Although our scientific reasoning now appeared to be watertight, I was still nervous that there might be a flaw in our version of European prehistory that even our most persistent and vociferous critics had overlooked. They had done a good job in making us test and prove every conceivable aspect of our principal tool – mitochondrial DNA itself. We had checked and rechecked the mutation rate. We had spent weeks running different versions of our evolutionary network programs and they all gave the same results. We had ridden the storm of recombination. We still felt sure that main chapters of the genetic history of Europe were written in the time of the hunter–gatherers, long before the farmers arrived. To be sure, agriculture had added some important extra paragraphs; but it had definitely not erased the original text. We felt very confident that most living native Europeans traced their maternal ancestry back to the hunter–gatherers who lived before the dawn of the Neolithic and the coming of agriculture.

Nevertheless, even though we were very sure of our data and the way we had interpreted them, our conclusions were still only inferences about past events: inferences built on large amounts of data and robust statistical treatments, but inferences none the less. So I was still slightly anxious. Perhaps we had made a mistake about the dates. I didn’t think we had, but suppose we were out by a factor of two? Suppose that events we had dated to fifty thousand years ago actually took place only twenty-five thousand years ago? More importantly, suppose the dates for the major mitochondrial clusters which we had placet! at or around the end of the last Ice Age, between fifteen and twenty thousand years ago, were out by the same factor and were really less than ten thousand years old? That would bring them too close for comfort to the Neolithic period, and mean that they might have been part of the wave of Near Eastern farmers after all.

What we needed was a direct test on DNA taken from a human fossil which was known to predate the arrival of farming. If we could only find DNA that fitted into one of these crucial clusters in the remains of a hunter who lived thousands of years before farming was ever thought of, then we would be home and dry. We would not need to rely exclusively on reconstructions from the modern sequences. We would have found the real thing in Palaeolithic Europe. These mitochondrial clusters then had to have arrived in Europe thousands of years before farming ever reached it, and our dates must be right. Conversely, if the DNA from a very old fossil was unlike anything we now found in Europe then we were on shaky ground. We could not then be sure that the ancestors of the major modern clusters were in Europe before farming.

Human remains from the Upper Palaeolithic are few and far between. For one thing, ten thousand years is a very long time, and only in the very best of circumstances do bones last that long. Any that do survive are jealously guarded specimens, and rightly so. We would have to make an exceptionally good case to persuade a curator to allow us to take a sample from such a rarity. In my favour I did at least have a track record in getting DNA out of old bones. With my colleagues, I was the first to do so, with the Abingdon bones in 1989, although in that case the material was only a few hundred years old. Our work a few years later on the Iceman had become widely known, and was well thought of. But that was a unique case – a completely frozen body. At five thousand years it was old, but not old enough to predate agriculture. Although the Iceman’s DNA belonged to one of the key clusters, it couldn’t be used to strengthen our case because he was living two thousand years after farming had reached the Alps. We were looking for remains that were at least twice as old as the Iceman. Even so, he was the oldest human by far to have had his DNA successfully extracted, and as a deep-frozen body he was an exceptional case. There was no assurance that an ordinary skeleton would retain its DNA for five thousand years, let alone ten thousand.

Although DNA is obviously a much tougher molecule than anyone ever thought when they were scared to take it out of the refrigerator for fear of its decomposing, it cannot survive very long on its own. It needs to be in a skeleton to survive for thousands of years. What distinguishes bones, and teeth, from all other tissues is the hard, calcium-based mineral, hydroxyapatite. This protects the proteins and the DNA from decay by shutting out the bacteria and fungi that feed on the soft tissue in the rest of the corpse. So long as the mineral is intact, there is a chance that the DNA will have escaped being gobbled up. Once the calcium goes, the DNA is exposed and soon disappears. Calcium is alkaline and survives much better in an alkaline soil than anywhere else. In neutral and particularly in acid soils, DNA is much shorter-lived. The spectacular peat-bog bodies of northern Europe, where even the hair and skin are intact, always have a collapsed and deflated look about them because the calcium in the bones has dissolved in the acid bog. A lot of the protein survives and is protected against decay by the acid, which kills bacteria and fungi. However, because of its molecular structure, DNA is cut to shreds by even dilute acid very quickly. So, unfortunately, bog bodies are not a good source of ancient DNA.

Heat is also bad news. Egyptian mummies were an early, high-profile target for those in search of ancient DNA and, sure enough, some was found. But these were the carefully embalmed bodies of the wealthy, sheltered from decay not only by the natural preservatives in the embalming fluid but also by a succession of wood and stone sarcophagi which sealed the body in an underground tomb away from the baking heat of the sun. There are thousands of much less elaborate burials for the less well off in shallow graves just beneath the sand; but, even though these mummies are only two or three thousand years old, they are almost totally devoid of protein or DNA. The inorganic calcium is unaffected by the heat, but the organic molecules are long gone, broken down and leached away by the scorching heat of the desert.

We knew, then, that we had to avoid burials in hot countries and acid soils, and so we turned our attention to the limestone caves of northern Europe. Within these caves the temperature remains cool and, importantly, constant throughout the year. The daily fluctuations of heat and cold in the Egyptian desert probably do more damage to the DNA than the heat alone. A cool, stable temperature was much more promising. But what really recommends limestone is the alkaline nature of the surroundings. Bone mineral and limestone are chemically very similar. They are both compounds of calcium. The water that drips its way through the caves, forming stalactites and stalagmites, and covering the walls in sheets of flowstone, is rich in dissolved calcium. There is calcium everywhere. A bone left in a limestone cave does not have its mineral leached away. And if the mineral stays, and the temperature isn’t too high, the DNA will stay as well.

The caves in Cheddar Gorge are the most famous in Britain. A small, winding road threads its way down from the top of the Mendip Hills about twenty miles west of Bath. At first it is like any other wooded valley in that part of the world. Ash and hawthorn trees flank the road and, in the spring, the woods are full of the white flowers and pungent smell of wild garlic. As you descend further, the sides of the valley get higher and higher and the trees retreat up the increasingly steep slopes until, only a couple of miles from the top, you are staring up at vast walls of limestone three hundred feet high. Except at the very bottom of the gorge there is no sign of the river which formed it. This disappeared underground long ago, where it dissolved caves and caverns out of the rock. As the roofs collapsed and collapsed again, so the gorge was formed. The newest caves are still there, not yet obliterated by the forces of water and gravity. In the bustling tourist town of Cheddar at the foot of the gorge, the caves are big business alongside the cheese for which the town is famous. On the left hand side of the gorge, directly opposite the Cheddar Caves Fish and Chicken Bar, and with its entrance partially obscured by the Explorer’s Cafe-Bar and a shop, is the biggest and most spectacular cave of them all – Gough’s Cave. And in the museum near the entrance to the cave stands a cast of its most celebrated former inhabitant: Cheddar Man. He was excavated in 1903 and subsequently carbon-dated to about nine thousand years ago, at least three thousand years before farming reached Britain. The cast is a copy of the original skeleton, which is stored in the Natural History Museum in London, in the care of Chris Stringer, head of the Human Origins Group. I rang him and made an appointment.

I knew Chris by reputation and had met him once at a scientific conference in Sardinia. The Natural History Museum I had known since my childhood. It was always a treat for my brother and me to be taken there by my mother in the school holidays. As I made my way up towards the immense and towering Victorian Romanesque entrance I felt a real excitement to be going to the Museum again not as a schoolboy but as a professional scientist. To reach Chris Stringer’s office I had to walk past the skeleton of the huge dinosaur, Diplodocus, that dominates the magnificent entrance hall. Then I turned right into a wide corridor, its walls hung with the skeletons of Ichthyosaurs and other marine reptiles, still embedded in the blue clay of the Dorset cliffs where they were found. But when I went through the door into the palaeontology department, the atmosphere and the decor changed abruptly, from the dramatic to the professional. Row upon row of anonymous sliding cabinets concealed the treasures which lay catalogued within them. Chris Stringer’s modern office led off from this priceless yet strangely silent testament to the wonders of the natural world.

Over a mug of tea, it didn’t take long to explain any reason for wanting to sample human fossils from the Palaeolithic. He had read about the controversy which our work on European prehistory had sparked, and quickly saw the sense in testing the DNA from a pre-farming skeleton. He wanted to know what the chances were of our being able to recover any DNA if he were to give us permission to sample. I could not give a definite answer. After all, the Iceman was so unusual that I could not promise that because we had been successful with him we were guaranteed a good result with an unfrozen bone twice that age. Without that assurance Chris was understandably reluctant to give permission for us to take a destructive sample from something so precious as Cheddar Man. Remembering that we had also been successful with animal bones from the Mary Rose, I made a suggestion that 1 hoped would get us over this impasse. If there were any animal bones from Gough’s Cave of approximately the same age, could we try them? If that worked, we could be fairly confident that the conditions within the cave were good enough to preserve I)NA for ten thousand years. Happily, there were scores of animal remains from Gough’s Cave and I went back to Oxford with a small piece of deer bone.

Within a month I was back in Chris’s office with the good news. There was plenty of DNA in the deer bone. Chris agreed that this was sufficiently good proof to allow me to sample the human material. On the table in his office he carefully laid out the actual remains of Cheddar Man, each one enclosed in a cardboard box and supported by cotton wool. The skull had its own made-to-measure wooden case, with foam rubber supporting the delicate reconstruction from a dozen or more fragments cemented together. I didn’t dare to touch it. Eventually we settled on the hallux, the solid-looking bone of the big toe. Chris packed it into a small cardboard box and I took it back to the lab.

Next day, I carefully drilled into the bone. What appeared from the outside to be solid bone was not. In no time I had punctured the thin shell of the cortex and was into the honeycombed interior. Black specks fell into the small pile of brownish bone powder from the drillings. These black bits certainly didn’t look like bone; most likely they were bits of soil that had found their way into the middle of the toe-bone through a crack. I picked them out one by one with watchmaker’s forceps and put them to one side. I had exactly 17.8 milligrams of Cheddar Man bone powder. It would just have to do; I didn’t want to make another hole. By the following day I knew it was not going to work. There was no sign of any DNA. The control experiments had worked perfectly. Bright orange fluorescent spots, indicating the presence of amplified DNA, were in all the positive controls. The blanks, always run at the same time with water and not bone extract to control for contamination, were all blank. And so was the extract of Cheddar Man’s toe. This was bitterly disappointing.

I went back up to London to talk things over with Chris. We knew from the success with the animal bone that the environment of Gough’s Cave was good enough to preserv e DNA for at least ten thousand years. Maybe the fact that the bones had been outside the cave for the best part of a century had something to do with it. Maybe the resin that was used to stabilize the bones had interfered with the DNA extraction. Or maybe there just wasn’t any DNA there at all. Just so that we could have a focus for our thoughts as much as anything, Chris brought the skull back into his office and laid it out on his desk once more. I don’t find it particularly easy to relate a skull to a living person but, as I looked at the pieces displayed on the desk, I began to imagine the flesh and the skin of the head building up on the reconstructed skull. As I write this it sounds distinctly macabre, but at the time it wasn’t in the least. In my imagination, these were no longer just lifeless fragments of bone but a real person. I had no clear impression of what he looked like – no idea whether he had black or fair hair, brown eyes or blue – but I did have a very strong feeling that this was a person. Strange, remote, from a far-off time, but a person none the less. What stories he could tell about his life, his family. I picked up the lower jaw and looked at his teeth, the teeth he used to crush hazelnuts and tear into the flesh of freshly caught deer. The enamel was worn down, but the teeth were not rotten. In fact, they looked pretty healthy compared to my own set, which are full of fillings. When I idly mentioned this to Chris he turned and said, ‘Well, if you think these are good, come and have a look at this.’ He led me out of his office and into the large room with the storage cabinets. We walked to a distant part of the room and Chris brought out another small wooden box. He opened it and inside, nestling on its bed of foam rubber, was the lower jaw of a younger male. The teeth were absolutely perfect. White, regular and with no sign of decay. They could have come straight out of a toothpaste ad. I imagined they must be only a few hundred years old at the most. But they were not. These were the teeth of a young man who lived more than twelve thousand years ago – over three thousand years before Cheddar Man – and whom Chris had excavated himself from Gough’s Cave in 1986.

Back in the brightly lit office, the teeth looked even better. Could it be that, inside the teeth, the dentine and the pulp cavity would be much better protected even than the bone? Could the few molecules of DNA, which were all we needed to test our theory, be hiding inside the teeth encased in an unbroken shield of enamel? Even though we had failed with Cheddar Man’s toe, we agreed it would be worth a shot. But no-one had any experience of extracting DNA from teeth, especially teeth still embedded in the jaw, and there was no question of being allowed to remove them to make it easier. I promised to go away and devise a method of drilling into teeth in a way that did not mark the enamel and allowed them to remain in the jaw. If I could do that, then Chris would allow me to take a sample from the Gough’s Cave specimen.

I was back within a fortnight, having practised on some teeth given to me by my dentist, Mr Miller. I had perfected a way of drilling into a molar tooth and getting the dentine out’with the tooth still embedded in the jaw, and I brought with me some samples of my handiwork for Chris to inspect. After trying and rejecting a straightforward dental drill (the compressed air blew the powder all over the place) I had found a small modelling drill which had been recommended by a colleague at work and which I bought from an ironmonger on the Tottenham Court Road in London. It was just perfect for making the small entry hole just below the enamel. Once I had got inside the tooth, another, longer drill bit could be attached and wiggled to and fro, reducing the soft dentine to a fine powder. I rigged up a suction device and, using this, it was very easy to remove the powder from inside the tooth into a small test tube. The hole then only needed to be filled with a colour-matching cement and the tooth looked as good as new – as it were. And the dentine, at least in my trial teeth, was full of DNA.

To avoid the ever-present possibility of contamination with modern DNA I needed to drill the teeth from the Cheddar fossils in my own laboratory, where we had recently installed a filtered-air clean room. We had bought it as a ready-made unit constructed for the silicon chip industry. The incoming air was filtered and maintained at a positive pressure, which meant that there was no chance of dust or flakes of skin getting into the room when you went in through the air lock. It was an expensive and elaborate precaution, but well worth it. So I had to take the jaw back with me to Oxford – which was a nightmare. I had come into London on the bus, and it was on the bus that I returned with this priceless and irreplaceable specimen in its box on the seat beside me. Every few seconds I would turn to make sure it was still there, trying to imagine what I could say if I lost it. Thank God, I didn’t; and by late afternoon it was safely locked away in the specimen cabinet back in Oxford.

The next day I started the extraction. It couldn’t have gone better. The drill sank into the second molar easily, but not too easily – that would have been a sign of bad preservation – and there was a slight smell of burning in the air. This was the collagen being vaporized by the speed of the drill, a smell I used to hate during my own visits to the dentist but one I had now come to love as a sign that there was plenty of protein left in the specimen – and where there is protein there is usually DNA. When I switched on the suction pump, the pale cream powder came flying out of the tooth into the tube. There was lots of it – just under 200 milligrams. I took 50 milligrams, so as to leave plenty for a repeat, and started up the extraction process.

By the following evening I knew I had mitochondrial DNA from the tooth. Over the next two weeks I read through the sequence, checked it again and confirmed it by a second extraction. I was looking at the DNA sequence of the oldest human fossil, by far, that had ever been successfully extracted anywhere in the world. But that wasn’t the most important thing. The crucial piece of information we were looking for was embedded in the details of the DNA sequence itself. Was this the same sequence as a thoroughly modern European, or was it an obscure relic that was now extinct?

The answer was crystal clear. The ancient DNA from Gough’s Cave was also completely modern. The sequence lay at the centre of the largest of the seven mitochondrial clusters. It is by far the commonest sequence in modern Europe; and here we had found it in the tooth of a young man who had lived fully seven thousand years before the arrival of farming in Britain. Here was the proof that this sequence, this cluster, and, by extension, the others of a similar estimated age were well and truly established in Europe long before the farmers. The Upper Palaeolithic gene pool had not been fatally diluted by the Middle Eastern farmers. There was more of the hunter in us than anyone had thought.

Though I had got no further than drilling into his big toe, this was not the last encounter I had with Cheddar Man. We were re-introduced, so to speak, as part of a television documentary. Philip Priestley, an independent producer, was setting up a series of archaeology-based programmes for a west country TV station, and one of them was built around the excavation of a Saxon palace in Cheddar. By now our work on the genetic continuity between the Palaeolithic and the present day was reasonably well known, and it occurred to Philip that it would make good television if he could relate, through DNA, some of the present-day residents of the town with Cheddar Man himself. This seemed both fun and worthwhile; but I explained that we had already had a go at getting DNA from the Cheddar Man remains without success. If he could get permission from Chris Stringer, I was willing to try again, this time with the teeth, not the toe-bone, but only on condition that if nothing came of it we would not be filmed. I always work on that basis. 1 have seen too many programmes that begin with a big build-up anticipating a great scientific discovery at the finale, only to peter out in an inconclusive or unsuccessful experiment. So, with all the ground rules agreed, and after another nerve-racking journey on the bus, this time with an even more famous fossil in a box beside me, I drilled into Cheddar Man’s first molar.

Out came the powder – not quite as clean as the earlier Gough’s Cave material, but in sufficient quantity for an extraction. We found enough DNA for a reasonable sequence and were not surprised when it fitted comfortably into one of the seven clusters. Philip, understandably nervous as the deadline for filming got closer, was delighted and immediately organized the second strand of the piece, the sampling of the Cheddar residents. The site of the Saxon palace, featured in another programme in the series, is in the grounds of the local secondary school, and it made good sense to approach the school to see if they would agree to their pupils taking part in the programme. By now we had refined our DNA sampling procedure. We no longer used blood samples; instead we found that a small brush rubbed gently against the inside of the cheek picks up enough cells from the surface to give us plenty of DNA. After a short visit to the school, we had twenty samples from the sixth form volunteers and some of the teachers. Knowing how often we had found Cheddar Man’s sequence in modern Britain, I reckoned there was a fifty–fifty chance of getting a close match in the twenty samples we had taken. Within four days we had the results. We knew the names, and (crucially, as it turned out) the ages of the volunteers. Philip was on the phone.

‘We’ve got a match,’ I told him.

‘Who is it?’ was his first question.

This wasn’t part of the deal. While we had agreed to see if we could find a match among the twenty residents, I had not agreed to identify any individuals, for a very good reason. Although the children, and their parents, had signed forms consenting to have their DNA sampled and to take part in the television programme, I felt there was a risk that they might not have realized what they were letting themselves in for if the story broke in a big way. Though there is no way of knowing beforehand how big a story is going to become, the experience of Marie Moseley and the Iceman was an indication of its potential.

At this point Philip became distinctly agitated. He thought the story would be worthless without an individual identification. He immediately faxed me a copy of the consent form, but as far as 1 could see it was just a standard release – not, in my opinion, sufficient as a basis on which to claim consent to a possible worldwide media intrusion into the life of a teenager. I checked our list of sequences against the names and ages of the volunteers. There was not one match but three: two exact matches with Cheddar Man, and one with a single mutation; and while the two exact matches were children, the close match was a teacher, in fact the head of history who was organizing the filming in the school, Adrian Targett. I made the decision that I would identify Adrian Targett but not the two children. As it turned out, it was one of the best decisions I ever made. Unknown to me, Philip and his publicity team had organized a public ‘reveal’ where Adrian Targett would be identified as Cheddar Man’s relative in front of the cameras and in the presence of a television news crew. They, too, were beginning to sense the potential magnitude of the story. The next day, when I went to the newsagent, I could not believe my eyes. The story of Adrian Targett and Cheddar Man was in all the papers: from the London Times to the tabloid Daily Star, there was Adrian on the front page, posing beside his famous fossil relative. I bought the lot.

In the following days and weeks the story of Cheddar Man spread around the world. I met Adrian Targett on a TV chat show. He told me how one tabloid newspaper, famous for its pictures of topless women, had offered him a five-figure sum (so at least £10,000) to pose in a fur loincloth beside his ancient relative. Being a sensible man, conscious of his standing as a teacher, he declined. But it did make me wonder what the newspaper would have offered a teenage girl to wear the same outfit – or less. Even now, years later, people still remember the Cheddar Man story, if not always accurately. I was talking to an American audience in 2000 on something completely different when a woman asked me: ‘Are you the one who did the DNA from the Cheese Man?’ At the time, not surprisingly, I had a full postbag for weeks after the story broke. Many letters were complimentary, including a very well-informed one from the inmates of San Quentin gaol in California, who were keen to discuss the findings at the next meeting of their anthropology study group. But the one that stood out came from the secretary to Lord Bath. It turned out that Cheddar Caves are part of Lord Bath’s estate. Evidently he had read the story (though whether in the Times or the Daily Star I never discovered) and wanted to know if he too was related to Cheddar Man.

Alexander Thynn, Lord Bath, is the owner of Longleat, one of the most beautiful houses in England. It is famous for the safari park in the grounds, where visitors can watch the famous Longleat lions and other dangerous animals from the alleged safety of a car. Lord Bath himself, affectionately referred to as the Loins of Longleat, is well known for his idiosyncratic personal life. In addition to a legitimate wife and children, he has a stable of what he calls his ‘wifelets’, many of whom live on the estate. This was definitely one to follow up, and the next weekend I was on the way to Wiltshire. I was led upstairs to the penthouse suite on the top floor of this magnificent Elizabethan house. Lord Bath himself, now in his sixties but with a youthful twinkle in his eye, was dressed in one of his collection of brightly coloured kaftans that bulged from a wardrobe close to an absolutely enormous wooden desk. The life clearly suited him. He poured out two large glasses of rose from a tap on the wall as I went through the genetics with him. A few glasses later we got round to the test itself, and he brushed the inside of his cheek. During the course of the morning several other people passed through the penthouse, and each was encouraged to give a sample, which they cheerfully did. He was evidently very popular with his staff. By lunchtime we had half a dozen DNA brushes and it was time for me to leave.

When we got the results back it came as no surprise that Lord Bath was not closely related to Cheddar Man. There was no particular reason why he should be. But his butler, Cuthbert, one of the other people who had given a sample during my visit to Longleat, was an exact match. At a stroke he could claim an ancestry which stretched back nine thousand years, making the five-hundred-year pedigree of the Thynns look distinctly nouveau. I asked Lord Bath how Cuthbert had received this piece of news. Had it made him reassess his attitude to the aristocracy? ‘Well,’ he replied with a smile, ‘he has been feeling very confident lately.’

We had now done about as much as we could to establish our claim that the maternal ancestors of the majority of modern Europeans were already living in Europe well before the arrival of farming. We could not say anything about other genes, only about mitochondrial DNA; but on this basis we had a clear picture of European prehistory, built up from both modern and fossil DNA, not of a massive replacement of the hunter– gatherers by the farmers but of a strong continuity back to the days of the Palaeolithic. There was only one of Cavalli-Sforza’s criticisms that we could not answer. Whatever way you look at it, mitochondrial DNA is only one gene and, as such, subject to statistical fluctuations that might make it unrepresentative of the human genetic legacy as a whole. I did not think this very likely; but what was needed to substantiate our version of European prehistory was confirmation from another gene altogether.