The appellant seeks rectification of a claimed inaccuracy in the registered title of his property and appeals against the refusal of the Keeper to rectify. The problem arises in the delineation of the boundary between the appellant’s land and that of the interested party, who does not have a registered title but opposes rectification of the appellant’s registered title. It is agreed that the position of the “true” boundary depends on interpretation of a boundary created in a 1962 disposition to the interested party’s predecessor in title. Although the disputed area is a relatively narrow strip of ground, it is of particular significance to the appellant because of the close proximity of the boundary in his registered title to a house which he has recently built on his land. The primary dispute is whether the correct interpretation of the boundary depends on identifying on the ground the position of a track shown on a 1911 Ordnance map referred to in the description of the land conveyed in that 1962 disposition. That is the appellants’ position. If he is correct on that matter, the interested party accepts that the Land Register plan is inaccurate. In that event, however, there would still be some dispute, on the evidence, as to the exact extent of the inaccuracy and accordingly as to the position of the boundary to be shown in any rectified Land Register Title Plan.
 In the event of inaccuracy being established, the interested party also claims the protection of section 9(3) of the Land Registration (Scotland) Act 1979 as a “proprietor in possession”, but consideration of that claim, if it arises, was deferred, so that this hearing only considered the claim of inaccuracy. The Keeper, apart from lodging written submissions helpfully assisting in identification of the issues, took no part in this hearing.
 In summary, the Tribunal has not accepted the appellant’s primary argument and has accordingly refused this appeal. If we had accepted that there was inaccuracy, we would have put the case out for further consideration of the extent of the inaccuracy.
 The lands and farm of Nethanfoot, Lanarkshire, were conveyed in a Disposition by the Rt. Hon. The Earl of Home in favour of Mrs Agnes Mitchell or Gilchrist and Miss Mary Mitchell recorded on 27 April 1923. The foundation of the appellant’s title was a Disposition by Mrs Annie Gilchrist in favour of the appellant and Mrs Janet Welsh recorded on 5 June 1980. That disposition conveyed the subjects of the 1923 Disposition under exception of several subjects previously conveyed in a number of dispositions. In particular, the subjects conveyed in a Disposition by Thomas Gilchrist in favour of William Jackson (Carluke) Limited recorded on 27 December 1962 (“the Jackson disposition”), the foundation of the interested party’s title, were excluded.
 The Jackson disposition conveyed some 24 fields or enclosures, or parts thereof, extending to 143.564 acres, including, as relates to this dispute:-
“fields and enclosures marked Numbers … fifty nine (part) (59 part) [and] fifty eight a (part) (58a part) … on the Ordnance Survey Map (Scale Twenty five inches and three hundred and forty four decimal or one thousandth parts of an inch to one mile) Edition 1911 Lanarkshire Sheet XXIV, 8; all as the said plot or area of ground hereby disponed is shown coloured … blue … and outlined within the boundaries coloured red on the Plan thereof being an Excerpt from said Ordnance Survey Sheet annexed and subscribed as relative hereto”.
 The appellant subsequently acquired title of the subjects in the 1980 disposition as sole owner, his title now being registered as LAN140617. It is agreed, and of no significance to the issue, that the 1980 disposition (which could not convey land previously disponed in the Jackson disposition) shows the disputed boundary well to the north of the boundary as established by the Jackson disposition, i.e. well into the interested party’s property, but this clear mistake did not find its way into the appellant’s registered title and has no bearing on this appeal.
 The appellant claims that the true boundary of the property conveyed in 1980, being the boundary established by the Jackson disposition, is to the north-west, by a distance varying up to around 6.5 metres, of the north-west boundary on the Title Plan on the title sheet LAN140617. The issue is whether, as a matter of interpretation of the Jackson disposition, there is any inaccuracy in the Title Plan of LAN 140617 at this boundary, which also forms the south-east boundary of this part of the interested party’s property. The plan attached to the Jackson disposition is not a copy of an Ordnance map and is on a scale of 1:5000 (the Ordnance map being 1:2500). Boundaries were drawn on the plan at a thickness representing 4 to 10 metres on the ground. The appellant’s contention is that the correct interpretation of the Jackson disposition is that it delineates this boundary under reference to the position shown on the 1911 Ordnance map of a track running along the boundary between enclosures 58a and 59 (the position of the track apparently having changed slightly over the years). Although the Keeper’s approach was not specifically founded upon or explored in any detail, it had evidently been different. The appellant produced an ‘Appeal Plan’ showing what, on his interpretation of the disposition, he claimed to be the true boundary in comparison with the Register Title Plan.
 The interested party contends that the appellant has failed to establish inaccuracy. In particular, the interested party does not accept the appellant’s contention as to the correct interpretation of the Jackson disposition. In the event of the appellant’s contention being upheld, the interested party would accept that there is some inaccuracy but, on the evidence, would not accept that the true boundary is as shown on the ‘Appeal Plan’.
 The issue of inaccuracy was considered at an oral hearing at which the appellant was represented by Mr Upton, Advocate, instructed by Gillespie Macandrew WS, Edinburgh. He led evidence from Ian Jillott BSc FRICS, Lanark, who had been instructed to prepare a report and plans showing the correct boundary. The respondent did not appear at this hearing. The interested party was represented by Ms Ross, Advocate, instructed by Morton Fraser, Solicitors, Edinburgh. She led evidence from Robin Steel, MRICS, of C & D Property Services, Carlisle.
 Both sides lodged productions, which included a number of acetates enabling maps of the same scale to be superimposed over each other in the course of the evidence. Mr Upton objected to certain detailed evidence on behalf of the interested party, on the ground of lack of fair notice. The interested party’s pleadings and Note of Arguments had taken the line that it was not possible to determine the position of the disputed boundary with any certainty, as opposed to advancing any positive case on the actual position of the boundary. We admitted the evidence and have taken the view that the appellant’s position on the primary argument was not prejudiced but that, as discussed below this would affect the position on the secondary argument, if it arose, so that it would not be appropriate to determine the secondary argument without giving the appellants an opportunity to address the evidence further.
 In the particular circumstances of this case, the Tribunal did not consider that it would derive any assistance from inspecting the locus, it being clear that there had been considerable changes on the ground since 1962 and earlier.
Investors Compensation Scheme Ltd v WestBromwichBuilding Society  1 WLR 896
David Runciman & Sons v Scottish Borders Council 2003 SLT 1405
PMP Plus Ltd v Keeper of Registers of Scotland 2009 SLT (Lands Tr) 2
Land Registration (Scotland) Act 1979, ss 9, 25
Registration of Title Practice Book, 4.27
Rankine, Land Ownership, 4th Ed’n, 104
Halliday, Conveyancing Law and Practice, 2nd Ed’n, vol 2, Ch 33
Gretton & Reid, Conveyancing, 3rd Ed’n, 12–19, 25
Lewison, Interpretation of Contracts, 3rd Ed’n, 7–16
 On the basis of the evidence and submissions, we find the following facts established.
 The dispute concerns the boundary between a building plot, to the south-east, owned by the appellant and on which he has erected a dwellinghouse, and an undeveloped irregular wedge shaped area, to the north-west, owned by the interested party. The wedge shaped area provides access from Lanark Road, Crossford, at its north-east, narrow, end, where an old track leads off the road between No 4 and No 6, to farmland owned by the interested party at its south-west, broader, end. The area to the north-west of the wedge shaped area now includes three relatively modern houses, the nearest being ‘Craignethan Lodge’.
 The 1911 Ordnance map referred to in the Jackson disposition showed an apparently straight track running from the same point between Nos 4 and 6 Lanark Road to a point at or around the south-west end of the disputed boundary. The north-west side of the track was shown with a continuous line, separating fields 58a and 58 to the north-west. The south-east side was shown by a dotted line. A ‘tie’ showed the track to be part of field 59 which, apart from the houses Nos 6 and 8 at the road end, was shown undeveloped beside the track. The track shown extended approximately 96 metres. A line across the track, some 67 metres from the road end, i.e. about two-thirds of the way along it, may be taken as having shown a gate. Field 58, nearer the road, was shown as including the House No 4 but otherwise undeveloped. Field 58a had the word ‘Reservoir’ written on it, with some tiny markings which might have been huts or the like towards the south-west end.
 These areas were all within the lands and farm of Nethanfoot conveyed in the disposition of 1923 by the Earl of Home in favour of Mrs Agnes Mitchell or Gilchrist and Miss Mary Mitchell.
 Between 1911 and 1962 the line of the track apparently changed. The reservoir disappeared. Extensive nursery glasshouses occupying parts of fields 58 and 58a were built, along with a brick wall abutting the south-east side of the glasshouses and running across these two former fields, to the north of the track but converging with it at the corner of No 4 Lanark Road. Near the middle of the track, on its south-east side, a rectangular barn was built.
 Parts of the subjects of the 1923 disposition were conveyed away at various dates. Ground extending to 2.260 acres, immediately to the north-west of the wedge shaped area and including the glasshouses, was conveyed in a Disposition by Thomas Mitchell Gilchrist in favour of Messrs A & T Gracie on various dates in June and July and recorded on 23 August 1962 (“the Gracie disposition”). This ground is particularly described in a bounding verbal description, with full measurements. The ground so described is also shown on a detailed and apparently accurate plan annexed and subscribed. That plan is at a scale of 1:2500 and is apparently based on the 1911 Ordnance map with additional detail showing the glasshouses, wall and barn.
 The Gracie disposition also makes access provisions in relation to an irregular wedge shaped area, which is also shown, apparently accurately delineated and shaded, on the Gracie disposition plan. There is no bounding description of this area. The north-western boundary of this area is most of the south-eastern boundary of the ground disponed. This boundary runs along the wall described above and the side, south-east wall of No 4 Lanark Road. The south-easterly boundary of this area starts along the side, north-west wall of No 6 and continues in a straight line, shown first as a dotted line and then as the north-west wall of the barn, in a south-westerly direction, for approximately 70 metres. Then, at the end of the north-west wall of the barn, it takes a right-angle turn to the north-west, along which it extends around 3 metres, shown as a solid line, before turning back to run, shown as a solid line, in a south-westerly direction for approximately 20 metres. At the other (eastmost) end of the north-west wall of the barn, a solid line runs at right angles in a north-westerly direction across the wedge shaped area towards the glasshouses and meets the wall on the north-west boundary of the ground disponed.
 The Jackson disposition, later in 1962, was another partial disposition by Thomas Mitchell Gilchrist of the subjects of the 1923 disposition. This was a disposition of some 143 acres of farmland described, as narrated at Para  above, under reference to various field or enclosure numbers on the 1911 Ordnance map. Reference was made to the plan annexed, described as “an Excerpt” from the 1911 map. This was, however, to a larger scale, of 1:5000, and was not a precise copy. It showed field numbers corresponding to those on the 1911 map. Only a few of the features on the ground are shown. The boundary was drawn roughly and approximately in a thick line around the outer edges, in varying degrees of thickness representing (allowing also for the margin of error) between 4 and 10 metres on the ground. The wedge shaped area leading to the gap between Nos 4 and 6 Lanark Road is part of the subjects disponed. The wedge shaped area is also the subject of particular access provisions in the Jackson disposition, and is accordingly, although contiguous with other areas disponed on its west side, also separately outlined and shaded on this plan. ‘58a’ is marked within this area. ‘59’ is marked immediately outside the disputed boundary, to its east. There is a clearly discernible “kink” approximately two-thirds of the way down the boundary line, apparently corresponding approximately with both the solid line across the track shown in the 1911 map and the approximately 3 metre stretch of boundary, at right angles, shown on the plan attached to the Gracie disposition. Parties were agreed that no track is shown at this particular part of the Jackson disposition plan. Nor does it show any buildings such as the barn or the houses at the road end. This wedge shaped area on the Jackson disposition plan, now the property of the interested party, appears to correspond with the wedge shaped area which is the subject of access provisions in the Gracie disposition and is shown on the detailed plan annexed to that disposition.
 There is no apparent indication on the Jackson disposition plan of any area which could correspond with any part of field 59 on the 1911 map except at the line extending in a north-easterly direction from the “kink” to the road end.
 The parts of the fields 58a and 59 on the 1911 Ordnance map disponed by the Jackson disposition could not be determined by reference to the 1911 Ordnance map. The south-eastern boundary of the wedge shaped area could correspond with the track shown on that map, with about two-thirds, up to the “kink”, disponed and the remaining one-third retained. The north-west and western boundaries do not correspond with any line on that map.
 Ordnance Survey Map NS8247, revised in December 1963, shows, as well as the houses Nos 4, 6 and 8 Lanark Road, the glasshouses and solid lines corresponding with the boundaries of the land disponed in the Gracie disposition, with the exception of a small stretch at the east corner of that land. It also shows the barn and a track shown with two dotted lines until the end of the north-west wall of the barn, where a line crosses the track, after which the south-east side of the track is a dotted line and the north-west side a solid line. A solid line appears approximately to coincide with the westmost boundary of the wedge shaped area shown on the Gracie disposition plan. The field or enclosure numbers and areas are completely different from those on the 1911 map.
 The Register Title Plan, LAN140617, extracted from the current Ordnance Survey sheet NS8246, shows the north-west boundary of the appellant’s land running from its northernmost point, beside the garden of No 6 Lanark Road, in a south-westerly direction, first along the south-east side of a track which is shown with dotted lines, then crossing the track and continuing on its north-west side to the end of that boundary. The south-east boundary of the land disponed in the Gracie disposition is shown as a solid line, apparently representing the wall which remains. The more modern houses within that land are shown. Nos 4, 6 and 8 Lanark Road are shown with No 6 now shown as having a garden wall running along the track before it meets the appellant’s land. No building is shown on either the appellant’s land or the wedge shaped area. The track is apparently shown as now curving very slightly to the north-west around the end of the garden wall of No 6, rather than running in a straight line as apparently shown on the previous maps.
 Mr Jillott is a chartered surveyor with 33 years’ experience in a range of commercial and residential property matters. He was instructed on the appellant’s behalf to prepare a report and plans showing the correct boundary on the basis of the Jackson disposition. As well as the disposition plan, which he noted was not an exact copy of the OS map and also at a reduced scale, he worked with a certified copy of the 1911 OS map referred to in the disposition, in order to ascertain the boundary and compare it with the Register Title Plan. He considered that there was a clear inference that the boundary followed the line of the track shown on the 1911 OS map and formed the dogleg as shown on the deed plan. The dogleg was formed across the width of the track specifically to show where the boundary lay. Due to the reduced scale of the deed plan and the thickness of the line, the exact point at which the turn occurred was obscure but the plan could be used to determine the proportion of the length of the line at which the turn occurred. This was 70.7% of its length from the northeastern end of the track, which measured 96 metres on the OS map so that the turn occurred 67.9 metres down the track. It was clear from the OS map that the only defining feature was the track itself. Mr Jillott prepared plans based on photographic enlargement of the three plans. He checked the plans so far as possible against fixed reference points on the 1911 map and on site, the northwestern edge of the track being an extension of the southeastern wall of No 4 Lanark Road. He checked the bearing on the OS map with a compass bearing of the alignment of the northwest edge of the track on the site and satisfied himself that these were the same. He measured the width of the track as shown on the 1911 map, at its north-eastern end at 4 metres and at its south-western end at 3 metres. He thus tracked the line of the track as shown on the 1911 map against the Register Title Plan, showing the latter to diverge from the former by an increasing distance, up to a maximum of approximately 6.5 metres. He demonstrated the divergence with an acetate overlay, and also had had a more precise ‘appeal plan’, showing the area of the appellant’s land claimed to have been omitted from the Register Title Plan, prepared by a land surveyor. His acetate evidence used enlarged, i.e. smaller scale, plans and maps at a scale of 1:500.
 Mr Jillott derived a measure of support from certain aerial photographs in 1954, apparently showing the track running “more or less” in a straight line and with the same bearing (229 degrees) as on the 1911 map, and 1975, when it appeared to curve slightly to the north-west as on the Register Title Plan. He accepted that there were limitations in this line of evidence. He had not used the Gracie disposition plan, which he said was not based on the 1911 map but was generally quite similar, with the line of the track straight.
 Under cross-examination, Mr Jillott maintained that the references in the Jackson disposition to the 1911 map meant that features on that map, in particular the track, were part of what was being referred to and thus encompassed in the disposition although not specifically referred to or marked on the disposition plan. He would conclude from the 1911 map that the track was an agricultural, unmade track. He had worked from only the 1911 map, not any later versions. The Jackson disposition plan may have been created by tracing the 1911 map and then reducing and adding some text such as the field numbers. It might have been based on an estate plan. He agreed that the surveyor would have been likely to visit the site to see what the ground was like and it would have been highly unusual simply to draw on the basis of the 1911 map, but said that ascertaining what was on the ground in 1962 was a much less certain and reliable way of interpreting the intention of the draftsman, who had referred only to the 1911 map. There was nothing on the deed plan to indicate that it was based on any survey. Reference to the actual track at the time would always leave greater potential for dispute at a later date. Mr Jillott agreed that when overlaying acetates, a larger scale, giving more reference points, was generally more satisfactory.
 Mr Steel is a chartered surveyor with around 30 years’ experience and since 1988 has worked for a property services company in rural practice including valuations of farms and development sites and preparation of deed plans. He initially prepared a report simply determining the disputed boundary under reference only to the appellant’s Land Register Title Plan (a matter not, in itself, in dispute in these proceedings). In oral evidence, however, he addressed the interpretation of the Jackson disposition plan. He said it was not a copy of the 1911 map, being not so clear “by a long way”. In his view the draftsman was aware of the intention that the boundary was to correspond with the track, but it could not be concluded that the intention was to fix the boundary under reference to the 1911 map. The 1911 map might have been the latest OS plan. From the plan, there was not much to indicate what surveying of the site had taken place. He thought that the line marking the boundaries was quite inaccurate. The draftsman would normally go to permanent features, such as buildings, stone dykes and hedgerows, to show a boundary on a new plan. The position of a track with no features could vary, and if the contract was based on a track the draftsman would have to plot that to some permanent feature, rather than relying on an old map. Even a bottomed track could move or be widened.
 Mr Steel spoke to his understanding of the updating of OS maps. He said that during the 1940s a new survey using the national grid and metric measurements was set up and by the 1960s most populated areas, including this one, should have had new published maps. He then used acetates comparing the December 1963 revision with the 1911 map and the Register Title Plan. He explained a slight discrepancy in the features shown in his acetate, which was actually (apparently) of a 1962 version of the map. The track, he said, was 3 to 4 metres south on the 1963 map of its position on the 1911 map. The Keeper’s plan could be seen to have been based on the position shown in the 1963 map, with the glasshouses replaced by the three modern houses and the house at No 6 Lanark Road either replaced or extended and now showing a large fenced off garden area on the west side. On visiting the site, he had observed the track as a hard track between Nos 4 and 6, but the north-west corner of this garden area had pushed the track back a bit to the north and the track had disappeared after a gate just to the west of the garden of No 6.
 In cross-examination, Mr Steel agreed that there was no indication on the Jackson disposition plan of any survey on the site, and that if that had not been done the next best thing would have been to refer to an existing map, but it depended how old the map was. Accepting that there would be less room for dispute if reference was made to a surveyed scaled map such as the 1911 map, he said that it was not always clear where the track had been in 1911. He understood that the boundaries shown on the 1963 revision would have been plotted in 1962 or 1963.
 Mr Upton, speaking to a helpful written note of submissions, concentrated on the correct legal construction of the Jackson disposition. Under reference to the statutory positions and PMP Plus Ltd v The Keeper, he submitted that the Tribunal was to consider whether the Keeper’s decision was correct in fact and law rather than review whether he was entitled to make it. He submitted that considering all relevant matters of fact or law, the Tribunal should find there to be inaccuracy in the entry on the Title Plan in respect that the disputed boundary did not follow the line shown on the ‘Appeal Plan’. The terms of the Jackson disposition did not warrant the boundary position shown by the Keeper. He referred to the ‘Hoffmann’ approach of looking at the known factual background, the question, once the relevant facts were found, being a question of law. It was partly by looking at the facts found in 1962, i.e. the position of the track then, that one knew that that was not what the plan was seeking to achieve. Where there was ambiguity, uncertainty or difficulty in interpreting the Jackson disposition, and where the 1911 map provided material sufficient to answer the question, there was no need to pursue the inquiry further. The plan was to be construed in the light of the line of the track marked on the 1911 map (which might be the same as the actual position of the track in 1911). The alternative guide was the physical position of the track on the ground in 1962.
 On the law, Mr Upton submitted that verbal descriptions and plans were to be read together; in case of inconsistency, the text prevailed unless the plan was stated to be taxative; the text and plan would be construed so as to be consistent with one another, as far as possible; the general principles of construing conveyances were the same as for other contracts and commercial documents; and where two constructions were possible, the more reasonable, practical and convenient was to be preferred.
 The breadth of the line on the Jackson disposition plan was such that for practical purposes it could not be used without some measure of interpretation. There was every reason to consider that the lines which form the disputed boundary were informed by and derived from the 1911 map and no evidence from which to conclude that they were derived from examination of what was physically present at this location. The fact that the northern boundary of the blue area had been adjusted to reflect the presence of the glasshouse conveyed to Gracie did not entail the conclusion that the boundaries of the blue area reflected a survey of the subjects. Whoever prepared the Gracie disposition plan, Mr Gilchrist had the means of knowing its contents when he executed the Jackson disposition. The existence of the Gracie disposition was acknowledged in two passages in the Jackson disposition. Accordingly it was unnecessary to hypothesise that this plan reflected a site survey or even a visit. The absence of features which had been shown on the Gracie plan and on the 1962 map used by Mr Steel, was more consistent with the lines having been derived from the 1911 map: if he was reflecting what was then physically present rather than using, as he expressly acknowledged, the 1911 map, why did the author of the plan define the western boundary by a line at a different angle and ignore the line shown on the Gracie plan and the 1962 map? More fundamentally, the 1962 map showed that the field or areas in the 1911 map had ceased to have any reality on the ground, yet the author of the plan had stuck to them. There was nothing to support the hypothesis that the author of the plan was informed by a site survey (except the western boundary of the blue area, an anomalous, neutral feature), and no basis for selecting this boundary as having received special treatment. The position could be explained entirely by two documents whose existence was acknowledged within the deed. The respondent, on the other hand, was using extraneous evidence of the physical reality to construe the plan. The authors of the deed told us that they used the 1911 map. Although the southern boundary of the track was not an enclosure boundary, the author had referred to the deed plan as “excerpted” and “reduced” from the map and if a line on the plan was sufficiently explicable by a line on the map defined as the source there was no reason to refer to extraneous evidence putting the line elsewhere and contrary to common sense to do so.
 Further, said Mr Upton, Mr Steel had acknowledged that there was no evidence that the plan was informed by a survey of the site. It was only by speculating that it was informed by some sort of survey that the interested party’s case had any basis: if the boundary was intended to reflect the actual position of the track in 1962, the author of the plan or an informant of his must have visited the site. The alternative was to proceed on the basis stated in the deed and plan.
 Mr Upton further submitted that the appellant’s construction was in any event to be preferred on general principles of construction. The 1911 map was a public document which was the result of a careful survey. The alternative source for the disputed boundary was the reality of where the track ran in December 1962, as evidenced by 1962 and 1963 revisions of Ordnance sheets, which were, however, not suggested to be the source of the deed plan and were not published until 1965. The alternative, therefore, was simply the physical reality which was at all material times an unfenced, unsurfaced agricultural track, which might wander and on the photographic evidence appeared to have done so or at least widened. The coincidence of evidence of an ordnance revision in 1962 was the kind of contingency on which a system of deeds could not rely. That was within draftsmen’s contemplation, as was the probable changeless endurance and clarity of the Ordnance Survey which they had used and to which they expressly referred. The more practical and convenient, and hence reasonable, way to resolve uncertainty was to refer to the authoritative public document, a fortiori when the deed and plan stated that the plan was based on that. That provided fixed points of reference, which were preferable guides over evidence about ill-defined physical features which might move. Mr Upton referred to David Runciman & Sons v Scottish Borders Council, per Lord Drummond Young at para. 7.
 Mr Upton also adopted Mr Jillott’s evidence.
 For the interested party, Ms Ross first considered the nature of the dispute. The primary issue was whether or not the Jackson disposition must be interpreted solely by reference to the 1911 map. She agreed that there was no complex law involved: a construction convenient and workable to both parties was required. The difficulty was in interpreting a very badly drafted plan. The Keeper’s two decisions – plotting the boundary where he did, and refusing to rectify – could be taken together, and it was not a question of scrutinizing the reasonableness or correctness of the decisions on the material before her. The onus was on the appellant to show the inaccuracy and that onus had not been satisfied. The track being between 3 and 4 metres wide, the alleged discrepancy at most 8 metres, and the margin of error on the line drawn on the Jackson disposition plan between 4 and 10 metres, the disputed area was within that range.
 Ms Ross considered the appellant’s reliance on the 1911 map, on which, she said, the appellant’s argument depended entirely. The dogleg in the boundary was an immediate departure or innovation: the line of the track did not by itself show the boundary. Mr Jillott’s deliberation had started and stopped with the 1962 deed and the 1911 map, although a letter from the appellant’s agents to the Keeper had referred to later maps. He had not used the Gracie plan. The appellants relied chiefly on references to the 1911 map in the text of the deed, but, apart from the reference to ‘excerpt’, these were in the context of field numbers, the main purpose being to convey certain fields and enclosures (and it being accepted that in those cases, if there had been any boundary changes since, the 1911 map boundaries would prevail). No tracks or other boundary features were referred to. Where parts were conveyed, it was necessary to go to the plan. As to what, if any, use could be made of the 1911 map in those cases, Ms Ross submitted that while it made sense to refer to field numbers as a convenient description, it was going too far to borrow other potential boundary features; the disposition did not mention any track; an ‘excerpt’ not being a copy, there was immediate difficulty in relying on a line drawn on the deed where there was nothing else in it to say that it was the line of a track; and there was nothing on the deed plan to indicate that the line was a track. There was nothing to support the contention that the draftsman intended the 1911 track to be the boundary. Instead there was elaborate surmising and supposition from the text.
 That, said Ms Ross, would be sufficient for her argument that inaccuracy was not established. However, she also looked at the position in 1962. The track was shown then to be in a different position from that shown in the 1911 map. Because reliance was placed on such evidence, it could be accepted that the boundary was track-based. The only track on which it could be based was the track on the ground in 1962, reflected in the Register Title Plan. It was overwhelmingly more likely to be in the position shown on the 1963 map than that on the 1911 map. It was highly unlikely that the draftsman would choose something which was not there, referring to a map out of date by 50 years, and much more likely that he would have regard to the facts on the ground. There was speculation on what Mr Cullen, who prepared the deed plan, did. Although it would have been good practice to inspect the site, it could not be said whether he had done so or not. However, he would have had the Gracie disposition plan. That plan did not assist interpretation in the light of the 1911 map, nor had Mr Jillott suggested that. On that plan, the track ran along beside the barn, as on the 1963 map, consistent with the Register Title Plan. Therefore, even if this were simply a desk exercise, and if it were accepted that the boundary was a track line, it would be that line, alongside the barn. It was not necessary to say that the current map line must have been intended. The current position about the track on the ground told nothing about its position in 1962, but it had been identified as having a particular line at that time. The photographs did not allow any plotting or useful conclusions, except that there had been a barn and a track alongside it.
 Finally, Ms Ross turned to the actual surveying evidence. Accepting that if the appellants were correct on the approach to interpretation, there was evidence of some inaccuracy, the difficulty with Mr Jillott’s evidence was that overlaying a 1962 map acetate on the 1911 map, the two tracks ran side-by-side, i.e. showed a difference of one track’s width, the Appeal Plan showed a difference of up to 7.25 metres and therefore could not be accepted.
 Replying on some points, Mr Upton referred to Mr Jillott’s evidence on re-examination: whereas he might seem to have agreed on Ms Ross’s last point in cross-examination, he had explained in re-examination that his exercise, leading to the Appeal Plan, was more reliable. The Tribunal should find the Appeal Plan accurate, but, if not, the matter should be put out for further consideration. He said that it was clear, and accepted, that the ‘dogleg’ represented switching from one side of a path to the other. He pointed out that there were a total of 5 references in the Jackson disposition to the 1911 map. It was not obvious that it made sense to refer to the 1911 map for the enclosures and boundaries but not in relation to parts. It was still appropriate in considering what parts of ‘59’ and 58a’ were referred to to use a line on the 1911 map which could explain the position. Whatever ‘excerpt’ meant, there was no dispute that the plan was taken from the 1911 map. On the fact that the plan did not show the track, this was at worst neutral but in fact supported the appellant’s position because if it were shown, it might be taken to refer to the existing track. He asked why two different sources should be used to define the boundary and divide the property when there was one which enabled both to be done. The draftsman had clearly not had regard to the position on the ground. Although it was the practice of surveyors to inspect the site, there was evidence that they did not always do so. There was no detailed comparison between the Gracie disposition plan and the 1911 map, and in any event the deed did not give the Gracie disposition definitive status. If it was not known how well defined the track was in 1962, that weakened the case for relying on the position then. It would be foreseeable in 1962 that a farm track was not a reliable line to use as a boundary. Finally, on using acetates with ‘blown up’ plans, Mr Jillott had testified that he had taken care to check the measurements after enlarging: the Tribunal would be able to check this for itself.
 The appellant’s case stands or falls on the proposition that the correct interpretation of the Jackson disposition in 1962 is that the boundary then created was determined by the position shown on the 1911 map of a track.
 We did not understand parties to disagree in any way on the law applicable. Firstly, the Tribunal is to consider the correct interpretation of that boundary, the so-called “true” boundary, in order to determine whether there is inaccuracy in the Register entry: as the Tribunal has recently confirmed, it is not a question of reviewing the Keeper’s decision or decisions on the material available (PMP Plus Ltd). Secondly, normal principles of construction apply, in particular that where two constructions are possible, the more reasonable, practical and convenient, or as Ms Ross put it, “workable to both parties” was to be preferred. Thirdly, both parties recognized the need for a so-called “Hoffmann” approach (Investors Compensation Scheme Ltd v West Bromwich Building Society) of taking the “matrix of fact” into account, i.e. considering the background or surrounding circumstances as known to the parties in order to assist in determining the intention of the deed. We should perhaps mention that there did at times appear to be a concentration on the intention of the person who drew up the deed plan, but we think it is the intention of the granter of the deed, or perhaps the parties to the underlying contract, that has to be determined on an application of the rules of construction.
 We should first look at the Jackson disposition itself. As Mr Upton pointed out, the text, in particular of course the verbal description of the property conveyed, would normally prevail over the plan. Most of the property was described as “field or enclosure Number … on the Ordnance Survey Map”, the particular scale and edition being specified. The deed plan is referred to as an “Excerpt” from the 1911 map. The deed plan itself tells us that it is scale 1:5000 “reduced from” the 1911 map, and carries the name of a firm of architects and surveyors in Hamilton. The plan is notable for the thick line, of variable thickness, round the boundaries. That line appears plainly to be drawn rather than accurately plotted. Reference to the deed plan would give only the roughest idea where the boundaries lay. Whatever “excerpt” means, the boundary cannot be plotted on the ground from the plan.
 The intention in relation to whole fields or enclosures is clear: it is necessary to look at, and only to look at, the 1911 map. As Ms Ross accepted, no question of looking at the position at the time of the deed arises. As we saw from the more contemporary maps, fields or enclosures might have disappeared, or the physical boundaries changed, but the boundaries shown on the 1911 map would still have to be applied.
 The position is inevitably at least slightly different in relation to “field or enclosure Number … (part) … ”. Although the 1911 map is still referred to, it does not enable the part conveyed to be identified. There was an overall area measurement but it was not suggested that that could assist at all. Nor was there any evidence of possession. How then is the disputed boundary to be determined?
 As far as the parts of ‘58a’ and ‘59’ are concerned, regarding their combined area as very approximately triangular, we find that the third, internal boundary, as well as the other two boundaries which are part of the overall boundary, is also marked on the deed plan with a thick line. The explanation for that is found in the deed, because the resultant triangle or wedge shaped area, which is shaded a different colour on the plan, is the subject of two access provisions within the deed: firstly, slightly strangely perhaps, it is identified as a particular area over which the disponees are to have pedestrian and vehicular access; and secondly, an access right is also constituted over it in favour of the granter and his successors as proprietors of the remainder of field 59 (now including the appellant), and also in favour of the proprietors of the adjoining subjects to the north north west (the disponees in the Gracie disposition and their successors). Thus the area consisting of parts of ‘58a’ and ‘59’, whose south-eastern boundary is the disputed boundary, is made the subject of access rights as well as being conveyed.
 The other thing to be noted about the wedge shaped area on the Jackson disposition plan is that despite the roughness of the boundary marking there is a clear kink or dogleg on the disputed boundary, consistent with part of a track or other kind of route being conveyed subject to an access right but the other part retained. However, although some tracks are drawn on the plan, with dotted lines, this track is, the parties agreed, not depicted. (We did wonder, since only copies of the deed and plan were available, whether it might in fact have been on the original, but the witnesses and parties appeared all to agree that it was not and we have proceeded on that basis.)
 Turning to the surrounding or background facts, there is of course the 1911 map to which the deed specifically refers. Holding the 1911 map beside the deed plan enables us to confirm that the latter, although roughly drawn, does show the approximate positions of the fields as they were shown in that map. We refer to our findings as to what the 1911 map does show, particularly of course the track running along the north-west boundary of ‘59’. This supports the theory that this boundary was fixed by reference to a track, part conveyed subject to the access right constituted in the deed and part retained, although it does not enable the position of the dogleg to be fixed.
 Next, there is the Gracie disposition, executed a few months prior to the Jackson disposition. This is of a very much smaller area. It includes a full bounding description of the land conveyed, with reference also to a detailed plan. It also specifically conveys, and shows on the plan, a range of glasshouses which both from the plan and from the deed can be seen to abut the south-east boundary of the subjects conveyed. It also conveys a right of access over an area not particularly described but again clearly shown on the plan. This area appears similar to the triangular or wedge shaped area identified in the Jackson disposition as the subject of access rights. No track is shown within this access area, but there is a corresponding dogleg and the plan does appear to show a track continuing. The barn, not shown in the Jackson disposition plan, is shown on this plan with the track apparently running along the length of its north-west elevation. The access provisions in the Jackson disposition refer back to those in the Gracie disposition. Both Mr Upton and Ms Ross proceeded on the basis that the Gracie disposition was known to the granter of the Jackson disposition, who was the same person, a few months later.
 Mention requires to be made here of some uncertainty in relation to Ordnance Survey maps. It is clear that there was revising activity, and indeed a change in the mapping basis, between the date of the 1911 map and the 1965 map produced to us, the latter recording that it had been revised in 1963. (Mr Steel in fact based an acetate on a map which he said that he had obtained from the library in Lanark and from which we were further informed that boundaries had been plotted in “1962/63”.) No intervening Ordnance map was produced. The Gracie disposition plan appears clearly enough to have been based on an Ordnance map, although with some figures and also some features in the relevant areas removed. It is, however, clear to us on closer examination that many minor features on it resemble those on the 1911 map much more closely than those on either of the 1960s maps produced. Further, the barn, and also the physical boundary at the west end of the triangular area, are not plotted in exactly the same positions as those shown in the 1960s maps. Nor is the track.
 So the Gracie disposition plan does not precisely show the position on the ground of the triangular or wedge shaped area in 1962. The precise boundaries of the access area would not be crucial, at least once one moved away from the road end, because this area was not being conveyed, nor was a precise access route being constituted. What the Gracie disposition and plan do provide, as it seems to us, is part of the setting for the Jackson disposition, viz the creation of a roughly triangular shaped access area whose south-eastern boundary apparently abutted one elevation of the barn and from its dogleg shape may also have followed a track, changing sides on the track at the westmost corner of the barn.
 Aerial photographs at different dates before and after the Jackson disposition plan, while not at all clear or definitive, appear to show the barn with a rough track running along beside it. From these as well as the Gracie disposition plan and the 1963 revision of the OS map, we can find another potential part of the matrix of fact, namely the track, which we can say was a rough agricultural track. For this to become part of the matrix, we require to establish that the granter, or the parties, knew of its existence at the time. As we see it, that also is established by linking the kink on the deed plan with the Gracie disposition plan: the track explains the kinks, in at least approximately the same position, in the two plans.
 The evidence also shows that the route of the track shown on the 1911 map is not the same as the position shown on the contemporary maps. There is, however, no evidence of that fact being known. The rough drawing of the Jackson disposition plan, together with the fact that this referred to the 1911 map, suggests to us that, whether or not the drawer of the plan inspected the site, no comparison is likely to have been made between the 1911 map or any contemporary map and the position on the ground. So the discrepancy which has now come to light was probably not known at the time.
 This is, however, consistent with an awareness that the position of such a track might well change over the years.
 There is, as it seems to us, an irrefutable inference that the north-west boundary of the wedge shaped area conveyed in the Jackson disposition coincides with the south-east boundary of the subjects conveyed in the Gracie disposition: apart from the unlikelihood of any gap between them, the access provisions would not otherwise work. In other words, it can be said with certainty that the boundary of that part of the field or enclosure 58a was determined, not on the basis of the 1911 map which cannot assist in relation to it, but on the basis of the boundary, including a wall and, for much of its length, one side of the glasshouses, established by the Gracie disposition and on the ground at the time.
 The west boundary of the wedge shaped area looks to us also very similar to the west side of the access area depicted on the Gracie disposition plan, the 1963 map also showing a solid line near that point, although, as we have seen, its position on the Gracie disposition plan does not precisely correspond with the 1965 map. That, however, cannot be affirmed with certainty. What can be said about that west boundary in the Jackson disposition is that it too cannot be based on the 1911 map, which has no corresponding line.
 It can therefore be said, without considering the disputed boundary, that the Jackson disposition requires two different methods of interpretation of boundaries: in the case of whole field or enclosures, reference to the 1911 map; and in relation to two boundaries of ‘58a part’ and ‘59 part’, extrinsic evidence.
 That brings us to the disputed boundary line, or more accurately boundary lines since there is of course the dogleg. Which method of interpretation is to be deduced? Is this boundary to be fixed by reference to the 1911 map, which does have lines of a track in approximately the right position, or by reference in some way to the position in 1962?
 We must avoid hindsight, in particular the knowledge that the position of the track was different from that shown in the 1911 map. We should also remember that the access area boundaries, as such, may not have been of much importance at the time.
 In our opinion, in the light of the surrounding circumstances at the time of the deed, the most likely intention or meaning of the Jackson disposition is that this boundary was based on the position on the ground at the time, for the following reasons.
 It seems to us that the scheme of access rights constituted by the two 1962 dispositions is likely to have involved the same area. The area had been plotted on the Gracie disposition plan. An area consisting of parts of two fields and at least similar in shape was the subject of special stipulations about access in the Jackson disposition. Its boundaries were also the boundaries of the land disponed. Its south-eastern boundary is likely to have been the same as that intended in the Gracie disposition. That view is fortified by the existence of the track. That boundary appears to have been determined by the position of the barn and the track, as shown (albeit inaccurately) on the Gracie disposition plan. The barn was not shown on the 1911 map. An intention to define the boundary by reference to the position of the track on a 1911 map appears unlikely. That would have risked diverging from the current position of the path and possibly even bisecting the barn.
 Such an intention also appears to us unlikely when it is remembered that the other two boundaries of the parts of the two fields could not be defined by reference to the 1911 map. The description of the area would have to be along the lines, ‘ground bounded by glasshouse and walls and, on the south side, lines of a track shown on the 1911 map.’ If that had been intended, we think that it would have been spelt out. At least the position of the kink or dogleg would have been set out.
 Mr Upton argued that the interested party’s position assumed that the author of the Jackson disposition plan had been informed by a site survey. In our view, however, the granter of the deed, and the parties to the contract, need only have known the approximate position of the barn (his own property and to remain so) and the track running beside it, to have decided that the boundary would be fixed under reference to these. Some attention was, after all, focused on this access area. The position was shown on the Gracie disposition plan.
 Then it was argued that reference to the 1911 map would provide a way to resolve uncertainty. In our view, however, such a recognition of potential uncertainty in relation to the line of a rough track must include a recognition that the position of the track might have moved since 1911. In other words, the boundary would be fixed along the line of a track, indeed crossing it at a point in the middle so as to allocate one section to one owner and the other to another owner, knowing that the boundary might not coincide with the lines of the track. If a new boundary, unrelated to physical features, was being created, it might well be appropriate to plot it under reference to an old map so as to fix the line authoritatively, but the intention to relate this boundary to a track, and (on the basis of the Gracie disposition), a barn, seems to us to make this approach unlikely.
 Again in relation to certainty, Mr Upton argued that the coincidence of evidence of an ordnance revision around the same date was the kind of contingency which could not be relied on when the physical reality was an unfenced agricultural track which might wander. There is some force in that, but it is, or at least was in the era of Sasine titles, not uncommon to have boundaries not clearly defined either by description or plan, thus requiring, in the event of dispute, evidence of one kind or another as to the position, at the time of the deed, of some feature which is no longer present or not necessarily still in the same position. Again, the existence of the barn, an immobile structure, was also known.
 We therefore consider that, on a proper construction of the Jackson disposition, the intention of the deed was to reflect the current position of the track and barn and not the line of the track shown on the 1911 map.
 That is sufficient to dispose of this appeal, as the appellant has failed to persuade us that the boundary should follow the 1911 map and have not advanced any other case for inaccuracy.
 Although the matter does not arise in view of our decision on the main issue, we have considered whether the appellants have established the correct position of the boundary if it required to reflect the position of the track shown on the 1911 map, i.e. whether in that event the appellants’ ‘Appeal Plan’ accurately showed the ‘true’ boundary. We have decided that we would not have accepted that plan as accurate but that, in the circumstances, we would put the case out for a further hearing on this matter. This is because we consider that the appellants did not have a fair opportunity to consider the effect of certain evidence led by the interested party.
 In case this matter is taken further, we should indicate our thinking on the material presently before us, although we could of course not reach any final view at this stage in the circumstances.
 It appears to us that the ‘Appeal Plan’ prepared on the basis of Mr Jillott’s report may significantly over-estimate the difference between the boundary on the basis of the position of the track in 1962 and its position shown on the map. When an acetate of the 1911 Scale 1:2500 Ordnance map is placed over the 1:2500 1965 Ordnance map the clear impression gained, if alignment starts at the east end of the track between 4 and 6 Lanark Road, is that the position on the ground has shifted to the south by about 3m at the west end of the track and by about 1m at the north west corner of the barn and where the kink appears on the Gracie disposition plan. The equivalent distances on the ‘Appeal Plan’ are about 6.5m and about 3m. That apparent discrepancy flows through to the comparison between the 1911 map and the Register Title Plan, because we have been able to confirm that the latter is very close indeed to the 1965 map (subject to a very small difference, possibly within normal margins of error, in the appellants’ favour, because the line of the track appears to have been pushed very slightly to the north by the corner of the garden of No 6 Lanark Road).
 Mr Jillott’s comparison of the 1911 map with the Register Title Plan, seems to deprive him, in the particular circumstances, of the benefit of two further fixed points of reference in the immediate locality, viz the wall along the north-western boundary of the wedge shaped area, which appears on the 1965 map and is still present, and the barn, which appears on the 1965 map but has since disappeared. The one-stage comparison between 1911 and the Register might seem attractive, since on the appellants’ view the position on the ground at the time of the deed is not relevant, but the two-stage comparison, 1911 to 1965 and 1965 to current, referred to in the last paragraph, enables advantage to be taken of more points of reference.
 Another possible reason for the difficulty with the ‘Appeal Plan’, as we see it, is that aspects of Mr Jillott’s methodology seem open to question. He enlarged the Register Title Plan from a scale of 1:2500 to a scale of 1:500 and plotted the line of the roadway from the 1911 map, similarly enlarged. It may be superficially attractive to work at a larger scale, and no doubt useful for illustrative purposes in considering the main issue. In our opinion, however, as a method of plotting, it does not add any accuracy but rather introduces the potential for error at the stages of enlargement and reduction. It would in the Tribunal’s view be preferable to make any comparison of the 1911 and 1965 map or the Register Title Plan at their original scale. The line of the track derived from the 1911 plan could thus be plotted onto a revised Register Title Plan.
 Further, Mr Jillott plotted the position of the track on the ground following his site inspection, on the basis of the bearing (229 degrees) which he measured on the map and for which he found some other support. We regard the measurement of angles on the ground, in this case by looking along a stretch of old masonry which is itself uneven, as an imprecise method – a difference of one degree would not be at all unlikely and could account for the difference which we have found. Comparing maps or plans by measuring bearings seems to us less reliable than by using acetates.
 Mapping techniques have improved over time. As we understand it, there can in fact be slight variations in successive editions of Ordnance Survey maps. Thus, for example, the positioning of buildings which are unchanged on the ground over time might be shown sited in very slightly different positions on successive editions of the map. When trying to make comparisons, one has to identify a hierarchy of points, looking for features nearer the subjects rather than more distant, to assist in obtaining the best fit. In this case we have regarded the houses at 4 and 6 Lanark Road as the first point of comparison and secondly, when available, the wall, or part of the wall, which formed the south-east boundary of the subjects in the Gracie disposition and which is shown on both the 1965 map and the Register Title Plan.
 Accordingly, while it is accepted – and we agree – that on the appellants’ approach there would be some inaccuracy (going beyond normal margins of error), we could not, on the material before us, accept Mr Jillott’s Appeal Plan as showing the extent of the inaccuracy. As we have said, however, rather than producing or instructing another plan which would reflect evidence introduced without fair notice by the interested party, we would have put the case out for further consideration of this matter.
 The appeal is refused. In the event of any disagreement in relation to expenses, the Tribunal would propose, unless parties request otherwise, to dispose of that matter on the basis of written submissions.