Joyce v Rigolli

Three issues arise on this appeal from the order of His Honour Judge Hull QC sitting in the Epsom County Court dated 3 April 2003:-

Chanton Drive is a cul de sac. 7 Chanton Drive comprises part of a plot of land taken from the end of the garden of 6 Chanton Drive and an additional piece of land acquired from a neighbour. The original plot had been separated from 6 Chanton Drive before the Joyces came to live there but Mrs Joyce acquired it thereafter. Like 6 Chanton Drive, 7 Chanton Drive abutted the cul de sac. The eventual purchaser of 7 Chanton Drive was Mr Rigolli, and he built a house and garage there in 2000. At the time of the sale of 7 Chanton Drive to Mr Rigolli’s predecessor in title, the boundary between 6 and 7 Chanton Drive was not pegged out. It was not the same plot as had originally been taken from 6 Chanton Drive as part of that plot was retained as part of the garden of 6 Chanton Drive, and the additional piece of land had been added to it.

The transfer of 7 Chanton Drive dated 27 April 2000 by Mrs Joyce to Mr Rigolli’s predecessor in title made provision for the boundary between 6 and 7 Chanton Drive as follows:-

“The distance between the points marked C and D on the plan annexed hereto being a line from the northern corner of the building at 6 Chanton Drive at right-angles to the boundary between the points marked A and B shall be 8.382 metres (27 feet 6 inches).”

The annexed plan was an office copy of the filed plan for 6 Chanton Drive before the sale of the plot and it showed the existing house built on 6 Chanton Drive. On to this plan, the new boundary and the old boundaries between 6 and 7 Chanton Drive had been added. The ends of the new boundary were marked A and B respectively and the corner of the house nearest the new boundary was marked C. The line from point C intersected the new boundary at point D. As far as I can see, the plan annexed to the transfer dated 27 April 2000 does not purport to show the additional piece of land added to the plot, but nothing turns on that.

The transfer to Mr Rigolli’s predecessor in title also provided that the transferee should not build on 7 Chanton Drive except in accordance with a planning permission dated 25 November 1999. The transferee also had to complete a crossover to the cul de sac in accordance with the requirements of the local authority.

The judge found that the old boundary and the new boundary were to be parallel. However, the old boundary was never fixed, so this finding, in the judge’s words “carried the matter no further”.

The judge then considered whether the paper title placed the new boundary. Unfortunately, the transfer dated 27 April 2000 did not state the angle at which the perpendicular line C to D was to leave the house built on 6 Chanton Drive. If a line of 8.382 metres was taken from the house an arc could be drawn, part of which cut across the boundary as ultimately marked out by Mr Rigolli. The judge found that the furthest point of this arc from the boundary was about 2.5 metres. At trial Mrs Joyce adduced the expert evidence of Clive Francis, a land and measured building surveyor. Mr Francis was able to ascertain the boundary between 6 and 7 Chanton Drive by extending the line of the boundary of the plot shown on the Land Registry plan before the sale of the plot so as to meet other buildings in the area and then finding an angle between those buildings and the house on 6 Chanton Drive. Mr Francis found this angle to be 63.5 degrees. Mr Francis then plotted the new boundary line on a new map using this angle and the buildings identified in the manner described above. Mr Rigolli adduced the evidence of William J Maskell, an architect, whose opinion was that it was not possible to fix the boundary between 6 and 7 Chanton Drive from the conveyance plan.

In 2000 Mr Rigolli built a bungalow on 7 Chanton Drive. He placed a close-boarded fence between his property and 6 Chanton Drive. He built a garage close to this fence. The appellant’s case is that part of this fence and part of the garage encroaches on 7 Chanton Drive.

The judge found in favour of Mr Rigolli. The judge noted weaknesses in Mr Francis’ work:-

“I shall not attempt to describe further Mr Francis’ very careful and impressive work. There are certain obvious weaknesses which follow from its application. First and foremost, under Rule 278 of the Land Registration Rules, the filed plan is deemed to indicate general boundaries only and the exact line is left undetermined. Rules 278 and 279 would appear to have been entirely overlooked by the parties and their advisers at the time of the material transfers. But if they are observed, then Mr Francis’ entire operation is logically and legally illegitimate or probably so. Secondly, to project lines less than half an inch long representing a width on the ground of two feet or more so as to arrive at angles accurate to half a degree appears to me to be contrary to all common sense. Thirdly, as can be seen from the plan at p.110, the angle is made with an assumed boundary line. Mr Francis admitted that the lines of the fences on the eastern boundaries of 6 and 7 Chanton Drive were not straight. They apparently run along the top of a slope perhaps 8 feet high, leading down to playing fields beyond, and tend to follow the top of the slope. Mr Francis also admitted that a difference of as little as 7 degrees would mean that the fence was in the correct position and did not constitute a trespass.”

The judge, however, did not accept the evidence of either expert. He held:-

“If I had to choose between the views of these two experts, I should, I think, be obliged to prefer the opinion of Mr Maskell, notwithstanding the very high quality of Mr Francis’ work. Even if Mr Maskell is correct, however, it follows that a very small part of the land in Mrs Joyce’s paper title has been taken by Mr Rigolli – see the arc of the circle on the drawing at 848-3 – unless it had been agreed that the radius 27 feet 6 inches was not to be perpendicular to the boundary.”

Having thus held that the boundary could not be identified from the title documents, the judge went on to making findings about a boundary agreement alleged by Mr Rigolli to have been made orally by Mr Joyce and himself in November 2000. Mr Rigolli’s pleaded case as to what was agreed on that occasion was as follows:-

“At the meeting in November 2000 referred to in the previous paragraph (1) the Claimant and Mr Joyce agreed the position of the boundary between No 6 and No 7 on the ground. A measurement of 26 feet 6 inches was taken from the northern corner of the Claimant’s house. That measurement reached a cherry tree shown marked Y on the plan attached hereto (‘the Defendant’s plan’). Mr Joyce said that the Claimant wished to retain the cherry tree and it was therefore agreed that the boundary fence should be erected so that the cherry tree would remain on the Claimant’s side of the fence, but that to the west of the cherry tree the fence would revert to the original line (i.e. the line of the boundary on the basis that the cherry tree was on the boundary). The line of the boundary running through the cherry tree is shown marked green on the Defendant’s plan. The line of the boundary fence as agreed at the meeting (and as constructed) is marked red on the Defendant’s plan. (2) Stakes were inserted in the ground and a string was run from the front (western side) of the plots to the rear (eastern side) of the plots to identify the boundary. The western end of the boundary was in line with a drain/manhole on Chanton Drive shown in the position marked X on the Defendant’s plan.”

The judge set out the evidence at length. Mr Joyce denied that there was any agreement. The meeting was attended by Mr French, Mr Rigolli’s contractor, who wanted to mark out the site for the garage foundations. Mr Rigolli’s evidence was that a measurement of 27 feet 6 inches from the north corner of the house on 6 Chanton Drive reached the middle of the cherry tree but he agreed to let Mr Joyce have the cherry tree so that there was a kink in the boundary. Mr Joyce’s evidence was that the boundary was in a straight line on Mr Rigolli’s side of the cherry tree, and that it was not agreed that the boundary would come out at the cul-de-sac at the two gulleys. Mr French used the drawing attached to the planning permission rather than the title plans but I note that on the judge’s findings he pegged out the boundary on the basis that a line of 8.382 metres from the north corner of the house on 6 Chanton Drive intersected the boundary at right angles (as mentioned in the transfer of 27 April 2000). Mr French had insisted on a meeting between Mr Rigolli and Mr Joyce in order that the boundary should be agreed. Mr French gave evidence that the boundary as provisionally pegged out was then moved back to Mr Rigolli’s side to allow the cherry tree to remain on 6 Chanton Drive. He did not refer to a kink in the boundary to accommodate the cherry tree. Accordingly, Mr French supported Mr Rigolli’s case that there had been an agreement as to where the boundary should be.

The judge accepted the evidence of Mr French in these terms:-

“I am satisfied by that evidence. I found Mr French an extremely frank and trustworthy witness. I am content to act on the evidence of this witness, corroborated to an extent, of course, by Mr Rigolli, though contradicted, of course, by Mr Joyce. As I say, I find quite clearly that Mr Joyce was the agent of Mrs Joyce and it was quite clear that he had authority to fix the boundary and that he did so. I find that the boundary was agreed.”

The judge concluded his judgment with the following paragraph:-

“I find that the boundary was agreed, and that that agreement having been acted on is binding on both claimant and defendant. It was admitted and it is self-evident that if the line of the boundary runs beside and parallel to the garage and is in line with the inspection cover and the two drains in the road, then the defendant’s development is within his boundary. In my judgment, that is precisely what happened. Moreover, Mr Joyce, who, as I have said, was plainly acting for his wife, did nothing to prevent the erection of the southern gate post, the drive, the fence, the garage or other work between November 2000 and May 2001, or at any time before October 2001. When Mrs Joyce raised her doubts about the gate post, then Mr Joyce told her there was nothing that could be done because they had sold the land. In these circumstances, having considered the judgment in Neilson v Poole (1969) 20 P&CR 909 it appears to me that the parties have reached a binding agreement , which the court must recognise, with regard to the line of the boundary between their properties, and the action therefore must be dismissed.”

Mr Hutchings, for Mrs Joyce, contends that the judge failed to consider whether the boundary between 6 and 7 Chanton Drive could be ascertained from the paper title, and that he should have accepted Mr Francis’ evidence, summarised above, as to where the boundary was. On that basis, extrinsic evidence as to the agreement in November 2000 was inadmissible.

Mr Howard Smith, for Mr Rigolli, seeks to uphold the judge’s judgment on this issue.

Mr Hutchings submits that a boundary agreement which actually purports to transfer an interest in land must be in writing and that the question whether a contract falls within section 2(1) of the 1989 Act does not depend on the intentions of the parties. He also submits that there was no consideration for the boundary agreement found by the judge. Moreover in making the boundary agreement, Mr Joyce would have been acting under a mistake that no disposition of land was involved.

Mr Smith submits that the question whether a boundary agreement falls within section 2(1) of the 1989 Act depends on whether the parties intended to dispose of any interest in land. The parties’ mutual promises were the consideration for the agreement. There was no evidence as to any mistake by Mr Joyce.

Mr Hutchings submits that Mr French’s evidence was that a straight line boundary was agreed in November 2000, not a kinked line as Mr Rigolli contended and as the judge found. The judge needed to explain this discrepancy. Alternatively, the judge misinterpreted the evidence.

Mr Smith seeks to uphold the judge’s judgment. The substance of Mr French’s evidence was consistent with that of Mr Rigolli. He also relies on the fact that Mr Joyce did not complain about the boundary for some five months.

In my judgment, it is clear that the judge rejected the possibility that the boundary could be ascertained from the title plans to 7 Chanton Drive, and that he was right to do so for the reasons which he gave. The detail provided by the plans was insufficient to fix the boundary with any degree of certainty. This aspect of the case is another example of the problem which arose in Scarfe v Adams [1981] 1 All ER 843 on the division of a house. In that case, this court held that extrinsic evidence was admissible because the Ordnance Survey map used in the conveyance to identify the relevant property had been wholly inadequate due to its small scale. Griffiths LJ cited the following passage from the judgment of Buckley LJ in the earlier unreported case of Kingston v Phillips ([1976] Court of Appeal transcript 279):

“It will be observed that the parcels as there set out are really almost devoid of any particularity; all that is said about the property conveyed is that it is part of the Chicklade Estate and part of the dwelling house thereon. Unhappily, the plan which was annexed to that conveyance is wholly inadequate to perform the function which the draftsman of the conveyance seems to have contemplated that it would. It is a very dangerous practice for a conveyancer to frame a conveyance with parcels which are not adequately described. Perhaps the most important feature of all the features of a conveyance is to be able to identify the property to which it relates; and, if the draftsman of the conveyance chooses to identify the property solely by reference to a plan, it is of the utmost importance that he should make use of the plan which is a on a scale sufficiently large to make it possible to represent the property and its boundaries in precise detail, giving dimensions and any other features which may be necessary to put beyond doubt the subject matter of the conveyance.”

In Scarfe v Adams, this court warned conveyancers that, where a property was divided, it was absolutely essential that the transfer or conveyance should describe the property to be conveyed with sufficient particularity and precision so that there was no room for doubt about the boundaries of each. As Cumming-Bruce LJ in that case said:-

“The facts of the present case are really very simple, but I hope that this judgment will be understood by every conveyancing solicitor in the land as giving them warning, loud and clear, that a conveyancing technique which may been effective in the old days to convey large property from one vendor to one purchaser will lead to nothing but trouble, disputes and expensive litigation if applied to the sale to separate purchasers of a single house and its curtilage divided into separate parts. For such purposes it is absolutely essential that each parcel conveyed shall be described in the conveyance or transfer deed with such particularity and precision that there is no room for doubt about the boundaries of each, and for such purposes if a plan is intended to control the description, an Ordnance map on a scale of 1:2500 is worse than useless. The plan or other drawing bound up with the deed must be on such a large scale that it clearly shown with precision where each boundary runs. In my view the parties to this appeal are the victims of sloppy conveyancing for which the professional advisers of vendor and purchasers appear to bear the responsibility. We are not concerned in this appeal with determining or apportioning that responsibility. This court has to try to reduce to order the confusion created by the conveyancers.”

In this case, the judge was rightly critical of what he also termed “sloppy conveyancing”. I agree with his comment that this case illustrates the time and trouble which can be caused by sloppy conveyancing. It would have been far more satisfactory to the vendor, Mrs Joyce, if the boundary had been fixed in a proper manner before she sold 6 Chanton Drive to Mr Rigolli’s predecessor in title. This case is an object lesson for conveyancers. Boundary disputes are costly in terms of the money, court resources, and the strain they impose on the parties individually and in their relations as neighbours. It is in the interests of consumers of legal services and the public generally that conveyancers should take careful note of the warnings about imprecise boundaries given and now repeated by this court in several cases.

Mr Hutchings criticises the judge’s reference to rules 278 and 279 of the Land Registry Rules. On his submission, this case falls within rule 279 rather than rule 278. These provide:-

“278 General boundaries

(1)

Except in cases in which it is noted in the Property Register that the boundaries have been fixed, the filed plan or General Map shall be deemed to indicate the general boundaries only.

(2)

In such cases the exact line of the boundary will be left undetermined – as, for instance, whether it includes a hedge or wall and ditch, or runs along the centre of a wall or fence, or its inner or outer face, or how far it runs within or beyond it; or whether or not the land registered includes the whole or any portion of an adjoining road or stream.

(3)

When a general boundary only is desired to be entered in the register, notice to the owners of the adjoining lands need not be given.

(4)

This rule shall apply notwithstanding that a part of the whole of a ditch, wall, fence, road, stream, or other boundary is expressly included in or excluded from the title or that it forms the whole of the land comprised in the title.

279 Where physical boundaries do not exist

Where, and so far as, physical boundaries or boundary marks do not exist, the fullest available particulars of the boundaries shall be added to the filed plan or General Map.”

However, as I read the judge’s judgment (set out above) the judge was doing no more than saying that the boundaries appearing in the Land Registry plan used as the basis for the plan annexed to the transfer dated 27 April 2000 were not definitive and, therefore, could not form the basis of a reliable fixing of the boundaries by the experts. These boundaries included the boundary referred to as the eastern boundary of 6 and 7 Chanton Drive, on which Mr Francis had relied and which appeared on the filed plan for 6 Chanton Drive.

The leading authority on boundary agreements is Neilson v Poole (1969) 20 P&CR 909. In that case, Megarry J held that the boundary could be ascertained from the title documents. However, he went on to hold that even if he was wrong in the construction that he placed on the conveyance, the plaintiff nevertheless succeeded on his alternative claim based on a boundary agreement. With consummate clarity of expression and understanding of ordinary life, in Neilson v Poole Megarry J analysed boundary agreements into two different types: those that constitute an exchange of land and those by which the parties merely intend to “demarcate” an unclear boundary referred to in title documents. He held that it is to be presumed that when parties informally agree a boundary they are making an agreement of the latter class. The latter class of agreements have as their purpose the identification of a boundary, not the conveyance of land. Accordingly, such agreements do not constitute “a contract … to convey or create a legal estate” for the purposes of section 10(1) of the Land Charges Act 1925.

Megarry J expressed these points in the following passage:-

“I turn next to the defendant’s plea that the boundary agreement is void against him for want of registration as a land charge. It was admittedly not registered, and the only question is whether it was registrable. The point is devoid of any direct authority. A boundary agreement may, I think, be registrable, or it may not, depending on the nature of the agreement. The only suggested head of registration is as an estate contract, Class C(iv). By section 10 (1) of the Land Charges Act 1925, this is defined as :

any contract by an estate owner or by a person entitled at the date of the contract to have a legal estate conveyed to him to convey or create a legal estate;

and then the definition continues with options, pre-emptions and other like rights.

Now a boundary agreement may constitute a contract to convey land. The parties may agree that in return for a concession by A in one place, straightening the line of division, B will make a concession in another place; and the agreement may thus be one for the conveyance of land. But there is another type of boundary agreement. This does no more than identify on the ground what the documents describe in words or delineate on plans. Nothing is transferred, at any rate consciously; the agreement is to identify and not to convey. In such a case, I do not see how the agreement can be said to constitute a contract to convey land.

In general, I think that a boundary agreement will be presumed to fall into this latter category.

…

There may, of course, be cases in which it is uncertain or doubtful whether a boundary agreement will convey any land. Thus, the configuration of the boundary may suggest that land will be conveyed, without demonstrating this beyond doubt. In such a case I would hold the agreement not registrable. Clause C(iv) applies to a ‘contract … to convey,’ and not to a contract which leaves it uncertain whether or not any land is to be conveyed. In short, in my judgment, a boundary agreement is presumed not to convey land; the presumption may be rebutted, but unless it is, the agreement is not registrable; and to point to circumstances of doubt or uncertainty is not to rebut the presumption.

In this case, the boundary on the conveyance, as I have construed it, coincides with the boundary on the agreement, and so the agreement is not registrable. If the two boundaries had not coincided, because, for example, the true construction of the conveyance yields a different boundary, then the agreement would have been an agreement whereby in fact it was agreed that land belonging to one should thenceforward belong to the other. Nevertheless, even in those circumstances, I should not hold that the agreement was registrable: for, in my judgment, it is not a ‘contract … to convey’ within clause C(iv). A contract merely to demarcate and confirm is not a contract to convey. No doubt the parties cannot go back on this agreement, and each in time will acquire a title by limitation to the land of the other which falls on his side of the agreed boundary. Even if each were to be entitled to demand a conveyance of that land from the other, I doubt whether the agreement would be registrable: for although the obligation to convey would no doubt arise out of their agreement to demarcate, the contract was merely a contract to demarcate and not a contract to convey.”

Megarry J then commenced his conclusion with a passage the opening sentence of which stresses the policy of the law in promoting boundary agreements. This is a passage which has been used as a steer by courts dealing with boundary agreements in many cases since Neilson v Poole. Megarry J held:-

“I must, too, bear in mind that a boundary agreement is, in its nature, an act of peace, quieting strife and averting litigation, and so is to be favoured in the law. I also bear in mind that many boundary agreements are of the most informal nature, and that the penalty of failure to register an estate contract is that the agreement will be void against a purchaser. These more general conditions, I think, support me in the view that I have expressed. In my judgment, no boundary agreement should be held to be registrable unless it can be seen with reasonable clarity to be an agreement to convey. Accordingly, whether or not I am right about the boundary shown by the conveyances, I hold that the boundary agreement is not void against the defendant for want of registration.

I may add that there was a further contention in relation to the boundary agreement by the defendant; and this, too, I reject. This contention was that once the boundary agreement was made, it superseded the conveyance in so far as the conveyance defined the boundary. Yet what the agreement did was merely to establish on the ground, by agreement, what it was that the conveyance showed. A boundary agreement that merely demarcates is, I think, an agreement that is ancillary to the conveyance; it does not supersede it.”

In this case, on the judge’s findings the parties came to an agreement to establish the boundary. However, the judge found that the boundary so established might have encroached on the appellant’s land. He further found in effect that Mr Rigolli had given up or thought he had given up a small triangle of land by the cherry tree. In my judgment, for the reasons given by Megarry J, the agreement was nonetheless an agreement that merely demarcated the boundary. It did not purport to be a contract to convey any land from the appellant to the respondent. The parties’ purpose was to fix the boundary at about the place where they thought it ought to have been.

There was no issue in Neilson v Poole as to whether the boundary agreement had to be in writing by virtue of section 40 of the Law of Property Act 1925, no doubt because under the law as it then stood the doctrine of part performance could have been relied on as obviating the need for writing. Since the decision in Neilson v Poole, section 40 of the Law of Property Act 1925 has been repealed and the doctrine of part performance has been abolished. Section 2 of the Law of Property (Miscellaneous Provisions) Act 1989 now provides instead as follows:-

“2 Contracts for sale etc of land to be made by signed writing

(1)

A contract for the sale or other disposition of an interest in land can only be made in writing and only by incorporating all the terms which the parties have expressly agreed in one document or, where contracts are exchanged, in each. …

(5)

… nothing in this section affects the creation or operation of resulting, implied or constructive trusts.”

We are informed by Mr Hutchings that there is as yet no reported authority on the question whether section 2(1) of the 1989 Act applies to a boundary agreement which “demarcates” the boundary rather than conveys land for the purposes of the classification of boundary agreements enunciated by Megarry J in Neilson v Poole. Megarry J’s approach to the words “contract … to convey” in section 10(1) of the Land Charges Act 1925 was essentially that, for a contract to be a contract to do something, the parties had to have as one of their purposes the intention to do that thing. Similar reasoning in my judgment applies to the words “contract for the sale or other disposition of an interest in land” in section 2 of the 1989 Act. As a matter of ordinary English usage, for a contract to be one “for” selling or disposing of land, it must have been part of the parties’ purposes, or the purposes to be attributed to them, in entering into such a contract that the contract should achieve a sale or other disposition of land. The fact that the effect of their contract is that land or an interest in land is actually conveyed, when that effect was neither foreseen nor intended nor was it something which ought to have been foreseen or intended, is not the acid test. Indeed, it would be a surprising result if section 2 applied merely because the effect of the contract was that an interest in land was transferred even if the parties had no intention to make any such transfer and could not have foreseen or intended that that would be the effect.

In this case, however, Mr Rigolli consciously thought that he was giving up a small triangle of land round the cherry tree. (I am prepared to assume that he was in fact giving up some land and that Mrs Joyce was giving up some land too although on the evidence it would not appear that she or Mr Joyce consciously thought about this at the time of the boundary agreement). Even so, the area of land disposed of by both parties was of a very small amount. It would be unrealistic to require the parties to execute a transfer of the land given up by Mr Rigolli (still less of that unconsciously given up by Mrs Joyce). In both cases the land would also be quite difficult to define without the disproportionate expense of a survey. Further, to make the validity of a boundary agreement dependent on the preparation and execution of a written contract would be contrary to the important public policy in upholding boundary agreements so powerfully identified by Megarry J in Neilson v Poole. In those circumstances, I do not consider that Parliament, which after all enacted section 2 against the background of Neilson v Poole, could have intended section 2 to apply to transfers of land pursuant to boundary agreements of Megarry J’s latter type (“demarcating” agreements) simply because a trivial transfer or transfers of land were consciously involved.

Section 2 applies as much to exchanges of land as to other dispositions of interests in land. Accordingly, in deciding what is trivial, the court should not “net” the transfers by either side but aggregate all the conscious transfers involved of either party, if more than one. It is to be presumed, until the contrary is shown, that any transfer of land effected by a boundary agreement of the demarcating kind is trivial for this purpose.

In reaching my conclusion on the application of section 2(1) in this case, I bear in mind that one of the functions of the court is interpret the legislation that Parliament has enacted. Section 2 has indeed in several respects presented particular challenges to the courts in discharging that function. In interpreting legislation it is a well-established part of the court’s role to see whether a situation has arisen in which the mischief against which Parliament has legislated does not apply or is outweighed by other policy considerations of the law, and if so, whether an acceptable way of dealing with that problem can be found. As Megarry J put it, a boundary agreement is “an act of peace, quieting strife and averting litigation”. If section 2(1) applies where trivial transfers of land are consciously involved, the expense to the parties will also be disproportionate to the value of the land involved. Accordingly, in my judgment, it can in this case properly be concluded that section 2 does not apply to trivial dispositions of land consciously made pursuant to an informal boundary agreement of the “demarcating” kind.

A further approach in this matter would be to consider whether the defence based on proprietary estoppel was made out. In Yaxley v Gotts [2000] Ch.162, this court held that proprietary estoppel involved a constructive trust and that, by virtue of subsection (5) of section 2, section 2 did not apply.

In this case, the judge’s findings at the end of his judgment was that there had been an agreement which was acted upon by the parties. The ingredients of proprietary estoppel were therefore established. I do not consider that the court could give effect to this proprietary estoppel in any way other than giving effect to the boundary agreement. There is no finding that Mr Joyce was in some way mistaken about whether the boundary agreement involved the transfer of the minor portion of land which the judge held became part of 7 Chanton Drive. Furthermore, although Mrs Joyce might have given up a very small portion of her land, she also acquired a small part of 7 Chanton Drive where the boundary was diverted to avoid a cherry tree. If there had been findings to support a case of mistake, the mistake can only have been of a very trivial nature and it could not, contrary to Mr Hutchings’ submissions, displace the equity which arises on the judge’s findings. This could only be satisfied by upholding the boundary agreement between the parties. Accordingly, if, contrary to the view which I have expressed, the boundary agreement was an agreement to convey land, section 2 would, in my judgment, still not apply because the requirements of proprietary estoppel are satisfied and in those circumstances subsection (5) of section 2 prevents subsection (1) of that section from applying.

I accept Mr Smith’s submission that the requirement for consideration for Mr Joyce’s agreement to the boundary agreement found by the judge was given by Mr Rigolli’s promise to be bound by it.

If the judge preferred the content of the evidence of Mr French, rather than simply treating it as the decisive evidence of the existence of a boundary agreement, then on the face of this there was a discrepancy between the evidence of Mr French and that of Mr Rigolli. Mr French’s evidence was that a straight line was fixed which came out at the manhole cover and gullies in the cul-de-sac. However, Mr French also accepted that the fence was diverted to avoid the cherry tree. Mr Rigolli gave evidence that the boundary line ended between the gullies but he deposed to the making of a kink in the fence round the cherry tree. Arguably, the judge relied upon the evidence of Mr French only for his conclusion that in November 2000 Mr Joyce and Mr Rigolli had in fact come to an agreement as to where the boundary should be. He did not necessarily rely on Mr French for the purpose of locating the boundary. The judge’s findings as to the location of the boundary are then in the final extract from his judgment, set out above. The judge there upheld Mr Rigolli’s case that the boundary was fixed along the line of the close-boarded fence which Mr Rigolli erected.

The true explanation is however, in my judgment, that the judge did not perceive there to be any real difference between the evidence of Mr French and that of Mr Rigolli. The substantial issue before the judge was whether the appellant was right in locating the boundary nearer to Mr Rigolli’s bungalow than the fence. The appellant’s case was that the boundary was very different from the boundary asserted by Mr Rigolli. That submission is supported by the maps produced by the appellant at trial and on this appeal. However, we have also seen photographs of the relevant part of the fence. These were available at trial. These suggest that there is little difference between the boundary as described by Mr French in his evidence and that as described by Mr Rigolli in his evidence since the photographs show that, contrary to the appellant’s maps, the fence wraps tightly round the cherry tree which the parties agreed in November 2000 should be on Mrs Joyce’s side of the boundary. Mr Hutchings sought to highlight other discrepancies between the evidence of Mr French and that of Mr Rigolli. However, the judge heard all the witnesses. The action lasted several days. The judge also had the benefit of a site visit. On that occasion he could look at the fence closely and he could examine the boundary as well as the fence to see if the boundary had a kink in it as the fence (constructed in panels) had to have. The judge could see to what extent the fence diverged from the boundary that he found had been agreed. The likelihood is that the judge saw no real difference. The judge was in a superior position to this court when making his findings of fact. It would not be right for the court to interfere with the judge’s findings of facts unless it was satisfied that they were not open to him on the evidence. For the reasons I have given, I am not persuaded that the judge’s findings of fact as to the location of the boundary agreed on in November 2000 were not open to him on the evidence.

Accordingly, I would dismiss this appeal.

I agree that this appeal should be dismissed for the reasons given by Lady Justice Arden. I add to those reasons only in regard to section 2(1) of the Law of Property (Miscellaneous Provisions) Act 1989, which, so far as material, provides:

“A contract for the sale or other disposition of an interest in land can only be made in writing…..”

In Neilson v Poole (1969) 20 P&CR 909, 918, in a judgment which has consistently been followed and approved at first instance and in this court, Megarry J held that an oral agreement between adjoining owners for the definition of the boundary between their respective properties, though it may sometimes constitute a “contract….. to convey…. a legal estate” for the purposes of what is now section 2(4)(iv) of the Land Charges Act 1972, will in general be presumed to do no more than identify on ````````````````the ground what the title deeds describe in words or delineate on plans:

“Nothing is transferred, at any rate consciously; the agreement is to identify and not to convey. In such a case, I do not see how the agreement can be said to constitute a contract to convey land.”

At p. 919 he added:

“If the two boundaries had not coincided, because, for example, the true construction of the conveyance yields a different boundary, then the agreement would have been an agreement whereby in fact it was agreed that land belonging to one should thenceforward belong to the other. Nevertheless, even in those circumstances, I should not hold that the agreement was registrable….. A contract merely to demarcate and confirm is not a contract to convey.”

In my judgment the reasoning of Megarry J in regard to section 2(4)(iv) of the 1972 Act applies equally to section 2(1) of the 1989 Act. A demarcation agreement as described by him is no more a “contract for the disposition of an interest in land” than it is a “contract to convey land”.

The boundary agreement found to have been made in the present case was in the classical mould of Megarry J’s demarcation agreement, subject possibly to these points: first, a very small part of the land in Mrs Joyce’s paper title had been taken by Mr Rigolli; second, Mr Rigolli had given up a small triangle of land beside the cherry tree. These discrepancies did not trouble Judge Hull. He evidently thought that the case was nevertheless covered by Neilson v Poole.

I agree with the judge. There are two ways of looking at it. Either the agreement was one “whereby in fact it was agreed that land belonging to one should thenceforward belong to the other” within the second passage I have read from Megarry J’s judgment or the de minimis principle applies. Either way the agreement was outside section 2(1) of the 1989 Act, and it is unnecessary to rely on section 2(5) and the decision of this court in Yaxley v Gotts [2000] Ch. 162.

I agree.