England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM TECHNOLOGY AND CONSTRUCTION COURT
HIS HONOUR JUDGE THORNTON Q.C.
[2003] EWHC 1142 (TCC)

		Royal Courts of Justice Strand, London, WC2A 2LL
		28th May 2004

Lord Justice Tuckey:

1. On 15 May 2001 a warehouse on the Deeside Industrial Estate was seriously damaged by fire. The claimant, Re-source America International Ltd., was a licensee of part of the warehouse in which it stored and refurbished combustible materials for use in the semi-conductor business. Re-source claimed over £1m. for damage to its property and consequential losses caused by the fire against Platt Site Services Ltd. Platt was employed as a sub-sub-contractor to carry out the steelwork involved in the construction of an addition to the warehouse and it was Platt's hotwork (grinding, welding and cutting) which started the fire. Platt joined the main contractors, Barkin Construction Ltd. as Part 20 defendants. On a trial of liability in the Technology and Construction Court His Honour Judge Thornton Q.C. found that Platt was liable to Re-source but entitled to be fully indemnified by Barkin. He found that Platt had been negligent but also that it was strictly liable for the fire under the rule in Rylands v Fletcher and the doctrine of Ignis Suus and that it was vicariously liable for Barkin's negligence. He found Barkin contractually liable to indemnify Platt under the terms of a letter of 3 May 2001, alternatively that it should make 100% contribution to Re-source's claim as a joint tortfeasor.
2. Barkin sought permission to appeal a host of the judge's findings of fact. Thomas L.J. granted permission to appeal one such finding which I will identify later. Buxton L.J. granted Barkin permission to appeal a number of issues of law, some of which, as I shall explain, are not in issue. The ones that matter are that the judge found Platt liable in negligence to Re-source and Barkin liable to Platt on grounds which were not pleaded and that he was wrong to find that Barkin was liable to Platt under the letter of indemnity or to make 100% contribution.
3. The trial took three days. The judge delivered a long judgment after six months. The judge's unexplained delay in giving the judgment is regrettable. No complaint is made of the delay as such, but it is said to explain why the judge made a number of unnecessary and erroneous findings. The hearing before us has focused rather more on the conduct of the trial than could be anticipated from the notice of appeal and I shall have to refer to this to deal with points made by Mr Livesey Q.C. for Barkin, who did not appear below.
4. The judge's judgment is reported at [2003] EWHC 1142 (TCC) to which anyone wishing to see a full account of the facts may refer. I only propose to refer to those necessary for the purpose of deciding this appeal.
5. The contract for the construction of the warehouse addition was made between the building owner and Barkin on one of the JCT forms. A similar steel-framed building was to be constructed alongside the existing building. The new steelwork structure was to be tied into the existing structure by horizontal stanchions bolted onto the existing vertical steel stanchions at eaves level. The way in which this was to be done required the existing horizontal channel running between the stanchions to be re-fixed to the top of each stanchion by means of angle irons welded either side of the stanchion to enable the old fixing and part of the channel to be cut away by flame cutting from the external face of the stanchion. Barkin subcontracted the fabrication and erection of the steel work to Henry Smith (Construction Engineers) Ltd. (HSW). HSW in turn sub-contracted this work to Platt by their order of 17 April 2001.
6. Health & Safety Regulations, guidance notes and codes of practice designed to create a structured method of identifying and avoiding risk applied to this project. The employer is required to appoint a planning supervisor and a principal contractor. For this project consulting engineers were appointed as the planning supervisor and Barkin was appointed as the principal contractor. The principal contractor is responsible for ensuring that risk assessments are made and methods statements prepared for all hazardous work before any such work is carried out. The judge said:

In summary, there was a contractual requirement imposed on Barkin to ensure that it prepared a risk analysis which analysed all risks associated … and with the cutting, grinding, welding and drilling work at high level in confined conditions with inflammable materials stored at high level in close proximity to this work and with the adjacent occupied area continuing to be used and worked in by those undertaking Re-source's on-going work. There was also a contractual requirement imposed on Barkin to ensure that appropriate method statements were prepared by either Barkin or Henry Smith covering the safe … cutting and welding work and which made provision, in conformity with HSE requirements, for the means of access to the hotwork in question and for all site precautions and measures needed to remove the risks that had been analysed in the relevant risk assessments. These method statements were not only to be prepared but had to be discussed with, and copies provided to, each contractor associated with this work, the Planning Supervisor and the site-based Health and Safety File. The necessary method statements would have to be available and included within the Health and Safety File at least two weeks before any hotwork was to start on site. They would need to take into account the advice and recommended precautions contained in the relevant HSE publications.

7. In the event these procedures were only followed in the most general terms. HSW prepared a risk assessment and method statement dated 14 March 2001 at which time they understood that the work to the stanchions in the part of the warehouse occupied by Re-source would be carried out after all the stock had been removed from it. The work would be done in the same way as that to the stanchions in the unoccupied part of the building from scissor lifts (cherry pickers) inside and outside (via holes to be cut in the cladding) of the building. However, by 23 April, when Platt arrived on the site, Barkin and HSW had agreed that Re-source's entire stock would not be removed: a 30 foot strip along the side of the building would be cleared of the stacks of plastic spools, drums and pallets stored there. This would reduce the fire risk and give access to the cherry picker. The adjoining stacks would be "barried" off and covered with fire protective materials.
8. Unfortunately no-one had previously discussed this plan with Re-source. When its manager, Mr Whitby, was informed of the plan on 27 April he said that he would not allow welders into his part of the warehouse because he had not seen the necessary risk assessments, method statements and safety equipment. Its stock could not be moved quickly and he would not be able to discuss the matter further until after he had returned from holiday on 10 May.
9. Barkin's managing director, Mr Andrews, was very put out by Mr Whitby's response which, as the judge found, had been perfectly reasonable. After unsuccessfully attempting to enlist the support of the building owner, he decided to carry out the work immediately with a view to its being completed before Mr Whitby returned from holiday. To this end he called a meeting on 2 May attended by Platt's principal, Mr Platt, and HSW. At this meeting Mr Andrews proposed that the hotwork to the existing stanchions should be carried out from the outside via larger holes in the cladding through which the bucket of the cherry picker could enter the building. The working area would be enclosed in a cocoon or shroud wrapped around the stanchion and fixed to the sides of the hole comprising an outer tarpaulin layer and an inner layer made of two welding blankets laced together. He personally would install the shroud at each stanchion with the assistance as necessary of the cladding sub-contractor. Platt's case, which the judge accepted, was that Mr Andrews said he would also provide a fire watch within the building whilst hotwork was being carried out.
10. Mr Platt, who was a very experienced steelworker, was very concerned with the safety of Mr Andrews' proposals. He said he would not be prepared to allow Platt to undertake the hotwork unless Barkin provided Platt with a written "damage waiver". Mr Andrews, who was desperate to get the work started, agreed. On the following day he wrote to Platt saying:

Please accept this letter as indemnity to yourselves against the site works as discussed, associated with the occupied part of the building.

You are to ensure fire blankets are used for protection of internal stock, together with a secondary tarpaulin. We do, however, appreciate this will be a difficult job and any damage to the tenants stock will be covered by our insurance.

This is the indemnity upon which Platt succeeded.

11. The hotwork to the stanchions in the unoccupied part of the warehouse and the first 12 of the 13 stanchions in Re-source's part of the warehouse proceeded uneventfully. The work to the 13^th and last stanchion in the corner of the building was due to be done on Friday 11 May. However, Mr Atherton, the Platt steel worker who had carried out most of the work on the other stanchions discovered that he was missing the angle iron which he needed to carry out the work to this stanchion. He reported this to Mr Platt the following Monday and the missing part was received on site the following day, the 15^th. That afternoon Mr Atherton was engaged in doing the hotwork to the last stanchion when the fire started at about 3 p.m. because, as the judge found, molten and red hot slag burnt off by his flame cutter fell onto and ignited Re-source's combustible materials stacked below.
12. There was a considerable conflict of evidence about the events leading up to the fire. Mr Andrews' evidence was that he had not fitted the shroud for the last stanchion. A hole had been cut in the cladding which had been temporarily covered over. That was still the position on the morning of the Tuesday when Mr Platt asked him to remove the temporary cover. He said he would arrange for this to be done and then fit the shroud as soon as Platt had put a cherry picker in position. At about 2.30 p.m. he decided to leave the site, having forgotten that he had promised to fit the shroud. However he saw the cherry picker in position, that the plate had been removed and that welding operations had started. He could see a tarpaulin in place but no welding blankets. Mr Atherton was on the ground having tea and he told him to go and get welding blankets from where they were stored and put them in place over the tarpaulin. He then left the site in his car and had not seen the fire start.
13. Mr Platt's evidence was that on the Tuesday Mr Andrews asked him when the hotwork to stanchion 13 would be done. He told him it would be done that afternoon. Mr Andrews told him that the shroud was already in place and that Platt should remove the temporary cover over the hole in the cladding. After lunch Mr Atherton removed the cover. Both he and Mr Atherton said that the welding blankets were in place. However, because the stanchion was in the corner of the building and the hole in the cladding was larger than the ones made for the other stanchions Mr Platt instructed Mr Atherton to lay a third blanket over the other two already in place, which he did.
14. Based on Mr Andrews' evidence and the fact that no trace of welding blankets was found after the fire, it was Barkin's case that at the time of the fire no welding blankets had been in place. The judge accepted Mr Platt and Mr Atherton's evidence that welding blankets had been in place and it is this finding of fact which Barkin have permission to challenge.
15. As I have said the judge found that Mr Andrews had agreed to provide a fire watch in the warehouse whilst the hotwork was being carried out. There was no fire watch at the time of the fire, but the judge found that this failure was not causative: no one could have got through the impenetrable piles of stock to put out the fire just below eaves level even if fire fighting equipment was available, which it was not.
16. The judge was severely critical of Mr Andrews of whom he said:

He was both evasive and overbearing and he rapidly became agitated when being cross examined. These mannerisms occurred when confronted with the difficulties of his evidence and Barkin's case and appeared to be the result of a need to cover up and exculpate his apparently unsatisfactory conduct in relation to the steelwork modification work. In short, he was a thoroughly unreliable witness.

Of Mr Platt, who was aged about 64, he said:

He gave the impression of one who was honest, forthright, experienced and safety conscious.

Mr Atherton was about 20 at the time and had worked for Platt for about eight months before the fire. The judge said of him:

On leaving school he had moved straight into steel hotwork, having worked part time for 18 months whilst undertaking a college course in welding and allied skills., He was, therefore, despite his relative youth, already experienced and skilled in hotwork. He also appeared to be both honest and careful. He had no particular reason to be partisan in his evidence having moved on to other welding employment since the fire.

17. The judge gave a number of reasons for rejecting Mr Andrews' evidence about the events leading up to the fire, some of which are challenged by Barkin. One notable finding however that the judge did make was that Mr Andrews had not left the site before the fire but after and because it had started and had then altered his diary to give credence to this evidence. This was one of the findings of fact which Barkin challenged. Although permission to appeal this finding had already been refused twice, Mr Livesey again sought to challenge it. We refused to allow him to do so.
18. The basis upon which the judge found Platt negligent was Mr Atherton's unprompted evidence during cross-examination that the welding blankets had holes in them. He had been asked whether he could account for the fact that Re-source's deputy manager had seen sparks from inside the warehouse. He said that you could not trap all the sparks. The ones which could not be trapped

… would just fall. They burnt through the fire blanket because I remember seeing the fire blanket and it had big holes in it from the previous ones I had done… molten metal would burn through.

He had not told Mr Platt about this as he should have done.

19. This unexpected revelation has given rise to much of the argument on the appeal to which I will come later in this judgment, but I think it is convenient at this point to consider Barkin's appeal against the judge's finding that the welding blankets were in place when Platt came to do the work on stanchion 13 because if this finding cannot stand much of the rest of the appeal becomes academic.
20. Mr Livesey prefaced his attack on the judge's finding by submitting that a number of things went wrong at the trial. Firstly, the judge made findings of strict liability based on the rule in Rylands v Fletcher, the doctrine of ignis suus and vicarious liability by Platt for Barkin's negligence which were not pleaded and were expressly not espoused by any of the experienced counsel (two Q.C.s and a senior junior) appearing before him. Likewise no one sought to support these findings on appeal. Another point which the judge decided, but no one contended for, was that the litigation would have been unnecessary if Monde had effected insurance under clause 22.C. of the JCT form (as the contract required) because Re-source would have been covered by such insurance as its stock was part of the "contents" of "existing structures… for which the employer was responsible".
21. Thus far Mr Livesey's criticism is justified. A judge must restrain his enthusiasm for deciding points of law, particularly difficult ones, if they are not in issue or properly argued before him. I express no view about the judge's findings of strict liability. The insurance position in this case is entirely unknown. Whether it explains why Re-source chose to sue Platt alone must remain a mystery. In any event it seems to me that the judge's construction of clause 22.C. was wrong. As was submitted to the judge "responsible" appears to mean "responsible for the care of", i.e. as bailee, rather than "legally liable for".
22. But Mr Livesey went further. He took us through what he called the shape of the trial and suggested that the judge had hurried and harried the parties to complete the trial by, for example, persuading them that it was unnecessary to call the experts when Re-source, at least, had said it wanted to call its expert, Mr Bullen and to cross-examine Dr Bland, who was Platt's expert..
23. I do not accept this part of Mr Livesey's criticism of the conduct of the trial. There is no general complaint of unfairness in the notice of appeal and having considered the whole of the transcript I do not think that any such criticism is justified. The trial was estimated to last four days. It was completed in three long days. If any party had felt that it was under unfair time pressure it was represented by counsel who were well able to stand up and say so. The transcript shows that the judge behaved courteously to all concerned, and that such interventions as he made were entirely appropriate. I shall have to return however to the way in which he dealt with the pleading points which arose during final submissions later.
24. The specific attack on the judge's finding of fact is that in making it he ignored, misconstrued or undervalued the expert evidence. He relied too heavily on his general view of Mr Andrews' credibility. He latched on to Mr Atherton's evidence about holes in the blankets which was not credible. This enabled him to explain how the fire had taken place and relieve Platt of liability. He also made other mistakes of fact and ignored other circumstantial evidence.
25. After the fire the scene was examined by a number of experts. Among other things, they were looking for traces of the welding blankets. None were found, either in the vicinity of stanchion 13 or a low adjoining building, the flat roof of which had been used for storage of the blankets. The evidence was that there were four blankets. On Barkin's case all four would have been on the flat roof; on Platt's only one.
26. Platt's expert, Dr. Bland, said that having regard to the fact that this was a severe and prolonged fire in which the steel work of the building had collapsed and copper cable ends had melted he was "not surprised that identifiable remains of the blankets were not found after the fire". Re-source's expert, Mr Bullen (on whose evidence Barkin sought to rely), did not dispute Dr Bland's evidence that the fire generated temperatures in excess of 1000° C, but said that he would "not be surprised to find recognisable residual material" from the blankets. In his view this was more likely than not.
27. The judge rehearsed this expert evidence (paras. 161 and 162) and concluded that it "was not indicative of the absence of blankets in the shroud". In reaching this conclusion it is arguable that he may have mischaracterised the difference between Dr Bland and Mr Bullen. But the difference was simply one of emphasis. Mr Bullen was not saying that there were bound to be traces of blanket. The judge's conclusion was therefore justified. What is more, if traces of welding blanket were expected to be found, it was far more likely that they would have been found in the area of the flat roof and yet none were found there either. The judge also referred to evidence that traces of glass fibre had been found at the base of stanchion 13 and this led him to conclude that "the absence of any traces of welding blanket near stanchion 13 was not proved". This was a misunderstanding on the judge's part: the traces of glass fibre found were clearly from a different source. However, this misunderstanding does not seriously undermine the judge's conclusion.
28. I do not think the judge relied too heavily on his view of Mr Andrews' credibility in resolving this issue. In the light of his finding that Mr Andrews had lied about it when he left the site and then falsified his diary to support his evidence, it would have been open to the judge to reject his evidence in its entirety. He did not do so. After reviewing the factual evidence he reached a provisional conclusion that Mr Andrews had erected the shroud on Friday 11 May and then tested his conclusion against his findings about the events of 15 May, subsequent events and the expert evidence.
29. Mr Livesey's criticism of the judge's acceptance of Mr Atherton's evidence about holes centred on what he claimed was the judge's misunderstanding of the fire resistance of the blankets. The manufacturers gave a temperature resistance of 205° C for continuous and 350° C for short-term exposure. Mr Bullen's evidence however was that this temperature rating was appropriate for the silicon coating of the blanket, but its main component, woven glass fibre fabric, could withstand much higher temperatures. He had exposed a sample of this fabric to a blow torch flame of 1200° C. It did not melt, but on prolonged heating became brittle and frail. There was no evidence, Mr Livesey submits, about the temperature of the hot products of flame cutting, although the judge said this would be in excess of 1000° C. Furthermore the joint statement of the experts (Mr Butler and Dr Bland) was that a shroud provided by "a tarpaulin and non combustible fire blanket" was an acceptable method of working provided there were no gaps through which hot products could fall and an effective fire watch was maintained. Mr Livesey submits that the experts would not have said this if the blankets were liable to become holed.
30. I do not think Mr Livesey's submissions compel the conclusion that the judge should have rejected Mr Atherton's evidence about holes. In his report Mr Butler had said that the process of flame cutting produced "a very hot fluid molten iron oxide slag … The temperature of this slag would be lower than welding splatter but its volume and heat content much greater". Although Mr Andrews had apparently purchased new blankets to enable this work to be carried out, they had been used a number of times before 15 May and the flame cutting involved was self evidently substantial: the whole process took about 1½ hours for each stanchion. With the blankets slung between the stanchion and the hole in the cladding any hot products would be likely to collect at the lowest point in the blankets. I think it is fair to say that the joint experts' statement only just accepts the method of working. It certainly does not rule out the possibility of holes being burnt in the blankets.
31. I can deal with the other points relied on by Mr Livesey shortly. The judge mistakenly said that on the Friday Mr Atherton had discovered that he was missing an angle iron when he finished work on stanchion 12 when in fact he said this was after he finished stanchion 11. The point about this was that the earlier on in the day this discovery was made, the less likely it was that Mr Andrews would have fitted the shroud for stanchion 13 that day. Mr Andrews denied that he had fitted any shrouds that day, but as Mr Atherton completed stanchions 10, 11 and 12 that day and there were only four blankets Mr Andrews must at least have fitted the shroud for stanchion 12 that day and there was no reason why he should not have fitted the shroud for stanchion 13 after Mr Atherton had finished stanchion 11. Next, Mr Livesey said that the shroud for stanchion 13 was fitted differently from the way Mr Andrews had fitted the shrouds for the other stanchions. But this can be explained, as the judge did, by the fact that stanchion 13 was different because it was at the corner of the building and a larger hole had been cut in the cladding to gain access to it. Further miscellaneous points are made about the evidence of other witnesses (none of whom gave direct evidence about the issue) with which it is not necessary to deal in detail, because I do not think they add anything to the attack on the judge's finding.
32. So, for the reasons which I have given, I think Barkin's challenge to the judge's finding that there were no welding blankets fails. I have also explained why I think it was open to the judge to find that the fire was caused by the hot products of flame cutting falling through holes burnt in the shroud.
33. The question which then arises is whether Platt could or should have been able to make their part 20 claim against Barkin on the basis that the welding blankets fitted by Mr Andrews for stanchion 13 had holes in them. A further pleading point arises, which it is convenient to consider at the same time, as to whether Platt had made their claim for contribution on any wider basis. This question arises because the judge's finding of negligence against Barkin was as follows:

192. Barkin required the hotwork to start without any appropriate risk assessment or method statements and without causing the area below the proposed hotwork to be cleared. No fire watch was arranged and no clearance around or access to the potential seats of fire resulting from that work was arranged and no suitable and working fire fighting equipment was arranged. Inadequate fire protection was arranged in that the welding blanket was not sufficiently incombustible for the temperatures in excess of 1000° C that would be created when cutting was in progress and no method was devised that would ensure that any red hot slag falling into the lap of the protective measures would be removed or extinguished and not allowed to remain attacking the welding blanket. The welding blankets were used even after holes were burnt into them. No thought was given to the safety of Re-Source's employees or warning system was devised and no barriering off was arranged. No permit to work system or similar fail safe system to ensure that the fire watch was in place when required was devised. Finally, Mr Andrews, without any qualification or training in fire protection, took it upon himself to supervise the hotwork, devise the fire protection measures and supervise their execution. On the critical day, rather than supervising these measures, he left the whole work to Platt and, on discovering that this system of work had led to a fire, rapidly left site in an attempt to cover up his obvious failings with regard to the hotwork.

34. Barkin say that the only case which Platt made against it was that the shroud had gaps in it and that it had failed to provide a fire watch. Gaps were not the same as holes and failure to provide a fire watch was not causative so the claim for contribution should have failed. For good measure Barkin also say that Re-source's claim against Platt should have failed because the only allegation against it was that it had done the work without any welding blankets at all.
35. We have been taken in some detail through the pleadings, the opening and closing written submissions of the parties and the transcript of the oral closing submissions. I shall attempt to summarise the position as far as possible so that this judgment is not over-burdened by extensive citation.
36. Re-source undoubtedly pleaded its case against Platt in wide enough terms to encompass an allegation of holes. But its leading counsel in opening the case made it plain that Re-source's case was limited to the allegation of no blankets at all. No one of course anticipated what Mr Atherton would say about holes. Platt had called him as a witness, but it no longer employed him and he said he had not said anything before about holes because no-one had asked him. In his closing submissions counsel for Re-source made it clear that he now did rely on what Mr Atherton had said and asked the judge to allow him to do so in spite of what he had said in opening. There was some opposition to this from Mr Moxon-Browne Q.C. for Platt, at least to the extent that Re-source should only be allowed to rely on holes against Platt if Platt could also do so against Barkin.
37. Platt's pleading, if nothing else, was clearly intended to pass on any finding against it to Barkin. It said nothing specifically about holes as one would expect, but it did allege gaps because that was one of the allegations made by Re-source in its pleading. In his closing submissions Mr Cannon, for Barkin, objected to a claim based on holes being passed on to his client. It would be prejudiced by such a change of case at that late stage. None of the witnesses, who might have been able to shed some light on the matter, including Mr Andrews, had been asked about this. Mr Atherton had given his evidence on the second day of the trial. No one had said at that time that they wished to rely on what he had said or that they intended to amend their pleadings. Understandably, because time was running short at the end of the last day of the trial, the judge was not invited to rule on these issues at this stage. Nor does the judge deal specifically with them in his judgment, although he does so inferentially by finding liability on the basis of holes both on the claim and the Part 20 claim.
38. Platt now accepts that the judge was entitled to find against it on the basis of holes. Barkin does not, and says that we should not be influenced by the fact that Platt may be stuck with a liability which it cannot pass on. Mr Livesey repeats the submissions about unfairness to Barkin which were made by Mr Cannon to the judge. He adds that if a claim based on holes had been made against it, it might have wished to call further expert evidence about the fire resistance of the blankets.
39. I turn to the more general point about the case which Barkin had to face. Platt's Part 20 claim set out at some length Barkin's contractual responsibilities to prepare risk assessments, method statements and so on. It said that Platt had no powers or duties in relation to the method of working, access to the steel fixing or protection of Re-source's stock. These matters were for Barkin (among others). Paragraph 7 of the pleading then said:

In the premises Barkin was contractually responsible under the terms of the main contract … for fire precautions in relation to the steel fixing work and for supervising the same so as to enure that it was carried out safely and in particular was responsible for ensuring the safety of the claimants' stock.

Paragraph 10 (1) said that all this meant that Platt was not liable to Re-source. Paragraph 10 (2) added:

Further or in the alternative if Re-source sued Barkin in respect of the consequences of the said fire, Barkin would be liable to Re-source in that regard.

40. Barkin's defence contained predictable denials including an assertion that Platt had failed to plead any basis for its liability, to which Platt replied that the basis for Barkin's liability was to be found in para. 7 of its claim and that "the fact of the fire demonstrates that the fire protection which Barkin had agreed to provide was inadequate".
41. There was then a procedural skirmish in which Platt was persuaded to amend its claim to amend its pleading to add:

7(2) For the avoidance of doubt, Platt will contend that Barkin owed Re-source a common duty of care to discharge its said contractual duties with due care, because at all material times Re-source stood in Barkin's contemplation as a party likely to be directly affected by the proper performance of such duties.

(3) Platt will rely upon the fact that the whole purpose of the matters Barkin agreed to undertake was to ensure the safety of the claimant's stock and Barkin knew or ought to have known that if such matters were not dealt with carefully, the claimant's stock was likely to be damaged by fire.

and particulars under para. 10 (2) as follows:

Particulars

The said fire occurred because the fire protection for which Barkin was responsible was fitted in such a way that it allowed hot particles from the steel-fixing work to drop onto the claimant's stock, and no or no careful fire watch was maintained by Barkin. These matters gave rise to a foreseeable risk that the claimant's stock would be damaged by fire, as in fact happened.

42. Another skirmish shortly before trial produced further information under paragraph 10 (2) which said that Platt's case was that the shroud appeared conventional and adequate, but the fact that unacceptably large products of welding had not been trapped by it must have meant that there were small gaps around the stanchion or where the blankets were laced together and this was Barkin's fault. But having apparently confined its case in this way to a case of gaps, in his written opening submissions Mr Moxon-Browne said:

There has been some interlocutory skirmishing with [Barkin] directed to the question whether by their pleadings [Platt] has properly addressed the basis upon which Barkin's contractual role on site is relevant to establishing the parameters of their tortious duties to the claimant, this being centrally relevant to the Part 20 claim. Lest there be any lingering doubt, on either of these issues Platt further submits as follows….

The submission then set out the well-known proposition that in complex undertakings such as a building project the participants' obligations to third parties arising in tort are likely to be defined by the terms of their particular contracts. After summarising those duties, under the heading "Breach of Duty by Barkin" the submission broadly made the points which the judge found against Barkin in para 192 of his judgment.

43. In his opening written submission Mr Cannon said that:

Platt tries to raise a new unpleaded claim that Barkin was negligent for failing to ensure that a safe system of carrying out the work to the claimant's premises was devised and implemented. Platt would have to plead and prove that it would have complied with some other safer system. Given that Platt would be liable to Re-source on Re-source's primary case for its own deliberate failure to comply with what had been agreed, this is not a case which Platt can run.

This, I suppose, is an objection to the broad case outlined in Mr Moxon-Browne's opening submission, but no objection appears to have been taken at the outset of the trial before the judge. Moreover in the course of the hearing Mr Moxon-Browne asked questions of a number of witnesses to establish Platt's case on the responsibilities of the parties involved without objection from Mr Cannon.

44. In the course of his final submissions Mr Moxon-Browne said that if holes in the blankets were the cause of the fire the method of working prescribed by Barkin was "asking for trouble". The method never descended to detail as to how the shroud was going to be fixed, what material was going to be used, and whether it was going to be fire-proof, and if so, to what temperature. Mr Cannon responded by referring to the pleadings and the fact that Platt had tied itself to the case which had been set out in the further information given shortly before trial.
45. It is easy to be wise after the event but, no doubt because of constraints of time, Mr Cannon did not insist on the judge making a ruling on this issue, explain why his clients were prejudiced by having to meet such a case or ask for an adjournment or at least an opportunity to put in further submissions if the judge was prepared to consider the case against Barkin on the wider basis submitted by Mr Moxon-Browne.
46. By far the more serious of Barkin's complaints about the pleadings is that it was not open to Platt to make its wide case. If Barkin are wrong about this, I think the debate about the difference between gaps and holes becomes academic. There can be little doubt that Platt intended to make a case that as Barkin assumed entire responsibility for fire protection of the claimants' stock, the fact that there was a fire showed that they had not discharged this responsibility (see for example its reply). Such a case did not depend at all on whether there were gaps around or holes in the blankets. In any event in litigation of this kind the unexpected often happens. Advocates in such trials often find themselves having to make or meet changing cases. If on close scrutiny of the pleadings in this case this court were to hold that, whilst Re-source had been able to make a case against Platt based on holes when it had earlier clearly confined its case in the way I have described, but Platt could not make the same case against Barkin, the result would be manifestly unfair to Platt. There was in fact no real prejudice to Barkin by the change of case. Mr Andrews gave evidence after Mr Atherton and said that although the blankets were discoloured "there were never any holes in any weld blanket" (day 2 p. 114). The judge obviously did not accept this evidence. As to expert evidence, the court already had the evidence of Mr Butler and Mr Bullen about the effect of flame cutting and the fire resistance of the blankets to which I have already referred.
47. Looking at the pleadings and the other material to which I have referred I think it was open to Platt to make its wide case against Barkin. Its case obviously could have been pleaded more clearly and I can well understand why the further information which it gave shortly before trial was taken to mean that it had tied its case to gaps. But its written opening is clear and its conduct at trial was only consistent with the wider case which it had made clear it wanted to make. Looking at its earlier pleadings I ask rhetorically what was the point of spelling out in detail that it had no responsibility and Barkin had every responsibility for ensuring that the hotwork would not damage Re-source's goods, if it was not for the purpose of showing that Barkin would, if sued, be liable to Re-source? That was the conclusion which it spelt out in the amended paragraph 10 of its Part 20 claim and its reply and one which I think it was open to the judge to reach looking at the matter as a whole.
48. For these reasons I reject Barkin's grounds of appeal based on the pleadings. Having said this I think it is unfortunate that the judge did not deal specifically in his judgment with the issues on the pleadings which had been raised in the final submissions and I agree with what Lord Justice Brooke says about this in para. 87 of this judgment. I am not however persuaded that this amounted to a serious procedural irregularity, but Barkin were entitled to be told why, despite its objections, the judge had decided the case against it in the way he did.
49. It is convenient next to consider the judge's finding of 100% contribution based on Platt's wider case. I have already referred to the judge's findings of negligence against Barkin. Under the heading Contributory Negligence the judge said:

201. Platt also seeks contribution from Barkin under the Contribution Act. That Act allows any proportion of the loss to be ordered by way of contribution including a complete indemnity. The over-riding principle is whatever is just, fair and equitable.

202. In this case, justice, fairness and equity all inevitably point to Platt being given an indemnity amounting to 100% of its liability to Re-source for damages and costs. It is only necessary to consider the long catalogue of Barkin's negligence, the lack of any direct negligence by Platt save for that that had been directed by Barkin and by its stipulated method of working and Mr Andrews' behaviour. This behaviour consisted of his instigating the hotwork originally in highly contentious circumstances, of his directing it in a wholly dangerous manner and of his deliberate decision to leave site as soon as he learnt that a fire had started so as to avoid criticism in these regrettable acts. This was then followed by a lengthy campaign in which he sought to show that he had left site earlier than he did and in innocent circumstances, that the fire was exclusively caused by the reckless conduct of Platt which he knew that Platt had not engaged in and which aimed to vindicate both Barkin and himself and unfairly leave Platt solely responsible, liable and culpable for the fire.

50. Mr Livesey complains that in para. 201 the judge mis-directed himself. Section 2 of Civil Liability (Contribution) Act 1978 provides that:

The amount of contribution recoverable from any person shall be such as may be found by the court to be just and equitable having regard to the extent of that person's responsibility for the damage in question.

In his summary of this statutory provision the judge omits the reference to causation and in para. 202 the judge relies on among other things Mr Andrews' behaviour after the fire which was obviously not causative of it. In any event, Mr Livesey submits, the judge should not have assessed contribution at 100%. Platt was the specialist contractor. It should not have carried out hotwork protected only by blankets with holes in them. It should bear, if not the whole, at least a substantial proportion of the blame.

51. Section 2 of the 1978 Act is not expressed exclusively in terms of causative responsibility for the damage in question, although obviously the court must have regard to this, as the section directs, and it is likely to be the most important factor in the assessment of relative responsibility which the court has to make. But in the result the court's assessment has to be just and equitable and this must enable the court to take account of other factors as well as those which are strictly causative. Such an assessment made by a trial judge will only be altered on appeal if it is clearly wrong.
52. The judge found, as he said, a long catalogue of negligence against Barkin. It had undertaken entire responsibility for the way in which the hotwork was to be done and the protection of Re-source's stock. Platt's fault, through its young welder Mr Atherton, was to acquiesce in the use of the unsatisfactory protection provided by Mr Andrews. Mr Atherton's evidence was that he had never used welding blankets before. His employer Mr Platt was unaware of the holes in the blankets.
53. In these circumstances, whilst many judges would have apportioned some part of the blame to Platt, I am not persuaded that the judge's apportionment of 100% to Barkin was clearly wrong. His findings of negligence in para. 192 justified such a finding. It was clearly just and equitable in view of the way in which Barkin had run its defence.
54. This conclusion makes it unnecessary to consider whether the judge was right to hold that Barkin were also liable under the letter of indemnity. I will however state my conclusion about this shortly.
55. The argument below, which the judge rejected, was that the indemnity did not cover damage caused by Platt's own negligence. This argument was founded on the principle laid down in cases such as Canada Steamship Lines v King [1952] AC 192 that if an indemnity is intended to cover the consequences of the claimant's own negligence it must say so in terms. This however is only a general rule, normally applied as part of the contra proferentem principle. I do not think it was applicable to Barkin's letter of 3 May which Platt sought as a condition of agreeing to carry out the work in the way in which Mr Andrews decreed, about which Mr Platt was unhappy.
56. However I think there is another reason, (which does not appear to have been fully developed before the judge) why Platt could not recover under the letter of indemnity. The indemnity was conditional upon Platt ensuring that "fire blankets are used for the protection of internal stock". It was conceded that if no fire blankets were used there would be no indemnity. It seems to me that holed fire blankets would not meet the condition either. They would obviously not provide protection of the stock. For this reason I do not think that Platt was entitled to recover under the indemnity.
57. However for the reasons I have given I would dismiss Barkin's appeal.

Lord Justice Laws:

58. I agree that this appeal should be dismissed for the reasons given by Tuckey LJ. I understand and respect Brooke LJ's view that the judge below was at fault in failing to grapple clearly with the proposition – on which the case in fact went forward to judgment – that it was fair to allow Re-source to change its claim against Platt and to allow Platt to change its claim against Barkin. However:

i) In his written opening submissions for Platt Mr Moxon-Browne heralded the case against Barkin which the judge ultimately accepted at paragraph 192 of his judgment.

ii) Mr Cannon for Barkin made no specific request of the judge for a ruling as to the reach of Platt's proper case against his clients, nor seek an adjournment or an opportunity to put in further submissions.

iii) There was no real prejudice to Barkin arising out of Platt's change of case, for the reasons given by Tuckey LJ at paragraph 46.

iv) As Brooke LJ (if I may say so) succinctly states, the fire was Barkin's fault, not Platt's.

59. At all events I am clear that no injustice was done.

Lord Justice Brooke :

60. In a criminal trial on indictment the matters the prosecution has to prove against the defendant are staked out by the wording of the indictment. The trial judge has power to allow the indictment to be amended, if the evidence turns out to be rather different from what the prosecution expected, and if it is fair in all the circumstances to allow the amendment.
61. In a civil trial conducted in accordance with the Civil Procedure Rules ("CPR") the matters a claimant has to prove are set out in its statement of case, supplemented by any further information it may give in response to a request. In McPhilemy v Times Newspapers Ltd (No 1) [1999] 3 All ER 775. Lord Woolf MR said at pp 792-3 that it was legitimate to provide further information about one's case by reference to what is said in a witness statement, but the overall result is the same. The opposing party will then be able to understand the case he has to meet, adduce appropriate lay and expert evidence, consider the legal difficulties in his path and, where appropriate, make offers of settlement after assessing the strength of the other side's case. The same principles govern the conduct of Part 20 proceedings as govern the main action.
62. As Tuckey LJ says, the unexpected often happens in the course of a complicated trial. If it does, it is the judge's duty to reappraise the situation, if any of the parties request it, and to permit a party to amend its statement of case if satisfied that this can be done without injustice to a party affected by the amendment. If necessary and appropriate, an adjournment may have to be granted to permit further evidence to be adduced to meet the new case.
63. The very acute difficulty which we face on this appeal stems from the fact that Judge Thornton felt that he was free to make relevant findings regardless of whether the issues were staked out in the parties' statements of case and, in relation to the three matters mentioned by Tuckey LJ in paragraphs 20 and 21 of his judgment, in the teeth of the parties' contentions that they were not relying on these matters and did not wish him to make findings on them. Fortunately, no issues arose on any of those matters on the appeal. The difficulty that now confronts us stems from the fact that the judge's adverse findings against Barkin which Tuckey LJ has recited in paragraph 33 of his judgment are simply not reflected in the allegations of breach of duty set out in Platt's statement of case against it, reinforced as it was by further information supplied pursuant to a judge's order. This statement of case was never amended to reflect any further complaint. Nor did Platt ask for it to be amended.
64. The problem arose in this way. For reasons we do not know, Re-source elected to bring this action against Platt alone. Re-source was the company whose goods were destroyed by fire and Platt was the company whose activities were the immediate cause of the fire. Re-source chose to frame its claim against Platt in negligence, and it is certainly correct that the "particulars of breach" contained in paragraph 15 of its statement of case, both before and after it was amended, were framed in wide terms. Platt, however, requested further information about Re-source's case, and the answer, dated 11^th February 2002, stated in very clear terms:

"Only a tarpaulin was in position at any material time. No fire blanket (also referred to as a welding blanket) was in position at any material time. It was [Platt's] responsibility to ensure that a fire blanket was in position and not to proceed or continue with work to the final hole without a fire blanket being in position."

This case was never changed before the trial started, and Mr Philipps QC (for Re-source) made his opening submissions to the judge along these lines.

65. By its defence Platt denied that it owed any duty of care to Re-source other than a duty to carry out the steel-fixing work carefully and in accordance with their sub-contract. This was why it was at such pains in its defence to explain the heavy burden of responsibility that fell on Barkin. In its defence Platt stated in terms the case which the judge accepted at the trial, viz that before its representatives started work on the final stanchion, they found that it was protected by two welding blankets supplied and fitted by Barkin, and that they added a third blanket so as to provide additional fire protection.
66. Re-source chose not to join Barkin as a second defendant when it received Platt's defence.
67. So much for the parties' statements of case in the main proceedings. In the Part 20 proceedings Platt originally put its case in one of three ways:

(i) it was not liable to Re-source (and therefore there was no liability to pass on);

(ii) alternatively, if Re-source sued Barkin, Barkin would be liable to Re-source, and therefore Platt was entitled to an indemnity and/or contribution from Barkin;

(iii) alternatively, Platt was entitled to a contractual indemnity from Barkin.

68. It is hardly surprising in these circumstances that Platt set out very fully the reasons why, if it was found liable to Re-source , it should be entitled to pass the blame onto Barkin. So far as the facts were concerned, Platt simply repeated its case that it had found two welding blankets in place and added a third before work started.
69. In its defence to the Part 20 claim, Barkin accepted that if fire protection was in place on the thirteenth stanchion, Platt would not have been negligent. Its case was that Platt chose to carry out its works without tarpaulins and welding blankets being in place as required, and that Platt was therefore 100% to blame. In answer to Platt's claim for contribution, Barkin observed that Platt had failed to plead any basis for such liability. At the same time as it served its defence on 9^th May 2002, it sought further information of Platt of the way it was putting its case.
70. This led Mr Moxon-Browne, who was not assisted by a junior, to respond to Barkin's rebuttal of the claim to contribution by settling a reply to the effect that the basis of Barkin's liability to Re-source was pleaded in paragraph 7 of Platt's Part 20 claim. At that time that paragraph merely stated what Barkin's contractual responsibility was. It said nothing about any duty owed by Barkin to Re-source, still less about any breach of that duty. The final sentence in this short reply is set out by Tuckey LJ in paragraph 40 of his judgment. The reply was served on 27^th June 2002, together with a response to a request for further information which restated Platt's case on the facts. In response to a request focused on the basis for the claim for contribution or an indemnity, Mr Moxon-Browne used the same language as he used in the final sentence of the reply.
71. This, then, was the condition of the parties' statements of case when Judge Seymour QC gave directions for the timetable leading up to a trial on 16^th July, and it was on this basis that the parties commissioned their experts' reports. On 2^nd October 2002 Mr Butler (for Barkin) and Dr Bland (for Platt) submitted a joint statement to the effect that the use of non-combustible fire blankets would have been an acceptable approach, provided that there were no gaps in the fire blankets and provided that an effective fire watch was maintained. On the basis of this joint report, liability rested on the judge's decision as to whether or not those blankets were in place.
72. We do not know what led Platt to seek from Judge Seymour QC on 12^th November 2002 permission to amend paragraph 7 of its Part 20 claim (and to add particulars under paragraph 10(2)) three weeks before the trial started. The judge granted such permission "conditional upon such particulars being proper particulars of the allegations sought to be added by way of amendment". Barkin's lawyers succeeded in persuading the judge that this precision was necessary.
73. It was against this background that Mr Moxon-Browne for the first time stated why Platt claimed that Barkin owed Re-source a duty of care (see para 7(2) of the Amended Statement of Claim) and, much more importantly, identified (in para 10(2)) the breach of duty on which Platt relied. The breach of duty was expressed, so far as is material, in these terms:

"The said fire occurred because the fire protection for which Barkin was responsible was fitted in such a way that it allowed hot particles from the steel-fixing work to drop onto the claimant's stock."

74. Further information initially supplied responded to a request concerning Platt's fire watch case. This must include an answer (iv), which had no corresponding request but referred back to Barkin's contractual responsibility for matters of this kind. The critical further information, dated 13^th November 2002, which the judge required, made it crystal clear that Platt's complaint was that Barkin had not paid proper attention to the need for a close spark-proof seal around the stanchion being protected and to the lacing of the weld blankets to prevent gaps through which hot products might pass.
75. In his judgment Tuckey LJ has summarised what took place at the trial. In this judgment I have endeavoured to show how the pre-trial formulations of Platt's case against Barkin went through a number of permutations and only crystallised shortly before the trial started in a manner in which Barkin could understand clearly what breach of duty (as between it and Re-source) Platt relied on.
76. In the event, the judge found:

(i) that Barkin had put the weld blankets in place;

(ii) that the fire did not occur by reason of either of the matters of which Platt gave notice on 13^th November, but for an entirely different reason, namely because the weld blankets had holes in them because of the treatment they had experienced from the work done on the first 12 stanchions.

77. This had never been part of Platt's case against Barkin, even though Mr Atherton, who gave the crucial evidence, was Platt's employee and witness (although no longer in its employ). Nor had it ever been part of Re-source's case against Platt (see para 64 above). Even if Mr Moxon-Browne was fairly relaxed when Mr Philipps very informally (see Day 3B, p 51) sought permission to rely on this evidence against Platt at the very end of counsel's closing submissions (because Mr Moxon-Browne anticipated being able to pass the liability down the line to Barkin), this does not mean that it was necessarily fair to require Barkin to answer this very late charge, which had never formed any part of anybody's case, without a very careful analysis being made by the judge of what was now being alleged against Barkin and whether it was fair to complete the hearing on this altered basis.
78. Mr Atherton gave the critical evidence on the Tuesday. When Mr Andrews gave evidence soon afterwards, the possibility of there being holes in the weld blankets was addressed in only one question, to which Mr Andrews gave a flat denial (Day 2, p 114E). The likelihood of brand new weld blankets being damaged so quickly was never explored with him or anybody else. At the end of that day there was a discussion between counsel and the judge as to whether the parties' experts need attend for cross-examination. Neither Re-source nor Platt had asked for permission to amend their case in the light of Mr Atherton's evidence, and none of the experts had been invited to turn their mind to the possibility that the weld blankets could have been so unfit for their purpose that they were in the state described by Mr Atherton after only being used on 12 occasions.
79. On that Tuesday evening only Mr Cannon, for Barkin, anticipated the difficulties that might arise if his clients were obliged to face a case which turned on the resistibility of the weld blankets to welding material. At the end of a very long day the judge does not seem to have understood the potential difficulties Mr Cannon was talking about (see transcript Day 2, p 144), in that this was an issue which nobody had had any reason to foresee. And Mr Cannon himself did not in the end resist the judge's decision that it was unnecessary for any of the experts to attend to be cross-examined. He merely said (Day 2, p 145C):

"My Lord, I am in the position where I have not got my own expert in this particular field, so I am in the hands of the other parties. I have made it clear that it might give me more prejudice but that is a matter for Thursday."

80. In this passage he was referring back to what he had earlier described as a "pleading point" or a "notice point", because one question had been put to Mr Andrews by Mr Moxon-Browne (in relation to holes in the weld blankets) of which he had had no notice and about which he wanted to ask questions. He appeared to be reconciled to accepting the judge's desire that the experts should not be called simply to give evidence on some entirely new issue which none of them had been invited to address, and to saying that his position in relation to any new case, if he might be obliged to respond to such a case, would be that much more severely prejudiced if there was not even any expert evidence on the topic.
81. In his closing written submission, prepared on the Wednesday, Mr Philipps adopted Mr Atherton's evidence as part of his case (at paras 2.2.2 and 50). Mr Moxon-Browne contented himself at that stage to responding to Re-source's pleaded case, which did not adopt Mr Atherton's evidence. Mr Cannon similarly ignored the effect of Mr Atherton's evidence, and in a 20-page closing submission gave his client's answer to the case it had come to court to meet.
82. On the Thursday Mr Philipps's closing oral submissions were punctuated by comments on the judge's part which indicated his belief that Re-source could establish its case by a much simpler route than Mr Philipps had ever adopted. It was the judge who initiated the possibility (at Day 3A, p 4) that a requirement for non-combustible fire blankets presaged a requirement for fire blankets without holes in them, but the main thrust of Mr Philipps's submissions was to the effect that there were probably no fire blankets in place at all, or that there were gaps between the blankets through which spatter and debris could fall (Day 3A, p 6). Mr Philipps added, in passing, that there were gaps in the blankets themselves, according to Mr Atherton (Day 3A, p 6D), but he did not really develop this point in his submissions and he certainly did not apply to amend his case against Platt so as to include this point.
83. In his closing oral submission Mr Moxon-Browne complained forcefully that no case based on holes in the welding blankets had ever been advanced against him, and some of the discussion with him initiated by the judge thereafter evidenced the judge's predilection to make findings on liability even if the relevant issue had never been formulated in anyone's statement of case. Mr Cannon, in his turn, explained to the judge how the pleaded cases stood, and the judge again evidenced his willingness (Day 3B, p 43) to make a finding that the fire was caused through material dropping through the hole in the bottom of the blanket, notwithstanding that it formed no part of Platt's case against Barkin as it then stood.
84. The question whether the judge was entitled to decide the cases on the basis that Barkin had indeed fitted the welding blankets but they were not fit for their purpose because they contained holes, was only really opened up when the judge invited Mr Philipps to make further submissions at the end of another very long day. Mr Philipps now stated clearly (Day 3B, p 51) that he wished to adopt this contention, although he maintained that all he was doing was to move away from the narrower approach with which he had opened the case. The critical answer to the request for further information (see para 5 above) was not mentioned in his discussion with the judge.
85. Mr Moxon-Browne's attitude was that if Mr Philipps was allowed to reformulate his case in this way, then his clients should be allowed to pass the liability on to Barkin. Mr Cannon, for his part, argued that whatever might be permitted in the main proceedings, it was just not fair to raise this allegation against his clients in the Part 20 proceedings at this very late stage. He took the judge (Day 3B, pp 55-57) to the order made on 12^th November, and to Platt's reformulation of its case which had followed the making of the order (see para 14 above). He then reminded the judge (Day 3B, p 58) that nobody had put to Mr Andrews in cross-examination that he knew that the blankets were full of holes (a point Mr Moxon-Browne accepted, because this was not then any part of Re-source's case he was having to meet). Mr Cannon argued that if anyone had wished to run a case against Barkin that its representatives should have spotted the holes, this should have been done before Barkin called its evidence on the Tuesday and not in the course of counsel's final submissions.
86. The matter tailed off unsatisfactorily that evening, with Mr Cannon continuing to protest that it would be unfair to find against his clients on a point that had never been put to his witnesses, and the judge not making any ruling at all on any amendments he might be willing to allow.
87. In these circumstances one would have been entitled to expect that the judgment would have contained a careful analysis of the reasons why the judge considered that it would be fair to allow Re-source to alter its claim against Platt, and to allow Platt to alter its claim against Barkin in relation to a matter which had never been effectively explored in evidence at all, and in the face of Mr Cannon's protests that it would be unfair to his clients if the Part 20 case against them was altered in this way at this extremely late stage without any prior notice. The judgment contained nothing of this kind. The judge merely set out what he regarded as the relevant principles of law, and made his findings of fact and his findings of liability without any regard to the question whether his findings of liability were based on any of the allegations of liability the parties had made against each other in their pre-trial statements of case.
88. For my part, I consider that a serious procedural irregularity occurred. The judge should not have approached his task with such a cavalier disregard of the matters the parties were inviting him to decide, or of the difficulty of having to decide whether it was fair to allow their cases to be altered in a material way at the end of the trial after the evidence had been completed. Although the notice of appeal does not refer to the criteria in CPR 52.11(3)(b), these provide the tests an appeal court has to apply when it is said that a judge's findings go beyond the case pleaded and opened against the appellant. This rule provides, so far as is material:

"The appeal court will allow an appeal where the decision of the lower court was –

(b) unjust because of a serious procedural … irregularity in the proceedings in the lower court."

89. The reason why I am unwilling to dissent from Tuckey LJ's judgment is that I do not consider, at the end of the day, that the judge's decision was unjust because of this irregularity such that we should set it aside. The reason I have come to this conclusion is not because I would readily brush aside the matters of which Mr Livesey made complaint, but because the justice of the case flows with the grain of the judge's other findings. It is quite clear that responsibility for carrying out the task in this very risky matter rested with Barkin, and that Platt was worried about what was being proposed. Mr Moxon-Browne explained all this in his opening, although his arguments were then directed to a different issue. Mr Platt and Mr Atherton made a favourable impression on the judge, and Mr Andrews made a disastrous impression on him. And I do not consider that a further visit to the effect of Mr Atherton's evidence about the holes would have any realistic prospect of altering those impressions. Put shortly, the fire was Barkin's fault, not Platt's. If I felt that justice demanded that these two parties should be put to the heavy expense of a new trial because of the matters I have discussed in this judgment, I would not hesitate in advocating that course. But I do not.
90. I agree with Tuckey LJ's judgment on other aspects of the case. I, too, would therefore dismiss this appeal.