The Pembrokeshire Murders: Catching the Bullseye Killer

by Steve Wilkins

  In addition to the vast number of burglaries, there were also four robberies. The first was in 1985 and from then on, the level of violence increased, culminating with the Sardis robbery in 1996. It had been noted that the robberies were very similar; lone women in isolated rural properties. Prior to the Scoveston Park murders the robberies did not involve a firearm, just excessive violence to control the victims. The offender was alone and always instructed the women to cover their heads with clothes or bed covers. Significantly, robberies committed after Scoveston Park always involved a sawn-off shotgun. In April 1985, a robbery occurred at an isolated house called The Briars, at Crossways in Milford Haven. The telephone was disabled, the occupant, a lone female, was threatened with violence and jewellery was stolen. The property was later recovered at Scoveston Fort, a disused military structure from the Napoleonic wars, very close to Scoveston Park Farm. In September 1985 there was a second robbery at a similar house called Primrose Hill Cottage, in Rosemarket. Entry was gained to the house and a lone female was threatened with violence. Her hands were tied behind her back with rope that the attacker had brought with him. Jewellery was stolen. In September 1994, nine years after the murders of Richard and Helen Thomas, an armed robbery occurred at an isolated house at Westfield Cottage, Sardis. A lone female occupant was threatened with violence and blindfolded. Her hands were tied behind her back with rope brought to the scene. The offender was wearing a balaclava and was armed with a shotgun. He disabled the telephone and jewellery was stolen from the house. Then in November 1996 came the armed robbery at Westwinds, in Sardis, when Sheila Clark was confronted by a man wearing a balaclava and carrying a double-barrelled sawn-off shotgun. It was a carbon copy of the other robberies. Significant research was carried out on the burglaries and it was considered the MO was very similar in most of these offences:

  • Properties were approached across fields

  • Evidence of surveillance by the offender

  • Hedge line had been used as cover

  • Stolen items had been discarded along the hedge line

  • Wire fences had been cut on the route taken by the offender

  • Keys had been stolen from the houses

  In the case of the robberies that had been committed in this area, the properties were very similar to those targeted in the burglaries but in addition:

  • Use of a shotgun to control the victim was evident

  • Victims were lone females

  • Hands of the victim were tied behind their back

  • Rope was brought to the scene by the offender and then taken away, if possible

  On a number of occasions police dogs had tracked the offender across fields and hedge lines. Fences were cut to allow easy escape at locations only known to him. Also, at some of these locations, he had left the top strand of wire intact, only cutting the lower strands, no doubt to injure any would-be chaser, unable to see it in the dark. It was also clear to the officers who investigated these escape routes that the North Pembrokeshire night sky was illuminated by many significant landmarks. The oil refineries, which lined the Milford Haven waterway and the Pembroke Power Station stack, provided beacons on the horizon from which he could fix his location at any one time; this was someone who knew the area like the back of his hand.

  Back in 1998, Cooper had fallen into the parameters of the Huntsman house-to-house inquiries and was required to give a voluntary sample of his DNA. He refused and was obstructive but remained an ‘interesting’ character and when the various dog tracks and cut fences were reviewed it was clear that he lived at the very epicentre of the criminal activity. On one occasion in March 1996, dog handler Mark ‘the bark’ Jenkins had followed a track from a burglary at Castle Pill Crescent, Milford Haven, which took him to the entrance of a field opposite Cooper’s home, and he had actually spoken to him. This offence took place a few days before the Milford Haven sex attacks and was within sight of the scene. DS Steve Matchett and DI Jim Morris were suspicious and went to his home. Cooper was uneasy and officers noticed property in his house, including videotapes, that were identical to items stolen in a number of the burglaries. In January 1998, following months of research and planning by the Huntsman team, John William Cooper was arrested at his home address of 34, St Marys Park, in Jordanston. His house, surrounded by fields, sat at the hub of the circle of offences. Fingertip searches were carried out at both his house and at the homes of his relatives, as well as on the surrounding land, including the tracks that the various dog handlers had previously followed towards his home address. It was a huge operation lasting more than thirteen days involving earthmovers and dozens of officers. In all, more than three thousand eight hundred items of property and exhibits were found in and around Cooper’s home. These included:

  • ROPES – Numerous pieces of rope were recovered, some of which appeared to have been constructed to be used as a hand restraint.

  • JEWELLERY – A range of items were recovered from his house, attic and hedges. Items were also found at the homes of his relatives which were later identified as coming from the burglaries and robberies. Some of these items had been burnt and discarded, probably with a view to preventing forensic testing and identification.

  • AMMUNITION – Two shotgun cartridge boxes and another box containing various loose cartridges were found under a duck run in the garden. The cartridges were similar to the types of cartridge that the ballistics experts had indicated were involved in both the Scoveston and Dixons murder enquiries (Eley and Maionchi/SMI). One of the boxes contained an old style cardboard cased cartridge, which again appeared to be similar to the type recovered from Richard Thomas’ gun cabinet back at Scoveston in 1985.

  • SHOTGUN – A well-oiled shotgun with its stock sawn off was found under the duck run at his home, covered by over a foot of soil. This had been preserved by being soaked in gun oil, placed in black bin bags and then enclosed in a sewage pipe. This was later identified as being stolen from a burglary at Castle Pill in Steynton on 1st March 1996.

  • GUN PARTS – Components of shotgun were found discarded in the hedge at 34, St Marys Park; one firing mechanism could be described as belonging to a hammer action shotgun. In the Scoveston enquiry, it is suspected that Richard Thomas’ gun or guns were stolen from the house and even though he owned a gun cabinet and other associated shotgun items, no shotguns were ever found.

  • KEYS – Over five hundred keys were recovered from Cooper’s possession, many of which were found in a cesspit on his land. Some of the keys would later be matched to mortise keys stolen from some of the burglaries. It’s important to note that the keys belonging to Richard Thomas’ Rover car were missing when police arrived at the scene of the house fire at Scoveston.

  During the investigation Cooper was linked to seventy-two offences of burglary and robbery. However, following early liaison with the Crown Prosecution Service, he was eventually charged with thirty-four offences consisting of thirty burglaries, one burglary with intent and three robberies. When he was questioned about the Scoveston and Coastal Path murders the officers were wrongly advised that Cooper would respond to their authority. Instead he refused to answer any questions and the interview lasted just thirteen minutes. What the officers didn’t know at the time, was how important to me those interviews would become almost twenty years later in influencing our approach to Cooper.

  We were now at a stage where we wanted to move evidence from the Huntsman inquiry into the new Ottawa investigation without formally naming Cooper as a suspect. The investigation was growing into a hybrid of a thematic forensic review and re- investigation of crimes over a twenty-five year period in North Pembrokeshire. We were required to go through a formal linking process and review the terms of reference for Operation Ottawa. It was decided because of the scale of the investigation and in line with ACPO policy that I would take over the role of SIO with Andy John as my deputy. We were now totally immersed in the investigation and I could not have been happier leading this
exceptional group of officers and support staff.

  In line with our decision to take an objective and independent approach Andy and I agreed to engage a new forensic service provider to carry out any examinations. The original material and exhibits had been examined by the Forensic Science Service, which at the time of the original offences had the monopoly on forensic examinations. The intention was to submit the final schedules to an experienced scientist who would review the schedules and give advice on any further examinations or forensic opportunities. The National Centre for Policing Excellence (NCPE) provided the ideal person in the shape of Joanne Ashworth, Head of Physical Science. Joanne was well known in the homicide arena and had been instrumental in developing many approaches to the recovery of trace evidence in major crime investigations. The examinations were to be carried out by LGC, a private company which had a growing reputation after their groundbreaking work on cases such as the murder of Rachel Nickell on Wimbledon Common.

  With the Huntsman exhibits now of great interest to Ottawa, all the exhibits needed to be found and logged in the same way as the other material. The issue of Cooper being named as a suspect had to be addressed as Ottawa and Huntsman were now so intertwined that I felt it needed transparency. There was another school of thought that suggested naming him would leave me open to accusations of tailoring the investigation to fit. This held no fears for me because I was sure that the forensic review process was not directed at any one individual and would either implicate, eliminate or identify another suspect. On 18 December 2006 the decision was made to name John William Cooper as a suspect in Operation Ottawa.

  The Prime Suspect

  WITH COOPER NAMED as a suspect I was able to focus on his long history of offending and include in my review all the forensic evidence from Operation Huntsman. It was clear by now that we were in for the long haul so I decided to tap into the expertise of Des James from NCPE. I had known Des for years; he was a retired Detective Superintendent from Gwent Police and very experienced in major crime investigations. Des was a gentleman and had a lovely way about him; he could make a suggestion in such a way that you felt it was your own idea. By now we had put together a presentation that showed the Ottawa crimes overlaid with Cooper’s convictions; it was crude but a great briefing tool. Des came to Carmarthen for an initial chat and we showed it to him. He watched in total silence and at the end I looked at him and said, “Ok Des, in your professional opinion as the Regional SIO advisor, what do you think?” Des looked at me and pondered his answer, “He’s fucking done it Wilkie!” I must say that I was hoping for something a bit more scientific, but this was the start of an excellent relationship with Des who became a good friend and advisor to the team.

  The terms of reference for Operation Ottawa needed to be adjusted to include Cooper as a suspect and my policy book named him as such. The changes to the terms of reference were simple:

  • Investigate further lines of enquiry identified during the forensic review.

  • Progress new investigative opportunities in respect of the suspect John William Cooper.

  • Prepare for an interview with the suspect John William Cooper or any other person identified from the evidence gathered.

  • Exploit any advances in investigative techniques and legislative changes in an effort to gather any other admissible evidence that would support a prosecution of the offences under review.

  The last point was unusual and would significantly impact on the investigation and court process. There had been a change in the law that allowed Bad Character Evidence to be admissible as part of the prosecution’s evidence of guilt. It has been a great source of frustration to the police, and I believe the general public, that during a trial they are not made aware of certain information about the defendant when it is directly relevant to the case and tended to show that the defendant had a propensity to behave in a certain way. In simple terms, prior to 2005 a defendant’s bad character would only be admissible in certain very narrow circumstances when they made the character of another witness a point of issue, and then only with the leave of the court. As you can imagine, defence lawyers steered well clear of such a position. In 2005, the Bad Character provisions changed for the better, allowing the prosecution to introduce a defendant’s bad character through so called ‘admissibility gateways’. In essence if a career criminal demonstrated through previous convictions that they carried out certain behaviour during offending or committed certain types of offences, the gateways allowed these facts to be given as evidence to the jury.

  In terms of the exhibits we needed to keep it simple. My initial thoughts were to build up comprehensive schedules of material and physical evidence and establish the provenance of each item, assess and consider its relevance and importance to the investigation and identify a system of prioritisation before submitting for forensic examination. We created a ‘one stop shop’ database of information relating to forensic exhibits that meant that we could view the original reports immediately. The database also contained images of the exhibits, either an original photograph or a more recent image prepared during Ottawa.

  This method of recording information gave the team a comprehensive research capability to interrogate the system. It also served another purpose in allowing the team to start to build up continuity schedules for the exhibits, mapping the journey each one had made. It was established that all investigations except the Scoveston Park review, which had been done on a card index system, had been recorded on the HOLMES computer system. Many versions of HOLMES had been developed and we experienced considerable challenges in trying to convert the old systems into the new. The work done by Glyn Johnson to crack this problem was exceptional and would become national best practice. I was very impressed.

  All the recovered exhibits were eventually transferred to the forensic laboratories of LGC in Oxfordshire where they would undergo fresh and independent examinations. A corresponding inventory was then recorded by LGC on receipt of the items to ensure continuity of the exhibits. It was also likely that some exhibits had been destroyed, returned to their owners or indeed lost. It was therefore important to establish if any exhibits were missing and to conduct the necessary searches to retrieve any that were outstanding. Any such searches needed to take place in forensic conditions and by specialists. The services of Police Search Advisor (POLSA) Inspector Nigel Hayes, as well as search officers, were used to ensure that the main historic storage venue, the firearms range in Haverfordwest, was clear of all relevant exhibits relating to our review. Again the forensic integrity of any exhibit was maintained and recovered in line with the original strategy.

  It was soon apparent that many exhibits had been destroyed with the authority of the then SIO, because forensic opportunities, especially DNA, were very limited. To review the submission of items to date and to make recommendations on any further forensic submissions the team needed access to various source documents to understand the ‘forensic’ history of each exhibit as well as its provenance. These documents included:

  • Major Incident Room documents

  • Scientific Support submission forms

  • Scientist General Examination Records

  • Original Photographs

  • Exhibit Books

  The original scientists’ notes from the FSS gave the team the opportunity to gather considerable and reliable information on the exhibits, some of which would have not been reported back to investigators at the time. The notes were light on the significance of the fibre evidence because at the time gathering it was very expensive and time-consuming. Some of the clothing belonging to the Dixons had gone missing within the Forensic Science building at Chepstow only to re-surface later when it was found in a storage bin. The exhibits were divided into the following categories:

  • Firearms / Shotguns

  • Ammunition

  • Ropes

  • Keys

  • Clothing

  • Jewellery

  • Sexual Elements<
br />
  • Forensic Samples

  • Miscellaneous exhibits

  Further research on the Eley ammunition threw up an unexpected contact. Graham Morris, who had been managing director of the Eley-Hawk company, was able to provide information on the dates of manufacture of certain cartridges as well as indicating the type of wadding and shot they contained just by looking at them. Some of the ammunition was actually dated back to the 1940s. There follows a summary of the significant findings:

  FIREARMS/SHOTGUNS

  It is suspected that a shotgun(s) were stolen from Scoveston Park and possibly used as the murder weapon. Both Richard and Helen Thomas suffered shotgun wounds as did Peter and Gwenda Dixon. Ballistics experts indicated that the murder weapon used during the Dixons attack was likely to be a double-barrelled sawn-off shotgun. Significantly, during Operation Huntsman, various shotguns and component parts were recovered from Cooper’s home address in Jordanston. A well-oiled shotgun was concealed under a duck run. This shotgun was identified as having been stolen from a local burglary in March 1996. The stock of the gun had been sawn off and a screw placed in the stock with a small washer attached. Other sections of older shotguns (barrels, trigger mechanisms) were also recovered from a hedge at Cooper’s home. In 1996, during the Sardis robbery for which Cooper was convicted, a sawn-off shotgun was discarded. It had been modified, including the introduction of a rope lanyard attached to the stock and barrel by metal clips and screws. A missing screw from this shotgun was eventually matched to a screw recovered from a container in Cooper’s garden shed. A similar robbery happened in the Sardis area in 1994, in which the victim was tied up, involved the offender brandishing a sawn-off shotgun.