French railway murders: More Clouseau than Maigret

Flaws in the hunt for the killer of Isabel Peake and two others have put French policing in the dock

On Tuesday morning, a gendarmerie detective in Bourges, in the centre of France, received a routine piece of information from the police in Dijon, 150 miles to the east. A woman had been savagely stabbed on an overnight train. One of the possible suspects was a young itinerant, a modern railway hobo, who specialised in travelling on night trains without a ticket, rather than hopping on freight trains.

Alarm bells rang, metaphorically, in the gendarmerie office in Bourges. The detective was one of the 10 officers working full-time on the unsolved murder of the British student, Isabel Peake, who was thrown from an overnight express just south of Chateauroux on 13 October. The name of the suspect circulated by the police in Dijon, was Sid Ahmed Rezala, 20, of no fixed abode, originally from Marseilles, with a criminal record for violence and rape - although in his case, the rape of young boys.

Rezala's name already figured on a list of 30 potential suspects, or material witnesses, identified in the search for Isabel's killer. Bingo. A breakthrough in the Peake case at last.

Despite hundreds of hours of detective work in the last two months, which is far more than normal on a case of this kind, despite an appeal on the internet - the first in French police history - the trail of Isabel's killer had gone cold.

The team of two investigating magistrates and 10 gendarmes had been working under intense political pressure. The British press had been howling about the number of unsolved British murders in France; a complaint partly justifiable, partly unfair, as we will see later. The Justice Ministry in Paris had made it clear that it wanted Isabel's murderer found rapidly.

Within hours two gendarmes from Bourges were in Dijon, comparing notes with detect- ives. The two forces in French policing don't always co- operate so well. Rezala became not just a possible suspect but the number one suspect for both night-time train murders.

On Monday night he had been found travelling without a ticket on the train from Calais to the Mediterranean coast on which a 36-year-old French woman, Corinne Caillaux, was later brutally stabbed. He had no money to pay the spot fine and had been forced to show his identity card to a ticket inspector to avoid arrest. He had been verbalise, or charged, 42 times for this same offence in the seven months since leaving jail.

Rezala was known to have been in Limoges on the day before Isabel began her fateful journey home to Staffordshire in October. He was ejected by ticket inspectors from a train from Paris at Limoges railway station on 12 October, 12 hours before Isabel caught a train in the opposite direction.

A coincidence? It's hardly likely. His description also fitted, not exactly but closely enough, the composite sketch of a young man seen chatting to Isabel as she climbed into the 3.08am train to Paris that same night.

On the surface, this was good and rapid police work. The gendarmes in Bourges had identified Rezala as a Peake suspect because they had taken the trouble to obtain the names of all people recently caught travelling without tickets on the Limoges line. If the matching name had not come up, police in Dijon might not have made Rezala their number one suspect as swiftly as they did.

And yet several questions beg to be asked without surrendering to the knee-jerk reaction of those in the British press, which seems to be perennially convinced that French police work is more Clouseau than Maigret.

Why did the Peake investigators not make more effort to track down Rezala before the second murder on Monday night? Was it not staring them in the face that he was the likely killer? Here was a man with a record for violence, who was known to be at Limoges railway station on the day before Isabel travelled. He answered the description of someone seen boarding the train with her. He was a man who haunted night trains.

Although Rezala was a suspect, he was one of 30 possible suspects or material witnesses, identified by investigators as having been on the train or near Limoges railway station that day. He was not regarded as a more likely candidate than several others. Efforts were being made to trace him, but nothing urgent or intensive.

On Friday afternoon a third victim was discovered in Amiens. The decomposing body of Emilie Bazin, 20, was found under a heap of anthracite in the cellar of a building where Rezala had been living. She had last been seen with a man answering Rezala's description on 29 October.

If the detectives and judges in the Peake case had been quicker, could Emilie Bazin's murder have been prevented? Probably not. A murder inquiry did not begin in Chateauroux until 26 October because of delays - partly justifiable, partly incompetent - in identifying Isabel's body. A more rapid response by the Peake murder team might, however, have saved Ms Caillaux.

Even more incomprehensible is the slow reaction of investigating magistrates and police in Dijon after Rezala became the leading suspect on Tuesday morning, within hours of the discovery of Ms Caillaux, dying from her wounds in a toilet on the Calais train.

Rezala's father later told police that his "psychologically disturbed" son had been present at his house in Marseilles from late Tuesday morning to late Wednesday afternoon. No arrest warrant was issued by the investigating magistrate in Dijon until Thursday morning at 11am. Police had been watching the house on Wednesday but had no power to search or arrest. They appear to have missed Rezala when he left at around 6pm on Wednesday. Why were the investigators so slow?

It is worth re-examining the standard British press indictment of French policing. It is traditionally stated that there have been 27 murders of Britons in the past two decades and that only two have been solved. It is alleged that in the past five years, there have been three murders of Britons in France and that none of the culprits has been found.

The figures are false and self-justifying. They include only "newspaper murders" - those killings which have been lurid enough to make newspaper headlines, which by their nature are often the hardest to solve. The real figure for the past five years, including less newsworthy murders, is seven Britons murdered in France, of which four have been solved. If Rezala is caught, and his guilt confirmed, the clear-up rate for British murders over the past five years will be five in seven, which is reasonably good. In the two outstanding cases - the rape and murder of the 13-year- old schoolgirl Caroline Dickinson in 1996, and the murder of the 54-year-old businessman Roderick Henderson on the Champs-Elysees in 1997 - there is undeniable evidence of police and judicial lethargy and incompetence in the early stages of the investigations. There is now, however, a promising new lead in the Dickinson case.

The French clear-up record for non-domestic murders - less than 50 per cent - is disturbingly low but there is no evidence that it is any worse in cases involving foreign victims. The 30 per cent clear-up rate for all crimes in France is disastrous. Many reasons for this are offered by French criminologists: a demoralised police force and gendarmerie; tensions and lack of co-operation between the two kinds of police; and the practice of having examining magistrates rather than detectives running investigations, which makes the system over-dependent on the competence and character of individuals not trained in police work.

Magistrate-led inquiries can be extremely effective for crimes such as fraud or political corruption. They bite hard as soon as there is a clear suspect who can be hauled in. Magistrates are trained in interrogation, not in detection. Complex murder cases without an obvious suspect require speed, intuition, local knowledge and painstaking inquiries. In France they often seem to proceed on a laborious, broad front rather than with any rapid and clear purpose in mind.

The Peake inquiry has made considerable progress in a case that was extremely difficult. Thirty suspects or missing witnesses had been identified; the special gendarmerie team of 10 was working diligently but slowly to trace all of them and eliminate them one by one. It had seen no special reason to focus on Rezala.

His previous sexual offences were with boys. The train from which Isabel was thrown was searched for his fingerprints and none were found. In hindsight, it was a mistake not to have pursued him more vigorously.

The ponderous response of magistrates and police in Dijon, which allowed him to slip through the net in Marseilles, is harder to justify. This seems to have been a straightforward case of bureaucratic slowness and poor co-ordination between police forces. It seems that the investigating judge in Dijon failed to issue an arrest warrant until the Thursday because police in Dijon had decided to pursue local angles before travelling south. They wanted the glory of making the arrest. They expected the Marseilles police to keep Rezala under surveillance for them but he got away.

There is still every chance that Rezala will be caught. Investigators believe he does not have the contacts, money or shrewdness to remain on the run for long. But what if he kills a fourth time before he is caught?