'Will all great Neptune's ocean wash this blood

Clean from my hand? No, this my hand will rather

The multitudinous seas incarnadine,

Making the green one red.'

Macbeth, II, ii

B

lood. It's the key to life. Without it, we die. It's the thread that runsthrough history, transferring property and power from one generation to thenext. From the earliest times, man has understood blood both as a tribal markerand as an individual blazon. In some societies, inheritance flowed not fromfather to son, but from father to sister's son, because you could be sure thatyour sister's son was of the same blood as you. You knew for a fact hisgrandmother was your mother; you couldn't be certain your own sons shared yourblood.

It's also been at the beating heart of crime fiction from the beginning. WhenDoctor Watson first lays eyes on Sherlock Holmes, he is bent over a tableperfecting a test for haemoglobin. Watson's slowness in grasping the test'sbrilliance makes the consulting detective fume. 'Why, man, it is the mostpractical medico-legal discovery for years. Don't you see that it gives us aninfallible test for bloodstains. Come over here now!' Then he stabs a needleinto his own finger and uses the resulting drop of blood to show the test inaction.

'Criminal cases are continually hinging upon this one point,' he declares. 'Aman is suspected of a crime months perhaps after it has been committed. Hislinen or clothes are examined, and brownish stains discovered upon them. Arethey blood stains, or mud stains, or rust stains, or fruit stains? That is aquestion which has puzzled many an expert, and why? Because there was noreliable test. Now we have the Sherlock Holmes' test, and there will no longerbe any difficulty.'

The very title of Arthur Conan Doyle's first novel, A Study in Scarlet, comesfrom Holmes's lecture to Watson on the meaning of detective work. 'There's thescarlet thread of murder running through the colourless skein of life, and ourduty is to unravel it, and isolate it, and expose every inch of it.' When,shortly afterwards, the pair discover a 'scarlet thread' beginning in a lonelyhouse off the Brixton Road, Watson is nearly sick at the scene, which seemsfrankly improbable given he's a medical practitioner who has served in theAfghan Wars. But then, I am a writer whose work features blood and gore and yetI am squeamish about blood.

But back to the book. A man has been stabbed in the side as he lay in bed andthe blade has pierced his heart. 'From under the door there curled a little redribbon of blood, which had meandered across the passage and formed a littlepool along the skirting at the other side.' This time, there is no need for hisnew test; instead Holmes assimilates all the physical evidence in the house,and listens to a policeman's take on the anonymous assassin. 'He must havestayed in the room some little time after the murder, for we found bloodstainedwater in the basin, where he had washed his hands, and marks on the sheetswhere he had deliberately wiped his knife.'

Reconstructing past events from blood found at a crime scene is known asBloodstain Pattern Analysis. Conan Doyle's imagination barely touched the edgesof what spilled blood can tell modern experts. Two years before A Study inScarlet was published, Eduard Piotrowski, an assistant at the Institute forForensic Medicine in Poland, took the first steps in the discipline when hewrote a paper about interpreting bloodstains to explain the course of violentaction itself, 'Concerning the Origin, Shape, Direction and Distribution of theBloodstains Following Head Wounds Caused by Blows' (1895).

Piotrowski put a live rabbit in front of a paper wall, smashed it over the headwith a hammer, and got an artist to paint the gory result. The colour illustrationsin the paper are as accurate as they are grisly. He battered other rabbits todeath using rocks and hatchets, varying his position and angle of attack to seehow it affected the shape and position of the bloodstains. We can't know how hefelt during the experiments, but in his paper he expressed a nobility ofpurpose: 'It is of the highest importance to the field of forensic evidence togive the fullest attention to bloodstains found at the scene of a crime becausethey can throw light on a murder and provide an explanation for the essentialmoments of the incident.'

Nevertheless, Piotrowski's pioneering work gained little attention until themid-twentieth century. In a key case in 1955 a handsome doctor named SamuelSheppard was convicted of bludgeoning his pregnant wife to death in the bedroomof their home on the shore of Lake Erie in Ohio. He maintained that a'bushy-haired intruder' had attacked his wife (and the back of his own neck,which was injured in a way that would have been extremely difficult toself-inflict).

At his trial, and again at his retrial in 1966, forensic scientist Paul Kirk ofthe University of California, Berkeley, testified for the defence: 'When aweapon hits a bloody head, the blood flies out like the spokes of a wheel,radially, in all directions.' Kirk showed photographs to the court of a blankarea on the wall to the side of the bed where the killer had stood and batteredMrs Sheppard. 'It is entirely certain,' he said, 'that the murderer receivedblood on his person, and no portion of his clothing that was exposed could havebeen exempt from bloodstaining.' When police had first arrived at the houseSheppard was shirtless and in a state of shock. The only bloodstain they foundon him was on the knee of his trousers. He couldn't remember how he had come tobe shirtless: 'Maybe the man I saw needed one. I don't know.' A torn t-shirt ofSheppard's size was later found near the house, with no blood on it. Kirk'sconvincing testimony at the retrial helped overturn Sheppard's conviction. Hewalked free after eleven years in jail.

From top left: Samuel Sheppard after the alleged attack, his wife Marilyn ReeseSheppard and Sheppard testifying at his trial in a neck brace. He served tenyears of a sentence for second-degree murder; he was found not guilty at asecond trial in 1966

Five years later the US government published the first modern handbook onbloodstain analysis, Flight Characteristics and Stain Patterns of Human Blood(1971). The handbook, and its sixty colour photographs, showed CSIs thatbloodstaining could reveal how and where a fatal blow was delivered, the kindof weapon used, the likely bloodstaining on the murderer, whether the murdererbled, too, whether they moved the victim post-mortem, or whether the victimthemselves moved before dying.

The police still use blood spatter analysis every day: to date, it has helpedsolve thousands of crimes. But the seismic change in the significance ofbloodstains came in the 1980s with the discovery of genetic fingerprinting. Thequestion of 'who' could now be added to the list of 'what', 'where' and 'how'.Since the early twentieth century, scientists had been able to identify theblood type of a suspect from a sample of blood or semen. Though this was usefulin narrowing down the pool of potential suspects, the frequency with which someblood types occur in the general population meant it could usually only be usedas circumstantial evidence. Blood typing was a far cry from the forensicpossibilities offered by DNA.

For thirty-two years Val Tomlinson has been investigating bloodstains at murderscenes, and analysing DNA in laboratories, first with the British ForensicScience Service (FSS) from 1982 until it closed in 2011, and since then withLGC Forensics. She is a mild-mannered, genial woman whose appearance belies herintimate relationship with blood – the way it moves, its inner chemicalstructure, the messages it carries with it – and her profound understanding ofthe genetic codes that underpin every human life. 'There is a logic to DNA.Scene work is more an art than a science, bizarrely.'

Dr Paul Kirk examines blood spatter on Marilyn Sheppard's pillow

By the time Val arrives at a murder scene with a pad of blank paper under herarm, CSIs have usually photographed and videoed every square inch. 'Many a timeI've had a debate with the bobby on the door who goes, "Why are you drawing it,Ms Tomlinson? There's no need."' But – like an artist painting a landscape –Val wants to completely immerse herself in the scene. 'I can take 200photographs on holiday and when I come home they're just snapshots. But if Istand and draw the scene I'm drawn to specific aspects of it. Very slowly Ibuild up a picture and eliminate irrelevant things. All of the items might be completelyirrelevant except one and I can highlight that. A photograph just showseverything that's on the table. There is no emphasis on, say, an item that'sturned over, a bloodstained item, a coffee mug.'

Once Val has been in a scene 'for five or six hours' she has ordered it, madeit logical. Thus the act of drawing is more important than the drawing itself.'Even if I haven't got all the answers I can at least give a discourse aboutwhat I've seen and the possible sequence of events.' She imparts this narrativeto the Senior Investigating Officer and then later to the court, where she usesher scene drawings 'probably just as much as photographs, because the jury canget hold of them, be taken away from all the things that might distract them inthe room, and brought into what matters.'

More than anything else at the crime scene, blood matters to Val. Like anyother liquid, its dynamics are subject to the laws of physics. If it strikes afloor at right angles it produces a circular stain, often because it hasdripped slowly from a person or object. If it travels at an angle it willproduce an elliptical stain, usually caused by a punch or a blunt weapon. Thelonger and thinner the ellipse the more acute the angle of impact. If a groupof bloodstains on a surface radiate 'like the spokes of a wheel' they probablycame from a blow (or blows) inflicted in one place. A blood spatter expert likeVal can calculate the angle of impact of the stains, then attach a piece ofstring to each one and spool it back at the appropriate angle. The strings willconverge at the point where the blow was made. So if, for example, the point ofconvergence is located close to the floor, the victim could not have beenstanding when they were struck. Photographs of this 'stringing model' can thenbe used in court. And, increasingly, angles of spatter impact can also be putinto a computer program, such as 'No More Strings', to make a 3D model of blowsgiven at a crime scene.

Cause of death isn't always a mystery: at a scene of battery or stabbing it canbe pretty obvious, in which case the SIO might find Val's analysis sheds morelight on the incident than the pathologist's post mortem. Is the blood spatterconfined to one area, showing that the victim dropped immediately to the floor?Did he stay on his feet and put up a fight, in which case blood might havedripped down his clothes? Did the murderer drag the body for some reason,causing the hair to spread out backwards or the clothing to ruck up, perhapsspreading a trail of blood across the floor? Are the ankles crossed over,indicating that the body has been turned over? The answers to these questionscan give the SIO useful information about the actions of the suspect and eventssurrounding the death of the victim.

SIOs want Val to tell them as quickly as possible how bloodstained the suspectis likely to be. 'The last one I went to, there was an awful lot ofbloodstaining at an old Victorian house with lots of rooms. You could see theway the assailant had gone out because every doorway had smears of blood downwhere the clothing had touched it. Ultimately it turned out they'd burnt theclothing but that was recovered and it was still bloodstained.'

The police are racing against the clock to find the suspect before they disposeof vital evidence. But bloodstaining – like much physical evidence – can besurprisingly hard to get rid of. Val is sometimes called away from the scene ofcrime to the home of a suspect to examine doors and clothes. 'Often they've hada clean up, so we look at the contents of the washing machine.' Forensicscientists don't give up on evidence easily, something John Gardiner found outto his cost when he tried to dispose of vital evidence in the manslaughter ofhis wife in 2004 (see p.174).

But blood analysts can't always report so usefully, especially when they aren'tgiven five or six hours to form an artistic relationship with the crime scene.'I've heard horror stories of scientists going to scenes and being told, "Iwant you to look at the blood pattern over there and that's it,"' Val admits.'To me, that's a disaster waiting to happen. We need to be part of the wholepicture.' In some cases analysts testify in court without having visited thescene at all, as a happened in a tragic and complicated case that began on 15February 1997, in the coastal town of Hastings, East Sussex.

In the late afternoon 13-year-old Billie-Jo was putting a lick of paint on thepatio doors of her foster parents' home. Siôn Jenkins, her foster father anddeputy head of a nearby school, returned from a trip to a local DIY store withtwo of his own daughters. One girl walked round to the patio to talk toBillie-Jo, and let out a scream. Billie-Jo was lying on her front with her headcaved in. Siôn pushed her shoulder up to get a better look at her face, and sawa bloody bubble appear at her nostril, which then popped. He called 999, andparamedics pronounced Billie-Jo dead at the scene.

CSIs found a bloodied metal tent peg near the patio, measuring 46 cm by 1.5 cm.The autopsy showed that the attacker had inflicted at least ten ferocious blowsto Billie-Jo's skull. The following day a bloodstain analyst came to examinethe scene and found radiating spatter on the wall next to the patio, the innersurface of the patio doors and the dining room floor.

When a child dies in suspicious circumstances, the police often begin to lookvery carefully at those closest to them. Siôn Jenkins' clothes and the tent pegwere sent to the FSS for analysis. On 22 February scientists discovered 158tiny blood spatters on his trousers, jacket and shoes – too small for the nakedeye to see. Were the spatters there because Jenkins had battered his daughter?Or did Billie-Jo breathe a mist of blood on to him with her dying breath?

A number of days after the murder the bloodstain analyst concluded that theblood on Jenkins' clothes was consistent with him being the attacker, butcouldn't be certain there wasn't another explanation.

The police arrested Jenkins on 24 February and his trial began on 3 June. Ascientist instructed by the prosecution had made bubbles with a blood-filledpipette and burst them next to a white surface. The 'pop' produced a finespatter that travelled downwards and sideways up to 50 cm – but no spatter roseupwards. Next he filled a pig's head with blood and beat it with the same typeof tent peg found near Billie-Jo. This left a fine spatter on his overalls.

A scientist instructed by the defence had done some of his own experiments. Heput some of his own blood in his nose and exhaled over a white piece of paperan arm's length away. He also found a fine spatter.

The prosecution argued that Billie-Jo was already dead when Jenkins pushed hershoulder to see her, and so couldn't have made a breath. Paediatrician DavidSouthall testified, 'Anybody approaching a child with an injury who is gaspingwould be in no doubt whatsoever that the child was breathing and still aliveand would report that because it would be so obvious to an observer.' However,neuroscientists had not reached a consensus on exactly when a brain was tooinjured to cause the respiratory system to produce one more breath.Pathologists for the defence thought that Billie-Jo could have survived longenough to exhale on to her foster father. Under cross-examination the twobloodstain analysts testifying as part of the defence case agreed that thespatter on Jenkins' clothes may have come from the impact of the tent peg.

Siôn Jenkins continued to protest his innocence, but was convicted of murder on2 July 1998 and was sentenced to life imprisonment. Some rejoiced at theverdict. Others were shocked at how little evidence it was based on, believingthat the police had relied too heavily on the assumption that the murderer waslikely to come from within the family. In the previous two years there had beeneighty-five reports of prowlers and suspicious characters near the Jenkins'home in Hastings. The New Statesman railed at the conviction, claiming that'The police had a redhot suspect: someone with a psychiatric history and aknown record of violence towards children, whom a number of people sawloitering nearby on the afternoon of the murder. When the police went tointerview him he seemed strangely to have disposed of most of his clothing.Whoever the true murderer is, he now has the opportunity, as a result of thevagaries of British justice, to kill someone else's daughter.'

When Siôn Jenkins appealed against his conviction in 2004, the pathologistinstructed by the defence presented new evidence about the state of Billie-Jo'slungs. The original autopsy had found them to be hyperinflated, which meantthat something (probably blood) was preventing some air escaping. Thepathologist suggested that if the blockage was in the upper airways it couldsuddenly have been released and caused the spatter on his clothes whetherBillie-Jo was dead or alive. Two retrials followed, both of which ended in thejuries failing to reach a verdict, and, in 2006, Jenkins was acquitted. In July2011 he got a PhD in Criminology from the University of Portsmouth. Now heworks with pressure groups trying, among other things, to ensure that expertswho appear in court are properly experienced and impartial. The real murdererof Billie-Jo Jenkins has never been found.

In 1984, Alec Jeffreys was in his lab at the University of Leicester when heexperienced a 'eureka moment'. He had been checking X-rays of a DNA experimentcomparing members of his technician's family: looking at the results, it wasimmediately obvious that he had stumbled upon a technique which could revealthe unique variations in the DNA of any individual. Since this chancediscovery, DNA profiling (or genetic fingerprinting, as it is sometimes called)has become the 'gold standard' of forensic science. When Sherlock Holmes dreamtup his test for haemoglobin he could proudly state, 'It appears to act as wellwhether the blood is old or new. Had this test been invented, there arehundreds of men now walking the earth who would long ago have paid the penaltyof their crimes.' Within one hundred years of those words being published, realdetectives would be able to know whose blood they had found at a crime scene.Such knowledge might indicate guilt or, just as importantly, make a compellingcase for innocence. For example, if blood found at a rape scene does not belongto the victim or the suspect then, at the very least, you are looking foranother person, someone who might have vital information – or who might be thereal culprit. In the USA alone 314 people who were languishing in jail, some onDeath Row, have been exonerated because of new DNA evidence.

Genetic fingerprinting astonishes people even more than physical fingerprintingdid at the turn of the nineteenth century. In the public imagination it standstriumphantly astride other physical evidence. Forensic scientist Angus Marshallremembers 'a legendary case in the States where a jury came back to the judgeand said, "We're not going to accept blood splatter evidence, we want to seeDNA." They were practically dealing with a confession but they still didn'tbelieve it. It was ludicrous.'

As this suggests, DNA profiling has not always been seen as a purely positivedevelopment. But when Alec Jeffreys was asked on the twentyfifth anniversary ofhis discovery whether genetic fingerprinting was now being used in a way he wasno longer proud of, he replied, 'Catching large numbers of criminals,exonerating the innocent – some of whom have spent more than thirty years injail – immigrant families reunited ... I would argue the good heavily outweighsthe bad.'

To understand the pros and cons of genetic fingerprinting we need to revisitthe first crime it helped to solve, in the tranquil and ancient village ofNarborough in Leicestershire. On 22 November 1983, the body of 15-yearold LyndaMann was found strangled and raped near a footpath. She was naked from thewaist down and her face was bloodied. Biologists established that a semensample taken from her body belonged to someone with Type A blood, and aparticular type of enzyme secretion, a combination shared by only 10 per centof men. But with little else to go on, the case went cold.

Three years later, on 31 July 1986, Dawn Ashworth, also aged fifteen, wentmissing. Her body was found near to Lynda's, just off Ten Pound Lane. Again,she had been strangled, raped and left naked from the waist down.

The main suspect was Richard Buckland, a 17-year-old hospital porter withlearning disabilities. Buckland had a troubled past and had been spotted nearthe scene of the crime. When interviewed, he revealed details about Dawn'smurder, and about her body, that weren't publicly available. Before long heconfessed to her murder. But he vehemently denied killing Lynda three yearsearlier.

Convinced that the same man had murdered both girls, police approached AlecJeffreys at the University of Leicester, five miles from Narborough, who hadrecently appeared in a local news story about 'genetic fingerprints'. Hisanalysis of the semen samples revealed that the police were right: the same manhad committed both murders, but he was not Richard Buckland. Despite hisconfession, Buckland was exonerated – the first person to be proven innocentbased on DNA evidence.

The police now had the genetic fingerprint of the killer but they had losttheir only suspect. They asked all 5,000 adult men in Narborough and thesurrounding villages to volunteer blood or saliva samples. Of the 10 per centwith the particular blood type taken from Lynda and Dawn's bodies, Jeffreysestablished full DNA profiles. This was a huge and unprecedented undertaking.But, six months and considerable expense later, there was still no match, andagain the case went cold.

The following year, a woman sitting in a local pub overheard a local man calledIan Kelly boasting to friends that he had made £200 by posing as his mate,Colin Pitchfork, at the sampling. Pitchfork, a cake decorator – quiet but proneto bouts of temper – had asked Kelly, a colleague at the bakery where theyworked, to take the DNA test for him. He said he'd been charged with indecentexposure in the past and wanted to avoid being harassed by the police. Theexcuse was shaky but £200 was cash enough to stop Kelly asking questions. Thewoman went to the police, who arrested Pitchfork and took his DNA. It matched.Finally, the detectives had their answer.

In 1988 Pitchfork was sentenced to life imprisonment for both murders. Lawenforcement agencies and scientists around the word sat up and took note. GillTully was an undergraduate biology student at Cardiff University at the time,and it took her breath away to see such a savage – and seemingly unsolvable –crime exposed by so sophisticated a scientific process. She finished her firstdegree and then went on to do a PhD at the Forensic Science Service, where she landeda job afterwards. There she was involved in some extraordinary developments ata time when the UK was the powerhouse for genetic research. Val Tomlinson hadalready been at the FSS for six years when Gill arrived, and she recalls theatmosphere of the pre-DNA days:

'It was very hands-on. Personal protective equipment hadn't really beeninvented. We rarely used gloves. One of the tests for semen staining waswhether or not it was stiff to the touch. We didn't have separate offices. Yourbench was your office, so you wrote your reports on the same bench where youexamined all your dirty knickers and bloodstained items. 

Colin Pitchfork, the first person in the UK to be convicted on the basis of DNAevidence

'It is quite hilarious to think back about when we started DNA profiling. Itwas bucket chemistry, literally bucket chemistry, you would make whole vats ofsalt solutions and you had radioactive substances and you needed a bloodstainthe size of a ten pence piece if not bigger to get a DNA profile.

'In the early stages of my career there were no formal training courses otherthan a very initial one, and then you worked alongside an experienced scientistwho took you along with them and you did everything from blood alcohol testing,semen staining, fibres analysis, hairs. I've done pheasant feather cases, I'vedone salmon poaching cases. I've done leek slashing cases.'

When Gill arrived at the FSS during her sandwich year, while she was still anundergraduate at Cardiff, most of the geneticists approached their work withenthusiasm, but without taking much notice of the revolution they were settingin motion. 'At coffee break the main issue was whether there was still a jamdonut left,' Gill says now with a rueful laugh. While the Colin Pitchfork case hadshown the world how useful DNA could be, she admits, 'We thought it would onlybe for those occasional really highprofile cases.'

But over the years each innovation has broadened the application of DNA. 'Everytime you think, "Oh yes, that would be really good, it's a bit too expensivefor routine but that occasional high-profile crime, that could be the thingthat makes a difference." And yet a number of those have then become cheapenough and routine enough to use even in burglaries.'

The most significant move away from 'bucket chemistry' was made by Kary Mullis,a Californian surfer and LSD enthusiast who went on to win the Nobel Prize inChemistry. In 1983 Mullis was driving along Highway 128 when he had arevelation. If he added an enzyme called polymerase to DNA it would, in hiswords, 'reproduce the hell out of itself'. Using the polymerase chain reaction(PCR), Mullis could take a very small amount of DNA and make it significantenough to interpret. Before long, scientists were using PCR to understandcriminal cases that had been cold for up to seventy years, as well as thegenealogy of fossilised dinosaurs and buried royalty, and the diagnosis ofhereditary diseases.

When Gill Tully started working at the FSS she and her supervisor were the onlytwo people working on refining and using PCR; she regards herself as being'deeply privileged to have been there from the beginning'. Traditional geneticfingerprinting had relied on bodily fluids and hair, but by 1999 the team thatGill was part of was using PCR to develop a much more sensitive method, knownas 'low copy number (LCN) DNA profiling'. To get an LCN profile they neededonly a few cells from a potential suspect. Whether it was a speck of dead skin,the sweat from a fingerprint or the dried saliva from a postage stamp, therequired amount of bodily substance had spiralled down from the size of a tenpence piece to one millionth of a grain of salt.

LCN profiling has had a seismic impact on the way crimes are investigated inthe UK. But its road to acceptance has been a long one. Controversial trialsinvolving LCN DNA evidence have provoked reactions from judges and commentatorsthat have forced forensic geneticists to defend and redefine their methods.

One particularly contentious trial that helped to shape the role of LCN DNA incourt was triggered by a massive bomb in a small town in Northern Ireland. In1998, the Good Friday Peace Agreement was reached, supposedly marking the endof hostilities between Unionist and Republican paramilitary organisations. Buton 15 August the Real Irish Republican Army detonated a bomb which rippedthrough a busy street in Omagh, County Tyrone. Responding to a phoned-inwarning from the bombers naming the local courthouse as the location, thepolice had actually moved people into the path of the blast, located in thecentre of the town. Twentynine people were murdered, including several childrenand a pair of unborn twins. Over 200 were injured. The Secretary of State forNorthern Ireland at the time, Mo Mowlam, described it as 'mass murder'.

Three years later, Colm Murphy, a building contractor, was convicted of causingthe explosion and sentenced to fourteen years in prison. It was to be the startof a long and painfully unresolved judicial process. In 2005 his conviction wasoverturned when it emerged that police had forged notes of interviews they hadconducted with him. The following year, the police arrested Colm Murphy'snephew, an electrician called Sean Hoey. At his trial the prosecution's caserested on LCN DNA found on the bomb timers used in the attacks, which aforensic geneticist said was one billion times more likely to belong to SeanHoey than to an unknown individual. But in the absence of eyewitness testimony,or any other compelling evidence, the case fell apart.

When Mr Justice Weir delivered his judgment on 20 December 2007, he criticisedthe way the prosecution had made LCN DNA the crux of their case, rather thanusing it as a guide towards finding other substantial evidence. He complainedabout the 'slapdash approach' of the police and some forensic experts. He evenalleged that the evidence had been 'beefed up' by police, who were guilty of'deliberate and calculated deception' in their efforts to secure a conviction.He pointed out that the only published papers validating LCN DNA profiling werewritten by its inventors at the FSS. Ultimately Weir found the method too noveland recommended an urgent review of its use – a bad end to an investigationthat cost the state £16 million.

The day after Weir's verdict the Crown Prosecution Service (CPS) suspended LCNDNA profiling, and commissioned a review of its fitness for purpose. Since 1999it had been used in 21,000 serious crime cases in the UK and abroad –particularly in cold cases. The CPS ordered all live cases involving LCN DNA tobe re-examined. One of them involved the brothers David and Terry Reed ofTeesside in the north-east of England.

On 12 October 2006 a friend of the former boxer and hard man Peter Hoe receiveda 4-minute-long voicemail message of the New Age music of Mike Oldfield. Butwhen he played it back and listened carefully he could hear the muffled groansof Hoe, who was bleeding to death from five deep stab wounds in the living roomof his house in Eston, near Middlesbrough. The police arrested and chargedprime suspects David and Terry Reed. David, the older brother, was known to bejealous of Hoe's tough reputation, and in court Hoe's brother claimed that theattack was a retaliation for a pub scuffle some days earlier: 'They went up tomy brother's house and murdered him because David couldn't handle the hiding.'

When Val examined Peter Hoe's living room she found nothing to indicate thatthe perpetrators themselves had bled, but noticed two small pieces of plastic.'We see it all the time when knives are used in stabbings. The vibrations andforces project down the knife blade and hit the hilt with such force that itbreaks.' Back at the laboratory Val looked more closely at the plastic piecesand, based on her experience, decided they had originated from cheap knives.Traces of DNA were found on them. LCN profiling revealed that it matched theReed brothers'.

At the trial the defence called an eminent professor of plastics, 'a lovelygentleman from Newcastle University', who had been to Argos and bought a cheapknife with a plastic handle. Then he had put the knife into a machine that bentit slowly until the hilt broke. He explained in court that he'd measured theforces, and was satisfied that a human wrist was not capable of producing them.He declared that the pieces of plastic were unlikely to have come from astabbing. 'I was sitting in court listening to this,' recalls Val, 'and it wasjust fundamentally wrong. We had another murder in the laboratory at the same timewith four knives. Three of the four had broken in exactly the same way at thehilt.'

The plastics expert had investigated a dynamic event of life and death, steelon bone, plastic on flesh, in a controlled but unrealistic laboratoryenvironment. For Val, that is a scenario fraught with problems. 'Murder is nota replicable experiment. Every one is unique.'

All the time maintaining their innocence, the Reed brothers were both sentencedto a minimum of eighteen years' imprisonment. As they were led from the court,they grinned and thanked the judge while Hoe's mother Maureen cried in thepublic gallery.

Not long after their conviction, Justice Weir acquitted Sean Hoey of the OmaghBombing and LCN profiling was put under intense scrutiny. Though its use wasreapproved by the CPS in January 2008, enough doubt had been sown that, on 20October 2009, the Reeds appeared at the Court of Appeal. Their lawyer arguedthat Val Tomlinson had overstepped the mark when she speculated at the originaltrial on how the Reed brothers' DNA got on to the pieces of plastic recoveredfrom the scene.

At the appeal of the Reed brothers, in October 2009, the court heard from BruceBudowle, a former FBI forensic scientist. Budowle argued that LCN DNA profilingwas inherently flawed and that its results were not always reproducible. 'Theconfidence in it has not been assessed,' he said. He accepted that the bits ofplastic came from the knives of the murderers, but the Reeds' DNA could havebeen the result of secondary transfer – that is, they could have touchedsomeone who then touched the knives.

As well as knowing the latest research-based theories, forensic scientists likeVal must draw on a database of professional experience to understand what theysee. Gill Tully says, 'There have been some interesting judgments from theCourt of Appeal in recent years which have really pointed forensic scientiststo give opinions informed by experience rather than giving statisticalevaluations, which is slightly bizarre for the scientist, although you can seewhere Their Lordships are coming from.' But, as Sherlock Holmes knew way backwhen, 'There is a strong family resemblance about misdeeds, and if you have allthe details of a thousand at your finger ends, it is odd if you can't unravelthe thousand and first.' Val's testimony, both about the shattering of theknife handles and the DNA traces they contained, was based on years ofexperience with evidence. It was data and opinion; art and science. And,ultimately, the court believed her: while the review made some recommendationsabout external validation, ultimately it found the method robust and reliable.The three judges at the Reed brothers' appeal decided that the circumstantialevidence was powerful enough to make doubt unreasonable, and upheld theirconvictions. They thought that Val's professional opinion on how the DNA hadgot on to the plastic was 'not only possible ... but essential'.

The case against the Reed brothers had solid corroborating evidence – such asthe fact that Peter Hoe had inflamed the pride of David Reed by knocking him tothe floor with a light punch in a pub a fortnight before the murder – asopposed to the case against Sean Hoey, which relied nearly solely on LCN DNA.Valuable lessons had been learnt about the place of DNA in criminalinvestigations as a key component of a case, but nevertheless, only acomponent. More such lessons were to come.

In 2011, a woman was brutally raped in Plant Hill Park, Manchester. DNA takenfrom a swab of the victim linked the crime to Adam Scott, a 19-yearold fromPlymouth, who was duly arrested. He was incarcerated in a special segregationwing for rapists and paedophiles, and verbally abused by inmates. But he wasadamant that he'd been hundreds of miles away in Plymouth on the night of thecrime, and had never even set foot in Manchester.

After four and a half months in jail it emerged that Adam Scott had been theunfortunate victim of laboratory cross-contamination. Some months previously hehad been involved in a 'spitting incident' in Exeter, after which police took aswab of his saliva. Scientists placed the swab in a tray at the LGC Forensicslaboratory, which was re-used for the swab of the Manchester rape victim.Scott's mobile phone records confirmed that his phone had been in Plymouth whenthe rape took place.

Andrew Rennision from the government's Forensic Science Regulator said, 'Thecontamination was the result of human error by a technician who failed tofollow basic procedures for the disposal of plastic trays used as part of avalidated DNA extraction process.' Adam Scott's case echoed the strange case ofthe 'Phantom of Heilbronn', a seemingly superhuman female serial killer whoseDNA was found at the scene of robberies and murders across Austria, France andGermany in the 1990s and 2000s. In 2009, when the DNA appeared on the burnedbody of a male asylum seeker in Germany, the authorities concluded that the'phantom' was simply the result of laboratory contamination: the cotton swabsused for DNA collection were not certified for the purpose, and were eventuallytraced to the same factory, which employed several Eastern European women whofitted the DNA profile of the 'phantom'.

As with a real fingerprint, a genetic fingerprint should not be enough tosecure a conviction on its own. According to Gill, 'DNA doesn't lie. It's anexceptionally good lead and exceptionally strong evidence but there is humaninteraction in the process [of profiling]. So the error rate is exceptionallylow but it's not zero ... DNA shouldn't be a lazy way to not do aninvestigation.'

If in some cases DNA has become a crutch for the police to lean on, in manymore it has opened up outlets for their energy, giving them the chance to solvecases both new and old. If DNA found at a crime scene doesn't produce a perfectmatch when it's run through the national database, it's no longer the end ofthe line. Because blood tells more than just one person's story.

Familial DNA searching was developed at the FSS by Jonathan Whitaker when here-examined a grim cold case. In 1973 three 16-year-old girls had been raped,strangled and dumped in woods near Port Talbot in South Wales. After anexhaustive investigation into 200 suspects, police had made no arrests. Then,in 2000, Whitaker used the 28-year-old crime scene samples to develop a DNAprofile of the suspect. He ran the profile through the national database, andturned up a blank. Then, a year later, he was struck by an interesting idea.Could there be a family member on the database with a similar profile? Hesought permission to search and found a profile with a 50 per cent match. Theoffender was on the database for car theft, but Jonathan Whitaker was convincedhis family tree housed a far more heinous offender. Joseph Kappen, the carthief's father, who had died from lung cancer ten years previously, became theprime suspect. An exhumation order was granted and Whitaker was able to analyseDNA from his teeth and femur. It matched and, although the criminal could notbe punished, the triple murder was finally solved.

The first live case to be cracked using familial searching came in 2004.Michael Little was driving his lorry under a motorway overpass when someonethrew a brick from overhead. It crashed through the windscreen and struckLittle in the chest. He managed to steer his lorry on to the hard shoulderbefore succumbing to a fatal heart attack. When scientists fed the LCN DNA fromthe brick into the database, it produced no direct match, but a familialconnection led them to Craig Harman, who admitted his crime and was sentencedto six years for manslaughter. For Detective Chief Inspector Graham Hill ofSurrey Police, there was only one reason a conviction was secured: 'There is nodoubt in my mind that without this groundbreaking technique this crime wouldhave remained undetected.'

In the aftermath of the Harman conviction Alec Jeffreys said that familial DNAsearches raised 'potentially rather thorny' civil liberty issues. The responsehas to be proportionate to the crime, striking the right balance between anindividual's civil rights and the need to identify a perpetrator. Familial DNAsearching for forensic purposes remains illegal in most countries. In the US itis only allowed in California and Colorado, though a familial search on DNAextracted from a discarded piece of pizza may have helped to find the 'GrimSleeper', a serial killer and rapist who terrorised Los Angeles from the late1980s to the early 2000s. In the UK it's only used for investigations intomurder and rape. Since the Harman conviction it has led the police to a suspectin fifty-four serious crimes – producing thirtyeight convictions.

Ethical issues persist. Troy Duster, a sociologist at New York University,points out that because incarceration rates in the US are eight times higherfor black people than for white people (for socio-political reasons, includingalleged racism on the part of the authorities), familial searches are much morelikely to help convict black criminals. The profiles of around two in five of theblack men in the UK are on the national DNA database, compared to around one inten white men. In US, around 40 per cent of the DNA profiles in the federaldatabase are African-Americans, who make up about 12 per cent of the nationalpopulation. It is predicted that the DNA profiles of Latinos (about 13 per centof the population) will soon show a similar skew, mainly due to crimessurrounding immigration.

One way to gradually level the playing field would be to profile everybody.Already the UK national DNA database has over 6 million profiles, a higherproportion (10 per cent) of its citizens than any other country in the world.DNA from everyone arrested (whether convicted of a crime or not) was held onthe database indefinitely until a decision by the European Court of HumanRights in 2008 forced a change. In 2012–13 profiles of 1.7 million innocentpeople were deleted from the database. Alec Jeffreys had called for this in2009: 'My view is very simple ... innocent people do not belong on the database. Brandingthem as future criminals is not a proportionate response in the fight againstcrime.'

Because so many crimes are committed by repeat offenders, the national databaseis a powerful police tool. In 2013, 61 per cent of DNA profiles found at crimescenes found matches in the database. The Home Office doesn't record how manyof these matches led to a conviction, but it's a formidable help to the policeforces, some of whom have advocated mandatory profiling. But others believethat would lead to more false assumptions. DNA from several people is oftenpresent at one crime scene for perfectly innocent reasons, particularly sincescientists can now produce results from such minute quantities.

This nightmare scenario, along with issues of personal privacy and the hugebureaucratic cost of profiling 60 million people, is probably enough to put thematter to rest for now. In addition, some worry that mandatory profiling wouldmake it easier for criminals to frame innocent people. A defence barrister onceput the framing idea to Val Tomlinson in court, claiming that his client's LCNDNA had been planted at the scene by an anonymous other. To prove it, he askedVal a hypothetical question:

'If you were going to set somebody up for this how would you go about doingit?'

'I don't think I could,' said Val.

In Val's experience most set-ups fall down on basic points. 'Children go overthe top when they are trying to cover up for their mistakes. And you tend tofind that people who frame others distribute too much blood in the wrong way,or a whole bucketful of glass instead of the two little pieces which is whatyou would expect to persist on a piece of clothing recovered a week after acrime.' Like any powerful tool, DNA can be misused. But, as always, the analysisof evidence is not simply about the collection of data – whose DNA is or isn'tthere – but also about the interpretative skills of the scientists who dealwith it. This is what should – and mostly does – protect the innocent.

Of course, not all criminals want to hide their identities: when politicalfighters or terrorists commit crimes, they want the world to know who did it.In the Madrid Train Bombings (see p.34), DNA and politics were central to thecase from the very beginning. The timing of the attack, three days before thegeneral election, was significant. In the immediate wake of the bombings, theincumbent government claimed that evidence had been found that implicated theBasque separatist group ETA, perhaps hoping to quash speculation that thebombings were a result of Spanish involvement in the Iraq war. But three dayslater the self-proclaimed 'military spokesman for Al-Qaeda in Europe', AbuDujana Al-Afghan, declared responsibility. 'This is a response to the crimesthat you have caused in the world, and specifically in Iraq and Afghanistan ...You love life and we love death.'

A month later, seven suspects on the verge of a police raid detonated bombs intheir apartment, wiping out four of them and one police officer. Scientistscould not match the LCN DNA found at the scene (including on a toothbrush), andfrom other locations, with profiles on national databases. A judge ruled thatthe scientists should use the DNA to determine whether the suspects who werestill at large were of North African or European descent. This would helpinvestigators finally settle whether their targets were members of Al-Qaeda orETA.

But intermarriage between southern Europeans and northern Africans on bothsides of the Mediterranean made differentiating between the two all butimpossible using the technology current at the time. Forensic geneticistChristopher Phillips developed a new technique, and was able to conclude thatone DNA profile, which did not belong to any of the dead or arrested men,'almost certainly' belonged to a North African. Familial DNA searching laterindicated it belonged to Ouhane Daoud, an Algerian whose fingerprint was alsofound on unused detonator caps in a Renault Kangoo near the site of thebombings.

While conducting his research into ethnicity, Christopher Phillips was alsoable to deduce 'with around 90% predictability' that DNA from a scarf found ina van used in the bombings belonged to someone with blue eyes. More and more,scientists can discern details about a suspect's physical appearance from theirDNA: traces left at a crime scene can describe the people who were there almostas accurately as an eyewitness.

It all started with ginger hair. In the early 2000s scientists at the FSS foundthat if a gene (the melanocortin 4 receptor) is switched off in both parentsthen the child will have red hair. Gill Tully is cautious about the ethicalimplications of DNA profiling in this way, but overall, she says, 'It's aboutusing things in the right way: when we were developing the red hair test, wehad some detectives in Scotland phone us and say, "There's been a shooting andwe know from the ballistics what window the shot was fired from. Around therewe found some cigarette butts and got a DNA profile from them. We also have aneyewitness account saying a red-haired man ran away from the building. Sobefore we start doing a mass DNA screen of individuals to see if we can findthe person who smoked the cigarette butts, can you tell us if they were smokedby a red-haired man?" We weren't quite able to do it at that stage but it wasquite a nice example of how these things can be used in an ethical andappropriate way to help direct an investigation so that you don't spend lotsand lots of money analysing cigarette butts that are entirely irrelevant andwere smoked by somebody months ago.'

Genetic fingerprinting is a powerful indicator of guilt or innocence; thesingle biggest advance in forensic science since William Herschel and HenryFaulds developed fingerprinting a century before. Much of forensic science isbased on subjective interpretation: as explored in the Fingerprinting chapterof this book, experts are sometimes good at finding patterns where they wish tosee them, like all human beings. That is a useful skill for a forensic investigator,as long as its intuitive nature is recognised and expressed in court.

Although human error can always creep in, in its simplest form, DNA pulls usout of the trap of subjective bias, interpreting empirical data using objectiveprobabilities that have been refined for over thirty years. When Gill hasunmixed DNA from a crime scene which she matches to a suspect she can safelytell jurors that the 'probability of observing that profile if it was fromsomeone other than the person of interest would be one in a billion. That's aconservative estimate that an average juror can get their head around. If youstart going into trillions it means nothing.' But life – and crime scenes – arerarely simple. Where, as Gill points out, 'you have DNA mixed from two people,which is often the case, then you need to do a more thorough evaluation ofstrength of evidence and look at the probability of observing that particularset of mixed peaks both if the prosecution hypothesis is true, and if thedefence hypothesis is true.'

There remains a lot more for forensic scientists to learn from DNA. At themoment Val and Gill look at much less than 1 per cent of a person's DNA tojudge whether it matches a profile on the national database. As it get quickerand cheaper, 'you could theoretically analyse someone's entire genome'. Thepossibilities are endless, 'but there are very, very significant ethical andpractical issues to answer before you would want to do that. You certainlydon't want to be using forensic samples to generate information about people'spredisposition to commit crime.' That's a profoundly disquieting thought. Wealready know, for instance, about the existence of a 'warrior gene' – presentmainly in men – which is linked with violent and impulsive behaviour understress. We do not, in the twenty-first century, want to return to CesareLombroso's nineteenth-century uomo delinquente, or 'criminal man', or theVictorian discipline of phrenology, which diagnosed a predisposition tocriminality from lumps on the skull. A nightmare scenario by any measure.

But, if used proportionately, the future of the genetic fingerprint is moreexciting than scary. There are now instruments that can analyse DNA in lessthan an hour and a half, making it possible to run the profile of an arrestedsuspect through the national database before they are released from custody. Ifthe search produces a match to profiles found at scenes of unsolved crimes,police have halted a serial offender. Gill explains, 'Habitual burglars, if theyare caught, sometimes know the DNA is going to get them so when they are out onbail they will commit more crimes to look out for their family while they areaway. Then they'll ask all those crimes to be taken into account and they'llserve their sentences concurrently. There are a few really key cases where itcould have potentially prevented serious crime, where people had come intopolice custody then gone out and committed a serious crime. Whereas if thepolice had had a DNA result quickly they would never have been bailed.'

At the moment analysing the minute quantities of DNA usually found at crimescenes takes quite a bit longer than an hour and a half, but 'the time willdefinitely come, and it won't be that far away, when you'll be able to identifya suspect, and not only identify them but potentially go around to theiraddress before they have fenced off all the gear that they've just stolen. Itcould then be returned to people, things of sentimental value and so on. Thepotential to do this really, really quickly is not that far away. It's notgoing to be long.' Let the burglar beware.