Issue no 41, February 1999

New dating methods are shedding light on the mysteries of hill-figures, writes Paul Newman

Why did people carve the shapes of enormous men and horses into the chalk hillsides of southern England? In ancient times, one assumes these figures had a religious or totemistic signifance, connected to a cultural ethos that we can only guess at now by following scattered clues.

Yet the craze continued well into the modern era. Why did people do it? Perhaps they could think of no better way to spend their time. Our communal pleasures, after all, like gala days, sports days and carnivals, have a kind of exuberant senselessness about them. The cutting of a hill-figure similarly demands a lively co-ordinated effort. Periods of intense digging and dumping chalk waste are followed by intervals where beer, sandwiches and laughter are passed around.

In creating such an artefact, a community is engraving its signature on the land. Folk in Wiltshire and Dorset can look upon their white horses knowing that an ancestor helped carve it. And for what? - a shared memory, a legacy that will gallop down the centuries and still invoke the capacity to wonder?

In recent times, archaeology has turned its attention to these figures. They are no longer seen as isolated gestures but as integral to the landscape. The small Iron Age enclosure above the Cerne Giant's head is said to bear a direct relationship to the figure, as is St Augustine's Well at the foot of Giant Hill. Similarly the Uffington Horse is coupled with the flat-topped spur of Dragon Hill and the chambered mound of Wayland's Smithy. The Long Man has been related to Windhover burial mound and other features in the Neolithic landscape. This method is both informative and fallible. In any town centre, a medieval church can be seen nudging up to a building society office or a public convenience. But they are unintentional juxtapositions whose sole relationship is that of proximity. Probably the most renowned carving is the Uffington Horse in Oxfordshire. It is, artistically speaking, a triumph of imaginative omission. The lines that define it are landscape lines; they curve and melt into the greens and browns of scarp, dip and glacial terrace. White Horse Hill is crowned by Uffington Castle, a double-walled hillfort dating from the 7th century BC. Many think that its occupants carved the Horse as a tribal totem or ensign.

The status of the Horse as a mysterious dateless artefact, not provably Bronze Age, Iron Age or Saxon, was challenged in 1995 by Optical Stimulated Luminescence or OSL. The technique of silt-dating relies on periods of neglect when the trenched outlines of hill-figures become overgrown and fill with soil. It is essential to locate an early cut or trench, so that a layer of `original' soil can be tested to determine when it was last exposed to sunlight.

With this in mind, an old cutting was opened by the Oxford Archaeological Unit in the Horses's beak. Successive layers of `beaks' - some over a metre longer than the present projection - were traced, and a trench through the body showed that it had once been a metre or so wider, but never strikingly different from its present design. The angle of the body had changed, climbing the hill over the centuries, so it now occupied the flat upper slope and was less visible from a distance. Samples taken from between two of the lower layers of the body, and from another cut near the base, produced three dates of approximately 1400-600BC, indicating a Late Bronze Age origin.

When published, the figures raised a few eyebrows. The Iron Age origins of the Horse had become official. To some, the Horse seems so typical of La Tène art that a Bronze Age dating de-stabilised some widely held cultural assumptions.

Another well-known horse is at Westbury in Wiltshire. This is a slightly wooden and formal-looking horse, like a docile creature staring at you over a farmer's fence. But its setting, on the slope of Bratton Camp, is impressive and tradition ascribes it to King Alfred's victory over the Danes at Ethandune. The Horse is, if you like, a badge of victory. But the present creature is the result of drastic cosmetic surgery performed in 1778 on an older, sag-bellied beast, of pathetic appearance, with a tail resembling a dolphin's and a large ringed eye. So now the debate still rages: how old was the original beast? Was it a cousin of the weird white horse of Uffington?

None of the other Wiltshire horses presents a historical enigma. The elegant horse at Alton Barnes was cut by a tenant of Manor Farm, Mr Robert Pile, in 1812. The apologetic-looking Marlborough Horse was cut by schoolboys in 1804, and the giraffe-necked, originally glass-eyed Cherhill charger by Dr Christopher Alsop in 1780. Standing about two hundreds yards from the top of Labour-in-Vain hill, he boomed instructions through a megaphone to his workforce, ordering them to move around until a credibly equine shape was achieved and then the turf was pared away.

Even more than the horses, hill-figures of human form capture the imagination. Of these the Cerne Giant of Dorset is the most famous, on account of his swaggering immodesty. An enlargement occurred when the navel became overgrown and was, in the subsequent scouring, mistakenly identified as the tip of the penis. About three years ago, I attended a conference at Cerne in which academics pored over the chalk shaft of the 30ft penis, charting any enlargement or diminishment down the centuries. Poets attended the event, too, exposing many a raunchy stanza to the naked ear.

Many scholars regard the Giant as a Romano-British portrait of Hercules. Others disagree. The Bristol historian Ronald Hutton argues that if the Giant is pre-medieval, why is there no allusion to him in medieval documents relating to Cerne Abbey? The earliest reference occurs in the parish records of 1694 in which the sum of `3 shillings' is put aside for `cutting' or restoring the outline. Hutton goes on to suggest that the Giant may be a 17th century fake. The Restoration was renowned for filth and frolics - what better example than a naked giant impudently disporting himself! Unfortunately, Hutton's complaint about the lack of documentation strangulates his own thesis. If a prominent landowner created such a gigantic, comical obscenity in the 17th century, why do we find no mention in any of the thousands of diaries, letters and gossipy broadsheets of the period?

Evidence from soil probes indicates what we see today is not the complete portrait. Various irregularities beneath the left wrist suggest the Giant might be holding something. This is the thesis of historian Rodney Castleden, who conducted a resistivity test on the figure a few years ago. The readings were fed into a computer and, eventually, a pattern swelled up on the monitor - namely, a severed head with dangling dreadlocks and death-set eyes. The Cerne Giant may therefore be a naked fighting-man brandishing his spoils dating from as early as 500BC.

But no sooner does this challenging interpretation emerge than another appears. The latest - advanced by the historian Joe Bettey - identifies the Giant as a 17th century lampoon of Oliver Cromwell. He claims to have unearthed some correspondence of the Dorset historian, the Reverend Hutchins, in which this suggestion is raised.

The Long Man of Wilmington, Sussex, offers even fewer clues to his identity. Only his staves vaguely speak of something: a traveller, a sage, a magus - all these suggestions have been advanced. Like most of the hill-figures, he is vaguely evocative of many things in general but nothing in particular. His staves are 231 and 235 feet high, and it has been claimed that he is Europe's largest representation of the human form (the sexless giant of the Millennium Dome will be a pygmy by comparison).

Until recently the earliest known sketch was made by Sir William Burrell in 1776, showing a clothed, shambling figure holding a rake in his right hand and a scythe in his left. However, in 1993, a new drawing was found on a map at Chatsworth House dating from 1710. This sketches a slightly flabby figure with a conical head and bulges where his ears should be. Eyes, nose and mouth are marked; kneecaps and pectorals are even hinted at faintly. The posture holds a hint of challenge or confrontation. There is more of the warrior about him than the farmer or haymaker and the staves are significantly longer.

In the view of the late Christopher Hawkes (1967), the rake and scythe were the remnants of Christian crosses imposed on the original spearheads of Odin in much the same way as early medieval saints made their marks on the standing stones of prehistory. He believed that a resemblance to the Long Man was found in the Finglesham belt-buckle showing a Viking warrior in ceremonial attire.

In a visionary interpretation, however, Rodney Castleden has restored the Neolithic case. He sees the Long Man as the sun-god opening the dawn portals and letting the ripening light flood through.

I have summarised several theories because they underline a basic point about these carvings. Hill-figures possess inexhaustible symbolic significance. They refuse to be fixed to a century or epoch or specific intention. They trigger the romantic dreamer in us all. Some scholars glory in this confusion. But many think it is no bad thing that the technology is available to narrow the time band. In the next few years, I have little doubt that OSL will probe the anatomies of other figures, and I for one will enjoy having the clay washed from my eyes as long-concealed silts yield their secrets.

A new edition of Paul Newman's book on hill-figures, Lost Gods of Albion, was recently published by Sutton at £19.99

Modern diagnosis of ancient disease

Sophisticated techniques are now used to extract the secrets of ancient illness, explains Tony Waldron

The occurrence of disease in human remains has attracted the interest of archaeologists and anthropologists for over a century and many diseases have been found to have a long history, including osteoarthritis and a number of infectious diseases; while evidence of trauma seems as old as man himself.

Until recently, palaeopathologists have relied on their own observations of the changes taking place in the skeleton, aided where possible by x-ray evidence. In the past few years, however, several advances have taken place largely in the wake of the observation that the organic matrix of bone survives much better after death than had previously been supposed. This has led to a new area of investigation into what are now called ancient biomolecules - molecules which have survived sufficiently well to be extracted from human remains.

Until a few years ago it was believed that bones recovered from excavation consisted entirely of the inorganic matrix - the bone crystal. But then some intrepid investigators attempted to extract DNA from bones, and to everyone's surprise (probably including their own) they succeeded. At first it seemed possible only to extract mitochondrial DNA (mDNA), a form of DNA found in the mitochondria within cells and exclusively inherited down the maternal line. Although it can be used to determine such matters as racial affinity, it can tell nothing about inherited diseases or genetic mutations which are passed down via the chromosomes.

In the past two or three years, however, genomic DNA - DNA derived from the chromosomes, inherited from both parents - has been recovered. Using this it is possible, for example, to determine the sex of skeletons. This is particularly important in the case of infants and juveniles, since there is almost no other way reliably to do so.

Using genetic techniques, it has also proved possible to pinpoint the genetic abnormality responsible for causing a rare disease, porphyria, in a skeleton belonging to one of the descendants of George III, which gives great weight to the suggestion that he himself had this condition, and that it was responsible for his episodic madness.

Encouraged by these observations, attempts were made to extract bacterial DNA from mummies and skeletons thought to have had infectious diseases during life. The first successful extraction was made from the spine of a juvenile skeleton from Turkey, dating to the Byzantine period, which had lesions suggestive of tuberculosis. This was done using the polymerase chain reaction (PCR), by which the short fragments of ancient DNA which survive in bones can be amplified and identified by comparison with samples of modern DNA. DNA from the tubercle bacillus was also later recovered from a Peruvian mummy and from several other skeletons.

The bacterium which causes leprosy is closely related to the tubercle bacillus, and DNA from this organism has been recovered from a skeleton with the typical bony signs of the disease. This skeleton was excavated from the grounds of the monastery of St John the Baptist, on the spot on the River Jordan where St John was thought to have baptised Christ.

This work has a number of important implications. Firstly, it allows for a positive diagnosis of infectious disease in the past, and there is no doubt that DNA from other bacteria will be isolated in due course.

It will be of great interest to see whether DNA from bacteria which commonly cause gut disease can be detected - Salmonella spp for example. DNA from these organisms could be recovered from cess pits or perhaps from coprolites (ancient excrement). The use of PCR will also allow palaeopathologists to diagnose diseases such as plague or malaria - both the subjects of current research - which leave no traces on the skeleton and so cannot be diagnosed by `traditional' methods.

PCR could also provide an answer to the debate about the origin of human tuberculosis. Humans are affected by two forms of tubercle bacillus, the bovine form and the true human. They have different clinical courses and it is generally supposed that humans first contracted bovine tuberculosis from close contact with domesticated cattle - although they may have contracted it from other animals such as deer - and that the human bacillus evolved from the bovine. Since the DNA of the two forms differs, it should be possible to distinguish between the bovine and human forms in skeletons, allowing us to determine the earliest date at which the human bacillus appears; if it were to be contemporaneous with the bovine form at all periods, then, clearly, it is unlikely to have evolved from it.

In many diseases, antigens are produced to bacteria or multicellular parasites, and in theory these antigens should be capable of survival and extraction. Similarly, in a number of malignant diseases, various abnormal proteins are produced and these so-called biomarkers are used in clinical practice to diagnose disease and monitor the progress of treatment.

In multiple myelomatosis, a fatal malignant disease of plasma cells in the bone marrow which affects elderly individuals, large amounts of an abnormal protein are formed. There is some skeletal evidence that this disease occurred in the past but it is difficult to differentiate from other kinds of malignant disease in which there is diffuse secondary spread of the tumour to bone - cancer of the breast, for example.

Now, myeloma protein has been extracted from the skeleton of an elderly woman who was buried in a medieval cemetery in Germany, the first biomarker ever to be recovered from ancient human remains. This opens the way for researchers to try to extract other biomarkers, including perhaps prostatic specific antigen (found in large amounts in prostatic cancer), HLAB27 (related to ankylosing spondylitis) and rheumatoid factor (seen in patients with rheumatoid arthritis).

Another project utilising ancient biomolecules is the investigation of the epidemiology of schistosomiasis in ancient Egypt. This disease is caused by parasitic worms which lay eggs that are passed in the urine or faeces. During life, antigens are produced to these eggs and the diagnosis of schistosomiasis will be made by extracting the soluble egg antigen from mummy tissues.

The work on ancient biomolecules is certainly the most exciting development in palaeopathology for many years but the more `traditional' kind of palaeopathology has not been without its successes either.

Among rheumatologists there has been a widespread belief that rheumatoid arthritis is a relatively recent disease. The first clinical description did not appear until 1800 and the disease seems now to be on the decline in this country. This has prompted the suggestion that some environmental factor was present during the past two centuries, producing the disease in those genetically predisposed to it, and that this factor - whatever it is - is now disappearing. Until very recently, there had been no cases described in skeletons in this country, which seemed to support the notion.

In 1994, however, the first case was described in the skeleton of an elderly female from a medieval cemetery in London. As usually happens, once one case is reported, others quickly follow, and cases have now been described in Anglo-Saxon skeletons, as well as in skeletons from the Viking period in Sweden and from the Roman period in France. Rheumatoid arthritis has also been found in a skeleton from the Jomon period in Japan, dating to 2,000-3,000 years ago. This shows with certainty that we are not dealing with a new disease.

Another suggestion relating to rheumatoid arthritis was that it may have come to Europe from the New World. There seems to be an unusually high prevalence of the disease in native American skeletons, but the finding of the disease in European skeletons dating to well before contact with the New World lays this hypothesis firmly to rest. It seems likely that Europeans cannot blame the New World for syphilis either, since several instances of skeletons with the skeletal stigmata of this disease have been found recently from pre-Columbian periods in Europe.

Osteoarthritis is by far the most common disease to affect the skeleton. Skeletal evidence shows that, in the past, the disease shared many of the characteristics found nowadays - it is age dependent, for example, and slightly more common in women than men. Recent research, however, has shown that it appears to have changed its character since the medieval period. Before about 1500, osteoarthritis of the hip was more common than of the knee, whereas the converse is the case now. Moreover, osteoarthritis of the medial compartment of the knee - which is the most common form seen in clinical practice - may be of very recent origin, arising perhaps no more than 200 years ago.

Osteoarthritis of the hand also seems to have changed. In the medieval period, the tendency was for only a single joint of the hand to be affected, whereas since then it is much more common for multiple joints to be involved. The reason for these changes is not known, but probably relates either to changes in activity or in diet and nutrition; further work should help to decide which is the more important.

Dr Tony Waldron is a palaeopathologist at the Institute of Archaeology, University College London

You can recreate a Classical garden wherever you live. Linda Farrar describes how

The Romans are known to have enjoyed all the luxuries in life, and what could be more pleasant on a warm summer's day than reclining on a couch sipping wine under a vine-covered pergola?

This and other idyllic garden scenes were described by Pliny the Younger in his Letters, and depicted in numerous wall-paintings. Now archaeology is beginning to trace out the remains of Roman gardens, and to show how important the garden was to Roman life from Britain to North Africa.

Gardens were created in town and country, and although there was more space and a greater freedom of design in rural areas, many garden features in towns reflect the villa way of life on a smaller scale. In towns such as Dougga in Tunisia, even properties with just a tiny courtyard were decorated with plant troughs, reflecting a powerful desire to domesticate nature. Sunken urban gardens are known, as well as gardens on terraces. Gardens were also created for public use, and their surroundings enhanced many a public monument. An example is the portico garden discovered by the theatre at Mérida (Spain) which has now been replanted and is believed to mirror the famous Portico of Pompey in Rome.

In towns the garden was usually enclosed by walls or colonnades, and in the Late Republic and Early Empire was generally sited to the rear of the house. Many later houses were constructed around the garden, a design that allowed the open space to serve as a light-well for a number of rooms. Important reception/dining rooms opened onto the vista of greenery, and in many cases specially prized garden features would be placed in full view of diners. Garden sculpture would include urns, herms (a bust on a post), pinakes (marble relief panels on a short pillar), oscilla (small suspended pieces of marble relief), figures of garden deities and animals, and sundials. Shrines and altars were also sited in the open, or protected in a niche or grotto, and on feast days these would have been decorated with garlands.

Water played an important role in Roman gardens and where there was an adequate supply a pool or a fountain was installed. Of the seven principal types found across the Empire, three have been found on sites in Britain - the basic rectangular type, the rectangle with square or semi-circular niches in the sides (as at the Governor's palace in London), and the semi-circle (as in mansiones at Wall in Staffordshire and Wallsend in Tyne and Wear, the latter sometimes mistakenly regarded as a plunge-pool connected to the bath-house).

The Romans introduced their customs to the many peoples they conquered and the possession of a garden seems to be part

of this process of Romanization. We find that gardens in the Mediterranean counties are very similar to those of Rome and Pompeii. However, in the hot-test regions like North Africa precautions were taken to preserve water - in the use, for example, of small water basins rather than larger pools - and in the northern provinces we assume differences in plant-use and in the extent of exposed pipework and delicate statuary vulnerable to frost.

Garden archaeology is a relatively new field and has seldom been used for the Roman period in Britain. Exceptions are at Fishbourne Palace in Sussex, Frocester Court Villa in Gloucestershire and Bancroft Villa near Milton Keynes. The architecturally inspired bedding trenches for topiary hedging found at Fishbourne are well known, but rectangular plant beds have also been discovered at Bancroft and Frocester. In the latter, independent archaeologist Eddie Price with the Gloucester and District Archaeological Research Group identified spade cuts in the subsoil, as well as a discarded iron cutting blade from a wooden spade.

Landscaped terrace gardens are now suspected at a site near Swindon (see BA, June), and nymphaea or fountains have been identified at Chedworth Villa (Glos) and Caerleon (Gwent). There, the garden was in the public portico of the baths. Water flowed through a figure of a dolphin (this belonged to a group with Venus but only the dolphin survives), and cascaded down a series of steps into an elongated pool. The large nymphaeum that housed this statue recalls the larger version at Welschbillig in Germany, which was really a garden pavilion sited next to a huge ornamental pool (65m long) that was surrounded by a decorative low wall which used herms as posts.

Landscaped grounds of villas could also contain belvederes (viewpoints), dovecotes, arbours and little rest rooms, all of which are mentioned in ancient texts, but apart from those at the imperial villa of Hadrian at Tivoli, and the ones found in the towns of Campania affected by the Vesuvius eruption in AD79, they have rarely been identified. Alignments of post pits could however indicate a vine-covered trellis or a row of espalier trees, as were found at Fishbourne. Aerial photography and geophysical surveys could reveal further details such as buried fishponds.

At one villa site partly excavated recently by the Association for Roman Archaeology in the Og valley in Wiltshire, aerial photography indicated that a pergola or a path lined with statue bases may have led down to the river. The use of ephemeral items such as wooden trellis fences is, however, largely assumed from ancient wall-paintings, which also provide clues to Roman garden plants.

Enthusiastic Romanists may wish to try to recreate a Roman-style garden of their own. Many Roman gardens were small, so size should be no barrier. Roman gardens were usually approached through a portico, and the addition of a pergola, colonnade or loggia to the rear of a house with a south-facing garden would provide an ideal alternative to a conservatory.

Otherwise a pergola covered by vines or ivy, or even roses, could be placed on one side of the garden. If it was sited near a terrace, garden furniture could be brought out for Bacchanalian al fresco feasting - the hanging grapes would really set the scene.

A special feature should be incorporated in the garden to provide a focal point; possibilities include a recess for a shrine, a statue, an urn or a fountain. Venus was the ancient protectress of the garden, with her son Priapus, and her presence would ensure the fertility and well-being of plants grown there. The phallic figure of Priapus might be rather crude for today's taste but a figure of a satyr or Bacchus would not be out of place. A Bacchus figure was found at the Spoonley Wood villa in Gloucestershire.

Paths should be made of pebbles or stonechip and could be lined with herms as at the Villa of Oplontis near Herculaneum or at Tivoli. If space permits, a fishpond can be added, but as the Romans preferred formal designs it should be a symmetrical shape. Roman gardens sometimes employed trompe-l'oeil wall-paintings of rural scenes on garden walls, to give an impression of increased space, and this is another feature that might be considered.

The Romans are thought to have been responsible for introducing many new species of plants to Britain - examples include box, plum, damson, cherry, walnut, mulberry, vine, leek, garlic, onion, fennel, parsley, turnip, cabbage and rose - but we cannot assume that the modern varieties are quite the same species as existed 2,000 years ago as there have been innumerable hybrids made and new varieties discovered. Occasionally discoveries of macrofossils and pollen in excavations can give direct clues to Roman plants - the pollen of rose, for example, has been found at Silchester, while clippings of box are known from Fishbourne and Frocester.

The Romans were fond of topiary and trimmed bushes, such as box, bay, laurustinus and myrtle. Colour contrast could be achieved with juniper, or the feathery grey-leaved southernwood which, during a recent Italian-led excavation in one garden at Pompeii, was found to alternate with rose. The glossy-leaved acanthus mollis would add a statuesque element to any bedding plan. The Romans also liked scented plants and species that could be used to make garlands and floral crowns, such as rose, lily, violet, narcissus, parsley and many of the herbs.

A fruit tree could also be included, as these were particularly welcome then as now, and species such as the quince are so scented that fruits were brought in on a dish expressly to scent the interior of a Roman house. The Romans also experimented and refined grafting techniques to such an extent that they produced two or more different fruits on one tree, in what we call today a family tree. This perhaps emphasizes the point that the Romans knew a thing or two about gardens, and as we tend our little plots today we could do worse than look back for inspiration to the simple but luxurious designs of the Classical world.

Linda Farrar is an ancient historian with a special interest in gardens. Her book, Ancient Roman Gardens, was published in December (Sutton, £25.00)