Traces of Guilt Catching History's Criminals: The Forensics Story

In the act of murder, there is a weapon...

GUNSHOT

..a crime scene...

..and a body.

All vital evidence in the hunt for the killer.

It's a game of cat and mouse between police and murderer

that used to favour the criminal,

but then something happened that swung the odds in favour of justice.

The arrival of forensic science.

I'm Gabriel Weston.

As a surgeon and writer,

I'm fascinated by the work of the forensic scientist

and the murders they've helped to solve.

In this series...

..I'll explore the cases that transformed criminal investigation.

Through poison and acid...

..fingerprints and blood...

..from the earliest days,

to the cutting edge of modern forensics.

This is the story of the crime scene...

..four murders which reveal

how science became central to crime scene investigation.

But it's a troubled history.

The science is improving all the time

and can tell us more and more about what happened,

where, to whom and who did it.

But the science has to be applied by humans -

and humans always have been and always will be far from perfect.

In June of 1908,

the police in the small town of Rockenhausen, Bavaria, were stuck.

The body of a woman in her early thirties -

an unmarried housekeeper, elegantly dressed -

had been discovered in the woods.

Margarethe Filbert had been strangled and decapitated

and her head was nowhere to be found.

Now, in this small, tightly-knit, rural community,

everyone knew everyone else

and the investigation was only a few days old

when the police realised they had a suspect firmly in their sights.

There was only one problem.

They had absolutely no evidence.

Nobody had witnessed the murder,

but as the police knocked on doors and asked questions,

one name cropped up again and again -

that of a local farmer, Andreas Schleicher.

They paid him a visit.

Schleicher simply denied all knowledge of the crime.

The police asked to see the clothes he'd worn that day,

because the killer's clothing must have been drenched in blood.

But the jacket and trousers seemed clean.

The shoes had some mud on the soles, but why shouldn't they?

Schleicher lived on a farm.

He insisted he'd spent the whole day in his own fields.

That was his story, and he stuck to it.

But the police took the clothes and shoes away, just in case.

They were certain he was Margarethe's murderer,

but all they had was gossip and hearsay.

They needed evidence.

The idea that science had something to offer crime investigations

had been gently stewing in Europe for at least 20 years,

ever since Conan Doyle had first had Sherlock Holmes

make his collections of different cigar ashes.

Whole books had been written

about the ways in which science could shed light on criminal cases

and it was the local District Attorney who remembered

that there was a chemist in Frankfurt, Georg Popp,

who had helped the police before.

He sent Popp a telegram.

Popp's normal business was more ordinary.

His company specialised in food hygiene

and the purity of mineral water.

Crime science was just a sideline.

But for Popp, it was also a passion.

He replied to the telegram at once.

Yes, he would be more than happy to help,

but the police weren't happy to have him.

They resented the meddling District Attorney

inviting this amateur to get involved

and repeatedly refused Popp's requests

that they send him the clothes and shoes.

Instead, they searched the valley again and again,

looking for Margarethe's purse, her parasol

and her severed head.

In the cellars of a ruined castle

they found a secret cache,

including a shotgun and some trousers which were damp,

but apparently free of blood.

Could these items have anything to do with the murder?

Could they belong to Schleicher?

Once again, Schleicher denied all knowledge.

Several weeks after the discovery of the secret cache,

a lowly constable was inspired to turn the trousers inside out.

Once word got around,

the police were a laughing stock.

Now, the District Attorney insisted that Popp must be involved.

At long last, the clothes and shoes were sent to Georg Popp

in his Frankfurt laboratory.

Margarethe Filbert had been dead two months.

Popp began with some chemical tests

and made his first revelation.

Schleicher's jacket and trousers

were indeed both stained with human blood.

A good start, but there was no way to prove whose blood it was.

It didn't connect Schleicher to Margarethe,

so Popp stopped looking at the clothing

and he started looking at the shoes.

And he had a simple but amazing insight

that would prove pivotal to the case.

It occurred to him that the mud on Schleicher's shoes

was a record of where he'd been on the day of the murder.

I'm in a muddy spot here, myself.

There's mud all around me at my feet,

it's squelchy and there's all sorts of stuff

growing in and around the mud - there's moss and plants.

There's twigs and all sorts of rotted matter.

It's evident, then, that if I were standing here for any time at all,

I'd be gathering evidence of this mud all over my footwear.

Popp had good reason to think

that the shoes held the evidence he needed.

He knew the police had had them

unworn and uncleaned since the murder.

He knew that Schleicher insisted

he'd only walked in his own fields on that day.

But what if the mud on the shoes

could prove he'd walked somewhere else entirely?

Late in September, Popp travelled to the valley.

He collected soil samples

from every location that was connected to the case.

Popp compared these soil samples with the mud on Schleicher's shoes.

First, he had to cut into the mud to make a section from top to bottom.

Now, I'm a surgeon and like to think I have a steady hand.

Let's see how I get on with this task.

It's very difficult.

It crumbles incredibly easily.

All I can get off is these chunks.

Popp was more successful

and what he learnt from his microscopic analysis was this -

he found nothing that resembled the soil of Schleicher's fields.

Instead, stuck directly to the sole itself were goose droppings,

like those outside Schleicher's house.

They must have got onto the shoe first.

Then, a second layer contained grass,

like the meadows on the way to the woods where the body was found.

The third layer was a sandy soil,

flecked with granules of quartz and specks of leaf mould,

exactly like the soil where Margarethe's body had been found.

And the final, outermost layer

was a soil containing fragments of dust,

exactly like the castle cellar

where the secret cache had been discovered.

By now, he had connected Schleicher with the crime scene,

but he didn't stop there.

In the third layer, he found tiny fibres -

wool and cotton, reddish brown.

And in the fibres of Margarethe's petticoat,

he found a perfect match.

He'd established that Schleicher

hadn't visited his own fields at all that day.

Popp's analysis had tracked him all the way to the scene of the crime,

to the secret cache in the cellar

and had proved that he had even stood on Margarethe's petticoat.

The soil on his shoes had shown he was the killer.

Schleicher was jailed.

After a year behind bars, he confessed.

He'd killed Margarethe because she looked well off,

then found her purse was empty

and cut her head off in a fit of rage.

He told the police where it was hidden.

Georg Popp had invented a new discipline -

forensic geology.

To see what's possible today,

after filming for this story,

I sent my Wellington boots by courier

to a forensics laboratory in Aberdeen.

We filmed this story in Mugdock Country Park, north-west of Glasgow,

and I want to see if the scientists,

using the mud stuck to my boots and nothing else,

can tell me exactly where in Scotland's 78,000 square kilometres

that muddy evidence came from.

'At the Hutton Institute Labs in Aberdeen,

'one of my boots is having its mud removed for analysis.'

You can tell specifics about what those species are....

'Professor Lorna Dawson and her team

'have instruments and techniques at their disposal

'which Popp could only have dreamt of.

'After two days of tests,

'Lorna takes her new data to her colleague, David Miller.'

So, we use the national mapping of resources of Scotland....

'David has maps of Scotland

'which contain information about minerals,

'plant distribution and the soil itself.

'But have they learnt enough

'to tell me where the mud on my boots came from?

'It's a process of elimination.'

Lorna, can you give me

some idea of what we're going to be looking for first,

or what we cannot be looking at?

Well, what we knew from the gas chromatography.....

'The first result arose from gas chromatography,

'which analyses traces of organic matter and chemicals.

'It revealed the absence in our sample of this

'pink line of hydrocarbon pollution, caused by cars.

'We didn't film in a town or city.'

-So we exclude the urban areas of Scotland.

-OK.

So, that's all the black areas now, are not part of the search?

Yes, exactly. Next...

'Using techniques that Popp would have been very comfortable with,

'Lorna studied the soil itself

'and the botanical traces it contains under a microscope.'

You can see that there's a fibrous texture to that soil.

It's got very small granular structure.

It's not a single grain,

as you would find generally in an arable soil.

It's not got the peat fibres that you would expect in a peatland.

'So, Lorna concluded that we didn't film in peatlands and uplands -

these purple sections.

'And she excludes all of Scotland's arable land,

'in light green, from the central belt to the north-east.

'But Lorna did find fragments of beech leaves.

'Gas chromatography confirmed

'there were no chemicals connected with pines and conifers.

'It's deciduous forest they're looking for.

'Scotland's pine forests, in dark green, are eliminated.'

We're really narrowing down now,

to really what looks like a tiny amount left of Scotland,

where this could have been, this walk.

-Well, we hope it is.

-LAUGHTER

So, you were telling me something about the underlying geology.

'Lorna's team has two powerful resources for geological analysis.

'X-ray crystallography identifies the minerals found on my boot

'by their crystal structure...

'..and the National Soil Archive - soils from all over Scotland -

'has been used to make a database

'for comparison with crime scene samples.

'In our case, the results point to

'an area of volcanic soils that surrounds Glasgow.'

So, the area we're zooming in

is to the north and the west of Glasgow, here....

'Now they're looking not at where I wasn't,

'but at where I was.'

..Then what we're interested in are the areas...

'Lorna and David's software highlights areas in yellow

'where the right kind of volcanic soil

'can be found with deciduous trees.'

Look just a little bit to the north of Glasgow,

there's quite a big area here which fits the characteristics......

'They soon home in on a place that looks very familiar.'

It's got the right soil characteristics and....

'I think I'm about to be astonished.'

That could be an area that you picked up the soil and that

would certainly be an area that we'd prioritise to the police to search.

And where is this area?

This is Mugdock Castle, at Mugdock Country Park,

just north of Glasgow, that we're looking at just now.

-Mugdock Loch.

-I'm completely blown away.

Absolutely amazing, it's exactly where we were.

-Was that where you were?

-Exactly where we were.

So, out of 78,000 square kilometres, to what -

what would this area be?

This area will be a few hundred square metres.

-Amazing.

-Can you point where you walked?

-Well, we were right by the castle...

-Gosh.

..and we went down into the forest, there was a valley...

Lorna and her team would certainly have caught Schleicher too.

Crime scenes stick to criminals,

and Georg Popp was one of the first to prove it.

After Popp, criminals would have to contend with investigations

that could determine what they'd done

and even where they'd been with ever greater clarity.

With the help of scientific techniques,

crimes that might have taken months to solve -

if they'd ever been solved at all -

could sometimes be done and dusted in just a few days.

ENGINE STARTS

Within three decades,

one such crime was to prove just how far crime scene analysis had come.

Manhattan, New York.

10th April, 1936.

Good Friday.

Theodore Kruger and his employee John Fiorenza

were delivering a reupholstered chair to one of their customers.

The third floor apartment at 22 Beekman Place

was home to Lewis and Nancy Titterton.

He was a publisher.

She was an aspiring crime writer,

who usually worked at home.

They had an appointment with Mrs Titterton,

but there'd been no answer when they rang the bell.

The street door, however, was open -

so Kruger and Fiorenza had climbed the stairs

to find the apartment door was open too.

Why wasn't she here?

Kruger pushed the door open and they carried the seat inside.

At the very least, they could put it back where it belonged,

in the living room.

There was no sign of Mrs Titterton.

Kruger decided to leave a note for her

and in looking around for pen and paper,

the two men glanced into the bathroom.

There, Kruger saw Nancy Titterton

and why she hadn't answered.

She lay in the bathtub, with a piece of clothing tied round her neck.

She was dead.

Anyone could have killed her.

It was the sort of crime that might once have defied solution,

but by 1936,

wherever there was a large, modern city

with a well-funded police force,

there were crime scene investigators.

The New York Police Department had its own crime scientists on staff

and the Chief Medical Officer had a team as well.

Soon, men from both appeared on site.

Nancy had been strangled with her own pyjama top.

When her body was lifted from the bathtub,

a 13 inch length of cord was revealed.

Nancy's wrists had been bound

and the killer had tried to remove the evidence,

unaware that this section lay beneath her body.

It was string, really.

Nothing fancy, just a piece of binding twine.

Clearly, the murderer had brought it here,

but leaving it behind wasn't part of the plan.

But could it tell the investigators anything more?

It was sent to the laboratories

of the Chief Medical Officer for analysis.

The rest of the crime scene was secured and analysed.

The carpets, the bedclothes

and even the bed itself

were taken away for microscopic scrutiny.

Their attention to detail soon paid off.

In the bedclothes they found a single hair,

about half an inch long, too stiff to be human

and too heavy to have blown into the room on its own.

When they examined it under a microscope,

they found it was horsehair -

commonly used for padding seats.

Attention turned at once

to the two men who claimed they'd only discovered the body.

But if one of them had transferred it to the bedclothes,

when had that happened?

Kruger said he'd only glanced through the door.

As for the cord,

on 17th April, about a week after Nancy's death,

careful microscopic analysis showed

it was indeed highly unusual.

Made mainly of hemp and jute,

it also contained a rare fibre called istle.

In the three states closest to Manhattan,

there were 25 manufacturers of rope and cord.

The NYPD telegraphed them all.

Only one, the Hanover Cordage Company of York Pennsylvania,

responded that it made cord containing istle.

Officers took the cord to the company for comparison.

Although it had expanded in the bathwater,

it was clearly identical to the company's product.

Officers then asked to see the company's dispatches

and found that some of the cord

had been delivered the day before the murder

to an upholstery company in Manhattan,

run by Theodore Kruger.

It was alibi time.

Theodore Kruger said he'd been in his own premises all morning,

with witnesses to prove it.

But John Fiorenza claimed he'd been at an office,

which the police soon found out

had been closed on the day of the murder.

They brought him in for questioning

and after having a bit of fun with his impossible alibi,

one of the detectives simply laid the piece of cord

down on the table in front of him.

Fiorenza went silent.

After a night in the cells, he confessed.

On 9th April, the day before the murder,

he'd gone with Theodore Kruger to collect the seat

and had managed somehow, during this brief encounter,

to persuade himself that Nancy Titterton was attracted to him.

The following morning, he'd gone back to see if he was right,

with a coil of twine in his pocket in case he wasn't.

The NYPD were justly proud.

They paraded John Fiorenza for the cameras -

white fedora, cigarette in mouth,

headed for the electric chair.

He'd raped and strangled Nancy Titterton

and the NYPD had proved it.

It had taken them just ten days.

The moral of the story was simple.

All investigating officers anywhere in the world needed to do

was do what had been done in Manhattan in 1936 -

use science.

Preserve and analyse the crime scene in microscopic detail

and act on the evidence generated.

By the middle years of the 20th century,

faith in forensic science was limitless.

It seemed there was no crime it couldn't solve.

Some crime scientists would become personally famous.

The American Paul Leland Kirk was one of the most famous of all.

As he wrote in his book, Crime Investigation,

physical evidence cannot be wrong.

It cannot perjure itself.

It cannot be wholly absent.

Only human failure to find it,

study and understand it,

can diminish its value.

It sounds like an astonishingly confident statement

on the power of forensic science -

and yet, there's that last sentence with its emphasis on human failing.

Paul Kirk wrote those words in 1953

and it's almost as if he knew what would happen next -

knew that he was about to become involved in a case

where the simple rules - the basic procedures -

would be forgotten, abused, ignored.

On January 22nd 1955,

Paul Kirk arrived in the small town of Bay Village,

near Cleveland, Ohio, on the shores of Lake Erie.

He was America's most revered forensic scientist,

come all the way from California to analyse a crime scene,

but it was a cold case

and Dr Kirk was more than six months too late.

The crime had taken place on July 4th of the previous year,

and Kirk was here to help the defence lawyers prepare an appeal.

Their client was already in jail,

convicted of second degree murder.

The victim, as usual, was a woman.

Her name was Marilyn Sheppard

and the convicted killer was her husband Sam.

She'd been beaten to death in her bedroom...

..and this was the room that Kirk had come to see.

Those black spots, those splashes,

are blood - all blood.

All four walls were spotted and splashed with blood.

This is the bed,

once her body had been removed.

Until that day,

the Sheppards had seemed to be the ideal American couple -

college sweethearts, happy, wholesome, prosperous.

Dr Sheppard was a surgeon at the local hospital -

a business that his family owned -

and at the time of her death,

Marilyn had been four months pregnant with their second child.

But at the trial, the prosecution had revealed that Sam Sheppard had

had an affair two years before.

The happy marriage, it seemed, was an illusion.

So this was the story the jury chose to believe.

In their luxurious lakeside house on 3rd July 1954,

the Sheppards had friends round for dinner.

They watched a film.

Marilyn sat on Sam's lap.

Later, Sam fell asleep on a day bed

at the bottom of the stairs that led up to their bedroom.

Their guests left and sometime in the early hours of the morning,

Sam woke up, went upstairs to where his wife was sleeping

and beat her to death, leaving her body in her bed...

..fractured skull,

broken and missing teeth,

mattress soaked with blood and urine.

Sam Sheppard's version was different.

He said that he'd been woken from his deep sleep on the day bed

by strange, loud noises.

He'd gone upstairs to find someone in their bedroom -

someone who'd knocked him unconscious.

When he came to, he discovered

Marilyn lying on the bed, soaked in blood.

He knelt beside her to see if she was breathing

and at this point, he heard noises coming from the floor below.

Rushing downstairs,

he was just in time to see someone leaving through the back door.

From the back of the house,

a set of stairs led all the way down to the shore of Lake Erie.

Sam chased this figure down the stairs to the lake,

where they struggled

and Sam Sheppard was once again knocked unconscious.

When he awoke, he was lying partly in the water.

His trousers were soaked and his T-shirt had been stolen.

He returned to the house and at around five in the morning

he called not the police, but his neighbours,

telling them "Come quick, I think they've killed Marilyn."

The jury weren't convinced.

Now, nearly seven months later,

Sam Sheppard was in jail.

All that stood between him and a life sentence

was Dr Paul Kirk.

In the bedroom where Marilyn had died,

Kirk set about the painstaking task of rebuilding the crime scene.

The bloodstained walls and doors were still there,

still smeared with fingerprint powder.

He had obtained the sheets, the pillowcases,

the mattresses of both beds.

He proceeded carefully...

..measured everything...

..referred constantly to the original crime scene pictures.

He was here to see if he could make sense of it all.

Were there signs of the intruder?

Or signs only of a husband's fatal assault on his wife?

But there must have been another question in Paul Kirk's mind

as he stood here, studying these photographs

with their spattered walls and horribly bloodstained sheets.

Why was it only now, more than six months after the murder,

that a crime scene analyst of his skill and experience

was standing here where Marilyn had died?

Why hadn't this been done before?

It looked as though it had.

The original crime scene photographer

had certainly worked hard.

But most of what he recorded -

Sam Sheppard's medical bag overturned,

signs of disturbance in several other rooms,

drawers pulled out and their contents dumped -

was never really analysed at all.

Bay Village was in Cuyahoga County.

It had its own small police force and its own county coroner.

It would have been very simple for that coroner, Dr Samuel Gerber,

to call the nearby city of Cleveland and ask for crime scene analysts,

but he didn't.

From the moment he arrived at the Sheppard house,

Gerber was certain that Sam Sheppard was guilty.

He never entertained any other possibility

and under his control,

the investigation unfolded as a catalogue of errors.

The crime scene was not properly secured.

Journalists were allowed access on the day of the discovery.

Sam Sheppard had injuries to the back of his neck.

Self-inflicted, or the work of the intruder?

Gerber never bothered to study them.

The room was covered in blood,

but Gerber would ignore it all,

except for the stains on Marilyn's pillow,

in which Gerber thought he could see the outline of a medical instrument.

He became convinced that Sheppard had killed his wife

with one of the tools of his trade.

He wasted months in the search for a weapon that matched the stain.

But worst of all,

that search for something that fitted the mysterious outline

was the only analysis he ever conducted of the spatters and stains

that covered the walls and beds.

All of it was evidence and it was all ignored.

Paul Kirk would focus on precisely what Gerber had neglected -

the patterns of bloodstaining -

and he had noticed something significant.

All four walls had blood on,

but as his diagram shows,

in what he called the north-east corner,

he found no blood at all.

A blood void.

Why was there no blood there?

'At Harperley Hall Police Training College near Durham,

'there are facilities for training in blood pattern analysis.'

If you only had three drops and I really hit it like that,

it's going to break it up....

-'That's what Pete Smith teaches...'

-If you then apply the same....

'..and he's going to show me what a blood void is -

'and what it means.'

..Keep my mouth shut for this one.

Right, so I'm going to...probably use that amount of blood.

'In the corner of the training room,

'Pete puts horse blood in the palm of his hand...'

One, two, three.

'..and punches it several times to create spatter across the walls.'

One, two, three.

Like that, OK? So, one punch.

Spread it onto my upper front here

and if you look over my left shoulder,

you can see a lot of spatter. This is called impact spatter,

that's been projected from the punch, up onto the wall

and indeed, onto the ceiling.

And if I actually move away, you can see that

there's actually what we call a "void area" on the wall...

-Yes.

-..which I think most people would equate with

the position of my body at the time of the beating.

'Here was strong evidence that Marilyn's killer

'had stood on the right-hand side of her bed,

'in the blood void.

'Now, Paul Kirk wanted to try and understand

'the patterns of blood spatter themselves.

'He soaked a sponge rubber pad with blood

'and struck it in different ways,

'from different angles,

'with different weapons.'

So I'm just going to load some blood, using the pipette,

onto the surface here.

Just a little bit more....

'Pete Smith uses a joint of pork.'

So, after three -

one, two, three.

-That's good.

-Wow.

'I'm used to blood, but not to violence.'

Feels horrible, I have to say.

It depends how many times you've done it.

One, two, three.

Excellent.

'Different actions make different spatter.'

In the photographs from the bedroom,

Paul Kirk had observed that the largest drops of blood

were on the bedroom and wardrobe doors,

both on Marilyn's right.

In his experiments, he found that these large drops

were usually associated with "cast off", or "throw off" -

blood that had accumulated on the weapon

and was thrown off as it was raised to strike.

The blood drops on Sam's bed, to her left, were much smaller.

These were blood drops produced by the impact of the weapon.

So this, Paul Kirk concluded,

was the arc the weapon had been swung through.

He was more and more certain

that Marilyn had been struck by a man standing on her right,

holding the weapon in his left hand.

Sam Sheppard was a right-handed man.

What do you say about the possibility from blood spatter

of being able to determine something

like whether an assailant was left-handed or right-handed?

I personally would exercise some degree of caution,

because you could have somebody who's right-handed

swinging from here to here,

or they could swing from left to right

and similarly, with a person who's left-handed,

left to right or right to left.

It's all about observing,

inferring something from what you see

and then passing your opinion.

Somebody else can always have a different opinion.

Blood pattern analysis has been practised for decades

and these days, professionals like Pete Smith have become aware

that what they can offer is at best an interpretation -

a probability.

But Paul Kirk believed he could be very definite

and after two months of experiments,

he presented a detailed report to Sam Sheppard's defence lawyers.

His conclusions were very powerful.

He excluded Sam Sheppard from the scene of death,

apart from as a distressed husband.

He put great emphasis on his belief

that the killer was a left-handed man.

Ten years passed

while Sheppard's lawyers struggled to obtain a retrial,

so it wasn't until 1966

that Paul Kirk finally went on the stand to present his evidence.

When he did, the prosecution had nothing to rebut

his analysis of the spatter patterns.

Sam Sheppard was acquitted and left court a free man...

..but a broken one.

He died of liver failure in 1970.

Was he innocent?

Did he struggle with an intruder by Lake Erie,

or did he beat his pregnant wife to death?

It's impossible to say.

The original investigation left too much evidence untouched,

and some people have always felt that the jury at the retrial

were blinded by Paul Kirk's scientific authority.

What we can say with absolute certainty is this -

without Paul Kirk's evidence, without his analysis,

Sam Sheppard would have died in jail.

Forensic science had secured his freedom.

The Sheppard case reminds us

that there are two kinds of forensic science.

Some is definitely science -

Georg Popp's analysis of soil.

The NYPD's careful scrutiny of an apparently ordinary cord.

But some, like Paul Kirk's investigation,

is really about opinion and interpretation -

and without proper safeguards,

it can borrow the authority of science

and disguise those opinions as hard, scientific fact.

It's a particular risk for a kind of crime scene evidence

that we've sworn by for more than a century.

We leave them everywhere we go.

We can't help it - unless we're wearing gloves, of course.

Fingerprints.

They're one of the oldest forms of crime scene evidence we have.

We trust them implicitly.

The idea that everybody has different, unique fingerprints

is an article of faith.

The gold standard of crime scene evidence.

Incontrovertible proof

that the suspect had been exactly where they said they hadn't.

But a little more than 100 years ago,

things were completely different.

On 2nd May 1905,

prosecuting counsel Richard Muir arrived at the Old Bailey.

He had just two days in court to convict two men.

The only evidence of their presence at the crime scene was a fingerprint

and the idea that fingerprints were unique to each person was new.

His fingerprint evidence was itself on trial.

The crime had taken place at number 34 Deptford High Street,

where Thomas and Ann Farrow, an elderly couple,

managed a shop selling inks and paints.

Two men had gained entry to the shop and attacked the Farrows.

Thomas died on the scene,

Ann from her injuries a few days later.

The thieves, Alfred and Albert Stratton,

found the shop's cashbox in the bedroom.

It contained about £13.

The Strattons must have been hoping for more,

but they took the money and fled,

sure that their masks had protected their identities.

In fact, they should have been wearing gloves.

In 1905, there was absolutely no reason

why two amateurs like the Stratton brothers

would have known that their own hands could give them away.

Scotland Yard's Fingerprint Bureau had only been founded in 1901

and fingerprint evidence had only ever been used

in a low profile burglary case,

so the Strattons hadn't worn gloves

and that's why they left behind a crucial piece of evidence.

A thumbprint, in fact.

The Fingerprint Bureau were certain it had been left

by one of their leading suspects, the Stratton brothers.

They were sure they had a match with Alfred Stratton's right thumb

and there was no innocent reason for that thumbprint to be there.

As long as you accepted the newfangled idea

that fingerprints were unique,

then it proved his presence

and his criminal intent at the scene of the crime.

But what if you didn't accept that idea?

Muir had to defend his fingerprint evidence.

Over 40 eyewitnesses had seen the Strattons

near the shop on that morning.

Muir called them all to the witness stand.

He must have rehearsed them all for speed.

He had to allow as much time as possible for his expert witness,

Detective Inspector Charles Collins of the Fingerprint Bureau,

because when Collins took the stand,

he couldn't just say he'd found a thumbprint

and it belonged to Alfred Stratton.

He would first have to explain to the jury

what fingerprints were

and why they were proof of anything at all.

'We don't have to travel back in time to hear what Collins said.

'The methods of fingerprinting haven't changed at all.'

The most important thing is not to be rushed...

'Stephen Hughes is a fingerprint analyst

'of over 30 years' experience.'

So, Stephen - talk me through some of the basic principles.

When we make an identification,

there are two levels of identification we have to make.

The first level of detail is the patterns.

Now, as you can see here,

there are three basic patterns.

That is the arch type pattern,

which is fairly clear to see why it's called an arch type pattern.

You have a loop type pattern...

..and you have the whorl.

And then we take it to the next level.

Fingerprints are formed by ridges.

These ridges are the black lines you see on these patterns

and at the summit of these ridges are microscopic pores.

The microscopic pores exude sweat.

The sweat runs along the ridges and then,

when we place our hand and grasp something,

we find an image of the ridge detail upon it in sweat

and that's a latent fingerprint.

The ridges will suddenly end

or bifurcate into two, like here.

This is what makes us individuals.

What would be the smallest percentage of a print

on which you would be able to confidently make an identification?

I would like to have at least nine or ten ridge characteristics

in coincident sequence, without any in disagreement.

So, Stephen's looking for nine or ten places

where ridge lines break or split in two

in exactly the same way in both the crime scene mark

and the fingerprint taken by the police.

He could find what he needs in an area this small.

In 1905,

DI Collins had worked in the same way.

In the courtroom,

he gave the jury and the Stratton brothers a whistle-stop tour

through these basic principles of fingerprint analysis.

At last, he produced two images -

an enlargement of the mark taken from the cashbox

and an enlargement of the print

taken from Alfred Stratton's right thumb.

And he drew the court's attention, one by one,

to 11 ridge details at which the prints agreed.

"This is definitely the fingerprint of Alfred Stratton",

DI Collins insisted.

Muir had made his case.

Now, he would have to watch as the defence attacked

the very idea of fingerprint evidence.

First, they called Dr John George Garson,

a man who knew contemporary police practices well,

but had only a passing knowledge of the principles of fingerprinting.

But the jury didn't know that

and he certainly looked convincing

as he conducted a point-by-point rebuttal of Collins' evidence.

Garson said fingerprints proved nothing at all,

but when Richard Muir took over to cross-examine,

he stopped Garson in his tracks.

He produced a letter,

in which Garson had offered to testify for the prosecution.

He would give evidence for whoever paid him most.

The judge told the jury to ignore his evidence.

The defence were so horrified

that they decided not to call their second expert at all,

which, for Richard Muir and the prosecution,

was the very best thing that could have happened.

They would have had trouble with the questions this expert witness

never got a chance to ask.

He thought all ten fingerprints were needed to prove unique identity -

and as for the crime scene finger mark,

the result of a brief contact or a glancing touch....

How good could such an image possibly be?

That unasked question has recently come back

to haunt the field of fingerprint evidence.

In 1997, Detective Constable Shirley McKie of Strathclyde Police

was working on a murder case in Kilmarnock.

The fatal stabbing of Marion Ross, aged 51.

A print was found on the bathroom doorpost inside the house.

It was logged as "Fingerprint Y7",

and identified by fingerprint analysts

as the left thumbprint of Shirley McKie.

But her duties had never required her to enter the house.

It was a serious breach of protocol if she had.

Shirley McKie insisted that she hadn't been inside the house

and Strathclyde Police were now faced with an interesting decision.

Did they trust in the science of fingerprints,

or did they trust Detective Constable McKie?

It wasn't a difficult decision.

They chose fingerprints.

Shirley McKie would be charged with perjury and lose her job.

But fast forward to 2011.

Shirley McKie isn't on trial.

Fingerprint evidence is,

and this is the verdict.

One - there is no evidence other than the mark Y7

to suggest Miss McKie at any time

entered Miss Ross's house beyond the area of the porch.

Two - the mark Y7 on the doorframe of the bathroom in Miss Ross's house

was misidentified as the fingerprint of Miss McKie.

Three - Miss McKie did not make the mark Y7.

It had taken 14 years,

but here, at last, was clarity.

Shirley McKie had told the truth.

The fingerprint analysts, who said the mark Y7 was hers, were wrong.

You can trace that case back to a Scottish judgment,

a Scottish appeal court case, sometime in the '30s,

where the judge described fingerprints as

"practically infallible".

It's a very curious phrase

and it's certainly my impression

that as a consequence of that,

that notion of infallibility...

Gradually, that idea took hold.

That then became evidence that was essentially incontestable.

It was error-free.

It didn't merit any kind of argument or challenge,

and it became a matter of routine.

So, it became embedded as a presumption

in the legal process in Scotland,

that fingerprints were simply incontrovertible evidence.

Not just in Scotland, of course.

There were other cases in other countries

where fingerprints were misidentified.

The errors were always human.

The cases where we have seen mistakes are often to do with

particular situations in the case,

and it is usually one or two things in combination.

The first one is often that it's a difficult mark, a complex mark.

In other words, a finger mark that might have been twisted or distorted

or laid on an unusual surface that affected it.

And the second one is that they tend to be in very high profile cases.

The McKie case was typical.

The crime scene mark was complex.

Two fingerprints overlapped each other...

..and the investigators were under pressure,

surrounded by intense media interest.

Under these circumstances,

analysts can start to see what they expect or hope to see,

not what's really there.

It isn't very scientific,

but it is entirely human.

In 1905,

Detective Inspector Collins made his match

and went unchallenged -

and no analyst has ever suggested since

that the fingerprint evidence in the Stratton brothers case

was anything other than a decent fit.

The Stratton brothers were executed in Wandsworth

at 9am on 23rd May, 1905.

It was the first murder case in a British court

in which fingerprint evidence had helped secure a conviction.

There's no doubt at all.

We've seen it for ourselves that science continues to deliver

new techniques for the capture of evidence.

But machines and chemical tests don't interpret evidence.

People do -

and they can get it wrong.

Next time, I'll reveal why the murder weapon

is such a vital piece of evidence.

Firearms,

poisons,

knives...

Weapons can hold all the clues needed to catch a killer.

Delve deeper with the Open University

and find out more about the science behind forensics.

Go to:

..and follow the links to the Open University.