Pleistocene depositional history in a periglacial terrane: A 500 k.y. record from Kents Cavern, Devon, United Kingdom

Joyce Lundberg* Department of Geography and Environmental Studies, Carleton University, Ottawa ON K1S 5B6, Canada Donald A. McFarlane† W.M. Keck Science Center, The Claremont Colleges, 925 North Mills Avenue, Claremont, California 91711, USA

ABSTRACT past 500 k.y. is probably unique in the karst information in the deposits has largely been literature and establishes Kents Cavern as a ignored. Here we focus on the reconstruction of The signifi cance of the stratigraphic record site of international scientifi c interest. the Pleistocene depositional events of the past in Kents Cavern, Devon, United Kingdom, to 500 k.y. that are recorded in this inter nationally the interpretation of the British Quaternary Keywords: Kents Cavern, middle Pleisto- important site. is confi rmed on the basis of a thorough reex- cene, Britain, cave bear, interglacial, MIS 11, The south Devon karst is one of the few amination of the deposits in concert with 2 Hoxnian, speleothem, periglacial, Acheulian, limestone areas in Britain that was not directly new Al-Be cosmogenic and 34 new thermal cave sediment. affected by the Pleistocene ice sheets. Deposi- ionization mass spectrometry U-Th dates. tion of both clastic and crystalline material in The deposits show evidence of complex INTRODUCTION the cave was abundant, such that much of the reworking in response to periglaciation, and site was barely accessible to early explorers the main fl owstone deposit is a multilayered Kents Cavern, Torquay, Devon, UK (NGR (MacEnery, 1859). The fi rst systematic excava- complex spanning marine isotope stage (MIS) SX 934 642, Fig. 1), has been a focus of scien- tions began in the early 1800s; since then, the 11–3. The lowermost unit of fl uvial sands is tifi c interest for at least the past two centuries. deposits have mostly been removed from the Cromerian or older. The second deposit, a The extraordinarily rich record of Pleistocene cave so that the cavity now exposed is consid- muddy breccia of surfi cial periglacial solifl uc- mammals and human artifacts has focused erably larger than that accessible in the early tion material containing Acheulian artifacts, interest on the paleontology and archaeology nineteenth century. The major excavation was entered the cave during MIS 12 from high- of the site (e.g., MacEnery, 1859; Pengelly, carried out in 1868–1880 by W. Pengelly (see level openings to the west. Cave bears denned 1869, 1884; Campbell and Sampson, 1971; reconstruction of Pengelly’s excavations in in the cave during MIS 11, the Hoxnian inter- Cook and Jacobi, 1998), while the sedimento- McFarlane and Lundberg, 2005). MacEnery glacial; their bones are capped by an MIS 11 logical, climatological, and geomorphological (1859) and Pengelly (1884) described the basic calcite fl owstone layer. From MIS 11 onward, each interglacial period and the warmer interstadial periods (MIS 11, 10b, 9, 7, 6b, 5, Devensian glacial and 3) produced calcite fl owstone deposition limit: MIS 2 Babbacombe in the cave; MIS 9 was particularly active. Bay S. WALES Each glacial or stadial period (MIS 10c, 10a, 8, 6c, 6a, 4, and 2) caused periglacial activity Anglian glacial TORQUAY in the cave, during which the thinner layers limit: MIS 12 Hope’s of calcite were fractured by frost heave and Bristol Channel Nose redistributed by solifl uction. This sequence S. W. ENGLAND PAIGNTON Kents Cavern was interrupted during MIS 3–2 with the 51 N introduction of sandy and stony clastic sedi- ments from entrances to the east, and fi nally Tor Bay N cemented by the uppermost layer of MIS 1 Kents Cavern fl owstone. This is the fi rst publication of well- Torquay Berry Head dated and clearly documented evidence of 3W frost heaving in interior cave passages. The A 100 km English Channel B 1km Kents Cavern record of continuous, repeated sedimentation events followed by frost shat- Figure 1. (A) Location of Kents Cavern, southwest England, in relation to the glacial limits tering and remobilization events over the for the Devensian, MIS 2, and the Anglian, MIS 12, glaciations (glacial limits after Culling- ford, 1982; Croot and Griffi ths, 2001). (B) Location of Kents Cavern at the northern end of *[email protected] the town of Torquay, and outcrops of Torquay limestone in the region (shaded areas; after †[email protected] Scrivener, 1987).

Geosphere; August 2007; v. 3; no. 4; p. 199–219; doi: 10.1130/GES00085.1; 19 fi gures; 3 tables.

For permission to copy, contact [email protected] 199 © 2007 Geological Society of America Lundberg and McFarlane sedimentological sequence, from oldest to U-Th dates; second, we present the fi rst direct genesis may have involved some rejuvenation youngest, as Breccia, Crystalline Stalagmite, radiometric dates on the Breccia and its ante- of late Carboniferous–Permian paleokarstic Loamy Cave Earth, Stony Cave Earth, Granu- cedent stratum, providing a well-supported features (Campbell et al., 1998b). The geomor- lar Stalagmite, and Black Mold, and these terms time window for the middle Pleistocene fauna phological and hydrologic setting must have have been retained by later workers. While this and human artifacts from the cave. been different during speleogenesis, because the stratigraphy appears to be adequate for the 14C modern catchment for Kents Cavern, set high on dated material, the clastic deposits, and the Geological Background a small hillside, is too small to provide substan- upper parts of the calcite (Fig. 2; Oxford Radio- tial volumes of water. carbon Accelerator Unit, 2006), it is insuffi - The geological setting was described by Dur- The Quaternary deposits in the region include cient to describe the Crystalline Stalagmite and rance and Laming (1982) and Scrivener (1987). raised beach deposits, river terrace gravels, and hides a wealth of complexity. Proctor (1994) The cave is developed in the Middle–Upper periglacial deposits (Campbell et al., 1998a). expended considerable effort to date the calcites Devonian Torquay Limestone (Fig. 1B). The The ice sheets of the Quaternary glaciations did by uranium series disequilibrium alpha count- underlying Nordon Slate is a gray, well-cleaved, not reach the Torquay district, but peri glacial ing (alpha U-Th) and by electron spin resonance calcareous mudstone with some bands of slaty conditions were widespread during glacial (ESR), and continued this simple designation, limestone. This gives way to a complex of vol- periods (Cullingford, 1982). The two glacial ice with some minor variations. Proctor et al. (2005) canic tuff, agglomerate, and lava interspersed sheets that came closest to Kents Cavern were added three thermal ionization mass spectro- with limestone, much of which is reefal in struc- the last, during marine isotope stage (MIS) 2, metric U-Th (TIMS U-Th) dates to the data set ture; the whole unit is designated as the Torquay the Devensian, peaking ca. 20 ka, and MIS 12, in an attempt to clarify the interpretation of the Limestone. The overlying Upper Devonian the Anglian, ca. 430 ka (Fig. 1A). Periglaciation lowermost unit, the Breccia. Gurrington Slate is a gray-green and purple triggered extensive frost action and produced Improvements in the technology of both mudstone. These rocks were severely fractured thick periglacial slope deposits or solifl uction U-Th dating and the development of cosmo- and folded by the Variscan orogeny ca. 300 Ma. deposits (locally called “head”). The slates and genic isotope dating for cave sediments now New Red Sandstones of Permian age uncon- mudstones were particularly susceptible to frost allow a more defi nitive dating of the older formably overlie the uplifted and eroded post- action and resulted in thicker than average head calcite and clastic deposits. The aim of this orogenic surface. The Devonian rocks generally deposits (some sections documented as >20 m study is twofold: fi rst, we clarify the deposi- show substantial weathering underneath the thick; Cullingford, 1982). tional sequence and its paleoenvironmental sandstone beds. context based on 34 new, high-precision TIMS Many relatively small caves have formed METHODS in the Torquay Limestone; Kents Cavern is the third largest, with <1 km of mapped pas- Sampling (Fig. 3) was undertaken with the 45 sages. Kents Cavern preserves evidence of both support and cooperation of the cave owner and phreatic and vadose activity, but phreatic under permit from English Nature. Particular 40 Cave Earth features generally dominate. The cave prob- care and attention was paid to environmental 35 ably formed in the early Pleistocene, but speleo- impact, and the work was completed with mini-

30 MIS 3

25 Cave of Rodentia Kents Cavern, Vestibule 20 Wolf’s Torquay, Devon Cave North

C-14 age (ka) 15 Entrance Charcoal Cave 10 South Granular Entrance 5 Stalagmite Rocky MIS 1 MIS 2 Chamber 0 North Gallery Sally 14 Long arcade Port Figure 2. The distribution of C dates on Southwest artifacts from the Cave Earth and Granular Clinnick’s Gallery Chamber South Sally Port Stalagmite (Oxford Radiocarbon Accelera- High Level Chamber Inscribed Lake tor Unit, 2006) arranged in order of age for Boss N each type of material. The dates fi t into two Water Gallery Labyrinth clearly separated populations. The Cave Swallow Hole Gallery Bear’s Den Earth dates are between ca. 22 and 40 ka In-Between Boss Bear’s Den Boss

(MIS 3), and the Granular Stalagmite dates Hedges Boss are between ca. 4 and 16 ka (MIS 1 and late Terminal Chamber 30 m MIS 2). This second set of dates is largely on bones cemented into the base of the fl ow- stone and thus is biased toward the begin- Figure 3. Sampling sites and names of main passages in Kents Cavern. Outline survey is ning of deposition of the fl owstone. based on Proctor and Smart (1989), with modifi cations.

200 Geosphere, August 2007 Kents Cavern Pleistocene History

mal or no visible trace. Samples were taken of visible traces of detritus, vugs, or intercrystal- ferent parts of the hand specimen because they calcite that had already been damaged by historic lite voids were removed with a dentist’s drill had visible detrital contamination. blasting and excavations. In order to minimize under the microscope. The very low U content Two of the samples with highest detrital the collection of undatable specimens, all poten- (0.05 ± 0.02 ppm) dictated the use of relatively thorium concentration, SW1-O and LB-O, tial sites were examined in the fi eld for evidence large sample sizes (~2 g). Samples (n = 45) were were sampled 4–6 times for isochron dating of open-system behavior such as vugs, hiatuses, ultrasonically cleaned, ignited for 5 h at 875 °C (Schwarcz and Latham, 1989; Ivanovich et al.,

loose or friable or random fabric, recrystalliza- to remove organics, dissolved in HNO3 and 1992); each subsample was chosen for its vis- tion, or dissolution. Sample sites were recorded spiked with 233U-236U-229Th tracer. Apart from the ibly differing detrital content. The principle of on site using photographs and photographic samples chosen for isochron dating (see follow- isochron dating is that the subsamples are depos- mosaics. In view of the complicated stratigraphy ing), no sample showed visible detrital contami- ited at the same time but with differing amounts and in order to avoid subsequent misinterpreta- nation. U and Th were coprecipitated with iron of detrital contamination. The carbonate frac- tion of photographs, sequences were logged and hydroxide, and purifi ed twice on anion exchange tion 230Th is presumed to be constant for all sub- photographs annotated in the cave using a por- columns (Dowex AG1-X 200–400 mesh). samples, but the detrital fraction 230Th varies; table computer. Measurement of U and Th isotopic ratios 232Th is used as an index of detrital content. The was mainly done with TIMS, 14 using the VG slope of the regression line of 234U/232Th against 26Al-10Be Dating 354 TIMS at McMaster University, Hamil- 238U/232Th gives the 234U/238U, and the slope of ton, Ontario, and 19 using the Triton TIMS at 230Th/232Th against 234U/232Th gives 230Th/234U at Recent advances in dating cave sediments have the Isotope Geochemistry and Geochronology zero detrital contamination (Fig. 4). These ratios achieved considerable success with 26Al-10Be Research Centre, Carleton University, Ottawa, are then used to calculate the isochron age. dating of buried quartz (Granger et al., 2001; Ontario. The 8 samples for isochron dating A 230Th/232Th ratio of <20 is normally consid- Anthony and Granger, 2004). Cosmic ray sec- and 4 repeats were measured using the multi- ered to indicate detritally contaminated material ondary neutron bombardment of exposed or collector inductively coupled mass spectrometer (e.g., Schwarcz and Blackwell, 1992). However, superfi cially buried quartz generates 26Al from (MC-ICP-MS) at Géotop, University of Quebec isochron dating is prohibitively expensive for silicon atoms and 10Be from oxygen atoms in at Montreal, Quebec. Each suite of measure- many samples. A simplistic way to adjust the a constant ratio of ~6:1, regardless of absolute ments was accompanied by the processing of 230Th/234U to account for detrital contamination is dose. After transport and burial in a cave, these uraninite in secular equilibrium to ensure accu- to estimate an initial 230Th/232Th activity ratio for nuclides decay at differing rates (half-lives: rate spike calibration and fractionation correc- the detrital fraction. For example, Kaufman and 26Al = 1.02 Ma; 10Be = 1.93 Ma), causing a shift tion. Precision of isotopic ratio measurement Broecker (1965) used a value of 1.7 from measure- in the nuclide ratio that can be used to date sedi- is limited by extremely low U content and by ment of modern detritus. In our case the sample ment burial ages in the range of ca. 100 ka to high 232Th content. The typical 2σ error of the SWO had an age of 183 ka, but a 230Th /232Th ratio 5 Ma (Granger and Muzikar, 2001). Two mate- TIMS measurement of 234U/238U is 0.12% and of of 3. It was then dated by isochron on another rials were chosen as potentially amenable to 230Th/234U is 0.46% (TIMS instrumental repro- four fractions (SW1-O) to 152 –30/+45 ka. If this technique (Table 1): the quartz sand from ducibility on spiked uraninite standards is 0.06% this is taken as the more correct age, then the the basal Red Sands of the Gallery (cf. Camp- for 234U/238U and 0.11% for 230Th/234U). The pre- 230Th/234U ratio for SWO can be adjusted to give bell and Sampson, 1971) and quartz pebbles cision on the resultant ages varies with age; the an age of 152 ka using an initial 230Th/232Th ratio from the exposed Breccia in the Bear’s Den and average 2σ precision on these mid-Pleistocene of 1.00. This estimate for initial 230Th/232Th was Labyrinth areas. Samples were prepared and dates is 2.2% (Table 2). then used for all the samples with a 230Th/232Th analyzed at the Purdue Rare Isotope laboratory, Three of the hand specimens were sampled ratio of <20. It is clearly only an approxima- Purdue University. twice. CDF and B20 were measured on the tion, but is an improvement over the alternate McMaster TIMS and CDF2 and B20R on the practice of simply using a standard value. If the U-Th Dating Géotop MC-ICP-MS. The coincidence of the estimated initial 230Th/232Th ratio is low, and/or if numbers (within 2σ) suggests that isotopic the measured 230Th/232Th ratio is high, the adjust- The calcite samples were dated by standard measurements on TIMS and ICP are com patible. ment does not alter the original date. U-Th disequilibrium techniques (e.g., Ivano vich Samples CEB and CEBR were not from the same In Table 2, the age for the samples with and Harmon, 1992). Specimens were sliced section of hand specimen: CEBR was deliber- detrital contamination is shown fi rst as the into 2-mm-thick slivers and examined under a ately chosen from a different part in the hope simple calculated age; beside it in parenthe- binocular microscope with back lighting. Only that it would not be leached. Similarly, LOBO ses is the adjusted age; the age used is high- the cleanest parts were used for dating. All and LOAN were deliberately chosen from dif- lighted in bold. In most cases, the adjusted age

TABLE 1. DATA FROM 26Al/10Be ANALYSES OF SANDS FROM GALLERY AND QUARTZ CLASTS FROM BRECCIA IN LABYRINTH AND BEAR’S DEN Sample Mass Be Al 10Be/9Be 26Al/27Al 10Be(at/g) 26Al (at/g) 26Al/10Be Burial Age (g) (mg) (mg) ±1σ ±1σ ±1σ ±1σ ±1σ (Ma) ±1σ 51 34 163,920 386,726 2.36 2.32 Gallery sand 12.1 0.755 7.26 ± 4 ± 14 ± 14,610 ± 210,621 ± 1.30 ± 1.74 93 14 94,878 420,568 4.43 0.95 Breccia quartz 40.0 0.695 5.51 ± 8 ± 26 ± 8388 ± 87,379 ± 1.00 ± 0.55 Note: Ratios are shown as atomic ratios with 1σ error.

Geosphere, August 2007 201 TABLE 2. DATA FROM U-Th ANALYSES OF CALCITES Age U Th 230Th/234U 234U/238U 230Th/232Th 234U/238U initial Sample +2σ –2σ (ka) (ppm) (ppm) ±2σ ±2σ ±2σ ±2σ G2TR † 47.4 (20.0) 3.1 3.0 0.037 0.623 0.356(18) 1.192(13) 1.6(1) 1.22(1) G2BW † 79.5 2.6 2.5 0.031 0.028 0.523(11) 1.118(10) 41.1(8) 1.15(1) CAT † 80.2 0.8 0.8 0.042 0.055 0.524(3) 1.082(4) 27.5(2) 1.103(4) CBN † 93.3 1.9 1.8 0.060 0.050 0.582(7) 1.129(3) 49.9(7) 1.167(3) CCT † 94.6 (91.0) 4.0 3.9 0.030 0.070 0.588(15) 1.130(5) 17.9(5) 1.169(5) CDF † 148.2 9.5 8.7 0.044 0.097 0.757(21) 1.116(12) 24.1(6) 1.18(1) CDF2 # 150.4 5.0 4.8 0.044 0.099 0.763(12) 1.121(2) 23.9(4) 1.185(2) CEB † Leached 0.042 0.029 1.63(5) 1.084(3) 161(5) CEBR # Leached 0.038 0.230 1.21(1) 0.955(1) 16.1(2) B20 † 210 2 2 0.078 0.010 0.985(4) 2.705(8) 1274(6) 4.08(1) B20R # 197 8 8 0.067 0.027 0.960(18) 2.752(1) 531(11) 4.057(1) WUXU † 417 104 52 0.051 0.026 0.996(12) 1.053(10) 130(1) 1.17(1) WTXT † 132 6 6 0.046 0.114 0.713(16) 1.105(13) 20.3(4) 1.15(1) WTXT2 # 150 6 6 0.047 0.114 0.759(15) 1.086(2) 21.0(4) 1.131(2) LA6R † 1.2 (–2.8) 0.2 0.2 0.048 0.323 0.011(2) 1.257(4) 0.3(1) 1.258(4) BD1 § 326 7 7 0.036 0.271 0.981(3) 1.111(1) 29.9(2) 1.279(2) BD2 § 439 (432) 23 19 0.018 0.334 1.029(4) 1.144(2) 14.0(1) 1.498(2) CGTB § 193.3 1.5 1.4 0.040 0.267 0.857(2) 1.169(1) 40.3(2) 1.291(1) CICD § 131 (100) 3 3 0.027 1.400 0.716(8) 1.169(2) 2.73(4) 1.246(2) CIE2 § 239 (220) 4 3 0.050 2.648 0.909(3) 1.094(3) 5.18(3) 1.185(3) CIF § 310 6 6 0.045 0.331 0.978(3) 1.140(1) 36.4(2) 1.336(1) CIG § 312 10 9 0.042 0.362 0.985(5) 1.161(1) 34.3(2) 1.389(2) CIH § 289 (273) 20 17 0.027 1.093 0.966(14) 1.147(1) 5.9(1) 1.332(2) HB2B § 408 (400) 15 14 0.048 0.793 0.996(3) 1.060(1) 14.0(1) 1.190(1) HLC2 § 121 1 1 0.048 0.108 0.664(2) 0.942(1) 47.4(3) 0.918(1) LB3 § 296 (288) 9 8 0.037 0.687 0.960(5) 1.102(2) 12.8(1) 1.235(2) LB4 § 315 (307) 7 6 0.034 0.652 0.964(3) 1.068(1) 12.8(1) 1.165(1) LB5 § 293 6 6 0.024 0.126 0.964(4) 1.125(2) 43.4(3) 1.286(2) SW3 § 293 (281) 14 12 0.063 2.092 0.969(9) 1.150(2) 8.2(1) 1.344(2) SW5 § 254 (245) 3 2 0.056 1.415 0.935(2) 1.143(1) 10.5(1) 1.294(1) SW6 § 238 (232) 4 4 0.054 0.907 0.917(4) 1.143(2) 14.5(1) 1.281(2) SWAN § 306 4 4 0.040 0.032 0.964(2) 1.088(1) 228(1) 1.209(1) SWIM § 352 6 6 0.068 0.708 0.993(2) 1.108(1) 27.6(1) 1.292(1) WG1U § 307 5 5 0.052 0.440 0.973(3) 1.124(1) 34.0(2) 1.295(1)

LOBO † 265 (256) 26 21 0.040 0.307 0.957(20) 1.206(12) 9.6(2) 1.44(2) LOAN † 302 (296) 30 24 0.041 0.180 0.983(18) 1.184(5) 16.7(3) 1.431(6) LB-1E # 329 31 24 0.041 1.078 0.989(14) 1.139(2) 3.6(1) LB-2F # 365 57 38 0.066 2.143 0.986(16) 1.066(1) 2.97(3) LB-3G # 362 26 21 0.059 1.257 1.020(10) 1.201(2) 5.0(1) LB-4H # 317 43 31 0.044 0.518 0.987(20) 1.157(8) 7.4(2) LB-O Isochron 311 28 22 1.6(1)

SWO † 183 (152) 23 19 0.047 1.062 0.850(41) 1.258(3) 3.0(1) 1.433(3) SW1-A # ∞ 0.107 6.611 1.006(17) 1.036(3) 2.33(4) SW1-2B # 221 6 5 0.115 2.847 0.889(7) 1.108(1) 2.49(2) SW1-3C # 305 16 14 0.075 3.249 0.952(8) 1.045(1) 1.98(2) SW1-4D # 570 8 ∞ 0.102 7.234 0.975(11) 0.956(3) 1.02(1) SW1-O Isochron 152 45 30 1.6(3) Note: The age in bold is the one used (see text for explanation). The value in parentheses beside the age is the calculated age if detrital contamination is assumed to have had an initial 230Th/232Th activity ratio of 1.0 (indicated by isochron date)—this is relevant only for 230Th/232Th activity ratios <20. Ratios are shown as activity ratios with 2σ error in parentheses. Half lives from Cheng et al. (2000). † VG thermal ionisation mass spectrometer (TIMS), McMaster University, Hamilton, Ontario. # Inductively coupled plasma mass spectrometer, Geotop, Uqam, Montreal, Quebec. § Triton TIMS, Isotope Geochemistry and Geochronology Research Centre, Ottawa-Carleton Geoscience Centre, Carleton University, Ottawa, Ontario.

202 Geosphere, August 2007 Kents Cavern Pleistocene History

is within error of the original age, so there is no that gradually decreases over the course of the well above the 3σ range. The one exception is justifi cation for using anything other than the Pleistocene, a very common pattern explained sample HLC2, fl owstone that is in situ, but has original age. Only three of the dates we have by depletion of the more soluble 234U in the an initial 234U/238U ratio below the 3σ range. This used are the adjusted ones. overburden (Fig. 5). The values for the mate- sample is from a high part of the cave, close to Initial 234U/238U ratios for the in situ fl ow- rials that are not in situ, but rather have been the surface with only a thin overburden. stone samples are all within a narrow range transported from another part of the cave, are RESULTS

4 Al-Be Dating SW1-O 20 LB-O An analysis of quartz sand from the basal 3 15 Red Sands of the Gallery (Figs. 6A, 6B) and quartz pebbles from the exposed Breccia in the

232 Bear’s Den–Labyrinth area (Fig. 6C) yielded 10

U/ Th 2 U/ Th fi nite 26Al-10Be dates, albeit with very broad

234 232

234 234U/ 238 U = 5 234U/ 238 U = 1 1.37 0.16 1.20 0.02 A 0 1.0 2.0 3.0 0 5 10 15 238U/ 232 Th 238U/ 232 Th

20 SW1-O LB-O 3 15

B Top of exposure 2 10 Fluvial Gravel

Th/ Th

Th/ Th

230 232

230 232 230Th/ 234 U = 5 230Th/ 234 U = Current-bedded 1 0.79 0.08 0.97 0.01 sands Fluvial clay 0 1.0 2.0 3.0 0 5 10 15

234U/ 232 Th 234U/ 232 Th Gravel fallen into pit from top 10 cm

Figure 4. Isochron plots of samples SW1-O and LB-O. C

B20: 4.07 In-situ flowstone Transported in breccia 1.6 SW-1 LB-O

ratio 1.4

1.2

/1.0 U activity HLC2 Figure 6. (A) The basal Red Sands from

234 238 lU 0.8 2rangeσ the Gallery. (B) Cross section through the 3rangeσ basal Red Sands showing, from the base up, Initia 0.6 fl uvial mud, current-bedded fl uvial sands, 0 100 200 300 400 and gravel. (C) The top of the Breccia, from the Water Gallery, looking up (stand- Age (ka) ing in the vacuity enlarged by Pengelly’s Figure 5. Change in initial 234U/238U activity ratio over time. The in situ fl owstone samples excavation in the Breccia), showing matrix- are shown as solid dots and the transported material as open circles. The dashed line shows supported angular clasts and bone arti- the 95% confi dence interval, and the dotted line shows the 99% confi dence interval. facts (white arrow).

Geosphere, August 2007 203 Lundberg and McFarlane error margins (Table 1). Unfortunately the Red Hedges Boss 9, 7, and 5 are presumed, based on data from Sands are only exposed in this one site and Of all the sites, Hedges Boss (Fig. 7A) other parts of the cave. Proctor et al.’s (2005) the relationship with Breccia is not explicit. has the simplest stratigraphic relationships. sample KC–90–2, taken from just below the The Gallery Red Sands yielded a date of The Breccia (>2 m thick here) is capped hiatus (a position confi rmed by the large saw- 2.32 ± 1.74 Ma (1σ range: 0.58–4.06 Ma) and by ~10 cm of orange-pink, laminated cal- cut remaining from their sampling), shows the Breccia yielded a date of 0.95 ± 0.55 Ma cite (sample HB-2B) that was dated to 408 evidence of recrystallization. Our sample was (1σ range: 0.4–1.5 Ma). +15/–14 ka, early MIS 11. The sequence con- not taken close to the hiatus and shows the tinues with a thin layer of red, detritus-rich clearly defi ned fi ne growth laminations of the U-Th dating calcite that in this cave indicates a hiatus in original fabric. Thus there is no reason to sus- deposition and/or very slow deposition. This is pect this date. Data are presented in Table 2. We present the capped by ~10 cm of white, laminated calcite results from each site. The cave stratigraphy is with several poorly expressed hiatuses, and Bear’s Den complex, such that no one site offers a continu- a thin layer that is currently active (MIS 1). The stratigraphic relationship of the Breccia ous sequence of events; instead the sequence Figure 7B shows the diagrammatic interpre- and the calcite is not simple here (Fig. 8A). must be reconstructed from the partial informa- tation of this section. We have no other dates Sample BD-1 (326 ± 7 ka, MIS 9) is a clean, tion from many sites. from this site; therefore the layers from MIS orange-pink, laminated calcite from immedi-

Active flowstone A MIS 1 B MIS5?

MIS7?

Active flowstone MIS9?White White laminated calcite calcite

Pink laminated calcite MIS 11 Hiatus Pink calcite Hiatus HB-2B 397– 423 ka HB-2B: 408 -14/+15 ka MIS 11 Breccia

Breccia Sampling sites in Hedges Boss Figure 7. (A) The remains of Hedges Boss perched on a pedestal of thick breccia. This is the only part of the Crystalline Stalagmite in this chamber that was thick enough to have not been fractured (see discussion in text). (B) Diagrammatic interpretation of depositional sequence.

MIS 1? B A MIS 5c MIS 5e Proctor’s KC4-83-C MIS 7 ? 115±8kaMIS5c MIS 1? drapery MIS 9 Proctor’s KC4-83-C Laminated calcite MIS 11 107–123 ka

Laminated calcite BD-1 326 Breccia BD-1 ±7 ka, MIS 9 Rock 319–332 ka BD-2: 439 Bedrock Flowstone shelf/slab -19/+23, MIS 11 BD-2: 420–462 ka cave wall Breccia Breccia Breccia Sampling sites in Bear’s Den Figure 8. (A) Bear’s Den and Bear’s Den Boss, where the fl owstone was thick enough to escape fracturing. The roof shows phreatic cusps with no evidence of collapse. (B) Diagrammatic interpretation of depositional sequence.

204 Geosphere, August 2007 Kents Cavern Pleistocene History

ately above the Breccia-calcite contact, ~6 m calcite fl owstone is a fl ake against the rock but The focus for the main site was on the broken to the right of the Bear’s Den Boss (inset separated from rock by a thin fi lm of mud, which fl owstone in the middle (Figs. 9A, 9D). The Fig. 8A, placed in correct stratigraphic posi- we interpret as Breccia remains. sequence shows that the upper layers of white, tion). Sample BD-2 (439 +23/–19 ka, MIS 11) laminated calcite (sample G2-BW, 79 ± 3 ka, is from a layer of white calcite coating bedrock Clinnick’s Gallery MIS 5a) were fractured and displaced, and that curves out into the Breccia. Proctor (1994) Here the stratigraphic sequence between cemented in place by thin layers of red calcite interpreted this as a calcite vein intercalating the two sampling locations is not continuous. (sample G2-TR, 47 ± 3 ka, MIS 3). These were between two beds of breccia. Having cleaned Figure 9B (shown in correct position relative in turn fractured and cemented in place by the the face with wire brush and water, we inter- to the main sampling site, Fig. 9A) shows the topmost layer of white vuggy Granular Stalag- pret this calcite as simply fi lling a narrow gap contact with Breccia. Immediately on top of mite of MIS 1 age. Figure 9C shows the dia- between the Breccia and the rock, probably the Breccia is ~25 cm of soft, vuggy calcite. grammatic interpretation. created by dripping, and thus postdating the We sampled the solid, cream colored, opaque Breccia (Fig. 8B). calcite above this (CG-TB:193 ± 1 ka, MIS 7). High Level Chamber The calcite is separated from the rock face by In the High Level Chamber (Figs. 10A, Water Gallery a thin fi lm of mud, which we interpret as the 10B), the remains of calcite fl owstone with a Because so little material remained, only remains of Breccia. No hiatuses are apparent natural fracture surface (rather than the typical one sample (WG1U: 307 ± 5 ka, MIS 9) was between the contact and the sampling site; we clean and scraped surface from excavation) are taken of the calcite layer immediately above the thus assume that the entire calcite layer was plastered to one side of the passage at ~1.5 m Breccia. All that is left of the white, laminated deposited during MIS 7. above the present fl oor. The lowermost 10 cm

C A G2-TR White Granular Stalagmite 47 ±3 ka MIS 3 Fractured blocks of red calcite White “Granular Stalagmite” Red laminated calcite coating G2-BW Broken red calcite Fractured blocks of 79 ±3 ka MIS5c? white laminated calcite Red calcite MIS 5a G2-TR 44–50 ka

MIS5e? White laminated calcite Broken slabs of white, laminated calcite CG2-BW 76–82 ka

CG-TB 193 ±1 ka MIS 7 White laminated calcite MIS 7 White vuggy calcite Bedrock cave wall Breccia Bedrock

30 cm B Bedrock Laminated calcite CG-TB: 192–194 ka Vuggy calcite

Breccia

D

Figure 9. (A) The main sampling site in Clinnick’s Gallery. Here the fl owstone is not in contact with breccia. (B) The secondary sampling site of fl owstone in direct contact with breccia. Mallet is 30 cm long. (C) Diagrammatic interpretation of depositional sequence. (D) Detail of broken fl owstone layers.

Geosphere, August 2007 205 Lundberg and McFarlane of fl owstone is not amenable to dating (labeled Figure 11A (a circumferential mosaic). Much drapery is currently active and assigned to MIS as sugary calcite in Fig. 10A). Sample HLC-2 of the most clearly exposed face to the left 1. Figure 11B shows our reconstruction of the (121 ± 1 ka, MIS 5e) comes from the fi rst layer could not be sampled for aesthetic reasons. Of sequence of deposits. of solid, laminated calcite. No hiatuses are the two samples taken in direct contact with the apparent between the contact and the sampling Breccia, CEB was leached and CI-HB dated to Labyrinth site; we thus assume that the entire calcite layer 289 +20/–17 ka, MIS 9. All three dates from This thick layer of fl owstone emerging from was deposited during MIS 5e. On close exami- this lowermost block of clean, white calcite are the Little Oven Exit that originally spilled into nation, it is apparent that it had been fractured not statistically separable, showing that depo- the Labyrinth (Figs. 12A, 12B) was diffi cult and cemented with overgrowth of red, muddy sition was rapid in MIS 9 (CI-G, 312 +10/–9 ka to sample. Much of the calcite is vuggy, and calcite. The site is further complicated by the and CI-F, 310 ± 6 ka). The thinner parts of this appears to have been deposited in shallow stand- remains of white, vuggy calcite both above the layer to either side of the main boss are cracked ing water. The dates (LB-3: 296 +9/–8 ka; LB-5: 5e deposit and just below it. Proctor’s (1994) (see following discussion). It was originally 293 ± 6 ka; LB-4: 315 +7/–6 ka) suggest rapid alpha-counted U-Th date on this material (the presumed that sample CI-E2 (220 +4/–3 ka, deposition in MIS 9. Two samples were taken lower layer) is 53 +6/–4 ka, placing it in MIS 3. MIS 7) would be stratigraphically equivalent from within the Breccia. The fi rst, a broken slab to CDF (150 ± 5 ka, MIS 6b), but the dates of fl owstone, LB-O, was isochron dated to 311 In-Between Boss prove otherwise. The remaining samples show +28/–22 ka, MIS 9. The second, a piece of sta- This boss, between Inscribed Boss to the north a clear progression through MIS 5 (CCD, at lagmite, broken and embedded in the Breccia, and Hedges Boss to the south, unnamed on the 100 ± 3 ka, CCT, at 95 ± 4 ka, CBN, at 93 sampled ~5 m to the left of this site (sample Proctor and Smart (1989) survey, is shown in ± 2 ka, and CAT, at 80 ± 1 ka). The overlying B20, shown in correct relative stratigraphic

Bedrock A cave wall Vuggy white calcite Laminated calcite B False Floor remnant

HLC-2: 120–122 ka HLC-2: 121±1 ka Laminated calcite MIS 5e Sugary calcite SugarySug calcite Proctor’s KC91-1 53 -4/+6 ka Vuggy calcite MIS 3 Proctor’s KC91-1 Vuggy calcite BreccBreccia Sampling sites in High Level Chamber

Figure 10. (A) High Level Chamber sampling site. (B) Diagrammatic interpretation of deposits. B MIS 1 A Hiatus CAT: 80 ±1 CAT: 79–81 ka CBN: 93 ±2 MIS 5a CBN: 91–95 ka CCT: 95 ±4 CI-F: 304–316 ka CCT: 91–99 ka CICD: 100 ±3 MIS 5c CI CD: 97–103 ka Hiatus CI-E2: 217–224 ka CDF: 150 ±5 Hiatus lined with CDF: 145–155 ka MIS 6b CI-G: 303–322 ka red flowstone CI-E2: 220 -3/+4 MIS 7 Hiatus CI-F: 310 -6/+6 CI-HB: 272–309 ka CEB: Leached CI-G: 312 -9/+10 Pale Breccia Contact of flowstone MIS 9 calcite with breccia Slab of flowstone within breccia CI-H: 289 -17/+20

30 cm Sampling sites at In-Between Boss, Cave of Inscriptions Breccia

Figure 11. (A) Photomosaic around In-Between Boss showing sampling sites. (B) Diagrammatic interpretation of sequence.

206 Geosphere, August 2007 Kents Cavern Pleistocene History

be radiometrically dated and the assigned age A B depends on a combination of paleontological, archaeological, and palynological remains, on stratigraphic position, and on sedimentology. Exit of Little Oven Passage Differentiation of the post-Anglian interglacial sites (e.g., Purfl eet and Hoxnian Interglacials) normally relies on aminostratigraphy. However, McCarroll (2002) suggested that this method is Vuggy calcite not sensitive enough to allow confi dent separa- Pool deposits Vuggy calcite Pool deposits tion of populations into different interglacials. It is now widely accepted that sites attributed to Hiatus the Hoxnian on the basis of pollen spectra may LB-4: 309–322 ka Hiatus Laminated LB-4: 315 -6/+7 ka, MIS 9 represent more than one warm stage (Scourse Laminated calcite calcite et al., 1999; Dowling and Coxon, 2001; Thomas, LB-5: 293 ±6 ka, MIS 9 Some pool deposits Some 2001). Schreve and Thomas (2001) suggested LB-5: pool deposits 287–299 ka Smallhiatus? that two episodes are recorded, each with a MIS 9 LB-3: 288–305 ka LB-3: 296 -8/+9 ka, MIS 9 Hoxnian-type pollen signature. As new data are published, the controversy lessens. For example, Broken Stalagmite B2O 210 ±2, MIS 7 Grün and Schwarcz (2000), from U-series/ESR LB-O: 289–339 ka ages of 403 +33/–42 ka on teeth, placed deposits Breccia Flowstone slab LB-O: 311 ka -22/+28, MIS 9 of the type locality of the Hoxnian Interglacial Breccia in MIS 11. Rowe et al. (1999), based on careful U-series dating of lake sediments, also assigned the Hoxnian to MIS 11. The Anglian is the most widely recognized Sampling sites in Labyrinth, under Little Oven Exit event in the mid-Pleistocene of Britain, yet Figure 12. (A) Sampling sites from the northeast side of Labyrinth. (B) Diagrammatic inter- its age is not conclusively established. The pretation of deposits. MIS 11 date for Hoxnian deposits from Rowe et al. (1999) also assigns the Anglian glacial deposits that underlie the Hoxnian deposits, with no evidence of any signifi cant break in position in Fig. 12B), was dated as 210 ± 2, dated as 152 +45/–30 ka, MIS 6b, and shows deposition, to MIS 12. The majority of publi- MIS 7. These last two have initial 234U/238U an initial 234U/238U ratio outside of the 3σ range cations refer the Anglian to MIS 12 (Schreve ratios well outside the 3σ range for the in situ (Fig. 5), indicating that it probably originated in and Thomas, 2001). fl owstone (Fig. 5, open circles), suggesting that another part of the cave. Cromerian sites from Britain have, in the past, they originate from elsewhere in the cave. been assumed to represent a single interglacial DISCUSSION stage, whereas in the Netherlands the Cromerian Southwest Chamber Complex has incorporated four interglacials and We sampled from both the northwest In order to understand the evidence of Qua- their intervening cold stages (Preece, 2001). (Figs. 13A, 13B) and southeast (Figs. 13C, 13D) ternary events in Kents Cavern it is important Recent molluscan evidence has supported sev- sides of this passage. The northwest side shows a to be aware of some of the complexities of cave eral distinct stages (Preece, 2001; Schreve and simple sequence: SW-3 (293 +14/–12 ka, MIS 9) sedimentology in general. The age of any one Thomas, 2001). An age of MIS 13 is gener- in direct contact with Breccia, up through SW-5 deposit cannot usually be predicted from its ally agreed as the younger limit, but there is (245 +3/–2 ka, MIS 7) and SW-6 (238 ± 4 ka, geomorphological position because the princi- no consensus about the age of the start of the MIS 7), and red vuggy calcite on top that was ple of superposition does not necessarily apply Cromerian; Parfi tt et al. (2005) proposed at least not amenable to dating. On the southeast side, in caves. In some sections deposition appears MIS 17 for the Cromer Forest bed, and by impli- the calcite remains plastered high on the wall to have obvious hiatuses, but dating shows that cation, the Cromerian type site at West Runton. consist of the lowermost layer of rapidly depos- they are in reality simply shifts of the drip point In the following discussion we assume that ited, dendritic fabric, full of intercrystallite voids (e.g., some parts of In-Between Boss). Rework- the Hoxnian stage correlates with MIS 11, the and not amenable to dating; an intermediate ing of nonindurated sediments may complicate Anglian stage with MIS 12, and the Cromerian layer of white, laminated calcite (WUXU, dated temporal relationships: we believe that the clas- is MIS 13 to at least MIS 15. imprecisely to 417 +104/–52 ka, MIS 11); and tic sediments in this cave have been repeatedly The distribution of the standard four units, white laminated calcite (SWAN, 306 ± 4, MIS remobilized since initial deposition, confusing i.e., the Breccia, the Crystalline Stalagmite, 9). The lower part of the southeast side shows the sequential relationships. the Cave Earth, and the Granular Stalagmite, a sequence of calcite in contact with breccia It is also important to be aware of the prob- was described by Pengelly (1884), Keith et al. (SW1M, 352 ± 6 ka, MIS 10b); a hiatus; a thin lems of inconsistent application in the literature (1931), Campbell and Sampson (1971), Proctor layer of redder calcite (WTXT, 132 ± 6 ka, MIS of British stage names and associations with (1994), Straw (1997), and Proctor et al. (2005), 5e); and vuggy, Granular Stalagmite (MIS 1). the marine oxygen isotope record. The mid- but without a reliable temporal framework for The sample taken from within the Breccia, of a Pleistocene record from British sites is some- materials outside the range of 14C dating. Proc- fractured fl owstone slab, SW1-O, was isochron what hazy because so many of the sites cannot tor (1994) provided detailed descriptions of

Geosphere, August 2007 207 Lundberg and McFarlane

Bedrock Sampling sites in Southwest cave wall Chamber, NW side A Red vuggy calcite B Red vuggy calcite Pool deposits Pale Whit Pool deposits laminatedish calci Whitish calcite te SW-6: 234–242 ka calci SW-6: 238 ±4 ka te Pale, laminated calcite MIS 7 Red calcite SW-5: 245-2/+ 3 ka Red calcite Pale, laminated calcite MIS 7 SW-5: 243–248 ka Pale, laminated Red calcite coating Bedrock Stalagmite enclosed SW-3: 293 -12/+14 ka cave wall calcite in flowstone MIS 9 Breccia Breccia SW-3: 281–307 ka Bedrock Stalagmite cave wall Breccia

MIS 7? Vuggy red calcite SWAN: 302–310 ka C D Bedrock SWAN: 306 ±4 ka, MIS 9 Haitus cave wall WUXU: 417 ka Laminated calcite WUXU: 417 -52/+104 ka, MIS 11 Dendritic Bedrock Dendritic calcite calcite cave wall Pink, vuggy Granular Stalagmite MIS 1 Pink, vuggy calcite Red, laminated calcite WTXT: 126–138 ka WTXT: 132 ±6 Hiatus MIS 5e SW1M: 346–358 ka SW1M: 352±6 ka Pale, laminated calcite MIS 10b Laminated calcite Flowstone slab SW1-O: SW-1 Old: 152 ka 152 -30/+45 ka Breccia MIS 6b Breccia Sampling sites in Southwest Broken flowstone slab Chamber, SE side

Figure 13. (A) Sampling sites on the NW side of Southwest Chamber. Mallet is 30 cm long. (B) Diagrammatic interpretation of deposits. (C) Sampling sites on the SE side of Southwest Chamber. The fi gure points to the upper level samples. The stratigraphic relationship between the upper level sequence and the lower level sequence cannot be traced. (D) Diagrammatic interpretation of deposits.

facies. Here we focus on the information from pit dug into the fl oor of the Gallery, but the rela- est possible age on the Cave Earth (i.e., end of the new dates and present a new interpretation tionship with the Breccia is ambiguous because MIS 5a, 79 ka) is only 0.09 (Monte Carlo simu- of the stratigraphy. the sands are separated from the overlying lation, 10,000 iterations). Thus Proctor’s (1994) Breccia by a substantial vacuity. The Red Sands hypothesis that the Gallery sands are a remnant Age and Emplacement of the Red Sands (Fig. 6A) comprise a basal unit of horizontally of the Cave Earth can be rejected. bedded mud; a middle unit of bedded sands that The 26Al-10Be data support a signifi cantly A sometimes-overlooked deposit, the Red dips steeply ~30° away from South Entrance; greater age for the Gallery sands versus the Sands, has been variously interpreted. Keith and an upper unit of poorly bedded gravel, the Breccia, and support Campbell and Sampson’s et al. (1931) reported the Red Sands as the basal clasts of which are subrounded, poorly sorted, (1971) view that the Gallery sands are basal to unit below the Breccia in the Water Gallery, but mostly 1–2 cm but as much as 5 cm in diameter. the Breccia and synonymous with Keith et al.’s this exposure is now buried by a cement path- We interpret this as a simple fl uvial sequence (1931) Red Sands, making them the oldest way and cannot be reexamined. Keith et al. of low-water mud, current-bedded sands, and dated deposits preserved in the cave. In addi- (1931) and Campbell and Sampson (1971) cor- high-water river gravel (Fig. 6B). tion, examination of the 26Al nuclide concen- related these sands with the Red Sands currently We address the relationship of the Red Sands trations in Gallery quartz versus Breccia quartz exposed in the Gallery, but Proctor (1994) inter- to the Cave Earth and the relationship of the Red reveals that they differ by more than 4 standard preted the unit as a wash facies of the loamy Cave Sands to the Breccia. Even with the very large deviations (Table 1), demonstrating that the Earth (well dated by 14C as late MIS 3, Fig. 2). 1σ error, the probability that the 26Al -10Be burial exposure history of these materials has been The Gallery exposure remains accessible, in a age of the Red Sands is younger than the earli- quite different.

208 Geosphere, August 2007 Kents Cavern Pleistocene History

We suggest that the Red Sands are simple (474–620 ka; Bassinot et al., 1994), but also consists of the chronospecies Mimomys plio- fl uvial deposits indicating a fl ow with a high incorporates the full glacial at MIS 16 together caenicus (early Pleistocene), M. savini (late-early clastic load, of variable discharge, moving up with at least one preceding interglacial (Parfi tt through early-middle Pleistocene), and A. canti- South West Chamber and diverging through et al., 2005). The Breccia is unlikely to date to the ana (late-middle Pleistocene to early-late Pleis- the Gallery toward the Long Arcade and North early Cromerian: if the cave were open and the tocene) (Lister et al., 1990). Neither Mimomys Entrance. While no defi nitive date can be Breccia had been produced in the MIS 16 gla- species is known from the Kents Cavern Brec- assigned (in view of the high error margin of cial, the Breccia and the cave would have to have cia. The appearance of unrooted molar teeth the cosmogenic date), it would be reasonable to remained completely unaffected by the 146 k.y. characteristic of the genus Arvicola is dated to assume that the Red Sands are Cromerian (MIS of Cromerian temperate conditions and another MIS 11 in continental Europe (Pevzner et al., 13–15, a long period of predominately temper- 46 k.y. of Anglian glacial conditions until the 2001), but to the late Cromerian (Sutcliffe and ate conditions), and represent the fi nal stages of initiation of calcite deposition in MIS 11. Thus Kowalski, 1976) or immediately pre-Anglian vadose activity after the phreatic activity that we conclude that the Breccia must have been (Andrews, 1990) at the Westbury site in Britain. carved the primary cave passages. produced in a post–MIS 16 cold stage. The only This sets the maximum age limit for the Kents stage that is cold enough between the oldest Cavern Breccia fauna at MIS 13. The extinct Age of the Breccia date on the calcite, MIS 11, and MIS 16 is MIS Pine vole, Pitymys gregaloides, is also present 12, the Anglian glacial period (MIS 14 was too at Westbury but has not been found in any Brit- The 26Al-10Be burial date on the quartz grains warm). The question is not resolved by other ish Hoxnian (MIS 11, sensu Schreve, 2001) site within the Breccia gives an age estimate of mid- head deposits, because so few have been reliably (Sutcliffe and Kowalski, 1976), and is generally Pleistocene. The oldest of the U-Th dates on the dated. Bates et al. (2003) indicated that most are presumed to be indicative of pre–MIS 11 age. capping fl owstone offers an upper window. How- from MIS 4 and MIS 2, some are from MIS 6 Thus the rodent fauna are MIS 13 in age (we ever, the possibility of a potential hiatus between or MIS 8, but in a few places with good dat- agree with Proctor et al., 2005). The deposit sediment deposition and calcite deposition must ing control, head deposits dating to the marine in which they are incorporated may also be be acknowledged in view of Stock et al.’s (2005) oxygen isotope stage 12 can be recognized. For MIS 13, but could equally be younger. If we fi ndings that U-Th dates on speleothem over- example, head deposits from cold stages back accept that the Breccia represents a cold-climate lying clastic sediments are often considerably to MIS 12 have been found (with ages rang- (periglacial) deposit, and that the oldest calcite younger than cosmogenic 26Al-10Be burial dates ing from MIS 11, on amino acid evidence, to capping is MIS 11, then the only possible age of on the sediment. In Kents Cavern, the Breccia MIS 13, on mammalian biostratigraphy) in the emplacement would be MIS 12. contains additional evidence of the time of its Hampshire-Sussex coastal plain. Murton and formation in the form of sedimentological char- Lautridou (2003) also dated periglacial deposits Evidence From the Bears acteristics and paleontological remains. along the English Channel coastlands (by radio- The Kents Cavern Breccia is most notable for We dated 11 samples of calcite fl owstone carbon, luminescence, and mammalian biostra- yielding large numbers of teeth and bones attrib- in direct contact with the Breccia: 2 of these tigraphy); they place most in MIS 2, some in uted to the cave bear, a term that is applied to ani- yielded MIS 11 dates, 5 had MIS 9 dates, 2 had MIS 6, and some in other cold stages. mals in the Ursus savini–Ursus deningeri–Ursus MIS 7 dates, and 1 had a date of MIS 5. The speleaus chronospecifi c lineage. The exact posi- simplest interpretation from the capping calcite Paleontological Evidence tion of the Breccia bears along this lineage has is that the Breccia formed before MIS 11 and Although their U-series ages provide only been ambiguous, in part because the material the younger dates represent hiatuses of various an upper window of MIS 9, suggesting that the has not been formally reviewed and also because lengths. Proctor et al. (2005) tried to date the Breccia could represent either MIS 12 or MIS the systematics of European middle Pleistocene Breccia using the calcite cap, but they took the 10 deposition, Proctor et al. (2005) argued that bears has not been fully resolved. Bishop (1982) average (MIS 9) as an indication of minimum the faunal remains in the Breccia suggest a late placed the Westbury (MIS 13) bears at full age. Any averaged time series will yield a mean Cromerian age, consistent with many examples U. deningeri grade. Schreve (2001) considered that is signifi cantly younger than the actual age of well-established pre-Anglian faunas in UK bears of full U. speleaus grade to be Hoxnian of the initiation of the series. sites and in the Netherlands. The paleonto- (MIS 11) in age. A preliminary analysis of the logical evidence comes from both the rodent Breccia bear teeth (McFarlane et al., 2006) indi- Sedimentological Evidence remains (generally disseminated throughout the cates that they are of an advanced U. deningeri– Sedimentologically, the Breccia is a poorly Breccia) and bear remains (generally at the top early U. speleaus grade, consistent with an age sorted diamict of angular to subangular clasts of the Breccia). of MIS 12–11, but clearly younger than MIS 15 (red sandstone, siltstone, slate, quartz, and rarely (U. savini; Kurtén, 1968), and probably younger limestone) in a matrix of red mud (Fig. 6C). Evidence From the Rodents than the Westbury MIS 13 bears. A small pro- Proctor (1994) noted that the clasts are as large Pengelly’s notes (1868–1880) do not address portion of generally less well preserved bear as 20 cm and are matrix supported, and that little the microvertebrate fauna of Kents Cavern, but material is found disseminated throughout the or no fabric is visible in the generally homoge- workers have identifi ed sparse remains of the deposit, but the majority of the bear material is neous deposit. This material is typical of frost- voles Pitymys gregaloides, Arvicola “greeni” found within the uppermost 25 cm, immediately shattered regolith or head deposits that abound (= A. cantiana; Sutcliffe and Kowalski, 1976) below the calcite cap, indicative of a superfi cial on the hillsides of the area as a result of former (Campbell and Sampson, 1971), and Microtus emplacement, so much so that MacEnery (1859, periglacial activity (Cullingford, 1982; Scriv- oeconomus (Proctor, 1994). The evolution of p. 27; our brackets) reported “The fi rst fl ag [of ener, 1987; Croot and Griffi ths, 2001). This voles is of great importance in the middle Pleis- overlying fl owstone] that was turned over exhib- requires a long period of cold conditions. The tocene biostratigraphy of Europe, and provides ited, in relief, groups of skulls and bones adher- Cromerian represents a long period of temper- time constraints on the Kents Cavern Breccia ing to the stalagmite.” The majority of the bones ate conditions from MIS 13 to at least MIS 15 deposits. The British extinct water vole lineage are from bears using Kents Cavern as a hiber-

Geosphere, August 2007 209 Lundberg and McFarlane naculum, on top of the Breccia, and the bones reiterated in Proctor et al., 2005). This suggests (1986) stated that excavators were too ready to are incorporated into the topmost muddy layer. a warm or at least warming climate. Collcutt blame processes such as cryoturbation. While Thus most of the bear material postdates the (1986) noted that some sorting occurs during this many be true, we suggest that not all matrix- MIS 12 emplacement of the Breccia. debris fl ow, the larger particles moving to the supported, relatively homogeneous deposits in sides and base; that for high-water-content fl ows caves must necessarily now be reinterpreted as Evidence From the Artifacts there is a tendency for orientation of particles water-saturated debris fl ows, and that solifl uc- Much of the historic interest in Kents Cav- with fl ow; and that larger clasts or artifacts and tion clearly continues to produce sediments with ern has resulted from the recovery of very early faunal remains may be signifi cantly worn and features similar to those in the Breccia. Acheulian fl int and chert artifacts. Acheulian damaged by the fl ow. While there is little clear The time frame available for breccia emplace- bi-face tools are known elsewhere in Britain evidence of sorting, damage to the artifacts dis- ment was relatively narrow: the head deposit from Boxgrove (MIS 13) and Pakefi eld seminated throughout the Breccia (Cook and that is the basis of the Breccia was formed by (MIS 17; Parfi tt et al., 2005). Campbell and Jacobi, 1998) supports this view. However, it is frost action during MIS 12; the bears moved in Sampson (1971) noted that Pengelly recovered apparent to us that the features of this deposit and used the top of the Breccia for a den; the these artifacts throughout the thickness of the could equally be interpreted as a solifl uction top of the Breccia was partly encrusted in cal- Breccia, including its lowest levels. Moreover, fl ow (see following). Croot and Griffi ths (2001) cite during MIS 11 while the bears continued to the largest number of artifacts (31%) came observed that the head material on the surface use the area. Thus the bulk of the Breccia had from the “4 ft” level within the Breccia. Several has been moved downslope by solifl uction, or to have been emplaced at the end of MIS 12 authors (Campbell and Sampson, 1971; Cook by rapid “slushfl ows” and slides, or by torren- before the bears moved in. If the Breccia is a and Jacobi, 1998) have commented on the poor tial fl ash-fl ood events. We suggest that the head wet debris fl ow, then the time for emplacement condition of the artifacts, which show evidence material could equally have moved into the cave had to have been during termination V or very of both rolling and rotting, consistent with long by such processes. early MIS 11. exposure on the surface and subsequent entrain- Bertran et al. (1997) defi ned solifl uction as ment in the Breccia debris fl ow. slow mass movement caused by freezing and Post-Breccia Calcite Flowstone The archaeological and paleontological evi- thawing, combining small-scale downslope frost dence suggests that the Acheulian artifacts accu- creep and viscous fl ow from rapid release of The appellation of all the Breccia-capping mulated on the surface in pre-Breccia time, and meltwater during thawing. It produces lobes and calcite as the Crystalline Stalagmite implies a were subsequently carried into the cave entrained sheets, usually with a preferred clast orientation simple sequence of synchronous events. In fact, in the Breccia debris fl ows. In contrast, the bear parallel to slope. Debris fl ows are rapid mass except for the relatively clear case of Hedges remains are both spatially and taphonomically movements of poorly sorted solids and water, Boss, the relationship of the Breccia and the consistent with having been derived from a bear usually triggered by rain events but also reported overlying fl owstone is rarely simple and rarely hibernaculum in the cave, on top of and post- as resulting from rapid melting of ground ice synchronous. dating the Breccia. Subsequent remobilization triggering retrogressive thaw slumps. Fabrics In the Bear’s Den Boss, the visible stratig- of the Breccia (see following) has incorporated for both these deposits have many analogies: raphy suggests a lower layer of breccia, an some of this bone material in its upper layers. both show clast orientation parallel to slope, and overlying layer of fl owstone, a second layer of The provenance of the rodent remains is not sometimes imbrication in clast-rich deposits. breccia, and a cap of fl owstone. Proctor et al. known with certainty, and these animals may Data on clast orientation show that debris fl ows (2005) suggested that the lowest layer of fl ow- include specimens both coeval with and postdat- and solifl uction deposits are not clearly differ- stone was broken up, accompanied by local ing the emplacement of the Breccia. entiated, having a similar range of a-axis vec- reworking of the Breccia. An alternative view tor magnitudes. Both may exhibit simple shear is that the lowest layer simply represents an Emplacement of the Breccia deformation. Some differences between the exposed fl owstone ledge (typical of the edge two deposits can be seen, in that fabric strength of a boss) that was buried by postdepositional Generally the Breccia shows very little inter- is often lower in debris fl ows than in solifl uc- movement of the Breccia creating the complex nal structure. It is clearly not fl uvially emplaced, tion deposits, and that fabric shapes for debris pattern of breccia-rock-ledge that is outlined in but rather is a mass-movement deposit. Proctor fl ows are usually moderately developed girdles, Figures 8A and 8B. (1994) and Straw (1997) suggested that it is a whereas in solifl uction deposits clusters and gir- While the fi rst breccia-topping calcite debris fl ow, on the basis of the thick, structure- dles occur in equal proportions. Millar (2006) deposit was from MIS 11, the majority of dates less beds with poor sorting, chaotic to sub hori- observed that debris fl ows often produce imbri- indicate a signifi cant episode of calcite deposi- zontal clast orientation, and matrix support. cate fabric depending on the clast frequency and tion in MIS 9, e.g., almost 1 m of MIS 9 fl ow- Collcutt (1986) defi ned debris fl ows as water- slope, but in solifl uction the materials behave as stone in the In-Between Boss and in the Laby- saturated materials moving at speeds detect- a viscous fl uid and may also produce imbricate rinth. There was an even longer hiatus between able to the observer, up to tens of kilometers fabric, depending on clast frequency and slope. breccia emplacement and calcite capping in per hour. He considered them to be common Proctor’s (1994) description of breccia fabric other areas. The history of the breccia-calcite events in caves and the dominant mechanism is not detailed enough to allow application of contact from Clinnick’s Gallery (Fig. 9B) con- for lateral displacement of deposits. Initiation of Bertran et al.’s (1997) criteria. While the mode nects with that from Rocky Chamber, at its movement requires slopes of at least 20° and a of emplacement of the Breccia does not really northern end. The absence of any calcite cap- sudden input of large quantities of water. Thus, matter to its subsequent history, we suggest that ping the Breccia until MIS 7, and the presence Proctor (1994) suggested that the Kents Cav- the Breccia is a standard head deposit that was of red mud lining the junction of the calcite and ern Breccia debris fl ows would have required a moved into the cave (through suffosion dolines, the wall, suggests that the passage was com- high water content and that emplacement in the such as the Swallow Hole Gallery, and/or rifts) pletely blocked by breccia. The occurrence of cave was accompanied by small streams (a view either by solifl uction or by debris fl ow. Collcutt previously unreported paragenetic pendants

210 Geosphere, August 2007 Kents Cavern Pleistocene History

and anastomoses in the roof of Rocky Chamber Sequence of Events age of 20 ka. However, a further constraint on (Fig. 14A), together with the remains of a frag- the time window for cracking is the subsequent ment of breccia in the roof (Fig. 14B), suggests Any scenario for the sequence of events must deposition of Cave Earth that has been 14C dated that Rocky Chamber was also fi lled to the roof explain the formation and emplacement of all as ca. 23–35 ka (Fig. 2). Thus the original age with breccia. The paragenetic erosion suggests the sedimentary units, but it must also explain is the more likely, and the timing for cracking fl uvial action after emplacement of the MIS 11 two unusual observations. The fi rst is the evi- most likely to be MIS 4, the fi rst cold stage of Breccia and before deposition of MIS 7 calcite. dence for widespread cracking of fl owstone lay- the Devensian glaciation. We also have to fi t into We suggest that the most likely time was the ers and the second is evidence for the incorpora- this timing the cracking of the red calcite and MIS 9 interglacial period. The delay in depo- tion of fractured material of younger ages within its subsequent cementation with Granular Sta- sition of the calcite cap in High Level Cham- the upper layers of the Breccia. lagmite, well dated by 14C to MIS 1, ca. 4–16 ka ber (Fig. 10A) until MIS 5 is also probably (Fig. 2). The second episode of cracking must explained by the passage being largely blocked Cracking of Flowstone have occurred between MIS 3 and MIS 1. by breccia, although we observed no direct evi- Many examples can be seen where a layer of The evidence from High Level Chamber sug- dence of paragenesis here. fl owstone is cracked and shifted, and the cracks are gests that the MIS 5 fl owstone was fractured and Southwest Chamber (Fig. 13) shows evi- cemented with a thin coating of younger calcite. largely removed, and the Breccia surface level dence of an even more complex suite of events. Most of the evidence of cracking was removed lowered, before the subsequent deposition of The high-level remnants of MIS 11 and MIS 9 during excavations, but Straw (1997) recon- MIS 3 calcite under and slightly overlapping the fl owstone indicate that the Breccia must have structed, from Pengelly’s original reports, the dis- MIS 5 fl owstone. This cracking must also have been at least 2 m deep in order to provide a tribution of Crystalline Stalagmite. Pengelly dif- occurred during MIS 4. substrate on which the fl owstone grew. This ferentiated between Crystalline Stalagmite that Another example of cracking that we dated is must have been subsequently largely removed was cracked in situ and Crystalline Stalagmite from the In-Between Boss, but the evidence is not (in MIS 10) to allow deposition of the MIS 9 that was cracked and relocated, and was found as so clear cut. Here the thick basal fl owstone that fl owstone at the lower level. Drips from the detached fragments in the Cave Earth. The in situ was dated to MIS 9 was cracked and lined with roof continued to deposit calcite on top of the cracking was centered on the Bear’s Den, Laby- red calcite, but only in its thinnest part, where it is original MIS 11 layer (now a false fl oor or rinth, Cave of Inscriptions, and Clinnick’s Gal- ~20 cm thick. At the left side of the boss the basal ledge), but also deposited calcite on top of the lery. The relocated fragments found in the Cave layer of MIS 9 calcite is cracked and overlain by newly eroded Breccia at the same time. Thus in Earth were moved from the Vestibule southward MIS 7 calcite (sample CI-E2 220 +4/–3 ka). The a single passage the thin remnants of fl owstone and into North and South Sally Ports. cracking here must have occurred during MIS 8. represent two layers that were contemporary Only one attempt to date the time of cracking However, at the right side of the boss, the layer but vertically separated by several meters. has been published. Proctor (1994) was able to above the red calcite that cements the cracks in The evidence is clear that the Crystalline delimit the date of cracking of the edge of Bear’s MIS 9 material is dated as MIS 6b (sample CDF, Stalagmite represents a complex of fl owstone Den Boss to after 115 ka (±4 ka), but this exam- dated twice, 150 ± 5 ka and 148 ± 9 ka), giving a layers, and that each part of the cave requires ple has no younger overlying intact calcite that wide window for cracking from MIS 9 to MIS 6b. detailed study. It is also relevant to note that could provide an upper age limit. We studied While the evidence is not unequivocal, we argue the appellation Granular Stalagmite is not several examples. The best is from Clinnick’s that if the cracking on this side had occurred dur- really adequate to represent the complex of Gallery (Figs. 9A, 9B, 9C). We dated the white ing MIS 8, then we would expect the crack to Holocene calcite deposits. The cave opened up fl owstone layer that was cracked and the thin red have fi lled with clastic material or calcite. The much more in the Holocene, so that deposition calcite that coated the broken fragments. This fact that the fi lling material dates to MIS 6b sug- in the entrance zones and main passages was gave a window for cracking and recementation gests that the cracking on this side had occurred thick, porous, and vuggy, as is characteristic of of 79 ±3 ka (the end of MIS 5a interglacial) to immediately before this, perhaps during MIS 6c, rapid, evaporative deposition. However, Holo- 47 ±3 ka (the MIS 3 interstadial). Observant the fi rst cold period of MIS 6. cene deposits in less open parts of the cave are readers will note the high detrital content of the The evidence is for several episodes of crystalline in fabric and still active. uppermost red calcite coating and the adjusted cracking, and each of the times indicated, MIS 8, MIS 6c, MIS 4, and MIS 2, suggests a cold period. Further evidence for cracking is dis- cussed in the following. AB10 cm Incorporation of Younger Material The second of the unusual events that must be explained in any sequence of events is the incorporation of fractured material of younger ages within the upper layers of the Breccia (in addition to blocks of angular bedrock). The three we sampled yielded dates of MIS 7, MIS 6b, and MIS 9 (B20—210 ka, SWO—152 ka, and LBO—311 ka). The standard explanation for mixing of materials of different ages is reworking Figure 14. (A) Rocky Chamber paragenetic anastomoses in roof (looking straight up). of the sediment. Thus the Breccia (at least the sur- (B) Remnants of breccia adhering to paragenetic pendants. Scale shows inches to left and face few decimeters) must have been remobilized centimeters to right. some time after its original emplacement.

Geosphere, August 2007 211 Lundberg and McFarlane

The evidence for postdepositional move- carrying MIS 11 bear bones from the Bear’s Mechanism for Fracture of the ment of the Breccia is clear. MacEnery (1859), Den toward the South Entrance, and was then Flowstone Sheets describing the cave before the major archaeo- covered by MIS 9 fl owstone. (This sequence is logical excavations, noted several examples of corroborated by the evidence in the Bear’s Den The fractured fl owstone has triggered much the Breccia having been higher, having pulled and Southwest Chamber, discussed next.) It speculation: MacEnery (1859) suggested tec- away from the overlying crust, leaving the roof is also interesting that the MIS 9 layer had not tonic activity as the cause, Pengelly (1876, overhead with bones sticking out of base of the been fractured; the Breccia must have been intact p.176) suggested hydraulic pressure as the calcite. The Water Gallery illustrates this. Fig- when this thin layer was deposited. Some time cause, and Proctor (1994) reverted to the earth- ure 15 shows our reconstruction of the cross after MIS 9 the Breccia was partly removed to quake theory. Straw (1997), based on Proc- section at the Lake–Water Gallery, based on create the Vacuity in the center of the passage. tor’s one date, assumed that it must have been Pengelly’s original fi eld notes (1868–1880; still If the Breccia underneath it was removed before caused by a single strong earthquake between detailed at this early part of his excavations; the next fracturing episode, then the MIS 9 ca. 100 ka and 75 ka. In response to Straw’s however, they became less thorough as the layer would remain intact (see discussion of (1997) article, Ford (1997) collected several dig progressed to other parts of the cave). Our mechanisms of fracturing): this second episode references to fractures observed in other caves date of 307 ± 5 ka (sample WGIU) places the of Breccia remobilization was thus probably in in Devon, and speculated that all of them are lower layer of crystalline stalagmite in MIS 9. MIS 8. There is no evidence of fl uvial activity in attributable to earthquake activity; however, he The dimensions and positions of the blocks of this movement of the Breccia: the redeposition offered no substantive support for the fracture Crystalline Stalagmite distributed throughout the of breccia plus bones farther down-passage still events. For example, Ford reassessed Sutcliffe’s Breccia were carefully documented in Pengelly’s bears the marks of mass movement. (1960) frost-heave interpretation of the fractured notes (1868–1880). The thin horizontal slab of The Southwest Chamber is one of the best- calcite sheet in Joint Mitnor cave as earthquake calcite just beneath the bone-rich layer of breccia documented passages because it was one of the damage. The evidence Ford invoked is largely may have been MIS 11, deposited in situ, and fi rst places to be excavated by Pengelly. We have negative; he argued that frost heave is unlikely then fractured during the MIS 10 glacial period. reconstructed the cross section from Pengelly’s 100 km from the glacial margin and in a region The Breccia became mobilized during MIS 10, notes (1868–1880), dated many of the calcite lay- under maritime infl uence, and, assuming that ers, and then reconstructed the most parsimonious frost heave requires the presence of permafrost, suite of events required to explain the deposits the only explanation is earthquake damage. Reconstruction of Water Gallery looking along Pengelly’s datum to South. Series 11, Parallel 1 to 4 (Fig. 16). The high-level remnants of MIS 11 and We present unequivocal evidence that the MIS 9 fl owstone indicate that the Breccia must fracturing occurred repeatedly throughout the have been at least 2 m deep in order to provide middle to late Pleistocene history of the cave, a substrate on which the fl owstone grew. Some and that it occurred during each cold episode of the Breccia must have been removed during since the fi rst layer of calcite was deposited in Granular Stalagmite 1m MIS 10 in order to make space for the lower level MIS 11. The arguments that the region could not MIS 9 calcite deposition. The Breccia must also have been cold enough have very little empiri- have been quite mobile, because artifacts from the cal support. Croot and Griffi ths (2001) mapped Crystalline Stalagmite: The Lake 112-127 cm thick den area were transported up this passage embed- fossil polygons and stripes within ~10 km of the ded in the top of the Breccia. A mobile breccia is cave and many examples of frost-related fea- The lower layer of Crystalline Stalagmite required until MIS 6a in order to incorporate the tures both on Dartmoor and at many coastal sites Thin layer of Breccia 7.5 cm thick on underside of very securely dated SW1-O MIS 6b fl owstone all around Devon and Cornwall. Some recon- WG1U: 307 ±5 ka Crystalline Stalagmite MIS 9 slab in the top of the Breccia. Another cracking structions of conditions at 20 ka in the region event, during MIS 6a, was required to break the ( Murton and Lautridou, 2003) show the whole The “Vacuity” MIS 6b slab. Pengelly’s notes (1868–1880) docu- region, including Torquay and Dartmoor, under mented a thick layer of Crystalline Stalagmite continuous permafrost, while others show perma- across the passage; this must have represented frost only on Dartmoor (and presumably sea- the complex of deposition from MIS 11–5. The sonal frost elsewhere), and some show discon- Breccia became indurated after MIS 6a, and fur- tinuous perma frost just a few kilometers north of ther movement stopped. Torquay. Evidence that Devon probably under- Thus the evidence supports additional cracking went discontinuous permafrost, at least during Slabs of Crystalline Stalagmite and bedrock in Breccia events, during MIS 10 and MIS 6a, and remobili- the coldest periods, was provided by Ballantyne zation of the Breccia during MIS 10, MIS 8, MIS and Harris’s (1993) fi nding of a fossil pingo on Granular Stalagmite flowstone Cave Earth 6c, and MIS 6a. Straw (1997; Figure 1 therein) West Dartmoor. Croot and Griffi ths (2001) sug- Crystalline Stalagmite flowstone mapped “detached Breccia pieces in Cave-Earth”; gested that during glacial periods the climate of Breccia: coherent, rock-like, very rich in bones this implies that the Breccia was also cracked, Devon may have been similar to that of Svalbard Breccia: Incoherent, poor in bones probably during MIS 2 just before the Cave Earth today (which displays excellent examples of peri- Void Rock moved from the Vestibule down toward the North glacial activity and frost-related features). and South Sally Ports. Some of the breccia in the Considering that frost heave does not require Figure 15. Reconstruction of the sedimen- Bear’s Den appears to have been frozen, cracked, the presence of permafrost, either continuous tary fi ll of Water Gallery looking along slightly faulted, and then cemented with calcite or discontinuous, and that it only required one Pengelly’s datum, to the south. Information (not dated, but stratigraphically in line with the signifi cant frost-heave event per cold interval taken from Pengelly’s fi eld notes (1868– dated MIS 9 sample, suggesting another breccia- to fracture the relatively thin calcite sheets, we 1880) for series 11, parallels 1–4. cracking event in MIS 10). argue that the evidence for frost shattering is

212 Geosphere, August 2007 MIS 9

MIS 11 MIS 11 MIS 11 MIS 9

MIS 9 MIS 12 MIS 12

MIS 12, 11 MIS 10 MIS 9

Late MIS 12: Breccia introduced to cave. MIS 10: Upper part of breccia + bones mobilized, MIS 9: Deposition of flowstone over earlier calcite and Early MIS 11: Bears move into cave; bears den on top creating vacuity; over bone-rich breccia. of breccia; bones embed into muddy breccia. Material moved from Bears Den into SW Gallery. MIS11: Flowstone cap protects some parts.

MIS 7 MIS 7 MIS 9 MIS 9 MIS 9

MIS 11 MIS 11 MIS 11 MIS 9

MIS 9 MIS 9 MIS 9

MIS 8 MIS 7 MIS 6

MIS 8: Frost heaving of breccia; cracking of flowstone; MIS 7: Deposition of flowstone. MIS 6a: Cracking of flowstone. mobilization of breccia. MIS 6b: Deposition of flowstone elsewhere in cave. MIS 6c: Remobilization of breccia, incorporation of rock and flowstone blocks.

1m

MIS 1 MIS 7 MIS 7 MIS 7

MIS 9 MIS 9 MIS 3 MIS 9 MIS 3

MIS 11 MIS 11 MIS 11

MIS 3 MIS 1 MIS 3 MIS 5 MIS 5 MIS 5 MIS 9 MIS 9 MIS 9

MIS 5 MIS 4,3,2 1m MIS 1

MIS 5: Induration of top layer of breccia; deposition of MIS 4: Cracking of flowstone. MIS 1: Rapid deposition of vuggy calcite, incorporation of flowstone. MIS 3: Cementation with thin flowstone, deposition of broken flowstone and rock slabs. cave earth. MIS 2: Cracking of flowstone + breccia, remobilization of cave earth. Pengelly’s categories Granular Stalagmite flowstone Void Cave Earth Breccia: very rich in bones and rock fragments Rock Crystalline Stalagmite flowstone Breccia: Incoherent, poor in bones Southwest Chamber: Presumed sequence of events. Cross section reconstructed from Pengelly’s field notes (1868-1880). Figure 16. Reconstruction of the sedimentary fi ll of Southwest Chamber, and the sequence of events required to explain the sediments.

Geosphere, August 2007 213 Lundberg and McFarlane overwhelming. The occurrence of at least seven green) occurs only in a very thin line on the the cave by natural processes through suffosion earthquakes of suffi cient magnitude to crack the western side of Clinnick’s Gallery. The com- dolines during MIS12. Subsequently, each inter- sheets, all coincidental with glacial periods, is bination of thick breccia and intact fl owstone glacial period produced calcite deposition in the extremely improbable. As further argument we (purple) occurs only under the bosses, where cave, and each glacial period caused periglacial cite Forti’s (2004) discussion of tectonic effects the fl owstone is too thick to fracture. Thus Fig- activity in the cave, during which the calcite was on speleothems: tectonic stresses can be recog- ure 17C offers support for the assertion that the fractured by frost heave, incorporated into the nized by the characteristic fracture, the most fracture is associated with the thickest breccia mud, and the mud moved by solifl uction. typical being breakage of stalagmites along sub- and is thus most likely to represent frost heave. The full sequence is as follows (see Fig. 19 horizontal planes, by the consistent breakage Freezing of the Breccia could easily have been and Table 3). in certain directions, and by breakages being achieved in a single cold season, and suffi cient 1. MIS 15–13: late Cromerian warm stage: grouped together in time. Forti (2004) was ada- aeration of the cave was likely, in view of the extensive surface weathering of hillside; deep mant that all other possible causes of breakage numerous open or semi-open routes for entry of regolith forms; vadose transport into cave of must be discounted, such as simple mechanical sediments and/or animals. fl uvial mud, sand and gravel, the Red Sands; failure from increase in loading, from sliding fabrication of Acheulian artifacts. of stalagmites and columns on unconsolidated Mechanism for Remobilization of Breccia 2. MIS 12: Anglian glaciation: formation materials, and from tongues of ice during glacia- and emplacement of breccia with entrained tion. All Forti’s tectonic examples are of broken Evidence for removal and remobilization of artifacts. stalagmites or columns; none are of fl owstone the Breccia is clear. Some removal may have 3. MIS 12 (terminal phase)–early MIS 11: sheets, and all examples come from regions been from fl uvial activity, but the only clear Bears hibernate in cave. of known and signifi cant tectonic activity. We evidence of this is in the paragenetic activity 4. MIS 11: Deposition of calcite layers; bears suggest that the subhorizontal shear operating on the roof of Rocky Chamber. The upper few hibernate in cave. on a fl owstone sheet would produce low-angle decimeters of breccia has remained mobile such 5. MIS 10: Cracking of fl owstone and of greenstick fractures: the cracks we see in the that artifacts and fractured pieces of calcite were breccia; removal and mobilization of breccia. cave are subvertical. moved from their source (e.g., Bear’s Den) to 6. MIS 9: Deposition of calcite. If frost heave is responsible for fl owstone new locations (e.g., Southwest Chamber) and yet 7. MIS 8: Cracking of calcite, remobilization fracture, then there should be an association are still emplaced within the Breccia matrix with of breccia. of the relative thickness of both the potentially no evidence for fl uvial deposition. We reject the 8. MIS 7: Deposition of calcite. heaving material and the potentially fracturing theory that water-rich debris fl ows effected this 9. MIS 6c: Cracking of calcite. material. So, a thick layer of water-saturated remobilization on the basis that little evidence 10. MIS 6b: Deposition of thin calcite fl ow- breccia will expand more than a thin layer, and a can be found for fl uvial orientation, sorting, or stone. thin layer of fl owstone will fracture more easily lamination. Long sections of the routes taken 11. MIS 6a: Cracking of fl owstone, remobili- than a thick layer. Figure 17A is a simple map by the Breccia (Fig. 18) show slopes of ~3°. zation of breccia. of the Breccia thickness. Both Proctor (1994) Debris fl ows generally require slopes of >20°, 12. MIS 5 e-a: Deposition of calcite. and Straw (1997) mapped the distribution of the but solifl uction fl ows can operate on only a few 13. MIS 4: Cracking of calcite; remobiliza- Breccia from Pengelly’s original reports. This is degrees. The slopes of the entrance regions are tion of breccia. the basis for Figure 17A, but we have divided steep enough to allow for either debris fl ow or 14. MIS 3: Deposition of calcite, emplace- the Breccia according to thickness, recon- solifl uction. However, the gentle slopes of the ment of Cave Earth. structed partly from sections in Proctor (1994) main passages would preclude debris fl ows as a 15. MIS 2: Cracking of calcite and breccia; and partly from fi eld observations. Figure 17B mechanism for breccia remobilization. We sug- mobilization of Cave Earth. shows the distribution of intact fl owstone and gest that the most likely mechanism for Breccia 16. MIS 1: Deposition of Granular Stalagmite. fractured fl owstone, a simplifi ed version of and Cave Earth movement is solifl uction. In addition to providing the most complete Straw’s (1997) map. The accuracy of this recon- Pleistocene record yet documented from any struction is partly limited by the detail and accu- CONCLUSIONS British cave, this report is also the fi rst publica- racy of Pengelly’s reports; we are aware that the tion of well-dated and clearly documented evi- level of detail deteriorated as his dig progressed, The sedimentary history of Kents Cavern is dence of frost heaving in interior cave passages. the earlier fi eld notes being considerably more considerably more complex than has previously The magnitude of the internal cave response to detailed than the later ones (1868–1880). For been recognized. The deposits record a rich, major global climatic shifts is of interest. This example, we are aware of some areas, such as cyclic history of events that track all the major history of repeated sedimentation events fol- the Labyrinth, that show evidence that Pengelly climatic cycles of the past 500 k.y. of British lowed by frost shattering and remobilization blasted through intact crystalline stalagmite, Pleistocene history—a situation not known for events is probably unique in the karst literature. yet this is not shown in Straw’s (1997) map. any other European cave or subaerial site. Most The uniqueness of the Kents Cavern sequence Figure 17C shows the distributions overlain: the of the deposits show some evidence of complex is likely an artifact of the relative lack of study two main areas of shattered fl owstone (yellow) reworking, and the fl owstone layer designated in that cave sedimentary sequences—rather than coincide largely with thick breccia (blue), pro- the literature as Crystalline Stalagmite is shown simply isolated speleothems—have received. ducing the pattern of green in the Bear’s Den to be a multilayered complex spanning MIS 11 We hope that the Kents cavern record will and the Hedges Boss–Inscribed Boss region. to MIS 3. In summary, the fi rst deposit is of serve as a demonstration of the potential of The intact fl owstone (red) in many areas is not fl uvial sands. The second deposit, of muddy these sequences, and focus future attention underlain by any breccia: otherwise it is under- breccia, incorporates the famous Acheulian on the conservation and study of other poten- lain by thin breccia (gray). The combination artifacts that were most likely fabricated during tially important cave sedimentary sequences in of thin breccia and fractured fl owstone (gray- MIS 13 (latest Cromerian) and transported into Britain and elsewhere.

214 Geosphere, August 2007 ABBRECCIA FLOWSTONE

Thin breccia Intact flowstone Fractured flowstone Thick breccia

BRECCIA + Cave of Rodentia C FLOWSTONE Vestibule Wolf’s Cave North Thin breccia and Thin breccia and Entrance intact flowstone fractured flowstone Charcoal Cave Thick breccia and Thick breccia and South fractured flowstone intact flowstone Entrance

Rocky Chamber North Gallery Sally Long arcade Port Southwest Clinnick’s Gallery Chamber South Sally Port High Level Chamber Inscribed Lake Boss

Water Gallery N Labyrinth Swallow Hole Gallery Bear’s Den In-Between Boss Bear’s Den Boss

Hedges Boss Terminal Chamber 30 m

Figure 17. (A) Simplifi ed survey (after Proctor and Smart, 1989) showing the distribution of major breccia deposits in the cave. The location of the Breccia follows Proctor (1994) and Straw (1997), both of whom based their mapping on fi eld notes of Pengelly. The division into Thick (breccia exposed is >~1.5 m thick) and Thin (breccia exposed is <~1.5 m thick) is based on sketched sections in Proctor (1994) and our own observations of the remnant breccia in the cave. (B) The distribution of the main fl owstone unit, the Crystalline Stalagmite, following Straw’s (1997) reconstruction from Pengelly’s notes. The division into Intact, representing fl owstone that has not been broken, and Fractured, representing fl owstone that is largely in situ but crazed or cracked with minor displacement, also follows Straw (1997). (C) The distributions from A and B superimposed so that four categories emerge from all possible combinations of breccia and fl owstone.

Geosphere, August 2007 215 Kents Cavern: Cave of Rodentia Wolf’s Routes for Cave Breccia Route B North Entrance Breccia flow Charcoal Cave South Entrance Breccia Route A Rocky Chamber

Gallery Long arcade Southwest Clinnick’s Gallery Chamber

High Level Chamber Inscribed Lake Boss N Water Gallery Labyrinth Swallow Hole Gallery Bear’s Den Breccia Route C

Hedges Boss Bear’s Den 30 m Inlet Rift

Breccia Route A: Extended Long Section from Swallow Hole Gallery to Rocky Chamber

Entrance route of breccia from surface

High Level Chamber

Rocky Chamber 9 Hedges Boss Clinnicks Gallery Swallow Hole 10 11 Gallery 7 6543 60 m O.D.

Slope = 3.2° ? To small phreatic maze passages

Breccia Route B: Extended Long Section from Swallow Hole Gallery to Long Arcade and Wolf’s Cave

Entrance route of breccia from surfact

High Level Chamber

9 Hedges Boss Swallow Hole 10 Long Arcade 11 Gallery 12 13 14 15 Wolf’s Cave 17 60 m O.D. Cave of Rodentia

? Slope = 2.9° To small phreatic maze passages

Breccia Route C: Extended Long Section from Bear’s Den to SouthWest Chamber to Lecture Hall to South Entrance Doline/Rift ?

Bear’s Den Inlet Rift 30 Lake, Water Gallery Bear’s Den Great Chamber SouthwesT Chamber Lecture Hall Maze going 29 33 36 31 35 35 South Entrance down to 34 38 Hedges Boss 60 m O.D. Cave of Rodentia 19 18 17

The Gallery Slope = 2.9° Charcoal Cave ?

Slope= 2.8°

Figure 18. Probable routes for the main breccia fl ows in the cave, shown in plan view and in extended long sections. The long sections are reconstructed from the numbered cross sections in the survey of Proctor and Smart (1989). At least two entrances appear to have admitted breccia, the one in Swallow Hole Gallery and the other in the Bear’s Den area, most likely the inlet rift shown in section 30 of the survey. O.D. refers to ordinance datum.

216 Geosphere, August 2007

Prod u cti on of Ach eu lia n artifa ct s .

Sa nds .

F orm a tio na n dde positi on of R ed

W ide sp re a d w ea th e rin g , de ep re g o lith . Stage 13-17 Cromerian complex from from in pink. ue for the ue for error bars. error

σ For m ati nand on deposition of B re c cia

Anglian

Stage 12

Oc up c to yc by ation av e bear s

Dep osition fcalcite of flowstone

Hoxnian Stage 11

P robabl e cra cking of cal cite flo w stone

R em obil iza ion t or er osion of B recci a

Deposit ion of cal cit e flo w st one

Crack ng i fca of lcite flo wsto ne

10 11 12 13 Stage 10 Remo b iliza ion t o Breccia f

9

Fluvial erosion? Paragenes is?

Deposition of calcit e flowstone

Stage 9 Purfleet

rcigo act flowstone calcite of Cracking

R em obilization of Brecci a

Stage 8

Deposition of calcite flow sto ne Aveley Stage 7

than thickness of calcite growth. ects sampling effort rather

rcigo act flowstone calcite of Cracking

Rem o bilization of Brec ci a

Deposi inof tion act flowstone calcite

C ra ckin go f cal cit e flo ws to ne

Rem o biliz ation fBrecci of a

Stage 6

D e p o sitio n o f c a lcite flo wsto n e Stage 5

Ipswichian

σ

C ra ckin g o cal f cit e flo ws to ne

4

Rem o bilization of Brec cia

Figure 19. The Pleistocene depositional history of Kents Cavern summarized on the time line. The marine oxygen isotope curve is The Pleistocene depositional history of Kents Cavern summarized on the time line. 19. Figure shown with 2 The radiometric dates are triangles). as red the stage markers (shown here Bassinot et al. (1994), as are Periods of clastic sedimentation and remobilization of the Breccia are shown in yellow. Periods of calcite deposition are shown Periods of calcite deposition are shown in yellow. are of the Breccia Periods of clastic sedimentation and remobilization Periods of fracturing are shown in gray. The British names for stages are shown, in red for the interglacial periods, and in bl for shown, in red stages are The British names for shown in gray. Periods of fracturing are glacial periods. The distribution of dates refl glacial periods. Depo sition of calcit e flowst one σ

35

Depo sition of C av e Eart h

Cra ckin g of calcite and Bre ccia

et al.

U-Th date, isochron date, 2 range U-Th date, 2 range Isotope stage boundaries

Rem o bilization of Cave Eart

h 2678

Devensian Stage 2

eoiino Granular of Deposition talagmit S e 1 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 Stage 1 BassinotBritish names for stages from www.nhm.ac.uk/hosted_sites/ahob/overview_time_chart.gif (1994) Normalized oxygen isotope values Age (ka) Flandrian 0 1 2 3 -2 -1

Geosphere, August 2007 217 TABLE 3. SUMMARY OF EVIDENCE FOR EVENTS Age Event Site Evidence Cromerian Red Sands deposition Gallery >580 ka Al-Be MIS 12 Breccia deposition Labyrinth >400 ka Al-Be MIS 12 Breccia deposition Hedges Under MIS 11 calcite MIS 11 Calcite deposition Hedges Calcite on top of breccia MIS 11 Calcite deposition Bear’s Den Calcite on top of breccia MIS 11 Calcite deposition Southwest On upper level, former breccia surface MIS 10c Breccia movement Southwest Breccia surface lowered between MIS 11 calcite and MIS 10b calcite MIS 10b Calcite deposition Southwest On top of breccia MIS 11-9 ? Fracture + movement Water Gallery MIS 11? slab broken and covered by breccia, covered by MIS 9 calcite MIS 11-9 Breccia fracture + remobilization Bear’s Den Breccia remobilized to cover MIS 11 calcite, but under MIS 9 calcite; Breccia frozen, cracked and shifted; older calcite (MIS 11?) cracked and moved MIS 11-9 Breccia remobilization Water Gallery Breccia had been higher on wall, removed before MIS 9 calcite deposition MIS 9 Calcite deposition Bear’s Den Calcite on top of reworked breccia MIS 9 Calcite deposition Water Gallery Calcite on top of thin red muddy film on wall; remnant of breccia MIS 9 Calcite deposition In-Between Rapid, thick flowstone deposition above breccia MIS 9 Calcite deposition Labyrinth Rapid, thick flowstone deposition above breccia MIS 9 Calcite deposition Labyrinth Broken slab within breccia MIS 9 Calcite deposition Southwest Calcite flowstone on top of MIS 11 calcite and on top of breccia Post–MIS 9 Fracture + movement Labyrinth MIS 9 calcite slab in breccia broken and moved MIS 8 Fracture In-Between Thin edge of MIS 9 flowstone boss cracked, filled by MIS 7 calcite Pre–MIS 7 Breccia remobilization Clinnick’s Breccia had been higher on wall, removed before MIS 7 calcite deposition MIS 7 Calcite deposition In-Between Calcite on MIS 9 calcite, cementing crack MIS 7 Calcite deposition Labyrinth Slab in breccia MIS 7 Calcite deposition Southwest On top of MIS 9 calcite Post–MIS 7 Fracture + movement Labyrinth MIS 7 calcite slab in breccia broken and moved MIS 6c Fracture In-Between MIS 9 calcite cracked, cemented by MIS 6b calcite but no MIS 7 calcite MIS 6c Fracture + movement Southwest MIS 7 calcite fractured and removed to make way for MIS 6b calcite MIS 6b Calcite deposition In-Between Calcite on top of MIS 9 calcite, cementing crack MIS 6b Calcite deposition Southwest Slab within breccia MIS 6a Fracture + movement Southwest MIS 6b calcite broken and moved, cemented by MIS 5; induration of breccia? MIS 5 Calcite deposition In-Between Sequence from 5c to 5a MIS 5e Calcite deposition High Level Calcite on top of breccia MIS 5e Calcite deposition Southwest Calcite flowstone MIS 5e–3 Fracture + movement High Level MIS 5e calcite fractured, removed, covered with thin red calcite coating MIS 5a Calcite deposition Clinnick’s Top of ~30 cm laminated flowstone, probably MIS 5e–a MIS 4 Fracture Clinnick’s Fracture of MIS 5a calcite, cemented by MIS 3 calcite MIS 4 Fracture + movement High Level MIS 5e calcite cracked, removed, cemented by MIS 5a calcite MIS 3 Calcite deposition High Level Calcite on top of fracture surface of MIS 5e calcite MIS 3 Calcite deposition Clinnick’s Red calcite cementing broken slabs of MIS 5a calcite MIS 3–1 Fracture Clinnick’s Fracture of MIS 3 red calcite MIS 3–2 Cave earth deposition Radiocarbon dates MIS 2 Fracture Clinnick’s MIS 3 calcite cracked and cemented by MIS 1 calcite MIS 2 Fracture of breccia and calcite + Breccia and calcite fragments incorporated in cave earth; movement of cave earth Pengelly’s notes (1868–1880) + Straw’s (1997) map MIS 1 Calcite deposition Clinnick’s Granular stalagmite cementing broken slabs of MIS 3 calcite MIS 1 Calcite deposition In-Between Active stalagmite and curtains at top of boss MIS 1 Calcite deposition Southwest Granular Stalagmite, radiocarbon dates

218 Geosphere, August 2007 Kents Cavern Pleistocene History

ACKNOWLEDGMENTS Cullingford, R.A., 1982, The Quaternary, in Durrance, E.M., northern Europe: Nature, v. 438, p. 1008–1012, doi: and Laming, D.J.C., eds., The geology of Devon: 10.1038/nature04227. We are indebted to Nick Powe, owner and managing Exeter, University of Exeter Press, p. 249–290. Pengelly, W., 1868–1880, Typed transcripts of original director of Kents Cavern, for his encouragement of this Dowling, L.A., and Coxon, P., 2001, Current understanding handwritten fi eld notes, held in Torquay Museum, project. We also thank Barry Chandler of the Torquay of Pleistocene temperate stages in Ireland: Quaternary Devon, UK. Museum for access to specimens and manuscripts in Science Reviews, v. 20, p. 1631–1642, doi: 10.1016/ Pengelly, W., 1869, The literature of Kents Cavern: Part II: his care, and two anonymous referees for helpful com- S0277-3791(01)00028-2. Devonshire Association Transactions, v. 3, p. 191–482. ments on an earlier version of this manuscript. We are Durrance, E.M., and Laming, D.J.C., eds., 1982, The geology Pengelly, W., 1884, The literature of Kents Cavern: Part V: of Devon: Exeter, University of Exeter Press, 346 p. Devon shire Association Transactions, v. 14, p. 189–343. grateful to Darryl E. Granger and the staff of PRIME Ford, T.D., 1997, Stalagmite sheets shattered by earthquakes: Pevzner, M., Vangengeim, E., and Tesakov, A., 2001, Qua- Lab, Purdue University, for cosmogenic dating sup- Cave and Karst Science, v. 24, p. 140–141. ternary zonal subdivisions of Eastern Europe based on port and advice. Permission to collect samples from Forti, P., 2004, Paleotectonics from speleothems, in Gunn, vole evolution: Bulletino dell Società Paleontologica this scheduled Ancient Monument was granted by J., ed., Encyclopedia of cave and karst science: Lon- Italiana, v. 40, p. 269–274. English Heritage Secretary of State for Culture, Media don, Fitzroy Dearborn, p. 565–566. Preece, R.C., 2001, Molluscan evidence for differentiation and Sport under permit # HSD 9/2/7663. Partial fund- Granger, D.E., and Muzikar, P., 2001, Dating sediment burial of interglacials within the “Cromerian Complex”: Qua- ing was provided by the Andrew Mellon Foundation with cosmogenic nuclides: Theory, techniques, and ternary Science Reviews, v. 20, p. 1643–1656, doi: under a sabbatical improvement grant through Scripps limitations: Earth and Planetary Science Letters, v. 188, 10.1016/S0277-3791(01)00032-4. p. 269–281, doi: 10.1016/S0012-821X(01)00309-0. Proctor, C.J., 1994, A British Pleistocene chronology based College (to McFarlane), and through a research sup- Granger, D.E., Fabel, D., and Palmer, A.N., 2001, Pliocene– on U-series and ESR dating of speleothem [Ph.D. port grant from Pitzer College (to McFarlane), and Pleistocene incision of the Green River, Kentucky, deter- thesis]: Bristol, University of Bristol, 353 p. Natural Sciences and Engineering Research Council mined from radioactive decay of cosmogenic 26Al and Proctor, C.J., and Smart, P.L., 1989, A new survey of Kents grant (to Lundberg). This is Ottawa-Carleton Geo- 10Be in Mammoth Cave sediments: Geological Society of Cavern, Devon: University of Bristol Spelaeological science Centre, Isotope Geochemistry and Geo- America Bulletin, v. 113, p. 825–836, doi: 10.1130/0016- Society Proceedings, v. 18, p. 422–429. chronology Research Facility contribution 44. 7606(2001)113<0825:PPIOTG>2.0.CO;2. Proctor, C.J., Berridge, P.J., Bishop, M.J., Richards, D.A., Grün, R., and Schwarcz, H.P., 2000, Revised open system and Smart, P.L., 2005, Age of middle Pleistocene fauna REFERENCES CITED U-series/ESR age calculations for teeth from Stratum and lower Palaeolithic industries from Kents Cavern, C at the Hoxnian Interglacial type locality, England: Devon: Quaternary Science Reviews, v. 24, p. 1243– Andrews, P., 1990, Owls, caves and fossils: London, Natural Quaternary Science Reviews, v. 19, p. 1151–1154, doi: 1252, doi: 10.1016/j.quascirev.2004.07.022. History Museum, 231 p. 10.1016/S0277-3791(00)00085-8. Rowe, P.J., Atkinson, T.C., and Turner, C., 1999, U-series Anthony, D., and Granger, D.E., 2004, A late Tertiary ori- Ivanovich, M., and Harmon, R.S., eds., 1992, Uranium- dating of Hoxnian interglacial deposits at Marks gin for multilevel caves along the western escarpment series disequilibrium: Applications to Earth, marine Tey, Essex, England: Journal of Quaternary Sci- of the Cumberland Plateau, Tennessee and Kentucky, and environmental sciences (second edition): Oxford, ence, v. 14, p. 693–702, doi: 10.1002/(SICI)1099- established by cosmogenic 26Al and 10Be: Journal of Clarendon Press, 910 p. 1417(199912)14:7<693::AID-JQS477>3.0.CO;2-X. Cave and Karst Studies, v. 66, p. 46–55. Ivanovich, M., Latham, A.G., and Ku, T.-I., 1992, Uranium- Schwarcz, H.P., and Blackwell, B.A., 1992, Archaeological Ballantyne, C.K., and Harris, C., 1993, The periglaciation series disequilibrium applications in geochronology, in applications, in Ivanovich, M., and Harmon, R.S., eds., of Great Britain: Cambridge, Cambridge University Ivanovich, M., and Harmon, R.S., eds., Uranium-series Uranium-series disequilibrium: Applications to Earth, Press, 330 p. disequilibrium: Applications to Earth, marine and envi- marine and environmental sciences (second edition): Bassinot, F.C., Labeyrie, L.D., Vincent, E., Quidelleur, X., ronmental sciences (second edition): Oxford, Clarendon Oxford, Clarendon Press, p. 513–552. Shackleton, N.J., and Lancelot, Y., 1994, The astro- Press, p. 62–89. Schreve, D.C., 2001, Differentiation of the British late nomical theory of climate and the age of the Brunhes- Kaufman, A., and Broecker, W., 1965, Comparison of middle Pleistocene interglacials: The evidence from Matuyama magnetic reversal: Earth and Planetary 230Th and 14C ages for carbonate materials from Lakes mammalian biostratigraphy: Quaternary Science Science Letters, v. 126, p. 91–108, doi: 10.1016/0012- Lahontan and Bonneville: Journal of Geophysical Reviews, v. 20, p. 1693–1705, doi: 10.1016/S0277- 821X(94)90244-5. Research, v. 70, p. 4039–4054. 3791(01)00033-6. Bates, M.R., Keen, D.H., and Lautridou, J.-P., 2003, Pleis- Keith, A., Myres, J.L., Burkitt, M.C., Favell, R.V., Garfi tt, Schreve, D.C., and Thomas, G.N., 2001, Critical issues in tocene marine and periglacial deposits of the English G.A., Garrod, D.A.E., and Sollas, W.J., 1931, Kents European Quaternary biostratigraphy: Quaternary Channel: Journal of Quaternary Science, v. 18, p. 319– Cavern, Torquay: British Association for the Advance- Science Reviews, v. 20, p. 1577–1582, doi: 10.1016/ 337, doi: 10.1002/jqs.747. ment of Science, Report of the 98th Annual Meeting, S0277-3791(01)00024-5. Bertran, P., Hétu, B., Texier, J.-P., and Van Steijn, H., 1997, p. 266–267. Schwarcz, H.P., and Latham, A.G., 1989, Dirty calcites 1: Fabric characteristics of subaerial slope deposits: Kurtén, B., 1968, Pleistocene mammals of Europe: London, Uranium series dating of contaminated calcite using Sedimentology, v. 44, p. 1–16, doi: 10.1111/j.1365- Weidenfeld and Nicolson, 317 p. leachates alone: Chemical Geology, v. 80, p. 35–43. 3091.1997.tb00421.x. Lister, A.M., McGlade, J.M., and Stuart, A.J., 1990, The Scourse, J.D., Austin, W.E.N., Sejrup, H.-P., and Ansari, Bishop, M.J., 1982, The mammal fauna of the early middle early middle Pleistocene vertebrate fauna from Little M.H., 1999, Foraminiferal isoleucine epimerization Pleistocene cavern infi ll site of Westbury-sub-Mendip, Oakley, Essex: Royal Society of London Philosophi- determinations from the Nar Valley Clay, Norfolk, UK: Somerset: Palaeontological Association Special Paper, cal. Transactions, ser. B, v. 238, p. 359–385. Implications for Quaternary correlations in the south- v. 28, p. 1–108. MacEnery, J., 1859, Cavern researches, or Discoveries of ern North Sea basin: Geological Magazine, v. 136, Campbell, J.B., and Sampson, C.G., 1971, A new analysis organic remains, and of British and Roman reliques, in p. 543–560, doi: 10.1017/S0016756899002812. of Kents Cavern, Devonshire, England: University of the caves of Kents Hole, Anstis Cove, Chudleigh, and Scrivener, M.F., 1987, An introduction to the geology of the Oregon Anthropological Papers, v. 3, p. 1–39. Berry Head: London, Simpkin, Marshall, and Co., 78 p. Torquay District: Torquay, Devon, Torquay Natural Campbell, S., Hunt, C.O., Scourse, J.D., and Stephens, N., McCarroll, D., 2002, Amino-acid geochronology and the History Society, 18 p. 1998a, Quaternary of south-west England: London, British Pleistocene: Secure stratigraphical framework Stock, G.M., Granger, D.E., Sasowsky, I.D., Anderson, R.S., Chapman and Hall, 439 p. or a case of circular reasoning?: Journal of Quaternary and Finkel, R.C., 2005, Comparison of U-Th, paleo- Campbell, S., Keen, D.H., Currant, A.P., Collcutt, S.N., Science, v. 17, p. 647–651, doi: 10.1002/jqs.690. magnetism, and cosmogenic burial methods for dating and Stuart, A.J., 1998b, Pleistocene cave sequences, McFarlane, D.A., and Lundberg, J., 2005, The 19th century caves: Implications for landscape evolution studies: in Campbell, S., et al., eds., Quaternary of south-west excavation of Kents Cavern, England: Journal of Cave Earth and Planetary Science Letters, v. 236, p. 388– England: Geological Conservation Review Series, and Karst Studies, v. 67, p. 39–47. 403, doi: 10.1016/j.epsl.2005.04.024. v. 14, p. 129–153. McFarlane, D.A., Lundberg, J., and Sabol, M., 2006, Middle Straw, A., 1997, Kents Cavern—Whence and whither?: Cave Cheng, H., Edwards, R.L., Hoff, J., Gallup, C.D., Richards, Pleistocene cave bears from Kents Cavern, England: and Karst Science, v. 24, p. 35–40. D.A., and Asmerom, Y., 2000, The half-lives of uranium- Geological Society of America Abstracts with Pro- Sutcliffe, A.J., 1960, Joint Mitnor Cave, Buckfastleigh: 234 and thorium-230: Chemical Geology, v. 169, p. 17– grams, v. 38, no. 7, p. 214. Torquay Natural History Society Transactions, v. 13, 33, doi: 10.1016/S0009-2541(99)00157-6. Millar, S.W.S., 2006, Processes dominating macro-fabric p. 3–28. Collcutt, S.N., 1986, Contextual archaeology: The example of generation in periglacial colluvium: Catena, v. 67, Sutcliffe, A.J., and Kowalski, K., 1976, Pleistocene rodents debris fl ow in caves, in Collcutt, S.N., ed., The Palaeo- p. 79–87, doi: 10.1016/j.catena.2006.03.003. of the British Isles: British Museum (Natural History) lithic of Britain and its nearest neighbours: Recent Murton, J.B., and Lautridou, J.-P., 2003, Recent advances in Bulletin, Geology, v. 27, p. 1–147. trends: Sheffi eld, University of Sheffi eld, p. 57–58. the understanding of Quaternary periglacial features of Thomas, G.N., 2001, Late middle Pleistocene pollen bio- Cook, J., and Jacobi, R., 1998, Observations on the artefacts the English Channel coastlands: Journal of Quaternary stratigraphy in Britain: Pitfalls and possibilities in from the Breccia at Kents Cavern, in Ashton N., et al., Science, v. 18, p. 301–307, doi: 10.1002/jqs.748. the separation of interglacial sequences: Quaternary eds., Stone age archaeology: Essays in honour of John Oxford Radiocarbon Accelerator Unit, 2006, Results data- Science Reviews, v. 20, p. 1621–1630, doi: 10.1016/ Wymer: Lithic Studies Society Occasional Paper 6, base: http://c14.arch.ox.ac.uk/results.html. S0277-3791(01)00026-9. p. 77–89. Parfi tt, S., Barendregt, R.W., Breda, M., Candy, I, Collins, J., Croot, D., and Griffi ths, J.S., 2001, Engineering geological Coope, G., Durbridge, G., Field, M., Lee, J., Lister, A., signifi cance of relict periglacial activity in South and Mutch, R., Penkman, A., Preece, R., Rose, J., Stringer, MANUSCRIPT RECEIVED 5 JANUARY 2007 East Devon: Quarterly Journal of Engineering Geology C, Symmons, R., Whittaker, J., Wymer, J., and Stuart, REVISED MANUSCRIPT RECEIVED 13 APRIL 2007 and Hydrology, v. 34, p. 269–281. A., 2005, The earliest record of human activity in MANUSCRIPT ACCEPTED 25 APRIL 2007

Geosphere, August 2007 219