Carsington Pasture Cave, Brassington, Derbyshire.
A Prehistoric Burial Site.
Dr. Andrew Chamberlain, Department of Archaeology and Prehistory, University of Sheffield.
INTRODUCTION
Carsington Pasture Cave is located at national grid reference SK 2415 5368, about 1 km southeast of Brassington (Figure 1). The cave entrance is situated at an altitude of 300m on a promontory ridge that curves south and then westwards from the main limestone escarpment, giving a distant viewshed in all directions. The low "natural" or walk-in entrance to the cave is hidden in a north-facing hollow, and a second artificial entrance 6 metres to the south (possibly an access shaft created by lead miners) provides the current vertical entry to the cave. Immediately inside the walk-in entrance the cave widens into a 5m diameter chamber (hereafter the "Entrance Chamber") which has flowstone decorated walls and a clay floor. An unsuccessful attempt to dig through the floor of the Entrance Chamber was made by members of the Peak District Mines Historical Society in 1965 (Tune, 1967), but no other speleological investigations of the cave have been recorded.
This report provides a preliminary account of recent archaeological discoveries in Carsington Pasture Cave made by members of the Pegasus Caving Club, assisted by archaeologists from the University of Sheffield. The report describes human and faunal remains recovered from the surface of the cave deposits and bones and artefacts revealed during operations to gain access to the deeper parts of the cave system. Exploration of the cave has led to partial disturbance of the deposits, but wherever possible the sediments have been left in situ in recognition of their importance for future archaeological and palaeontological research. At an early stage of the investigation it was recognised that the cave could not be secured easily against unauthorised entry, and it was decided to remove most of the human remains that were visible on the surface of the floor of the second chamber. This recovery exercise was undertaken after obtaining a Licence for the Removal of Human Remains from the Home Office and with the agreement of the county archaeologist and the land owner. Exploration of the cave is continuing, and the present report documents discoveries made up to July 1999.
Geology and Mining History
Carsington Pasture Cave is formed in dolomitised Bee Low Limestone of the Carboniferous limestone series, near the southern margin of the Derbyshire lead-zinc orefield. The limestone dips gently to the ENE, and the presence of reef facies and dolomitisation renders the limestone resistant to weathering, producing steep-sided hills with localised vertical exposures of bedrock. The limestone is extensively mineralised, mainly with lead and zinc ores together with gangue minerals. A dense network of mineral veins, principally trending WNW-ESE, permeates the limestone underneath Carsington Pasture and the local area has been mined intensively for lead probably from as early as the Roman period (Willies, 1995). A large number of mine shafts are located within 100 metres of the cave: Great Rake Mine, a substantial lead mine originally developed in the 17th century, is situated about 250 metres WSW of the cave (Slack, 1987).
Immediately to the west of the cave is Wester Hollow, a dry valley floored with Tertiary sands, clays and gravels of the Brassington Formation. These deposits are believed to have formed by the infilling of large solution hollows in the surface of the Carboniferous limestone, and they are sometimes associated with the formation of cave systems, as for example at Golconda Caverns (Ford, 1977). The presence of large numbers of quartzite pebbles in Carsington Pasture Cave provides evidence that the Wester Hollow pocket deposit was formerly more extensive and outcropped at a higher level than it does today.
Regional Archaeological Context
Carsington Pasture Cave is located in an area that is rich in archaeological finds of the Romano-British period. Nearby settlements of this date have been excavated at Rainster Rocks, 1 km WNW of Brassington, and at two sites alongside the former Scow Brook in an area which is now inundated by the creation of Carsington Reservoir (Ling and Courtney, 1981; Dearne et al., 1995). The line of the Roman Road from Buxton to Derby is conjectured to run about 300 m to the east of the cave (Willies, 1995) and sherds of Romano-British pottery have been found near the limestone outcrop 200 m north of the cave (Derbyshire SMR Number 3208). Roman lead pigs have also been found near Carsington, and the area is believed to have been an important centre of lead production in Roman times (Branigan et al., 1986).
There are two known archaeological caves within 3 km of the site (Gill and Beck, 1991), these being Rains Cave (SK 226 553) and Harborough Cave (SK 242 552). Rains Cave was excavated by J. Ward in the 19th century and produced Neolithic and Bronze Age pottery and the skeletal remains of at least six human individuals. Harborough Cave was excavated successively by W. Storrs Fox, A.L. Armstrong and by D. Bramwell, and has produced Neolithic, Bronze Age, Iron Age and Romano-British artefacts and the skeletal remains of two human individuals.
Figure 2: Plan of the Entrance Chamber and 2nd Chamber in Carsington Pasture Cave. Locations of human bones are depicted as open circles for remains of neonates and as closed circles for remains of children and adults. Numbers indicate micromammal sampling locations.
Recent Excavations
In 1998 members of the Pegasus Caving Club created a passage through the clay and rubble fill of the Entrance Chamber of Carsington Pasture Cave (Steans & Scothon, 1998). The dug passage follows the sloping roof of the cave steeply downwards in a southerly direction for a distance of about 10m, terminating in the newly-discovered 2nd Chamber (Figure 2). The 2nd Chamber, which is also referred to as ‘Yorick Chamber’, is about 10m long E-W and 4m wide N-S with a low horizontal roof of limestone bedrock decorated with stalactite straws and pendants. The unconsolidated rubble floor of this chamber is formed of large limestone blocks and quartzite and gritstone boulders, partially encrusted in some areas by speleothem. Interspersed among the boulders were numerous human and faunal remains, some of which were also encrusted with speleothem. Boulders attached by speleothem to the cave wall indicate a former elevated position of the floor of the cave: the unconsolidated cave fill has slumped in places by as much as 1 metre below this former level. However, the presence of extremely fragile neonatal cranial bones indicated that at least some of the human skeletal remains had not moved very far from their original site of deposition.
In May 1998 all faunal and human skeletal remains visible on the surface of the 2nd Chamber were photographed in situ, recorded in three dimensions and lifted in order to prevent damage occurring to the archaeological material during further exploration of the cave system. One human cranium which was embedded in a thick stalagmite deposit at the side of the chamber was left in situ. Discrete deposits of small mammal bones were noted at various locations on the surface of the deposits near the middle of the 2nd Chamber (Figure 2 above, locations 103, 105 to 109, 112). These deposits were bulk sampled by James Williams for species identification and taphonomic analysis.
Monitoring of finds made during further exploration of the cave has been limited to the identification of bones revealed by the cavers’ excavation and the recording of their approximate location within the cave system. Three artefacts and numerous animal and human bones were exposed during the excavation of a 4m deep shaft to gain access to the 3rd Chamber (Figure 3 below).
This chamber is of a similar size to the 2nd Chamber and is roofed by a 1 metre thick flowstone deposit. The flowstone ceiling of the 3rd Chamber formerly rested directly on top of an unconsolidated deposit of silts, gravels and boulders: partial collapse of the unconsolidated deposits has created the 3rd Chamber as a 1 m high void space below the hanging flowstone ceiling. The 3rd Chamber showed evidence of prior entry by lead miners in the form of a dug trench through silt deposits, hob-nailed boot prints, pick marks on the cave walls and the bowl of a clay pipe. It has not been established how the miners gained access to this chamber, as there was no evidence for lead mining activity within the 2nd Chamber. It is possible that a mine shaft or adit breached the 3rd Chamber but if this was the case the point of entry must now be concealed by debris.
In 1999 a new excavation in the floor of the Entrance Chamber led to the discovery of further chambers of the cave system extending in a northerly direction. Both faunal remains and a small number of human remains are present in this part of the cave system, as well as prehistoric and Roman pottery and a few Roman coins.
Human Remains
The locations of human remains recovered from the 2nd Chamber are indicated on Figure 2 above. A minimum number of 20 individuals is represented in the assemblage and the demography of the sample suggests that all ages and adults of both sexes are included (Table 1, see below). The frequency of recovery of human remains declines as one moves vertically downwards through the cave system: of the 115 skeletal elements recovered, 66 (58%) were located on the floor of the second chamber and 38 (34%) were recovered from the excavation of the shaft between the second and third chambers. Only 6 human skeletal elements have been retrieved from the deep shaft excavated below the floor of the third chamber. This spatial pattern confirms initial impressions that the primary context for the deposition of the human remains was located in or adjacent to the second chamber, with the distribution in the lower parts of the cave system resulting from carnivore disturbance and/or partial collapse of the cave sediments. Within the second chamber there is a non-random spatial distribution of skeletal remains, with all of the neonatal remains being clustered near the centre of the chamber (Figure 2 above).
Table 1: Age and Sex Structure of the Human Remains from Carsington Pasture Cave.
The skeletal parts recovered are dominated by cranial remains together with some of the larger parts of the postcranial skeleton. Smaller bones are almost entirely absent: no hand bones, only a single foot bone and very few vertebrae have been recovered. This pattern is not attributable to degradation of bone tissue, as the state of preservation of all of the human and faunal remains is excellent. A similar pattern of loss of small sized skeletal elements is seen among the faunal remains, and it is possible that the smaller skeletal elements have either fallen out of sight between the larger boulders of the cave infill or have simply not been detected during the exploration of the cave. An alternative explanation for the pattern of skeletal part representation is that bodies were defleshed and disarticulated outside the cave, or in as-yet unexplored parts of the cave system, and only parts of the skeleton were then brought into the 2nd chamber. However, some of the individuals originally must have been deposited within the cave as intact or at least partially articulated skeletons: in the excavated area below the second chamber the hip bones, sacra and lower lumbar vertebrae of two individuals were discovered in close proximity. Similarly some of the neonatal remains in the second chamber include partial skeletons with cranial and postcranial remains from the same individual.
Movement of fragments of the same individual within the cave system is demonstrated by the spatial distribution of parts of a single skull, which were separated by a distance of 3.5 m within the second chamber. In all cases the mandibles were separated from the crania, but the damage to the facial bones of the crania has prevented successful cross matching of mandibles to the upper jaws of the same individuals. A number of the human remains exhibit damage caused by medium-sized carnivores (probably foxes, dogs or wolves). Some long bones show characteristic gnawing damage to the delicate epiphyseal areas, and some skulls and mandibles have been chewed in the region of the mastoid processes and the mandibular condyles.
All remains have been inspected for cutmarks, but these features are confined to two femurs from the same adult individual. The left side femur shows multiple transverse cutmarks on all four aspects of the distal metaphysis, beginning just above the articular surfaces about 40 mm above the level of the knee joint (Figure 4). The cutmarks are up to 20 mm long and have well defined v-shaped cross sections with some parallel striations within the marks. Several of the cutmarks follow slightly divergent courses indicating the use of multiple strokes. The right side femur has a few cutmarks of similar morphology but confined to the medial and antero-medial aspects of the distal metaphysis between 10 and 20 mm above the articular surface. The location of the cutmarks is consistent with disarticulation of the body at the knee joint. It is unlikely that the aim was to remove muscle from the thigh as no cutmarks are visible at the muscle attachment regions on the proximal and midshaft areas of the femur. Cutmark evidence for defleshing and disarticulation is present at other prehistoric mortuary sites in Britain, especially during the Neolithic when human bones, which may have been treated as relics, were transported between various locations in the landscape (Edmonds, 1999).
Figure 4: Cutmarks on posterior surface of adult human femur.
Faunal Remains
The macrofauna from the cave have been studied by Josephine Paxton (1999) and a summary faunal list of the mammalian species is presented in Table 2 below. The macrofauna are dominated by domestic species, principally cattle (Bos taurus) and pig (Sus scrofa), with dog (Canis familiaris), horse (Equus caballus), sheep (Ovis aries) and goat (Capra hircus) occurring less frequently. There are a small number of bones of wild species, including red deer (Cervus elaphus), roe deer (Capreolus capreolus), wild cattle (Bos primigenius), badger (Meles meles) and red fox (Vulpes vulpes): the large dog present in the assmblage may in fact represent wolf (Canis lupus). Two skeletal elements of brown bear (Ursus arctos) have been recovered from different parts of the cave.
Table 2: Vertebrate Fauna Recovered from Carsington Pasture Cave.
The faunal remains show few signs of butchery or other processes of food preparation or secondary product utilisation. There is copious evidence of gnawing by medium sized carnivores, with splintering of long bone shafts and puncture marks on the epiphyses and other regions of fragile bone. There are also two instances of an unusual pattern of incisor "chiselling" caused by a large rodent. Among the pig remains there are several long bone elements that appear to be left and right antimeres from the same individual, suggesting that whole carcasses rather than individual skeletal elements were entering the cave system. This hypothesis is supported by the low frequency of butchery on the bones of this species: just 2 out of 139 pig bones exhibit cut marks. We infer that the pigs may have fallen into the cave through vertical fissures, or have been brought in by carnivores as partial or complete carcasses.
The microfaunal remains sampled from surface locations in the second chamber have been the subject of a preliminary study by James Williams. The microfaunal concentrations are dominated by the remains of water vole (Arvicola terrestris) and field vole (Microtus agrestis), and are likely to have accumulated from the pellets regurgitated by owls that at some time utilised the chamber for roosting or nesting.
Artefacts
A decorative bone pin, a fragment of worked antler and a flint flake were discovered during excavation of the floor of the second chamber in order to gain access to deeper sections of the cave system. The complete perforated bone pin (Figure 5 ), made on a split roe deer metatarsal, is 148 mm long and is 15 mm wide at the flattened proximal end, tapering distally to a finely worked point with a cylindrical cross section. The proximal end incorporates a smooth unmodified portion of the articular surface of the metatarsal. The drilled perforation is 2 mm in diameter and is positioned 10 mm from the proximal end. The pin appears to have been crafted using stone tools, and it is comparable with similar artefacts found in Bronze Age burial mounds in Derbyshire and in East Yorkshire. The flint artefact is a cortical flake on translucent grey flint exhibiting a clear striking platform and a bulb of percussion. The dimensions of the flake are 34mm from striking platform to distal end, 23mm maximum breadth and 4mm maximum thickness.
Figure 5: Bone pin recovered from deposits in floor of 2nd Chamber
The worked antler (Figure 6 ) is the basal 150 mm of the beam of a naturally shed left side adult red deer antler with part of the brow tine and an incipient bez tine. The artefact has three deep parallel grooves characteristic of the ‘groove and splinter’ method of antler-working as described and illustrated by Clark & Thompson (1953). In this technique, longitudinal blanks with a triangular cross section are prepared using a sharp stone instrument to score parallel V-sectioned grooves through the hard outer wall of the antler. The grooves are positioned so as to follow the natural longitudinal furrows in the surface of the antler, and often three or four separate splinters were cut from a single antler beam. The specimen from Carsington Pasture Cave is broken distally, but a length of about 100 mm of the grooves is preserved including their shallow terminations close to the base of the antler. Two of the grooves penetrate almost as far as the medullary cavity of the antler, and the preparation of the splinter must therefore have been near completion. The edges of these grooves are contiguous and define a splinter with a steep triangular cross-section that is about 8 mm broad on its interior margin. The third groove is positioned on the posterior surface of the antler about 15 mm from the nearer of the two completed grooves. The third groove is about 3 mm deep and was therefore incomplete when the artefact was discarded. Fine incised striations curving away from the margins of the grooves show where the stone tool slipped during the process of shaping the grooves.
Figure 6: Antler worked with the groove and splinter technique
Figure 7: A selection of antler "combs" from Neolithic sites in Britain.
Caution must be exercised when attempting to apply typological dates to worked bone and antler, as some manufacturing techniques have changed little over the last 35,000 years, and the groove and splinter technique in particular has been documented from the Upper Palaeolithic right through to present-day arctic hunter-gathering cultures (Clark & Thompson, 1953). Within Europe, antler working using the groove and splinter technique is recorded at Late Glacial and Early Post-glacial sites in Spain, France and Germany, but in Britain the closest parallel to this artefact is the "antler comb", an object encountered on several Neolithic sites (Figure 7 Above).
Discussion and Conclusion
The presence of human remains in Carsington Pasture Cave is consistent with the widespread prehistoric practice of cave burial (Chamberlain, 1996), but in the absence of stratified artefacts or radiocarbon dating evidence it is impossible to determine when the remains were placed in the cave, or even whether multiple episodes of deposition may have occurred. The stalagmite formation covering several of the human bones is suggestive of an early date, but speleothem deposits can develop rapidly in some cave systems and cannot be relied upon as an indication of antiquity. Preliminary examination of the fauna in the cave suggests that the human remains have a maximum age of early Neolithic (about 4000 B.C.), but they are equally likely to date to the later prehistoric period. The few artefacts recovered from the cave suggest a wide range of dates (from Neolithic to post-medieval), but none of the artefacts are necessarily associated with the deposition of the human remains and this highlights the insecure nature of cave deposits, which often consist of a palimpsest of different phases of activity that took place at widely separated time intervals. Radiocarbon dating is being planned to establish the date of deposition of the human remains.
A remarkable number of individuals are represented in the assemblage of human remains recovered from the cave, and it is likely that many more human bones would be discovered if the cave was fully excavated. Within the Peak District only Ravencliffe Cave has yielded a comparable number of human remains. This site, excavated early in the twentieth century by W. Storrs Fox, contained an estimated 20 individuals together with Later Neolithic, Bronze Age and Romano British artefacts. Unfortunately the size of the human skeletal assemblage at Carsington Pasture Cave does not provide any indication of its date, as large assemblages of human remains in caves are characteristic of both the earlier and later prehistoric periods in Britain. The presence of cutmarks on the human skeletal remains is particularly unusual, and is reminiscent of the marks discovered on Late Glacial human remains at Gough’s Cave in Somerset (Cook, 1991) and marks recently discovered on human remains from the Neolithic site of Hambledon Hill (McKinley, personal communication).