England and Wales Court of Appeal (Criminal Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CRIMINAL DIVISION)
ON APPEAL FROM THE CROWN COURT AT BRISTOL
(MR JUSTICE AULD)

		Royal Courts of Justice Strand, London, WC2A 2LL
		4th October 2005

Lord Justice Scott Baker:

Introduction.

1. On 30 March 1994, after a lengthy trial before Auld J and a jury in the Crown Court at Bristol, Jonathan Probyn, the appellant, was convicted of the murder of his wife Tanya. She had been reported missing on Wednesday 3 September 1992. On the afternoon of Tuesday 8 September 1992 police divers discovered a Renault motor car upside down on its roof in the River Severn, about 100 to 150 metres south of some tyre marks where it was suggested the vehicle may have entered the river. The vehicle was 21 yards out from the bank and in 20 feet of water. Tanya's body was in the back of the car. The appellant was the last person to see Tanya alive on Wednesday 2 September 1992. The Crown's case was that the appellant, having made Tanya unconscious, put her in the car with intent to kill her and propelled the car into the river at Stonebench, near Gloucester, causing her to drown.
2. The verdict of the jury in March 1994 was unanimous. The appellant was refused leave to appeal by the Full Court. Kay J, in giving the court's decision, said:

"A great deal of the evidence, particularly the pathologists' evidence, was the subject of dispute, so were the inferences that could properly be drawn from the evidence. All these matters were for the decision of the jury and no one else. We can see no basis for concluding that the verdict was unsafe or unsatisfactory."

3. There had been an earlier trial that had begun the previous October, but after seven days, leading counsel for the defence told the court that after he had received revised instructions from the appellant he and junior counsel were placed in a position of professional embarrassment and were unable to continue to represent him. The judge concluded that it would not be possible for the trial to continue with fresh counsel, nor would it be appropriate for the appellant to represent himself. The jury was discharged and a retrial ordered.
4. The conviction now falls for reconsideration by this court following reference by the Criminal Cases Review Commission ("the CCRC"), the CCRC being of opinion that there is a real possibility the conviction will not be upheld. Central to the reference is an invitation to the court to accept the fresh evidence of David Parry, a principal consultant in the Investigations and Risk Management Group, and team co-ordinator for the Accident Investigation and Reconstruction Team at the Transport Research Laboratory Ltd ("TRL"). He specialises in traffic accident reconstruction and, in particular the use of computer systems to analyse and visualise traffic accidents using reconstruction and simulation programmes.
5. Mr Parry's written report has been considered by Dr Ashton, who has been instructed by the Crown Prosecution Service. There is no relevant disagreement between the two men. Indeed they have helpfully provided a joint report setting out areas of agreement.
6. The fresh evidence relates to the reconstruction of what happened or may have happened to the vehicle immediately before it entered the river. The appellant's argument is that this reconstruction is inconsistent with the appellant having propelled the vehicle into the river. Had the jury had the benefit of this fresh evidence it may not, it is submitted, have convicted him and therefore his conviction is unsafe and should be set aside.

The background facts.

7. The appellant married Tanya in May 1985. They had two daughters, Natalie and Bryony, who were aged 7 and 4 respectively at the time of their mother's death. Tanya was 27. The appellant was an engineer employed by British Gas. He had for some years suffered with intermittent pain in his left knee and back and this had resulted in his taking a considerable amount of sick leave from his job.
8. The marriage appears to have been happy until January 1992 when Tanya began an affair with Robert Cross, a local boat builder. Tanya and Mr Cross planed to move to Derbyshire taking with them Natalie and Bryony, but the appellant obtained an injunction preventing this.
9. There was a point at which Tanya decided to remain with the appellant, but just before Easter 1992 she started seeing Mr Cross again. The appellant discovered this and Tanya left the matrimonial home, Grove End House, and went to live with Mr Cross, who had left his wife, at Vaisey Fields, which was a few minutes drive away.
10. According to Mr Cross, his relationship with Tanya was happy. The appellant suspected Mr Cross was being unfaithful to Tanya and clandestinely recorded some conversations that took place between Mrs Cross and Carol, with whom Mr Cross had been having an affair some years prior to his affair with Tanya. These tapes have been referred to as "the Carol tapes."
11. Tanya applied to the Court for a residence order for the children. A preliminary hearing was fixed for Thursday 3 September 1992. The appellant said his relationship with Tanya was improving by August 1992 but that there was no question of a reconciliation as he was having a relationship with another woman.
12. By prior arrangement both Natalie and Bryony stayed with the appellant on 2 September. At about 10.10 pm the appellant telephoned Tanya and said that Natalie was upset and wanted to see her. He accepts this was a ploy to persuade her to come. Tanya left Vaisey Fields at about 10.25pm in a white Renault 19 and drove to Grove End House which would have taken approximately 10 minutes, depending on her route.

The Crown and defence cases at trial.

13. The Crown's case was that Tanya arrived at Grove End House at about 10.35 pm. Shortly thereafter the appellant assaulted her, rendering her unconscious. He put her unconscious body into the Renault and drove it to Stonebench, a rough grassy bank on a bend on the River Severn about two minutes drive from his house. There he propelled the car, with his unconscious wife inside it, into the river and she drowned. The appellant then returned to Grove End House on foot, a journey that would have taken about 15-30 minutes.
14. Stonebench is a well known spot from which to view the Severn Bore, a phenomenon caused at high tide by great rush of water upstream.
15. Both the clock in the car and Tanya's wrist watch had stopped at or shortly after 11 pm. The Crown relied upon this as proving the time at which the Renault entered the river.
16. The appellant first gave a detailed account of events in a telephone conversation with Tanya's mother which began at 11.50 pm that night. He repeated that account to a number of other people, to the police and, with one significant addition, in his evidence at the trial. The significant addition related to a tape recording of a conversation he said that he had had with his daughter Bryony in which – according to the appellant - Bryony described how she had been sexually abused by Mr Cross. The first time the appellant referred to the so called 'Bryony tape' was when he gave evidence at his trial before Auld J in March 1994. We shall refer to the 'Bryony tape' in more detail when we recount the appellant's evidence.
17. What was said by the appellant to Tanya's mother included the following:

(1) Tanya had arrived at about 10.30 pm and almost immediately there had been an argument. This centred on Mr Cross and the appellant's allegation that Mr Cross was still seeing Carol. Tanya challenged him to prove this. The appellant played her a number of passages from the clandestine recordings he had made of the conversations between Mr Cross and Carol. He said Tanya became very quiet and seemed to realise that what he was saying was the truth.

(2) Tanya then attacked him, kicking and grabbing him by the hair. They fell against a wall. The appellant said she might have hit her head against it. Tanya then calmed down and they went into the kitchen. He played her more of the 'Carol tapes'. Tanya asked if there was any chance of resurrecting the relationship with him and the appellant said he did not know if he wanted her back. They talked about the possibility of Tanya renting another house. The appellant went to the shed and retrieved a plastic bag containing some of her clothes. He gave this to her together with a red Adidas sports bag containing some of his old work clothes and a clean towel.

(3) Tanya was in a hurry to go. She walked to her car and they argued again. She pushed him back; he pushed her back. She fell on her bottom on the drive. He tried to help her up. They then went to the car and he put the two bags inside. He returned to the house. The appellant then noticed Tanya had not driven off, so he went back out to the car. He squatted down beside the open driver's door and tried to perch on the edge of the driver's seat. He then went round and sat on the front passenger's seat. They argued again. The appellant put the car lights on because he thought he saw some eyes glinting in the dark. In doing so he tried the other switches in the car. They had a final row and Tanya drove off at speed shouting at him through the driver's open window. She left at around 11.30 pm. He had a bath and washed some of the clothes he had been wearing that day. Then, at 11.47 pm he began to record a television programme for a friend. From 11.50 pm to 12.27am he spoke on the telephone to Tanya's mother. About 12.30 am the appellant telephoned Vaisey Fields. He said he put the telephone down when Mr Cross answered with a sleepy voice, because this suggested Tanya had not returned.

18. Between 12.45 am and 1 am Mr Cross drove to Grove End House looking for Tanya but could not find her. He returned home and telephoned the appellant who told him Tanya had left about 11.30 pm and that she was very angry when she left. Mr Cross then drove back to Grove End House looking for Tanya but did not find her and returned home between 2 am and 2.30 am.
19. On the morning of 3 September the appellant telephoned Mr Cross asking where Tanya was because she had not collected Bryony as arranged. The appellant went to the court for the preliminary hearing about the children, but the case was adjourned on Tanya's non attendance.
20. Tanya was reported missing to the police by Mr Cross at 4 pm that afternoon. WPC Collings went to Grove End House and saw the appellant who said he had last seen Tanya at 11.30 pm the night before. In his evidence he said he had told the WPC: "11pm, half past 11, call it 11.30 pm." The following day in a witness statement he told the police he estimated Tanya had left the house at 11.30 pm. In evidence he described his witness statement as a rough account of what had happened.
21. On Saturday 5 September the police searched Grove End House. During a discussion with Chief Inspector Preston the appellant added three matters to the account he had given in his statement. First that he had recorded a television programme, second that he had done the washing and third that he had telephoned Mr Cross at about 12.30 am. The appellant said that during the course of the conversation with the Chief Inspector of Preston he admitted he and Tanya had had a fight but he said that he had not caused her any harm.
22. On Sunday 6 September police divers went to Stonebench. PC Evans noticed tyre tracks leading from the road towards the water. Between the tyre tracks there were two pieces of plastic; there was another piece near the water's edge. After the vehicle was recovered it was found that they together matched moulding missing from the underneath of the front bumper. The officer also noticed that a small willow sapling on the sloping riverbank had been bent over and damaged. The scene was photographed and, while at Stonebench, police officers noticed the appellant driving past. In evidence the appellant said this was the route he usually took when visiting his mother. Although the river was searched for eight hours the vehicle was not found. It was eventually found on the afternoon of Tuesday 8 September between 100 and 150 metres down stream of the tyre marks.
23. On the afternoon of Sunday 6 September DS Hobday again spoke to the appellant at Grove End House. After he had left, the appellant was seen at Stonebench walking slowly around the grass area looking at the ground. He said the police had told him they were diving at Stonebench; he was curious and had stopped to have a look.
24. The Crown's case at the trial depended at circumstantial evidence. It was submitted that if the circumstances were taken together they pointed unerringly to the appellant having propelled the unconscious Tanya in the Renault into the river albeit it was not possible or indeed necessary to show the precise mechanism by which he had done so. Both pathologists agreed that the cause of death was drowning. The defence case was run on the lines that it was impossible to rule out various possibilities that the vehicle may have gone into the river in some way other than at the instigation of the appellant. These were accident, suicide or that someone else murdered Tanya. The appellant's case was that he knew nothing about how or why the car went into the river and he was not present when it did so.

The vehicle and the scene: factual and expert evidence at trial

25. At the trial there was expert and factual evidence about the state of the vehicle when it was recovered from the river and about the scene at Stonebench where it was said the car went into the river. The Crown called Dr Grant and Police Inspector Bottomley.
26. Dr Grant went to Stonebench on 9 September 1992. He described the area as rough ground partly covered with grass, weeds and other vegetation. From the road the area of ground was generally flat for a short distance, rising to a small hump with some rocks, stones etc protruding. The ground then fell gently to another relatively flat area on the approach to the bank of the river. The relatively flat area consisted of grass and then an area of tyre-ploughed mud followed by flat, bare, fairly firm, sandy mud turning briefly to grass just short of the abrupt commencement of the steeply sloping riverbank. He noticed the tyre tracks leading from the road that had been seen by PC Evans three days before. He examined the start of the tracks close to the road and saw no evidence of skidding to suggest that the car had swerved off the road at speed and out of control.
27. When the Renault was inspected at the forensic science laboratory, amongst the matters noticed were the following:

• The cassette player was switched on.

• The handbrake was off.

• The ignition key was in the engine running position.

• The gearlever was in neutral.

• The front windows were open.

• The seatbelts were retracted but functioning.

• The clock showed 11 0'clock.

• The stalk light control was off.

• The bumper moulding was damaged.

28. Police inspector Bottomley attended the scene on 14 September 1992; there was evidence that there had been flooding at Stonebench on 10 September. He identified the tyre tracks and prepared a scale plan. At the trial he described the route to Stonebench with reference to an album of photographs. The tyre tracks went from the road to the ridge but thereafter there were no clear tracks. The tracks were consistent with having been made by Pirelli P2000 tyres which were the type of tyres on the Renault. There was an acceleration mark in the mud and grass starting 0.4 metres from the road surface and continuing for a length of 3.1 metres. It was, on his evidence, possible to move the vehicle without acceleration across the bank in first gear but it stalled in second gear. It was possible to tip the Renault over the edge of the riverbank. On 14 September 1992 he did not notice any disturbed stones on the route across the bank.
29. It was Dr Grant's evidence at the trial that if the vehicle grounded the gear mechanism could be knocked out of second and fourth gear into neutral but not out of first or third gear.
30. Inspector Bottomley was present when Mr Lorriman carried out a survey of the scene on 12 February 1993.
31. At the time of the original investigations tests were carried out on behalf of the prosecution using a Renault 19 over the actual riverbank following the route it was believed the Renault had travelled on 2 September. Those results were made available to the defence in the unused material and were referred to in the course of the cross-examination of Inspector Bottomley. They showed that the Renault would ground at speeds of 10 and 20 mph. The prosecution tests were concerned with:

(i) Whether it was possible to propel the car from outside using the ignition in first or second gear – which it was.

(ii) Whether the car could be tipped over the edge of the riverbank – which it could.

(iii) The speed of the vehicle leaving the roadway and whether loss of control was a likely explanation – which it was not.

Other evidence at trial.

32. After the vehicle had been recovered with Tanya's body in the back of it, Professor Knight, an eminent forensic pathologist, was instructed to carry out a post mortem. He concluded that Tanya had been alive when she entered the water but had died from drowning. The defence pathologist agreed with this.
33. On Wednesday 16 September the appellant was arrested. The same day the police found Tanya's bracelet in the drive at Grove End House under the off side of his white Vauxhall Astra which was parked there. The bracelet was found in the area where the appellant had told the police Tanya had fallen to the ground. It was accepted by both sides that it had a loose fastening and it could easily have come off when being worn.
34. The appellant was interviewed extensively over the next two days. He answered all the questions put to him and was charged with Tanya's murder on the evening of 17 September 1992.
35. Before analysing the evidence at trial and the new evidence, we set out the legal principles which we must apply.

The legal principles relevant to the appeal.

36. By section 9(1)(a) of the Criminal Appeal Act 1995 the CCRC referred the appellant's conviction for murder to the Court of Appeal. By section 9(2) the reference is to be treated for all purposes as an appeal against conviction by the appellant under section 1 of the Criminal Appeal Act 1968. The appellant then applied to the court to receive evidence not adduced at the trial, namely the report from Mr Parry whose evidence is agreed by Dr Ashton in the joint statement to which we have referred.
37. Although the appellant made several different submissions to the CCRC as to why his conviction was unsafe, the reference was made solely on the basis of the fresh evidence and how the car went into the river. That has been the single ground of appeal pursued by the appellant.
38. Section 23(1)(c) of the Criminal Appeal Act 1968 provides:

For the purposes of this part of this Act the Court of Appeal may, if they think it necessary or expedient in the interests of justice –

c) receive any evidence which was not adduced in the proceedings from which the appeal lies."

Subsection (2) provides:

"The Court of Appeal shall, in considering whether to receive any evidence, have regard in particular to –

(a) whether the evidence appears to the Court to be capable of belief;

(b) whether it appears to the Court that the evidence may afford any ground for allowing the appeal;

(c) whether the evidence would have been admissible in the proceedings from which the appeal lies on an issue which is the subject of the appeal; and

(d) whether there is a reasonable explanation for the failure to adduce the evidence in those proceedings."

39. There is no dispute that the fresh evidence meets criteria (a), (c) and (d) of subsection (2). The question is whether it meets criterion (b) i.e. might it afford a ground for allowing the appeal. This leads on to asking whether the conviction is safe, whether if the evidence had been before the jury at the trial the jury might have come to a different conclusion. The correct approach was set out by Lord Bingham of Cornhill in R v Pendleton [2002] 1WLR 72, 83.

"Thus the House in Stafford v the Director of Public Prosecutions [1974] AC 878 where right to reject the submission of counsel that the Court of Appeal had asked the wrong question by taking as the test the effect of the fresh evidence on their minds and not the effect that that evidence would have had on the mind of the jury. It would, as the House pointed out, be anomalous for the court to say that the evidence raised no doubt what ever in their minds but might have raised a reasonable doubt in the minds of the jury. I am not persuaded that the House laid down any incorrect principle in Stafford, so long as the Court of Appeal bears very clearly in mind that the question for its consideration is whether the conviction is safe and not whether the accused is guilty. But the test advocated by counsel in Stafford and by Mr Mansfield in this appeal does have a dual virtue to which the speeches I have quoted perhaps give somewhat inadequate recognition. First, it reminds the Court of Appeal that it is not and should never become the primary decision-maker. Secondly, it reminds the Court of Appeal that it has an imperfect and incomplete understanding of the full processes which led the jury to convict. The Court of Appeal can make its assessment of the fresh evidence it has heard, but save in a clear case it is at a disadvantage in seeking to relate that evidence to the rest of the evidence what the jury heard. For these reasons it would be usually be wise for the Court of Appeal, in a case of difficulty, to test their own provisional view by asking whether the evidence, if given at the trial, might reasonably have affected the decision of the trial jury to convict. If it might, the conviction must be thought to be unsafe."

40. Mr Owen QC, for the appellant referred us to R v Bowler 24 July 1997 (unreported) in which the Court of Appeal, presided over by Lord Bingham CJ, with some considerable hesitation, concluded that a conviction could not be regarded as safe where fresh evidence suggested it was possible, albeit unlikely, the deceased could have met her death by accident rather than at the hand of the defendant. We will return to this case later in this judgment.

Analysis: The nine features and the new evidence.

41. In summing up the judge drew the jury's attention to nine main headings or features of the evidence on which the prosecution relied. He pointed out that the appellant's case was that, although he accepted Tanya died from drowning in her car in the River Severn at or near Stonebench, he denied he was with her when the car went into the water. He did not know how or why it went into the water and invited the jury to consider various possibilities including accident, suicide and someone else being responsible for her death.
42. Before us, although other possibilities have neither been pursued nor abandoned, Mr Owen has really focused on accident.
43. The judge told the jury that the first four of the nine headings related to the issue of murder by someone, the next three to murder by the appellant and the last two fell into a different category. We shall deal with them in turn, examining how they are affected by the new evidence. The Crown's argument, is, as we have said, that taken together they point unerringly to the guilt of the appellant as found by the jury. The Crown submits that the fresh evidence is insufficient to disturb that conclusion. In short what the Crown are saying is that there was an overwhelming case based on circumstantial evidence against the appellant. That being so, it does not matter if there is uncertainty about precisely how the Renault entered the river. The Crown's case at trial – to put it at its lowest - was that the expert evidence did not show that the appellant could not have propelled the car into the river. The fresh evidence does not show this either. Thus the Crown says that there is no reason to think that a jury might reach a different conclusion at a new trial.
44. Accordingly there are three questions which are, or may be, raised by this appeal. The first concerns the Crown's contention that the circumstantial evidence at trial demonstrates that the appellant must have propelled the car into the river by one means or another. Is that evidence so overwhelming that it can only be answered by showing that such a course was completely impossible? If so, the second question is whether the fresh evidence is capable of showing that such a course was completely impossible? If the answer to the first questions is no, then the third question arises: is the fresh evidence, albeit not so strong as to make the Crown case impossible, sufficiently cogent to be capable of countering the circumstantial evidence?
45. With those questions in mind we examine each of the nine features and the position overall.

1. Injuries to Tanya

46. The evidence of Professor Knight was that some of Tanya's injuries could have been caused by an assault by someone rendering her unconscious before she entered the water. Injuries to her nose and eyes were typical of a blow or blows to the face; there was an extensive internal haemorrhage across the top and back of the head with satellite areas of haemorrhaging on either side, probably caused by several impacts by a very blunt or flat object such as a plank or the roof of the vehicle. Her hand injuries suggested she had been gripped. Dr Heath, the defence pathologist, thought all the head injuries were caused by the impact of the car on its bumpy approach to or it to drop over the bank. The appellant said the hand injuries could have been caused when he grabbed Tanya's hand in the hallway or when he grabbed her to pull her up from the gravel drive.

2. Seat belt.

47. Dr Grant examined the front seatbelts of the Renault and found them fully retracted. Mud marks on them indicated that they had been retracted for the whole of the time that the car was in the water. The belts were in proper working order. Mr Cross's evidence was that Tanya always wore a seatbelt when she drove and that she wore it 90 or 95% of the time when she was a passenger. Some six other witnesses confirmed she always wore a seatbelt when she drove and insisted on her passengers doing so, especially if they were children. The appellant's evidence was that she normally wore a seatbelt when she drove with the children but that she mostly did not when she drove without them. He said that she did not like wearing a seatbelt. The defence suggested it was possible that either she was not wearing the seatbelt on the night in question or that she had stopped at Stonebench for some purpose and unfastened it.

3. Car windows.

48. Both front windows were open when the car went into the river. Tanya did not normally drive with the windows open and the Crown submitted that the open windows were indicative of an intention that the car should sink rapidly. It was suggested that Tanya may have been smoking, either when she was driving or as she sat alone in the car at Stonebench. The appellant's evidence was that she did not like the smell of smoke on her clothes and may have had the windows open for this reason. However, as the judge pointed out, Mr Cross's evidence was that she had left her cigarettes behind on the kitchen work surface.

4. Car controls.

49. The prosecution case was that the position of certain of the car controls was consistent with it having been started by someone standing outside it. Dr Grant found the ignition key was in the engine running position, the engine lock functioned correctly. The lighting switch, which was a rotating stalk switch on the steering column, was in the "off" position. The handbrake was in the "off" position and functioned correctly. The footbrake worked, the steering functioned correctly and the pedals were muddy but normal. The gear lever was in neutral but functioned correctly. Because the lighting switch was a rotating stalk control, it was unlikely it could have been turned off accidentally. Because the gear lever was in neutral, if it had been knocked into that position by some impact under the car when it was moving, it would have been in either second or fourth gear. There was evidence that it would have been difficult to propel the car over the bank and into the river by placing it in second gear from outside since it would have been likely to stall.

5. Timing.

50. The Renault's clock and Tanya's wrist watch indicated by the time that they had stopped that the car had gone into the river at about 11pm. The Crown's case was that this demonstrated the appellant's story was untrue. First he had made repeated assertions to everyone that Tanya had left Grove End House at or about 11.30 pm and second, all that he said had taken place at Grove End House after her arrival at 10.30 pm simply could not have happened if the car was in the river at 11 pm or thereabouts.
51. The appellant's evidence of what had occurred was this. It repeated and to some extent expanded on what he had told Tanya's mother on the telephone on the night of 2/3 September. In summary, what he told the jury can be described as follows:

(1) Tanya arrived about 10.30 pm. A row began almost immediately about Mr Cross and his continuing relationship with Carol. Tanya told him to prove Mr Cross was still seeing Carol whereupon the appellant took her into the music room and played one of the 'Carol tapes' on the Hi Fi. He did so by flicking through it and fast forwarding the tape to play particular passages indicating meetings between Mr Cross and Carol and telling her at the same time intimate things which Carol had told him that Mr Cross had told her about Tanya. When the appellant gave evidence he was unable to remember or locate the precise passages from the tape that he had played.

(2) The appellant said Tanya seemed to realise that all the things he had been saying about Mr Cross were true. She said she had heard enough and was going to go. She walked out. He followed her with a Fisher Price tape recorder and the 'Bryony tape'. He said: "You'd better listen to this." In the hallway he played her the tape and then accused her of aiding and abetting a paedophile and of living with a child molester. She said it was untrue and set about him. She kicked him on the left leg just below the knee. She tried to hit him and grabbed him by the hair. He tried to stop her by holding her wrists. This occurred in the hallway. They fell against the wall and he thought that she hit her head.

(3) Tanya calmed down and they went into the kitchen. He played her some more of the 'Carol tapes', this time on a Loitron tape recorder on the work surface. He also, as he put it, had another go at her about Bryony and Mr Cross. He asked her if she was going to go back to "that paedophile". She replied that she could do with some time and space for herself. He asked if she was going to go away and said questioningly "I suppose there is no chance for us?" He said he did not know if he wanted her back and that whatever she did she had better get rid of Mr Cross. He told her he would do everything he could to prevent her from living with Mr Cross and unless she left him he would, with the help of Mrs May, ensure that she did not keep the children. There was then some discussion about Tanya possibly renting another house.

(4) The appellant then said he had some of her clothes in the shed outside. He left her in the kitchen and went to get them. He took from the shed two bin liners, one inside the other, containing some of her clothe. Whilst doing so he noticed a red Adidas bag which he knew contained some of his work clothes and a clean towel. He gave her the bag and its contents because he thought the bag and the towel would be useful to her if she went away for a few days. He did not think his old work clothing would be of any particular use to her but he just left it there.

(5) Tanya was in a hurry to leave. She went outside. The appellant followed with the bags. Another row developed. She pushed him; he pushed her. She fell down on the corner of the drive on her behind. She cursed him and she was crying. He tried to pull her up and there was an incident that he said could have caused the pressure marks on her hand. This incident happened in the area where the bracelet was eventually found.

(6) They went to the car. He threw the two bags in the back. There was a further argument. The appellant returned to the house. He looked out of the window and saw that Tanya was still there. He picked up the Fisher Price tape recorder still containing the 'Bryony tape' and returned to the driver's door of the car which was open. He squatted beside the door; they talked. Then he tried to sit next to Tanya on the driver's seat but was unable to, so he went round to the passenger's seat, got in and played her a bit of the 'Bryony tape'. He asked her why she had not gone and what she was going to do about the court hearing the following day. He said he had to get the children away from Mr Cross and would play the 'Bryony tape' at court if he had to. He also said he had tape recorded her that night admitting that Bryony had told her Mr Cross had indecently assaulted her.

(7) There was another row. Either just before that row or an earlier one the appellant saw some eyes glinting in the dark and flicked some switches to turn on the lights. They had a final row. Tanya either wound down the window or it was already down. She shouted at him and drove off fast, revving through the gears.

(8) The appellant said the happenings at the house and in the car had seemed to take a long time – about an hour – and when she left it was approximately 11.30 pm, but he added he could be wrong about that. A timescale of just 30 minutes between Tanya's arrival at Grove End House and the vehicle entering the river simply does not square with the detailed events the appellant says took place at Grove End House which included the playing of the tapes.

6. Lies.

52. The judge gave the jury the appropriate 'Lucas' direction on lies and then proceeded to deal with the various alleged lies relied on by the prosecution. The first was timing. As we have illustrated, it was the Crown's case that the appellant's account of events would impossible on the timing and that if the appellant was lying the only explanation was that he had killed Tanya.
53. The second alleged lie was the 'Bryony tape' and the appellant's claim that he had destroyed it. The Crown's case was that the 'Bryony tape' was a figment of the appellant's imagination, brought into existence because there was nothing in the 'Carol tapes' that could have possibly have caused the reaction in Tanya that the appellant claims she had that night. Something was needed to justify her driving off in a rage in the manner suggested by the appellant. Why, if the 'Bryony tape' existed was it (i) first mentioned at such a late stage and (ii) destroyed by the appellant. The appellant's answer was that he never wanted to mention the 'Bryony tape' because of his concern about the subsequent investigations to which his daughter would be subjected. He destroyed the tape some few days after Tanya's death when there was no longer an issue about Bryony living in the same household as Mr Cross.
54. The third alleged lie relates to the Adidas bag. The Crown's case was that the only reason why the bag, with his old clothes and towel in it, was in Tanya's car was because he took a change of clothes in case he needed to use them in manoeuvring the car and Tanya into the river and that he forgot to retrieve them. His explanation about going to the shed to get Tanya's clothes (which incidentally were never found) and the Adidas bag being with them was simply an unconvincing lie to explain away an embarrassing piece of evidence.
55. The fourth alleged lie was the appellant's statement to the police that Tanya never wore a seatbelt. The fifth alleged lie was detail volunteered to various witnesses as to why his fingerprints might be found on the car and why Tanya might have driven away without putting on her seatbelt. The most obvious example is touching the light switch because he saw what he thought were the eyes of an animal.

7. The Adidas bag.

56. Quite apart from the alleged lies about the Adidas bag the Crown's case was that its presence in the car with the appellant's working clothes in it was a mark of preparation for the murder.

8 and 9 the remaining issues, tyre tracks and vehicle damage.

57. Whilst the first seven issues remain untouched by the new material sought to be introduced by the appellant, issues 8 and 9 are closely related to it. The judge said that issues 8 and 9 had generated much debate and speculative examination in the course of the case and that days had been spent on them. He dealt first with the tyre tracks. He said this:

"The prosecution maintains that it has found the tyre tracks made by the Renault as it entered the water, and that they indicate that the car was driven onto the bank slowly, that it stopped, and then started again, so as to approach the bank slowly and in low gear and go over the bank. It suggests that that is consistent with its case that the defendant drove Tanya to the scene in the car, stopped somewhere on the bank, put it in gear, got out and reached through the opened window, turned on the ignition so that the car could make its own way into the river.

…..Dr Grant, the forensic scientist, investigated the scene at Stonebench on Wednesday 9 September. He told you that he saw a partial, and not particularly defined, pair of tyre tracks leading from the road across the humped bank and in the direction of the river edge. He said that where the tracks started close to the road he saw no evidence of skidding suggestive of the car swerving off the road at speed and out of control.

He said that he was able to trace the general path of the tracks over the hump and across the tyre ploughed mud, but from there the tracks were so unclear that he could not determine whether they reached the edge of the river bank. He did, however, as you will remember, notice damage to the vegetation down the steeply sloping bank, including some small bushes pushed over which he said was consistent with the vehicle having gone down the slope into the water. He told you that there was sufficient of the tyre marks to enable him to determine the track width of the vehicle that had caused them, and the pattern of its tyres……

Dr Grant continued. He said that the car tracks width and the tyre tread pattern were consistent with those that he had seen on the bank at Stonebench, and that therefore those marks could have been made by the Renualt…..

You have heard a great deal of detailed evidence from Dr Grant and Inspector Bottomley about tyre marks at the scene, and the conclusions or lack of them that can be drawn from them. The importance of such evidence, if any, is not as to whether the car went into the river at about this point, but as to how it went in. You may conclude, it is a matter for you, that the car with Tanya in it did go into the river thereabouts on the night in question. The question whether it made the tracks upon which the prosecution relies in the various sketches and photographs is only important if those tracks tell you how the vehicle that made them moved across the bank. The important question, therefore, is whether they support or throw doubt on the prosecution case that someone standing outside the car murderously propelled it into the river."

58. The judge went on to discuss the possibility of the tyre marks relating to Tanya's car, pointing out the lapse of time between 2 September and discovery of the marks, the popularity of Stonebench and the number of vehicles that could have visited the spot in the interim as well as the common tread pattern. He then directed the jury that it would be dangerous to conclude that any of the marks described were made by Tanya's Renault car. He said:

"In short, members of the jury, however much you may speculate about the probability or the possibility of the various marks described, drawn and photographed being made by Tanya's car you cannot be sure of it. Accordingly, I advise you not to draw any conclusion adverse to the defendant from such marks or their condition as described. Even if you could be sure the tyre marks were made by Tanya's car where would it take you?"

59. The CCRC in its reference describes this direction as overcautious. The judge, however, had had the advantage of hearing the extensive evidence. More significantly, the appellant's argument is that it is one thing for the judge to direct the jury to disregard the tyre marks as assisting the prosecution case but quite another now to disregard the tyre marks if the likelihood is that they were made by the Renault when that is integral to the appellant's case about how the vehicle went into the river. Mr Owen's argument is that if these were the Renault's tyre marks, the fresh evidence shows it was highly likely, putting it at its lowest in the light of the fresh evidence, that the appellant did not propel the vehicle into the river.
60. The judge put the issue of damage to the Renault this way:

"The prosecution relied on this evidence not so much as to show how the car was propelled into the water but to show where it went in. As I have said, there is no real dispute that it went into the river somewhere close to where the broken pieces of plastic were found. The evidence of damage to the Renault, including those broken pieces of plastic from its front bumper moulding, is relied upon by the defendant to support his contention that the car is likely to have been, or may have been, moving at a moderately fast speed across the bank before diving into the river. His case is that such damage to the vehicle is inconsistent with the car having moved off from a stationary position on the bank and falling over the edge of it into the river before getting up much speed."

61. Judge said of the damage to the plastic bumper moulding that it would undoubtedly have required some sharp impact with the ground, whether jutting rock or firm surface. He went on:

"It suggests, as Mr Solley submitted to you, some speed of the vehicle over rough terrain as it approached or went over the edge. But what speed? More than required from a sudden start nine or ten yards back from the edge? Or that of a vehicle passing without stopping in the road onto the bank? What speed?"

A little later he said:

"Movement across the bank at some speed, Mr Solley argued, was the only possible explanation of the damage to the plastic mouldings on the front underside of the car and the unusual damage to the roof and windscreen."

62. The manner in which the Renault passed from the road into the river was plainly an issue at the trial and there was considerable argument about the inferences that could be drawn from whatever were proved facts at the scene, including the damage to the front bumper moulding and the tyre marks. The judge said this to the jury:

"Mr Solley on behalf of the defendant, suggested that the tragedy could have occurred, could have been caused, by Tanya for some reason turning onto the bank and continuing at a moderate speed across it. It could, he suggested, have been a sufficiently fast speed over the rough ground to bump her into unconsciousness and to damage the underneath of the car in the way that it did, culminating in the car diving over the edge and possibly causing more injury to her on its descent into the river."

A little later he added:

"The defence suggest that accident, in the way that I have just summarised its case for you, was entirely possible.

Inspector Bottomley's experiments on the tipping also have to be considered alongside his experiments about the starting of the car from the outside.

The prosecution maintains that the Renault, in about the position shown in the photographs, could have been propelled by somebody outside the car over the bank by putting it in low gear and then starting the ignition from the outside."

The first question: Our assessment of the evidence at trial.

63. The fundamental question for the jury at the trial was whether they were sure that the appellant in some way propelled the Renault into the river or whether it may have gone into the river for some other reason. Precisely what happened to the vehicle between leaving the road and entering the water was, as is apparent from the passages of the summing up that we have recited, a matter of extensive debate at the trial. No witness saw the vehicle and its movements were it at best a matter of inference from established facts. That in itself posed the question whether supposed underlying facts were in truth established. Furthermore, the detail of how the vehicle moved between the road and the riverbank was of little consequence unless it was established with sufficient clarity to undermine all the other evidence that pointed to the appellant as the perpetrator.
64. This brings us to our first question identified earlier, concerning the Crown's contention that the circumstantial evidence at trial demonstrates that the appellant must have propelled the car into the river by one means or another. Is that evidence so overwhelming that it can only be answered by showing that such a course was completely impossible? It seems to us that the combined effects of the features described by the judge point inescapably to this conclusion. The appellant's behaviour after the event of itself was damning, explicable only as a desperate attempt to create an exculpatory story. The evidence that the car had gone into the river at or shortly after 11 p.m. was impregnable, and thus it would have been impossible for anything like the events he asserts took place to have occurred in the time available. Particularly important is evidence relating directly to the vehicle. First, both the driver's and the front passenger's windows were open. This is not only inconsistent with an accident but strongly indicative of a desire that the vehicle should sink quickly. Second the lights were extinguished, again inconsistent with accident and indicative of the vehicle having been put into the river deliberately. The evidence about Tanya's injuries, the seat belt, the car windows and the car controls all point to murder, and the appellant is the only person with any real motive. His alleged lies and the Adidas bag add to what is already an overwhelming case that the appellant must have propelled the car into the river by one means or another. We conclude that it can only be answered by showing that such a course was completely impossible.

The second question: the evidence at trial and the fresh evidence.

65. It is not suggested by the Appellant that the evidence at trial showed it was impossible for him to have propelled the car into the river by one means or another. Is the fresh evidence, taken in conjunction with the evidence at trial, capable of doing so? In order to answer this question we turn to discuss the fresh evidence.
66. Mr Parry of TRL was appointed by the CCRC to investigate the movement of the Renault on the riverbank at Stonebench prior to entering the river. Dr Parry produced a report in December 2003. Dr Ashton had been instructed by the prosecution during the original investigation. He had examined the Renault in which Tanya had died and carried out tests using a similar vehicle at the scene. His findings were set out in a statement dated 13 June 1993 but he did not give evidence of at the trial. In a statement dated 4 May 2005 he reviewed Mr Parry's work. He and Mr Parry have prepared for the assistance of the court a joint statement setting out the areas of agreement between them. There are no matters on which they disagree. The matters on which they have focused are:

• The nature of the vehicle debris found at the site.

• What can be inferred from the location of the debris.

• The interaction of the Renault with the riverbank.

• The conclusions that follow from the above.

67. It is necessary, before discussing the reconstruction experts' conclusions to recite briefly some salient background facts. The tyre tracks were first noticed by PC Evans on 6 September, that is four days after the Renault went into the river. They were approximately 10 metres in length. Photographs of the bank were taken the same day. According to PC Pickering the pieces of plastic that matched the front bumper moulding were about five feet from the bank edge. On Sunday 30 August there was a high tide that would have flooded above the bank and into the road at Stonebench. The following high tides' were lower and diminishing. The first high tide after 11 pm on 2 September would have been 18 inches below the top of the bank. On 12 February 1993, that is over five months after the incident, Mr Lorriman made a survey of the scene using electronic measuring equipment in order to prepare survey plans. The measurements in those plans were used to construct a three dimensional model used by the prosecution at the trial. They were also used by Mr Parry as the basis of a computer analysis. Neither this analysis, nor the computer programme on which the analysis is based, were in existence at the time of the trial. Thus the fresh evidence now tendered by Mr Parry was not available at the trial.
68. The experts proceed on the footing that the location of the pieces from the bumper moulding is of critical importance in determining the speed at which the Renault was probably travelling as it went across the riverbank and entered the river. If the debris had come to rest on the 'upper' section of the riverbank (as shown in the photographs) there must have been significant contact between the underside of the bumper section and the ground prior to the vehicle going over the riverbank edge. The contact must have been sufficient to cause them to break and detach before the vehicle passed over the riverbank edge.
69. The experts' evidence is that if such contact had occurred it would have occurred as a result of the lower part of the bumper contacting the ground and that this could have occurred as the vehicle was going over the ridge on the upper section of the riverbank. The experts could not exclude the possibility that contact could have occurred as the vehicle left the road but this did not fit with where the debris was found. It is obviously critical to the experts' assessment that the debris had not been moved between 2 September and when it was discovered on 6 September.
70. The most important feature of the experts' fresh evidence is that the speed of the vehicle, in order to break the pieces of bumper moulding, would have to have been at least 15 mph and this was faster than the speed at which the appellant could have propelled the vehicle over the edge in the manner suggested by the prosecution.
71. In short, it is the appellant's case that the damage to the front bumper moulding was caused as the vehicle travelled over a ridge in the upper bank at a speed of at least of 15 mph and that it did not stop before reaching the riverbank edge. The minimum distance between the end of the ridge and the riverbank edge is said to be approximately 5 metres. We have been shown a table of stopping distances produced by TRL which shows how the stopping distance varies according to the rate of deceleration, the deceleration rate being dependant on the nature of the ground. The available information suggests that if the Renault was travelling at 15 mph when it crossed the ridge and caused the bumper moulding to fracture, it is unlikely [our emphasis] it would have stopped within 5 metres i.e. before going over the edge.
72. Mr Owen succinctly puts his argument as follows. There is now agreement as to the minimum speed at which the car had to be travelling to cause the damage to the front bumper moulding. The Crown's scenario cannot produce sufficient speed to have caused the damage. The judge would never have given the jury the direction about disregarding the tyre marks had he heard the fresh evidence.
73. One possibility that we have considered is whether the bumper moulding was indeed fractured at some point after the vehicle left the road but that it stopped before the edge of the riverbank and that the appellant propelled it over from there. There is no evidence on the ground that there the Renault grounded at any particular point. There are no dimensions given for the ridge and assuming that the vehicle did ground on the ridge, no evidence of the distance between the precise point at which it grounded and the riverbank edge.
74. The computerised simulation on which the fresh expert evidence depends assumes Mr Lorriman's February 1993 survey is an accurate reflection of the state of the ground on 2 September 1992. In our judgment, bearing in mind the lapse of time of over 5 months and that the locality was prone to erosion by high tides and frequent vehicular access, the Lorriman survey cannot be regarded as a sufficiently sound foundation on which to base precise conclusions about the movement of the Renault. It is also critical that no one moved the pieces of bumper moulding between the Renault entering the river and PC Evans finding them four days later. Bearing in mind the popularity of Stonebench, we do not think this is an assumption that can safely be made.
75. Where this leads, it seems to us, is this. The critical fresh evidence, based on the computer programme, that the Renault must have been travelling at, at least, 15 mph when the pieces of bumper moulding became detached and that this occurred when it passed over the ridge does not establish the accused cannot have propelled the vehicle into the river from outside. Nor indeed does Mr Owen submit that this is the case. What the evidence does show is that based on a number of assumptions it is unlikely. One of those assumptions is that the tyre marks were made by the Renault, something about which the judge felt sufficiently unsure to tell the jury that it would be unsafe to rely on the tyre marks having been made by the Renault as assisting the Crown.
76. The jury knew at the trial that the vehicle must have been travelling at some speed to have detached the pieces of bumper moulding and deposited them where they were found. But, as the judge said, at what speed? And, as the judge pointed out, it was obvious that the faster the vehicle was going the more likely it would have gone over the bank under its own steam. The new evidence, so it seems to us, provides no more than a refinement to an issue that was already before the jury. What we now know is that if the assumptions on which the fresh evidence is based are correct, and in our view it is by no means clear that they are, then the vehicle must have been travelling at least 15 mph when it went over the ridge which would have made it very difficult to stop before it reached the riverbank edge.
77. Lord Bingham, in the passage cited earlier from Pendleton, is not suggesting that the Court of Appeal should consider the fresh evidence in isolation from the other evidence in the case. In the present case there was no eye witness who saw the Renault go into the river. The question whether it was propelled in some way by the appellant or, as Mr Owen now submits, an accident in which the appellant had no involvement, is one to be resolved by inference from proved facts. Based on certain assumptions the two experts Mr Parry and Mr Ashton are agreed that the Renault must have been travelling at, at least 15 mph when travelling over the ridge to have caused the pieces of bumper moulding to have become detached at the point they were found. If so, it is unlikely that the car stopped before it reached the riverbank edge. This cannot in our view help the appellant. First, for the reasons given earlier it is by no means clear that the assumptions are correct. Second, even if they are correct, the conclusion drawn by the experts is not that the appellant could not have propelled the vehicle over the edge. It is unlikely in the sense that, if isolated from all else that we know, one would not have expected it to happen. Such a conclusion does not reach – and is not capable of reaching – the degree of cogency needed to answer the case against the appellant.
78. It is in our view relevant to look at the issue to which the fresh evidence relates and to consider the extent to which that issue was before the jury at the trial and the impact the fresh evidence may have had on that issue.
79. In our view Bowler is distinguishable from the present case on the facts. On our analysis this is not a case where the fresh evidence – if considered with all the other features of the evidence – suggests it is possible that the deceased met her death otherwise than by the actions of the appellant.
80. As in Bowler, we heard the fresh evidence de bene esse. We also remind ourselves that the overriding consideration for the admission of fresh evidence is the interests of justice test and that it is still possible to entertain fresh evidence even though one or more of the criteria in section 23(2) are not met. We bear in mind the warning of Lord Bingham that the test is not whether we think the appellant is guilty but whether the conviction is safe. Whilst we cannot look into the minds of the jury we cannot disregard the very considerable circumstantial evidence which remains untouched by the fresh expert evidence. The fresh evidence is based on assumptions, which in our judgment provide an insufficiently firm foundation for that evidence to have affected the jury's decision to convict.

The third question

81. On our analysis the third question does not arise.

An irrelevant consideration.

82. After his conviction the appellant produced a sixteen page handwritten document giving a different account of events from his evidence at the trial. The thrust of it was that he drove the car with Tanya in the passenger seat and that he drove into the river at Stonebench by accident. In a letter to the CCRC dated 4 April 2001 the appellant said that his account in that document was not true. Initially the prosecution sought to adduce these documents as fresh evidence to support the conviction. However, Mr Atkinson Q.C for the prosecution did not pursue this in oral argument. In our judgment he was right not to do so. This evidence would have been relevant only to the credibility of the appellant and not the commission of the offence. Accordingly we have had no regard to this material.

Conclusion.

83. Although we have given most careful consideration to the fresh evidence which we have heard de bene esse we have in the end come to the conclusion that it should not be admitted under Section 23 of the Criminal Appeal Act 1968 because we do not think it would afford any ground for allowing the appeal. In our judgment, had the jury heard the fresh evidence they would have inevitably reached the same conclusion. The conviction is safe and the appeal must be dismissed.