LORD JUSTICE LAWRENCE COLLINS: This is an appeal with the permission of this court against the order of HHJ Hamilton QC made on 10 April 2006 at the conclusion of a trial in the Bromley County Court. The action concerned a dispute between neighbours as to the location of the boundary between the front gardens of their houses, Nos. 10 and 11 Courtney Drive, Beckenham, Kent.

The properties are on an estate developed by the Wates Group. The houses were built in the early 1990s. Mr and Mrs Haycocks, the claimants below and the respondents in this court, bought No. 10 from the developers in 1992. No. 11 was originally owned by a Mr and Mrs Campbell. The houses have open front gardens with no walls, fences or hedges between them. Shortly after the houses were occupied there were some discussions between the Haycocks and the Campbells about the position of the boundary between the properties. Mrs Haycocks’ uncontradicted evidence was that when she and her husband moved into No. 10, they, together with the Campbells, requested the site manager of the developers to provide clarification as to exactly where their respective boundaries lay.

As a result, a plan was prepared by Colin Wykes and Associates, which has been referred to in this appeal as the Wykes Plan, and a copy of the plan was given to the Haycocks and the Campbells. Mrs Haycocks’ evidence was that she and her husband believed that the Campbells were of the view that the Wykes Plan accurately set out the boundary between Nos. 10 and 11. Subsequently the Campbells planted some conifers between the properties. The Haycocks’ case was that the trees were planted along the boundary shown on the Wykes Plan. In 1996 the Campbells replaced the conifers with leylandii.

Mr and Mrs Neville, the defendants below and appellants here, purchased No. 11 in November 2000 from the Campbells and moved into the house in mid-January 2001. From the correspondence it seems that an issue as to the boundary arose almost immediately between the Haycocks and the Nevilles. Mr and Mrs Neville apparently believed at one stage that all of the lawn and the borders at the front of their house formed part of their property. In May 2001 the Haycocks sent the Nevilles a copy of the Wykes Plan. It seems that by July 2001 the Nevilles had already instructed solicitors, since in that month Mr Neville wrote to the Haycocks to say that their solicitor was steadfast that the Wykes Plan was incorrect.

In July 2002 the Haycocks engaged solicitors and through the solicitors suggested that the boundary problem should be solved by jointly employing a surveyor to physically mark up the boundary as shown on the Land Registry plans and suggested Garner and Co as surveyors, but stated that they were willing to consider firms of other surveyors to be put forward by the Nevilles. The Haycocks said they were prepared to be solely responsible for their fees. The Nevilles did not agree to a joint instruction and Mr Garner, the Haycocks’ surveyor, went onto the land on 6 September 2002 to set out the boundary. His report was sent to the Nevilles’ solicitors, who took the position that the report was flawed by the failure of Mr Garner to scale off the length of the boundary from the back part of the Nevilles’ property through to the pivot point as against the dimensions on Plan J, to which I shall refer later.

In December 2002, the Haycocks said that it was imperative that the boundary was established with clear and permanent markings as soon as possible either by an agreed resolution procedure or, if necessary, by court process. In February 2003 the Nevilles’ solicitors said that the Nevilles would not engage in any further correspondence and if the Haycocks wished to take the matter to court they should do so. In May 2003 the Haycocks suggested that there should be a meeting between surveyors instructed by both parties, but this was refused by the Nevilles.

The nature of the dispute is as follows. There has never been any dispute that the boundary runs in a straight line from the point where the back gardens of Nos. 10 and 11 meet the fence of the gardens to the rear of the properties, shown as point A by the experts, to a pivot or hinge point at the front of the respective properties. The boundary then turns at an angle and runs through to Courtney Drive. The dispute concerns the location of the pivot point, i.e. the point at which the boundary between the properties turns at an angle and runs down to the front boundary of the properties adjacent to the road. The only question for the court to decide was where on the ground the pivot point should be located. Once that point was determined, the line of the boundary was not further in dispute and could be marked out. The experts agreed just before trial that the boundary should run at an angle of 45 degrees from the pivot point to the road.

A number of plans were before the judge and before this court. In the first category is what is called Plan J and also the Land Registry Plan. Plan J was prepared by the developers and is stamped “Legal Plan” and is on the scale 1:500. On its face it appears to be drawn originally in 1988, and amendments were made in August and September 1989 before any construction work had been carried out on the development site. Plan J shows a boundary between what are now Nos. 10 and 11. The boundary is a straight line from the rear between the houses, until it reaches the pivot point at what seems to be the south-western corner of the hammerhead of what was to be No. 10’s driveway. It was common ground between the parties that the boundary between the properties was as shown on Plan J; the problem was to ascertain where on the ground the pivot point shown in Plan J in fact fell. Plan J was the plan from which the Land Registry title plans were subsequently drawn. The plan attached to the Land Registry entry is on a scale of 1:1250 and is of no assistance whatever in relation to the present dispute.

Next is the Wykes Plan. As I have said, the Wykes Plan was produced in 1992 following discussions between the Haycocks and the Campbells. A surveyor from Colin Wykes & Associates pegged out the boundary and ran a line between the pegs. The end of that line was marked by a nail in the kerb; subsequently the pegs were removed, but the nail was left. The Wykes Plan is dated May 27, 1992 and is on a scale 1:200. It does not show the pivot point as such, but shows the angle boundary meeting the curvature of the hammerhead of the Haycocks driveway, a short distance northwards from the south-western corner of the hammerhead. It also records that the end of the boundary where it meets Courtney Drive is marked by a nail in the kerb.

The next is Plan K. Plan K is very similar to Plan J but with amendments dated June 5, 1992, after the properties had been built and the driveways laid. A note to Plan K states that the driveway to what is now No. 10 is set out “in accordance with the C Wykes survey”. Plan K shows the pivot slightly north from the edge of the hammerhead, but this may be simply a matter of copying and otherwise it is in all respects, to the naked eye, identical to Plan J. There are also other plans produced by Laser Surveys Limited, who were instructed by both parties to carry out a detailed survey of the properties and their plans have been used by the experts to transpose their measurements.

The experts were Mr Garner for the Haycocks and Mr Stimpson for the Nevilles. Both experts relied upon scaled measurements to arrive at their respective points. Both experts agree that for the exercise of scaling and measuring the appropriate plan to use was Plan J. Mr Garner’s measurements were originally made on September 6, 2002, prior to these proceedings. In his report, Mr Garner said that he had scaled off the position of the pivot point from the 1:1250 plan, but this is almost certainly an error -- he must have meant 1:500 from Plan J -- from the positions of the front walls of the two houses, on the basis that they were the shortest distances from any available fixed reference points. He located the pivot point three metres from the front wall of No. 11, which he said corresponded with the centre of the fourth brick of the edge of the driveway running from the south-western corner of the hammerhead of the driveway. He discovered the nail in the kerb placed at the time of the Wykes Plan and said that it corresponded exactly with the position where he considered the boundary met Courtney Drive. He said that his opinion was supported by the Wykes Plan and Plan J, which appeared to show the pivot point to be located on the outer edge of the driveway to No. 10, even lower than the measurement of the outer edge of the driveway to No. 10 and even lower than his measurement, and therefore even more favourable to the Haycocks.

He used a scaling-off method, but it is plain from his report that he considered that the intention of the original developers had to be gleaned from the relationship of the physical features on Plan J and that if scaling was used, a cautious approach had to be adopted and dimensions used had to be as short as possible. He rejected the approach of merely scaling off 25 metres from point A at the rear of No. 10 and No. 11’s gardens. He considered that this could not have been contemplated by the developers. Mr Garner said that the pivot point coincided with the south-western end of the hammerhead to No. 10’s drive and that he considered that to be a logical position.

Mr Stimpson made his site visit on August 23, 2005. He relied solely upon Plan J and said that the correct way to deduce the position of the pivot point was by scaling off Plan J from the rear of the back garden at point A. He thought that this was the correct approach because Plan J was begun long before the houses were built and the position of the houses could not be relied upon because a survey had shown that the distances from No. 11 and the rear garden fence, and the distance between the rear wall of the house and the garden fence, and the distance between the front corner of the house and the rear garden fence differed by up to a metre. The pivot point identified by Mr Stimpson did not end at any physical feature such as the hammerhead or any of its sides, but in an area of grass to the north of trees planted on number 11’s land.

The judge’s decision was as follows. Plan J was the plan from which an attempt was to be made to find the boundary and that Plan had to be taken with any other helpful evidence. Mr Stimpson's method of measurement from point A was preferable to measuring from the front of No. 11, since point A could be defined when Plan J was drawn up. Because the house was not built exactly on Plan J, measuring from the house (as Mr Garner did) could not be accurate. On Plan J the intention was to put the pivot point on the south-western corner of the hammerhead at No. 10, to ensure that the driveway and hammerhead fell entirely within the ownership of number 10, but on either surveyor’s estimate of the pivot point a small corner of the Haycocks’ hammerhead would inevitably fall within the land owned by the Nevilles.

After construction the pivot point is shown on Plan K, which identified the drive actually built on No. 10, was set out in accordance with the Wykes survey, where the pivot point was only slightly removed from the position on Plan J and still coincided with a part of the hammerhead to No. 10’s drive. Mr Stimpson’s pivot point was totally reliant on scaling off and ignored the representation on Plan J, Plan K and the Wykes Plan, as to where the pivot point coincided with or was intended to coincide with the hammerhead of No. 10’s drive.

When Plan J was drawn up it was a development plot with no features and there were therefore no topographical features to help with the delineation of the front boundary between Nos. 10 and 11, but when Plan K was drawn from Plan J, or over Plan J or on Plan J, the drive of the houses were on site, and the position of the drive relative to No. 10 was taken from the plan specifically drawn up by the developer to define the boundary between Nos. 10 and 11, which was accepted by the owners of Nos. 10 and 11, until the Nevilles purchased No. 11 in 2000. On Plan K and the Wykes Plan, the pivot point touched the hammerhead, albeit not at the south-west corner, but at a short distance to the north of it.

The judge accepted that the trees planted by the Campbells were intended to follow the Wykes Plan line. She found that the topographical features, namely the hammerhead and the trees planted by the Campbells, although coming after the drawing up of Plan J, were helpful in ascertaining the boundary, since the boundary exercise had been carried out to the satisfaction of Nos. 10 and 11 and the site developer in 1992, and was identified on the ground, but the boundary on the Wykes Plan should not be ignored and gave an indication where the boundary was to be found on the developed site.

The Nevilles were not bound by any earlier agreement and were entitled to rely upon the deeds for the position of the boundary, but the position on the deeds and measurements there from were questionable in view of the problems with Plan J. Her conclusion was that the point, as shown by Mr Garner, was a little too far south of the pivot point on the Wykes Plan and Plan K derived from it, and ran through three trees, which on the evidence of Mrs Haycocks were on the land of No. 11. The pivot point was closer to the point, as found by Mr Garner, but not by virtue of scaling from the front of No. 11. From Plans J, K, the Wykes Plan, the later constructed hammerhead and the tree planting, the pivot point should be at a point which could be derived from the Wykes Plan and Plan K and would coincide with a part of the hammerhead, although not exactly at the corner. It would be closer to Mr Garner’s point and the red line, than it would be to Mr Stimpson's purely scaled line to his point.

She concluded that the surveyors should therefore attempt to take from the Wykes Plan where the pivot point was shown on it and place it on one of the laser plans. The case was adjourned to allow the experts to plot the position of the pivot point. In fact there is no pivot point as such on the Wykes Plan and the experts therefore interpreted the judge’s direction to mean that they were to extend the common boundary line along the fencing to a point where it intersected the curvature of the hammerhead. The pivot point was plotted by the experts as Point J, which was about 80 mm from Mr Garner’s point. The judge accordingly declared Point J to be the location of the pivot point and ordered the Nevilles to pay the costs of the proceedings and £10,000 on account.

The photographs illustrating the effects of the determination show that the effect of using that method is almost the same as Mr Garner’s. The effect is that a strip of land, which is at no point wider than about a meter, is on the Haycocks’ property rather than the Nevilles’ property.

The Nevilles’ position on the appeal can be summarised as follows:

I come to my conclusions. As Lord Hoffmann said in Alan Wibberley Building Ltd v Insley [1999] 1 WLR 894 at 895:

This is just such a case. As HHJ Hamilton QC said, this case arises out of an extremely minor boundary dispute, on which nothing turns and the existence of which would make the sale of either property fraught with difficulty. The exact position of the boundary makes no real difference to either neighbour’s enjoyment of their land.

There are some general points to be made. First, the initial step in the event of a boundary dispute is to look at the title documents, but as Lord Hoffmann pointed out (at 896), plans attached to a conveyance may not necessarily delineate precise boundaries and “the scale is often so small and the lines marking the boundaries so thick as to be useless for any purpose except general identification”. If it becomes necessary to establish the exact boundary, the deeds or maps and plans in the Land Registry will almost invariably have to be supplemented by such inferences as maybe drawn from topographical features which existed or may be supposed to have existed when the conveyances were executed.

Secondly, an agreement to demarcate an unclear boundary is binding on the parties and binds successors entitled without the need for a written agreement: Neilson v Poole (1969) 20 P&CR 909, recently applied in Joyce v Rigolli [2004] EWCA Civ 79. The evidence was that the Wykes Plan had been drawn up to clarify the boundary following discussions between the Haycocks and the Campbells and was drawn at a scale of 1:200. If the Haycocks had pleaded and proved that they and the Campbells had agreed in 1992 that the Wykes Plan represented the boundary, then that would have been binding not only on them but on successors in title, such as the Nevilles.

Third, it was accepted that the solutions for which the experts contended, the Wykes Plan line and the line determined by the judge’s direction were all within the tolerance of the line on Plan J.

Fourth, in exceptional circumstances a judge is entitled to reject expert evidence on both sides, but the present case was not a case where the judge rejected the expert evidence on both sides and substituted her own expertise. The experts both proceeded on the assumption that the task they had was to scale off from Plan J and they differed only in the methodology used for that exercise. It was for the judge to determine the boundary on the totality of the evidence including Plan J, Plan K, the Wykes Plan, the evidence of the circumstances in which the Wykes Plan was made and the evidence of the location of the houses. She was not bound to start from the same point as the experts, namely that the right method to determine the boundary is to scale off from Plan J.

The judge was entitled to reject the sole use of Plan J, if she gave sufficient reasons for taking such a course and if such a decision flowed from the evidence before her. In Richardson v Macnab , 20 April 1999 (unreported), this court had a rather similar case in which it was not possible to deduce the line of the boundary from the conveyancing documents. In that case the boundary area in dispute was never more than 2 ½ feet wide and in some places less, and the dispute in the words of Buxton LJ, excited “in the outside observer feelings bordering on despair”, in some respects similar to this case. The Court of Appeal in that case decided that the judge was entitled to reject the methodology of both experts and to substitute his own line of reasoning provided that his conclusion properly flowed from the evidence, which in that case it did not.

In the present case the judge rightly took Plan J, which formed the basis of the Land Registry Plan, as the starting point. She also rightly said that Plan J was the plan from which an attempt was to be made to find the boundary and the plan must be taken with any other helpful evidence. She then preferred Mr Stimpson’s method of measurement from point A to Mr Garner’s measurement from the front of No. 11, since point A, and not the position on the house, was a fixed point when Plan J was drawn up and because the house was not built exactly as shown on Plan J. She decided, however, that she could not place sole reliance on Plan J because of the scale of 1:500, the lack of dimensions on the plan, the fact that it was not an original, but had been photocopied many times and the fact that a line of 2 millimetres on the plan corresponded to 1 metre on the ground.

It was argued for the Nevilles that it was impermissible of the judge to take account of events subsequent to the conveyance to determine the boundary. In Watcham v Attorney General of East Africa Protectorate [1919] AC 533, the Privy Council decided that conduct subsequent to a transfer of land could be used to demonstrate that one of two inconsistent descriptions of the property in a conveyance was an error and that the true construction of the document was in the sense of the other description. In Beale v Harvey [2003] EWCA Civ 1883, [2004] 2 P&CR 18, this court, speaking through Peter Gibson LJ, declined to follow that decision on the basis that it was inconsistent with the modern authorities to the effect that subsequent conduct is not an admissible guide to interpretation. Beale v Harvey was a case where a wall and fence had been placed beyond the boundary shown on the plan and it is not surprising that, in the absence of material on which to base a case for rectification, it was held that the true intention of the parties to the transfer as to what they intended the boundary to be should not be ascertained by reference to what they did after the transfer at a time when they did not know that the wall and fence had been wrongly positioned.

In Ali v Lane [2006] EWCA Civ 1532, this court distinguished Beale v Harvey and applied Watcham v Attorney General of East Africa Protectorate in a case like this one, where there was uncertainty about the position of the boundary. The court held that in the context of the conveyance of land where the information contained in the conveyance is unclear or ambiguous, it is permissible to have regard to the evidence of subsequent conduct, subject always to that evidence being of probative value in determining what the parties intended. I find no fault in the judge’s reasoning that the topographical features, namely the hammerhead and the trees planted by the Campbells, although coming after the drawing up of Plan J, were helpful in ascertaining the boundary, since the boundary exercise had been carried out to the satisfaction of Nos. 10 and 11 and the site developer in 1992 and the boundary was identified on the ground. Accordingly, the boundary on the Wykes Plan should not be ignored and gave an indication of where the boundary was to be found on the developed site.

I am therefore satisfied that the judge was entitled to reject both Mr Garner’s point and Mr Stimpson's point as the position of the pivot point and that the finding on the intermediate point was open to her on the evidence. I agree with the judge’s conclusion that any solution in this case would be imperfect and that it was a case which should never have come to court. It should have been dealt with, as suggested by the Haycocks, by jointly appointing a surveyor to provide the best possible determination.

I would therefore dismiss the appeal.

LORD JUSTICE WALLER: I agree.

MR JUSTICE CHARLES: I also agree.