Wallingford Bridge spans the Thames at the edge of the Chilterns, near where the Ridgeway runs down Grim’s Ditch before following the river to Goring. A very few houses in Wallingford have a river frontage near the bridge, among them the parties to this appeal. This is Mr Cattermole’s appeal from the decision of the First-tier Tribunal on his application for the determination of the exact line of the boundary between his land and that of the respondents, Miss Jordan, Ms Mullen and Ms Bamber.

I heard the appeal, which was a review of the FTT’s decision, at the Royal Courts of Justice on 20 April 2022. Mr Cattermole was assisted by Professor Amy Smith, who is a professor of classical archaeology. The respondents were represented by Mr Matthew Cannings of counsel.

In 1997 the first respondent, Miss Jordan, bought Lodge Cottage, Thames Street, Wallingford together with her partner. Later he died and his share was inherited by his daughters, which is why there are three respondents; Lodge Cottage remains Miss Jordan’s home. Thames Street is set back from the river and in 1997 Lodge Cottage had a large garden in the shape of a narrow strip between the surrounding properties and then opening out to form a broad area beside the river. In 2016 the respondents sold the larger area to Mr Douglas Rogers of East Gate House, 1-2 High Street, retaining the strip down to the water’s edge so that she could still see her own land from her sitting room. Before his purchase Mr Rogers’ property had no river frontage and the addition of the garden, now registered at HM Land Registry under title number ON325953, must have added considerably to the amenity and value of his home. The appellant Mr Cattermole bought East Gate House from Mr Rogers a few months later in October 2016.

I have reproduced on the next page the title plan of Lodge Cottage, showing the portion sold. The boundary that is in dispute is that between ON325953 and the strip of garden that still belongs to Lodge Cottage. There is no physical boundary, because the parties take the view that to put up a fence or wall would be a breach of a restrictive covenant imposed on the land in 1994 pursuant to an agreement made with the local planning authority under section 106 of the Town and Country Planning Act 1990. So the boundary is invisible on the ground. But the two owners each need to know where to stop mowing, where they can walk and sit, where they can moor a boat, and so on.

It seems to have become apparent to Mr Cattermole not long after his purchase that his view of where the boundary lay differed from Miss Jordan’s. In 2019 he applied to HM Land Registry for the boundary to be determined, and the resulting dispute was referred by the registrar to the FTT.

It is agreed that the western end of the southern boundary of ON325953 – described by everyone as point C – is at the intersection of the south and east walls of the churchyard. The quarrel is about the line it then takes to reach the river. The essence of the dispute before the FTT was whether the line is as depicted on the transfer plan, making an angle at C (as can be seen on the plan above), or whether the boundary continues the straight line of the churchyard wall as the respondents contended. The respondents’ line would give Miss Jordan 6.68 metres more river frontage than would Mr Cattermole’s.

The FTT decided that the boundary is as depicted in the transfer plan. The issue in the appeal is where that line lies on the ground.

Because the boundary is a straight line, it is convenient to distinguish the various lines that have been discussed in the course of the proceedings by their end point on the river bank, and to list those points from south to north. The southernmost candidate, giving most land to Mr Cattermole and least to the respondents, is shown on the plan submitted with Mr Cattermole’s application for a determined boundary (“the DB plan”), and is referred to as point D. The expert witnesses instructed by the parties agreed that the line on the transfer plan ends at point F, 40mm north of D at the northern edge of a buttress in the river wall, and Mr Cattermole was and remains content to adopt that point in preference to point D. The FTT decided that the line ends at point X, being the northern edge of a hole in the river wall (visible from the bridge or across the river when the river is sufficiently low), between points D and E but a little nearer E. Point E is the end of the respondents’ line, 6.68 metres north of D.

It will be helpful at this point to provide a summary of the legal principles that govern the determination of boundary disputes.

The legal principles are not, I believe, in dispute and I can summarise them briefly. In order to decide where the boundary lies the judge must start with the deed that created it. If it is unambiguous, and indicates for example by measurements the position of the boundary on the ground, then that is the end of the matter. If the deed is ambiguous or does not tell the full story then extrinsic evidence may be considered, and in particular the physical features on the ground. As Mummery LJ put it in Pennock v Hodgson [2010] EWCA Civ 873:

The court may also look at evidence of the subsequent conduct of the parties to that deed, but only insofar as that is evidence of what they intended in the deed: Ali v Lane [2006] EWCA Civ 1532.

In Cherry Tree Investments Limited v Landmain Ltd [2013] Ch 305 Lewison LJ reiterated at paragraph 130 the well-established principle that “collateral documents” cannot be used to influence the construction of a document that has been entered on the register of title. That casewas not about boundaries, and the cases relating to boundary disputes allow a little more leeway; in Scarfe v Adams [1981] 1 All ER 843 a wholly ambiguous transfer plan was construed with the help of the contract plan. But other documents will be considered only if they do not contradict the deed itself, and subject to the important principles that both the pre-contractual negotiations and the subjective intentions of the parties are irrelevant to the construction of the transfer.

Turning to matters of land registration, title plans depict general boundaries only. They do not purport to be entirely accurate, and in any event a line on a plan may be a metre or so thick on the ground. Section 60 of the Land Registration Act 2002 enables a registered proprietor to make an application for the exact line of a boundary to be determined. Such applications are usually supported by a plan; HM Land Registry’s practice guide indicates the registrar’s requirements for the scale of that plan but those are not statutory requirements, and indeed the statute does not require a plan at all; it would be possible for a boundary to be determined by reference to precise physical features, although it would be unusual for that to be possible.

The crucial issue before the FTT was whether the boundary was as depicted on the transfer plan, so that the respondents’ retained strip gets narrower at the river end, or as the respondents argued should follow a straight line produced from the south wall of the churchyard so that it gets wider towards the river.

The respondents’ line was obviously inconsistent with the transfer plan, which shows the strip getting narrower, and one might have thought that – in view of the legal principles set out above – they were bound to have an uphill struggle. For the respondents it was argued that the transfer should be construed by reference to a plan drawn up before the contract for sale was drafted, by David Barrington Limited and referred to as “the Barrington plan”. Both Miss Jordan and Mr Rogers gave evidence that the Barrington plan was the basis of Mr Rogers’ agreement to buy, before contracts were exchanged, but that once instructed to draft a contract the respondents’ solicitor used a plan based on the OS map which showed the land to be purchased shaded in red up to a dashed line, and gave rise to the shape indicated on the title plan and reproduced above. The respondents and Mr Rogers signed the plan without noticing that it was not the same as the Barrington plan.

Mr Cattermole argued that the line was as shown on the transfer plan, with an angle from the churchyard wall; he drew the DB plan himself (he is an engineer) and claimed that it was an accurate depiction of the line created by the transfer.

The two expert witnesses, Mr Carl Calvert FRICS for Mr Cattermole and Mr John Rackham MIRICS for the respondents, each produced a report and then a joint statement in which they agreed that the line depicted on the transfer plan started at point C and ended at point F, 40mm north of Mr Cattermole’s point D. The experts were not required by the parties to attend the FTT hearing because they had agreed that line.

The decision of the FTT reflects the valiant endeavours of counsel for the respondents (not Mr Cannings) to persuade the judge that the Barrington plan could be used as an aid to construction. The judge referred to Cherry Tree Investments Limited (see paragraph 13 above), and recounted the arguments made by counsel for the respondents to the effect that this was obiter and that collateral documents should be admissible. She observed at paragraph 34 that “boundary disputes are different” and that it might be useful to refer to other documents because land is registered only with general boundaries, and because physical features on the land might change and might be unknown to later purchasers. But she conceded that the authority of the Court of Appeal was against this view and decided that “little weight should be given to collateral documents when determining a boundary.”

The judge then summarised the evidence given to her about the sale from the respondents to Mr Rogers, and said that she thought that Mr Rogers and the first respondent may both have thought that the Barrington plan would form part of the contractual documents. She recounted Mr Cattermole’s evidence about what he was told by both the first respondent and Mr Rogers when he purchased, about which the parties were not in agreement. She described what she saw on the site visit, including the hole in the river wall of which she saw just a little, and referred to a photograph in the bundle taken when the water was lower so that more of the opening is visible.

Turning to the construction of the transfer, the judge rejected arguments that anything in the conduct of Miss Jordan and Mr Rogers after the transfer, for example in conversations with Mr Cattermole, shed any light on their intentions in the transfer. Instead, she indicated at paragraph 61 that she must imagine herself standing on the ground with the transfer plan in 2016 and work out what a reasonable person would conclude was intended to be transferred. She added that that would involve allowing no more than minimal weight to the Barrington plan.

She therefore (at her paragraph 62a) rejected the respondents’ line C-E because the transfer plan depicted an angle at point C.

Turning to Mr Cattermole’s line C-D she commented at 62b that although it produced a shape that was consistent with that on the transfer plan, it left the respondents with “a rather odd parcel of land” which was largely unusable at the river frontage because of the number of trees there. She said it seemed unlikely that the parties would have intended that, bearing in mind that the price paid for the garden by Mr Rogers, £75,000, “would appear (from what Mr Cattermole told me) to be a low price for the loss of a usable river frontage.”

At paragraph 62c the judge observed that neither of the parties’ lines attached any significance to the dashed line on the transfer plan. She explained later, in her refusal of permission to appeal, that what she meant was that neither party had suggested that the dashed line corresponded with any physical feature. She found that it was intended to delimit the shaded area, and therefore asked herself what it represented, in light of the fact that dashed lines on OS maps tend to show minor features such as changes in surface or features overhead of underground. She took the view that on the balance of probabilities the dashed line represented a drain running from the corner of the churchyard to the outlet in the river between the first and second buttresses.

She added at point 62d that “The colouring appears to stop short of the dashed line so I consider that the best interpretation of the plan is that the boundary runs immediately north of the drain” to a point which she called X on the river wall “immediately north of the drain” between points D and E, and about 3.6 metres from point F; this she felt (paragraph 62e) was within an acceptable tolerance of the experts’ point F and was close enough to justify a conclusion that this was what the parties objectively intended.

Mr Cattermole was successful in persuading the FTT that the boundary was as depicted on the transfer plan. Indeed any other decision would have been manifestly wrong. But he disagrees about where the line on the transfer plan lies on the ground. He appeals, with permission from the Tribunal, on the grounds

There is no cross-appeal; for the purposes of the appeal it is accepted that the boundary is as shown in the transfer plan, and the issue is where that lies on the ground so that an exact line can be determined for the purposes of section 60 of the Land Registration Act 2002.

The first ground of appeal relates to the judge’s findings about the price of the land and the position of the drain. These are findings of fact, and I am aware that the Tribunal will rarely interfere with such a finding when the FTT has seen or heard the witnesses. But that is not to the point here, where the problem is one of relevance, so far as the price is concerned, and of an absence of evidence of the presence of a drain.

As to the price, it was suggested that the judge misunderstood what Mr Cattermole had said about the value of the land. He had given evidence that the price of East Gate House with the additional garden was £725,000, in contrast to the £355,000 it had fetched without, but that should not have been translated by the judge directly into the value of the garden itself.

I agree, but the pertinent legal point is that the exercise prescribed in Pennock v Hodgson does not involve assessing the price or value of the land concerned. Mr Cannings argued that the price was stated in the transfer and could be taken into account by the judge in construing the transfer, but he was not able to refer me to any authority where the value of the land had formed part of the decision about the boundary. It is in my judgment irrelevant. Had it been a relevant consideration then the judge’s finding would have been unsafe because she did not hear any expert evidence about the value of the land, but in any event the value or price of the land transferred was irrelevant.

Mr Cannings also sought to argue that the judge’s comment about the price of the land sold was not material to her decision, but I reject that suggestion. Her point at paragraph 62b was not simply that the C-D line left the respondents with an odd parcel of land, but also that the price they had received was insufficient for what they had lost; her comment about the price was integral to the rejection of the C-D line.

As to the judge’s finding that the dashed line represented a drain, Mr Cattermole has sought to assert that the drain simply could not be in the position the judge put it, and I make no comment about that. The real difficulties here are first that the judge had heard no evidence about whether there was a drain from point C to the hole in the river wall, and second that the possibility was not mentioned at the hearing. I take it that the judge came to her conclusion after the hearing, but she did not then give the parties or their expert witnesses the opportunity to make submissions about it.

Mr Cannings argued that the judge saw the hole in the wall at the site visit and was entitled on the basis of that evidence to come to the conclusion she reached, but that argument cannot possibly succeed. The judge’s decision took into account, and indeed she rested her conclusion upon, a finding of fact that the dashed line represented a drain. The existence of a hole in the river wall and the existence of the dashed line on the map are together insufficient to lead to an inference that there is a drain, or that any other structure, under the ground in a line from point C so as to account for the dashed line. That finding was not open to her on the evidence before her; if it had been there would still have been an issue of procedural unfairness because she did not allow the parties to make representations about it.

For the two reasons given in the first ground of appeal the FTT’s decision must be set aside.

I need not therefore say very much about the second ground. I reject Mr Cattermole’s suggestion that the judge did not read the experts’ reports; but what she did not do was to explain why she rejected their methodology. Mr Cannings argued that the experts’ methodology was so poor that the judge was right to disregard it, but the judge did not say that. She rejected the conclusion because of her own, unevidenced, view about the drain, but that was not enough. It is a rare blessing in a boundary dispute to have agreement by the expert witnesses, unchallenged by the parties, as to the position on the ground of the line on the transfer plan, and the judge should have explained why she rejected that agreed evidence. That she did not do so is a further reason why the decision must be set aside.

Thus far then the appeal succeeds and the decision of the FTT is set aside insofar as it relates to the position of the boundary on the ground. I asked the parties at the appeal hearing whether they would wanted me to remit the matter to the FTT or to substitute the Tribunal’s own decision, and they agreed that I should decide the determined boundary application, on the evidence that was before the FTT.

I can do so quite briefly.

My starting point is the unappealed conclusion that the boundary is depicted by the transfer plan and not by the Barrington plan. I would comment only that the parties’ evidence about the Barrington plan was evidence of their subjective intentions at the time of the sale and was irrelevant. Moreover the Barrington plan was inconsistent with the plan in the transfer. It should have been given no weight at all, and I give it none. It would have had central relevance to an action for rectification of the transfer, which the respondents and Mr Rogers might have brought by consent had they spotted the problem before the sale to Mr Cattermole, but it has no relevance to the construction of the transfer that created the boundary.

As did the FTT, I must start by looking at the transfer plan. It is unfortunate that the parties did not draw a line on the plan to indicate the boundary of the land transferred, because that has left room for suggestions, which the judge accepted, that the shading stopped short of the dashed line. In my judgment, looking at the copy that the parties agreed, at the hearing of the appeal, to be the best copy of the transfer plan in the bundle, it does not. That makes sense, because the reason why the transfer plan did not have a drawn line for the boundary was that the intention was to use the dashed line. Accordingly the boundary does not lie north, or “immediately to the north” (whatever that means, see paragraph 24 above) of the dashed line, and what I have to decide is where the dashed line is on the ground, bearing in mind that that is not a simple matter because the thickness of the dashed line on the plan denotes a much thicker area on the ground.

There was some discussion in the course of the appeal about how the dashed line got on to the plan in the first place. It was suggested that it derives from a deed imposing restrictive covenants on the land in 1994. The origins of the line are irrelevant; the line is the one the parties to the transfer to Mr Rogers chose to use, on an objective consideration of their choice and intentions.

So where is it on the ground? I take the judge’s point that lines, solid or dashed, on the Ordnance Survey map usually indicate a physical feature, but there is none visible; the reasonable person with the plan in their hand gets no assistance from anything in the garden. The dashed line may derive from an older edition on the OS map and may originally have represented a physical feature, but there is nothing there now. I attach no significance to the opening in the river wall; it could be the end of a drain or a culvert but there is no evidence as to the position of whatever lies under the ground.

The only evidence before the judge was the experts’ agreed conclusion that the line on the transfer runs from C to F, where F is the northern edge of a buttress on the river wall. That conclusion was unchallenged by the parties. Mr Cannings told me, on instructions, that Miss Jordan does not agree that that is where the line on the transfer runs on the ground, but the respondents were legally represented before the FTT and chose not cross-examine Mr Calvert or Mr Rackham.

Mr Cannings argued that the experts’ methodology was flawed. Mr Rackham reached his conclusion as to where the line runs by scaling up from the plan on the transfer, and the notes on the plan itself give warning that that may not be reliable. Mr Calvert looked for a physical feature that would denote the extremity of the line of the boundary, and found the buttress; Mr Cannings says that there is no reason why there should be such a feature at the end of the boundary.

Nevertheless that was the best evidence available to the FTT. I would add that although scaling up a land registry plan is not generally a reliable exercise, it is less unreliable where the line in issue is a straight one so that if the angle from point C is right then the line will be correct.

Accordingly the exact line of the boundary is from point C, at the intersection of the south and east walls of the churchyard, to point F, the northern edge of the buttress identified by the experts in their joint statement and on their agreed plan.

What is then the outcome of the applicant’s application for a determined boundary? Rule 40 of the Tribunal Procedure (First-tier Tribunal) (Property Chamber) Rules 2013 provides:

Accordingly the Tribunal can give to the registrar one of the two directions set out in rule 40(2),with a condition attached as to an entry on the register. I shall direct the registrar to give effect to Mr Cattermole’s application as if the objection had not been made, on condition that instead of his original DB plan the plan entered on the register shall be the plan produced by the experts indicating that the line runs from C to F. If either party wishes to make any representations about the form of that order they may do so within 14 days of the date of this decision, following which I will make the direction taking any representations into account.