By Anne Eastham
Introduction.
Overlooking the Ritec valley at Penally, near Tenby, two caves, Hoyle’s Mouth and Little Hoyle, sometimes known as Longbury Bank cave, penetrate the north facing slopes of the limestone embankments. Both have been of intermittent interest to antiquarians and archaeologists since before the middle of the nineteenth century. These are just two of a number of cave sites in Pembrokeshire that have revealed traces of human presence during the Late Upper Palaeolithic Final Palaeolithic and Mesolithic periods. Most of the evidence for human activity comes from the stone tools recovered during excavation, occasionally there are the marks made by the same flint tools on the bones of the animal they killed and butchered and even more rarely the skeletal remains of the people themselves.
Within Pembrokeshire a site, like Priory Farm cave, Pembroke, was found to contain a mixture of Upper Palaeolithic tool types and, close by, on what is now Caldey Island, a group of caves, Nanna’s cave, Ogof-yr-Ychen, New cave and Potter’s cave, all contain Late Glacial and early Post Glacial deposits and stone tool artefacts, indicating that there was quite a group of human hunter-gatherer settlements overlooking the expanse of grassy plains and marshlands that stretched as far as the Gower during the periods pre- and post the Last Glacial maximum. 1 Another group of cave occupation sites on the Gower, including the celebrated Paviland cave, overlooked a similar habitat reaching out towards the Severn estuary, then much reduced, because of the lower sea level.
Within this general pattern, the human and animal remains at Hoyle’s Mouth and Little Hoyle belong either to just before or just after the maximum of the Last Glaciation. This paper aims to demonstrate something of their significance for the environments of the time through a study of the bird bones and the niche ecology of each species recovered on the two sites in relation to the intermittent hunter-gatherer occupation. 14C (carbon 14) dating on animal bone within the main deposition levels of the breccia underlying an upper layer of stalagmite, and below the breccia a silt layer, gave results of between 12,000 and 10,220 years BP (horse) for the breccia and 26,200-26,600 (reindeer) on bone from the silt layer. This would place the occupation periods, either prior to 22,000 or post 12,000 years BP, interspersed by the period of re-advance of icy conditions known as the Younger Dryas from around 11, 000 until around 10,000 years BP, when the site became somewhat inhospitable. Later, there was an economic change in the Pembrokeshire prehistoric record towards a pattern of Mesolithic subsistence.
The nineteenth century finds of unstratified woolly rhinoceros bone and an Aurignacian busked burin pointed to some level of occupation of the cave by humans and, or hyaena during the Middle Devensian c.30,000 years BP.
Hoyle’s Mouth site study.
Plan of Hoyle’s Mouth Cave, showing the location of excavated trenches (after S. Aldhouse Green, 1996)
The earliest recorded work on the cave was by Col. Jervis in 1840. He was followed by the Rev. G.N. Smith (1863), the Rev. H.H. Winwood (1865) W.Boyd Dawkins (1874) and E.L. Jones in 1882. The disturbance caused by these explorations and the need to confirm the contexts of the artefacts disclosed by them prompted the rescue excavations by Savory in 1973. He explored the areas outside and within the cave entrance trenches 1, 2 and 3 and apparently recovered very few bones of birds.
The paucity of avian bone in this sector was confirmed during the years 1986 -1990 by the work of Aldhouse Green. Excavations at the entrance of the cave included the entrance chamber, trench 1, and the recess, trench 2, followed by trench 3. No avian bone was recovered from Trench 2, a few bones survived from trench 1, level 2 at depths of 35-75 centimetres. They represent the remains of wildfowl, which would be consistent with the late Pleistocene bird fauna at Little Hoyle and two passerine bones, indicating a wintering thrush and a starling. The three bird bones from trench 3 may be more equivocal: two from the SW quadrant and one recovered during cleaning from the SE quadrant at a depth of over 1 metre. This last, a robin, Erithacus rubecula, might be accepted as a late glacial specimen but may equally have been deposited during the Holocene or very recent times.
Excavations during the years 1990 – 1996 were carried out in the interior sectors at Trench 4 in the East fissure, Trench 5, the Bear chamber and Trench 6, the Reindeer chamber.
Trench 4, the East fissure, was rich in avian bone with a diversity of species. Most of the bone came from layer 1 and the sample was dominated by passerine species, whose bones appear extremely fresh and lack patination or soil derived colouration. Apart from a vertebra that was probably from a domestic fowl, all the small passerines, from the distribution of skeletal elements, appear as though they might have been washed into the fissure from the outside at some time during the Holocene, or even quite recently. The list includes
Branta leucopsis barnacle goose *
Cf. Tadorna sp.shelduck species *Starred species have a dark patina on the surface.
Anas platyrhyncos mallard *
Cf. Gallus gallus domestic fowl
Lagopus lagopus willow grouse *
Anthus cf pratensis with another Anthus sp. meadow pipit or similar
Troglodytes troglodytes wren
Erithacus rubecula robin
Turdus merula blackbird
Turdus iliacus redwing
Turdus philomelos song thrush
Sturnus vulgaris starling
Passer domestica house sparrow
Carduelis carduelis linnet
Emberiza cf. cirlus bunting cf cirl
The number of bones from any single species is small. Apart from the robin, Erithacus rubecula, only a pipit, possibly a meadow pipit Anthus pratensis, an indicator of open grassland on the slopes above the cave, may be represented by more than a single individual. It seems doubtful whether they were deposited either by anthropogenic intervention or by a carnivore like a fox, that would have consumed the entire carcase and there are certainly no signs of digestion on the bone. It is possible that the ten bones of a robin all belonged to a single individual. Territorial competition is very strong in this species and one bird will pursue a rival to its death in order to preserve its own patch. A recent visit to the cave has shown a further accumulation of bone and other debris at this point, including the skull of a small thrush and a collection of acorns. It seems that this section, still within the zone of daylight penetration, has accumulated and continues to accumulate vegetation and animal material over quite a long period of time. The single vertebra that may have belonged to domestic fowl is almost certainly a recent intrusion, perhaps the lunch remains of a nineteenth century excavator.
There were, however three bones from level 2, 2a and Unit 4 of trench 4 that carry a quite different patination: HM2513, a sternum fragment of barnacle goose, Branta leucopsis, that was stained dark brown in a way that is similar to the goose bones at Little Hoyle; a willow grouse, Lagopus lagopus ulna [HM 2576]; and a mallard, Anas platyrhyncos, coracoid, [HM 2755]. These seem to form a separate group and may possibly have belonged to an earlier deposition phase.
The Bear and Reindeer chambers, Trenches 5 and 6, present a different situation.
The range of species in the Bear chamber, Trench 5, over 30 metres beyond the daylight zone within the cave, demonstrates the rich diversity of habitat niches available in the Ritec valley at that period.
The list includes:
Ardea cinerea, grey heron
Anser cf. erythropus lesser whitefront goose
Branta leucopsis, barnacle goose
Tadorna tadorna shelduck
Anas platyrhyncos, mallard
Anas acuta pintail
Anas penelope/ strepera gadwall/ wigeon
Anas crecca, teal
Cf. Aquila sp. ? eagle
Charadrius dubius little ringed plover
Tringa totanus redshank
Cf. Scolopax rusticola woodcock
Sterna sp. cf. Chlidonias niger tern species, cf. black winged tern
Cf. Tyto alba, medium sized raptor cf. barn owl
Sturnus vulgaris starling
Carduelis sp. cf. cannabina/flavirostris linnet/twite
Most numerous in this trench were the bones of mallard, Anas platyrhyncos, but there were other species of dabbling ducks and marginal feeders, waders and the occasional small passerine and raptor. It indicates the presence of quite a complex avifauna that equates closely with other assemblages of dates around 12,000 BP from south west Britain, the Rhineland, Belgium and Aquitaine. Unfortunately, for a few of the species, the only evidence we have is a phalange from the foot. These are reasonably reliable indicators but not totally definitive, though useful as a guide. One such is the heron, a bird that appears only rarely in the Palaeolithic record. It was noted at the Abri Dufaure in the Landes of south west France [c.10,000 years BP], also in a Magdalenian level at Hohler fels beischelklingen, Baden Wurtemberg, West Germany and in Ireland in the Holocene deposits of Ballymintra cave, co. Wexford. 2 Other wading birds, the redshank and little ringed plover, are species that feed on the margins of lakes and shorelines. The wildfowl, the geese, barnacle and lesser white fronted and the dabbling ducks, Anatidae and the shelduck are all dependent on the availability of open water and their presence suggests that the present day ponding of the Ritec valley was more extensive at the time. This probability is further reinforced, as will be noted below, by the recovery of several goose species at Little Hoyle.
Trench 6, the Reindeer Chamber
By contrast, very little in the way of bird remains was recovered from trench 6 in the Reindeer chamber but the few were all bones of duck, mallard and teal, once again showing the importance of the Ritec and its valley in the subsistence patterns of the late glacial occupation. Flint artefacts in this deposit show that there was a human presence some 40 metres into the cave, even though the animal remains may have been introduced by mammalian carnivores like hyaenas. Nineteenth century disturbance may have affected this sector of the cave, as a date of 1817, engraved on the wall of the chamber, asserts.
Little Hoyle, or Longbury Bank Cave
Introduction
Like Hoyle’s Mouth, Little Hoyle has been explored and excavated by antiquarians and archaeologists since the second half of the nineteenth century and at all stages bird bone was noted among the finds. Research was begun in 1866 by the Reverend Winwood and then the Reverend G.N. Smith, followed by Edward Laws, Wilmot Power and Professor Rolleston and the majority of the material from their excavations, along with the itemised description of the finds depicted in the watercolour of E.L.Jones, remain in Tenby Museum. 3 Amongst the bone from the eastern chamber mentioned on the drawing are black cock, Tetrao tetrix, cormorant, Phalocrocorax phalocrocorax, gull, Larus sp. and ‘Eagle’. The latter has more recently been identified as common crane, Grus grus and will be referred to again later.
In 1958 and 1959 McBurney 4 undertook excavations on the North Platform of the cave talus, mainly in Trench 1, and recovered a large amount of material together with a significant number of bird bones from a diversity of species. The stratigraphy of the bone is insufficiently clear for any definite correlation between his findings and the results of subsequent excavations on the North Platform and in the interior of the cave by Aldhouse Green between 1984 and 1990. 5 However, ecologically, the two assemblages are not incompatible.
Later excavations in the interior of the cave revealed another rich deposit of goose remains from a range of species. Originally these were considered to have been the remains of birds culled from a breeding colony of barnacle geese, Branta leucopsis, during the Spring and Summer season. The phenomenon was even quoted in the local press. A revised determination indicated that several species of the Anser family were also represented. During the breeding season each of these occupies a different habitat in the arctic and would therefore have arrived together in the Ritec valley, not to reproduce, but on migration or as a winter refuge.
Site study
North Cave, Trench 5
The list of bird bones from all levels in Trench 5 consisted almost entirely of geese:
Anser brachyrhyncos pink footed goose
Anser albifrons white fronted goose
Anser erythropus Lesser white front goose
Perdix perdix grey partridge
Larus fuscus lesser black backed gull
Sturnus vulgaris starling
Almost all the avian bone from Trench 5 in the cave interior belongs to geese and amongst them four species are represented, pink foot, white front, lesser white front and barnacle. The barnacle goose is present in by far the larger number of bones and variety of skeletal elements, but the representation besides of three species of Anser is of considerable significance for the ecology of the site at the time and for its seasonal use by the Upper Palaeolithic occupants of the cave. Most of the bone from layer 2 and 2b in this context was described as coming from a stratified level of green silt with limestone.
Out of a total of 122 identified goose bones from the two parts of levels 2 and 3 of Trench 5, 97 could be determined as barnacle goose but there were also a few examples, some 25 in all, that were bones of members of the Anser goose family. The morphological distinctions were verified in detail with aid of micro-photography. The only other species from this level were woodpigeon, of which was found a fragment of pelvis and a tibiotarsus and a single tarsometatarsus of starling. All the bone from this trench was dark in colour with apparently a similar patina to some of Hoyle’s Mouth sample.
The bone assemblage from level 3 of trench 5, as level 2, contains mainly geese, though they were fewer in number than in level 2, a total of only 10 bones, 8 of which were from a barnacle goose. The only other species were a grey partridge, and a gull humerus, probably a lesser black backed gull. 14C dating taken from a barnacle goose bone in level 2 of trench 5 gave a date of 22.800 +/-300 years BP, prior to the early onset of the Last Glaciation. The presence of collared lemming, Dicrostonyx torquatus, indicates that conditions were becoming less temperate for human occupation, at least during the winter.
North Platform Trench 1
A large part of trench 1 was excavated by McBurney in 1959 and, as already noted, the relative sequence of layers is not entirely clear. This is unfortunate because there is a considerable diversity of avian species amongst them, many more than were revealed by re-excavation of the trench years later and some may well belong to the Holocene, or even quite recent deposits. If the lists of the species most likely to have been predated by carnivorous mammal or anthropogenic action are compared the discrepancy between the two stages becomes apparent:
Avian bone 1959 Avian bone 1986 and 1990
Anser anser grey lag goose/domestic
Anser albifrons/brachyrhyncos white front or pink foot goose
Anser erythropus lesser white front goose
Branta leucopsis barnacle goose
Anas penelope wigeon Anas penelope wigeon
Anas platyrhyncos mallard
Lagopus lagopus willow grouse
Tetrao tetrix black grouse Tetrao tetrix black grouse
Perdix pedix grey partridge
Scolpax rusticola woodcock Tringa sp. cf. sandpiper
There were, in addition, a number of bones of domestic fowl, that were almost certainly post Roman and may even have come in from the 1959 excavation and with them a number of small passerine species that may, like those from the East Fissure in Hoyle’s Mouth, have been introduced in a variety of ways, perhaps unrelated to human or other mammalian intervention.
North Platform, Trench 4
Although part of it was disturbed material from the 1959 excavations, Trench 4 yielded the clearest sequence of dateable levels on the North Platform. Of the birds, there were some 18 species represented, the majority from the layer 3 of the upper scree, with a probable date of around 9.980 +/- 800 years BP, meaning that they belong to a deposit at the tail end of the climatic changes following the Last Glaciation . The basal silt layers below the screes contained a smaller quantity of avian material, whose date, based on samples taken from reindeer bone, was around 23,340 +/- 350 years BP, from prior to the Last Glacial maximum, although there may have been indications of deteriorating climate. A comparison between the bones from the scree and the silt layers demonstrates the degree of change.
Scree, layer 3 Silt layer 4
Puffinus puffinus, manx shearwater *
Anser albifrons, whitefront goose
Branta leucopsis, barnacle goose
Anas platyrhyncos, mallard
?Haliaeetus albiclla, white tailed eagle
Accipiter nisus, sparrowhawk
Falco tinnunculus Kestrel
Lagopus lagopus, willow grouse
Terao tetrix, black grouse
Perdix perdix, grey partridge Perdix perdix, grey partridge
Alca torda, guillemot *
Alauda arvensis, skylark Alauda arvensis, skylark
Turdus, thrushes: blackbird/ring ouzel, redwing
Sturnus vulgaris, starling
Pyrrhocorax pyrrhocorax, chough Pyrrhocorax pyrrhocorax, chough
Corvus corone, crow
Despite the apparent species diversity shown in this list, it amounts to very few individuals, a single bird, perhaps of any one species. And even though the sea level was beginning to rise 10,000 years ago the appearance of a manx shearwater (starred), that is entirely pelagic outside the breeding season, and in any case spends most of its life on the wing, is something of a surprise. Superficially also, the guillemot would be considered an oddity. However, it is a frequent resident in inshore waters during the winter and the presence of waterfowl at all levels in both Hoyle’s Mouth and Little Hoyle suggests that the Ritec valley enclosed a considerable expanse of water.
Among the group of raptorial birds the single claw of the white tailed eagle poses another curiosity. A claw is hardly a diagnostic element but in this case it is larger than golden eagle and it is the most likely of the larger raptors to appear on Britain at this period. Bones were found at Church hole in Derbyshire and at Soldier’s hole in Somerset from around 12,000 years BP and appeared at Cathole cave on the Gower in a context related to the Middle Devensian. Finds of this fishing eagle persist into the Neolithic, where they occur in association with tombs in northern Scotland and Orkney, like that of Isbister, where the motivation and symbolism of their cull remains undecided in the realms of funerary practice. There is no doubt that this claw matches perfectly the skeletal remains in the Tankerness museum at Kirkwall,Orkney. The other two small raptors, sparrowhawk and kestrel, will have had little to do with any human activity in and out of the cave unless they hung around to profit from small birds and rodents who would have gathered around the debris left by a hunter-gatherer or hyaena occupation.
The South Cave, Trench 6
Since Little Hoyle penetrates through the Longbury Bank cliff by a narrow fissure, a trench was made to explore the south entrance. Although some mammal bone, notably hyaena was recovered, there were only two avian bones, both of black grouse.
Tenby Museum Collection of nineteenth century finds.
Many of the archaeological recoveries from Little Hoyle are held in Tenby Museum, together with the watercolour commissioned by Laws and Rolleston from E.L.Jones. Each recorded find from Roman or Medieval pottery to bone fragment is noted in its approximate position and a note on the drawing indicates:
The entrance to the eastern chamber in which were found: human remains, waterworn hammer, flint flakes and bones of ox, goat, dog, brown bear, badger, fox, polecat, cat, eagle, blackcock, gull and cormorant.
The cormorant is actually a shag and, besides those mentioned, the collection includes other specifically maritime species, herring and greater black backed gull and the common scoter, usually pelagic during the winter. Otherwise, a ‘partridge’ bone becomes a kestrel and there were also barnacle geese, just as they were found on both sites during later investigations.
The ‘eagle’ is interesting. It is indeed a large bone, a coracoid, an element that links the upper wing, humerus with the sternum and scapula. But it belonged to a common crane, Grus grus, and is significant for the interpretation of the ecology of the Ritec valley in the past. Currently, the preferred summer habitat in northern Europe is between the tundra, taiga and steppe zones, often around marshland and swampy wetland or damp meadows, migrating by stages to southern Spain or North Africa in winter. Conceivably confusing the crane with the heron, Gerald of Wales described large flocks of a hundred or more when he visited Ireland in 1185 but they have not bred in Britain since 1600. In Pembrokeshire, occasional pairs have been observed over several weeks during the winter in recent years, presumably when blown off course on migration. They are infrequent in the Pleistocene and early Holocene record of Western Europe and sparse in Britain. Remains from Thatcham in Berkshire were dated to between 10 and 9,000 years BP; traces of them were found embedded in the silts of the Severn Estuary and bone found in the Glastonbury lake dwellings and in the lake dwellings of Ireland.
The Enigma of the Geese
It has already been mentioned that the original identification work on the goose bones found only one species, the barnacle goose and from this arose the idea of there being a breeding colony in the Ritec valley prior and post the last Glacial Maximum, the periods either side of the coldest conditions when the southern land mass of Britain and northern Europe became once more just about habitable. This assumption has had to be revised for two reasons. The first is the morphology of the bones and the second is the ecology and distribution behaviour of the different species of geese.
Goose families in northern Europe are divided between the Anser geese and the Branta. The shape of the bones of the wing differ perceptibly in the details of their construction between the two groups. The distinction between members of the Anser family is mainly to be found in size variation, with very small adaptive differences. The adaptive separation between the character of Anser bones and those of Branta geese are considerably more marked and show up particularly well using micro-photographic procedures. The basic skeletal research done in the University of Munich in the 1960’s was, and is, of considerable assistance to the archaeozoologist in identifying fragmentary material but recent computer aided photographic methods have improved its reliability. In consequence, it has been possible to claim a lesser white front goose, Anser erythropus, and a barnacle goose, Branta leucopsis, as being present at Hoyle’s Mouth and small numbers of three species, pink foot, lesser white front and white front, A. brachyrhyncos, A. erythropus and A. albifrons with larger numbers of barnacle geese, Branta leucopsis, in the North Cave and Platform at Little Hoyle.
The other factor is the geographical and habitat diversity of breeding location of each of the species. The breeding distribution of the Greenland pink foot is concentrated on the Atlantic side of the Western Palaearctic. Colonies breed in central Iceland, east Greenland and west Spitzbergen, using snow free hummocks and rocky ledges as protection against predators and floods. The breeding grounds of the white fronts is considerably further east in the arctic tundra of Russia and Novaya Zemlya and over into north Greenland, migrating in winter either towards Mesopotamia or western Britain. The lesser whitefront, apart from being considerably smaller than other Anser geese prefers more southerly breeding sites in shrubby areas close to wooded tundra in Fennoscandia and Russia. The inaccessible rocky cliff ledges of the high arctic and its islands provide the nest sites for the barnacle geese, which, unlike the other species are gregarious at all times of year.
The aggregation locations for all these species are, therefore, not on the breeding sites but on the staging points and winter feeding grounds, where they will congregate in groups feeding on pastures and marshland vegetation. Based on a count of left side and right side skeletal elements, it would appear that the number of individual birds was never large. 87 barnacle goose bones from one level of the North Cave probably represent not more than 9 birds with possible additions from other parts of the section. This cannot be considered a major cull of wildfowl congregating in the valley, more a subsistence use of a temporary resource.
Hunting
In order to decide whether the predator was man or animal, marks of butchery on the bones give a lead. Mammalian attack may take the form of tooth marks and there are tooth marks on large animal bone from both caves. At the Middle Palaeolithic cave of Coygan at Laugharne, even human bone was found gnawed by hyaenas. The mammal report on Hoyle’s Mouth suggests that much of the large animal bone was attacked by carnivorous mammals. In many ways birds are most easily dismembered by tearing the joints apart and this seems to have been the normal method. Indeed, at Little Hoyle there are a few signs of human hunter-gatherer intervention in the form of cuts and scratches on the goose bones from the North Cave made by flint tools some four or so were noted on bones of barnacle and some of the other geese, suggesting that the wildfowler was probably human. There were also some indications that a few bones may have been subjected to fire. Sadly, however there is no actual evidence that bone was subsequently fashioned into tools or the feather re-used. Hunting methods appear to have changed very little over time. The same techniques of trapping, netting and even decoys seem to have been used from the Upper Palaeolithic until, with elaborations, the later Middle Ages and the invention of the shotgun. Trapping using nets was a likely method to have been employed by the hunter-gatherers here, combined perhaps with driving the birds into traps, or even during the later period in the transition towards the Mesolithic, by the use of dogs or trained geese to lead the wild birds into simple decoy tunnels. Nets or decoys would have been efficient methods for taking wildfowl without disturbing the remainder of the flock and images identified as drawings of nets appear on the walls of decorated Palaeolithic caves in France and elsewhere.
Conclusions
We tend to find that human hunter-gatherer groups used caves for shelter throughout the period of Palaeolithic, Old Stone Age cultures, particularly during the winter season and the evidence of any early occupation of Hoyle’s Mouth and Little Hoyle confirms this view. Avian seasonal indicators are predominantly towards the autumn, winter and early spring months, when many species besides waterfowl would pass through or take refuge in slightly more temperate climatic zones. Similar patterns of faunal distribution and exploitation at this period may be found in different parts of Britain.
The nearest comparisons may be found on sites that also overlooked the broad expanse of the Severn valley, whose nomadic settlers exploited the animal and avian resources that abounded amongst its plains, marshes and forests. Soldier’s Hole, Gough’s Cave and Gough’s Old Cave and Chelm’s Combe all belong to the later occupation phase and the birds they hunted were similar. The same is true of the much longer period of intermittent occupation at Pinhole Cave in Derbyshire. Continental parallels are equally consistent. Cave sites in Belgium and the Rhineland of Germany indicate that wintering wildfowl were universally exploited and, depending on the situation, game birds also, especially grouse, as found at the Penally caves. From France there is evidence of a much more southerly distribution of winter migrants like geese, swans and other aquatic species than may be seen at the present day. In the region of southern Aquitaine, cave and rock shelter sites, supporting hunter-gatherer groups towards the end of the Last Glaciation, contained considerable numbers of wintering wildfowl, including bewick’s and whooper swan and a variety of geese. Scrape and cut marks on the bones revealed that not only were some bones of the wings used to make tools and more rarely musical instruments but also that the valuable plumage was used, for clothing perhaps or decoration. Though the evidence has not been discovered to date, we may imagine that avian resources of every kind must have had a role in the cultural life of the hunter-gatherer people of south Pembrokeshire.
Notes
1. David, A. 1991 Late Glacial archaeological residues from Wales, a selection. In, Barton, N., Roberts, A.J. and Roe, D.A. (eds.) The Late glacial in North West Europe: human adaptation and environmental change at the end of the Pleistocene. London: Council for British Archaeology, Research Report No. 77, 141-159.
2. Eastham, A.S. 1995. L’Ecology Avienne. In, Straus, L.G. Derniers chasseurs du monde Pyrénéen, l’Abri Dufaure: un gisement tardiglaciare en Gascogne. Société Préhistorique Française, No.22, 219-234. Lambrecht, K. 1933. Handbuch der Paläornithologie, Berlin.
3. Laws E. 1885. In Arnett J.E., Mason’s guide to Tenby and the surrounding district. Publisher ?
4. McBurney C.B.M. 1959. Report on the first season’s fieldwork on British Upper Palaeolithic cave deposits. Proceedings of the Prehistoric Society No 25, 260 – 269.
5. Aldhouse Green, S.R.H. 1996. Hoyle’s Mouth and Little Hoyle Caves. Archaeology in Wales, No 36, 70-71.
Acknowledgements. Grateful thanks to Elizabeth Walker at the National Museum of Wales and the staff of Tenby Museum for their help and encouragement and to Michael Eastham for his patient assistance with digital images.