R - v- Barry Michael George [2002] EWCA Crim 1923

IN THE SUPREME COURT OF JUDICATURE

COURT OF APPEAL (CRIMINAL DIVISION)

ON APPEAL FROM QB DIVISION

CENTRAL CROWN COURT

The Hon. Mr Justice Gage

[2002] EWCA Crim 1923

Case No Case No: 200104011 Y4

Venue Royal Courts of Justice, Strand,, London, WC2A 2LL

Between

R - v- Barry Michael GeorgeAppellant

Before

Lord Chief Justice of England

Mr Justice Curtis

Mr Justice Henriques

Mr Michael Mansfield QC and Miss Maryam Syed

(instructed by Messrs Kean Etienne) for the Appellant

Mr Orlando Pownall QC and Mr Jonathan Laidlaw QC

(instructed by the Treasury Solicitor) for the Respondent

Date 29 July 2002

The Lord Chief Justice: INTRODUCTION After a trial before Mr Justice Gage at the Central Criminal Court, which commenced on 23 April 2001, the jury retired on 27 June 2001 to consider their verdict. On 2 July 2001, after deliberating for almost 32 hours, the equivalent of 5 days, the appellant was convicted by a majority of 10 to 1 and sentenced to life imprisonment. He appeals against conviction by leave of the single Judge, Morland J. Granting leave Morland J stated:

"Although much of the background and circumstantial evidence is consistent with guilt, I consider that the Court of Appeal should assess the general safety of the verdict having regard to the overall paucity of cogent evidence of guilt particularly in view of the weaknesses in the identification evidence (eg the so called underlying unity or disunity – see pages 97/98* and 116 – 120* of the summing up of 25 th June 2001) and the possibility of innocent contamination of the Cecil Gee coat. Exhibit 24, by the particle of percussion cap residue through insecure transmission (i.e. the arguably corrupted integrity of the coat, exhibit 24, see pages 77 et seq. 100 et seq. and pages 111 – 115 of the summing up of the 26 th June 2001)."

[2]The victim of the murder was Jill Dando, the well-known television presenter. On 26 April 1999 at about 11.30am she was shot in the head outside her front door at 29 Gowan Avenue, Fulham. Nobody witnessed the shooting. Within a short time a number of people saw her slumped and bleeding in the porchway. The police and emergency services were called. She was transferred to Charing Cross Hospital where at 13.05 she was pronounced dead. A post mortem examination was carried out. Two gunshot wounds were found, an entry wound and an exit wound. The bullet entry wound was located behind the top of the left ear and the exit wound was above the right ear. An impression of the muzzle was discernible in the area surrounding the entry wound indicating that the muzzle had been pressed firmly against the left side of her head when fired. The gun was not fitted with a silencer but the proximity of the muzzle to the left side of the head would have substantially reduced any noise from the weapon. A fired bullet and casing were recovered by the police from the vicinity of the doorstep. There followed a massive investigation in an attempt to identify the killer.

[3]The appellant was first drawn to the attention of the police two days after the murder but it was almost a year later, on 11 April 2000, that the appellant was first spoken to by police officers as a witness. He was arrested for this murder on 25 May 2000 and charged on 29 May 2000. THE GROUNDS OF APPEAL

[4]The appeal raises a number of issues, principally as to the quality of the identification and scientific evidence. As to this evidence the appeal gives rise to one issue of general importance as to the admissibility of evidence. It is as to the questioning which is permissible by the prosecution of a witness as to identification which is either a partial or a non-identification by the witness ("The Partial and Non-Identification Issue"). In addition prior to the jury being empanelled for the trial, Mr. Mansfield QC, who appeared for the appellant at the trial, made an application that the prosecution should be stayed on the grounds of abuse of process. He relied on the prejudice caused to the appellant by the delay in his being prosecuted and the adverse publicity he received when restrictions on reporting were lifted. Gage J dismissed the application. On the appeal it is contended this decision was wrong. ("The Abuse of Process Issue"). There is also a ground of appeal based on an irregularity which it is alleged occurred after the jury retired ("The Jury Issue"). The issue on which Mr. Mansfield particularly focussed at the hearing of the appeal is whether on the evidence as a whole this Court should come to the conclusion that the conviction is unsafe. If it is that would mean that Gage J was wrong not to stop the case; either by accepting a submission of no case to answer or by withdrawing the case from the jury. Alternatively Mr Mansfield argues that for the same reason this Court must have a lurking doubt about the safety of this conviction and allow the appeal for that reason. As Mr Mansfield accepted, in this case these different approaches can all be satisfactorily resolved so far as his client is concerned by asking a single global question as to whether the conviction is safe. It is primarily submitted that the conviction is unsafe because of the flawed identification evidence so that issue be described as "The Identification Issue". We turn now to consider the issues starting with the Abuse Issue. THE ABUSE ISSUE

[5]Mr. Mansfield QC relies on the prejudicial effect of: i) the delay in the appellant being interviewed and in consequence his prosecution; ii) the adverse publicity following the Judge lifting an order preventing publication of photographs of the appellant on 27 February 2001 which was followed by widely publicised photographs of the appellant appearing in the media. DELAY

[6]Complaint is made that the police received information between the 26 April 1999, the day of the murder, and the 14 June 1999 which should have caused the investigating team to appreciate that by that date the defendant was a person whom they needed to interview as a matter of urgency. Steps in fact taken to trace and identify the defendant between February and May 2000 should have been taken it is said between April and June 1999.

[7]The Judge heard evidence as to the cause of the delay from Detective Chief Inspector Campbell. The police received a vast amount of information. Between the 26 April and the 12 May they received 1,109 messages and raised 1,273 actions. A system of giving priority to various aspects of the enquiry had to be put in place. Three categories of suspects had to be investigated. So far as messages were concerned they were given four levels of priority. The messages relating to the defendant were given a low priority until February 2000.

[8]Mr Mansfield contended that it should have been glaringly obvious to the investigating team that witness statements should have been taken from a number of witnesses and that there had been a breakdown of the system. He relied upon a note dated 29/30 April 2000 written by Mr Campbell himself in which the word 'failure' was used. Mr Campbell did not accept there had been a failure of the system. An action of 15 May relating to the appellant had been downgraded to a lower priority in the queuing system and his view was that the officer who had taken that decision had acted perfectly reasonably in the light of stated priorities and the amount of information which had come in and was continuing to come in.

[9]The trial Judge concluded that in terms of an investigation of this magnitude the delay was small. It extended to a period of months, less than one year. In addition it was his opinion that the delay was not something that could be categorised as a failure of the system or inefficiency. It seemed to the Judge reasonable in the context of this massive investigation that at the time, the information should not have received a higher priority. That the decision turned out to be wrong does not mean that it was made through inefficiency or that it was evidence of a system failure.

[10]Mr Mansfield's submissions had been based on well established principles as stated in R v Beckford

[1996] 1 Cr App R 94

where this Court stated there were two categories of case where proceedings may be stayed on the basis that their continuance would be an abuse of process. The first is where(a) the defendant would not receive a fair trial and/or

(b) where it would be unfair for the defendant to be tried. The second is based on R v Mullen

[1989] 2 Cr App R 143

where this Court stated that proceedings may be stayed in the exercise of the Court's discretion not only where a fair trial is impossible but also where it would be contrary to the public interest in the integrity of the criminal justice system that a trial should take place.

[11]Mr Mansfield's submissions were based upon the first category of case. He submitted that the appellant was prejudiced in three ways namely, as to his alibi, the forensic evidence and the identification evidence.

[12]In support of an alibi, it was the appellant's case that he visited the Hammersmith and Fulham Action for Disability (HAFAD) premises on the day of the murder. The exact timings were important because they could make it impossible for the appellant to have been at the scene of the crime at the time of the murder. The Crown's case is that his visit to both HAFAD and premises belonging to Traffic Cars was partly a matter of comfort and partly of a deliberate ploy by the appellant to be seen wearing different clothes from those he had been wearing in Gowan Avenue and behaving in a manner inconsistent with someone who had just carried out a killing. There are indeed differences in the timings of witnesses from HAFAD and some have changed their own estimates of the time. Had they been visited nearer the event in question it is submitted that their recollections would have been clearer and would have provided an alibi for the defendant. The trial process it was said cannot now cure this prejudice.

[13]A similar argument is raised in relation to scientific evidence. The appellant's home was not searched until April 2000. Apart from one particle of firearm discharge allegedly found in the pocket of one of the appellant's coats, which the Crown rely on to establish a link with the bullet which killed Miss Dando and a fibre found at the scene which may have come from a pair of the appellant's trousers, there is no other scientific evidence to connect the appellant to the crime. Had a search taken place much sooner the Crown's argument that the defendant had had time to remove any incriminating evidence would have been diminished.

[14]As to identification evidence it was said that the delay has reduced the effect of the defence point that a number of witnesses have not been able to identify the appellant. The submissions in relation to identification evidence were repeated in a subsequent submission that the identification evidence should be excluded.

[15]The real thrust of the defence submission was that the Crown would use the delay unfairly to their advantage. They would use it to undermine the defence alibi asserting that the HAFAD witnesses could not be expected accurately to recollect times after a year. They would assert that the defendant had rid himself of incriminating articles and they would assert that any uncertainty on the part of an identifying witness was attributable to lapse of time rather than their inherent weakness. AS TO ADVERSE PUBLICITY

[16]The facts start with the appellant's arrest on 25 May 2000 and an order which was made by the Magistrates' Court under S.11 of the Contempt of Court Act 1981 prohibiting publication of photographs or drawings of the appellant's face. The making of such an order was at the invitation of both parties with a view to protecting the integrity and fairness of the identification parades that were to take place thereafter.

[17]On 27 February 2001 representatives of the Press invited the Judge to lift the prohibition, as the trial was about to commence. There were no further identification parades to be held and thus no apparent reason to prevent the publication of photographs. The prosecution adopted a neutral stance and the Judge lifted the order.

[18]The consequence of lifting the order was that on the following morning a number of vivid headlines appeared in the Press. They published numerous pictures of the appellant spanning 30 years under what the Judge himself described as "lurid headlines" and the accompanying publication of background information, which for example spoke of the appellant's habit of changing his appearance. The Judge re-imposed the section 11 order on 29 February 2001 and thereafter the press was allowed to use a sample photograph of the appellant during the trial, which had been approved by the defence. The Judge looked at every sketch of the proceedings before sanctioning its use. None of the sketches showed the face of the appellant. CONCLUSION

[19]The trial Judge accepted that there was some force in the arguments made on behalf of the appellant. He rejected them on the grounds that neither individually nor cumulatively did they amount to such prejudice as can properly be described as sufficiently serious to justify the exceptional course of staying the proceedings.

[20]A number of arguments support the correctness of the Judge's decision. First, in respect of the HAFAD witnesses there were a number of contemporary records made in the period 26 April 1999 to 12 May 1999 which assist with regard to disputed issues as to the time of a visit of the appellant on the day of the murder. As to the possibility of scientific evidence linking the appellant to the murder, if he wished to remove any traces he could have done so within hours or days of the event. So far as identification evidence is concerned the operation of S.78 of the Police and Criminal Evidence Act 1984 ("PACE") would permit exclusion of identification evidence if, in all the circumstances, it were unfair to admit it. Accordingly, as the Judge concluded, the trial procedures would be sufficient to deal with any possible prejudice. On the topic of delay the Judge stated that even if he had concluded that such delay as there had been was due to the inefficiency of the police or a failure of the system in the circumstances, he would not have stayed the prosecution.

[21]The trial Judge indicated that he found the submissions based on the possible effect of publicity the more troublesome. We have already made reference to the "lurid headlines", which followed the lifting of the S.11 order on 28 February 2001. At a pre-trial hearing in January 2001 a complaint had been made by Mr Mansfield that the defendant had been referred to in some sections of the media as an odd ball, a loner and a weirdo, such publicity being in late May 2000. Immediately following the lifting of the order on 28 February 2001 not only were there "lurid headlines" there was also reference to background information about the appellant's habit of changing his appearance. It was submitted that these reports contained the clear innuendo that the appellant was guilty of this crime. The Judge having reviewed the numerous relevant authorities was not satisfied on the balance of probabilities that the defendant could not have a fair trial. He indicated that the jury would be able to observe the defendant during the trial thus dispelling any unflattering impressions left by the press photographs. He stated he would give to the jury the appropriate clear and firm warning to put out of their minds anything they had seen or read about the defendant reminding them that their duty was to try the case on the evidence that they heard in Court. Further he concluded that the risk of any identification witnesses' evidence being tainted was remote. CONCLUSION

[22]We do not consider that the Judge's conclusions can be faulted. He came to the right decision. It would be his responsibility to ensure the fairness of the trial and he had ample powers to enable him to do so. He subsequently used his powers during the trial in an exemplary manner. Mr Pownall QC who appears on behalf of the Crown submits that he neither had any intention to nor did in fact use the fact of the delay which had happened unfairly. If he had, then that could have been tackled appropriately during the trial. We are satisfied that a stay in anticipation of possible problems was not the right answer.

[23]As to the Judge's decision to lift the S.11 Contempt of Court Act Order on 28 February 2001, if he had appreciated the lack of restraint that was going to be shown by a section of the media he would undoubtedly have not lifted the order. However, the fact that with the benefit of hindsight we know that there was irresponsible behaviour does not mean at the time the decision was made that it was wrong. The purpose for the order being made (to give time for identification parades to be held) had been fulfilled. The need for proceedings to be held in public means that a S. 11 Order should only remain in force for the time necessary to fulfil its purpose. It would be wrong to regard the decision of the Judge unfavourably because he did not take into account the danger of the media behaving irresponsibly. We do not consider that the cumulative effect of delay and adverse publicity was such as to render a fair trial no longer possible. The trial which took place was fair. We reject this ground of appeal. THE PARTIAL AND NON-IDENTIFICATION ISSUE

[24]Mr Mansfield submits Susan Mayes is the only witness who made a positive identification at the identification parade. He submits that all the other evidence of what occurred at identification parades should have been excluded pursuant to S.78 of PACE. This is both because of the nature and quality of the purported identifications and the inherent dangers arising from delay.

[25]It is the experience of this Court that when a witness to a crime has failed to make a positive identification on an identification parade the witness may nevertheless be called into the witness box to describe the offender and what occurred on the identification parade. A failure to make a positive identification is no bar to describing either the event or the offender. As the trial Judge commented, the lack of authority upon the point arises because it is axiomatic that such evidence is admissible.

[26]Mr Mansfield concedes that each witness could quite properly give a description of the man observed, including the clothes he was wearing, but he submits that when a witness fails to make an unqualified positive identification no evidence can be given other than there has been a non- identification. In particular no evidence should be given as to why the witness only made a qualified identification or failed to make an identification since this would encourage the jury to treat the witness as having made an identification when this is not the case. While this restriction applies to the Crown when there is not a positive identification Mr Mansfield concludes a defendant can cross-examine a witness so as to show the positive identification was mistaken.

[27]As we have already pointed out, the point raised by Mr Mansfield is of a general nature and in assessing the correctness of his submission we will assume that Susan Mayes was the only witness who made a positive identification. In addition, we will assume that what happened was at a "live" identification parade, (although in this case after the first identification parade at which he was not identified, this appellant would not agree to stand on an identification parade) subsequent identifications were based on video pictures of individuals who had been made up to look like the appellant.

[28]Mr Mansfield submits there is no place for qualified identifications, in a system manifestly aware of the dangers of mistaken identification. He says sufficient errors have been made by those who have not qualified their identification for this to be the position. Unless there has been a positive identification recorded as such, he submits, the jury should hear no more than that the witness failed to make a positive identification.

[29]We have been referred to the report of the Departmental Committee set up under Lord Devlin in 1976 to review all aspects of the law and procedure relating to identification (Report to the Secretary of State for the Home Department of the Departmental Committee on Evidence of Identification in Criminal Cases, Hansard C. 338 1976).

[30]That Committee examined the possibility of putting three questions to identifying witnesses; 1. Can you positively identify anyone on the parade as the person you saw? 2. If not, does anyone on the parade closely resemble the person you saw? 3. If not, can you say that the person you saw is not on the parade?

[31]The Committee rejected the possibility of asking questions two and three. The chief reason for this being that it might confuse a witness. Whilst the Committee decided against any further questioning of witnesses it did approve the recording of the spontaneous observations of witnesses. The forms used by the Parade Inspector require a note to be kept of witnesses conduct and statements throughout the parade. The Report highlighted the inherent dangers of identification evidence repeated in R v Turnbull 63 Cr App R 132 and R v Forbes [ 2001] 1 Cr App R 442. Mr Mansfield contends that permitting partial or qualified identification greatly increases the risks, which accompany positive identifications. Mr Pownall replies that Mr Mansfield's submission accords neither with established practice nor with any precedent. There is no reported case in which this Court has ruled that in the absence of a positive identification, evidence of the circumstances of identification cannot be admitted.

[32]If evidence is relevant and admissible and passes the quality test set by S.78 of PACE it is submitted by Mr Pownall that a Judge should admit it having fully evaluated the fairness of doing so. Mr Pownall argues that the consequences of the ruling contended for by Mr Mansfield is that a large number of witnesses may be 90% sure but without being positive and yet no evidence of this would be admitted. Secondly, if a denominative feature is obscured and a witness cannot identify or give a qualified identification e.g. where a plaster is over a scar no evidence would be admitted. Thirdly, a witness may say that he is sure at a parade but under cross examination may say he was 95% sure which logically under the regime proposed by Mr Mansfield would result in the jury being told to disregard that identification evidence in its totality.

[33]The Crown, it is submitted, must be allowed to investigate the circumstances in which a failure to identify has occurred in the light of all the circumstances, including time and change of appearance. A video parade provides an inferior opportunity to examine and assess a suspect. This appellant chose not to co-operate with a conventional parade. Eight people were made up to look like the appellant and where the appellant has chosen to decline to participate it is argued that he cannot complain if, subject to a test of fairness, evidence of a qualified or partial identification is admitted. CONCLUSION

[34]We fully recognise the dangers involved of wrong convictions occurring in identification cases. This is the reason for the requirement that in all identification cases clear Turnbull directions must be given. We also accept that counsel for the defence is usually faced with a difficult task in challenging an honest witness who has made a mistaken identity. We also agree that prosecuting counsel must be cautious and avoid conducting his examination of a witness who has failed to make a positive identification in a manner which suggests to the witness that but for this fact or that fact that the witness would have made a positive identification. An identification which is qualified cannot be transformed into one which is unqualified by careful questioning. It remains qualified and the jury should be aware of this. Equally a defendant must not be convicted on the evidence of a qualified identification alone.

[35]However, there are at least two situations where a qualified identification may in appropriate circumstances be both relevant and probative. First, where although the weight of the evidence will still be less than a positive identification, it supports or at least is consistent with other evidence that indicates the defendant committed the crime with which he is charged. Secondly, the explanation for a non or qualified identification may help to place the non or qualified identification in its proper context and so, for example, show that the other evidence given by the witness may still be correct. Otherwise, a non or qualified identification could be used to attack the credibility of other evidence given by a witness when the explanation for this may show that such an attack is unjustified.

[36]In each case it will be for the Judge to decide whether the evidence is more prejudicial than relevant and probative bearing in mind the importance of protecting the position of a defendant against unfairness. In this case as we will see part of the case for the prosecution is based on the pattern of identification evidence, including the build, the complexion and clothing which the appellant was wearing. Subject to the jury receiving appropriate warnings, which were given in this case, the general evidence of the witnesses who saw a man who the prosecution say was the appellant was highly probative.

[37]Certainly, where a witness makes a spontaneous remark at a parade and it is recorded we are satisfied that such a remark is admissible in evidence if relevant and probative subject to the trial Judge's discretion to exclude it pursuant to S.78 of PACE. In many cases where there is an early and conventional identification parade and a witness fails to make a positive identification any spontaneous remark may either lack relevance or probative value or fall to be excluded under S.78.

[38]However, there are also circumstances in the absence of a positive identification where it is fair to permit the prosecution to adduce evidence of a spontaneous observation by a witness attending upon an identification parade. The same approach applies to a witness being questioned in court after not making a positive identification. THE IDENTIFICATION EVIDENCE ISSUE

[39]In this case, the prosecution relied on the evidence of 9 witnesses who described a man in or in the vicinity of Gowan Avenue either on the evening of the 25 April 1999 or on the following morning up to the time of the murder. We set out in an appendix to this judgment a table summarising their evidence. We emphasise the table provides no substitute for the transcripts of their evidence which we have read. It does however show the pattern of evidence of the witnesses on which the prosecution rely.

[40]The video parades in question were on 14 August 2000 and the 5 October 2000 and were 16 and 18 months after the observation. On 26 May 2000 the appellant indicated he would not attend upon any further conventional parade. Between the murder and the video parades the appellant had grown a volume of facial hair. The task of any witness was accordingly greatly increased. They were not comparing like with like and they were deprived of the opportunity of observing the height and build of the suspect. It is hardly surprising that, for example, Miss Normanton could not remember a moustache, nor is it surprising that Miss Scott found the parade difficult partly because of the time that had elapsed, partly because each picture was of a man with facial hair and also she could only see the head and shoulders of each person. Each witness must be considered separately and with care. This is what we propose to do.

[41]As we consider each witness in turn it is useful to have a description of the appellant in mind. On the day of the murder he was 39 years of age. He was about 5ft 11 tall. He was fairly heavily built. He had dark hair.

[42]Mr Geoffrey Upfill-Brown in his first witness statement described a person running from 31 Gowan Avenue very shortly after the murder as white, between 30 and 40 years old. He was about 5ft 10 of medium build. He had a mop, very definitely a mop of thick black straight hair, collar length. It was such a mop it gave the impression it might have been a wig. He had a round face, which was of a sallow complexion and he was clean-shaven. He was wearing very dark, tending towards black, clothing which was baggy, therefore, he could not be sure of his build.

[43]At trial the witness described him as 5ft 8 to 6ft tall. By his reckoning about 5ft 10. Because of his loose clothing he seemed to be fairly thick set. Age was somewhere between 35 and 40. He had black hair, long almost down to collar length, and curly. He described it, as a mass of thick black hair and at the time thought it could have been a wig. It came almost to the back of his collar. He had no facial hair, was clean-shaven, no glasses and of sallow complexion. His clothing was loose fitting, the jacket almost to his knees; dark in colour with, he thought, a tinge of brown in the darkness. The trousers were baggy and loose fitting and of the same colour as the top. He had black shoes.

[44]Mr Upfill-Brown's observation was of a man exactly opposite his house. The position could be seen by the jury on photographs. He saw the man running for about 20 yards towards Fulham Palace Road and he then stopped and turned around. The man looked at the witness and then broke into a semi-jog form of running before going behind a lorry. The witness had an opportunity to look at the man's face for one or two seconds. He was able to see at least three-quarters of the man's face. When he was running he was able to see the side of his face. When asked if he had any reason to recollect what happened he said "I was expecting to have to remember it because of the way the man was behaving. I watched him very carefully and made every mental note I could."

[45]On 14 August 2000 Mr Upfill-Brown failed to identify the defendant at a video identification parade. So far as delay is concerned the witness instantly realised the importance of the observation. He made a statement on the day of the murder and thereafter could be expected to appreciate the importance of remembering what he had observed.

[46]Mr Richard Hughes lived at number 31 next door to Jill Dando. Between 11.30 and 11.45am on the morning of the murder he was upstairs and heard a sound like his wife's car alarm. He heard footsteps approaching for about 5 seconds then a scream, which he did not consider significant. He heard the gate click then looked out of an upstairs window and saw a figure walk left towards Fulham Palace Road and had a side view of his face for a few seconds. In his statement dated 29 April 1999 he described the man as "white in colour, with thick set broad shoulders, thick black hair, pushed back, collar length, mid to late 30's, he was not wearing any glasses". In an earlier statement made on the day of the murder he said the man was wearing a darkish wax/Barbour jacket and the man immediately reminded him of a comedian he had seen on television by the name of Bob Mills.

[47]In evidence he described the man as between 30 and 40 years old above average height. He said he was 5ft 7 and estimated that the man was taller. The man was of a thick set build and his hair was dark brown, to the ears and collar. He described it as a thick dark mop of hair, which, from what he could see, did not appear to have curls or waves. The face was full and 'jowly'. There was no facial hair or glasses. The man wore a dark colour Barbour style jacket. Some five minutes later he heard of the killing. On 26 May 2000 he attended an identification parade but was unable to make a positive identification.

[48]It was conceded certainly for the purposes of any submission, that Mr Hughes and Mr Upfill-Brown had not only observed the same person but that such person was the gunman. Indeed Mr Hughes had seen Mr Upfill-Brown come out of his gate just as he saw the man move off briskly. We can well understand why Mr Mansfield at trial was anxious to exclude the Bob Mills reference but it having been excluded we can see no further possible objection to the admissibility of Mr Hughes's evidence.

[49]There followed evidence from a group of witnesses, six in number, who had seen a man in the Gowan Avenue area or close to the deceased's home during the evening before, and the morning of her death.

[50]Helen Scott saw a man in the area of the junction of Gowan Avenue and Fulham Palace Road at about 8pm on Sunday 25 April 1999 (the evening immediately preceding the murder). He was looking down Gowan Avenue and appeared to be behaving in a bizarre manner. In her witness statement dated 4 May 1999 she described the man as white though there was a hint of Mediterranean about the skin but still of olive complexion. He was 35 to 40 and had thick well-groomed black hair; straight, brushed forwards shorter on the sides, thicker on top, collar length. The suspect was clean-shaven; his eyebrows were very thick and dark. His eyes were dark, his face was wide shaped and also round, but not chubby. He was wearing a three-quarter length coat, 5ft 10 and of medium build. She believed she would be able to recognise him again. She attended a video identification on 23 January 2001 and having seen the tape once asked to see No.s 1, 2 and 8 again, and finally selected No. 8 (not the appellant who was No. 2). She said she was 80-85% sure. We have seen, as the jury would have seen, the faces of the persons used for the video identification. As Mr Mansfield accepted, the photographs of the persons shown as No.s 2 and 8 were very similar in appearance.

[51]In evidence Helen Scott said it was a bright sunny evening when she left home to collect a pizza. She saw a man standing alone looking up towards Gowan Avenue. As she walked north, she saw him for probably a couple of minutes. He continued to look up Gowan Avenue until she was almost level with him. Even though he crossed the junction he continued looking up Gowan Avenue. She was in the shop for 15 minutes and he still looked up Gowan Avenue. She described him as late 30's possibly early 40's. Average height about 5ft 10, average build and not overweight. He had very dark hair, possibly black, but fairly short at the sides. It was slightly longer and thicker on top. It seemed quite well cut with a slight fringe just above the collar. His eyebrows were quite full and dark and there was a small gap between the eyebrows. His face was quite dark and she described him as being "sort of Mediterranean, with a slightly olive complexion." She noticed a Barbour style three- quarter length waxed jacket.

[52]Mr Mansfield submitted at trial that the passage of time made any identification very difficult and that it was quite wrong for the prosecution to put this evidence of what occurred at the identification parade before the jury when the witness failed to make a positive identification. He relied on R v Forbes

[2001] 1 Cr App R430

in submitting that evidence should be excluded pursuant to S. 78 of PACE.

[53]It will be noted that this video identification took place on 23 January 2001 whereas Mr Hughes and others had attended upon a conventional parade on 26 May 2000. It took time to make a compilation video, which was shown to groups of potential witnesses on 14 August, 5 October and 11 December 2000, and 23 January 2001.

[54]However, we note that the witness made a statement nine days after the event and that the importance of accurately recollecting her observations must have been obvious to her from the moment she heard of Jill Dando's death. In any event, she had very good reason to remember the man continuously looking up Gowan Avenue for two minutes and still doing the same fifteen minutes later. When cross-examined she agreed that she had heard reports that a dark haired man had been seen before she made her statement and she had seen an E-fit after she had made her statement. This E-fit had been produced by a member of the public called Sappleton, who became suspicious about a man he had seen sweating at a bus stop in the Fulham Palace Road shortly after the killing. The description given of this man by Sappleton made it quite clear that he was not the man seen by Messrs Upfill-Brown and Hughes and it certainly was not the Crown's case that the E-fit was of the killer. Accordingly, it is difficult to see how the appellant could have been unfairly or adversely affected by the witness seeing the Sappleton E-fit. She had also seen posters with the E-fit and three more faces on posters before she made her statement. She said she found the video identification parade difficult, partly because of the time that had elapsed, partly because each picture was of a man with facial hair, and also that she could only see the head and shoulders of each person. The Judge gave a very strong Turnbull direction pointing out in the clearest terms that she failed to identify the defendant at the video parade adding "in fact she identified another person. For this reason, I direct you that it would be extremely dangerous for you to rely on the evidence of what happened at the video parade in relation to this witness as in any way supporting the Crown's case. The fact is she did not identify the Defendant."

[55]We can see no basis for an exclusion of Helen Scott's evidence or of her recorded replies during the identification process. Her observations were in close proximity to the man she saw. She saw him first for a couple of minutes and then 15 minutes later and she had a good reason to memorise his description. She gave an explanation for identifying a volunteer, which was for the jury to consider.

[56]Susan Mayes is the witness whose evidence Mr Mansfield did not invite the Judge to exclude. There was not any possible basis for doing so. She lived in Gowan Avenue and left for work at precisely 6.57am daily. On the morning of the murder she saw a maroon car double-parked outside 28 Gowan Avenue, almost directly opposite number 29. The car could have been parked legitimately had the driver so chosen. She saw a man standing on the north side of Gowan Avenue just looking at the houses on the left. He then looked at the houses on the opposite side of the road and Susan Mayes had him in her view for about one minute. She was wondering what he was doing as he could have parked next to the pavement. She believed he was responsible for the car either as a chauffeur or more probably a mini cab driver. When she drew level with him he looked at the ground and he then started to clean the windscreen with his hand with his left arm against his face. He had stopped when she turned and looked back. Her description of the man was: age late 30's, height about 5ft 9 tall, stocky and slightly overweight; he looked Mediterranean and had olive skin; hair black in quite long layers with the longest bits over his collar; a heavy style. He was wearing a black suit with a white open-necked shirt, not smart and as she got closer he looked scruffy. In cross-examination she agreed that when seen by the police on 27 April 1999 the police record showed that she described his hair as "with a short and smart haircut." She did not know if those were her words or a police interpretation. She also agreed that her statement had referred to hair swept back very straight and layered. She attended a video identification parade on 5 October 2000, 18 months after the murder. She picked out No. 2 (the appellant) having asked to look at No.s 1 and 2 again and she said was very sure of that identification. She agreed she used shops in the locality but did not think she had seen the man before. When it was suggested to her that she had made a mistake in her identification, she said she was not mistaken. The car, a saloon was not an estate and not old. It was between her and the man and obliterated quite a lot of him. She had looked closely at him but did not stare. She agreed she might have been wrong in denying she had seen E-fits on the television before the parade. It was his hair that made her think he was Mediterranean. Re-examined she said he appeared 'thuggish' and had looked at his face three times for 5-6 seconds in total. He appeared nervous and it was obvious he did not want to be seen. There was an agreed admission that the police had failed to trace any maroon car and that the appellant had no car.

[57]The prosecution submitted that Miss Mayes's identification was unassailable. She had a close view of the man. Her evidence was clear and compelling and her description tallies with others that know him. The defence contended she was mistaken. No maroon car has ever been traced and a witness, Miss Waldock, who was in Gowan Avenue at 7.15am saw no car double-parked. There being no argument on the admissibility of Miss Mayes's evidence, we will return to it when reviewing the more general submission as to lack of evidence. Lord Chief Justice

[58]We turn now to the evidence of Stella de Rosnay and her daughter-in-law, Charlotte de Rosnay. Stella saw a man as she looked out of her bedroom window at 55 Gowan Avenue at about 9.30am on the morning of the murder. He walked towards her and she had an unobstructed view of him for about 10 to 12 seconds. She commented because he looked like her son (married to Charlotte). Her description was aged 35-40, height 5ft 10 to 6ft. Heavy build or strong build. Hair dark brown normal cut neither short nor long. Face strong and round – no particular feature. Skin high coloured pinkish and English. She did not think he looked Mediterranean. Wearing a suit – medium grey but she could not remember the colour of his shirt or tie. She attended the video identification parade on 5 October 2000 even though she said in her statement she had not considered that she would be able to recognise the person. She was surprised when she found she did recognise someone. She had asked to look at No.s 2 and 8 again having said, "for me for the moment it would be No. 2 or No. 8". She went on to say she could not tell between No.s 2 and 8 but her gut feeling was No. 2. It was frustrating. It was the colour of his skin. Having looked at No. 8, she was recorded as saying, "I would say No. 2." She could not make a positive identification.

[59]It is submitted that the Judge should have excluded this evidence pursuant to S. 78 because the witness said in a witness statement on the day of the parade that she (after she and Susan Mayes had picked out No. 2 – see next paragraph) was unable to pick anybody out and at the parade itself she said "I can't make a positive statement." It is important, however, that (after she and Susan Mayes had picked out No 2 she made a further witness statement on 6 October 2000 in which she said "I watched the video twice and immediately ruled out seven of the males. When I saw No. 2 on the screen, it suddenly brought something back to me. I recognised his face from that of the man I'd seen passing by my window on 26 April 1999, mentioned in my previous statement. The male numbered 8 on the video also brought back some form of recognition, although not as strongly as No. 2. The man I saw on 26 April was between 5ft 10 and 6ft. If I had been able to see No. 2 standing and the opportunity to see his build then I would be able to eliminate him. However, with the video footage of just his face I was deprived of the opportunity to do this."

[60]A further objection to the admissibility of this evidence was that both the de Rosneys had been transported home in the same vehicle after the parade together with Susan Mayes and in the course of that journey they had discussed the parade and learned that Miss Mayes had identified No. 2. It is convenient to deal with Charlotte at this stage since arguments of admissibility relating to these two witnesses necessarily overlapped. Charlotte de Rosnay lived at 55 Gowan Avenue. At about 9.40am she was with her mother-in-law, Stella, when she looked out of her bedroom window and saw a man crossing the road jogging. She saw him for about 5 seconds. She paid attention because he was similar to her husband - 33 to 37 years old, 5ft 10 to 6ft in height, stocky, not overweight but quite solid, dark brown hair quite long and quite wavy. His hair was blowing about and had a lot of movement. His face was quite square with no facial hair other than a sort of 5 o'clock shadow. She described him as having a sort of 'Desperate Dan' appearance. His skin was pale in tone. So far as clothing was concerned he had a suit of navy blue and a lighter blue shirt and dark shoes and she believed that she would have noticed if he was not wearing a tie. One arm was concealed, as if in his pocket, and overall his appearance was smart, not as smart as a City man more an estate agent. In cross-examination she said he did not have olive skin and was not of Mediterranean appearance. She had made a statement on 30 September 1999 and did not mention the man jogging. She thought the police were only interested in what happened when she was outside her house.

[61]She described attending the video identification parade on 5 October 2000 and said height and build was very important. She was not expecting a moustache and beard. These factors had a large effect on her ability to make an identification. During the parade she said, "I'd like to see No.2 again and there was a another half way through" – that was in fact No. 8. After viewing these two again, No.s 2 and 8, she was asked if the person she had seen on 26 April 1999 was on the parade and she said, "No I don't think so".

[62]Complaint was also made on the voir dire of the fact that Charlotte de Rosnay had an affair with one of the officers in the case and that evidence was excluded at trial. The Judge ruled the evidence of both women admissible. The complaints made were capable of evaluation by the jury and went to weight. In seeking to exclude the evidence of the de Rosnays, Mr Mansfield relied on a breach of Code D: 16 to PACE submitting that it would be unfair to admit the evidence in the light of such a breach. Code D: 16 relates to conventional identification parades and requires the police to inform the suspect and his solicitor, if present, if a post parade identification occurs. Where it does the police are required to consider giving the witness a second opportunity to identify the suspect. In a conventional parade it is, of course, extremely difficult, on many occasions quite impossible, to reconvene a parade when the volunteers have left the police station. Under Annex B there is no similar provision for the two reasons identified by Mr Pownall in his skeleton. Firstly, a suspect is not permitted to be present at a video parade and secondly no difficulty arises in reconvening a video parade. Accordingly, there was no breach of the Code made out.

[63]Before either witness had spoken to Susan Mayes each had independently and without prompting indicated an interest in or preference for No. 2 (the appellant). The most significant weakness in Charlotte's evidence was that in her statement of 30 April 1999 she stated, "I did not notice anyone in Gowan Avenue" at 9.40 and "again I did not notice anyone" at 10.30.

[64]Charlotte agreed that she had heard that Barry George had been arrested before she took part in the parade. She agreed that it was in all the newspapers. She agreed that she now knew that the person at No. 2 was the defendant and that he had been picked out by Susan Mayes and was also the preferred candidate of Stella. She said it was difficult to make an identification due to the presence of facial hair on the people participating in the video parade.

[65]We consider that the Judge correctly admitted the evidence of Stella and Charlotte de Rosnay. We recognise the shortcomings of the identification by these witnesses, but the jury was made aware of the weaknesses in these witnesses' evidence. The jury was able to assess the effect that the conversation with Susan Mayes must have had upon them and they could make full allowance for it and the other shortcomings in their evidence.

[66]Belinda Normanton lived in Gowan Avenue and left home at 9.50am on the morning of the murder. When she was near number 29 she saw two women with children (namely the de Rosnays) but she had also seen a man by a black car using a mobile phone. He was looking in the same direction as she was walking and he was in her view for two minutes. She had a clear view of him for one minute. Her description was: aged 40; height 5ft 8; medium build; black hair, long and slightly wavy but not down to his shoulders; no facial hair; face not narrow; and black suit and no tie. She had passed within 6-8 feet of him. She did not tell the police about this man on two house-to-house visits and did not record it until a year later in a statement. She said she had a condition which made her slow to connect thought process and speech. She said that after the incident she saw an E-fit of the appellant and said she would have identified it as the man she had seen. The defence contended that this undermined the accuracy of her evidence. The video parade she attended on 14 August 2000 was recorded and she said "I am sure, I am not quite sure. I cannot remember a moustache." Afterwards she made a statement that she believed No. 2 might have been the man referred to in her statement. When cross-examined she said she was sure she had not made a mistake and the person she saw on the parade was definitely the man in the street. In re-examination she said that she was sure that the man at No. 2 was the man she had seen in the street and she agreed that her initial description had included the fact that the man could have been Mediterranean or sun-tanned and of medium build. She said that was her initial impression of him and also one year later.

[67]Complaints about her evidence were that her first statement was a year after the event. The identification parade was also a year after the event. There were discrepancies between her evidence and her witness statement. She could not have had the man in her view for more than a 'shortish' space of time. She said No. 2 might have been the man referred to in her statement. She indicated that if the Stappleton E-fit had been on the parade she would have identified him. She also said in evidence that she gave a description when the police first visited. It is clear however that there were two police visits before any description was given. She was asked to cast her mind back to the parade and to say whether she had the knowledge at that time to say whether the man she had seen was on the parade. Again the jury were well able to evaluate her evidence and we consider that the Judge was perfectly correct to permit them to hear it.

[68]Terry Griffin, the postman, was delivering letters to the deceased's home at about 10.10am when he saw a man in the road looking directly at 29 Gowan Avenue. In his witness statement dated 30 April 1999 he described the man as of Mediterranean appearance, slight brown skin, black straight collar length hair, about 5ft 10 to 5ft 11 tall, medium build, wearing a dark top, may be a jacket of some sort. He was aged between 37 and 38 years old. He attended the video parade held on 14 August 2000 and when asked if the man he had seen on 26 April 1999 was on the parade he said, "No". In a further witness statement dated 14 August 2000 he said "the man at No. 2 on the identification procedure is the man I saw a few weeks later in Gowan Avenue who approached me and said he had seen the man who killed Jill Dando." In a subsequent statement he said that he was not able to say whether the man he identified was the same man he saw on 26 April 1999. He was not saying it was not him.

[69]In evidence he said that at about 10.15am to 10.20am, having delivered at number 29, a man was standing in the road opposite number 17. He said he looked quite strange and it happened very quickly. He gave a description of him, 5ft 10 tall, aged 38 to 40, and black hair just above shoulder length and just below the collar. It was straight hair. He was of medium to large build and looked Mediterranean with tanned skin. He was wearing a dark top or jacket. He was unable to recall what else he was wearing. Six weeks after the incident he was approached in the street by a man who said he had just seen the man who looked like the 'bloke' who killed Jill Dando in a café on the junction of Wardo and Munster Avenue and asked what they should do. Mr Griffin did not wish to get involved and pointed the man in the direction of a phone box. He described this man as aged 38 to 40, 5ft 10, and medium to stocky build; thin 'tache', dark black hair and wearing a white shirt and black trousers carrying a brief case. There is apparently no dispute concerning this meeting six weeks or so after the murder. It was the appellant who approached Mr Griffin. The defence of course make the point that here is a witness who was able to identify No. 2 as the man seen in June and thus he was able to identify the defendant and yet failed to identify the defendant as the man in Gowan Avenue on the morning of the murder. Again we can see no arguable basis for excluding this evidence.

[70]The final identification witness was Janet Bolton (who was asked to attend an identification parade but did not do so) who parked her car just into Gowan Avenue at 11.35 to 11.40am – she put it five minutes earlier in her statement of 30 April 1999 – and being anxious that a traffic warden might catch her on a yellow line she looked up Gowan Avenue and saw a man she thought might be a warden. She described the man saying he was about 100 yards away on the other side of the Sidbury Street junction. He was walking quickly towards Fulham Palace Road, not running. In her statement, dated 30 April 1999, she described him as a white male, tall, dark haired and was unable to estimate his age. His clothing appeared dark with a possible suit jacket undone and a lighter shirt. In her statement, dated 28 May 1999, she added that he possibly had a red tie on. In evidence, she said he was 35 to 40 years old, 5ft 10 with dark hair, not long and a dark coat, lighter shirt and possibly a red tie.

[71]Criticisms of her evidence made by the defence are that she could not give his age in her witness statement. She said she had seen the E-fit picture. She thought it similar to the man she had seen when it was not the defendant depicted. Having said in a witness statement, she was almost certain that he was the man she saw in Gowan Avenue on 26 April 1999, she was not asked to go to an identification parade.

[72]The prosecution contended that this lady was describing the appellant relying on his own witness statement in which he asserted that he had been wearing a dark suit and tie and a black overcoat. She also, like Mr Hughes and Mr Upfill-Brown, mentioned a three-quarter length coat. Again, we can see no sound basis for excluding this evidence pursuant to S. 78. The weaknesses in the evidence were highlighted for the jury and it was for them to assess them. CONCLUSIONS

[73]We now consider the overall position in relation to the identification evidence.

[74]We are satisfied that there was evidence properly admitted and properly left to the jury for their consideration. It was evidence from which the jury could conclude that each witness saw the same man and that the man was the appellant, Barry George.

[75]We recognise that Mr Mansfield has grounds for criticising the evidence given by each individual identifying witness. Allowance has to be made in favour of the defendant for the delay, the difficult circumstances of the identification and the danger created by the publication of the identikit drawing. However, when the identification evidence is looked at as a whole, it provides compelling evidence that the appellant had been at the scene of the crime at the relevant time. An examination of the schedule, set out in the appendix, confirms the picture of the evidence as a whole. The majority of the witnesses describe a person of the same age, build and height and general appearance as the appellant. He is of Mediterranean appearance. There are also the references to a Barbour and three-quarter length coat. Other witnesses in the case who knew him and describe him, namely Sophia Wallington, Louise Newall, Susan Bicknell and Ramesh Paul. They gave descriptions of height, build and hair which were remarkably consistent except it has to be noted that Ramesh Paul spoke about the appellant's blonde hair. In addition, there is the fact that Susan Mayes made a positive identification at the identification parade, Helen Scott, Stella de Rosnay, and Belinda Normanton also made a partial selection of the appellant at the identification parade as described in the schedule.

[76]We accept the Crown's submission that there was an underlying unity of description, a unity not only in the descriptions but in the circumstances of what each witness saw the man doing. There was a general consistency in the identification evidence. The Crown submitted that it was inconceivable that there could have been two such men of similar appearance behaving in a similar manner in Gowan Avenue in a period of time so proximate to the murder of Miss Dando. This Court accepts that submission having fully reminded ourselves of the dangers of mistaken identification and having looked at the other evidence capable of supporting and inconsistent with the identification. Apart from the 'identifying witnesses' there was a considerable volume of other circumstantial evidence, independent of the 'identifying witnesses', which suggests there was no error made and coincidence is not the explanation for similarity in what the identifying witnesses stated they saw.

[77]The starting point is the fact that the appellant lived in the vicinity of the crime and was admittedly not far away at the time of the crime. Next there was evidence, the value, if any, of which was hotly disputed, which on one view might establish a link between the appellant and the crime. The evidence is that which related to:(a) (A) the murder weapon and the bullet which killed Miss Dando, the appellant's familiarity with firearms and a small particle of firearm discharge residue which was recovered from his coat,

(b) (B) a fibre which was also recovered from the victim's clothes which could have come from the appellant's clothing. Then there was

(d) (D) his fascination for celebrities.

(A) The Murder Weapon

[78]A cartridge case was recovered at the scene. It was found to be made by Remington (an agreed fact). It had been fired only once and, as there were no rifling marks, it had been fired from a smooth bored handgun. The weapon had also probably been modified since the ammunition was 9mm short round ammunition and the only 9mm handguns [called "Parabellum and "Short"] which exist are not made with smooth barrels.

[79]The agreed expert evidence was that the source of the handgun used in the murder in the murder was, either(i) a deactivated or reactivated 9mm (handgun) or

(ii) a converted 9mm handgun, originally capable of firing blanks only, or

(iii) a handgun which had a smooth barrel, originally over 24 inches in length, which may or may not have been shortened.

[80]Among the documents found in the appellant's possession was a hand-written note by the appellant of the details of three handguns including a Colt Bruni handgun.

[81]The Prosecution also discovered a picture of the appellant wearing a gas mask and holding a handgun, with his finger on the trigger, which was identified at trial as a Colt Bruni. Later, the appellant said that this photograph was not one of him and denied it was taken inside his flat when it clearly was. He said the gun was a replica which had been in David Dobbin's possession. The jury could properly conclude that the appellant was lying to the police on these matters.

[82]The police also found amongst military and para-military papers, an advertisement for "deactivated" weapons showing that they could be purchased. However, they never found any evidence of tools appropriate to, or instructions on any method of, altering handguns. The appellant denied, in interview, knowing how to modify guns and there is no evidence to conflict with this, though this does not mean he could not have acquired a gun which was already modified.

[83]The Cartridge also had crimping marks on it. The appellant relied on this since this provided some support for Mr Mansfield's suggestion that the gun could have come from an eastern-bloc country which would support the suggestion that the gunman was an assassin from that part of the world with a political motive. Finally, a shoulder holster was found at the appellant's house with the other items we have mentioned.

[84]While in his Police interviews the appellant played down his interest in firearms and military matters. On the day of the murder he struck up a conversation with Miss Moorhouse about his service in the Territorial Army.

[85]The expert evidence showed that the deceased had been shot at very close quarters, since there was a small cut on the left-hand side of her head consistent with contact with the foresight of the gun. The point of entry was above and behind the left ear. At the time of her death, which would have been instant, the deceased's head would have been on or near the step with her killer bending or crouched over her.

[86]According to the written opinion of the Pathologist, Dr West, the killer was used to and experienced in handling handguns. When asked at interview the appellant denied he had ever owned a gun and said he had no access to them. However, he did say he had fired guns at a gun-club, a presumed reference to the Kensington Pistol Club which he last attended on 15 September 1982 and that he had been in the Territorial Army (in fact in 1981-1983).

[87]The Judge reminded the jury of an incident which was described by David Dobbins which occurred in 1985 when the appellant came to the door of his house and discharged a pistol at the door; he said it was a blank gun and he saw the cartridge ejected. He identified the pistol as the one held by the appellant in photograph no. 18.

[88]Susan Coombe gave evidence that the photograph was taken in the appellant's room and said that earlier in 1985 she had twice seen the appellant with handguns. On the second occasion, the appellant himself had shown it to her and then put it under the bed. She drew a picture of it which was in exhibit before the jury. (ii) The Firearm Discharge Residue Evidence ("FDR")

[89]Found in a search, at the appellant's home on 17 April 2000, was a coat hanging on the kitchen door which the appellant admitted was his. PC Cain found it and gave it to DC Isaacs who sealed the coat in an exhibits bag which was taken to a secure cleaned room at the local police station. Both officers gave evidence of the precautions taken by each of them individually as to cleanliness of attire and the system used, e.g. the use of surgical gloves to ensure the integrity of the exhibits. On 24 April 2000 the exhibit was transferred to the Amelia Street Police Photographic Studio and then returned to store. On 2 May 2000 it was taken to the Forensic Science Laboratory. At each stage a body of police officers gave evidence of the precautions taken personally and the precautionary system adopted on 17 April 2000. This was likewise at Amelia Street.

[90]No police were involved in the 17 April 2000 search who had been involved with firearms, except a PC Sankey who had found some ammunition in November 1999.

[91]In May 2000 Mr Keeley, a Senior Forensic Officer specialising in FDR, examined the appellant's overcoat. Mr Keeley also gave evidence of the system at the Laboratory. A taping from inside the internal right pocket of the coat revealed a small particle 11.5 microns in size. Two particles, which matched one another, had been found in the victim's hair near the wound, and on her raincoat. They were FDR and contained aluminium, lead and barium. Very little residue had been found since probably because most of it went into the head wound due to the proximity of the gun to the victim's head. Mr Keeley compared them with the particle found in the overcoat and they matched.

[92]The expert's evidence was that, when a gun is fired there is a discharge of 'percussion primer cap' ("CAP") residue, and propellant residue. CAP residue does not degrade. There are about five common manufacturers of CAP containing three basic chemical compounds; barium, lead and antimony to which aluminium is often added. Remingtons was one such manufacturer. The micron particle contained three of the basic chemical compounds in the CAP, as had the cartridge used to kill the victim.

[93]Tests to exclude other possible sources for the particle were carried out. As a result Mr Keeley's opinion was that the particle was not residue from a firework. Secondly, in his opinion it was not from a blank cartridge as they have a different form of primer. The Defence expert, Dr Lloyd, a Forensic Scientist of 30 years (1) experience, said it possibly was not firework residue.

[94]Mr Keeley's opinion was that the fact that only one particle of FDR was found was not significant. This was not an unusual situation. In Mr Keeley's experience CAP residue would more often than not be found on the firer of the gun, but would not be found on ordinary members of the public unless they had been associated with firearms. His evidence was that the micron particle was consistent with having come from the cartridge used in the killing. Dr Renshaw, equally well-qualified, reviewed Mr Keeley's findings and agreed with them.

[95]The jury had also to consider Dr Lloyd's evidence, who said that the particle was so small that to rely on it, one year after the killing, was "incredible". Its size "cast doubts on where it came from"- it could be the result of casual contamination. However, the main part of his evidence was directed to the places where innocent contamination of this exhibit could have taken place. Dr Lloyd was of the view that the police procedures had been flawed and contamination could have occurred at any stage, even before the events surrounding the victim's death. However, the possibility of contamination narrowed down on the evidence to whether PC Cain could be responsible by putting his hand into the pocket or the integrity of the exhibit was compromised by the photographic session of the coat at Amelia Road.

[96]In considering the question of contamination it has to be remembered that the appellant admitted to the police that he may have been wearing this coat on the day of the killing (26 April 1999) and little residue would be left by the killing which would be then available to 'contaminate' any body or thing.

[97]A suggestion that it was possible for the particle to have got into the accused's coat pocket before the relevant events by the appellant having visited a military clothing shop on 9 April 1999 was assessed by Mr Keeley as most unlikely. Dr Lloyd disagreed, in general in terms, he said that nothing could be ruled out.

[98]The next opportunity for contamination was when PC Cain found the coat. However, there was abundant evidence from DC Isaacs, PCs Cain and Williams as to the care being exercised and it was Mr Keeley's opinion that the only possibility of contamination which he could envisage would have occurred when PC Cain opened the coat pockets. He thought contamination from the search was unlikely and the risk of contamination by an officer bringing in the particle on his special clothing, though in theoretical terms "possible", was low.

[99]Dr Lloyd did not agree. The police vehicle used to take the police to the appellant's flat had not been sampled as was "proper procedure", and the gloves worn should have been 'forensic' gloves not 'rubber' gloves.

[100]In our judgment, the jury could safely be left to assess the risk of contamination. The main thrust of the appellant's case was that the photographic sessions had corrupted the integrity of the coat, the police procedures were flawed and precautions should have been taken at the studio to demonstrate that no contamination had occurred. Dr Lloyd dismissed the findings of the February 2001 examination by Mr Keeley (which showed no particles), that examination of Amelia Street, as being too late.

[101]In his evidence, Dr Renshaw identified six events which would have had to have occurred at the Studio for the particle to have been deposited on the coat from FDR at the studio. It was his opinion that as the particle was in the inside of the pocket, that contamination could not have happened unless a hand upon which the particle was deposited had placed it into the pocket. He regarded the probability of this occurring as being "low"; Mr Keeley's evidence was to like effect though he conceded he would not have wanted the photograph session to have taken place until after his examination.

[102]The prosecution called substantial evidence designed to show that no contamination could have occurred during the session but despite this evidence all possibility of contamination could not be removed. It could only be reduced. It was still, however, evidence that the jury were entitled to consider in coming to their overall conclusion.

[103]Mr Mansfield, in his written submissions on this aspect of the case, advanced a total of five grounds: In Ground 1 (iii) and (iv) : That the Judge should have acceded to a submission of no case to answer, in as far as this evidence related to the identification, and/or was wrong in directing the jury that the FDR residue was capable of providing "corroboration" or support for the identification.

[104]We are of the view that there clearly was a case to answer and that the Judge's ruling to that effect was correct. Further, in our view, the Judge was right in leaving to the jury whether any of the firearm's evidence (or evidence other than identification) of which they were sure, did in fact support any identification they were considering.

[105]In Ground 37 : That S78 of PACE should have been used by the Judge to exclude this evidence. The Judge considered S.78 and rightly concluded no judicial exclusion of the evidence was called for.

[106]In Ground 38 : That the firearm's evidence was not a matter for the jury. We reject this submission: it was entirely for the jury having been properly directed.

[107]Further Mr Mansfield suggests that the Judge's direction at pages 16, 34 and 35 was wrong about the status of this evidence. In our view a fair reading of that direction accurately reflected the evidential position at the end of the evidence. If it was as a result of further thought by the Judge as is suggested this does not matter.

[108]In Grounds 40 and 41 it is submitted(a) that there is an evidential rule that there is a distinction between primary and secondary facts

(b) the existence or non-existence provenance of a primary fact is a matter for the Court to determine as a condition precedent to the admissibility of "certain types of evidence" – (which includes FDR evidence). In our view, there is no such rule. The safeguards against irrelevant material entering the arena are in place under S.78 PACE and the trial Judge's common law powers.

(B) The Fibre

[109]A fibre was found on the deceased's raincoat. It was a common one that could have come from the appellant's trousers. If in fact it came from the deceased's raincoat that would provide strong evidence linking the appellant with the crime. However, the Forensic Expert had managed to carry out three of the four tests but not the fourth - viz extracting the dye from it, because the fibre was too short. In addition, the defence expert, Mr Webster, considered the finding insignificant. This is not more than weak support for any connection to the "raincoat" fibre but it is still a minor part of the whole picture.

(C) The Flawed Alibi And Lies

[110]The appellant provided a signed alibi statement dated 10 January 2001 in support of the alibi to which we referred in paragraph 12 of this judgment. In it he said he had been at his home until he had attended the offices of HAFAD in Greswell Street. He said he had no recollection of the precise time he left home. He believed it must have been sometime in the late morning –10.30-10.45am. He gave the route and he indicates that the journey would have taken about half an hour. He stated that he stayed at the offices until he left at approximately 1pm. Thereafter, he went to the offices of Traffic Cars at approximately 1.15pm.

[111]The Crown submitted that the appellant visited HAFAD after committing the murder. The Defence case was that he was at HAFAD at the time of the murder. The Crown submitted he must have arrived at around 12.45pm at the earliest. He stayed for 10-15 minutes and arrived at the offices of the taxi firm at 1.15pm which was only a few minutes away. Four witnesses at HAFAD indicated he only stayed a short time because he had made no appointment though there was confusion as to his time of arrival. However, the evidence from CCTV stills and as to a telephone call made by the appellant on his mobile phone supported the case for the Crown. It was also established that the telephone call could not, as the appellant suggested, have been made from HAFAD. The route the appellant states he took to Greswell Street was also a surprising one and, as the Crown argued, designed to avoid his being in Gowan Avenue. The evidence relied on by the Crown effectively destroyed the appellant's alibi. It also demonstrated that the appellant's police interview contained lies.

[112]Two days later, the appellant returned to the premises of HAFAD. It was left to the jury that there was no sensible purpose for the second visit, other than to ensure that there would be support for his wearing different clothes from those which were seen by the witnesses at Gowan Avenue.

[113]The appellant stated that he visited HAFAD on 28 April 1999 because he was concerned that his friends said he bore a striking resemblance to a published E-fit. In fact the E-fit was not published until 30 May 1999. LINKS WITH GOWAN AVENUE

[114]The appellant claimed, in interview, not to know where Gowan Avenue was. When asked to describe his route from home to HAFAD, on the day of the murder, he described a circuitous route, considerably longer than a direct route, which would have taken him along Gowan Avenue.

[115]He had a card in his flat from a doctor's surgery in Gowan Avenue. He had in his flat a photograph of a woman walking past the same surgery. He told a friend that he had a special friend who lived in Gowan Avenue and it was a lady-friend.

[116]The Crown's case was that the false alibi and the other lies which he told were evidence of the appellant's guilt. The Judge correctly directed the jury as to the significance of lies and what weight they should attach to lies and that it was a matter for them.

(D) Fascination With Celebrities

[117]A volume of material was recovered from the appellant's flat including newspapers with photographs of the deceased going back to 1990. The appellant had claimed at interview that he had not heard of Jill Dando.

[118]There were four copies of the magazine 'Aerial' (BBC internal magazine dated 27/4/99) with a full-page photograph of the deceased on the cover and inside the story of the murder. The appellant had in his flat several pictures of celebrity female newscasters, some had been taken by him, from the television.

[119]The appellant told the postman he had seen the killer. He made statements that he looked like the killer and took a considerable interest in the aftermath of the killing. Fascination With Firearms

[120]Most of the evidence supporting the Crown's suggestion that the appellant was fascinated by firearms has been referred to already. The appellant had in his possession a document setting out the recent legislation in respect of de-activating firearms. A photograph, taken in the appellants flat, depicted him wearing a respirator and holding a Bruni pistol. Part of the firearm had been removed deliberately. A witness, Susan Coombes, gave evidence that she had seen him with two other firearms. A photograph was found of the appellant wearing a mask and holding a Kalashnikov rifle. A list of guns in the appellant's handwriting was found. A copy of a birth certificate was found of a man who had acted heroically in the Hyde Park, Iranian Embassy siege.

[121]The appellant had a gunsmith's card in his possession and had various entries in Exchange & Mart relating to deactivated firearms, private investigators and ID cards. CONCLUSION

[122]In its totality, which we have not set out in this judgment, this supporting evidence was capable of being powerful support of the identification evidence. The jury could well have thought it could not have been a coincidence that the person identified should: live close to the scene of the murder; be in the vicinity of the murder on the day and at the time in question; be interested in firearms and Jill Dando; and, when interviewed, have told lies about his movements at the time. To this has to be added the scientific evidence which involved further coincidences. THE SAFETY OF THE CONVICTION

[123]As we indicated earlier in this judgment, it is not necessary to deal with the correctness of the numerous decisions, which the Judge had to make during the course of the trial, on a number of individual applications which were made both before and during the trial. However, we have read the decisions of the Judge and it is right that we should record that we consider his decisions cannot be faulted. There is no challenge to the general fairness with which the trial was conducted, rulings on Mr Mansfield's applications apart, and in fact the Judge excluded evidence on the grounds that it would be more prejudicial than probative which other Judges might have admitted. His summing up was a model of clarity and balance. He warned the jury against drawing adverse inferences from the fact that the appellant did not give evidence. He took the jury, with great care, through the identification evidence. He explained the dangers about which Mr Mansfield was concerned.

[124]The case, however, remained a difficult one for a jury, as is confirmed by their majority verdict and the time which the jury took to reach a verdict. Mr Mansfield contends that we should have at least a lurking doubt as to the correctness of the verdict. We were referred to R v Cooper

[1969] 1 QB 267

and R v Pope 85 Cr App. R 201 for this residual role of this Court.

[125]There are cases where an individual witness's evidence is critical. This is not such a case, though it is probable that if there had been no positive identification of the appellant, there would never have been a prosecution. The important thing in this case is the evidence as a whole. Looking at the evidence, as a whole, we have no doubt as to the correctness of the conviction. THE JURY ISSUE

[126]After the trial had finished one juror made an attempt to contact the solicitors acting for the appellant. She was clearly very upset. We do not know whether she was the juror who dissented. There are limitations on what can be discussed with a juror relating to a trial and everyone involved acted with perfect propriety but nothing was revealed which established that anything had gone wrong with the trial process.

[127]In addition, a book relating to the trial was published. This book had passages which Mr Mansfield submitted indicated that some impropriety had taken place involving deliberations outside court hours and unauthorised deliberations outside court hours and conversations with the jurors prior to the trial beginning on 30 April 2002, Lord Justice Mantell, Mr Justice Leveson and myself heard an application as to this matter. A transcript is available as to what occurred and of the short judgment I gave. The court did not consider that any directions were called for and no inquiry was ordered. It was, however, indicated that it would be in order for the police to make any inquiries they thought were appropriate. The author of the book was then interviewed and we have his statement. We are also in possession of other communications of which the parties are aware. Nothing has however been provided to us which leads us to believe that there has been any irregularity. We do not believe it is necessary or would be helpful to say anything more on this subject. We dismiss this appeal. TABLE OF WITNESSES FOR THE CROWN For your ease of reference, I set out below a table of identification witness specifying:(a) the name of the witness;

(b) the time of the sighting;

(d) the description given to the police;

(e) the result of any identification parades; and

(f) any inconsistencies and/or reliability issues. NAME HELEN SCOTT SUSAN MAYES STELLA DE ROSNAY CHARLOTTE DE ROSNAY BELINDA NORMANTON THE POSTMAN RICHARD HUGHES GEOFFREY UPFILL BROWN JANET BOLTON TIME OF SIGHTING Evening of 25/04/99 7am on 26/04/99 9:30am on 26/04/99 9:30am on 26/004/09 9:50 am on 26/04/99 10:15 to 10:20 am on 26/04/99 11:30 to 11:45am on 26/09/99 11:15 to 11:29 on 26/04/99 11:35 to 11:40 am on 26/04/99 LOCATION OF SIGHTING Junction of Gowan Avenue, and Fulham Palace Road. Gowan Avenue, opposite no.29 Gowan Avenue, crossing outside no.55 Gowan Avenue, crossing outside no.55 At no.29 Gowan Avenue Gowan Avenue, outside no.17 Gowan Avenue Gowan Avenue Gowan Avenue HELEN SCOTT SUSAN MAYES STELLA DE ROSNAY CHARLOTTE DE ROSNAY BELINDA NORMANTON THE POSTMAN RICHARD HUGHES GEOFFREY UPFILL BROWN JANET BOLTON DESCRIPTION GIVEN IN EVIDENCE: (a) age Late 30's/early 40's Mid to late 30's 35-40 33-37 30-40 38-40 30-40 35-40 35-40 (b) height 5"10 5"9 5"10 to 6"0 5"10 to 6"0 5"8 5"10 Taller than 5"7 5"8 to 6"0 5"10 (c) build Average, not overweight Stocky, slightly overweight Heavy or Strong Stocky and solid, but not overweight Medium Medium to Large Thick Set Fairly thick set None (d) hair Very Dark hair, sort of black, cut short, fairly short on sides, slightly longer and thicker on top, slight fringe, cut above collar Black hair, shoulder length, "short layers but quite long layers", over the collar, thick and heavy Dark brown hair cut in normal fashion, not very short, not hippy, good cut and straight Dark brown hair, quite long and wavy, messy sort of blown about Black and slightly long hair, wavy, down to collar, brushed to the left Black hair just below the collar, above shoulder length and straight Dark brown hair to ears and collar, a "thick mop without curls or waves" Black hair, almost collar length and curly, could have been a wig Dark hair not long (e) face Quite wide, not fat or round, no facial hair, eyes dark, eyebrows quite full and dark, possibly s small gap between them No facial hair Strong and round with nothing angular Quite square face No facial hair, face not narrow or thin, could not recall eyebrows though in statement to police had said they were normal eyebrows not bushy No facial hair Full and jowly face, no facial hair or glasses Clean shaven. No glasses None (f) skin Sort of Mediterranean look, slightly olive complexion, but white man Looked Mediterranean and slightly olive skinned High coloured pinky complexion, not olive or Mediterranean No facial hair but 5 o'clock shadow - a "Desperate Dan Look" Pale skin in tone, did not have olive skin nor was he of Mediterranean appearance Man could have been Mediterranean or sun tanned Mediterranean appearance, tanned skin None Pale complexion None

(g) clothes Three quarter length Barber style jacket, not waxed, brown black or green Black suit, open necked white shirt, no tie Medium grey suit, wearing a tie Suit of navy blue collar, lighter blue shirt and dark shoes, not "City Smart" but "Estate Agent Smart", probably wearing a tie Black suit, no tie, possibly "bluey" shirt Dark top/jacket Dark waxed coat, possibly Barber style Loose clothing, dark jacket almost down to knees with tinge of brown, black shoes, dark baggy trousers Dark coat, lighter shirt, possibly a red tie

(h) other None None None None Mobile Phone held in right hand, bracelet on right hand, could have been a watch face, not thin fingers None Possibly carrying a mobile None None HELEN SCOTT SUSAN MAYES STELLA DE ROSNAY CHARLOTTE DE ROSNAY BELINDA NORMANTON THE POSTMAN RICHARD HUGHES GEOFFREY UPFILL BROWN JANET BOLTON ID PARADE Date: 23 January 2001 Picked Volunteer (No. 8 on Parade). 1,2,8, then 2 and 8, then 80-85% sure of no.8 (Appellant no.2) Date: 5 October 2000 Picked Appellant (No.2 on parade). Very sure of identification Suggested made a mistake - not a mistake Date: 5 October 2000 No positive identification but thought it was Appellant (No.2 or 8, gut feeling no.2 - said she can't make a positive identification Date: 5 October 2000 No identification Date: 14 August 2000 Picked Appellant (No.2). Said was sure, then not quite sure, it was a year ago, I don't remember a moustache Date: 14 August 2000 Did not pick out man described above but picked out Appellant (No.2) as being man who approached him 6 weeks after the murder on the street (see below) Date: 14 August 2000 No identification Date: 14 August 2000 No identification Invited but did not attend HELEN SCOTT SUSAN MAYES STELLA DE ROSNAY CHARLOTTE DE ROSNAY BELINDA NORMANTON THE POSTMAN RICHARD HUGHES GEOFFREY UPFILL BROWN JANET BOLTON INCONSISTENCIES None Description of hair on 27/04/99 as short and smartly cut On re-examination, said he looked "thuggish" and nervous - not mentioned before None None None None None None Did not mention age or tie in initial (first) statement. Said she could not remember specifically if it was he or someone else that day she had seen wearing a red tie RELIABILITY View of man for about 2 minutes. Walked past him on way to shops. Saw him again 15 minutes later on return from shops. Total of 4 1/2 minutes Had heard reports/seen posters of dark haired man being sought prior to giving description Shown E-fit after made statement. E-fit not same, had rounder and wider face. View blocked by maroon car Saw his face momentarily before started cleaning the windscreen Looked around again when gone some steps on Looked three times for 5-6 seconds in total Had seen E-fit prior to giving description. No influenced in any way by the E-fit. Saw E-fit on Crimewatch, not the same as the one at page 6, face is thin and hair not long enough. Overall likeness reasonable View from bedroom window of No.55 Gowan Avenue See ID PARADE section, above Total view of approximately 10-12 seconds Saw 3/4 profile Did not see E-fit In witness statement states that does not believe she would be able to recognise the man again View from bedroom window of No.55 Gowan Avenue Glimpse of no more than 5 seconds. Diagonal View Had seen posters, e fit and spoken to Stella de Rosnay prior to giving description, but said e fit, not the man she saw E-fit jogged her memory but did not look like the man. Similarities in colour of hair and shape of face. Man in street had dark hair slightly longer On cross examination said sure he was not Mediterranean looking. At parade said " I'd like to say no.2 again and another one half way through - no, I don't think so" Though ID parade was difficult due to moustaches beards and half heights Accepted that her answers to the ID officer, being that she could not make an identification reflected her state of mind at the time of the parade 2 minute view, one minute very clear. 6-8 feet away, good look at the face gave description one year after the murder did not describe man at time of police house-to-house visit has seen E-fit on Crimewatch prior to giving description initially uncertain of identification at parade In re-examination said sure man at no.2 was person without the moustache or beard Noticed 1st man in the road as he was delivering post Did not identify anyone Approached 6 weeks later by man claiming he had seen Dando murderer in nearby café. This man was 38-40, 5"10, medium build thin "tache" wearing white shirt and dark trousers, and carrying a briefcase. Good view of second man as engaged in conversation View from upstairs window (No.31) Heard click of gate, went to window, had view of man for a few seconds and saw half face Man walking away from him but looked over his left shoulder Only a glancing view Unable to make a positive identification Saw man running down Gowan Avenue towards Fulham Palace Road, as he came out of the gate to his house Only a 1-2 second view of 3/4 face As man turned around, another brief glimpse of the side of this face Said watched man very carefully and took care to take every note that he could Learned Trial Judge described this witness as observant and careful E-fit on page 6 not similar, face too narrow and hair not floppy. E-fit page 7 similar Saw man walking Down Gowan avenue Towards Fulham Palace Road No good view Was looking for traffic wardens and once realised man was not a traffic warden, stopped looking at the man Seen E-fit and said in witness statement that she was certain that he was the man she saw

R - v- Barry Michael George [2002] EWCA Crim 1923

(A) The Murder Weapon

(B) The Fibre

(C) The Flawed Alibi And Lies

(D) Fascination With Celebrities

Cited in 9 later judgments