Lindow Moss

[SJ 820 805]

Potential GCR site

S. Gonzalez and D. Huddart

Introduction

Lindow Moss was originally an extensive (600 ha), lowland peat bog that accumulated in a kettlehole formed during the Late Devensian deglaciation in Cheshire, but now it is about one-tenth of its former size and covered mainly by birch scrub (Brothwell, 1986). The moss margins were gradually reclaimed for agriculture and the moss cut for fuel (Norbury, 1884; Turner, 1995a). Sandy islands remain rising above the moss and are thought to be windblown dunes formed in a periglacial environment after the ice melted. During the 1980s the moss was the location of a remarkable series of four discoveries of well-preserved human remains. The discoveries of Lindow I (Lindow Woman) in 1983 and Lindow II (Lindow Man) in 1984 have been described by Turner (1986). Lindow I consisted of a skull that retained its outer membrane and some hair. The vault contained a decayed brain and part of the left eyeball and optic nerve were also identifiable. Lindow II was the archaeological discovery of the 1980s according to Turner (1995a), and almost a complete body was excavated from a peat block (Figure 8.66). Subsequent discussions established its antiquity and a wide variety of scientific investigations were undertaken (see Connolly, 1985; Stead and Turner, 1985; Brothwell, 1986; Stead et al., 1986), including pollen analysis (Oldfield et al., 1986), a study of the insects (Girling, 1986), the radiocarbon dating of the peat and the remains (Ambers et al., 1986; Gowlett et al., 1986, 1989; Otlet et al., 1986), an analysis of the man's last meal (Hillman, 1986; Holden, 1986) and chemical analysis of the skin and bone (Pyatt et al., 1991a, b). It was established by forensic investigations that the body was that of a c. 25-year-old male, who had been struck twice on the head with an axe-like weapon, garroted and had had his throat cut (West, 1986). In 1987 over 70 pieces of a body (Lindow III) were recovered (Brothwell and Bourke, 1995) and in June 1988 the skin of the buttocks and part of the left leg of an adult male were recovered about 15 m west of where Lindow III was found. In September 1988 the right thigh and ends of the right femur were recovered close to the June finds. Collectively these samples are referred to as Lindow and represent the missing parts of Lindow II. A discussion of the relationships can be found in Turner (1995b). The upper Sphagnum-rich peat has all but disappeared from the commercially worked area, substantially reducing the chance of any more finds. However, considerable amounts of this type of peat still survive at the moss, most notably to the east and to the south, where a fragment of the moss is managed by the Cheshire Wildlife Trust as a nature reserve. The locations of Lindow Moss, findspot locations and excavation trenches exposed during 1987 are shown in (Figure 8.67).

Lindow Moss is important because it has provided the most detailed analysis of bog bodies at any site in Britain (Brothwell, 1986, 1995; Stead et al., 1986; Turner and Scaife, 1995). There is a wealth of associated palaeoenvironmental analysis (Birks, 1965b; Girling, 1986; Oldfield et al., 1986; Branch and Scaife, 1995; Dinnin and Skidmore, 1995; Leah et al., 1997) and the stratigraphy and bodies have been dated. There were initial problems with this dating (Ambers et al., 1986; Gowlett et al., 1986, 1989, Otlet et al., 1986) but the site has been important as the work has shown that ¹⁴C dating of human bodies from peat bogs presents special problems because of the biochemical processes involved in the interaction between the peat and human tissue. Nevertheless, there now seems to be agreement on the dating (Housley et al., 1995) and both Buckland (1995) and Barber (1995) have presented a variety of mechanisms by which the ¹⁴C dates and stratigraphical positions may be resolved. The archaeological significance of the Lindow bog bodies has also been much discussed (Connolly, 1985; Stead and Turner, 1985; Ross, 1986; Ross and Robins, 1989; Buckland et al., 1994; Briggs, 1995; Magilton, 1995; Turner, 1995c). Specialized techniques have been used to obtain detailed information, for example, related to the food eaten (Hillman, 1986; Holden, 1986; 1995; Scaife, 1995) and the significance of the skin geochemistry (Pyatt et al., 1991a, b, 1995; Cowell and Craddock, 1995). There seems no doubt that the studies on the Lindow bog bodies have helped to advance the investigative standards and techniques used on such remains across the world (Brothwell, 1995).

Description

The stratigraphy and pollen analysis

Lindow Moss was first investigated by Birks (1965b) from the south-western moss [SJ 820 807], where the stratigraphy revealed 3.5 m of gyttjas, reed peats, carr brushwood peats, Eriophoruni and Sphagnum peats. The pollen revealed a segment of vegetational history between c. 8000 and 6000 BP and culminating some time in the early 19th century.

Following the discovery of the Lindow II body much interest was focused on the past environment and development of Lindow Moss (Barber, 1986; Oldfield et al., 1986). After the later discoveries, stratigraphical analysis of the peat using plant macrofossils and pollen has been used to give a detailed palaeoenvironmental record for the context of the bog bodies (Branch and Scaife, 1995), although the maximum peat depth proved to be 8 m in trial boreholes. Lateral variation in the peat stratigraphy is marked and a complex of interweaving bands that coalesce to form level horizons is evident, typical of ombrotrophic peat bogs. The pollen diagram (Figure 8.68) has been divided into five recognizable pollen-assemblage zones from the base of the analysed sequence at 190 cm:

Leah et al. (1997) recorded the peat stratigraphy in the active peat cuttings from 38 cores. The basal organic deposits tend to be composed of reed and sedge peats or fen-carr. In a few places Scheuchzeria-dominatedassemblages are prominent. Fen-carr peat follows these initial stages. Following this, many locations show wood peats, succeeded by peats characterized by Eriophorum/Calluna, with frequent Polytrichum, Hylocomium and Aulocomnium palustre. In turn they are succeeded by S. imbrica-turn-dominated peats at the top. The deepest peats recorded reached 7.15 m, although this was at a point situated in the peat workings where perhaps c. 3 m had been removed, suggesting original depths approaching 10 m. Away from the central area though peat thicknesses typically were between 2 and 5 m.

Insect assemblages

Two peat samples were taken from around the Lindow III body to give information about the condition of the corpse and the details of the immediate environment at the time of burial, and from a monolith through the peat in the location from which the body is believed to have come. The detailed fauna is given in Dinnin and Skidmore (1995).

Radiometric dating

An extensive ¹⁴C dating programme was undertaken to date the bog bodies and the associated sediments and detailed discussions related to the techniques and problems are given in Gowlett et al. (1986,1989), Otlet (1986) and Housley et al. (1995). The radiocarbon determinations are given in (Table 8.12). These indicate differences in age between the bog bodies and the surrounding peat, e.g. Lindow II collagen determinations are about 1800 years BP, whereas the peat surrounding the arm is about 2500 years BP.

Chemical analysis of the skin

As part of the study of the Lindow II and Lindow III bog bodies, geochemical investigations were carried out on skin, bone and associated peat (Pyatt et al., 1991a, b; 1995). The skin samples of Lindow III were obtained from the shoulder region, whereas those of Lindow II were from under the right side of the body. Bone samples from Lindow II were from the upper right orbit. The investigations revealed an excess of aluminium, silica and copper, together with traces of titanium and zinc, although the general pattern of inorganic components differed between the two individuals. Cowell and Craddock (1995) analysed further samples from Lindow III from the left heel, the palm of the right hand, a finger and part of the torso and a sample of peat for copper, initially using X-ray fluorescence and then, once copper had been detected, quantitative analysis using atomic absorption spectrometry. The torso samples had 36 ± 4 μg g⁻¹, whereas the other samples ranged from under 2 to 10  μg g⁻¹.

(Table 8.12) Radiocarbon determinations from Lindow Moss (data from Ambers et al., 1986; Gowlett et al., 1986; Otlet et al., 1986; Housley et al., 1995; Leah et al., 1997)

Food residues from the Lindow bog bodies

To investigate the dietary characteristics of the Lindow bog bodies a preliminary analysis of a sample from the fundus and body of the stomach of Lindow II was made by Hillman (1986) and Holden (1986). This indicated that the major part of the last meal was made up of cere al (the bran of wheat or rye and the chaff of barley). Chaff fragments further indicated that both emmer and spelt wheats were present and minor components included fragmentary seeds of several common cultivation weeds. Further samples were taken from the upper part of the intestinal tract and the results again included the bran of wheat/rye and the chaff of barley (Holden, 1995). Several samples were taken from the gastrointestinal tract of Lindow III and although the results were disappointing (Holden, 1995) the food was dominated by fragments of the testa and brown inner layer of the pericarp of the hazelnut and there were smaller quantities of cereal bran. Nineteen samples were examined for pollen analysis from faecal, colonic, gastrointestinal, duodenal and rectal residue from Lindow III (Scaife, 1995) to compare with the results from the stomach and intestinal contents of Lindow II (Scaife, 1986). The results from Lindow III show largely cereal grains, although they were degraded, which contrasts with the good preservation from pollen in Lindow II (Scaife, 1995).

Interpretation

The bog bodies

There seems little doubt that Lindow IV represents the lower limbs of Lindow II cut off by the peat digging machinery in 1984 (Turner, 1995b) and it seems highly likely that the body parts allocated to Lindow III are all from the same adult male, buried close to Lindow IV Both died violent deaths and were buried naked, except for Lindow II's fox fur armband and the cord around the neck (Budworth et al., 1986). Both are much the same age and build and there are suggestions of traces of body painting (Pyatt et al., 1991a, b, 1995) from the geochemistry of the skin. However, although the copper content of the torso sample is higher than the other skin samples and also the peat sample (Cowell and Craddock, 1995), the authors suggest that the elevated skin content may be at least partly explained by the substantial, and possibly selective, loss of organic body weight through decay, leading to a preferential enhancement of the inorganic components. Hence the moderately higher concentrations on the torso could be explained by factors other than deliberate painting. The neatly trimmed hair, beard and fingernails suggest that both were of high status. As there is no evidence for predation by insects or mammals it would appear that both bodies were thrust or buried in the peat bog, a fact confirmed by the same difference in ¹⁴C date between the bodies and the layers from which they were retrieved. The cause of death was different, as Lindow II had multiple injuries whereas Lindow III was just beheaded. The depositional environment seems to have been different because Lindow II was found in a layer indicating a bog pool, whereas Lindow III was on a peat surface. The calibrated date range for Lindow II with a 95% confidence limits based on the Oxford ¹⁴C dates is 2 BC to AD 119, whereas the equivalent unified Oxford and Harwell date range for Lindow III is AD 25–230, so the two bodies may be broadly contemporaneous, or deposited up to 200 years apart. The date for Lindow III differs from the conclusion by Stead (1986), who then favoured a Middle Iron Age date. Turner (1995b) suggests that the most likely explanation for the presence of the two bodies in the bog is that they represent sacrifices, probably for religious reasons. As such they belong to a well-established practice across northern Europe, Britain and Ireland (Van der Sanden, 1995; Turner 1995c; Ó Floinn, 1995a, b). Much has been made of Lindow II's triple death and tripalism is one of the commonest Celtic religious symbols (Magilton, 1995), but Ross (1986) and Ross and Robins (1989) have gone further in linking the triple death to the appeasement of three Celtic gods by a ritual sacrifice, also suggested by the presence of mistletoe and blackened bread in his gut. Beheading as suffered by Lindow III also is a central part of Celtic mythology and religious practice (Ross, 1967) and has local parallels from Worsley (Garland, 1995), Red Moss (Smith, 1988) and Pilling Moss (Edwards, 1969).

The palaeoenvironment and mire ontogeny

Birks (1965b) interpreted the pollen record post-dating the Elm Decline as representing the effects of seven clearance episodes, but his pollen sampling intervals were wide by modern standards. However, the record demonstrates qualitatively that human impact had a major impact on the surrounding landscape from the Neolithic onwards. This has been confirmed by a detailed analysis of the palaeoenvironment, which was concentrated in the upper 200 cm of peat, which spans the later prehistoric period (Branch and Scaife, 1995). The main stratigraphical feature here is the change from well-humified Sphagnum and monocotyledonous peat to fresh, unhumified and largely Sphagnum imbricatum peat, which equates with the boundary between the LIN:2 and LIN:3 pollen-assemblage zones. This horizon has been seen across most of the moss and appears to be a major recurrence surface and it has been ¹⁴C dated to between 2447 years BP and 2345 years BP (Table 8.12). It probably represents a period of gradual climatic deterioration after 2900 years BP and a reduction of 1°C for Sphagnum pools to commence at the recurrence surface (Barber, 1982).

During the lower two pollen zones prior to the recurrence surface, Sphagnum was dominant in wetter areas of pools and damper hummocks, with Gramineae and Cyperaceae. Calluna and Erica probably were growing on drier hummock areas. Alnus and Salix probably were growing in carr woodland along the nutrient-enriched marginal areas of the bog. The drier, adjacent land was a dominantly wooded environment, with some areas of open agriculture. Quercus and Corylus were dominant throughout the pollen record, but with many other woodland species. Openness is shown by Ilex and Corylus, which require light for flowering. Branch and Scaife (1995) suggested that although woodland was dominant in the period pre-dating the recurrence surface, there had been anthropogenic disturbance and secondary woodland (Fraxinus peaks and Calluna). Agricultural activity is noted by the presence of Plantago lanceolata and other herbs. Archaeologically this disturbance has been substantiated by the presence of Neolithic artefacts from one of the sand islands (Turner, 1995a) and these could be contemporary with the ¹⁴C date from the charcoal-rich soil (Table 8.12). Charcoal is not restricted to the immediate environs of the sand islands, however, and it has been noted throughout the mire's history; apart from its very earliest growth stages (Leah et al., 1997). Burning was likely to have affected the local vegetation, as in many raised mires in north-west England (Huckerby and Wells, 1993; Hall et al., 1995; Middleton et al., 1995). The numerous pine stumps exposed on the cut surface of the moss also often appear to show signs of burning (Leah et al., 1997).

Above the recurrence surface there was no cessation of peat growth but a rapid environmental change, although dry-land vegetation continued to dominate the tree pollen, with an open-canopy woodland of mixed deciduous type, but with less lime. From 120 cm in LIN:3 there is a marked and progressive increase in the diversity and percentages of many herbs, especially disturbed and open-ground species, which correlates with the Iron Age to Romano-British period (Branch and Scaife, 1995). There also are sporadic but more abundant cereal-type records, which reflect increased arable and possibly pastoral agriculture, adjacent to the moss. Increases in the pollen of Ericaceae and bracken spores may be indicative of soil deterioration, especially on the sandy soils nearby. This phase of human disturbance also has been documented by Birks (1965b) and Oldfield et al. (1986) and it is clear that there was a progressive dis turbance from the early Iron Age, which attained a maximum extent during the Romano-British period. Branch and Scaife (1995) suggest that this disturbance was a regional phenomenon and perhaps can be applied to north-west England (Oldfield, 1960a; Mackay and Tallis, 1994; Middleton et al., 1995). The boundary between LIN:3 and LIN:4 pollen-assemblage zones is where there is a reduction in the percentages of herbs and the taxonomic diversity noted in LIN:3, and a late Romano-British date is suggested ((Table 8.12), UB-3238). In LIN:5, Quercus, Alnus and Corylus regain their earlier importance, showing a return to woodland dominance and reduced agriculture in the region. From the pollen analysis of the Lindow III body Branch and Scaife (1995) suggest that it was placed on the growing bog surface, some 10–15 cm above the recurrence surface and not in a peat cutting and they suggest an age of c. 2300–2200 BP.

From the stratigraphical work of Leah et al. (1997) the moss began life as an extensive area of reedswamp, fen and fen-carr occupying hollows within glaciofluvial and aeolian deposits, which gradually became carr woodland. The deepest deposits probably go back to the early Flandrian. The fen-carr phase ended with an increase in wetness, evidenced by the flooding indicator Scheuchzeria in places. A reversal of edaphic conditions immediately after this phase led to much drier conditions and the establishment of a slow growing mire and trees, notably pine, with several phases of pine forest establishment, rather than one discrete arboreal phase. Radiocarbon dates from a pine layer suggest that the species formed a significant component of the mire flora at various times between 7780 ± 70 years BP and 4060 ± 70 years BP (Table 8.12). This relatively dry, tree-dominated stage was followed by a return to wetter conditions and the establishment of Eriophorum–Calluna-dominated vegetation across almost the whole moss. Conditions became even wetter and S. imbricatum became the dominant peat-former.

The beetle fauna (Dinnin and Skidmore, 1995) from around the Lindow III bog body suggests that the corpse was deposited in a wet Sphagnum bog with pools of acid water, with this water surrounded by a typical oligotrophic bog flora of cotton grass, mosses and heather. The complete absence of any carrion fauna leads to the conclusion that the body was rapidly submerged. This was the same for the Lindow II body (Girling, 1986). Despite the stratigraphical recurrence surface, the insect assemblages above, below and around this feature revealed no appreciable change in the moisture status of the bog, or any indicators of climatic change.

The dates for the Lindow II and III bodies and the peat, the state of decay of the corpses and the insect evidence require some reconciliation. Buckland (1995) suggests that a human body could have been pushed down into a deep pool and never reappeared, but Barber (1995) refutes the deep-pool hypothesis, especially as at Lindow the evidence from the Cladocera and the Chironomidae point to a shallow pool (Dayton, 1986). He suggests that the body was inserted into the upper layers of the bog, where the felted pool peat could be rolled back, the body laid on the exposed surface and the upper peat rolled back over it. This would create minimal disturbance to the stratigraphy and its included biota and would be impossible to detect.

The results of the analysis of the gut contents of the Lindow II body show that emmer and spelt wheats and barley were the main constituents (Holden, 1995), which probably had been prepared as an unleavened bread. The Lindow III body on the other hand seems to have consumed hazel nuts and a smaller amount of wheat/rye. The palynological evidence (Scaife, 1986, 1995) illustrates a diet composed largely of farinaceous products, although the presence of mistletoe gave rise to speculation as to the possible druidical practices. Scaife (1986) suggests it was more likely that mistletoe was used for medicinal purposes.

Although the Lindow bog bodies are the best known British examples, research has illustrated (Briggs and Turner, 1986; Turner, 1995b, d) that northern England is particularly rich in such remains (Figure 8.69). For example, they have been documented from Austwick Common (Denny, 1871), Scaleby Moss (Turner, 1988), Seascale Moss (Turner, 1989), Whixall Moss (Turner and Penney, 1996), Grewelthorpe Moor (Turner et al., 1991), Red Moss (Smith, 1988), Pilling (Edwards, 1969) and Worsley (Garland, 1995).

Conclusions

Lindow Moss has proved to be the best-documented site for bog bodies in Britain and has provided a stimulus for new research on these remains elsewhere. The palaeoenvironment of the bog throughout its history has provided much information on mire ontogeny and the influence of humans on the vegetation in the later prehistoric period. The stratigraphy and the bog bodies have been dated extensively and the problems associated with ¹⁴C dating of such remains discussed in detail. The archaeological significance of such bog bodies has been debated extensively. Owing to the extensive commercial working of Lindow Moss and landfill development it is unlikely that further bog body remains will be found, but the site, especially the nature reserve, is important as a representative type locality of the large lowland moss in northwest England.

References