Andrew Riddell Dunlop v Rostislav Rostislavavitch Romanoff

We have to begin with the principles relating to the construction of contracts in general and of conveyances in particular. The modern starting point is the Supreme Court’s judgment in Wood v Capita Insurance Services Limited [2017] UKSC 24. In paragraphs 10 to 13 it was said that the courts must:

Furthermore:

In Pennock v Hodgson [2010] EWCA Civ 873 Mummery LJ explained the approach to be taken to the construction of conveyances in order to ascertain the extent of the land conveyed:

Accordingly, extrinsic evidence may be considered where the conveyance itself and the physical features on the ground do not clearly indicate what has been conveyed: Nielson v Poole (1969) 20 P & CR 909, Clarke v O’Keefe (1997) 80 P & CR 126, and Ali v Lane [2006] EWCA Civ 1532. Off limits, however, is evidence of the negotiations between the parties; in Investors Compensation Scheme Limited v West Bromwich Building Society [1998] 1 WLR 896 Lord Hoffmann said:

Sometimes it is clear that something has gone wrong in the drafting of a conveyance. Where there is a clear mistake, and it is clear what the correction should be made in order to cure it, then that correction can be made as a matter of construction without the need for an application for rectification: East v Pantiles (Plant Hire) Limited[1982] 2 EGLR 111; Chartbrook Limited v Persimmon Homes Limited [2009] UKHL 38..

Those are the general principles, and they are not in dispute. The issue in the present case is the inconsistency between the words of the conveyance and the plan.

Everyone who has studied unregistered conveyancing knows that where a plan is said by a conveyance to be “for the purposes of identification only” they should look to the wording of the conveyance, and that a more precise description in that wording will prevail over the plan. By contrast if the plan is said to be one on which the land is “more particularly delineated” then it will prevail if the words are unclear. Difficulties arise where the conveyance describes the plan using both of those phrases, or neither, but here it is beyond dispute that the conveyance is said to more particularly delineate the land conveyed (the same phrase is used for each of the three parcels).

The primary authority on the effect of such a plan is Eastwood v Ashton [1915] AC 900. The question was whether a strip of land, about one-twelfth of an acre in size, had been conveyed. The land was said to be “more particularly delineated” on the plan. The words of the conveyance were problematic – either inaccurate or imprecise – but the plan clearly showed that the disputed strip was included. The plan prevailed.

Earl Loreburn, at page 908, said:

Lord Parker of Waddington at page 912 said:

It is important that the words of the conveyance in Easton v Ashton did not provide a clear answer. It is also important that the House of Lords did not say that a plan that more particularly delineates the land will always prevail over the wording of the conveyance where the wording is unclear.

Again, I believe that thus far what I have said is uncontroversial. We now have to look at two difficult cases, both decided at first instance in the High Court.

In Network Rail Infrastructure Limited Mr Nicholas Strauss QC, sitting as a deputy judge of the High Court, had to decide the physical extent of leasehold property. The lease referred to a plan “delineating” the demised premises, and it was not suggested that the omission of the words “more particularly” was significant. The lease was of a parade of shops on a platform built over a railway, and the issue was whether the demised premises included an area beneath the pavement between the front of the shops and road, referred to as the “infill”. It was argued that it did because a previous lease of the same premises had included the infill. Mr Strauss QC went through the authorities, starting from Eastwood v Ashton and noting at paragraph 31 that:

At paragraph 41 the deputy judge said:

Mr Strauss QC decided that the infill was not included; it was not shown on the plan, the exclusion of the infill was not obviously unrealistic or impractical, and that although it might have been a mistake it was not so clearly a mistake that he could apply a corrective interpretation so as to amend or discount the plan.

Network Rail Infrastructure was not a case where the lease plan was in conflict with the words of the lease. The difficulty was that the words were imprecise, and that the previous lease (to the same lessee) had clearly included the infill. But the plan was clear. So the circumstances were very similar to those in Eastwood v Ashton, where only the plan gave clarity.

Mr Maynard for the respondent pointed out that the FTT was bound by the decision in Network Rail Infrastructure Limited, whilst this Tribunal is not. Mr Mills for the appellant argued that Mr Straus QC was wrong to suggest that the verbal description will only prevail over the plan if that plan “is not clear enough to show where the boundary lies” (quoting from the deputy judge’s paragraph 41, quoted above). But that quotation does not properly set out what he said. He said that the plan will normally prevail unless it is insufficiently clear. He did not say that the plan will always prevail unless it is insufficiently clear. And he specifically pointed out (at his paragraph 31, quoted above) that the House of Lords in Eastwood v Ashton did not say that a plan introduced by the words “more particularly described in” will automatically prevail even in a case where the words of the conveyance are unclear.

The proposition that a plan on which land is more particularly delineated “will normally take precedence over a verbal description … unless it is not clear enough to show where the boundary lies” does go further than the House of Lords went in Eastwood v Ashton, insofar as it relates to a situation where the wording of the conveyance is clear. But that seems to me to be a legitimate extension of what the House of Lords said, provided the word “normally” is not forgotten.

The issue in Wesleyvale was the position of a right of way. The conveyance stated that it was granted over a strip of land 35 feet wide. The plan on which it was said to be more particularly delineated showed a much narrower strip. On the ground, there were buildings which meant that for part of the route a 35-foot wide strip was physically impossible. Lewison J pointed out that there was “no interpretation of the conveyance that can give full effect to every part of it”. He was not willing to give effect to the plan alone, because that would have given “no effect at all to the only dimension mentioned in the conveyance”. Instead he said this:

That was an ingenious solution which avoided a direct conflict between the plan and the words of the conveyance. But the fact remains that the plan, despite the words “more particularly delineated”, did not prevail over a clear verbal description; instead, the judge came up with a construction that worked practically without entirely rejecting either the plan or the verbal description.

As I noted above, the judge in the FTT carefully set out the factual background, much of which was not in dispute, and he made important findings of fact about the ownership of the application land at the start of 6 June 1918. He made a careful survey of the authorities relating to the construction of documents, in his paragraphs 40 to 59. He noted what was said in Network Rail Infrastructure about the high standard of proof required before the court can make a “corrective interpretation” of the document. He regarded Wesleyvale as less on point than other cases because it was about an easement, and did not think that he could adopt the solution that Lewison J devised on the basis that it did not extend beyond the “particular case” of an uncertain right of way.

In the judge’s paragraphs 91 and following he set out the reasoning that led to his conclusion.

First, although the conveyance of Westcott was relevant and admissible, the judge took the view that it did not assist Mr Dunlop because it did not answer the question: what was actually conveyed in law to Thomas Dunlop? The antecedent contract had “merged into the conveyance” and so it was the latter that was the key title document. The judge also referred to the 1933 conveyance (paragraph 31 above) and said that the wording of the conveyance was ambiguous as to whether it was the original grant of a right of way or merely conveyed the benefit of an existing easement. In any event he took the view that the conveyance was at best evidence that in 1933 Thomas Dunlop subjectively believed he owned the lane. The judge did not refer to the conveyance of 12 June 1918; in the appeal Mr Mills explained that the register of title to The Hundreds was in the bundle before the FTT, and was referred to in argument.

Second, he held that he could not ignore and discount the “very clear plan” which “more particularly delineated” what was conveyed. From just one look at the plan, the judge explained, it was clear that the application land was excluded. The acreage of 0.389 in the conveyance did not trump the plan. He accepted that that acreage was qualified by the phrase in the parcels clause “or thereabouts”, so that the acreage was not supposed to be definitive but must yield to the plan. He considered that the margin for error imported by “or thereabouts” was sufficient to indicate not just a marginal variation in the total of what was conveyed but also variations or reductions within individual plots. The judge was unmoved by the argument that where the parties intended that part of a plot was conveyed, they said so. He was unimpressed by the argument that the conveyance granted rights of way over land outside the area conveyed where a right was needed, as it certainly would have been if the application land was not conveyed, because the property would (as both counsel accepted) have had a right granted by section 6 of the Conveyancing Act 1881 (now section 62 of the Law of Property Act 1925) if Thomas Dunlop had used it when he was tenant of Lunsford Farm.

At his paragraph 95 the judge said:

Nor did the judge consider that the position was so clear that the conveyance admitted of “corrective interpretation” so that this aspect of the plan could be ignored. He said at paragraph 97 that he considered that the red edging looked very deliberate, and that there might be a number of possible explanations for why it was drawn thus, other than mistake. The parties might have changed their minds at a late stage. Lt Lucas-Shadwell might have decided to retain the application land, either in order to keep it until the local authority adopted it or in order to retain control for the purposes of neighbouring development.

Furthermore, the conveyance plan did not produce an absurd or arbitrary result. The evidence and circumstances were not sufficient for the judge to reject the plan on the basis of a “clear mistake” or on the basis that “something must have gone wrong”.

For those reasons the judge rejected Mr Dunlop’s application on the basis of the 1918 conveyance.

Permission to appeal was granted by this Tribunal on the basis that it was arguable that:

I am going to address the first two grounds together. They are two different aspects of the central question about the construction of the conveyance, which is whether the judge was right in his paragraph 95 to find that the authorities required him to follow the plan in preference to the wording of the conveyance. As will be seen, in my judgment the appellant succeeds on the construction of the conveyance. The third ground refers to the judge’s paragraph 97 (summarised at paragraph 58 above) and is about the related question of “corrective interpretation”, with which I deal briefly under a separate heading below.

Mr Mills, for Mr Dunlop, took as his starting point the judge’s paragraph 95, which bears repeating because it was crucial to his reasoning:

That proposition appears to have persuaded the judge that he could not reject the plan in the present case. But, said Mr Mills, he was wrong about the decisions cited. In particular, Wesleyvale Limited v Harding Homes (East Anglia) Limited [2003] EWHC 2291 (Ch) was such a case, and the judge had been wrong to distinguish it on the basis that it was confined to the context of an uncertain right of way. The same legal principles apply to the interpretation of easements as to the interpretation of conveyances and contracts generally.

Mr Mills also referred to three cases from Hong Kong (to which the FTT was not referred) where a clear depiction on a plan that was said to more particularly delineate the land did not prevail over the verbal description. In Secretary of Justice v Wing Lung Wai Community [1999] 3 HKC 580, the schedule to the conveyance prevailed over the plan because it gave a “clear and unambiguous” measurement of the land and the rent reflected that measurement. The other two cases were Harvest Rise Development Limited v Ling Yau Yung [2002] HKCFI 467 and Lintock Company Limited v AG [1985] HKCFI 310 where again extrinsic evidence led the judge to prefer the verbal description of the land to what was clearly shown on the plan.

Mr Mills further argued that the deputy judge in Network Rail Infrastructure Limited v Freemont Limited [2013] EWHC 1733 (Ch) was wrong to suggest that the verbal description will only prevail over the plan if that plan is not clear enough to show where the boundary lies.

On the basis of those points Mr Mills argued that when interpreted as a whole in the light of the relevant circumstances the 1918 conveyance did convey the application land to Thomas Dunlop. He based his argument on the verbal description, which he characterised as “incredibly precise”; on the plan, which he regarded as unreliable; on the features on the ground, where there is no physical division between the two parts of plot 210 where it is sliced by the red line on the plan; and on the extrinsic evidence, in particular the 1917 sales particulars, the Westcott conveyance, the 1933 conveyance, and a conveyance of 12 June 1918.

Finally Mr Mills relied upon business common sense (see Wood v Capita, paragraph 36 above), because there was no reason for Lt Lucas-Shadwell or Mr Harvey to retain the application land and every reason why Thomas Dunlop would buy it.

Central to Mr Maynard’s argument was the proposition that while the plan is stated to be one on which the land is more particularly delineated, the schedule of acreages is expressly stated to be approximate. The area of the whole farm in the parcels clause is described as “comprising in the whole One hundred and sixty three acres one rood and eight perches or thereabouts”, and there is a further statement of approximation in the heading of the third column in the schedule: “Approximate Acreage”. Indeed, the acreages in the schedule are said to amount to 163.305 in total, whereas 163 acres, 1 rood and 8 perches equals 163.300 acres. This, and the similar discrepancies in the acreage of the green land and the blue land, indicate that “the verbal descriptions by area are not more than approximate.”

Mr Maynard referred to Eastwood v Ashton, and quoted Lord Wrenbury’s words at 920:

Mr Maynard argued that the FTT was bound by the decision in Network Rail Infrastructure Limited and that although the Upper Tribunal is not, it should follow it in any event as a matter of “judicial comity”. The plan is the dominant description. He accepted that none of the authorities is so absolute as to admit of no possible exception to the normal rule that a plan on which land is more particularly delineated is the dominant description, but in this case he said there is no intrinsic feature of the conveyance that takes it out of the norm. The phrase “more particularly delineated” creates an “objective hierarchy of importance” so that the plan prevails.

Mr Maynard also argued that the terms of the main clauses to the conveyance must take priority over a schedule, which he regarded as being a document incorporated by reference. He argued that the OS measurements in the schedule represent “conflicting extrinsic source material inadvertently engrafted into the contract” and should be rejected. As a matter of last resort, where two inconsistent provisions cannot be reconciled any other way, the earlier one prevails over the later, and so the plan prevails over the schedule: Forbes v Git [1922] 1 AC 256 (PC).

Mr Maynard argued that Wesleyvale is clearly distinguishable. The dimension of 35 feet was not qualified by “thereabouts” or “approximately”. And the judge’s solution, to combine the verbal description and the plan, is not available where the issue is a binary one as here, namely whether or not a particular piece of land was conveyed.

Mr Maynard also sought to argue that the judge was wrong to find that the Westcott conveyance was admissible, on the basis that it was evidence of subjective intentions. Strictly he was not able to make that argument since there was no cross-appeal of the judge’s finding on admissibility, nor any cross-appeal from the judge’s finding that the land Thomas Dunlop had contracted to purchase included the application land. Had there been any such cross-appeal it would inevitably have failed. Certainly contractual negotiations are inadmissible, but a conveyance is not a negotiation. It is a deed, from which the parties cannot resile; the parties to the deed could not have been heard to deny the statements of fact in it. The conveyance of Westcott was clearly admissible. It is hard evidence of what the parties did on that date and of the transactions they had already entered into. It can be used as an aid to construction of the conveyance of Lunsford Farm insofar as the Lunsford Farm conveyance when read as a whole is unclear.

By way of clearing the ground I can begin by dismissing three of Mr Maynard’s points.

First, there is no authority for the notion that a schedule to a conveyance is a document incorporated by reference and that the main clauses automatically prevail over it. Conveyances and leases commonly use schedules as a way of organising material; schedules are not documents incorporated by reference and they do not have any less importance than any other part of the document.

Second, the proposition that an earlier provision in a document is to be preferred to a later one is inconsistent with the modern authority that the court must construe the document as a whole (Wood v Capita, see paragraph 35 above).

Third, Mr Maynard relied heavily upon the proposition that the 1918 conveyance created, by the use of the words “more particularly delineated” a hierarchy of importance, and that where the highest element in the hierarchy – here the plan – was clear then there is no reason to resort to extrinsic evidence. Only if there is no “unified truth” in the conveyance and no hierarchy of precedence within it can one move outside the four corners of the conveyance and look at extrinsic evidence. That, again, is inconsistent with authority. The conveyance is to be construed as a whole, and where as here there is an internal inconsistency it is right to look at extrinsic evidence rather than to simply disregard the schedule and its measurements.

It is now time to look again at the 1918 conveyance and ask whether the words or the plan are unclear.

Taking the plan first, I agree with the judge that the plan was clear in this respect. I note what Mr Mills said about inaccuracies in the plan; it was not to a particularly large scale and there are areas where for example the red edging does not accurately follow the line of the road. But it was perfectly clear about the application land: it was not within the red edging.

Turning to the words, it is a curious feature of the case that whereas Mr Mills regards the wording of the conveyance as “incredibly precise”, Mr Maynard regards it as imprecise and unreliable, and as expressly stated to be so.

I agree that there are two occasions in the 1918 conveyance where the apparent precision of the schedule (with its long list of plot numbers, descriptions, and acreage specified to thousandths of an acre) is expressly qualified. First, the parcels clause (quoted above at paragraph 27) described the land as “comprising in the whole One hundred and sixty three acres one rood and eight perches or thereabouts”. Second, the column in the schedule to where the area of each plot is set out is headed “Approximate Acreage”. Such qualifications, Mr Maynard argued, render the dimensions uncertain and subordinate to any precise description such as the plan in this case.

I do not agree that those qualifications have that effect. First, the words “or thereabouts” in the parcels clause refer, in my judgment, to the whole acreage of 163 acres 1 rood and 8 perches. A perch is 0.00625 of an acre and it is easy to see that it must be difficult to be precise to the last perch when measuring a whole farm. The parties therefore added that careful qualification, but I do not take it as an express warning that an individual plot might be significantly misdescribed. Rather, those words account for the difference between 163a 1r 8p, which is equivalent to 163.300 acres, and 163.305 acres which is the total recorded at the foot of the column of acreages in the schedule.

Second, the words “Approximate Acreage” at the head of the column in the schedule have to be read in their context, which is a list of acreages going to three decimal places, and including some very small plots – plot 186 is only 0.113 acres. The application land measures approximately 0.15 acres, almost 40% of the area of plot 210. The heading “Approximate Acreage” might give warning that a measurement might be out by some thousandths, but not that it might be nearly 40% out – that is implausible in a context where the parties were obviously taking great care about numbers. I disagree with the view taken about that by the judge in the FTT, who thought that the words of qualification were sufficient to give warning of this level of discrepancy.

I agree with Mr Mills that the areas conveyed were precisely described in the schedule. Where the parties intended to convey part of an OS parcel they said so, and I take that (again in respectful disagreement with the FTT) to be a strong indication and that where they did not say so they meant the whole parcel.

Accordingly this is not a case where the verbal description was in any way unclear. It is therefore not on all fours with the situation in Eastwood v Ashton. This is a case where there is an inconsistency between clear words and a clear plan. Eastwood v Ashton was not about such a case and therefore cannot determine the outcome of such a case. Nor, as we have seen, was Network Rail Infrastructure about such a case; nevertheless, Mr Strauss QC did say in more general terms that where a plan more particularly delineates the land it “will normally take precedence over a verbal description, and over any physical features of the property, unless it is not clear enough to show where the boundary lies” and that seems to me to be a legitimate extension of what was said in Eastwood v Ashton. Strictly speaking that proposition was obiter because the words of the lease in Network Rail Infrastructure did not answer the question whether the infill was demised, but as a commonsense extension of Eastwood v Ashton it seems to me to be correct.

So, stripped to the bare essentials, the law is that where a plan more particularly delineates the land conveyed and either the wording is unclear or the wording is clear but inconsistent with the plan the plan will generally, or normally, prevail. The judge’s summary of Network Rail Infrastructure at his paragraph 95 (see paragraph 60 above) omitted the crucial word “normally”.

I also agree with Mr Mills that Wesleyvale is not to be distinguished on the basis that it is about an easement; the same principles of construction apply. It is the only authority from England and Wales where there was a conflict between clear words and a clear plan; and the plan did not prevail. That does not in itself provide an answer, but it demonstrates that where there is such a conflict, as there is in the present appeal, the plan does not have to prevail.

Therefore in the present case, where the plan conflicts with clear words, I have to decide where there is anything to take the case outside general rule that the plan prevails.

Two general principles are of great assistance. The first is the principle that the court can look at extrinsic evidence where the conveyance does not tell the full story. In looking for something that takes this case out of the general rule I am not limited to the four corners of the conveyance, as Mr Maynard argued, because to accept that limitation would be to accept that the acreage in the schedule is to be disregarded, which is inconsistent with the proposition that the conveyance is to be construed as a whole.

And in the present case the extrinsic evidence is overwhelmingly in favour of the schedule to the conveyance.

I have already explained why the Westcott conveyance was admissible and relevant. The judge in the FTT thought it was too, but then discounted it on the basis that it “merged into the conveyance”. The principle of the merger of the contract into the conveyance is about the obligations in the contract, which are no longer relevant once the conveyance has been completed. That does not mean that the contract did not happen. In fact the Westcott conveyance gives us two valuable pieces of evidence; one is that Thomas Dunlop had contracted to purchase Lunsford Farm including the application land and was therefore its beneficial owner at the start of 6 June 2018 and entitled to call for a conveyance of it; the other is that none of the parties to that contract had changed their minds about it at the point on 6 June 1918 when the Westcott conveyance was executed. That evidence points overwhelmingly to the schedule prevailing.

So does the conveyance of 12 June 1918. Admittedly we do not have a copy of it, but the register of title in referring to the conveyance, setting out the parties, and recording the grant of a right of way leads inexorably to the conclusion that Thomas Dunlop joined in in order to grant that right. If on 6 June 1918 Lt Lucas-Shadwell and Mr Harvey and Thomas Dunlop did change their minds about the conveyance of the application land to Thomas Dunlop, they had forgotten that 6 days later. That is implausible.

The 1933 conveyance points to the same conclusion. I disagree with the judge about the construction of this conveyance; it was clearly an original grant of a right of way. A conferral of a right to share an easement would have referred to that easement and to its source, and there is no such reference. The only way in which Thomas Dunlop could have had such an easement, if he did not own the application land, would be by the operation of what is now section 62 of the Law of Property Act 1925, then section 6 of the Conveyancing Act 1881, by virtue of his use of the application land when he was tenant of Lunsford Farm. But again if that was the case one would expect the origin of that easement to have been recited in the conveyance. Thomas Dunlop granted an easement over the application land to Beatrice Maud Eves in 1933. Even if that is evidence only of Thomas Dunlop’s subjective understanding it shows that if he and the other parties to the 1918 conveyance did change their minds at some point on 6 June 1918, after the conveyance of Westcott and before the conveyance of Lunsford farm, he had forgotten all about it in 1933. Which is again implausible.

The second general principle is from Wood v Capita (paragraph 36 above): if in doubt, the court is to prefer the construction that makes business common sense. Given that Lt Lucas-Shadwell and Mr Harvey were not free to change their minds on 6 June 1918 and were contractually bound to convey the application land to Thomas Dunlop, it made no business sense for them not to do so.

I conclude that the 1918 conveyance of Lunsford Farm, properly construed as a whole with the benefit of extrinsic evidence, conveyed the application land to Thomas Dunlop.

In case I am wrong about that, and it is not possible to read the conveyance as a whole in that way, then I take the view that a “corrective interpretation” is legitimate. It is clear from the extrinsic evidence, and in particular the conveyance of Westcott, and the conveyances of 12 June 1918 and of 1933, that insofar as the plan to the 1918 conveyance of Lunsford Farm cut out the application land that was a mistake. And it is perfectly obvious that the only correction possible or needed is to read the plan as if the red edging and pink shading included the application land.

The judge did not think that was possible. At his paragraph 97, referred to above at paragraph 61, he said that there were reasons why Lt Lucas-Shadwell might have wished to retain title to the application land. Mr Dunlop has permission to appeal on the ground that the FTT wrongly speculated about other possible explanations for why the plan excluded the lane.

In the appeal Mr Maynard pointed to the deletion and insertion in the parcels clause in the Lunsford Farm conveyance (paragraph 27 above); the original text described the whole property as “Lunsford Farm”, but the amendment applied that name only to the farmhouse. By contrast the Westcott conveyance refers to the property to be purchased by Thomas Dunlop as “Lunsford Farm” (see paragraph 21 above). So something did change, he said, between the two conveyances. In my judgment that minor drafting amendment goes nowhere near to showing that the parties re-negotiated their contractual obligations. In fact it indicates that they were alive to drafting details and that if they had indeed re-negotiated, and decided to exclude the application land, they would have made an amendment to the conveyance – probably to the schedule of acreages – in order to indicate that.

The difficulty with the judge’s reasoning in his paragraph 97 is that it wholly ignored the fact that Lt Lucas-Shadwell was not free to change his mind. The judge himself found that he had contracted to sell the application land to Mr Harvey, and the 1918 conveyance recites that Mr Harvey had paid in full. The judge found as a fact that Mr Harvey had contracted to sub-sell Lunsford Farm, including the application land, to Thomas Dunlop. The judge’s guesses as to why Lt Lucas-Shadwell might have changed his mind go nowhere near to explaining how he could possibly have done so. Accordingly the judge was wrong to reject the possibility of corrective interpretation and insofar as it is necessary I interpret the 1918 conveyance by correcting the plan to include the application land.

The appeal succeeds for the reasons I have given, and I shall direct the chief land registrar to give effect to Mr Dunlop’s application to be registered as proprietor to the application land as if Mr Romanoff’s objection had not been made.