Worcestershire Record No. 20 April 2007 p. 43-42

P. F. Whitehead

For many years from 1970 it was possible to glean from excavations in the bed and on the floodplain of the River Avon, evidence for environment and biota from the sedimentary successions revealed. Opportunities at some sites were lost, and at the time the facilities required to work on some groups, such as fossil insect remains, were not in place. Nevertheless, useful data was gleaned and this account deals with a somewhat limited amount of it, mostly of Bronze Age date.

Shotton (1978) proposed a typical section in alluvium. This was to establish a general working principle and it was stressed that alluvial sedimentology was, in fact, rarely typical and often entirely atypical. The River Avon now runs in a channel incised into this alluvium. The key features that may be looked for in a typical section through alluvium are, from the top down:

1. Red to reddish brown silty clay perhaps 2-3m thick aggraded rapidly during the Bronze Age to Iron Age
2. Grey organic sediments frequently Bronze Age
3. Coarse gravel of Mesolithic age (8460 BP) at Evesham
4. Lower Lias Clay bedrock

This is a highly simplistic convention. Whilst it may well occur as such (Shotton, 1972), it should be appreciated that as a post-glacial river cuts across its floodplain, sediments may be shifted and reassembled and older aggradations incised (Sandford, 1965). This was demonstrated in the valley of the River Alne at Alcester (Shotton, Osborne & Greig, 1977). Viewed in vertical section, evidence for the sedimentary aggradations that follow these erosive processes may create something that appears not to be sequential at all. Gravels underlying well structured alluvial clays and silts at Evesham contain animal bone dated by radiocarbon assay to 8460 BP (Birm-655) (Whitehead, 1979), and similar gravel beds occur elsewhere in the Avon valley (e.g. Fig. 4). They result from the post-glacial erosion of older aggradations prior to the development of binding vegetation, and to the effects of rising sea-levels following ice-sheet ablation. There is some evidence of a hiatus in alluviation in the Avon valley with recommencement about 6000 BP, supported by radiocarbon dates from the River Arrow at Ipsley and Ashmoor Common downstream on the River Severn, of between 5000 BP and 6000 BP (years before present) (Shotton, 1978; Brown, 1981).

Sections through River Severn alluvium were established in 1981 at sites including Callow End and Ashmoor Common, by the researcher Anthony Brown (Brown, 1981). At some sites on the River Severn, for example south of Diglis, Worcester, alluvium is at least 7m thick, which may be a consequence of river bed lowering associated with the cutting of the Ironbridge Gorge. When Tewkesbury Docks were constructed in the 19th century, exceptional thicknesses of alluvium attaining 9.2m were recorded overlying lacustrine organic beds themselves 2.3m thick, containing shells, pottery and the 'horn' of a Red Deer (Symonds, 1861a); a measure of high biological productivity at the confluence of two major lowland rivers which, judging from the thickness of the deposits, must have supported an extraordinary wetland biota.

Similar alluvial sequences occur widely in the valley of the River Thames. Between Abingdon and Reading, Red Deer (Cervus elaphus L., 1758), and more rarely Wild Boar (Sus scrofa L., 1758), are represented in basal organic-rich clayey sediments (PFW, pers. obs.). Goudie & Parker (1996) describe sections in the Thames alluvium near Lechlade, and Red Deer-bearing alluvium has also been recognised higher up the Thames headwaters (Gardiner, 1938).

The aggradation of alluvium that brought the floodplain surface up to its modern level (Figs 1, 2, 4) is a response to sediment influx associated with woodland clearance of the valley sides (Shotton, 1978; Brown & Barber, 1985; Morgan, 1986). This major episode of alluviation was completed rapidly, probably by the end of the Iron Age (Whitehead, 1989a). Overbank flooding and more minor episodic alluviation has occurred ever since but alluvial aggradation by overbank flooding at the present time is relatively insignificant because the amount and nature of mobile sediment in the catchment is greatly reduced. The introduction into the catchment of the volume of sediment required to bring the floodplain up to its present level must have produced an environmental catastrophe of some proportions with the wide scale obliteration of a rich diverse floodplain biota. The claim that this alluvium is inorganic (Shotton, 1978; Brown, 1983), now known to be inexact even if broadly indicative, is a biologically frightening one. That this major period of alluviation occurred universally by organised sequential processes of overbank flooding is unlikely. Rapid dumping and extensive recycling and reassortment are likely in many places when water levels fluctuated as the river responded to localised variations in load and flow rate. This may explain the presence of the bone of a Giant Ox (Bos primigenius Bojanus, 1827) from Pensham, discussed later, which is possibly significantly older than the alluvium that contained it.

The mud of the modern river bed contains evidence of all of the rejectamenta of post-Iron Age life and culture that found its way into it. In some places, such as was established at Bidford-on-Avon during the 1971 bed clearance, it is remarkably rich and diverse. Its organic mud supports a contemporary biota, some representatives of which are scarce and localised. Evidence of pre-Roman Celtic life and culture occurs rarely due to lateral erosion of the alluvium.

Fig. 1. Evesham, SP045444, section through alluvium at bankside, 5 May 1985

Fig.2. Evesham, Aldington, SP049448, section through alluvium near river, 27 February 1986

Alluvium at Evesham, Worcestershire, 1985-1986

Figs 1 and 2 show two alluvial sequences at Evesham some 400m apart, which conform to the 'Shotton norm' (vide supra). Fig. 2 resulted from an examination made for the Department of the Environment and Transport during the construction of the Evesham bypass. Fig. 1 shows brown alluvium 1m thick whilst in Fig. 2 it is slightly more variable in structure and 4 m thick; both confirm a lower grey organic alluvium bearing remains of Red Deer (C. elaphus). At Aldington, a Red Deer tibia had been gnawed by domestic dogs.

At SP045444 (Fig. 1), molluscs were extracted from 10 gms of clayey sediment from the shaft of a femur of a Red Deer. The 21 species recovered appear to have been living in a sluggish watercourse that was evidently somewhat exposed (there was no evidence to suggest woodland close at hand) and not markedly plant rich, the bed of which infilled with grey calcareous clay. Two nutlets of Bulrush Schoenoplectus lacustris L. and a utricle of sedge Carex sp. were isolated from the deer bone. A single specimen of Valvata was ascribed to the Red Data Book-listed Valvata macrostoma Mörch, 1864, constituting a new regional record. This specimen is juvenile which poses identification difficulties. It was examined by Dr D. T. Holyoak who suggested (in litt., 20 May 1985) that because it was not fully grown, a longer succession of specimens including adults would have been a desirable ideal, but confirmed that the specimen did show the characters of V. macrostoma. Since Valvata cristata Müller, 1774 and V. piscinalis (Müller, 1774) were both found with it, and comparisons were made with them, the record would seem to be supportable. Today V. macrostoma is a rare inhabitant of rich marsh drains in southern and eastern England (Kerney, 1999).

It has often been claimed that the upper red-brown alluvium of the River Avon is inorganic, but the presence of numbers of the aquatic snail Bithynia tentaculata (L., 1758) (with nothing else obviously organic) amongst comminuted shells in sections at Aldington, Evesham (Fig. 2), demonstrates that it aggraded in a sluggish river in which most of the hydrophytes would have been eliminated by turbidity.

Fig. 3. Evesham, SP045444, 5 May 1985, River Avon bankside. Molluscs from 10 grams of sediment from shaft of femur of Red Deer, basal alluvium, prehistoric, probably Bronze Age (see Fig. 1).

Excavations at Pensham, Worcestershire, during February 1984, enabled a section through alluvium (Fig. 4) to be constructed. The essential sequence was similar to those observed at Evesham (Figs 1,2), but organic sediments preserved beneath 2m of red-brown silty alluvium were intensely black, bearing remains of contemporary animal life and large amounts of wood, and overlay coarse gravel. These gravels rest on Lower Lias Clay and had been stained black as humic acids from the younger organic sediments permeated through them, and some mammal bones from the gravels were mottled black and orange. These are generally broken and heavily abraded as may be expected in a river transporting coarse gravel.

Mollusc shells were observed in large numbers and a small bulk sample of 1.04 kgs (Fig. 4) from towards the base of the sediment was passed through successional sieves, producing a remarkable 48 species (Fig. 5). As might be expected in a lacustrine context, these shells showed no trace of stratigraphic bedding, and seemed to be in primary death assemblages. Insect remains, which were not apparent in large numbers, were generally not recorded, although hazelnuts colonised by the nut weevil Curculio nucum L., 1758 were noted, as well as a rove beetle Anotylus rugosus (F., 1775) which remains widespread (Whitehead, 1992). A mite Euzetes nitens (Johnston, 1856) is of a species often found around the margins of lush vernal pools with shaded stable muddy herbage-covered margins. The plant remains, molluscs and ostracods from the bulk sample confirm that they originated in and around a clear water body with well-developed aquatic vegetation. A large cleft fragment of wood from the crown of an oak tree (Quercus robur L.) from the base of the organic bed provided a radiocarbon (14C) date of 3540+-90 BP (Birm-1228), placing it in the later part of the early Bronze Age. The difference in age between the base and top of the organic bed, and therefore the overall time period of its development, is unknown but there is no a priori reason for it to be very marked. Remains of man, together with those of Red Deer that probably did not die naturally, suggest that the waterbody was a focus for Bronze Age people. The lateral extent of it underneath the floodplain was not determined.

That the grey organic silts and clays in the Evesham sections (Figs 1, 2) are of comparable age is likely. Their structure implies some river bed stability and they contain the remains of Red Deer. The association of Red Deer with organic grey to black well-structured silts or clays in the basal alluvium is indicative, and in conjunction with Giant Ox Bos primigenius diagnostic, of the Bronze Age in the Severn and Avon valleys. There was clearly a substantial population of Red Deer in the riparian woodlands of the lower Avon valley at that time, which evidently suffered a major regional population collapse from which they could never fully recover, when the woodlands were cleared and then obliterated by alluviation. The same comment applies to Giant Ox. A Neolithic stone axe head, an agent of such clearance, was found close by at Birlingham during 1983 (see Appendix to this paper). These riparian Bronze Age alder woods are probably up-river extensions of the presumably vast post-glacial alder-dominated woods that were a feature of a number of larger west-coast estuaries (Smith, 1984).

As the Bronze Age lacustrine sediments at Pensham aged and consolidated, some of the Red Deer antlers contained in them were exposed and subjected to erosional processes before or during the aggradation of overlying alluvial clay, so that in some cases, although they were not dislodged from their primary contexts, what projected was abraded by attrition downwards to what is now the surface of the Bronze Age bed.

In this account nomenclature mostly follows Stace (1997). The 1.04 kg bulk sample yielded evidence of herbaceous and arboreal vegetation from a reasonable spectrum of niches and enables a tolerably clear reconstruction of the vegetation of the lake and its surroundings to be made. The information provided below is based on macrofossils only; pollen samples were not taken, although a sample of sediment has remained in a sealed tube since 1984.

There is a suggestion that trees may have overhung part of the water body, dominantly alder Alnus glutinosa (L.) Gaertner, of which several strobili-axes were recovered together with 18 seeds. Other trees growing close by included Cherry Prunus avium (L.) L., Hawthorn Crataegus sp., Pedunculate Oak Quercus robur L., and Wych Elm Ulmus glabra Hudson. In the midland river valleys Ulmus glabra is extant as a riparian element. The presence of Hornbeam Carpinus betulus L., attested to by a single nut and part of a bract, is an interesting regional record of a thermophilous tree now localised in Worcestershire, but which forms stands at one or two sites. Raspberry Rubus idaeus L., Blackberry Rubus fruticosus agg., and rose Rosa sp. grew in the wildwood. Elder Sambucus nigra L. and Hazel Corylus avellana L. were abundant and there is some impression of impenetrability; Hazelnuts were found abundantly.

The pool itself contained an abundance of Yellow Water-lily Nuphar lutea (L.) Smith, and the evidence for charophytes, implying clear clean water, was confirmed by Miss A. P. Connolly as Chara sp. Two species of Potamogeton were identified but neither was determined to species. Bulrush Schoenoplectus lacustris L. grew in the shallows and sedges Carex sp. occurred. Two mericarps of Fine-leaved Water-Dropwort Oenanthe aquatica (L.) Poiret were recovered and buttercups Ranuculus acris L. and Ranunculus repens L. grew along the fringes of the pool with three species of dock including Rumex sanguineus L, Rumex conglomeratus Murray, and Prickly Sowthistle Sonchus asper (L.) Hill. The presence of these plants may imply some marginal disturbance, probably by visiting large mammals, and some of them suggest that part of the marginal area was relatively open. This could also have been due to anthropo-impacts. Fool's Watercress Apium nodiflorum (L.) Lag. also grew along the fringes of the pool; this is a stoloniferous plant that withstands some trampling. Marsh Violet Viola palustris L. was identified by Miss A. P. Connolly who also cited Greater Water Parsnip Sium latifolium L.; evidence for this last was retained by Miss Connolly and presumably remains at the University of Leicester.

Curiously, the most frequently represented plant group in the 1.04 kg bulk sample is Goose Foot Chenopodium spp., which provided 81 seeds. There is a possibility that their growth was promoted, if not as a crop by man, then by the fertilising activities of other large mammals. 79 of these seeds belong to Fat-hen C. album L. but one belongs to Upright Goose Foot C. urbicum L. and a second much larger seed is also apparently that of C. urbicum.

Like the molluscs, plants provide useful opportunities for the reconstruction of habitat and ecology, but not of time, which is why absolute dating is invaluable. There is a near-perfect analogy for the vegetation at Pensham from sediments in the order of 115000 years older near St Ives, Cambridge (Preece & Ventris, 1983).

In this account nomenclature mostly follows Reischütz, 1998. Forty nine species of mollusc were recorded from the Bronze Age organic bed, of which forty eight were identified from the bulk sample. Excluding opercula, 3433 (95.4%) are aquatic and 158 (4.6%) are terrestrial (Fig. 5). Terrestrial molluscs are often poorly represented in relatively still-water assemblages since adjacent terrestrial habitats are neither scoured nor eroded. This is a rich fauna dominated by aquatic herbivores with planorbids well represented. Bithynia leachii (Sheppard, 1823) and Radix peregra (Müller, 1774), a dominant in the assemblage, tend to prefer slow flow regimes. The sphaeriid bivalves have virtually no high-energy flow tolerants, with just a single valve of the riverine Pisidium amnicum (Müller, 1774) and 10 of Pisidium henslowanum (Sheppard, 1823). The impression is of a closed or sluggish water body isolated on the floodplain, although this does not imply that it was not subject to changes of level. The organic sediments were unbedded and the molluscs were a primary thanatocoenosis. There is evidence that trees, especially alder, Hazel and oak, fringed and overhung part of the water body creating moist shaded conditions favoured by Ena montana (Draparnaud, 1801), Discus rotundatus (Müller, 1774), Helicigona lapicida (L., 1758) and Cochlodina laminata (Montagu, 1803) of which the latter species still persists in ancient woodland nearby (Whitehead, 1984).

The marsh snail Zonitoides nitidus (Müller, 1774) points to a muddy margin that is likely to have been subject to episodic fluctations of water level and Trichia hispida (L., 1758), the commonest terrestrial species, would have occurred there. Vallonia pulchella (Müller, 1774) is characteristic of wetter marginal habitats, Vallonia excentrica Sterki, 1892, of drier more open grassland, and Vallonia costata (Müller, 1774) most typically of yet drier situations (I have exceptional records of populations of this species on trees). Although the floodplain and its environs were clearly wooded, there is some evidence of a habitat mosaic, possibly on small spatial scales, and some of these terrestrial molluscs imply, as does the vegetation, that part of the waterbody was relatively exposed or at least not densely shrouded by trees. The very small numbers of succineids may suggest that marginal emergent plants (heliophytes) were limited in extent, but this could also reflect the position of the bulk sample in an extended horizontal sequence.

There is a complete absence of terrestrial wetland molluscs of the genus Vertigo. The Pensham site is situated at the base of an outcrop of Lower Lias Clay that forms a well-marked valley side ascending steeply above the modern meander loop. Vertigo spp. often demonstrate a range of habitats on open floodplains and the suggestion is that the topography of the valley side, likely to have been wooded, constrained the development of the floodplain on that side.

Aquatic mollusc faunas are generally good indicators of environment (Sparks, 1961) but not of time. The aquatic assemblage from Pensham with its many herbivorous planorbids, lymnaeids, and bithyniids can be mirrored almost exactly in Ipswichian sediments 120,000 years old (Sparks, 1964), and to a large measure even further back in time, in the Cromerian of 500,000 years ago (Gilbertson, 1980).

Fig.5. Pensham, Worcestershire, 19 February 1984. Molluscs from Bronze Age bulk sample 1.04kg. Sphaerids = valve count/2

It is observed that modern culture-favoured river molluscs (i.e. those that benefit from reasonably stable conditions in a canalised river following the termination of a major episode of alluviation) such as unionids, Viviparus viviparus (L., 1758), and Sphaerium rivicola (Lamarck, 1818) are totally absent from the Pensham Bronze Age fauna. This agrees with another Bronze Age mollusc fauna from Pilgrim Lock, Bidford-on-Avon (Shotton, 1972).

Samples of molluscs were also isolated from the shaft of a human femur, from the cervical vertebra of a Red Deer, and from the neural canal of an ox vertebra, probably an immature female Giant Ox (Bos primigenius Bojanus). These total 21 species and align perfectly with the mollusc fauna from the bulk sample, except that they add the River Nerite Theodoxus fluviatilis (L., 1758), which requires hard substrates to adhere to, and which most probably employed bones for that purpose.

One species of ostracod was recorded in the organic bed in large numbers. This was Herpetocypris reptans (Baird, 1835), a species of still plant-rich waters that swarms around hydrophytes.

Bithynia leachii (Sheppard, 1823) Bithyniidae. This is a somewhat localised herbivorous snail generally of larger stable sluggish biologically productive water bodies. It was scarcer than B. tentaculata and occurred with it in proportions of about 1:18. In the case of B. tentaculata, shells outnumbered opercula by a factor of something in the order of 4:1; as the snails died and the opercula were released, slow currents would have segregated them. On this basis about 27 of the somewhat rounder opercula of B. leachii should have been found. In reality none were found which may simply be due to chance.

Ena montana (Draparnaud, 1801) Buliminidae. Red Data Book 3 (rare) status. The same comments apply to this species as to H. lapicida (L.) below, but rather more so since the English range of E. montana is more southern and restricted (Kerney, 1999). According to Boycott (1939), it occurs amongst scrub in the Mendip Hills but examples seen by the author from the Somerset Mendips, are, like H. lapicida, associated with broken ancient woodland in humid defiles, where it especially favours fallen crown wood in medium shade. Kerney (1968) found that it was able to persist for some time following significant disruption of its habitat during the Bronze Age.

Helicigona lapicida (L., 1758) Helicidae. This is now a rare species in Worcestershire, and one that has declined in many parts of England (Kerney & Stubbs, 1980). Although it is still extant on Bredon Hill the populations are relict and threatened and the species is likely to be approaching extinction there (P.F. Whitehead, pers. obs.). At Elmley Castle (SO94) a shell believed to be from an individual dead for less than five years was found at the base of an old hawthorn enclosure hedge in the village on 12 November 1982 (P. F. Whitehead, pers. obs.), but such findings are very rare and may not now be repeated. On the Cotswold Hills small populations survive in dry stone walls (P.F. Whitehead, pers. obs.). The essential point about H. lapicida is that it is a shade-preferring mollusc (Limondin-Lozouet & Preece, 2004) adapted to the kind of climatic constancy that extensively wooded tracts of land may provide, and one which resents small-scale or regional vacillations of climate, particularly with reference to desiccation. Confirmation for this comes from stable populations in the Mendip Hills of Somerset, and in the Wye Valley of Gloucestershire, which occur in ravines or deep gorges in a local climate totally dissimilar to the ambient (P.F. Whitehead, pers. obs.).

Pisidium moitessierianum Paladilhe, 1866 Sphaeriidae. This is a localised species in Worcestershire and a temperate indicator in Britain which does not occur in any of the 'cold' Pleistocene mollusc assemblages of the Avon valley (Whitehead, 1989b). It is an indicator of comparatively 'good' fluvial conditions. Although it was not recorded in the Bronze Age lacustrine sediments at Pensham (for it is dominantly a slow river species), it was isolated from an assemblage of mollusc shells in a seam of buff silty fine sand resting on Lower Lias Clay at Pensham at SO933437 on 31 March 1984. As the lowest alluvial member these sediments are unquestionably prehistoric and probably Bronze Age. This fauna is not discussed here in detail as it contained species typical of fluvial conditions, although it was noted that almost all of the aquatic gastropods were juveniles. There was little evidence of rich plant growth and Hippeutis complanatus (L., 1758) was represented by one example in 400 molluscs. Bithynia leachii was absent. This confirms that P. moitessierianum occurred in the earlier post-glacial history of the River Avon, as it did in the Kennet valley of Berkshire (Holyoak, 1983).

Opportunities to describe primary lowland river mollusc assemblages of medieval age in the River Avon are very rare because the floodplain had been brought up to its modern level prior to that. However, on 19 July 1972, a small remarkably curved horn-core of a domestic ox (Fig. 6) was retrieved from the bed of the River Avon at Evesham. This yielded a radiocarbon (14C) date of 690 +- 100 BP (Birm-1051).

Fig. 6. Bed of River Avon, Evesham, Worcestershire, 19 June 1972. Small very strongly curved right horn-core of medieval (690+-100 BP) domestic ox.

Fine clayey sediment removed from the horn core produced the following molluscs: Valvata piscinalis 3, Bithynia tentaculata 3, Galba truncatula (Müller, 1774) 1, Radix peregra 1, Planorbis carinatus (Müller, 1774) 1, Gyraulus albus (Müller, 1774) 4, Acroloxus lacustris (L., 1758) 1, Succinea putris (L., 1758)/Oxyloma pfeifferi (Rossmässler, 1835) 1, Cochlicopa lubrica (Müller, 1774) 1, Musculium lacustre (Müller, 1774) 1, Pisidium casertanum (Poli, 1791) 1, and Pisidium moitessierianum 7. Although it is not suggested that the molluscs also date from 690 BP (the radiocarbon date is given to a convention which implies that there is a 2:1 likelihood that the true date falls between 790 BP and 590 BP) there is no question that they will be medieval, having entered its interstices shortly after it was thrown into the river. The dominance of P. moitessierianum indicates that the river was generally favourable then, in places more sluggish than at the present day, and that P. moitessierianum is likely to have persisted in the River Avon for a long period of Flandrian time. Shotton (1972) regarded a specimen from Harvington as modern and it may well be that P. moitessierianum is extant in River Avon.

Mammal bones were isolated in the field, both from material raised by the excavating equipment, and from sections revealed by bank side excavation (Fig. 3) on the inside of a large river meander. Teeth and bones of rodents were isolated from the 1.04 kg bulk sample. Some of these bones were clearly in primary death positions and included complete unabraded antlers and bones of Red Deer, some of which demonstrate evidence for both human and carnivore activity. The mammals represented are hedgehog Erinaceus europaeus L., 1758, shrew Sorex araneus L., 1758 (also on the basis of characteristically damaged shells of the land snail Trichia striolata Pfeiffer, 1828), Wood Mouse Apodemus sylvaticus L., 1758, Water Vole Arvicola terrestris (L., 1758), Field Vole Microtus agrestis (L., 1758), man Homo sapiens L., 1758, Red Deer Cervus elaphus and Giant Ox Bos primigenius, this last only by parts of the appendicular skeleton. Man was represented by a lumbar vertebra and a femur, both articulations of which have been removed by dogs, and they may explain its presence in the pool. Fish also lived in it but the species were not identified.

Wolf Canis lupus L., 1758 Canidae. The only canid bone located at Pensham was not found in situ so could not be assigned to an alluvial provenance. It is a humerus which probably represents wolf (Canis lupus L.). The distal articulations and condyles are very well developed, bringing with it implications for a free-living, rather than domestic animal, and the maximum width of the distal extremity is 44 mm. The shaft of the bone contained coarse sand.

Giant Ox Bos primigenius Bojanus, 1827 Bovidae. Giant Ox (or Aurochs) is a woodland animal that became extinct during the Bronze Age in Britain but persisted throughout the Later Prehistoric Period in Europe (Bartosiewicz 2005). The last truly wild individual died in Poland in AD1627, but ancient breeds of cattle survive (Whitehead, 1963). Whitehead (1979) collated dated records of Giant Ox in Britain that extend from 9557 BP to 3245 BP, and that chronological framework has remained essentially constant. This last date is from Charterhouse Warren Farm in Somerset and remains the most recent British example, although another, dated more recently from Galloway in Scotland, is placed at 3315 BP (Kitchener & Bonsall, 1999, who also give calendar-corrected dates). Although a cervical vertebra from the base of the organic bed at Pensham is from a mature bull, the radius of an old bull found in situ in the overlying alluvium (Fig. 3) is of particular interest. This had been forcibly disarticulated and cleft for its bone marrow, although there was nothing to suggest that it was located on a human activity horizon. This technique of longbone cleavage for bone marrow extraction was practised, to an almost uniform pattern, for >120,000 years of contact between humans and large grazing mammals in the Avon valley. By stratigraphic principle this bone should be more recent than 3540 BP (Birm-1228), but my expectation is that it may prove to be somewhat older than the alluvium that rather surprisingly contained it. This also further dispels the notion that this alluvium is inorganic (vide supra).

At Barton, Warwickshire, remains of Giant Ox, dredged out of black organic beds deep in alluvium on 4 December 1971, included the massively thickened frontal bones of a bull, together with a talus bone, also from a bull, from which it was possible to compute a live weight of 586 kg (sensu Noddle, 1971). This is an underestimate because the ratio of muscle to fat on which the live-weight calculation depends is different in wild animals to that of the domestic ones that were employed to create the baseline data. Dawkins (1866) refers to the remains of a bull Giant Ox hauled from the bed of the River Avon near Bath, as being from an animal likely to exceed 12 feet in length and 6.5 feet in height, with a horn-core span of 39 inches, truly a Giant Ox. The association of Giant Ox in post-glacial England is strongly with biologically productive riparian forest. At East Ham, London (Franks, 1961), the arboreal environment is likely to have mirrored that of Pensham. According to Symonds (1861b): "Mr Strickland, of Apperley Court, possesses a fine head of Bos primigenius and also of a large Bison priscus from the ancient lake beds of Cropthorne, Worcs." The cranium of Bos primigenius is in Worcester City Museum and is post-glacial. I do not recall seeing the head of the bison so cannot comment on that, but this clearly demonstrates that bone-bearing alluvial deposits comparable to those at Pensham exist at Cropthorne.

The relationship of prehistoric man with Giant Ox need not have been simply that of hunter and hunted. Special relationships between the two could have existed, and it would be unwise to suppose that in rapidly changing environments the objective of resource conservation was never an issue (Jarman, 1969).

Red Deer Cervus elaphus L., 1758 Cervidae. Red Deer was represented by a minimum of 14 animals, including two complete antlers from the same animal. In all determinable instances none of the antler-bearing stags had shed their antlers, indicating that death occurred in winter. If they were hunted, and it is difficult to conceive that such a resource was not, that may suggest that it was carried out seasonally, at a time when the stags congregated. One antler from a mature stag had a circumference just above the burr, or coronet, of 265mm; another from Aldington (Fig. 2) had a circumference of 252mm. Clearly these were robust animals. One Red Deer antler exhibits a series of distinctive tooth marks which may be ascribed to the incisors of Water Vole Arvicola terrestris. The Red Deer bones also showed evidence for human and carnivore activity. Metatarsal bones had been split vertically and extensively gnawed at both ends by domestic dogs or wolves. A Red Deer scapula had been gnawed, probably by domestic dogs, and the spine had been cut away leaving traces of parallel grooving on the blade. The cranium of a stag, otherwise wholly unabraded, had been forcibly shattered, probably to gain access to the brain.

Although some of the people mentioned here are no longer with us my gratitude to them is undiminished. I thank Miss Anne Connolly, Dr Juliet Clutton-Brock, Dr Andrew Currant, Dr Caroline Grigson, Dr Michael Kerney, Dr Eric Robinson, Professor F.W. Shotton FRS, Mr B.W. Sparks, Mr Roger G. B. Williams and Mr John G. Woodford. I had also better thank my wife Joan for tolerating the unremitting mix of pleasure and chaos that goes hand in hand with environmental study.

References:

On 15 May 1983 a polished Neolithic stone axe head was picked up by Mr J. G. Woodford, then of Malvern, close to the edge of the river at Birlingham, Worcestershire (SO9543). The site is only a couple of kilometres from Pensham. It is 115 mm in length and 56 mm wide with a maximum thickness of 26mm and a blade length of 52mm. I visited the site and found old river dredgings worked into the surface of a cultivated field by the river, some teeth of Red Deer and part of the skull of a large domestic dog. The Red Deer teeth are likely to be prehistoric; the dog remains more recent. The axe head was submitted to Professor F.W. Shotton for petrological analysis.

Professor Shotton (in litt., 7 July 1983) gave it the registration number Wo 37/C and confirmed that the rock was of great interest, being a Neogene flow basalt from Western Scotland, or just possibly from Antrim, with feldspar laths all in a preferred direction, not of a type previously recorded as an axe head. Quite probably some of its journey to Birlingham was made as a glacial pebble. The majority of the regional Neolithic axe heads cited by Whitehead (1979), and many of those found subsequently, are fashioned from Cretaceous Flint.