonga (Tay­ lor and Bloom 1977) and New South Wales (Mar­ shall and Thorn 1976). But the large statistical errors of thorium-dating (cf. Veeh 1966) incline one to discount occasional markedly aberrant dates (Shackleton 1971); the 20 000 years separa­ tion of Reef Vllb from VIla at Huon Peninsula can be bridged within the statistical limits (Chap­ pell 1974a). In view of the emergence of the South African coast through the Pleistocene, the identification of the +8-metre stage with Eem I needs to be proved. If it is correct, emergence had ceased by the Eemian stage. At Klasies River Cave bones on the floor at +8 m were dated by racemisation to 89 000-110 000 B.P. (Bada and Deems 1975). This method is under trial, and, until a full report is published, it cannot be proved what unconformity separated the incision of the cave floor from the Middle Stone Age occupation. Butzer's (1978) apriori curve of sea level in Eem contradicts all curves from the rest of the world. Thorium-dates have not been obtained from South Africa owing to lack of suitable material, and no archaeological research has been done on the coral fringed East African shorelines. At Border Cave Middle Stone Age deposits occur below the base of a sequence which climatically may be equated with the Eem­ ian sequences of deep sea cores (Beaumont et al., 1978, this volume, Butzer et al., 1978). Near Ziway in Ethiopia a typologically Middle Stone Age collection has been dated by potassium­ argon as older than 181 000 ± 6 000 B.P. and at a higher level as 149 000 0 ± 13 000 B.P. (Wendorf and Schild 1974, Wendorf et al., 1975). These dates are dubious, owing to the very long half life of K 40 . But if a Middle Stone Age industry was estab­ lished in the South Cape before 125 000 B.P., the Ethiopian dates do not look outrageous. It is not certain whether Middle Stone Age tech­ niques reached North Africa before or during any part of the Eem interglacial. Hand axes and pieces claimed to be Mousterian of Acheulian tradition are reported from Morocco on the Harounian beach (+18m) and beneath the Uljian beach 164 (+8 m) (Choubert 1962, Biberson 1963). Camps ( 1974: 27) lists two sites in western Algeria where a very primitive unrolled Aterian industry occurs in a Strombus beach; but at Falaises Rouges the Aterian lay in a silt well above the beach dated to 140 000 ± 10 000 B.P., and the Aterian level was dated to 31 800 ± 1 900 B.P. (Roubet 1969). There was similar material at La Calle (Morel 1967). I have suggested that the artefacts could have been deposited on the beach between its formation and consolidation and worked down into it (Davies 1964: 41-46). In Lebanon Mousterian pieces occur between two Strombus beaches which may be Eem I and II (Sanlaville 1973). The general opinion in Lebanon and Israel is that the Mousterian started at the be­ ginning of Wiirm I (Wendorf and Marks 1975), but this seems based on a short chronology which equates the Neotyrrhenian with a Wiirm intersta­ dial. The evidence from Western Europe is very vague (Bourdier 1969, Agache 1969). Mousterian probably occurs on, but not below, Eem deposits. However, Shackley ( 1977) reports five sites in England where Mousterian of Acheulian tradition occurs within the Eem deposits at + 7,5 m. He dates this beach as Eem III, but its altitude sug­ gests an earlier stage of the Eemian. So there seems no longer to be any objection to the establishment of Middle Stone Age industries and techniques in South Africa before 125 000 B.P., provided that it is admitted that the Tugela or "Natal Sangoan" Industry is a coastal facies of the Middle Stone Age and is not part of the Acheulian complex. Occasional pieces of Late Acheulian type may have been made or collected by the Tugela people, since such artefacts have been found stratigraphically associated with the Tugela Industry. Since the Tugela Industry occurs regularly on the older red dunes of the Berea Formation in Natal, and not beneath them (Davies 1976), these dunes must be redated as pre-Eem. They were probably originally calcified, but in Eem I the critical point of warmth and humidity was reached to induce almost complete decalcification. In the second stage of dune building a calcified cordon was constructed in places nearer the sea, for example Durban Bluff and St. Lucia Promon­ tory. The cordon is still largely calcified. At the same time red sand was probably derived by collu­ viation or by wind action from the older dune and deposited on the +8-metre terrace and on artefacts of the Tugela Industry. The date of this stage may be after Eem III. HIGH SEA LEVEL AFTER EEM I An explanation and date for the Stage F trans­ gression to+ 12-15 m is still required. The racemi­ sation figures from Cradock imply that it took place a good deal later than Stages A-C, so was not connected with the pre-Eem + 18-metre transgres- sion. I have given some evidence from Natal for this stage. In South Majorca (Butzer and Cuerda 1962) and Lebanon (Sanlaville 1969) a lower transgression is older than a higher, but these may be fluctuations towards the peak of Eem I. I assume that the Quaternary uplift of the South African coastlands has until the Eem been fairly constant without sharp reversals of the trend. If so, the high sea levels of Stages A-C and F must be due to rise in absolute ocean level. Stages A-C can be explained by the warmth of Eem I, which is documented mainly from deep sea cores (Mcin­ tyre et al., 1972) and pollen sequences from temper­ ate Europe (Woillard 1978, Starkel 1977). Stage F cannot be thus explained, since there is much evi­ dence that Eem II and III were cooler than Eem I. Therefore, Transgression F implies a fairly rapid surge which has left traces mainly in sheltered es­ tuaries. Such a surge could be a tidal wave due to vol­ canic explosion, but no marker ash layer exists in the South Indian Ocean. It would also be too rapid to leave more than a veneer of silts and mol­ luscs washed up an estuary. At Cradock I could examine only dumps and ditches left by bulldoz­ ers. There seemed to have been not less than 30 em of silt in situ, and Stow ( 1871) implies banks of considerable thickness which must have accu­ mulated over decades or longer. Wilson (1964, 1969) and Hollin (1965, 1969, 1970, 1972, 1977) have proposed that the Eem in­ terglacial ended in a rapid ice surge from Ant­ arctica which temporarily raised sea level to near +18m. This theory has been generally rejected (Chappell and Thorn 1978) but cautiously ap­ proved by Kennett and Huddles tun ( 1972) and Richards ( 1972) to explain certain poorly dated faunal anomalies in North America. It would also explain anomalous faunal stratigraphy in the Bal­ tic and Netherlands (Vander Heide 1961) and the fact that in south-east England the highest trans­ gression to about +15m was a good deal later than the maximum warmth indicated by the pol­ len curve (Sparks and West 1972, West et al., 1964). A major difficulty is that this theory presup­ poses a single peak of the Eem interglacial, and Wilson ties the surge to an isolation minimum about 115 000 B.P. It therefore does not fit the racemisation figures from Port Elizabeth which imply that the + 12-15 metre transgression was some 40 000 years later than the peak of Eem I (one third of the time 125 000-0 B.P.). Thus the ice-surge might be dated at the end of Eem II (94 000 B.P.) or better of Eem III (79 000 B.P.) but not of Eem I. Wilson and Hollin do not sug­ gest what happened in these later interstadials of Eem and do not show an unconformity between the Eem I marine sequence and the overlying freshwater beds in the Thames estuary. Knox ( 1969) makes it clear that the highest Eemian sea level was at the very end of Eem after a climatic deterioration in which broad-leaved trees were re- placed by conifers. Furthermore, there is no evi­ dence from cores in the southern oceans for a great diffusion of Antarctic icebergs (Thorn 1973). Another possible explanation for the rise of sea level in Stage F in the South Cape and possibly elsewhere could be that after the Eem I peak, dur­ ing the considerable cooling of the Eem stadials, ice accumulated in Antarctica, and the ice-sheet became thicker and wider as sea level dropped. This would slowly cause glacio-isostatic depression under the continent. At the same time the process of ocean deformation described by Clark et al. ( 1978) would operate; and, if the focus was first in Antarctica before the northern ice reached ocean coasts, one may reverse their map (Clark et al., 1978: fig. 15) and expect transgression about 30-40° S and 40-60° N until ice accumulation removed so much water that sea level dropped again. Much slower rheology of the magma would later cause the land to emerge at these latitudes and so depress the coastlands. This could have taken place at the end of Eem III or in Wiirm I, which would fit the racemisation figures. This solution may be acceptable until we know more about rheological processes during an ana­ glacial period. · EEMIAN CURVE OF SEA LEVEL Within the Eem as I have defined it, there is dis­ pute on the sea level curve, due to lack of agreement on methods of temperature evaluation, quantity of global ice and the use of foraminifera and molluscs as indicators of water temperature (Ericson et al., 1961, 1964, Kennet and Huddles­ tun 1972, Zinsmeister 1974). It is widely believed that the 0 18 content of molluscs, foraminifera and corals is directly correlated with global ice (Shack­ leton 1977a). But Marner ( 1974b) points out that it depends on the quantity in ocean water, which is a mixture of several variables (Savin and Stehli 1974). Ocean water contaminated by melt water contains less 0 18 than ocean water elsewhere. Nor does ice-volume necessarily reflect temperature ac­ curately. Very rapid cooling may traverse the tem­ perature range so quickly where snow fall is abun­ dant that no glaciers grow. It appears that in Ireland there was little glacier expansion in the Younger Dryas because temperature fell too fast (Mitchell 1977). Temperature charts deduced from speleoth~ms seem markedly unreliable (Har­ mon et al., 1978). Several areas are quoted as standard indicators of sea level fluctuation during the Eem: Greenland, Barbados, New Guinea and California. Camp Century ice core, Greenland The base of the Greenland sequence was dated on a time scale which was subsequently reduced for unexplained reasons (Dansgaard et al., 1971). Temperature changes were deduced from increase of 0 18 in warmer periods. The logic of this is not clear. 0 18 in the core should decrease with heavy falls of deficient snow, but these would occur to- 165 wards the end of an interstadial and would dimin-· ish as the North American ice sheet spread (An­ drews et al., 1974). The base of the core is compressed, and apparently the stratigraphy is mixed and valueless (Thompson 1976). Marner ( 1974a) shows that the core cannot be used for the chronology of the Eem interglacial (cf. Kukla 1977). Barbados Calculations of absolute sea level from modern elevations of coral reef peaks on the slowly emerg­ ing island of Barbados have been used to construct a curve which is dated by thorium in the corals. Temperature has been estimated by 0 18 and global ice volume from temperature. It is claimed that down to -55 m polar ice had practically no effect on local temperature, so the 0 18 values are a reli­ able measure of global ice and of ocean level (Fairbanks and Matthews 1978, Matthews 1978). Dates of the calculated sea level curve are approxi­ mate, based on averages of a number of samples; statistical errors are 4 000-10 000 years (table 2). Dates by Broecker et al. ( 1968) and Ku ( 1968) vary within the statistical errors. Sea levels were the tops of the reefs about low water neap, i.e. about-2m M.S.L. (Russell 1967). TABLE2 Eemian Neap Low-Tide Sea Levels calculated from 0 18 at Barbados (after Fairbanks, R.G. and Matthews, R.K., 1978: Quat. Res. 10 181-96) Stage B .P. Coral Molluscs M.S.L. (m) 6 nadir pre-Eem 135 000 -125 5e peak Eem I 124 000 +5 +10 +7 5d nadir I-II 114 000 -71 5c peak Eem II 103 000 -43 -28 -41 5b nadir II-III 92 000 -75 5a peak Eem III 82 000 -42 -47 -40 4 nadir Wiirm I 69 000 -120 Statistical errors 4 000-l 0 000 years These figures make assumptions which it is diffi­ cult to justify from other parts of the world. First, that local temperature varied little throughout the Eem. One would expect masses of global ice, mostly on land in the northern hemisphere, to have modified winds and currents in the tropical Atlantic even if the ice sheets had not spread far enough to introduce melt water into the Carib­ bean. Secondly, that the uplift of Barbados has been uniform. This is by no means proved. The reefs are warped and not at constant levels (Neu­ mann 1971, Stearns C.E. 1976). A regression of 70 m should alter the isostatic pressure on the ocean floor and so lessen the uplift of ocean is­ lands. Different calculations from Barbados (e.g. Steinen et al., 1973) give very different figures for sea level except for Eem I which is widely accepted as around +8 m over much of the world. 166 Huon Peninsula, New Guinea Uplift since the beginning of Eem has exposed a series of reefs in north New Guinea which have been dated (Chappell 1974a, b), and from modern altitudes contemporary sea level has been calcu­ lated on the supposition that reef building oc­ curred only close to transgression peaks and that uplift has been constant (Bloom et al., 1974). This is unlikely because the reefs are much faulted and are tilted both to the west and east (Veeh and Chappell 1970, Ruddiman et al., 1977b). Dates are comparable with those from Barbados and Califor­ nia. Reefs V and VI are reasonably equated with Eem III and II; but for Reef VII there are two clusters of dates, about 117 000 and 140 000 B.P. By stretching these dates to the limits of their sta­ tistical errors, one can nearly combine them into a single rhase about 126 000 B.P. California Recalculated dates from Palos Verdes (Szabo and Rosholt 1969, Ku and Kern 1974) indicate Eem peaks about 130 000, 104 000 and 86 000 B.P., with large statistical errors. Dates which can be correlated have been obtained from other sites on this coast. The area is tectonically active; though divergent present altitudes of the terraces are known, it is not possible to calculate contem­ porary sea levels. General There is good evidence that in stable areas m the tropics as far as 35° Lat. mean sea level m Eem I was about +8 m (Shackleton 1977b). A lower platform about +6 m in South Africa (Da­ vies 1970, 1972) cannot be dated and may not be­ long to the Eemian. So global ice was in Eem I considerably less than today, and perhaps the ice sheet of west Antarctica broke up (Mercer 1978). Only at this stage did sub-tropical water in the Atlantic penetrate as far as 51° N (Mcintyre et al., 1972). On the other hand, recent studies (Clark et al., 1978) on the gravitational attraction of ocean water by an ice-field indicate geoid deformation of the ocean surface; and glacio-isostatic adjustment of the magma is a slow process which is not yet completed in the Arctic though we seem to be long past the peak of the present interglacial. In tem­ perate and sub-polar regions these trends would work in opposition. At a given latitude, as an ice sheet retreated ocean level would first fall and then rise as magma flowed poleward from the sub-trop­ ics, while in the tropics it would fall. It has been suggested that this would account for mid-Holo­ cene sea levels at +2-4m, for example the Nouak­ chottian in Mauritania and Senegal (Faure and Hebrard 1975), the South Cape (Davies 1972), Brazil (Fairbridge 1976) and south-east Australia (Gill 1970). Based on the Blake Event in loess, Eem I is taken to have ended about 115 000 B.P. (Kukla and Koci 1972). Contradictory calculations have been made regarding subsequent temperature curves on land and ocean and of sea level. Ap­ parently oceans reacted differently, presumably according to currents, to geographical access from the Arctic and for reasons which are not under­ stood. Moreover, dates by extrapolation and by thorium do not give the required accuracy. In Eem I-II the north-east part of the Atlantic tropical gyre west of Africa was cooled 6-8°, while in the Sargasso Sea there was little difference between stadial and interstadial (Crowley 1976). On the disputed assumptions of regular uplift of Barbados and that the 0 18 curve can be correlated directly with accumulation of global ice (Shackle­ ton and Opdyke 1973), sea level has been reck­ oned to have dropped to -71 m; but Shackleton ( 1969) considers that the drop in temperature indi­ cated by 0 18 was too rapid for ice sheets to have spread widely if indeed much ice was able to form ( cf. Mitchell 1977). If the temperature curve from the Orgnac stalactite is reliable (Duplessy et al., 1970, 1971, cf. Shackleton 1971), cooling in south­ ern France was slight. In Moravia between 115 000 and 104 000 B.P. there were several peaks of a sub-arctic stadial (Kukla and Koci 1972)_. The pollen diagram from Macedonia shows sharp cool­ ing, and broad-leaved trees disappeared in Hol­ land (Starkel 1977). There were a few icebergs in the mid Atlantic about 107 000 B.P., but they may have come from Greenland on a southward swing­ ing current and need not imply ice formation on continents now unglaciated (Sancetta et al., 1972). But though the great ocean-drop postulated for Barbados is supported elsewhere (> -55 m in South Crete, a tectonic area, Angelier et al., 1976; between -30 and -59 m off Corfu, Sardinas 1973 and 1974), it is calculated that an ice sheet 1 000 m thick in East Canada would depress sea level only 14 m (Andrews and Mahaffy 1976). Taking other northern ice sheets into account, sea level would drop at most 30 m. In Bermuda the drop is taken to be about 8 m (Harmon et al., 1978), in Guernsey possibly not more than 12m (George 1977). I have seen no evidence for coastal gullying which would be expected if sea level had dropped as much as 70 m, but it may be masked by renewed gullying during the deeper drop in the Wiirm glaciation. Eem II lasted about 104 000-94 000 B.P. 0 18 and foraminiferal evidence indicates that it was less warm than Eem I. On some coasts there is no emerged shoreline between Eem I and the Holo­ cene, for example Madagascar (Battistini 1976a), Indian Ocean islands (Veeh 1966, Thomson and Walton 1972) New South Wales (Thorn 1976), Marshall Islands (Szabo et al., 1977) and Hawaii (Stearns H.T. 1978). Its altitude has been calcu­ lated on the assumption of constant uplift: on the Ryukyu Islands, -10m (Machida 1975); on Bar­ bados, -28 to -41 m, which is unlikely (Fairbanks and Matthews 1978), or -10 to-13m (Steinen et al., 1973). It is not agreed whether it was cooler or warmer than Eem III. Marner ( 1971) suggests that it was the coldest of the Eem interstadials, but in Florida its shoreline is slightly higher than that of Eem I (Multer et al., 1968, Os­ mond et al., 1970). In the Mediterranean there are two emerged shorelines 5-10 m apart (Cuerda 1965-7, Sole Sabaris 1962, Cotecchia et al., 1969, Sanlaville 1969). On faunal and geological grounds the lower cannot be Holocene, and it car­ ries an impoverished Senegalese fauna but hardly ever Strombus bubonius. On the Atlantic coast of France neither the Haut nor the Bas Normannien are datable (Guilcher 1969, Ters 1968). Because the lower in places rests on periglacial material, the higher may be pre-Eem. In Asturia there were two transgressions to the same level (Mary 197 5). In Natal a platform just below M.S.L. occurs be­ low a higher beach at + 7-8 m. At Die Kelders Cave a platform at +8 m and a storm beach up to +2m are claimed without evidence on which is the older (Tankard and Schweitzer 1974). On the emergent Atlantic coast of the United States there seem to be two pairs ofshorelines (Colquhoun D.J. in litt., cf. Hoyt 1975) , but dates run on molluscs do not support this interpretation. It is reasonable to assume that the upper shoreline is Eem I , but there is less evidence whether the lower is Eem II or III. Dates from other regions suggest Eem II , for example the Antilles (Herweijer and Focke 1978), Bermuda (Harmon et al. , 1976, Vacher et al. , 1971), Victoria (Gill and Amin 1975), Timor (Chappell and Veeh 1978) and the Mediterranean (Bernat et al., 1978, Bonadonna and Bigazzi 1971, Sanlaville 1973, Lalou et al., 1971) . There is much evidence for a sharp drop of tem­ perature in the Atlantic Ocean about 92 000 B.P. (Ruddiman and Mcintyre 1976) with a rapid change of fauna in the Gulf of Mexico (Kennett and Huddlestun 1972), a temperature drop on the continents (Duplessy et al., 1970, 1971, Woillard 1975) and loess and gullying in Moravia (Kukla and Koci 1972). It has been suggested that large ice sheets in Europe provided the fresh water for the sapropel layer in the East Mediterranean (Wil­ liams et al., 1978). Conditions then slowly im­ proved near 83 000 B.P., a control date being pro­ vided by an ash fall over much of the Caribbean about 85 000. Elsewhere there has been little ev­ idence in this stadia! for sharp cooling or for a large marine regression (Butzer 1968, Butzer and Cuerda 1962, Mcintyre et al., 1972, Sancetta et al., 1972), and it has been thought that Eem II-III was less cold than Eem I-II (Suggate 1974). This 167 divergence of opinion may be due to the brevity of the cold snap. The last interstadial before rapid climatic de­ terioration with heavy snow and rain was Eem III. Its peak was probably a little after 80 000 B.P. It is dated by a tephra from the Bay of Bengal result­ ing from the explosion of Mount Toba in Sumatra. This occurs in sediments after the peak, about halfway to the nadir of Wiirm I , and there is a Kl Ar date of 73 500 ± 3 000 (Ninkovich et al. , 1978) 0 Eem III was as warm or perhaps warmer than Eem II as shown at Orgnac. Red earth developed in the Mediterranean (Butzer and Cuerda 1962), and some cores from there show greater depletion of 0 18 than in Eem I (Thunell et al. , 1977). The same is found in the north-east Pacific (Shackleton 1977a), but in the North Atlantic Eem III was slightly cooler than Eem I (Sancetta et al., 1972). On many coasts there are two Eem shorelines. A few dates, especially from the Mediterranean (Stearns and Thurber 1965, 1967, Dai Pra and Stearns 1977, Bonadonna and Bigazzi 1971), indi­ cate that the lower was Eem III, whereas most of the Red Sea dates point to Eem II (Berry et al., 1966, Veeh and Giegengack 1970, Hoang et al. , 1973). The unreliability and large statistical errors of Thorium dating make it difficult to distinguish the two interstadials. Equally unreliable calcula­ tions (Savin and Stehli 1974, Marner 1974b, Van Donk 1976) of glacier ice from oceanic 0 18 have been published to show that Eem III sea level was well below present (Shackleton 1977b, Steinen et al., 1973, Machida 1973, 1975). On the whole I am inclined to think that the absolute sea level of Eem III was below present. Where there has been slight uplift or where geoid deformation has changed in relation to the present because of the long preceding interglacial or semi-interglacial period the Eem II shoreline may be slightly emerged. ACKNOWLEDGEMENTS My earlier field work in the South Cape was done with grants-in-aid from the C.S.I.R. I would particularly thank Professor D.J. Colquhoun (Columbia) and Dr. D .R . Grant (Ottawa), officers of the INQUA Shorelines Commission, for ad vice; Dr. G .H . Miller (Boulder) for the racemisation­ figures; Dr.J.T. H ollin (Boulder), Dr. D .K . H obday (Austin), and Dr. R .R . Maud (Durban) for discussion of these figures, and Mr. R. Kilburn for his constant advice in identification of the Cape Fauna. REFERENCES AGACHE, R. (1969) . Gallia Prihistoire, 11, 267-309. ANDREWS, J .T. et al. (1974). Comparison of the glacial chronology of Eastern Baffin Island, East Greenland and the Camp Century accumulation record. Geology, 2 , 355-8. ----and MAHAFFY, M.A.W. ( 1976). Growth rate of the Laurentide ice-sheet. Quat. Res., 6, 167-83. ANGELIER, J . et al. (1976). Les deformations du Quater­ naire marin. Rev. Giogr. phys. et Giol. dynam. 18, 427-48. BADA, J .L. and DEEMS L. (1975). Accuracy of dates bey- ond the C 14 dating limit using the aspartic acid racemisa­ tion reaction. Nature, Lond., 255, 218-9. BARNARD, K.H. ( 1962). Revised list of South African Late Tertiary and Pleistocene marine mollusca. Trans. R. Soc. S. Afr., 36, 179-96. BATTISTINI , R . ( 1976a). Applications des m e thodes Th230-U234 a Ia datation des depots marins anciens de Madagascar. ASEQUA Bull., 49, 79-95. ---- ( 1976b). Datation par Ia methode Th230-U234 du 168 Pleistod:ne moyen marin de Madagascar. Soc. geol. France C.R. somm. Seances, 20 I. BEATER, B.E. (1967). Middle Stone Age implements on aeo­ lianite at Isipingo Beach: S. Afr. archaeol. Bull., 22 , 59. BEAUMONT, P.B. et al. (1978). Modern man in sub-Saha­ ran Africa prior to 49 000 years B.P. S. Ajr. }. Sci., 74, 409-19. BERNAT, M . et al. (1976) . Datation a !'ionium de quelques formations coralliennes emergees de Nouvelle Caledonie et des lies Loyaute. C.R. Acad. Sci. Paris, 282D, 9-12. ---- et al. ( 1978). lo-U dating of the Uljian stage from Torre Garcia. Nature, Lond., 275, 302-3 . BERRY, L. et al. ( 1966). Some radiocarbon dates and their geomorphological significance, emerged reef-complex of the Sudan. Z.f Geomorph., 10, 119-43. ElBERSON, P. (1961). Le cadre paleogeographique de Ia Prehistoire du Maroc atlantique. Serv. Antiquites du Maroc, Publ. 16. ---- ( 1963). Quelques precisions sur les classifications du Quaternaire marocain. Soc. geol. France Bull. , 7 Ser., 5, 607-16. BLOOM, A.L. et al. (1974). Quaternary sea-level fluctuations on a tectonic coast. Quat. Res. , 4 , 185-205. BONADONNA, F.P. and BIGAZZI, G. (1971). Stud! sui Pleistocene del Lazio VIII. Soc. geol. ital. Boll., 89, 463-73. BONIFAY, E. and MARS, P. (1959). Le Tyrrhenien dans le cadre de Ia chronologie quaternaire mediterraneenne. Soc. geol. France Bull. , 7 Ser., 1, 62-78. BOURDIER, F. (1969). Position chronologique du Paleoli­ thique de Sangatte Wissant et Wimereux: C.R. Soc. prehist. fran., 66, 230-l. BREUIL, H. MS report in Archaeological Survey, Johannes­ burg; File B I 7/7. BROECKER, W.S. et al. (1968). Milankovitch hypothesis supported by precise dating of coral-reefs and deep-sea sediments. Science, 159, 297-300. BUTZER, K.W. (1968). Comment on C. Emiliani, "The Pleistocene epoch and the evolution of man". Curr. Anthrop. 9, 31-2. ---- ( 1978). Sediment stratigraphy of the M.S.A. sequence at Klasies River Mouth. S. Afr. archaeol. Bull., 33, 141-51. ---- et al. ( 1978). Lithostratigraphy of Border Cave.}. archaeol. Sci., 5, 317-41. ---- and CUERDA, J. ( 1962). Coastal stratigraphy of southern Mallorca.J. Geol., 70, 398-416. CAMPS, G. (1974). Les civilisations prehistoriques de l'Afrique du Nord et du Sahara: Paris. Doin. CHAPPELL, J . ( l974a). Geology of coral terraces, Huon Pen­ insula. Bull. geol. Soc. Am., 85, 553-70. ---- (I974b). Problem of dating Upper Pleistocene sea­ levels from coral-surfaces. Proc. II Int. Coral-reif Sympos. 2, 563-71. ----and THOM, B. G. (1978). Termination ofLast In­ terglacial episode and the Wilson Antarctic surge hypothe­ sis. Nature, Lond., 272, 809-10. ----and VEEH H.H. (1978). Late Quaternary tec­ tonic movements and sea-level changes at Timor and Atauro. Bull. geol. Soc. Am. , 89, 35&-68. CHOUBERT, G. (1962). Reflexion sur les parallelismes probables des formations quaternaires atlantiques du Maroc avec celles de Ia Mediterranee. Quatemaria, 6, 137-75. CLARK, J .A. et al. ( 1978). Global changes in post-glacial sea­ level. Quat. Res., 9 , 265-87. COOKE, H.B.S. (1973). Pleistocene chronology, long or short? Quat. Res., 3 , 206-20. COTECCHIA, V. et al. ( 1969). Oscillazioni tirreniane e olo­ ceniche del livello del mare sui Golfo di Taranto. Geologia applicata e Jdrogeologia, 4 , 93-148. CROWLEY, T.J. ( 1976) . Fluctuations of the eastern North Atlantic gyre during the last 150 000 years. Ph.D. thesis , Dissert. Abstr., 38, 12IB. CUERDA, J. ( 1965-7). Abstract in Richards, H.G. , Bibli­ ography of Quaternary Shorelines, Suppl. 1965-9, 143. DAI PRA, G. and STEARNS C.T. (1977) . Sui Tirreniano di Taranto. Geol. Romana, 16, 231-42. DANSGAARD, W. et al. (1971). Climatic record revealed by the Camp Century ice-core. In : Turekian, K.K. , Ed., Late Cenozoic glacial ages, 37-56. New Haven, Yale U.P. DAVIES, 0. ( 1964). Quaternary in the coastlands of Guinea. Glas­ gow, Jackson. ---- ( 1970). Pleistocene beaches of T a tal. Ann. Natal Mus., 20, 403-42. / ---- ( 1971-2). Pleistocene shorelines in the southern and south-eastern Cape Province. Ann. Natal Mus. , 21 , 183-279. ---- (1973). Pleistocene shorelines in the Western Cape and South-west Africa. Ann. Natal Mus. , 21 , 719-65. ---- (1976). The older coastal dunes of Natal and Zu­ luland and their relation to former shorelines. Ann. S. Afr. Mus. , 71 , 19-32. DE LUMLEY, H. (1976). Les )ignes de rivage quaternaire de Provence et de Ia region de lice. In: De Lumley, H. , Ed., La Prehistoirefrancaise, Ia 311-25. Paris , C.N.R.S., DUPLESSY, J.-C. et al. (1970) . Continental climatic varia­ tions between 130 000 and 90 000 B.P. Nature, Lond., 226, 631-2. ---- et al. ( 1971). La Mesure des variations climati­ ques continentales. Quat. Res., 1, 162-74. ---- et al. ( 1976). Paleoclimatologie des temps quater­ naires a l'aide des methodes nucleaires. In : De Lumley, H., Ed., La Prehistoirefrancaise, Ia 352-61. Paris , C.N.R.S. ENGELBRECHT, L.N.J. et al. (1962). Die Geologie van die Gebied tussen Port Elizabeth en Alexandria: S. Afr. Geol. Surv., Explan. Sheets 3325D, 3326C and 3425B. ERICSON, D.B. et al. (1961). Atlantic deep-sea sediment­ cores. Bull. geol. Soc. Am., 72, 193-286. ---- et al. ( 1964). The Pleistocene epoch in deep-sea sediments. Science, 146, 723-32. EVANS. P. ( 1971 ). Towards a Pleistocene time-scale. In : Harland, W.B. and Francis, E.H., Eds., Phanerozoic time­ scale, supplement. Geol. Soc. Lond. Special Paper 5, 123-356. FAIRBANKS, R.G. and MATTHEWS, R.K. (1978). Marine oxygen-isotope record in Pleistocene coral, Barbados. Quat. Res., 10, 181-96. FAIRBRIDGE, R .W. (1976). Shellfish-eating Preceramic In­ dians in coastal Brazil. Science, 191, 353-9. FAURE, H. and HEBRARD, L. ( 1975). Variations des lignes de rivage au Senegal et en Mauritanie au cours de l'Holocene. War­ saw, Prace o Plejstocenie. FLEISCH, H. ( 1956). Depots prehistoriques de Ia cote liba­ naise. Quatemaria, 3 , 101-32. GEORGE, P.K. (1977). Quaternary shorelines in Guernsey. X INQUA Abstr., 160. GILL, E.D. (1970). Current Quaternary shoreline research in Australasia. Austr.}. Sci., 32, 426-30. ----and AMIN, B.S. (1975). Interpretation of7,5 and 4 metre Last Interglacial shore-platforms in Southeast Australia. Search, 6 , 394-6. GOODWIN, A.H.J. (1933). South African raised beaches. Man, 33 (50). GUILCHER, A. (1969). Le Quaternaire littoral et sous-marin dans I' Atlantique: in Etudes fran~aises sur le Quaternaire, INQUA, Paris, suppl. to Bull. AFEQ, 33-41. HARMON, R.S. et al. (1976). Sea-level changes in Bermuda during the last 200 000 years. 26 Int. Geol. Congr. Abstr., ii, 513. ---- et al. ( 1978). Late Pleistocene Paleoclimates of North America. Quat. Res. , 9 , 54-70. HAUGHTON, S.H. (1937). Geology of the country around Mossel Bay. S. Afr. Geol. Survey, Explan. Sheet 201. HERWEIJER, J.P. and FOCKE, J.W. (1978). Late Pleisto­ cene depositional and denudational history of the Nether­ lands Antilles. Geol. en Mijnb., 57, I 77-87. HOANG, C.T. et al. ( 1974). Les n!cifs souleves de !'ouest du Golfe d'Aden. C.N.R.S. Colloques int., 219, 103-14. HOBDAY, D.K. (1974). The Port Durnford Formation. Trans. Geol. Soc. S. Afr. 77, 141-9. ---- ( 1976). Quaternary sedimentation and devel­ opment of the lagoonal complex, Lake St. Lucia. Ann. S. Afr. Mus., 71, 93-113. . HOLLIN, J.T. (1965). Wilson's theory of ice-ages. Nature, Lond., 208, 12-16. ---- ( 1969). Ice-sheet surges and the geological record. Canad.J. Earth Sci., 6, 902-10. ---- (1970). Antarctic ice-surges. Antarct.J. of the U.S., 5 , 155-6. ---- ( 1972). Interglacial climates and Antarctic ice­ surges. Quat. Res., 2 , 401-8. ---- (1977). Thames interglacial sites, Ipswichian sea­ levels and Antarctic ice-surges. Boreas, 6, 33-5?.. HOYT, J.H. (1975). Intercontinental correlations of Late Pleistocene sea-levels. Nature, Lond., 215, 612-4. KAUFMANN, A. (1971) . Status ofU-series methods of mol­ lusk-dating. Geochim. et Cosmochim. Acta, 35, 1155-83. KENNETT, J.P. and HUDDLESTUN, P. (1972). Abrupt climatic change at 90 000 B.P. Quat. Res., 2 , 384-95. KERAUDREN, B. (1970-2). Formations quaternaires ma­ rines de Ia Grece. Mus. d'Anthrop. Prihist. Monaco Bull., 16, 5-153; 17, 87-169 and 18, 245-82. KILBURN, R.N. and TANKARD, A.J. (1975). Pleistocene molluscs from the west and south coasts of the Cape Pro­ vince. Ann. S. Afr. Mus., 67, 183-226. KLEIN, R .G. (1972). Preliminary report on 1970 excavations at Nelson Bay Cave. In: Van Zinderen Bakker, E.M., Ed., Palaeoecology of Africa, 6, I 77-208. KNOX, A.S. ( 1969). Glacial age marsh, Lafayette Park, Washington D.C. Science, 165, 795-7. KU, T-L. (1968). Pa23l method of dating coral from Barba­ dos .] . Geophys. Res. , 73, 2271-6. ---- and KERN J.P. (1974). Uranium-series age of the Upper Pleistocene Nestor Terrance. Bull. geol. Soc. Am., 85, 1713-6. KUKLA, G.J. (1977). Pleistocene land-sea correlations. Earth Sci. Rev., 13, 307-74. ----and KOCI, A. ( 1972). End of the Last Intergla­ cial in the loess record. Quat. Res., 2, 374-83. LAIDLER, P.W. (1933). Dating evidence concerning the Middle Stone Ages and a Capsio-Wilton culture in the south-east Cape. S. A(r.J. Sci., 30, 530-42. LALOU, CI. et al. (197 1) . Donnees geochronologiques actu­ elles sur l'Interglaciaire Riss-Wi.irm. Rev. Giogr. phys. Giol. d.ynam., 13, 447-61. MABBUTT, J.A. (1954). Cape Hangklip. Trans. R . Soc. S. Afr., 34, 17-24. MACFARLANE, D.R. (1937). On the East London Leval­ lois. S. Afr.]. Sci., 33, 893-903. ---- ( 1958). Causes of submergence and re-emergence at Plettenberg Bay. S. Afr.J. Sci., 54, 231-40. MACHIDA, H . (1973). Pleistocene sea-level changes ana­ lysed by tephrochronology. IX INQUA Abstr. 225-6. ---- ( 197 5). Pleistocene sea-level changes analysed by tephrochronology. Bull R . Soc. New Zealand, 13. MANGION, S. and MATTHEWS, R.C. (1974). Sea-level dynamics at the end of the Last Interglacial. Geol. Soc. Amer. Abstr. 6, vii, 855. MARSHALL, J.F. and LAUNAY, J. (1978). Uplift rates of the Loyalty Islands. Quat. Res .. 9 , 186-92. 169 ----and THOM, B.G. (1976). Sea-level in the Last Interglacial. Nature, 263, 120-1. MARY, G. et al. (1975). Le Quaternaire de Ia cote asturienne. Bull. AFEQ, 42, 13-23. MATTHEWS, R .K. (1973). Relative elevation of Late Pleis­ tocene high sea-level stands. Quat. Res., 3, 147-53. ---- (1978) . Barbados high stands of sea-level and the deep-sea b 18° ice-volume curve. Litoralia, 11, 5, 6. MciNTYRE, A. et al. ( 1972). Southward penetration of North Atlantic polar front. Deep-sea Research, 19, 61-77. MERCER, J.H. (1978). West Antarctic ice-sheet and C02 greenhouse effect. Nature, Lond., 271, 321-5. MITCHELL, G.F. (1977). Periglacial Ireland. Phil. Trans. R. Soc. Lond., 280B, 199-209. MOORE, W.S. et al. (1974). Age determination of fossil corals.]. Geophys. Res., 79, 5065-8. MOREL, J. (1967). Les formations quaternaires littorales de Ia region de La Calle. Acts VI Panafr. Congr., 408-19. MORNER, N-A. ( 1971). Position of ocean-level during the interstadial about 30 000 B.P. Canad. ]. Earth Sci., 8, 132-43. ---- (1972a). World climate during the last 130 000 years. 24 Int . Geol. Congr., section 12, 72-9. ---- (1972b). Cold-warm changes during the last ice­ age. Stockholm Contrib. Geol., 24(4), 51-77. ---- ( 1974a). The Greenland 0 18 curve. In: Labeyrie, J., Ed., Variations du climat au cours du Pleistocene. C.N.R.S. Colloq. int., 219, 39-42. ---- (1974b). Ocean paleotemperatures and continen­ tal glaciations. C.N.R.S. Colloq. int., 219, 43-49. MORTELMANS, G. (1945). Plages soulevees a industries li­ thiques de Ia region de Keurbooms River. S. Afr. ]. Sci., 41, 375-96. MOUNTAIN, E.D. (1962). Geology of the country around Port Alfred: S. Afr. Geol. Survey Explan. Sheets 3326D-3327C. ---- ( 1966) . Footprints in calcareous sandstone at Na­ hoon Point. S. Afr.]. Sci., 62, 103-10. MULTER, H .G. and HOFFMEISTER, J.E. (1968). Sub­ aerial laminated crusts of the Florida Keys. Bull. geol. Soc. Am., 79, 183-92. NEUMANN, A.C. (1971). Elevated Pleistocene sea-level fea­ tures in the Bahamas. Quatemaria, 15, 73. NINKOVICH, D. et al. ( 1978). KAr age of the Late Pleisto­ cene eruption ofToba. Nature, Lond., 276, 574-7. OSMOND, J.K. et al. (1965). Th2:ro/U234 age of Pleistocene corals and oolites of Florida.]. Geophys. Res., 70, 1843-7. ---- et al. ( 1970). Age of the Cape Kennedy barrier and lagoon complex.]. Geophys. Res., 75, 469-79. PERTHUISOT,J.P. (1972). La Sebkha el Melah. Rev. Giogr. phys. Giol. d.ynam., 14, 67-84. PORTER, S.C. et al. (1977) . . Chronology of Hawaiian glacia­ tion. Science, 195, 61-3. RICHARDS, H.G. (1972). Abstr. in Bibliography 1970-3 26 of paper in Bull. Cape May Georgr. Soc., 26, 6-7. ROUBET, F.E. (1969). Le niveau aterien dans Ia stratigra­ phie cotiere a !'ouest d'Alger. In: Van Zinderen Bakker, E., Ed., Palaeoecology of Africa, 4, 124-9. Cape Town, Bal­ kema. RUDDIMAN, W.F. (1977a). Late Quaternary deposition of ice-rafted sand. Bull. geol. Soc. Am., 88, 1813-27. ---- et al. ( 1977b) . Glacial-interglacial response-rate of sub-polar North Atlantic water to climatic change. Phil. Trans. R. Soc. Lond., 280B, 119-42. ---- and MciNTYRE, A. ( 1976). Northeast Atlantic paleoclimatic changes. Mem. geol. Soc. Am., 145, 111-46. RUDDOCK, A. (1957). Note on the relation between Chelles-Acheul implements and Quaternary river terraces in the valleys of the Coega and Sundays Rivers. S. Afr. ]. Sci., 53, 373-7. 170 RUSSELL, R.J. ( 1967). Aspects of coastal geomorphology. Geografiska Ann., 49A, 299-309. SANCETTA, C. et al. ( 1972). Climatic record in 1 orth Atlan­ tic deep-sea core V 23-82. Quat. Res., 2 , 363-7. SAN LA VILLE, P. ( 1969). Les bas niveaux marins pleisto­ d:nes du Liban. Miditerranee, 3 , 257-92. ---- ( 1973). Sur le Tyrrhenien libanais. Quatemaria, 15, 239-48. SAVIN, S.M. and STEHLI, F.G. ( 1974). Interpretations of oxygen-isotope paleotemperature measurements. In: La­ beyrie, J ., Ed., Methodes quantitatives d'etude des variations du climat au cours du Pleistocene. C.N.R.S. Colt. int., 219, 183-91. SCHUBERT, C. and SZABO, B.J. (1978). Uranium-series ages of Pleistocene marine deposits. Geol. en Mijnb., 57, 325-32. SCHWARZ, E.H.L. ( 1899). Knysna between the Gouw­ kamma and the Blue Krantz Rivers. Geol. Comm. Rep., Cape Town, 453-63. SHACKLETON, N.J. ( 1969). The Last Interglacial in marine and terrestrial records. Proc. R . Soc. Lond., 174B, 135-54. ---- (1971 ). In: Harland, W.B. and Francis, E.H., Eds., The Phanerozoic time-scale; suppl. Geol. Soc. Lond., Special Publ., 5 , 80-111. ---- ( 1977a). Oxygen-isotope stratigraphic record of the Late Pleistocene. Phil. Trans. R. Soc. Lond. , 280B, 169-82. ---- et al. ( 1977b) . Oxygen-isotope stratigraphy of Late Pleistocene coral-terraces in Barbados. Nature, Lond., 268, 618-9. ----and OPDYKE, N.D. (1973). Oxygen-isotope and palaeomagnetic stratigraphy of Equatorial Pacific core V 28-238. Quat. Res., 3 , 39-55. SHACKLEY, M.L. (1977). New evidence for the chronology of the 7,5 metre beach in southern England. X INQUA Abstr., 417. SOLE SABARIS, L. ( 1962). Le Quaternaire marin des Bale­ ares. Quatemaria, 6, 309-42. SORDINAS, A. (1973-4). Stone-age sites on offshore islets northwest of Corfu. Paper at Amer. Anthrop. Assoc. meet­ ing 1972, pub!. Kerkyraika Khronika, 19, 88-93. SPARKS, B.W. and WEST, R.G. (1972). The ice-age in Britain. London, Methuen. STARKEL, L. (1977). Palaeogeography of mid- and east Eu­ rope' during the last cold stage. Phil. Trans. R. Soc. Lond. , 280B, 351-72. STEARNS, C.E. ( 1976). Estimates of the position of sea-level between 140 000 and 75 000 years ago. Quat. Res. , 6 , 445-9. ----and THURBER, D.L. (1965). Th230-U234 dates of Late Pleistocene marine fossils. Quatemaria 7 , 29-42. ---- ( 1967). Th230-U 234 dates of Late Pleistocene ma­ rine fossils . Progr. in Oceanogr., 4, 293-305 ( = VII IN QUA Vol. 13). STEARNS, H.T. (1961 ). Eustatic shorelines on Pacific islands. Z.f. Geomorph. Supplband, 3 , 3-16. ---- (1974). Submerged shorelines and shelves on the Hawaiian islands. Bull. geol. soc. Am., 85, 795-804. ---- ( 1978). Quaternary shorelines in the Hawaiian is­ lands. Bull. Bernice P . Bishop Mus., 237. STEINEN, R.P. et al. (1973) . Eustatic low stand of sea-level between 125 000 and I 05 000 B.P. Bull. geol. Soc. Am., 84, 63-70. STOW, G.W. (1871 ). On some points in South African ge­ ology. Quart.]. Geol. Soc. Lond. , 27, 515-22. SUGGATE, R.P. (1974). When did the Last Interglacial end? Quat. Res., 4, 246--52. SZABO, B.J. et al. (1977). Uranium-series dating of coral samples from Eniwetok. X INQUA Abstr., 456. ---- and ROSHOLT, J.N. (1969). Uranium-series dating of Pleistocene molluscan shells from southern Cali­ fornia.). Geophys. Res. , 74, 3253-3260. TANKARD, A.J. (1975). Thermally anomalous Late Pleisto­ cene molluscs. Ann. S. Afr. Mus., 69, 17-45. ---- ( 1976) . Pleistocene history and coastal morpho­ logy of the Ysterfontein- Elands Bay area. Ann. S. Afr. Mus., 69, 73-119. ----and SCHWEITZER, F.R. (1974). Geology of Die Kelders Cave and environs. S. Afr.). Sci., 70, 365-9. TAYLOR, F.W. and BLOOM, A.L. (1977). Quaternary tec­ tonics and Th230/U234 dating of emerged coral reefs. X INQUA Abstr., 462. TERS, M. (1968). Quelques gisements du haut niveau marin entre le Loire et Ia Sevre niortaise. AFEQ Bull. , 5, xiv, 19-43. THOM, B.G. ( 1973). Dilemma of high interstadial sea-levels. Progr. in Geography, 5 , 1 70-246. ---- ( 1976). Coastal sand-barrier evolution in New South Wales. 26Int. Geol. Congr. Abstr., 2, 513. THOMPSON, L.G. ( 1976). Microparticles, ice-sheets and cli­ mate. Ph.D. thesis, Dissert. Abstr., 37, 2129B. THOMSON, J. and WALTON, A. (1972). Redetermination of chronology of Aida bra Atoll. Nature, Lond., 240, 145-6. THUNELL, R.C. et al. ( 1977). Late Quaternary Paleoclima­ tology, Stratigraphy and Sapropel history in the Eastern M editerranean deep-sea sediments. Marine Micropalaeon­ tology, 2, 371-88. VACHER, L.H. et al. (1971 ). Bermuda's glacioeustatic strati­ graphic section. Quatemaria, 15, 71-2. VAN DER HEIDE, S. (1961) . Problems of the manne Eemian in Europe. VIINQUA Acta, 1, 167-73. VAN DONK, J . (1976). 0 18 record of the Atlantic Ocean. Mem. geol. Soc. Am., 145, 147-63. VEEH, H.H. (1966). Th230/U238 and U 234/U238 ages of Pleisto­ cene high sea-level stand.). Geophys. Res. , 71, 3379-86. ---- and CHAPPELL, J. ( 1970) . Astronomical theory of climatic change, support from New Guinea. Science, 167, 862-5. ----and GIEGENGACK, R. (1970) . Uranium-series ages of corals from the Red Sea. Nature, Lond. , 226, 155-6. VOLMAN, T.P. (1978). Early archaeological evidence for shellfish collecting. Science, 201, 911-3. WENDORF, F . et al. ( 1975). Dates for the Middle Stone Age of East Africa. Science, 187, 740-2. ----and MARKS, A.E. Eds. ( 1975). Problems in Prehis­ tory, North Africa and the Levant. Dallas, S.M.U. Press. ----and SCHILD, R. (1974). A Middle Stone Age se­ quencefrom the Central Rift Valley, Ethiopia. Wroclaw, Polska Akademia Nauk. WEST, R.G. et al. ( 1964). Interglacial deposits at IIford. Phil. Trans. R. Soc. Lond., 247B, 185-212. WIE EKE, F. and RUST, U. (1975). Zur relativen und ab­ soluten Geochronologie der Reliefentwicklung an der Kiiste des mittleren Siidwestafrika. Eisz. u. Gegenw., 26, 241-50. WIJMSTRA, T.A. ( 1969) . Palynology of the first 30 metres of a 120-metre deep section in Northern Greece. Acta botanica neerlandica, 18, 511-27. WILLIAMS, D.F. et al. (1978). Periodic freshwater flooding and stagnation of the Eastern Mediterranean Sea during the Late Quaternary. Science, 201, 252-4. WILSON, A.T. (1964). Origin of ice-ages. Nature, Lond. , 261, 147-9. ---- ( 1969) . Climatic effects of large-scale surges of ice­ sheets. Canad.j. Earth Sci., 6, 911-8. WOILLARD, G. (1975). Recherches palynologiques sur le Pleistocene dans !'Est de Ia Belgique et dans les Vosges lorraines. Acta geographirrz lovaniensia, 14. ---- ( 1978) . The last interglacial-glacial cycle at Grande Pile. Trav. Labor. Palynologie Univ. Louvain. WYBERGH, W. ( 1919) . The coastal limestones of the Cape Province. Trans. Geol. Soc. S. Ajr., 22, 46-67. BP-L1 171 ZIJ SMEISTEK, W .J. ( 1974). A new interpretation of ther­ mally anomalous mollusc assemblages. }. Paleontol., 48, 84-94.