LORD JUSTICE PITCHFORD MRS JUSTICE DOBBS DBE and MR JUSTICE KENNETH PARKER

Royal Courts of Justice Strand, London, WC2A 2LL 14th January 2011

R v SIMON JOHN HALL

Mr Michael Mansfield QC and Ms Danielle Cooper (instructed by Stephensons - Solicitors) for the Appellant

Mr Simon Spence QC (instructed by Suffolk CPS) for the Respondent

Hearing dates: 7, 8, 9th December 2010

Lord Justice Pitchford :

On 28 February 2003 Simon John Hall, born 14 September 1977, was convicted of the murder of Mrs Joan Albert, a widow aged 79 years, following a trial before Mrs Justice Rafferty and a jury at Norwich Crown Court. The appellant was sentenced to a term of life imprisonment and ordered to serve a minimum period of 16 years. A renewed application for leave to appeal against his conviction was refused by the full court (Rose LJ, Vice President, Hughes and Gloster JJ) on 22 April 2004.

Mr Hall now appeals against his conviction following a reference to the court by the Criminal Cases Review Commission ("CCRC") dated 16 October 2009. The basis for review is that "there is now new evidence and associated argument to suggest that the fibre evidence relied upon at the trial and at appeal may have been misinterpreted and/or misrepresented". The new evidence upon which the appellant relies is that of Mr Tiernan Coyle an expert in fibre analysis.

The circumstantial case

Before we embark upon a consideration of the evidence and argument adduced at this appeal we shall describe in summary the prominent features of the circumstantial evidence at trial.

Mrs Joan Albert was found dead in the hallway of her home at 15 Boydlands, Capel St Mary, a village near Ipswich, at about 9.45 am on 16 December 2001. She had suffered multiple stab wounds, some of them penetrating with considerable force. The weapon remained at the scene. Entry had been gained to the rear of her home through a window broken for the purpose. The intruder had, in all probability, gained access to Mrs Albert's home via the back gardens of 3 Vine Walk and 17 Boydlands. This suggested that Mrs Albert's home was the specific target. The appellant lived two streets away with his parents at 8 Snowcroft. His mother, Lynne Hall, was friendly with Mrs Albert. The appellant had spoken briefly about Mrs Albert to a girlfriend, saying she was quite well off. The prosecution invited the inferences that the intruder had entered for the purpose of burglary, had been confronted by Mrs Albert and had killed her to avoid detection. On the night of the killing, 15/16 December, the appellant was in Ipswich. He spent the latter part of the night with Jamie Barker who was celebrating his 21st birthday. The appellant drove him home to Myrtle Road, Capel St Mary, arriving, according to Barker's mother who complained about the time, at about 5.30 am and, according to the appellant, 6 am. The appellant told the police in interview and the jury at trial that he arrived home at 6.28 am, the time he saw displayed on the microwave oven clock. He went straight home, a journey of some 5 minutes. His mother was up when he arrived and they spent 10 minutes over a cup of tea. Mrs Hall gave evidence that the microwave clock was faulty; her son arrived home, she said, at 6.10 – 6.15 am. There was, on the prosecution case, an unexplained interval of an hour.

Distinctive fibres were recovered from key locations in Boydlands including Mrs Albert's body and dressing gown, the immediate surroundings, the edge of the broken window and the garden fence, but not elsewhere in the house. These were identified as black nylon flock fibres. Among them were several "green" polyester fibres of two separate types. The investigating team carried out systematic examinations of 51 vehicles and 34 addresses. Found at two further addresses, associated with Mr Hall, 8 Snowcroft (where he was living at the time of the killing) and 25 Hill House Road (to which he had later moved), and two vehicles used by him, an Audi and a Citroen Saxo, were clusters of fibres said to be indistinguishable from those recovered at the crime scene. During a house to house inquiry the appellant had said that he was wearing a fleece during the night of 15/16 December 2001. At trial he said he had been mistaken. He was wearing a black leather jacket which had been left at a previous address and never retrieved. The jacket was never recovered, nor produced at trial. The appellant was arrested on 25 July 2002 and interviewed. He had already, in June when the police were taking tapings at his place of work, told a workmate, Lisa Gilhooley, and others that he was a suspect, and was concerned that fibres from Mrs Albert's clothing may be found in his Audi car because, he said, he had once given her a lift. That was not, however, the appellant's evidence. He said he had spoken to Mrs Albert twice, once on the telephone and once when he passed her home with his mother. The Crown acknowledged that the central feature of its case against the appellant was the evidence of fibre analysis.

Fibre analysis evidence at trial

Mrs Judith Cunnison was and is an employee of the Forensic Science Service ("FSS"), Hitchingbrooke Park, Huntingdon. She had been a specialist adviser on scientific strategy for the investigation of serious crimes since 1999 and had been employed by the service since 1980. She was in charge of the laboratory examinations which led to her statements of 24 July and 12 December 2002 and formed the basis of her evidence at trial. Mrs Cunnison carried out microscopic examinations of the fibres which she "sampled" from the tapings taken at the different locations. By this means she was able to judge the colour, dimensions and ends of the fibres, together with the appearance of "delustrant" particles. In addition she carried out an instrumental analysis (by MSP – see paragraph 17 below) of the composition of the dye and the chemical composition of the fibres. Chemical delustrant is added to the polymer during manufacture to reduce shine on the finished garment or article.

Mrs Cunnison examined fibres recovered from tapings taken from the top of a concrete post in the garden of 15 Boydlands (MH4), outside the kitchen window sill (MH7), the edge of the broken glass in the window (MH8), the body tapings (MH12 A-I), Mrs Albert's dressing gown (MH8), a fabric mark in the kitchen (MH17), a fence at 3 Vine Walk (MH28), a curtain at 15 Boydlands (JRG101), and a window blind at Boydlands (JRG111).

Among the fibres recovered were well over a thousand black nylon flock fibres which Mrs Cunnison found to be distinctive in appearance and indistinguishable from one another. In all the tapings save for MH8, MH17, and tapings from Mrs Albert's legs (MH12 A&B), there were found low numbers of polyester fibres of green appearance under the microscope.

Mrs Cunnison examined tapings from 34 buildings and 51 vehicles unconnected with the appellant. She found a single flock fibre in tapings taken at 3 of those addresses. None of them matched the fibres recovered from Boydlands. However, over a thousand similar flock fibres were found on tapings taken from 25 Hill House Road (NAM1), 8 Snowcroft (NAM5), the Audi (NAM3) and the Citroen Saxo (JC1A). Mrs Cunnison said that she took representative samples of fibres from each of these locations in order to examine them in detail and to compare them with fibres found at the crime scene. She was unable to distinguish between them. They were tiny straight fibres with distinctive splayed ends produced during the manufacturing process. The ends were slightly angular. She observed tiny particles of delustrant under the microscope. The chance of finding a number of fibres of a particular colour and a particular type at random in the population was, she said, "very small indeed". Evidential significance depended upon the number and variety of fibres found. In the present case the number of fibres found was extremely high. In her experience it was unprecedented. She concluded that there had been a profuse shedding of fibres en route to the house, inside the house and upon confrontation with Mrs Albert. Among the tapings from 8 Snowcroft and the Audi motorcar, Mrs Cunnison also found "green" polyester fibres of the two separate types found at Boydlands. In Mrs Cunnison's opinion, there was "extremely strong scientific evidence" of an association between the fibres found at the scene and the fibres found in locations associated with the appellant.

The FSS and the police traced an Italian company called Wonder SRL which prepared material by gluing flock onto backing for supply to clothiers and others. Wonder SRL purchased the flock from another Italian company called Cassarti which manufactured the flock from nylon. Wonder SRL provided the FSS with swatches of flock fabric for analysis from which Mrs Cunnison tweezed individual fibres. In Mrs Cunnison's view, one of the samples, AW33, did not compare microscopically either with the others or with fibres from Boydlands, but the remaining five samples were indistinguishable from one another and from fibres found at Boydlands. The relevance of this evidence was that jeans manufactured from flock of this type were sold in the local Tesco store, a store admittedly visited by the appellant twice in the 12 hours or so before the killing to obtain cash and fuel. There was, however, no evidence that the appellant had made a purchase of jeans or anything else inside the store, and Mrs Cunnison conceded that there were many other suppliers of flock fibre whose products she had not examined and who could have provided the source.

The defence had, at trial, the services of and a report from their own expert, Mr Webster, who was present during the evidence of Mrs Cunnison. No challenge was made to Mrs Cunnison's evidence that the fibres she had examined were indistinguishable. There was, accordingly, no challenge in the evidence to the techniques of her examination and the precise results of their application fibre by fibre. Mr Rouch QC, who appeared for the appellant at his trial, sought from Mrs Cunnison in cross-examination concessions as to the possibility of secondary transfer of fibres between a garment or other article from which they were shed to garments worn by the appellant at his home and in motor cars to which he had access. Mrs Cunnison was prepared to accept that in areas where there were fewer numbers of fibres found at 8 Snowcroft (4 nylon flock fibres from the living room sofa and three from Mr and Mrs Hall's bedroom; 8 polyester fibres from the hall, cloakroom and bedroom 3) secondary transfer was a possibility, if not probability. The appellant's mother gave evidence in his defence suggesting that she may have been the source of fibres within 8 Snowcroft. Large numbers of the black flock fibres were found in the appellant's wardrobe. Mrs Hall claimed to have stored a black jacket in the appellant's wardrobe. Accordingly, the task for the jury was not to examine the quality of the fibre examination but (1) the strength of the evidence of association and (2) the quality of the explanation tendered on behalf of the appellant as to how innocent transfer may have taken place. The jury was plainly satisfied that the fibre evidence together with the other circumstances proved that the appellant was the killer. No argument was addressed to the court during the renewed application for leave that the scientific analysis of fibres recovered was in any way flawed.

CCRC investigation

The appellant made an application to the Criminal Cases Review Commission on 12 May 2005 in consequence of which the Commission examined 23 separate features of the trial and the evidence adduced at trial. The Commission has confined its grounds for making its reference to the fibre evidence. In May 2009 the Commission contacted Mr Tiernan Coyle, principal of Contact Traces Limited, Begbroke Science Park, Oxford. Mr Coyle has a Masters degree in Forensic Science and has been a forensic scientist for 12 years. He specialises in the field of textile fibre analysis and on this occasion he worked with the assistance of a colleague with 4 years' post-qualification experience. Mr Coyle agreed to review the work done by FSS. On 14 July 2009 Mr Coyle notified the CCRC that he had some concerns about Mrs Cunnison's evidence that flock fibres recovered from 15 Boydlands were indistinguishable from those found at other locations associated with the appellant. He noted that "the first derivative of the spectral data" did not appear to have been produced and analysed. It is the use of this scientific technique which has formed the basis for the appellant's application under section 23 Criminal Appeal Act 1986.

Neither Mr Michael Mansfield QC nor Ms Danielle Cooper appeared at trial. Mr Simon Spence QC, who appeared as junior counsel with Mr Parkins QC at trial, now represents the respondent. It is conceded by Mr Spence that Mr Coyle's evidence as to the application of the first derivative was not reasonably available to the defence at the appellant's trial. Mrs Cunnison had not heard of its application in the context of fibre analysis when she gave evidence at trial and she was asked no questions about it. The respondent has conceded that the evidence would have been relevant and admissible at the appellant's trial. In consequence of Mr Coyle's work Mr Ray Palmer, Principal Forensic Scientist at FSS, Huntingdon, carried out a review of Mrs Cunnison's work including further analysis of his own. Mr Palmer has been a forensic scientist for 25 years spending most of his career with the FSS and its predecessor, 7 years with a police laboratory in Dundee (1990-1997), and 2-3 years with a private defence consultancy (1997-1999). Mr Roger Robson is director of Forensic Access, East Lockinge, Oxfordshire, a freelance provider of forensic science services. Mr Robson has 33 years experience as a forensic scientist, working for both prosecution and defence. He was instructed by the Suffolk Constabulary to review the work of both Mr Coyle and Mr Palmer. The court has therefore received the evidence of Mr Coyle with a view to making a decision under section 23(2) Criminal Act 1968 whether the fresh evidence (i) is capable of belief and, if so, (ii) may afford any ground for allowing the appeal. We have heard the evidence of Mr Palmer and Mr Robson adduced on behalf of the respondent.

Ambit of fresh evidence

Both in his witness statements and in his evidence Mr Coyle has expressed his opinion based upon the conventional analysis of fibres recovered. In this respect his evidence does not qualify for admission under section 23(2) of the 1968 Act since it was available at the time of trial. It is arguable that it is not now open to the appellant to seek to improve upon the expert evidence available to him at trial simply by relying upon evidence which treads old ground. Mr Mansfield accepts this. Nevertheless, the description by each of the witnesses of his microscopic examination of fibres has been a necessary prelude to the application of the first derivative to the analysis of fibres and we have received that evidence. Furthermore, submits Mr Mansfield, if the fresh evidence does provide a ground of appeal we are entitled to and should have regard to all the circumstances when considering whether the verdict returned was safe. We shall return to this subject later in our judgment.

The science of fibre analysis

There is common ground between the appellant and the respondent as to the methodology and techniques for examination. Not every laboratory will adopt the same process in every case but there is a large measure of consistency. Fibres are gathered from the crime scene by means of taping and are first examined with the naked eye or under a low powered microscope. The evidence in the present case was that the fibres recovered were so small that they bore the appearance of dust. Flock is a non-woven textile obtained by cutting lengths of fibres into sub-millimetre lengths. The fibres are electro-statically charged and sprayed onto a surface which has been previously coated with an adhesive. The result is a fabric which resembles velvet or suede. It has many applications not confined to clothing. Flock is produced from the original fibres by guillotining or grinding. The fibres can be dyed before or after the cutting process and it is not uncommon for a mixture of different yarns to be used in the process. In the presence of a large number of fibres, such as those recovered in the present case, the examiner will make a representative selection for individual examination. Under the microscope the examiner will be able to study the "morphology" of the fibre, that is, its colour, pigmentation, thickness, cross-section, and the presence of delustrant.

The second stage of examination involves the use of a high powered comparison microscope. Textile fibres, natural or synthetic, vary in their appearance under the microscope and in their reaction to light. The general morphology of the two samples is compared under white light. Further comparisons are made using other wavelengths of light in the rainbow spectrum, usually blue, and, with the use of more modern instruments, ultraviolet. By this means the light emitted from the fibres is compared both for colour and intensity. Mr Coyle explained in his statement of 11 February 2010 that when viewing two fibres side by side in "fluorescence" mode there are three potential outcomes: (i) the fibres will fluoresce at different wavelengths (seen as different colours); the fibres will fluoresce at different intensities (seen as the same colour but one brighter than the other); and (iii) the fibres will fluoresce with the same colour and intensity.

Third, modern laboratories have a more discriminating visual aide for matching or distinguishing fibres called microspectrophotometry (MSP). The fibre is exposed to light throughout the spectrum of visible light measured in wavelengths of 380 nanometres (nm) to 725 nm. FSS now uses an instrument capable of emitting and measuring the reaction to wavelengths also in the ultra-violet spectrum below 380 nm. The instrument will measure the amount of light absorbed by the fibre, called its "absorbance", when exposed to light throughout the wavelength spectrum. A computer plots each measurement in the form of a line on a graph in which the x-axis records wavelengths (the independent variable) and the y-axis records absorbance (the dependent variable). The examiner will thus see in the line graph the degree of absorbance of the fibre at each wavelength of light. The plot produced, viewed as a landscape on the graph, may be comparatively featureless (typically a parabola) at one extreme, or it may be punctuated with significant peaks and troughs at the other extreme. The appearance of the line plotted will depend upon the degree of absorbance at different wavelengths caused by the chemical composition of the fibre and its dye. The equipment permits the examiner to overlay in one graph the measurements produced at several different sections of the same fibre, or measurements produced by several different fibres.

By this means the experienced and expert examiner can study the plots produced by several different fibres for consistency and diversion. He will be able to judge whether there is any difference between degrees of absorbance at identical wavelengths along the x-axis. If so, that difference may signify inconsistencies between the chemistry of the dye applied to the fibres or the chemistry of the fibres themselves. A consistent pattern may cause the examiner to declare a match between fibres. Significant divergence will result in a mismatch and exclusion of the suspect fibre. Lines plotted on the graph either for different fibres or for different sections along the length of a single fibre are called by the experts "MSP spectra". Mr Coyle has referred to the data thus plotted as the "raw" data before application of the first derivative. In most cases the examiner will be comparing fibres recovered from the crime scene with fibres present in a reference sample, that is a known garment. In the present case the examiners were comparing one population of fibres at the crime scene with other populations found at locations associated with the appellant. This has implications for the methodology to be employed which we shall discuss later in this judgment.

First derivative of the spectral data

The experts are agreed that the examination and interpretation of graphs plotted in the form of spectra are susceptible to more detailed analysis by means of the application of an arithmetical algorithm delivered by computer software. The technique has been known to chemical analysts for many years but its application to fibre comparison is of more recent origin. By this technique the data measured and plotted along the "raw data" spectrum is subjected to further analysis. The first derivative measures the slope between two adjacent points on the MSP spectrum and its value is described as the ratio of the difference between absorbance values (y-axis) over the difference in wavelength values (x-axis) between these two points. The resulting first derivative values are plotted on a further graph whose line depicts the calculated ratios throughout the spectrum (In strict mathematical terms the derivative or derived function is the general expression for the gradient of a curve at any point, and it is found by a process called differentiation. If need be, any two points may be "smoothed" so that they can be fitted into a unique function whose derivative is revealed by differentiation). The calculated ratios may be positive or negative depending upon the relationship between the two chosen points (one higher than the other) on the raw data plot. Usually, Mr Coyle explained, the extremes of peaks and troughs and points at which the raw data plot meets zero on the x-axis represent the main and most dramatic features of the plot produced by the first derivative. Its effect is to magnify similarities and differences between rates of absorbance along the spectrum. In his statement of 24 March 2010, pages 10-11, Mr Coyle described his approach to interpretation of the first derivative spectra produced by reference fibres and a "recovered" fibre, which we do not understand to be in issue, as follows:

"Spectra taken of fibres from a garment can exhibit some variation...This may be because the dye has been taken up differently (such as in cotton fibres) or perhaps because the outermost fibres have been bleached more by sunlight in comparison to those from deeper within the fabric.... The criteria for comparing fibres to the reference population are that for a recovered fibre the spectrum should fall within the variation observed in the reference fibres. Whilst there are always areas within a reference sample that vary, there are also areas of the spectra which display a high level of consistency – such as the position of the lamda max [the wavelength at which the fibre exhibits its maximum absorbance]. When interpreting spectral matches, the entire spectrum is observed, with particular attention given to regions of the recovered fibre where the reference displays a high level of consistency. If the [recovered] fibre displays a feature in the[se] regions which has not been demonstrated in the reference population (for example, increasing in slope where the reference fibres decrease) then it is distinguished from the reference." The witnesses agree that the availability of the first derivative is now a necessary tool in the hands of the examiner. They disagree upon the issues whether it should always be used and, if it is to be used, how it should be applied.

Absence of reference sample

The witnesses all agreed that the absence of a reference source made the task of the examiner more difficult than usual. They were examining fibres recovered from the crime scene. Those fibres had, they agreed, almost certainly been shed from a garment worn by the killer. No garment was found whose fibres remotely matched the fibres recovered from Boydlands. Their task was, therefore, to examine and compare two separate "populations" of fibres, one from Boydlands and the other from Snowcroft, with a view to reaching a conclusion whether they had a common source. The experts agreed that within the Boydlands population was a large number of black nylon flock fibres which were of different sizes. It was, therefore, necessary, first, to form an opinion as to the range of fibres found in the Boydlands population. Only then could the experts embark on a comparison with fibres recovered from Snowcroft. If fibres recovered from Snowcroft were outside the range identified at Boydlands then they were "distinguishable" and of no evidential significance. Secondly, if there were significant differences between the population of black flock fibres found at Boydlands and the population found at Snowcroft the finding of fibres at Snowcroft provided no evidence that the appellant was the source of fibres recovered from Boydlands.

Mr Tiernan Coyle

Mr Coyle was first asked to review the work of FSS. He wrote to the CCRC on 14 July 2009 to say that he had viewed the files and collated and read the laboratory notes. He had "viewed the VIS-MSP spectra taken at the time of the investigation and additional UV-VIS MSP spectra taken by the FSS in response to a recent request for further work by the CCRC". Mr Coyle did not then and there express the view that the MSP spectra taken from the principal populations were distinguishable. Mr Coyle wished to carry out his own work on the FSS fibre samples. Some 3,000 flock fibres were captured in tapings taken from Mrs Albert's body and clothing. Mr Coyle re-examined selected fibres recovered from those tapings microscopically. He found that while there were very many black flock fibres, they came in different thicknesses and, in his view, they differed also in their levels of delustrant particles. By visual estimation, but not instrumental measurement, he classified the range at Boydlands as comprising:

(1) 10 microns (thousandths of a millimetre or "µm") semi-dull nylon;

(2) 10 µm bright to semi-dull nylon;

(3) Thicker than 10 µm (i.e. 12.5, 15 and 20 µm) semi-dull or dull.

Following his first examination Mr Coyle advised that further work of sampling was required. He then embarked upon the exercise of selecting his own representative sample fibres from the tapings at Boydlands. His selection may be represented in a table as follows:

Exhibit ref. Description of location No of fibres selected MH4 MH28-B MH7 MH8 MH17 MH12-A MH12-E MH12-H MH12-I Concrete post Fence post Outside kitchen window sill Edge of broken glass in kitchen window Fabric mark Exposed right leg Exposed left leg LH side of dressing gown RH side of dressing gown 10 10 10 10 2 10 10 10 13

Mr Coyle found within MH12-E both thick and thin flock fibres which he mounted separately for examination. Within MH12-I he found three flock fibres of different thicknesses attached together with glue. This is evidence that the population range at Boydlands was particularly wide. Mr Coyle found that his selection fell into the same categories as those he had earlier identified in the FSS samples, although he specifically targeted thicker fibres which he found were "bright to semi-dull".

Mr Coyle performed a similar exercise with the Snowcroft fibres as follows:

Exhibit ref. Description of location No of fibres selected NAM5-H Wardrobe floor left 10 NAM5-I Wardrobe floor right 10 NAM5-F Bean bag cover 10

Mr Coyle concluded that the majority of nylon flock fibres from Snowcroft were "approximately" 10 µm thick, black and semi-dull. The rest were "approximately" 10 µm thick, black and "bright to semi-dull". One flock fibre was recovered from NAM5-F which was 20 µm thick, black nylon and "semi-dull".

Mr Coyle proceeded to make microscopic comparisons, fibre by fibre, under white, blue and UV light conditions between his selections from Boydlands and his selections from Snowcroft. He compared:

(i) 10 x 10µm semi-dull fibres from MH28-B (point of entry) with 10 x 10 µm semi-dull fibres from NAM5-H and I (wardrobe);

(ii) 11 x 10 µm thick, bright to semi-dull, flock fibres from MH 4, MH 7, MH 8, MH12-E, MH12-H and MH12-I with 4 x similar 10 µm thick fibres from NAM5-H and I (wardrobe);

(ii) The single 20 µm thick fibre from NAM5-F with 2 x "thicker" flock fibres from Boydlands tapings.

When comparing fibres under white light conditions, Mr Coyle saw subtle differences in morphology. Under blue light he saw differences in "shading". He thought Boydlands fibres tended to red tint and Snowcroft fibres tended to blue tint. He concluded that with one exception none of the fibres from Snowcroft was indistinguishable from the Boydlands fibres "although the differences were very subtle indeed". It follows that Mr Coyle disagrees with the opinion of Mrs Cunnison expressed unchallenged at the appellant's trial. Mr Coyle's usual practice would have been to eliminate the Snowcroft fibres at this stage but he proceeded to MSP.

Thus, Mr Coyle produced 106 MSP spectra in all, 78 for Boydlands and 28 for Snowcroft. After examination of the results he concluded that:

"...it can be seen that they [Boydlands and Snowcroft fibres] are subtly, but consistently, different. Some differences are present on the raw data plot (see appendix figure 13) but more are obvious on the first derivative plot (see appendix figure 14)........With the exception of one flock fibre recovered from the fence post, all of the flock fibres recovered from 15 Boydlands were distinguishable from those recovered at 8 Snowcroft. In my opinion, the findings provide no support for the prosecution that the flock fibres...arose from the same source....As a single flock fibre indistinguishable from those found at 8 Snowcroft could have been transferred to the fence post via contacts other than a direct contact, I am reluctant to attach any weight to its finding." We have photocopies of Figure 13 and Figure 14 in our file of witness statements. We can follow what Mr Coyle described as subtle differences. The general profiles of the raw data and first derivative spectra from fibres recovered from Boydlands and from Snowcroft respectively appear very similar but at least some of the spectra in each population profile appear to be slightly different.

Mr Coyle further demonstrated his analysis of MSP and first derivative spectra to the court with a most helpful pictorial presentation. He illustrated the technique at page 12 of his presentation bundle where Mr Coyle had photographed side by side the spectral graphs of fibres taken from two different garments (balaclavas A and B) and pointed out the slight but noticeable difference between them. Mr Palmer, when he gave evidence, agreed; had he been examining these spectra in the laboratory he too would have concluded that they came from separate garments. Turning to the fibres recovered in the present case, at page 33 of his presentation bundle Mr Coyle had juxtaposed three spectral graphs, the first comprising raw data spectra, the second first derivative spectra and the third, a zoomed-in section of first derivative spectra. The equipment did not permit Mr Coyle to depict the spectra for all 106 fibres on one graph; the limit was 48. He chose a mixture of Boydlands and Snowcroft spectra, Boydlands in black and Snowcroft in red and overlayed the results. He has pointed out to the court areas in which the "red" and "black" line spectra appear not to follow a closely consistent pattern both in the raw data graph and in the first derivative graphs. They appear to confirm Mr Coyle's impression on microscopic examination that there are subtle differences between the Snowcroft and Boydlands fibres, similar to those which Mr Palmer agreed could be seen at page 12.

Mr Ray Palmer

When he first reviewed Mrs Cunnison's work with the help of Mr Coyle's statements, Mr Palmer said (witness statement 22 January 2010, page 9) that he would not regard as significant the "weak differences" in fluorescence between the Boydlands and Snowcroft populations described by Mr Coyle, which he was "unable to confirm" for himself. It was his view that there were "two populations of...fibres exhibiting the same degree of variation in microscopic morphology...in two ostensibly unrelated environments". If a slight difference in fluorescence existed between the two populations, it might be explicable "when the compared fibres have been recovered from very different environments over a significant time frame". The difference may be explained by washing, wear and tear or the differences in the adhesive tape used to recover the fibres (statement 30 March 2010, page 7). It would be a different matter if under the microscope the colour itself was different between the two populations. Furthermore, slight differences in fluorescence observed were "often due to slight variation in the fluorescent behaviour (brightness and intensity) between individual fibres comprising a particular garment (i.e. intra-sample variation) and is usually resolved through additional sampling of the donor garment in question". In the present case there was no donor garment available. Mr Palmer posed what he regarded as the key question for the examiner in his witness statement of 22 March 2010, page 7:

"Does the variation in morphology and results of instrumental analysis of the fibres in each population fall within the same observable limits or are they incongruent in that respect?" Mr Palmer regarded it as vital that the true range of the intra-fibre and inter-fibre variation in each population was ascertained. In other words, it was necessary to discover not only the range of absorbance values between fibres in the same population but also the range of variation in absorbance values along the length of each fibre in each population. The reason is that variation within the length of a single fibre is just as capable of influencing the judgement of range within a population as is variation between fibres within the same population. A single spectrum from a single fibre is taken from one position, but a single fibre may produce different spectra depending upon where along its length the measurement is taken. Accurate assessment of the intra-fibre and inter-fibre range within each population will critically affect the judgement whether both populations shared a common source. Mr Palmer agreed with Mr Coyle that the selection of sufficiently representative samples from each population is essential to the efficacy of the comparison between them. This is particularly so where the range of variation is clearly wide. Mr Palmer chose 18 of the fibres from Boydlands, sampled by Mr Coyle, and 18 from Snowcroft. He chose 12 fibres from Boydlands, sampled by Mrs Cunnison, and 10 from Snowcroft. He selected 30 more of his own from the Boydlands tapings and 30 more from the Snowcroft tapings. He used, therefore, a total of 60 fibres from Boydlands and 30 from Snowcroft. Mr Palmer agreed with Mr Coyle that most of the fibres in both populations were 10 µm in diameter. There were other thicknesses but Mr Palmer saw them in both populations.

When he viewed the fibres with and without magnification Mr Palmer reached the following conclusion:

"The fibres in each population are dark (black) in colour to the naked eye but appear almost purple under high power microscopy. I can confirm they are flocked and are derived from nylon fibres. I can confirm there is a variation [with]in each population in terms of the fibre diameter, delustrant concentration, morphology and colour intensity of each population; however, the range and degree of this variation is congruent between each population." Mr Palmer noted that Mr Coyle had described the microscopic differences between the populations as "very subtle indeed" and that he had made some eliminations "solely on the basis of weak fluorescence differences" (Coyle statement 15 September 2009, page 10). When the weak fluorescent differences were pointed out to him using Mr Coyle's instruments at a meeting of the experts at Contact Traces' laboratory on 22 February 2010, Mr Palmer's view was that they were not such as to exclude the Snowcroft fibres from the Boydlands range.

During his examination of the Boydlands fibres Mr Palmer discovered a pair, recovered from Mrs Albert's face, which were still attached by adhesive applied during the manufacturing process. These were photographed for presentation and formed pages 21 and 22 of the respondent's presentation bundle produced to the court. The first photograph depicts the appearance of the fibres under white light, still joined by adhesive, and the second depicts the difference between them in intensity of colour observed under blue light microscopic examination. The significance of this finding is obvious. Both fibres must have been shed from the same garment in Mrs Albert's home. Yet even fibres which were conjoined displayed a variation in intensity of colour or fluorescence. Mr Palmer observed that they reflected the differences and similarities present in the populations respectively from Boydlands and Snowcroft.

Mr Palmer accepted that Mr Coyle had also carried out MSP examinations along the lengths of individual fibres although he was unable to discern what was the size of his sample. Mr Coyle had concluded when taking spectra from Snowcroft fibres that their spectral shape varied according to the thickness of the fibre. When Mr Palmer performed a similar exercise, the results of which were illustrated to the court at page 3 (Boydlands) and page 10 (Snowcroft) of the respondent's presentation bundle, he found that a large proportion of individual fibres showed considerable variation in spectral results along their length. He found that the differences were, as he expected, "exacerbated" further by the application of the first derivative. In Figures 5A (raw data spectra) and 5B (first derivative spectra) appended to his witness statement of 30 March 2010 Mr Palmer demonstrated how two fibres each from Boydlands and Snowcroft produced variable spectra along their lengths under MSP. However, when each of the two fibres from Boydlands was compared with each of the two fibres from Snowcroft it was discovered that the range of intra-fibre variation at Boydlands compared closely with the range of intra-fibre variation at Snowcroft.

Mr Palmer moved to inter-sample variation. For this purpose he returned to the Boydlands population. There were two ways of examining the spectra created by Boydlands fibres. The first was to examine the composite picture as Mr Coyle had done at page 33 of his presentation; the second was to ascertain whether there was within the Boydlands population more than one "type" of response to MSP. Mr Palmer found that within the Boydlands population there existed two marked categories of spectra which he called Type 1 and Type 2. These are graphically illustrated at pages 5 – 8 of the respondent's presentation bundle. Mr Palmer demonstrated that both Types were produced by individual fibres; in other words, the MSP response varied by Type according to the place along the length of the fibre to which MSP was applied. Mr Palmer then carried out the same exercise on the Snowcroft fibres. The results are illustrated at pages 10 – 15 of the respondent's presentation bundle. They showed that among the Snowcroft population there were also two discernible Types of spectral response to MSP within individual fibres. Finally, instead of overlaying mixed samples of the population at Boydlands with mixed samples of the population from Snowcroft, Mr Palmer overlayed the samples by Type. He discovered a close congruence between Type 1 Boydlands and Type 1 Snowcroft and a further close congruence between Type 2 Boydlands and Type 2 Snowcroft (pages 16 – 17).

When, however, Mr Palmer overlayed Type 1 spectra from Boydlands with Type 2 spectra from Snowcroft, he found that there was a mismatch between them (page 18); likewise, when Mr Palmer overlayed Type 1 Snowcroft with Type 2 Boydlands (page 19). When all spectra were overlayed, one with the others (as Mr Coyle has demonstrated at page 33 of his bundle), there was an appearance of difference between some fibres in the two populations, as shown by Mr Palmer's cross-overlays of Types, when the real difference was in the intra-fibre variations within each population; the same intra-fibre variation appears in both populations demonstrating a match rather than an exclusion. Mr Coyle had explained the appearances at page 33 of his bundle as the product of different dye combinations in the populations respectively (statement 12 October 2010, page 8, para. 8) while, in Mr Palmer's view, the results were attributable to differences in dye uptake by, or differences in thickness of, individual fibres, or both.

The appropriate use of the first derivative

Mr Palmer was critical of Mr Coyle's use of the first derivative. He accepted that it was a useful and sometimes essential tool in the hands of the examiner but it needs to be employed with caution. Mr Palmer was joint author of an article published in the journal Science and Justice in 2007 entitled "An investigation into the use of calculating the first derivative of absorbance spectra as a tool for forensic fibre analysis". He explained that the first derivative had been used by analytical chemists for decades. The FSS had used it since the 1970s when instrumentation was not as sophisticated as it is now. By the late 1990s MSP instruments had improved significantly in quality. Aware of the use made of the first derivative by Mr Coyle and others, Mr Palmer and his colleagues at FSS carried out laboratory research to test its robustness upon a variety of known samples of fibres of different types. The resulting article was presented to Science and Justice in November 2005 and peer reviewed during the following year. For present purposes it is sufficient to quote the "Abstract" of its conclusions:

"A range of fibre samples was measured using J&M MSP400 and J&M MSP800 microspectrophotometers across the visible and UV/visible wavelength ranges respectively. The first derivative of the absorbance spectra was then calculated and studied. When the absorbance spectra produced for some samples were broad and featureless, the first derivative spectra provided more points of comparison that facilitated discrimination. For many of the samples, calculating the first derivative did not result in any additional discrimination due to the high number of points of comparison present in the absorbance spectra. However, for the samples that exhibited a high level of intra-sample colour variation (e.g. through uneven uptake common in cotton and wool, etc.) which was evident in the absorbance spectra, the associated first derivative spectra highlighted this variation between the fibres and could potentially have resulted in false exclusions. The results show that whilst calculating first derivative can be a useful aid in the comparison of spectra, a high degree of caution is required when applying this method to fibres which exhibit large intra-sample variation in colour." Mr Palmer had heard of no criticism of the conclusions reached, although it is clear that, while FSS and other UK and European laboratories are more discriminating in their use of the first derivative, some laboratories routinely employ it in fibre analysis. To Mr Palmer's knowledge the FBI does not use it at all.

Mr Palmer's view was that this was a case in which the first derivative offered no extra discrimination; it merely highlighted known intra-fibre variations within both populations of fibres. When, therefore, the first derivative spectra from fibres taken from Boydlands and Snowcroft were overlayed together, they caused Mr Coyle to make a false negative.

Thin Layer Chromatogrophy (TLC)

As we have said, MSP is a measurement which is made in consequence of the reaction of chemical dye in the fibre to wavelengths of light. Differences in colour or concentration of dye will be represented by distinguishable MSP spectra. TLC is a further technique by which the dye itself is extracted from the fibre and analysed. TLC is a technique used widely throughout the world and is supported by scientific literature and standard texts. In Mr Palmer's view TLC is of particular value where, as here, fibres are deeply dyed and opaque. On 4 March 2010, in Mr Coyle's presence, Mr Palmer performed the TLC test on 80 bulked fibres from Boylands and 80 bulked fibres from Snowcroft. Had there been a difference between the populations of colour or concentration of dye Mr Palmer would have expected that phenomenon to be revealed. The results are shown at page 38 of Mr Coyle's presentation bundle. They are indistinguishable. This, suggests Mr Palmer, is confirmation of his MSP analysis: the populations from Boydlands and Snowcroft match.

Mr Coyle disagrees. In his statement of 11 February 2010, pages 2-3, he said:

"In summary, TLC separates the dyes used in fibres. However, it does not inform as to the relative concentrations of individual dyes. For illustration purposes only, if a fibre from one scene is dyed with 50% dye A, 26% dye B and 24% dye C, a TLC plate will probably show three spots in three different positions on the plate. Another fibre from another scene but dyed with 50% dye A, 24% dye B and 26% dye C will also provide the same three spots in the same positions. It will be impossible to tell them apart by TLC – even though the dye combinations will be different." On the contrary, an MSP spectrum would represent a complex (and precise) reaction of the three components A, B and C. Mr Coyle said that when the fibres from the two balaclavas, whose spectra are depicted at page 12 of his presentation bundle, were subjected to TLC they produced indistinguishable plates; yet, as Mr Palmer agreed in evidence, the MSP spectra were distinguishable. The explanation, in Mr Coyle's view, is that in the case of the balaclavas MSP measurement was more discriminating and demonstrated that the balaclavas were manufactured from different batches of the same dyeing components. Mr Roger Robson When Mr Coyle joined Forensic Alliance at the commencement of his career Mr Robson was his manager. It was Mr Coyle who introduced the first derivative to fibre analysis at Forensic Alliance. Mr Robson clearly had respect for the quality of Mr Coyle's work. Mr Robson acted as an independent assessor of the work done by Mr Coyle and Mr Palmer on the Boydlands and Snowcroft fibres. He visited each laboratory in turn. Mr Mansfield was, in our opinion, right to observe that Mr Robson was a conscientious and careful witness.

Following Mr Robson's visits he expressed himself as follows in a statement dated 24 March 2010, page 5:

"My observations [at Contact Traces] tended to confirm what I had already observed using the FSS microscope, namely that there was a range of dye intensity, fibre diameter and levels of delustrant common to both populations. The variation in physical characteristics was typical of that expected from flock fibres. I had, however, some concerns over a subtle, yet consistent, difference in fluorescence between the populations which I believe was more easily observed using Mr Coyle's microscope than that I had used at the FSS. I also understand that both parties have observed what appears to be a greater quantity of thicker, darker flock fibres at the crime scene and some of the flock fibres at the scene appear with the gluing agent intact... Since the release of Mr Coyle's statements the FSS has had the opportunity to review his work and, as a consequence of the challenge, they have performed a considerable amount of extra work to understand matters of discrepancy for themselves... As a result...Mr Palmer has rebutted two main aspects of that of Mr Coyle's, namely whether the variation in fluorescence is sufficiently significant to discriminate between the two populations, and whether the application of the first derivative to black nylon flock fibres is a safe test for elimination purposes. I have read Mr Coyle's response, dated 11 March 2010 in which he appears to stand by his initial findings." Mr Robson proceeded to examine and compare the different methods employed by FSS and Contact Traces which, in his view, had brought about the differences in their conclusions. At page 6, he continued: "The FSS Microscope Comparison Method: a scientist examines a sub-sample of the population from or connected with a potential source – in this case the suspect's house – noting the range of physical characteristics. S/he then takes sub-samples of fibres from various locations at the crime scene, again taking note of the range of physical characteristics. By doing so s/he builds up a memory bank of the range. They then go on to compare directly a quantity of the fibres and take the range of characteristics into account. All positive findings are then checked by a second scientist and any disputes resolved by a third scientist acting as arbitrator – usually another reporting scientist. The Contact Traces Microscope Comparison Method: after sub-sampling from various areas related to the scene and comparing the samples to each other the examiner builds up a catalogue of the range. The examiner then takes sub-samples from the suspect's environment and duplicates the process. Finally, both populations are systematically compared; each fibre from one sample being compared with each fibre from the other. A fibre is excluded if it does not share very similar physical characteristics. A second scientist then checks the positive findings and further eliminates any individual fibres if s/he considers one to be different. There is no third adjudicator and any discrepancies are automatically eliminated. Therefore, only the very best of matches remain." Mr Robson concluded that both methods were valid since they each explored and compared the populations concerned. The Contact Traces method tended towards exclusion. The FSS method was more likely to be inclusive. Mr Robson tried both methods and found that the FSS method tended to obscure subtle differences in fluorescence, while Contact Traces tended towards over-elimination because of inherent variability within the sub-samples. Mr Robson expressed his opinion as to the subtle differences of fluorescence as follows (page 7):

"In my opinion, there are various mechanisms and degrees of contact that could lead to the sort of subtle differences between the two populations of black flock fibres in this case. For instance, it is perfectly feasible to propose that the greater number of thicker fibres at the scene could have been transferred as a result of increased levels of contact through a struggle and/or the moving of the deceased's body. This would be supported by the observation of occasionally finding the gluing agent still to be intact on fibres at the crime scene." He concluded:

"...I consider the black nylon flock fibres in question to be unusual source of fibres for examination in a forensic laboratory. The range observed in both populations is complex, more so in that observed at the crime scene. It is important to recognise that this is not a simple case of assessing whether the fibres match or not. Without the garment(s) as reference, this comparison and its evaluation was never going to be easy and remains open to challenge. I believe it to be unlikely that the comparison microscope used by FSS at the original trial, which probably reflected the technology then, would have had the capability to detect the subtle differences in fluorescence in as much detail." Upon the ultimate issue Mr Robson said (page 8):

"I am in agreement that there are some differences between the two fibre populations. I am in disagreement with Mr Coyle that these differences are sufficient on their own to exclude an association between the populations." Given Mr Palmer's further work on intra-fibre and inter-fibre variability, research into fluorescence, and the mechanisms of transfer discussed, Mr Robson's view was that such differences as there were between the two populations were explicable. This was particularly so where the available information is that it is highly unlikely that fibres will be found in large numbers just by chance. At this stage of his investigation Mr Robson took the view that the scientific evidence provided "moderately strong support" for the assertion that the flock fibres found at the appellant's address were associated with those recovered from the body of Joan Albert. Mr Robson said in evidence that if he had come to the work fresh, as Mr Coyle had done, he too would have used the first derivative. However, having seen for himself the range of intra-fibre variations within both populations, he had concluded that the first derivative served no useful purpose. Mr Robson found "a considerable amount of variation" particularly within the Boydlands population. There was a range of morphology which was similar between the populations and he found "congruence" between the variations in both populations. Mr Robson explained himself further in cross examination. He confirmed his view that the different microscopic results obtained by Mr Coyle on the one hand and Mr Palmer on the other was a product of the difference in their approaches. Both approaches were valid. However, Mr Robson was unable to confirm visually the different categories within the Boydlands population which Mr Coyle had identified (para. 23 above) visually but without measurement. What Mr Robson saw was one range. There was no control garment against which to compare the Boydlands fibres. For this reason Mr Robson concluded that Mr Coyle had been too exclusive. Under the microscope Mr Coyle was making a comparison fibre by fibre which "in a way" was more scientific, but this approach failed to give sufficient attention to the whole range of variation in fluorescence found in both populations. Mr Robson thought that a combination of both methods was the best approach.

Polyester fibres

We have referred at paragraph 9 above to Mrs Cunnison's evidence at trial about the discovery of "green" polyester fibres. These fibres, it is now agreed, are not green but carbon black. There were two separate types of polyester fibre found at Boydlands, one pigmented with carbon black only and the other pigmented with a combination of carbon black and blue. Mr Coyle reached the following conclusion (statement 10 August 2009):

"Whilst the fibres from the different areas at 15 Boydlands were deemed indistinguishable from each other, they were different when compared to those fibres recovered from 8 Snowcroft with the exception of one fibre found on the floor of bedroom 3 (Rec number 196, NAM52, slide 1, number 4). This fibre was indistinguishable from one of the types of carbon black pigmented polyester fibres observed. However, in my opinion modern forensic fibre analysis techniques are unable [to] discriminate carbon black pigmented fibres to a high level, and this, together with their widespread use as a cheap way of producing black fibres, makes carbon black pigmented fibres poor value as a target fibre. Any links based on carbon black pigmented fibres should be treated with a high degree of caution." Mr Palmer was not so dismissive. He examined the fibres recovered by Mrs Cunnison together with 134 further fibres selected from tapings at Boydlands and Snowcroft. He agreed as to the correct colour identification of the polyester fibres. Like Mr Coyle he found two microscopically distinct types of polyester fibre in the Boydlands population which he called Type 1 and Type 2. Both types were present on MH4 (concrete post), MH12 (body tapings) and MH28-B (fence post). He found one fibre indistinguishable from Type 1 in NAM3 (Audi), three Type 1 and two Type 2 fibres in NAM5 (wardrobe), and one fibre each of Types 1 and 2 in NAM52 (bedroom floor). He agreed that, of themselves, the polyester fibres were of limited evidential value because polyester fibres of this type were comparatively common; there were few discriminating features available and relatively few numbers had been recovered. However, they provided another link between the populations of fibres recovered from Boydlands and Snowcroft and, to that extent, they provided some support to his conclusions concerning the black flock fibres.

Mr Robson examined Mr Palmer's samples under the microscope in different colour conditions and agreed with Mr Palmer's findings. At page 9 of his witness statement of 13 May 2010 Mr Robson said:

"...given the black polyester fibres were recovered, in particular, from the deceased's skin in the same location as that of the nylon flock then it is reasonable to suggest that even if they did originate from more than one garment they were transferred together and at the same time. As such the findings may be considered in association to each other. Furthermore, if it is assumed that the black polyester fibres have originated from the same source as the black nylon flock fibres, as proposed by Mr Coyle, then the additional presence of two types of black polyester, in my opinion, provides further support for the assertion that the nylon flock fibres recovered from Simon Hall's wardrobe are associated with those recovered from the body of Joan Albert. Either way, in my opinion, the presence of three fibre collectives recovered from the body of Joan Albert and the wardrobe of Simon Hall provides strong scientific support for the assertion that the fibres are associated with each other and have originated from a similar source or sources." [emphasis added] In evidence Mr Robson confirmed this opinion.

Frequency

In the course of her evidence, which was accurately summarised by Rafferty J in her summing up, Mrs Cunnison told the jury that the number of fibres recovered in the present case was, in her experience, unprecedentedly high. To a scientist the finding of a dozen relevant fibres would be significant. In each of two places, Boydlands and Snowcroft, there were well over a thousand (we now know there were 3,000 or so recovered from Boydlands). During her 22 years of experience nylon flock fibre had been very infrequently met and she had herself only come across it once before. She concluded that the garment which shed these fibres shed them profusely en route to the scene and at the scene itself.

We are invited, when considering the safety of the verdict, to consider the witness statement of Josephine Jones, also of Contact Traces Limited, who carried out target research with a view to ascertaining the frequency of appearance of black flock nylon fibres on garments selected at random. She was particularly interested in the distribution on garments of flock which may have been transferred from the interior surfaces of motor cars, but we are not concerned with this aspect of her work. Her work was peer reviewed by Mr Palmer and the results are agreed. One hundred garments selected on the wearers at random were taped front and back. Ms Jones was looking to recover synthetic flock fibres nylon, polyester or other polymer types. At least one synthetic flock fibre was recovered from 82 of the garments; a single fibre was recovered from 19 garments; at least two were recovered from 63; black pigmented polyester flock fibres were present on 20 garments; carbon black nylon flock fibres were present on 13 garments; 16 garments contained at least two microscopically similar flock fibres; of these, 10 garments had one population of at least two synthetic flock fibres and 6 had two populations of synthetic flock fibres, each population comprising at least two fibres. Of the 82 garments from which flock fibre was recovered, 84% contained nylon fibres. The number of garments from which more than 10 nylon fibres were recovered was 8, four of whose fibres were black, two brown, one red and one colourless. There was one garment from which more than 50 nylon fibres were recovered and no garment from which more than 100 nylon fibres were recovered. We have emphasised in italics those findings which seem to us to be particularly relevant to the present appeal.

Industrial inquiry

We have described Mrs Cunnison's evidence concerning the supplier Wonder SRL at paragraph 10 above. Mr Coyle has also reviewed Mrs Cunnison's work on the fibres labelled AW34, 38, 39, 43 and 45. His conclusion was that the sample fibres were, unlike the Boydlands and Snowcroft populations, of uniform thickness (10 µm). AW38, 39 and 48 were microscopically and chemically indistinguishable from each other. AW34 and 43 were distinguishable from each other and from the other three. In Mr Coyle's opinion, none of them was indistinguishable from the fibres recovered from Snowcroft and Boydlands. As we understand it neither Mr Palmer nor Mr Robson has carried out his own examination of these fibres.

Mr Mansfield has submitted that while this is not fresh evidence it is evidence of a further discrepancy between Mrs Cunnison's and Mr Coyle's results which should be influential in a decision whether the verdict of the jury is safe.

Discussion

It is plain to this court that the judgement whether two or more textile fibres of similar dimensions and similarly dyed are distinguishable is, first and foremost, a matter for the experienced and expert examiner. There is no measurement which, by itself, is capable of making that judgement. Under the microscope subtle differences in the intensity of colour or brightness are likely to result in some disagreement between experts. MSP spectra, while based upon measurement by an instrument of wavelengths of light absorbed, will require interpretation by the scientist, as will spectra produced by the first derivative. The chemical composition of a dye extracted by TLC is represented by bands on a plate which also require the interpretation of the scientist. We accept Mr Robson's evidence that the probability is that (1) microscopic instrumentation is more sophisticated now than it was in 2002 and (2) differences in interpretation between Mr Coyle and Mr Palmer is the result of their differences in approach to the analysis of the populations of fibres at Boydlands and Snowcroft. While we are satisfied that Mr Coyle and Mr Palmer set out conscientiously to achieve accurate results and sustainable opinions, we have reached the firm conclusion that Mr Coyle's evidence must be rejected for reasons which we shall explain.

We are required by section 23(2)(a) Criminal Justice Act to consider whether the fresh evidence is capable of belief. We have no doubt that the evidence of all three witnesses is capable of belief.

Having seen for ourselves the MSP raw data spectra and equivalent first derivative spectra produced by Mr Coyle, we accept Mr Palmer's evidence that Mr Coyle has with the production of the first derivative simply "exacerbated" differences in the MSP which, if Mr Coyle's scientific approach was correct, were there to be seen in any event. We do not consider that the "fresh" evidence given by Mr Coyle provides the appellant with any ground upon which the appeal may be allowed under section 23(2)(b) of the 1968 Act.

We cannot, however, ignore the fact that the fresh evidence has occasioned, at Mr Coyle's invitation, a wholesale review of the fibre evidence, including the conclusions properly to be drawn from conventional examination and analysis. It is now agreed that the populations in these various locations had a range of microscopic appearance and reaction to wavelengths of light. It was therefore inevitable that comparison between unlike fibres in the separate populations would produce differences, however subtle. Nonetheless, that was not the way in which the fibre evidence was presented to the jury. Mrs Cunnison gave evidence that she found the individual fibres sampled from Boydlands had "similar width to one another, similar dimensions, and the thing of particular interest is the widened aspect of the fibre you can see through it because it's been squashed. You can see through the fibre, you can see tiny dots...They're called delustrant and they're part of the manufacturing [of the] fibre...It can be present or absent or present in differing amounts" (transcript 17 February 2003, page 13C-H). When she came to describe her comparison with the fibres associated with the appellant, Mrs Cunnison illustrated her opinion that they were indistinguishable from the Boydlands fibres by referring to photograph pages 58 – 61 of the jury bundle and to "[t]he general appearance, the uniform cross section, the colour, the appearance of the ends" and to the delustrant "apparent in both the fibres from Snowcroft and fibres from the body of Joan Albert" which appeared to be indistinguishable (transcript pages 19E-21A). Mrs Cunnison does not appear to have identified either the categories of thickness to which Mr Coyle has referred or Type 1 and Type 2 intra-fibre variation as identified by Mr Palmer. We have not heard from Mrs Cunnison and thus do not know for certain whether she had found any variation common between the populations. It seems probable to us that Mrs Cunnison was indeed comparing fibres of the same thickness in which case, having regard to Mr Coyle's evidence, any differences were likely to be very subtle indeed and, in Mr Robson's opinion, difficult to see on the instrumentation available at the time. It was Mr Coyle's view that the number of fibres sampled in 2002 was inadequate. We agree that the fewer the number of fibres analysed and compared the less likely it is that a range of variation would have been discovered.

Furthermore, all experts agree that the polyester fibres examined by Mrs Cunnison were not green but carbon black.

The work done by Mrs Cunnison was available for inspection by Mr Webster and it is not suggested that Mr Webster failed in his duty as an expert in any respect. No challenge was made to Mrs Cunnison's evidence of matching fibres and it is not suggested that defence counsel were at fault. We must assume that Mrs Cunnison had good reason to express herself as she did, but we accept that in so expressing herself she gave the jury an incomplete picture of the variety of fibres to be seen in each population.

The ultimate test for the receipt of evidence is, under section 23(1) Criminal Appeal Act 1968, whether it is "necessary or expedient in the interests of justice" to receive it. The Court of Appeal will only admit evidence which could have been adduced at trial in exceptional circumstances (see, for example, Solomon [2007] EWCA Crim 2633). We have no reason to doubt that the appellant was advised, at the time of his trial, of the apparent strength of the scientific evidence relating to fibres and that his defence was conducted in the light of that advice. Unknown to those representing the appellant at trial the scientific examination of fibres was, to put it neutrally, incomplete. For these reasons we consider that it is in the interests of justice that the evidence of the three expert witnesses should be admitted for its full effect notwithstanding that the conventional examination of fibres could have been but was not challenged at trial.

In Pendleton [2002] 1 WLR 72, the House of Lords, confirming its previous decision in Stafford v DPP [1978] AC 878, repeated that the court must ask itself whether the conviction is unsafe, and it does not have to consider what effect the fresh evidence would have had on the jury. However, Lord Bingham said (at [19]):

"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 … 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 which the jury heard. For these reasons it will usually be wise for the Court of Appeal, in a case of any 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." The test was again considered by the Privy Council in Dial v State of Trinidad and Tobago [2005] UKPC 4, [2005] 1 WLR 1660. Lord Brown of Eaton-under Heywood, expressing the view of the majority which included Lord Bingham, said:

"31. In the Board's view the law is now clearly established and can be simply stated as follows. Where fresh evidence is adduced on a criminal appeal it is for the Court of Appeal, assuming always that it accepts it, to evaluate its importance in the context of the remainder of the evidence in the case. If the Court concludes that the fresh evidence raises no reasonable doubt as to the guilt of the accused it will dismiss the appeal. The primary question is for the Court itself and is not what effect the fresh evidence would have had on the mind of the jury. That said, if the Court regards the case as a difficult one, it may find it helpful to test its view "by asking whether the evidence, if given at the trial, might reasonably have affected the decision of the trial jury to convict" (Pendleton at p83, para 19). The guiding principle nevertheless remains that stated by Viscount Dilhorne in Stafford (at p906) and affirmed by the House in Pendleton: "While ... the Court of Appeal and this House may find it a convenient approach to consider what a jury might have done if they had heard the fresh evidence, the ultimate responsibility rests with them and them alone for deciding the question [whether or not the verdict is unsafe]." 32. That is the principle correctly and consistently applied nowadays by the criminal division of the Court of Appeal in England — see, for example, R v Hakala [2002] EWCA Crim 730, R v Hanratty [2002] EWCA Crim 1141, and R v Ishtiaq Ahmed [2002] EWCA Crim 2781. It was neatly expressed by Judge LJ in Hakala, at para 11, thus: "However the safety of the appellant's conviction is examined, the essential question, and ultimately the only question for this Court, is whether, in the light of the fresh evidence, the convictions are unsafe." In Hakala, Judge LJ (as he then was) made the following statement which we have found of assistance:

"11...The judgment in "fresh evidence" cases will inevitably therefore continue to focus on the facts before the trial jury, in order to ensure that the right question - the safety, or otherwise, of the conviction - is answered. It is integral to [that] process that if the fresh evidence is disputed, this Court must decide whether and to what extent it should be accepted or rejected, and if it is to be accepted, to evaluate its importance, or otherwise, relative to the remaining material which was before the trial jury: hence the jury impact test. Indeed, although the question did not arise in Pendleton, the fresh evidence produced by the appellant, or indeed the Crown, may serve to confirm rather than undermine the safety of the conviction. Unless this evaluation is carried out, it is difficult to see how this Court can carry out its statutory responsibility in a fresh evidence case, and exercise its "powers of review to guard against the possibility of injustice". However the safety of the appellant's conviction is examined, the essential question, and ultimately the only question for this Court, is whether, in the light of the fresh evidence, the convictions are unsafe" We propose, therefore, to consider the effect of the evidence adduced in this appeal.

We accept Mr Coyle's evidence that he managed to distinguish fibres of different thicknesses in both populations. He has demonstrated the differences in Figure 10 of his witness statement of 15 September 2009. We are not, however, persuaded of his ability accurately and consistently to distinguish between bright, semi-dull and dull fluorescence. During cross examination Mr Coyle confirmed that had he been comparing the two conjoined fibres found by Mr Palmer, separately under a comparison microscope, he would have excluded them from association when it was known that they were in fact associated. "Yes", said Mr Coyle, "but I chose fibres which could be compared". So much confidence did Mr Coyle have in his ability to categorise by subtle or weak differences in fluorescence that he would not embrace the possibility that he had been excluding, on a fibre by fibre comparison, fibres which clearly belonged within congruent ranges. We have no doubt that the tendency of Mr Coyle's comparison was to exclude on inadequate grounds. We found compelling Mr Palmer's demonstration, by "typing" of MSP spectra, of the range of intra-fibre variation within both populations of fibres and the "congruence" between the two populations. It is clear to us that the ability of a single fibre in both populations to produce several different MSP spectra along its length renders extremely doubtful an exclusion based upon visually perceived subtle or weak differences in fluorescence. We accept as inevitable Mr Palmer's conclusion, supported by Mr Robson, that when there is no suspect garment from which to take a reference sample, but only two populations containing thousands of fibres, the principal task of the examiner is to define the range of variation within each population.

It seemed to us that when examining fibres under the microscope Mr Coyle must have been reposing absolute confidence in his ability consistently to match subtle variations in fluorescence within populations of 85 fibres from Boydlands and 30 fibres from Snowcroft respectively and then to sustain the accuracy of his observation when making over 200 comparisons between fibres in each population. For the reasons given by Mr Robson we find this confidence misplaced.

For similar reasons we do not accept that Mr Coyle's MSP spectra analysis is reliable. While we accept that he found differences in the absorbance values of Boydlands and Snowcroft fibres we find that Mr Palmer is undoubtedly correct when he asserts that we were seeing incongruence between at least two types of response in different fibres and not incongruence between two populations of fibres. It follows that we do not accept the scientific premise on which Mr Coyle's analysis was founded. In the face of Mr Robson's understanding of the implications of Mr Palmer's work, we found Mr Coyle's inability or unwillingness to consider this possibility puzzling.

We do not accept Mr Coyle's evidence that there was a material difference between the carbon black polyester fibres found at Boydlands (wrongly identified as green at trial) and those associated with the appellant. We received no pictorial demonstration to illustrate Mr Coyle's opinion and none is provided by way of appendix to any of his witness statements. Mr Robson commented upon the inconsistent conclusions of Mr Coyle and Mr Palmer upon the match between Snowcroft and Boydlands polyester fibres in his statement of 13 May 2010, page 5, as follows:

"Again, I feel it is necessary to record that I find it particularly unusual to observe two credible forensic examiners disagreeing on the outcome of a microscopic comparison in that, in basic terms, the FSS say the black polyester fibres match, whilst Contact Traces say they don't. This appears to mirror, in part, the outcome of the comparison of nylon flock fibres which play a key role in this case. This discrepancy is an area of particular concern. I can only suggest, on reflection, of the science involved, this has occurred as a result of differing operational procedures and/or reporting guidelines." This would also be a matter for concern for the court were it not for the fact that Mr Robson made his own examination of the polyester fibres and concluded that his findings matched those of Mr Palmer and not those of Mr Coyle. Indeed, he concluded that Mr Palmer had been somewhat conservative in identifying the number of matches.

During cross examination Mr Spence asked Mr Coyle whether it was his view that the presence of black nylon flock fibres and black polyester fibres within both populations was a matter of significance. Mr Coyle was not prepared to concede that it might be. He said, "Not really. You would expect to find polyester fibres, but they are not necessarily from the same source which deposited the flock fibres on the body". While it is common ground that the polyester fibres are alone of little evidential value their presence in both populations of fibres was quite plainly, as Mr Robson told us, a matter of significance. In his statement of 10 August 2009, page 9, Mr Coyle expressed the following opinion:

"As both the flock fibres and the carbon black polyester fibres are present on the body and at the points of entry, the most likely explanation for the presence of these fibres is that they were transferred as a result of contact between the offender and the scene." In view of this perfectly sensible expression of opinion, we found Mr Coyle's reluctance in evidence to make the obvious concession, even conditionally, surprising and inappropriate.

Mr Spence asked Mr Coyle questions about a possible explanation for the fact that glued fibres were present at Boydlands but not at Snowcroft, and for the fact that the proportion of the thicker fibres in the Boydlands population was greater than at Snowcroft. He asked whether body contact with objects such as a post or a fence and a struggle with the victim might explain the population at Boydlands, while simply hanging the same garment in a wardrobe might explain the population within the wardrobe. Mr Coyle replied, "We have no idea...Normal contact could have caused the shedding". In our view it is perfectly obvious that contact such as that proposed by Mr Spence might cause the garment to show shedding characteristics which were different from those likely to take place if it was hung in a wardrobe; we were surprised that Mr Coyle was not prepared to engage with the question.

In his responses to these questions we gained the impression that Mr Coyle was demonstrating the same inflexibility towards reasonable scientific alternatives as that which he had demonstrated towards methodology and interpretation of spectra. When Mr Coyle was asked what his reaction had been to the advice and opinion expressed in the Science and Law article as to the application of the first derivative, he replied, "I disagree with it absolutely". Again we were perplexed that Mr Coyle should have been so dismissive of carefully researched and peer-reviewed work, and should have expressed himself so sweepingly in his denunciation of it. This reaction differed markedly from the measured way in which Mr Robson informed us that while he too was a proponent of the first derivative he did recognise its limitations in certain circumstances.

Finally, we were not impressed with Mr Coyle's responses to Mr Mansfield's questions as to the random frequency of nylon flock fibres. When asked whether nylon flock fibres turned up on large numbers of garments, Mr Coyle replied, "A significant number. They were present in 80% of the garments examined". Mr Coyle was, of course, basing his answer on his colleague, Ms Jones', target research with which he was very familiar. While it was true that of the 82 garments on which Ms Jones found flock fibres, 84% had nylon flock fibres on them, Mr Coyle must have known that, left unqualified, his answer was utterly misleading. First, black nylon flock fibres were found on 13 garments only and, secondly, not one of the garments taped had upon it even remotely the number of fibres which Mr Coyle had been examining in his laboratory for the purpose of the present appeal. We are surprised that Mr Coyle lent himself to the implied suggestion that what was seen in the present case was in no sense a rarity. It is quite plain to us, as it was to Mr Palmer and Mr Robson, that Mrs Cunnison's evidence, namely, that the chance of finding these numbers of black nylon flock fibres in hypothetically unrelated populations was very small indeed, was fully justified.

We reject Mr Coyle's evidence of distinguishable populations of fibres at Boydlands and Snowcroft. In reaching this conclusion we found the 2002 and 2010 TLC results of no assistance for the reasons given by Mr Coyle. We have no means of assessing the accuracy of Mr Coyle's analysis of the Wonder SRL fibres and, in the light of our other conclusions, cannot find that in this respect his evidence has any material bearing on the safety of the jury's verdict. Furthermore, we assess the effect at trial of the Wonder SRL evidence as peripheral at best. It gave rise to no more than a speculative possibility (the prosecution said opportunity) as to the source of the fibres. The judge gave the jury a warning about the danger of speculation and reminded them, first, that the garment which shed the fibres was unknown, second, that there was no evidence of a relevant purchase and, third, that there was evidence of other suppliers in the market.

Upon the evidence presented to this court we conclude that Mr Coyle's evidence does not give rise to any ground for allowing the appeal. While we have concluded that the fibre evidence given at trial was incomplete in its description and analysis of the available source material, and in its identification of green polyester fibres, wrong, we are quite satisfied that the scientific support for the assertion that the appellant was the source of the fibres found at the crime scene is compelling.

We have no reason to doubt the safety of the jury's verdict and the appeal is dismissed.