196. These three post mortems formed the basis of the application to the Full Court for leave to adduce further evidence in the current appeal on the ground that the fresh material "... has a serious impact upon one aspect of Dr Williams' evidence and moreover will inevitably have affected the jury's assessment of Dr Williams' reliability".
197. The three post mortems have been identified as Baby Callum, Baby M and Female X. The post mortem reports and supporting material have been examined, and have been subject to evidence by, Professor Levin and Doctor Parsons, who did not give evidence at the trial, and Professor Luthert and Dr Keeling, who did give evidence at the trial. The area of medical expertise of these four witnesses is:

• Dr Levin is Associate Professor of Ophthalmology at the University of Toronto, Canada, and Fellow of the American Board of Paediatrics. He is one of 10 in the world with a dual expertise in ophthalmology and paediatrics.

• Dr Parsons is Honorary Consultant in Ophthalmic Pathology at the Royal Hallamshire Hospital Sheffield, Senior Lecturer in Ophthalmic Pathology at Sheffield University and he has a special interest in child abuse. He is one of six opthalmic pathologists in the UK.

• Professor Luthert is Professor of Pathology at London University and Honorary Consultant Neuropathologist to Moorfields Eye Hospital.

• Dr Keeling is Consultant Paediatric Pathologist at the Royal Hospital for Sick Children, Edinburgh.

198. Baby Callum was discovered dead in a dustbin liner. He has never been identified and is the subject of continuing police investigation. He died at birth in March 1998. Dr Williams identified in the post mortem report: 1) blood-stained fluid in the left eye; 2) large blotchy haemorrhage in the lateral part of the white of the globe of the left eye; 3) blotchy haemorrhages in the white of the globe of the right eye; 4) no petechial haemorrhages in the eyelids. The infant showed acute asphyxia due to impaction of foreign material in the larynx and mouth, bruising of the neck consistent with attempts at strangulation and numerous bruises not consistent with normal labour. Cause of death was given as asphyxia but the mechanism was not identified.
199. Baby M died at 7 weeks of age in March 1998 but was not the subject of any police investigation; in consequence no photographs were taken. There were no petechial haemorrhages in the anterior parts of the eyes or eyelids. Dr Williams found: 1) petechial haemorrhages on the periosteal surface of the scalp, 2) removal of the eyes showed petechial haemorrhages on the globe of the left eye. Cause of death was given as asphyxia without physical evidence of a cause, and there were macroscopic and microscopic changes associated with asphyxia. The conclusion was reached by Dr Williams that the infant died from an asphyxial mechanism which may have been due to having been in bed with his parents.
200. Female X did not die until after the trial and therefore did not feature as one of the comparable cases referred to in Dr Williams's evidence. The deceased died of strangulation on February 5th 2000. Dr Williams found: 1) large blotchy haemorrhages into the corners and lids of both eyes with petechial haemorrhages into the eyelids and white sclera; 2) dissection of the eyes showed large intra orbital haematoma with haemorrhage into the scleral surface of both eyes. The cause of death was given as strangulation by ligature.
201. The further evidence called before this court is relevant to three issues which this court has to consider:
a) whether the haemorrhages observed at the back of Harry's eyes were peri-mortem, or were artefacts arising out of the autopsy and therefore of no evidential value;
b) whether the cases of Baby Callum and Baby M, upon which Dr Williams placed reliance in evidence, could lend support to the proposition that such haemorrhages, if real and not artefactual, may be associated with death by asphyxiation;
c) whether such evidence, had it been before the jury, might have affected their verdict so as to render that verdict unsafe.
202. We propose to examine each issue in turn.

Artefact or real haemorrhage in Harry's eyes
203. Dr Levin opined that the haemorrhages may be artefactual, because he had never come across them before either in practice or in literature, and if an autopsy is carelessly performed it can involve a prick of the blood vessel which would produce a haemorrhage in the form of leakage of blood into the episclera or tenon's layer. In giving this opinion Dr Levin assumed inaccurately that a knife had been used instead of forceps and that the eyes had been shredded for dissection instead of teased out in layers. He rejected the suggestion that the haemorrhages found by Dr Williams could have been caused by blood dripping, because the blood was below the membrane and could not have soaked through to settle between the layers.
204. Professor Luthert repeated his evidence at trial, that he considered the haemorrhages were artefactual, but he changed his position in respect of causation in that he gave his opinion before this court that the vortex vein might have been nicked with consequential oozing of blood which tracked under the semi-opaque membrane known as tenon's capsule, thereby tracking between the layers. Before this court Professor Luthert considered that causation more likely than blood dripping onto the eye.
205. Dr Parsons was 99% sure that the haemorrhages were real and not artefacts. He identified from an enlarged photograph of Harry's eyes that a forceps and not a knife had been used, and he demonstrated from that photograph the reason for his opinion that the bleeding was within the tissue plane in respect of the larger haemorrhage of 7mm. He found it harder to identify the origin of the smaller haemorrhage, but he was still of the opinion that it was a true haemorrhage. Dr Parsons gave reasons for his opinion:
a) Dr Spillman was present at the autopsy and gave evidence that he "had seen the area of haemorrhage on the upper surface of the right eye. He saw it immediately the eye was exposed". This evidence was before the jury. The significance according to Dr Parsons was that it takes time for a dead body to ooze blood if a vessel is nicked.
b) Dr Williams would have known if he had cut a vein, but he photographed what he considered to be a true haemorrhage.
c) The enlarged photograph, which was not available at trial and which was produced for Dr Parsons' own examination, enabled him to demonstrate a clear difference between the colour, edge, and wetness of the small collections of wet blood and the haematoma in the eyes. The eye haemorrhages were covered by a thin semi translucent white tissue membrane such as tenon's capsule.
d) On the macroscopic photographs of the eyes no puncture marks or cuts appear in the membranes over the eye haemorrhages.
e) The position of the eye haemorrhages in the macroscopic photographs is on the uppermost surface of the eye. If blood had dripped onto this point it would run off, by reason of gradient, and accumulate in the lowest part of the cranial cavities.
f) If blood had been on the eye surface and dried, it would have a darker appearance, such as is the case in respect of dried blood collections seen in photographs 18 and 19.
g) The irregular outline of the eye haemorrhages would be extremely difficult to create by dripping blood onto the eyes.
h) The site of the haemorrhages, particularly in the right eye, is in the area of the vortex vein and the ciliary veins which supply blood.
i) Professor Luthert's interpretation of the microscopic findings of the right and left eyes, which had undisputed episcleral haemorrhages, did not take account of the technique used by Dr Williams which involved removal of the membranes, whereas an ophthalmic pathologist would not have removed tenon's capsule. By reason of that procedure by Dr Williams, the episcleral haemorrhages appeared inaccurately to be the same outer eye surface as on the macroscopic photographs.
j) The India ink test referred to by Professor Luthert was flawed, because the tenon's capsule, which is thicker in a child than an adult, would form a barrier, and the ink has a different carbon size to blood.
k) The reference by Professor Luthert to his experience of patches of blood, as opposed to haemorrhages, on the surfaces of eyes removed after death for donation was not comparable to Harry's condition because donor eyes are removed from the front, cutting through membranes which release blood.
l) Post mortem hypostasis cannot be a factor in the production of petechial haemorrhages on the eyelid as the face was pointing upwards after death and hypostasis was posterior.
206. Dr Keeling deferred to the ophthalmic experts on this topic, but she did give her opinion that she thought it unlikely that dripping blood could account for the appearance of the eyes because she would have expected blood to run around the surface of the eye which is convex and that it would have washed off.
207. Dr Levin, Professor Luthert and Dr Parsons all agreed that, although they had never seen such haemorrhages, nor read about them, and therefore Harry was unique, nevertheless the haemorrhages, if real, were in conjunction with the other haemorrhages consistent with smothering. Dr Levin and Professor Luthert expressed their view in the words that they could not exclude a connection with asphyxia, whereas Dr Parsons concluded that Harry suffered an unnatural inflicted asphyxial mode of death by reason of the different areas of haemorrhage and the input of Professor Green and Dr Keeling.
208. Dr Levin is not a pathologist, and he was subject to criticism by the other experts for suggesting that a pathologist should move the eyes in the socket before removal and dissection.
209. If this evidence had been before the jury at the trial, they would have had to consider its significance in the light of the following matters: 1) Dr Levin does not have experience as a pathologist; 2) Professor Luthert had shifted ground in supporting his reason for the opinion that the haemorrhages were artefactual; 3) credible evidence from Dr Parsons, accompanied by careful investigation and reasoning to support the proposition that the haemorrhages were real.
210. In contrast, at trial the jury had evidence only from Dr Williams that the haemorrhages were real, and he is not an expert on eyes.
211. In the opinion of this court, the additional fresh evidence would have been likely to strengthen the contention that the haemorrhages were real.

Comparability of the three cases
212. A general point arises in respect of terminology. Dr Williams is not an ophthalmic pathologist and it is clear from the evidence that he uses the term "back of the eyes" to denote such area as cannot be seen when looking at eyes from the front. He regards anything which can be seen as anterior, and anything which cannot be seen as posterior.
213. Harry's haemorrhages were undoubtedly in the posterior region as defined by the ophthalmic experts, i.e. behind the anatomical equator or posterior pole as measured from the optic nerve.
214. The first question which arises in respect of the three cases is whether the haemorrhages relied on for comparison were in the same region of the eye.
215. Baby Callum. Dr Levin and Professor Luthert described the haemorrhages in baby Callum as anterior. However, each agreed in evidence that they were in the same plane as Harry's haemorrhages and that was an important factor. Dr Parsons identified the most posterior haemorrhage in Callum at 9 o'clock in the left eye over an episcleral vessel with intense congestion of the choroidal blood vessels. Professor Luthert identified one haemorrhage in the left eye 7mm from the limbus, but did not note the more posterior haemorrhage 8mm from the limbus, identified by Dr Parsons. Both findings indicate raised intravenous pressure and venous congestion and Professor Luthert, Dr Keeling and Dr Parsons agreed that the findings were consistent with death by asphyxiation, or following child birth by reason of compression of the chest. Dr Levin did not agree because the information in the autopsy report was imprecise, the majority of the haemorrhages were in the sub conjunctival space which were absent in Harry's case, and he considered the episcleral haemorrhages could be artefact. Further he considered the mechanism of asphyxiation to be different from Harry's case although an asphyxial manner of death was not challenged.
216. Baby M. Dr Levin was not able to draw any conclusion by reason of the imprecise information as to the existence and location of haemorrhages and the absence of photographs. What is clear from the evidence is that M was observed to have petechial haemorrhages at the back of the left eye, which were only visible when the globes were removed. However they did not show up on the slides of the sections taken. There was unchallenged evidence of an asphyxial death.
217. Baby M had petechial haemorrhages whereas Harry did not. Professor Luthert described M's haemorrhages as discrete rather than a confluence of petechiae. There was in essence no disagreement between the experts in respect of M. The evidence is imprecise. Dr Parsons stated in evidence that the case merely demonstrated that you can have haemorrhages on the back of the eye as defined by Dr Williams in association with an asphyxial mechanism. Dr Keeling agreed.
218. Female X was not a comparable used by Dr Williams because she died at a later date. It was agreed by all the experts that infants respond differently to insults than do adults, and a baby is not a small adult in that e.g. the tenon's capsule over the sclera is thicker in infancy. Some of the haemorrhages in Female X were subconjunctival, which was not a feature of Harry's case.
219. Dr Levin did not consider this a suitable case in principle as a comparison. There was disagreement in evidence between Dr Parsons and Professor Luthert in the interpretation of certain photographs. Professor Luthert could not identify in photograph 30 whether there was shadow or possibly orbital haemorrhage and he was not assisted by photographs 31, 32 and 33. However Dr Parsons in evidence identified a small round red point on photograph 32 which was resolved in photograph 33 as an isolated round haemorrhage. He concluded that there were bilateral orbital haemorrhages and that the photographs demonstrated an episcleral haemorrhage in the posterior half of the left globe in a posterior position similar to that of the haemorrhage in Harry's left eye. Both Dr Parsons and Dr Keeling considered that there was a general comparability with Harry's case, it being common ground that Female X had been strangled.
220. If the evidence of comparability of Baby Callum and Baby M with Harry had been before the jury at trial then the jury would have had to consider the significance in the light of evidence of : 1) the agreement between the medical experts that when strangulation or suffocation occurs there is an increase in the venous pressure in and around the eye, and an increase in the pressure can cause rupture of vessels; 2) each case having an asphyxial mode of death with evidence of episcleral haemorrhages in the part of the sclera which cannot be seen on examination from the front; 3) the position of the haemorrhages by evidence in Baby Callum's case and description in Baby M's case accords with Dr Williams own definition of "back of the eyes"; 4) the evidence of Dr Keeling and Dr Parsons that the cases support Dr Williams's statement in court that he has seen haemorrhages in the back of the eyes in cases involving an asphyxial death.
221. In contrast, the jury had no evidence at trial to support Dr Williams' contention.
222. Further, the case of Female X, although not strictly relevant, upon the evidence of Dr Parsons could demonstrate that Dr Williams had noted an association of posterior haemorrhages of the eye with an asphyxial death.