This appeal is from an order made on 19 May 2003 by His Honour Judge Hallgarten QC, sitting at the Central London County Court, in proceedings brought by Mr Rupert Loftus-Brigham and his wife, Mrs Suzanne Loftus-Brigham, against the London Borough of Ealing.

Mr and Mrs Loftus-Brigham are the freehold owners of a dwelling-house at 38 Drayton Green, London W13. The house was built at or about the end of the 19 ^th century and has been in their ownership for the past 40 years. It stands at the corner of Drayton Green and Drayton Bridge Road, which – as appears from photographs in the bundle before the Court – are attractive residential streets lined with mature trees. Those trees are the responsibility of the defendant, as local authority. Until removed in July 1999, shortly after the commencement of these proceedings, four of those trees – three limes and a plane tree – were immediately adjacent to the claimants’ property.

The house at 38 Drayton Green is built on London clay of high to very high plasticity. In or about August 1995 cracks began to appear in the property; or, perhaps more precisely, minor cracks which had developed some years earlier became sufficiently obtrusive to be a matter of concern to the claimants. By April 1997 the claimants had been advised that the property was severely affected by subsidence caused by desiccation of the clay sub-soil. They took the view, on advice, that the desiccation of the sub-soil was attributable to the root activity of the Council’s trees.

These proceedings were commenced by the issue of a claim form in June 1999. The claim was advanced in nuisance and negligence. The relief sought included an order that the trees be felled forthwith; and that was done, without admission of liability, within a few weeks thereafter. There remained a claim for damages in respect of the works needed to repair the house.

The Council served its defence at the end of July 1999. It denied that there was any, or any significant, desiccation of the sub-soil. But, if there were, paragraph 19 of the defence put in issue the question whether such desiccation of the soil as had occurred was attributable to the Council’s trees:

It is not in doubt that there was, at the material time, a substantial growth of climbers - Virginia creepers and wisteria – up the walls and over the roof of 38 Drayton Green. This growth can be seen in the photographs. As the judge put it, in paragraph 3 of his judgment: ". . . by 1995 it would appear that these climbers had virtually engulfed the house covering every external surface, including the left hand flank wall, the roofs and indeed, obscuring some of the windows, at any rate at the first floor level".

Paragraph 24 of the defence asserted contributory negligence on the part of the claimants:

The action was tried over three days at the beginning of December 2002. The judge heard evidence from expert witnesses called by each side. At paragraph 32 of his judgment – handed down in May 2003 – he set out what he understood to be the extent of the agreement between those experts:

After considering the evidence of the structural and other engineers, the judge expressed the view, at paragraph 44 of his judgment, that the survey readings which had been taken were consistent with the claimants’ case "that the exterior walls of the property moved seasonally ‘en masse’, best to be explained by moisture extraction and recovery of the soil". He then turned to the arboricultural evidence. He referred, at paragraph 45, to what he described as "one solid piece of evidence" – namely a plane tree root which extended from Drayton Bridge Road to the far side of the house – and he noted, without appearing to reject, the claimants’ contention that "damage to the house – including the rear extension – was thus liable to have been caused by all or some of the Defendant’s trees". But, at paragraph 47 he observed: "does this mean that the climbers are to be dismissed as a realistic cause of subsidence? I do not think so".

Having set out the rival contentions and the evidence advanced in support of them in detail and at some length, the judge expressed his conclusion at paragraph 56 of his judgment:

In the light of his conclusion that the Claimants had failed to establish their claim against the Council contributory negligence was no longer an issue. But the judge indicated (at paragraph 58 of his judgment) that, if he had been required to make a finding on that defence, he would have rejected it:

The Judge dismissed the claim. The claimants appeal to this Court with permission granted by Lord Justice Latham on 4 July 2003. The principal ground of appeal is that, in asking himself whether the claimants had shown that the Council’s trees were "probably the dominant cause" of the damage to their property, the judge had applied the wrong test. What, it is said, he should have done was to ask himself whether, on the balance of probabilities, the claimants had established that the Council’s trees were an "effective and substantial cause". Had he asked that question he would have been bound – so the appellants assert – to conclude that the answer was "Yes".

The issues on the appeal, therefore, are (i) what is the correct test if causation is to be established in a claim against a neighbouring tree-owner in respect of subsidence arising from the extraction of moisture from the sub-soil and (ii) did the judge apply that test, or some other and different test.

As we have said, the claim was brought both in nuisance and in negligence. That there were two concurrent causes of action had no effect on the conduct of the case; and rightly so in view of the observation of Lord Cooke of Thorndon, speaking for a unanimous House of Lords in Delaware Mansions v Westminster City Council [2002] 1 AC 321[31]:

It follows a fortiori from that analysis that the rules of causation in the two torts will also be the same. That would be apparent without authority, but the proposition is strongly supported by the respected analysis of Lord Hoffmann in the Kuwait Airways case [2002] 2 AC 883 [128], cited with strong approval by Lord Bingham of Cornhill in Fairchild v Glenhaven Funeral Services [2003] 1 AC 32[12]:

Some hesitation was expressed at the Bar as to whether decisions on causal principles in the law of negligence could be applied to the present case, in view of its having been primarily argued in nuisance, and that it was a "tree-root" case. For the reason already given, that diffidence was misplaced. Further, it is neither necessary nor appropriate to look for special causal rules applying to cases involving trees. The judgment of Mr Toulson QC in Paterson v Humberside CC (1996) Const LJ 64 is of valuable relevance to the present case; but that is because it applies general rules to a factual situation similar to that in the present case, and not because it lays down special rules for cases falling into that narrow factual category.

If we then turn to the rule or formulation appropriate in the law of negligence, and thus of nuisance - and have particular regard to cases, such as the present, where the damage may have resulted from two or more sources - the surest guide is found in the judgment of this court in Banque Bruxelles SA v Eagle Star [1995] QB 375, in a passage at 406E to which Mr Toulson QC referred:

The appellants complained that the judge had not applied the rules and approach just set out, but rather had required the claimants to show that the tree-roots had been either the sole cause of the damage, or at least the preponderant or (in the words of the Scottish court criticised by Lord Reid in Bonnington ) most probable cause. As we have seen, that complaint principally focussed on what appeared to be the judge’s general formulation of the rule in paragraph 56(5) of the judgment, which because of its central role in the appeal we set out again:

If that complaint is made good, then the judge erred. Mr Brown QC for the respondents sought to uphold the judge by a series of related arguments.

First, he submitted that on a proper construction of the judgment the judge had indeed applied the Bingham test, even though his language in paragraph 56(5) was, in verbal terms, inconsistent with that. For that argument Mr Brown principally relied on the judge’s formulation of the nature of the claim in the first sentence of the judgment: that the claim was on the basis that desiccation of soil caused by the tree roots was the substantial and effective cause of the damage. We would not give much weight to the appellants’ criticism that by use of the definitive article this formulation appears again to seek a single cause. We are, however, impressed by their demonstration from other parts of the judgment that the judge not merely expressed his task in terms of seeking a single or preponderant cause, but also applied that standard to the evidence. Thus in paragraph 56(4), discussing subsidence to the right hand side of the back extension, the judge said that

Second, Mr Brown argued, as the judge had said in refusing permission to appeal, that the two verbal formulations, an effective and substantial cause, and a dominant cause, meant the same thing. We doubt whether that argument is open to him in view of the judge’s actual approach to the case, as set out in paragraph 20 above, but in any event we cannot accept it. Mr Brown emphasised that the judge had said "dominant", not "predominant": he was accordingly simply speaking of the causal status or force of the event, looked at on its own, and was not engaging in any comparison with other causes. We cannot agree. "Dominant" in its natural meaning plainly carries comparative implications, in a way that "substantial" does not; and as his judgment shows, that is how the judge in fact applied the concept.

In so saying, we have not overlooked cases put before us by Mr Brown in which judges of high authority have used the terms "effective" and "dominant" apparently interchangeably in describing causal questions. But in none of these cases was the court addressing an issue of the order that concerns us. Rather, the cases are concerned either with the notion of "proximate cause", as applied in insurance contexts to determine whether loss had been caused by the peril insured against; or dealing with the difficulties of cases in which novus actus is alleged: - on which see, most recently, the speeches in the House of Lords in Reeves v Commissioner of Police [2000] 1 AC 360. Accordingly, when Glidewell LJ in Galoo v Bright [1994] 1 WLR 1360 at p 1374G said, in a passage particularly relied on by the respondents, that a breach of contract, to found recovery, must be shown to have been

Third, Mr Brown pointed out that the judge had prefaced the whole of his paragraph 56 by holding that

We are therefore satisfied that the judge erred in his approach to the question of whether the tree roots in law caused the damage complained of. The question that he should have asked himself was that approved by Lord Bingham in Fairchild : whether desiccation from the tree roots materially contributed to the damage. His judgment therefore cannot stand. Difficulties however arise in relation to how the Court should now proceed.

Miss Taylor sought to persuade us that there were sufficient findings by the judge in the earlier part of his judgment to enable this Court, applying the proper test, to resolve the causal issue in favour of her clients. We have some sympathy with that view in the light of the judge’s observations that we have set out in paragraphs 8 and 9 above; but we have concluded that the judge’s account of the evidence - in which we have to say it is not always easy to distinguish narrative from findings - is insufficient to enable the court to proceed with the necessary degree of confidence. The matter will therefore have to be remitted.

Normally, and particularly in view of the interests of economy, that remission should be to the same judge. We are however persuaded that that course would be inappropriate, in view of the judge’s strong expression of view, when the issue was ventilated with him after the trial, that the test that he had applied was in fact the same as that which the law requires. We fully accept that, if the matter were remitted to him, the judge would now apply the law as directed by this court. However, the claimants are entitled to say that, in view of the history of the matter, it would be better for the matter to be approached afresh. There are, in addition, two reasons why the normal course of remission to the judge who has already heard the evidence is less compelling in this case. First, although the judge has heard the evidence, he is now a long way away from it. Six months elapsed between the trial and the delivery of judgment, and the appeal process has now added a further period (though, it should be noted, less than the six months just referred to) to that time. Second, much of the evidence seems to have been given on a mistaken basis as to the nature of the question to be determined. That is apparent from the joint experts’ report of October 2002, which stated that they were not in agreement as to "which of the vegetation was the dominant cause" of the movement and damage to the property.

We therefore consider that the action should be remitted to a different judge for retrial. If those giving evidence have clearly in mind the issue that they have to address, as set out above, the proceedings should be very materially shorter, and easier for the court, than was the original trial. This Court invites the parties to propose directions that might be given to that end. We would also very strongly urge both parties to consider whether, in the light of the guidance given by this Court, it is not now possible for them to resolve their differences. Any further trial, even one not so burdensome as the last, is plainly to be avoided if at all possible.

We should mention in that connexion a suggestion tentatively made by Mr Brown that even if this Court found that the judge had erred it was open to it not to remit the matter - that is, make no order at all - because the costs of a retrial would be disproportionate to the amount in issue. While a court will take considerations of proportionality into account when considering whether to give permission to appeal, once an appeal is launched and decided there is, in our understanding, no discretion in the court to prevent the law taking its course; or to prevent the successful party from obtaining the fruits of his success. But we do not pursue this issue further, because even if such discretion exists we would not be prepared to exercise it in this case. The sums in issue are not trivial, and if the parties cannot resolve their differences the costs of determining liability in respect of those sums must lie where they fall.

A recurrent suggestion on the part of the claimants was that this court might cut the Gordian knot by simply apportioning the loss amongst the parties, on the basis that since both had caused the damage both should bear some part of the cost of that damage. Quite apart from the absence of sufficient material to undertake this task, the Court has no power to take that step. Apportionment in law has to be based on liability , not simply on causation. Since the judge found that the claimants were not contributorily negligent - a finding that has not been appealed - they were not liable in law for any part of the damage, whether caused by them or not. Here, as in Paterson , Ealing has to take its victim as it finds it.

The appeal is allowed. The order of 19 May 2003 is set aside and the action is remitted to the Central London County Court for trial before another judge.