This is an application for an order declaring that a certain area of ground forms the whole area of croft 113 Borve, Isle of Barra. It is, therefore, in effect, an application to fix the boundaries of that croft. The applicant is Mrs Cecilia MacFadyen, the tenant of the croft.
 The application follows on from a number of regulatory applications made to the Crofters Commission, as it was at the time, in relation to a neighbouring croft which is, despite the numbering, 88 Borve and lies immediately west of croft 113. That croft used to be tenanted by Mrs Mary MacNeil and in or around 2009 she and her granddaughter, Mrs Michelle Galbraith, entered into an agreement whereby the tenancy of the croft was assigned to Mrs Galbraith, who then applied to the Commission for consent to divide the croft and, that having been done, re-assigned part of the croft to Mrs MacNeil. The result is that croft 88 is now divided into 88/1, which is tenanted by Mrs Galbraith, and 88/2, which is tenanted by Mrs MacNeil. Both Mrs MacNeil and Mrs Galbraith are respondents in the present application.
 The purpose of the foregoing applications was to allow Mrs Galbraith, who belongs to Barra but had been living in Glasgow, to return to the island upon her marriage and to build a house on what had been her grandmother’s croft. She therefore lodged an application with the local authority for planning permission to build a house on what she says is part of croft 88/1. She also applied to the Crofting Commission, as it had by then become, for an order decrofting the intended house site and an area of garden ground. That application, at some stage, came to the attention of Mrs MacFadyen who wrote to the Commission out of concern that the intended house was to be built on land which was in fact part of her croft rather than of croft 88/1. Notwithstanding her intervention, however, the Commission granted the decrofting direction and Mrs MacFadyen then appealed that decision to this court (SLC/9/14). The court having identified that the true dispute was not between Mrs MacFadyen and the Commission but between Mrs MacFadyen and Mrs Galbraith, we sisted that appeal and invited Mrs MacFadyen to bring the present application.
 In addition to Mrs MacNeil and Mrs Galbraith, Mr Donald Archie MacNeil, the tenant of 95 Borve, which, again despite the numbering, borders 113 Borve on its eastern side, is a respondent in this case. Mr MacNeil is Mrs MacNeil’s son and Mrs Galbraith’s uncle. The Scottish Ministers, who are the landlords, have not entered appearance in the case.
 In order to understand the scope of the disagreement it is helpful to look at production 5. It was lodged by the applicant and shows croft 113, as she asserts it to be, shaded in pink. The site for Mrs Galbraith’s intended house is outlined in red and it is this area and the surrounding area which is the immediate focus of dispute but the case also concerns the two pink indentations, as we came to call them at the proof, shown extending across croft 88/2 to the south of the indentation on which the house is proposed to be built. Finally, there is an area at the north end of croft 95 and adjacent to croft 113 which the respondents say is part of 113 but which the applicant maintains is part of 95.
 We heard the application at Castlebay on 4 February 2015, when the applicant was represented by Mrs Eilidh MacLellan and the respondents by Mr Simon Fraser, both solicitors. We carried out an accompanied inspection immediately after the close of the hearing.
 The applicant herself did not give evidence. Instead evidence was led on her behalf from her son, Mr Iain Roderick MacFadyen, and her daughter, Mrs Margaret Ann MacFadyen or MacIntyre. For the respondents, evidence was given by Mrs Galbraith, Mrs MacNeil and Mr Donald Joseph MacLean, the Township Clerk of Borve. We have no doubt that the witnesses all gave their evidence in good faith and we found them all credible. Inevitably we have had to choose between two competing accounts of matters but the only area in which we have not been able to reconcile the evidence given with other evidence which we find to be credible and reliable has to do with the late Roderick MacFadyen’s understanding of the boundaries, as spoken to by Mr Iain MacFadyen and Mrs MacIntyre. We explain how we dealt with that conflict later in this judgement.
For the applicant
Mr Iain Roderick MacFadyen (46)
 Mr MacFadyen gave evidence that he was the applicant’s son. He explained that he was giving evidence in place of his mother because he had a better understanding of the croft boundaries than she has, due to his father having explained the boundaries to him. He had been brought up in Borve and had remained on Barra until the age of 27 when he had left, first for Inverness and then Glasgow. He continues to live in Glasgow but holidays in Barra regularly. When holidaying in Barra he stays at the family home in Borve, which is situated on an apportionment from the common grazings rather than on the croft itself.
 He spoke to production 27A, an A3 version of production 27 which was lodged by Mrs MacLellan on the morning of the proof. He said it was a copy of a map his father had had. He did not know where his father had got it. He believed it to be the township map, showing the boundaries of the crofts. The map was dated 1957. That date appeared on production 28 which showed a key to the various hatchings and shadings used on production 27A.
 There had been a history of informal occupation of croft land in Borve. For example, Mr Michael MacNeil had had an agreement with the witness’s grandfather, whereby Mr MacNeil used the “top side” of croft 113 (that is to say, the part of the croft to the south of the public road). Michael MacNeil had been the tenant of croft 95. The witness’s father had taken over the croft from his own father whilst Mr Angus MacNeil had succeeded to the tenancy of 95. The witness’s father and Mr Angus MacNeil had had an arrangement, convenient for them both, whereby Mr MacFadyen had used the bottom end of croft 95 (that is to say, the part of the croft north of the public road) and Mr MacNeil had used the area within which Mrs Galbraith had sought her decrofting direction. The witness’s father and Angus MacNeil had been friends, the arrangement suited them both and there had been no disagreement over the years.
 The witness’s father had used the part of 95 we have referred to for sheep. That part had not been fenced. There was therefore, no division between it and the neighbouring part of croft 113. Mr MacNeil had used part of 113 for his cows.
 Asked about the present situation, the witness confirmed that there was no fence at the bottom (north end) of 113 and that sheep could still go backwards and forwards between it and the neighbouring part of croft 95. The part of 113 which Mr Angus MacNeil had used was still used by Mrs MacNeil for her cows. Neither Mrs Galbraith nor her predecessors had made any improvements on the land. A Nissen hut had been built, probably for storage. That was not a permanent structure; it could be unbolted and moved. The witness had been happy for this informal use of part of croft 113 to continue as it was; that is to say to continue being used for grazing.
 Asked why his parents had decided to build their house on the common grazings and not on the croft, Mr MacFadyen explained that it had been more convenient; his father had not had a lot of money and to create a suitable access to a house on the croft would have cost a lot of money. He did not know whether an apportionment order in respect of that part of the common grazings on which the house stood had ever been recorded in the Register of Crofts. He confirmed that the landlords were the Scottish Ministers. Communication with them was by way of email and phone contact with their representative in the Uists, Mr Ian Muncaster.
 The landlords also had a copy of the map, production 27A. Referred to production 22, a copy of a letter from Mr Muncaster (designated “W J Muncaster. Higher Agricultural Officer”) to Mrs MacFadyen dated 15 February 2012, the witness confirmed that he had been aware of his mother receiving that letter. He had discussed it with her. They had been trying to confirm the boundaries of the croft. Mr Muncaster said in the letter that an application for a grant for fencing had been refused because of a boundary dispute. The letter said that the application to divide croft 88 lodged by Mrs Galbraith had been based on incorrect plans.
 Following receipt of this letter he had gone to see Mr Donald Archie MacNeil (the respondent of that name). He had regarded Donald Archie as a friend and he had gone to sort things out between them. Mr MacNeil had said that he did not know who croft 88 belonged to but he had offered to exchange part of 95 adjacent to his house for the part of 113 on which Mrs Galbraith wanted to build her house. He identified Mr MacNeil’s house on production 5. That had not been an attractive proposition from the witness’s point of view because the area being offered was a bog of no use. He had told Mr MacNeil that. Their conversation would have been at the beginning of 2012 shortly after his mother had received the letter from Mr Muncaster, production 22.
 Following this conversation with Mr MacNeil, he had contacted Mr Muncaster and emails had been exchanged. The purpose of this had been to agree boundaries. He had not been aware of Mrs Galbraith’s application to divide croft 88 before his mother had received production 22.
 Mr MacFadyen was then referred to production 39, a letter from Mr Muncaster to Mrs Galbraith dated 18 January 2013. Attached to this letter are three maps, two of which appear to postdate the letter itself. Mr MacFadyen said that his mother had received the same letter, but addressed to herself, but two of the three maps which had accompanied the letter to his mother had been different from the maps attached to production 39. All the maps which had accompanied the letter to his mother had been dated 16 January 2013. In those letters Mr Muncaster had been trying to get agreement to reorganise the croft boundaries. Mrs Galbraith had not responded. He himself had responded on behalf of his mother, saying that he was broadly in agreement with Mr Muncaster’s proposals but the maps to which he had been responding were not the same maps as were attached to production 39. He had broadly been agreeable to splitting the part of 113 to the south of the road down the middle. The purpose of the intended reorganisation was to confirm agreed croft boundaries and he and his mother had wanted to fence these boundaries once agreed.
 Asked why he and his mother had not objected to Mrs Galbraith’s application for consent to divide croft 88 (production 37), Mr MacFadyen explained that in January 2009 his father had been terminally ill, having been diagnosed with a tumour more than a year earlier. That would have been at the beginning of 2008. His father had been attending hospital in Uist and then backwards and forwards to Glasgow for chemotherapy and, in the course of it all, had suffered a stroke as well. The witness himself lived in Glasgow but his mother or his sister were having to accompany his father from Barra to Glasgow for his treatment. They could be away from Barra for up to two weeks at a time.
 Mr MacFadyen was then referred to productions 23 to 25, being copy entries from the Crofting Commission’s Register of Crofts, 23 relating to 113 giving an area of 2.87 ha, 24 giving the same area for croft 95 and 25 giving an area of 1.522 ha for croft 88/1. The witness did not accept the figure of 2.87 ha for croft 113 as accurate. He did not know where the Commission had got it from. He had no knowledge as to whether the figures for the other crofts mentioned were correct.
 In addition to the conversation with Donald Archie MacNeil to which he had referred, the witness had been spoken to by Mrs Galbraith’s mother, Mrs Elizabeth MacNeil, who had approached him in the Castlebay Hotel sometime in 2011 to say that Michelle would like to build a house where the Nissen hut stood. According to Mr MacFadyen, Mrs MacNeil had said “We know its Roddy’s land, but we are sure you wouldn’t mind.” “Roddy” had been a reference to the witness’s father. He had thought this approach to be “a bit cheeky”. He had never discussed the croft boundaries with Mrs Galbraith herself.
 The only other person he had discussed boundaries with was Alasdair MacNeil, tenant of croft 108, a cousin of Mrs Galbraith’s who had come to see him last August whilst the witness had been fencing the lower part of 113. Alasdair had come to speak to him about the dispute and had said “you are correct in what you are saying”. Alasdair MacNeil’s reason for saying that had been that his late father, Donald John MacNeil, had explained the boundaries of 113 to him. His father had told him “there will be trouble over that yet”. The reason for him saying that had been that the MacNeil family had been using part of 113.
 Reverting to Mr Muncaster, that gentleman had phoned the witness’s mother in June 2013 to say that Donald Archie MacNeil had offered to exchange part of 95 for the area on which the Nissen hut stood on 113. The witness had sent Mr Muncaster an email explaining that they (the MacFadyens) were not exchanging any part of the croft. That email, he believed, was dated 27 June 2013.
 Mr Muncaster had also said that Mrs Galbraith had offered to exchange part of 88 for the land on which she wanted to build her house. That had been in July 2013. The part of 88 in question was at the north end of the croft and was ground which was no good to anybody. It was a bog.
 On 16 July 2013 the witness and Mr Muncaster had pegged out the top area (south of the road) of the croft to reflect what the 1957 map showed. They had pegged out two ways in which the crofts could be divided. One was per plan 2 on production 39 and the other was per the 1957 map. When they had finished Mr Muncaster had gone to see Mrs Galbraith and Mrs MacNeil. He had then returned to Mrs MacFadyen’s house to speak to the witness and his mother and had explained that Mrs Galbraith had said “Just forget about doing the splits. We’ll just leave the croft boundaries as they are, per the 1957 map”. Mr Muncaster had ended their discussions by saying that they would have to sort it out for themselves. He had then sent an email with three maps and again said that the parties would have to sort it out themselves.
 The witness’s understanding was that the croft boundaries were as shown on production 27A, the township map, and, more accurately so far as the detail is concerned, on production 5. The plan at production 5 was an accurate statement of the boundaries of 113. The additional land shown as part of 113 on productions 7 (lodged by Mrs Galbraith) and 9 (lodged by Donald Archie MacNeil) were not part of 113 but part of 95.
 Mr MacFadyen was then questioned by the court as to the use made of the various parts of 113 in his younger days. The Nissen hut had been built in or around 1985. Angus MacNeil had used that area whereas Roderick MacFadyen had used the bottom end of 95. There was a fence between the three indented areas shown on production 5 and the rest of 113. It had been erected by the witness’s father but before that there had been a stone wall running along that line. One could not access these indents from 113.
 There had never been a fence between the boundary of the area on which Mrs Galbraith wanted to build her house, where the dotted line is shown on production 5, and the rest of 88/1. Nor had there been a fence between the bottom of croft 95 and the neighbouring part of 113; sheep had always been free to come and go across both crofts at the bottom end.
 When the witness had been growing up his father had not used the long continuous strip undoubtedly forming part of 113 to the south of the road. That had been used by Mr Donald John MacNeil, of croft 107, for cattle. Where dotted lines were shown representing the division between the indents shown on production 5 and croft 88, there were no fences. On the west side of croft 88 there was a continuous fence which ran all the way from the bottom of the croft up to the hill. The MacNeil family (by which we understand the family of Mr Angus MacNeil) had used the most northerly of the three indent areas for cattle when the witness had been growing up. The next most southerly one was fenced as shown on the map and had also been used by the MacNeil family for cattle. The area south of the most southerly of the three indents was all open with no fences and the MacNeil family had, effectively, had use of all three crofts, 95, 113 and 88 as a result. His father’s crofting operations had been confined to the part of 113 to the north of the road and the adjacent part of croft 95 at the very bottom end (shown shaded pink on productions 7 and 9).
 In cross‑examination, and under reference to production 22, Mr MacFadyen confirmed that his mother had applied for a CCAGSS (Crofting Counties Agricultural Grants (Scotland) Scheme) grant to fence the croft. He agreed with Mr Fraser that the reference to some of the maps used in the application being an error must be a reference to that CCAGSS application. The letter referred to some information having come to light. The witness said that that was a reference to the township map of crofts, enclosed with the letter. He denied that this was the point at which this map had first come to his mother’s attention. It was not the case that it had appeared for the first time with Mr Muncaster’s letter. Mr MacFadyen would be surprised to be told that most people in Borve had never heard of it. He did not know who had drawn it up. He did not know where the information had come from. Asked if it might be one person’s view of what the boundaries should be, he responded that it would have been the map for the township in 1957. He did not know why it had been drawn up. He did not know when it had first come to Mr Muncaster’s attention. He agreed that Mr Muncaster would have known about this plan when the assignation and division applications for croft 88 had been ongoing because he was the landlords’ representative.
 Mr MacFadyen confirmed that the objection taken to Mrs Galbraith’s planning application in 2012 had been based on this map. Mr Fraser put it to the witness that his mother had been intent on fencing the croft pretty much as it was now but the witness was unable to answer that because he did not have the plan which his mother had submitted with her application.
 Mr MacFadyen confirmed that after carrying out the pegging exercise with Mr Muncaster referred to above, in July 2013, he had built a road down the south side of 113. He also confirmed having built a culvert entering the middle indentation on production 5 and having cut the fence there. He had left it open but there had been no cows in that part of 88 at that time. He had not asked Mrs MacNeil’s permission to cut the fence. The fence had since been repaired but not by him.
 He was aware that Mrs Galbraith had gone through the various regulatory applications in order to build a house. It was to be built on the decrofted area. He acknowledged that his mother’s subsequent appeal against the decrofting direction had prevented Mrs Galbraith from doing so. He was not aware whether Mrs Galbraith had lost a considerable amount of money as a consequence of this. Asked whether it would distress him to hear that she had, he replied that we were here to define croft boundaries.
 He remembered Angus MacNeil. He could not say when Mr MacNeil had come into the tenancy of croft 95. Mr Fraser referred him to production 35, being correspondence between the Department of Agriculture for Scotland (“the Department”) and others relating to an assignation of the tenancy of 88 Borve to Mr MacNeil in July 1960. Particular reference was made to the letter of 23 March 1959 which is part of that production and refers to an inspection carried out on 18 March 1959 having shown the extent of the croft to be 12½ acres arable and outrun with one share in machair and one share in hill grazing. Asked whether he had any reason to doubt the stated extent of 12½ acres, the witness’s opinion was that the Department’s records would be correct.
 Going further back in history, reference was made to production 33, being an application to the Scottish Land Court by Mr Donald MacFadyen, the witness’s great-grandfather. It is dated 11 May 1925. The witness agreed that the stock recorded as being kept was one horse, one cow and four sheep. He agreed that that did not seem to be a particularly large stock. He did not know what the stocking had been during his grandfather’s time but he believed his grandfather to have kept some sheep. Asked whether his grandfather had made any use of the three indented areas shown in production 5, the witness said that Michael MacNeil had used them at that time. That had been by permission of Mr (Neil) MacFadyen.
 Mr MacFadyen agreed that the crofts in Borve seemed to run from the Craigston boundary to the hill. Yet, somehow, croft 88 was divided into four unattached pieces of ground with no access between them. He did not, however, accept that that was unusual, when one looked at the other crofts.
 Mr MacFadyen confirmed that the Nissen hut was still in place. It could, however, be unbolted and moved. According to the documentation it had been grant aided. He disputed whether a reference to a grant for a cattle shelter on production 43 was a reference to the Nissen hut; he thought it was for a grant for a cattle shelter on 113.
 Mr Fraser then made reference to production 42A, which bears to be an extract from the OS 1880 map, surveyed in 1878, and, using laptop projection, superimposed production 16, being an aerial photograph of the township, onto it. Mr Fraser suggested that the results showed that the straight lines bounding each side of 88 and corresponding to the boundaries claimed by Mrs Galbraith had been there for a long time but the witness was unable to comment on the 1880 OS map.
 Mr Fraser then referred to production 31, which appears to be an extract from a schedule prepared by the original Crofters Commission in 1887 and contains, at entry 446, reference to an Alexander Buchanan as the crofter of ¾8 Borve with stated acreages which Mr Fraser worked out to be 10 acres and 24 poles, or 4.21 ha, in total. Production 32, a fair rent application to this court in 1916, shows, at entry 8, the name “Alexander Buchanan” scored out and the name “James MacDonald” written in, the croft referred to as 88 but the acreages quoted being the same as on production 31. Mr MacFadyen confirmed that this appeared to be the stated acreage of croft 88 recorded by the Land Court at that time. The witness was not aware of the precise extent of the three indented areas claimed to be part of 113.
 Reference was made to production 34, a fair rent application made to this court in May 1985, which shows at entry 8 of schedule B the extent of croft 88 as 6.825 acres but the only point which seemed to be made in this regard was that the extents given for croft 88 had differed in the various documents looked at over a period of time and were different again in the Register of Crofts; a proposition with which the witness agreed.
 Mr Fraser put it to Mr MacFadyen that croft 88 had been operated and worked within the existing boundaries by the tenants of that croft for at least 60 years. The witness said that he could not go back that far. He denied that current and previous tenants had always regarded croft 88 as extending to its current boundaries.
 When questioned by the court as to whether he remembered seeing the map of which production 27A is a copy at home when he was young, Mr MacFadyen said that they had had a copy in the house for many years but it was possible that they had also been sent another copy by Mr Muncaster. His mother would have been aware that the map was in the house but it had been his father who had taken to do with the croft, not his mother. The map would still be in the house. He was not aware whether it had been given to Mrs MacLellan to lodge as a production.
 Questioned again by Mr Fraser, Mr MacFadyen said it would surprise him to be told that the three indents now claimed as part of croft 113 had never been mentioned to the MacNeil family as being part of croft 113 by his father or grandfather.
 In re‑examination Mr MacFadyen confirmed that the external boundaries of croft 88 shown on the aerial photograph (production 16), superimposed on production 42A were not in doubt. The extent of croft 88 as between those boundaries was not an issue. It was the breaking up of that “long chunk”, as Mrs MacLellan put it, (the linear strip which would otherwise all form part of croft 88) which was in issue.
 Asked what his attitude would be to an advertisement relating to planning permission on somebody else’s croft in the township, Mr MacFadyen said that he would not be interested. He would however be surprised to be told that someone else could apply for a grant to build something like a Nissen hut on his croft.
Mrs Margaret Ann MacFadyen or MacIntyre (40)
 Mrs MacIntyre, the daughter of the applicant and sister of the previous witness, gave evidence that she lived at 8 Leanish, Barra. Shown productions 27A and 28(a), she confirmed that she was familiar with them. Production 27A showed the whereabouts of each croft and 28(a) showed the numbers and corresponding patterns for each of them. She did not know for certain where the plan of which those productions are copies had come from. She imagined that it had come from MacNeil of Barra. That would have been before MacNeil had handed over Barra to the State, which had taken place in or around 2004.
 Mrs MacIntyre confirmed that her mother was the tenant of 113 Borve. She identified it on production 27A. It showed what had always been her understanding of the family’s croft’s boundaries.
 However, it was not the case that everyone in the township stuck to their own bits of croft. There was always coming and going. The same happened in Leanish, where she now lived, with people having bits for potatoes, bits for peats and bits for other purposes. These individual areas would not be individually fenced but were open.
 Asked about specific arrangements between her family and neighbours in Borve, she said that she was not very knowledgeable about these matters but, so far as she was aware, in her grandfather’s time the MacNeils had used the upper part (south side) of 113 for their cattle and had done so in return for milk being supplied to the MacFadyen family. She imagined that her grandfather had used the bottom of the croft (north of the road) himself. That arrangement was with whoever had been tenant of croft 88 at that time. Her grandfather was Neil MacFadyen and the informal arrangement he had had was that rather than use the two indents south of the road shown on production 5 he allowed them to be used by the tenant of 88. So far as the third indent was concerned her grandfather had allowed Angus MacNeil to build a Nissen hut on it. She could remember her father saying that the Nissen hut was on his bit of land. Her father and Angus MacNeil had been friends.
 Asked what had led to us being in court today, Mrs MacIntyre thought it was a breakdown of communications and that the matter could have been sorted out a long time ago but had not been.
 Asked about correspondence with the landlords, she confirmed that she had been present when she and her mother had spoken to Mr Muncaster maybe two years ago, around 2012, although she was not sure of the date, when Mrs Galbraith had lodged her planning application. She had been the one to notice the planning application, she had seen it online.
 When they had gone in to see Mr Muncaster they had had the map on which 27A was based with them. Mr Muncaster told them he had sent a copy of the same map to James Ferguson of the Grazings Committee but there had been no response.
 She did not know where the copy of that plan held by her family had come from. One would have to ask her mother. The meeting with Mr Muncaster had taken place before the date of the letter which was production 22 (15 February 2012). Mr Muncaster had not said how the plan had come to be in his possession.
 With reference to the indent to the north of the public road, the area in particular dispute, Mrs MacIntyre was asked about her understanding of Nissen huts. She did not imagine they were permanent structures. They could be demolished.
 She confirmed that her father had died relatively recently. Asked whether her family had discussed whether the arrangement with the MacNeils was to continue or whether, alternatively, they wanted to have that piece of land back, Mrs MacIntyre made the point that it was her family’s land.
 Asked about the history of use of the other two indents she said that they had not been used as part of 113.
 Referred to production 37, Mrs Galbraith’s application for consent to divide the croft, she did not accept the plans attached thereto as correct. That application had been advertised on 5 February 2009. In 2009 her father had been dying of cancer and had been hospitalised both in Uist and in Glasgow. When he travelled to the Beatson Institute in Glasgow he was accompanied by either her mother or herself. They would be away for a few weeks at a time. So there was a reasonable chance that they might have been away at the time when the advertisement had appeared in Guth Bharraigh (the local community newspaper). In any event she would not have paid much attention to an advertisement of the division of a croft in itself; she would have assumed that the division would have been confined to that individual croft alone. She had not had any conversations with anyone in Borve about the boundaries.
 In cross‑examination Mrs MacIntyre was again asked about production 22. She confirmed that the letter from Mr Muncaster was referring to an application by her mother who had intended to fence croft 113. She was not sure which parts of the croft were to be fenced. So she was not able to say whether her mother had applied for a fencing grant for fences on lines different from those shown in production 27A. She did not know whether 27A corresponded with the plan her mother had submitted in connection with her grant application.
 Asked whether the informal arrangement between the two crofts might have been that the tenants of 88 had allowed the tenants of 113 to use the three indents informally, that was something which had never come to her knowledge. But it was possible.
For the respondents
Michelle Galbraith (30)
 The witness gave evidence that she was the tenant of part of croft 88 Borve. That was the part below the road, between the road and the Craigston boundary.
 She confirmed that she was the granddaughter of Mrs Mary MacNeil. She had been born in Stornoway but brought up at 88 Borve. When she was 17 she had gone to study. She had got married and she and her husband had been living in Glasgow but had talked about going back to Barra (her husband is also from Barra) and building a house on 88 Borve. So she had asked her grandmother for a piece of land for that purpose.
 She identified production 36 as the documentation relating to the assignation of the whole of croft 88 to her. That had been approved on 20 October 2008. It would have been advertised in Guth Bharraigh, probably twice. Production 17(a) showed one of these advertisements.
 That had been followed by an application to divide the croft. That in fact had been an error. Their original intention had been that her grandmother would apply to divide the croft and sign over one part of it to her but instead her grandmother had assigned the whole croft to her and she had had it divided. Production 37 was the application for consent to divide. Question 8 on that form gave the extent of each new croft. She could not remember where she had got those figures but she imagined that it would have been from the Department’s office in Benbecula. This application too had been advertised in Guth Bharraigh (production 17(b)). There had been no objections. The division had been approved by the crofting commission on 17 March 2009.
 Production 38 was the assignation application applying for consent to assign part of the croft back to her grandmother. That was the part to the south side of the road. It had been approved in September 2009. She and her husband had moved back to Barra in 2010.
 Having moved back, they had applied for planning permission for the building of a house and also for decrofting of the house site. They had also taken title to the house site. When she and her husband had applied for planning permission, in 2012, their application had been objected to by Mrs MacFadyen on the basis that the land concerned was part of her croft. However, her planning permission had been granted in May 2012; production 2.
 Mrs Galbraith had been surprised by Mrs MacFadyen’s objection. She had not been aware of any claim that the land in question was part of Mrs MacFadyen’s croft at any time until she and her husband had submitted their planning application.
 She had first seen the plan of which production 27A is a copy when Mr Muncaster had sent it to her at or around the time of Mrs MacFadyen’s objection. She had not seen it previously. She had never heard of any claim by the MacFadyen family to this land from her mother or grandfather or anyone else.
 The MacFadyens had also objected to the application for Commission consent to the decrofting of the land. However that too had been granted, on 9 December 2013 (production 3). However Mrs MacFadyen had appealed the decrofting direction decision and the house had never been built.
 Mrs Galbraith estimated that she and her husband had spent some £10,000-£11,000 pounds on things such as the deposit for the kit for their house, drawings, engineer’s report, putting a mortgage in place and so on.
 Production 27A did not accurately reflect the croft boundaries as she understood them. Nor were they the croft boundaries as the crofts were presently occupied and fenced. She showed where the current boundaries of 88 and 113 were. The boundary between them was presently fenced. As well as the fence there was stonework along the boundary dyke in the form of the remains of a stone dyke. It ran out in the direction of the hill. She herself had not been born when her grandfather had built the fence but she confirmed that in all the time that she had known of the existence of the fence there had been no opening in the fence between the two crofts until Mr Iain MacFadyen had created an opening. The fence had been continuous and had gone out in the direction of the hill.
 She did not know when her grandfather had moved back to Barra but she thought he might have been tenant of 95 before becoming the tenant of 88. Referred to production 35, she confirmed that the letter dated 23 March 1959 referred to an inspection of the croft on 18 March 1959 and contained a statement that the croft extended to 12 ½ acres of arable land with one share in the machair and one share in the hill grazing. The letter from the Department dated 19 July 1960, part of the same production, referred to a subsequent inspection on 13 July 1960, when Mr MacNeil was said to have a stock of two cows, five two year olds, two yearlings, one calf and 66 sheep on the croft. Mrs Galbraith confirmed that at present there were six cows on the croft, which was roughly equivalent to two cows and five two year olds. That was on crofts 88/1 and 88/2 combined.
 The witness was then referred to production 39, being a letter from Mr Muncaster to herself dated 18 January 2013, which refers, under reference to map 1 of three enclosed maps, to the boundaries of crofts 88/1, 113 and 95 Borve and which records that these areas had been enclosed by the tenants or their predecessors prior to the Scottish Ministers becoming landlords. The letter goes on to state that, with reference to croft 88/1, its boundaries were shown on said map “as they were confirmed as part of a division application and agreed to by the Crofting Commission following regulatory process”. She agreed with Mr Fraser that the letter confirmed acceptance by the landlords of the boundaries of croft 88/1 as stated by the Commission when granting the division application.
 In cross‑examination Mrs Galbraith confirmed that she had become the tenant of the whole croft in 2008. At that time she had not been living in Barra. Nor had she been there when the application for consent to the division of the croft was being considered. She had still been living in Glasgow.
 Referred to the second plan which is part of production 36, she confirmed that it showed the area which had been assigned to her by her grandmother. She thought that plan had come from IACS forms. The purpose of IACS forms was to enable a crofter to claim subsidy for cattle or sheep. She had heard it said that IACS forms did not necessarily delineate legal boundaries. The map had been used because it was the only map her grandmother had.
 She acknowledged that there was a difference in the acreages used by her in her application for consent to the division of the croft and those specified by the Commission when granting it.
 When the application for consent to division had been advertised, on 5 February 2009, she had not been aware of the MacFadyen family’s situation. In particular she had not been aware that Mr MacFadyen was ill. She accepted that if a member of one’s family was ill that was going to have priority over worrying about a neighbour’s regulatory application: “Family comes first”, as she put it. Had she seen an application for the division of another croft, her attitude would also have been that it had nothing to do with her.
 The 1957 plan had only come to light when she had lodged her planning application. She had seen it for the first time when Mr Muncaster sent it to her with his letter, production 22. She was pretty sure that that had been after she had submitted her planning application.
 With reference to the decrofting application, she had not been aware that the MacFadyens had requested that no decision be taken on that application until the boundary had been decided by the Land Court. She agreed that, had the Commission assented to that, she would not have lost her grant because the land would not have been decrofted.
 Referred to production 39, the letter from Mr Muncaster to the witness dated 18 January 2013, it was obvious from that letter that the MacFadyens were claiming part of croft 88 as theirs. She could not remember if she had responded to that letter. Mrs MacLellan referred to the letter’s conclusion that the boundaries as shown in the division application were being accepted as the croft boundaries by the landlords. She put it to the witness that a regulatory body such as the Crofting Commission could not fix croft boundaries. Mrs Galbraith responded that she had read something to that effect. She resisted a suggestion that the boundaries being referred to by Mr Muncaster were incorrect. These boundaries showed the way the crofts were in fact divided.
 The Nissen hut had been built in 1985 or 1986. The witness knew little about the construction of a Nissen hut, other than that it could be unbolted and moved. She imagined that a plan of its location would have been submitted with the application for a grant for its construction. She thought that the officials at the Department would match up such a plan with the plan they themselves had. However, she accepted a suggestion from Mrs MacLellan that it was quite common for crofters to apply for grants on land which was not theirs.
 Referred to production 7, Mrs Galbraith confirmed that it was not her position that that plan showed the entire boundaries of croft 113. As well as the part of the croft south of the road she thought croft 113 included the part to the north of croft 95 shown as being enclosed within the blue lines on production 7. That was the way it was fenced at the moment.
 Asked whether she was aware of the distinction between boundaries of land as occupied and legal boundaries, the witness said that she was. She agreed that it was the way the crofts had been fenced which informed her view of the boundaries.
 In re-examination Mrs Galbraith was referred to production 37, the application for consent to divide the croft, and to the areas stated in answer to question 8. She did not know where those areas had come from. She imagined they had come from the Department. With reference to the different areas stated on page 9 of the same production, she explained that they had received two maps one which showed the croft as stopping at the river (An Abhainn Mhòr shown on most of the map productions) and one which showed it going beyond the river.
Donald Joseph MacLean (55)
 Mr MacLean gave evidence that he was the Township Clerk of Borve, having been elected in 2011. This was his third stint in that post. He had been invited to give evidence in relation to a plan which had been lodged. Shown production 27A he confirmed that he had received a copy of it some two weeks previously. He had been asked if he had any plans of the township of Borve. He had none and he had asked people a lot older than himself if they had ever seen a township plan and they had not.
 He had been born in Borve and lived there all his life. Production 27A was far from accurate. Some of it was right but a lot of it was wrong. For example, his own croft was wrong; it showed him as having land on the other side of the road which he did not have. The entries for crofts 75, 76, 78, 82, 94 and 102 were all wrong. If reliance was to be placed on this map it could open up “a whole can of worms in the township” and a lot of people were worried about it. At present everyone’s croft was fenced. Some crofts were not fenced all the way to the Craigston border. Some were fenced only to the river. He was at a loss to understand where production 27 had originated. The oldest tenant in the village had never seen it.
 He could not comment specifically on the boundaries of crofts 88 and 113 as shown on that map.
 In cross‑examination he confirmed that there was in existence a Grazing Committee. There would be no minute to the effect that he was authorised to represent the Committee at this hearing. There were four on the committee; himself as Clerk, James Ferguson as Chairman, Alasdair MacNeil and Donald Archie MacNeil, although he seemed uncertain about whether the last mentioned was in fact a committee member. Before attending the hearing he had asked the Chairman and the other two. They had been happy for him to give evidence.
 He was not here to take sides. Asked whether he was here as a crofter in the township, he said that he was here as Grazings Clerk. Asked whether his evidence would have been any different if he was here as an individual, he could see no reason why it should be different. He was here because it concerned him that this map (production 27) was wrong. He was here in a duly authorised role as Grazings Clerk.
 It had been Michelle Galbraith who had given him the map. She had phoned him to ask if he had any maps of the township. That had been roughly two weeks ago.
 It was a fact that he was a business partner of Donald Archie MacNeil; they operated a fishing boat together.
 Mr MacLean denied having had a conversation with Mrs MacFadyen about house plans at Borve. They might have had a conversation at some point to do with her Nissen hut or agricultural shed. (This is probably a reference to Mrs Galbraith or Mrs Macneil, rather than Mrs MacFadyen.)
 Asked whether he considered the plan passed to him by Mrs Galbraith to be incorrect, he was adamant that he knew it to be incorrect. He knew his own boundaries, having been told about those by his father and the boundaries for his croft as shown on that plan were wrong. He was not prepared to say that something was correct when he knew it to be incorrect.
 Asked about the inaccuracies he had referred to and whether boundaries might have changed over the passage of time, he pointed out that the map was dated 1957, that he himself had been born in 1959 and that those changes had not taken place in his lifetime. He accepted that much of the plan was correct but he could not comment as to whether the boundaries of crofts 88 and 113 were correct or incorrect. He did know that the late Angus MacNeil had worked part of the land in question but nothing else. The boundaries of crofts 88 and 113 might well be correct on production 27; he just did not know. Asked to point out the croft boundaries which were correct, Mr MacLean referred to Anne Ferguson’s at 106, Angus John MacNeil’s at 91, James Houston’s at 90 and a croft belonging to a Mr MacInnes at 84. These were all correct.
 He accepted that it was possible that crofters had come to an agreement in terms of exchanging land. He referred to an agreement between himself and an Ann Galbraith in relation to his own croft. He and Donald MacNeil had swapped bits of land. That agreement had never been put on paper. But they each knew what belonged to them. There could have been a similar arrangement between 88 and 113 before his time. When he was young he had helped Angus MacNeil taking haystacks to the area of the Nissen hut. He could not, however, comment upon what the arrangement was whereby Angus MacNeil cultivated the land in question or occupied the land on which the Nissen hut was built.
 Asked by the court as to whether such informal arrangements were common or rare in Borve in his lifetime, he said that they were very rare.
Mrs Mary MacNeil (82)
 Mrs MacNeil, Mrs Galbraith’s grandmother, then gave evidence.
 She had been the tenant of the whole of 88 Borve. She had come to Borve in 1959. At that time the croft had been tenanted by her father-in-law, Mr Michael MacNeil. He had been succeeded in the tenancy by her husband. They had returned to live in Barra. They had built a house on the croft. It had been built in or around 1960. She had lived there ever since. They had kept cattle on the croft. She herself had looked after them because her husband was away at sea.
 Her father-in-law had also worked croft 95 and her husband had succeeded him as tenant of that croft also, before assigning it to their son, Donald Archie MacNeil.
 She could point out the boundary between crofts 113 and 88. There was a fence along that boundary. It had been erected by her husband. It was referred to as “the IDP fence” (a reference to the Integrated Development Programme in operation in the Western Isles at that time). It had been erected in the 1980s. It followed the line of an old dyke which was still visible. The IDP fence had been preceded by another fence which was already old when she had come to the croft. There had been no gates between the two crofts in the old fence, nor in the new fence.
 When she had come to 88 Borve in 1959 the tenant of 113 had been Neil MacFadyen, Roderick MacFadyen’s father. The house which Neil MacFadyen had occupied was no longer there. She and her husband had had a good relationship with Neil MacFadyen; he and his wife had been nice people and they had visited each other. Similarly Roderick MacFadyen had been very friendly with her husband and it had in fact been her husband who had first taken Roderick away to sea. Roderick had taken over his father’s croft and had kept sheep but not cattle. Roderick had built a new house near the machair. Roderick’s stock had been kept on the lower side of the road (north of the road), between the road and the burn. He had not kept any stock on the part of the croft on the other side of the road.
 Shown production 27A, she was not sure if she had seen it before. Shown production 5, she was clear that the three indentations shaded pink shown extending across croft 88 were not part of 113. She had never heard of croft 113 having those bits of land. She had been there for 56 years. Neither Neil nor Roderick MacFadyen had ever mentioned it to her. When her husband had built the fence from the road to the hill he had enclosed and occupied these areas. He had kept cattle in them. On the north side of the road, between the road and Craigston, where Michelle now had her croft, that was still used for grazing by their cattle.
 She denied a suggestion that there had been an agreement to occupy part of 113 in exchange for a supply of milk to the MacFadyen family. She had never given milk to anyone because she had a large family and in any event she did not think Mr and Mrs MacFadyen took milk.
 In cross‑examination Mrs MacNeil confirmed having lived in Borve since 1959 and the friendship between her husband and Roderick MacFadyen.
 Referred to production 37, she could not say who had prepared the plans included with that application for consent to divide the croft. She could not remember having given Mrs Galbraith these plans, nor indeed was she sure that she had ever seen them. Referred again to production 27A she still could not remember having seen it before.
 In re-examination she confirmed that her memory of older things was better than for recent things.
For the applicant
 Mrs MacLellan submitted that the evidence supported Mrs MacFadyen’s case. It was Mrs MacFadyen’s position that an area of her croft had been decrofted by a neighbour. There were certainly complicating factors in the background to the case, in the way of informal arrangements between crofters, but the real confusion had set in with Mrs Galbraith’s division application. It had been that application which had first, wrongly, included part of 113 as part of 88. That misrepresented what had been a friendly arrangement between Angus MacNeil and Roderick MacFadyen. However it had caused the landlords to say that they could not interfere with what the Crofting Commission had approved. However the original croft boundaries were now clear. They were as shown on the 1957 plan. Mr Muncaster’s letter, production 22, acknowledged these boundaries as the ones now recognised by the landlords as the original boundaries.
 Asked by the court as to why Mr Muncaster had not been called to give evidence, given the importance of the 1957 plan, Mrs MacLellan, as we understood her, said that she might have called him had Mr Fraser not been instructed so late in the day. This may be a reference to a need to respond to productions 30 to 43 which Mr Fraser had, with leave of the court and of consent, lodged late. In any event, in Mrs MacLellan’s submission, the evidence was clear that as far back as 1957 croft 113 had included the indents under discussion. Mr MacLean’s evidence as to informal arrangements and some croft boundaries being shown correctly and others incorrectly, in terms of how the crofts were now occupied, was compelling.
 The documentation from the Crofting Commission was carefully worded, reflecting what the Commission accepted as being the boundaries but avoiding any suggestion of a decision by the Commission to the effect that they were in fact the true boundaries. Such a decision would be ultra vires of the Commission.
 If the court were to accept that the indented areas had ever been part of 113 then they remained part of 113 because there had been no formal process to change their status.
 Mrs MacLellan referred to a number of authorities. Newlands v Assynt Trust Limited 2008 SLCR 276 was a decision of the Divisional Court holding that, although crofters could come to informal arrangements between themselves, for these to have any legal standing they required at least the consent of the landlord. The case also made the point that prescription did not operate in crofting law. That was of some significance here; whether an area of land had been occupied for 60 years or 160 years, it did not result in the acquisition of rights over the land.
 MacAskill v Taylor 1970 SLCR App 34 made the point that simply because rents were equal that did not mean that croft areas were equal.
 When applying for consent to divide the croft, Mrs Galbraith had relied upon an IACS plan but an IACS plan conferred no legal rights to the ground. Similarly occupation conferred no legal rights and Mr MacLean had been clear that his evidence was to do with occupational boundaries. The 1957 plan was better than an IACS plan. The case of Morrison v Murdoch 1997 SLT 381 was authority for the principle that the effect of crofting legislation was to prevent a tenancy of a croft being set up without the Crofters Commission’s consent. So occupation alone was not enough.
 Mrs MacLellan referred to the entries for crofts 113 and 95 in the Register of Crofts, showing them to be of the same extent. Mrs Galbraith’s submission seemed to be that if Mrs MacFadyen was right the area stated for 113 in the Register of Crofts would have been very much greater. But in Mrs MacLellan’s submission the entry in the Register of Crofts was wrong. The Register of Crofts was, in any event, nothing more than a useful database of information. It could be a help or a hindrance depending on its accuracy. Some entries in the Register were the results of regulatory decisions made by the Commission whilst others were based on information received from third parties. Where the Commission had made decisions over matters within their control, such as assignations of crofts, the information contained in the Register was reliable but in other areas, such as determining the extent of crofts, where the Commission had no regulatory role and depended on information received from third parties, the entries could be unreliable. The case of Ross v The Master of Lovat 1978 SLCR App 45 at page 50 made clear that, whatever information was contained in the Register, it could have no bearing on the status of a holding, its extent, rent or any such matter. So entries in the Register of Crofts had to be treated with a degree of caution.
 In answer to a question from the court Mrs MacLellan confirmed that her case was predicated largely on production 27 and on the understanding of Iain Roderick MacFadyen and Margaret Ann MacIntyre. She accepted that there was no evidence of the land having been occupied conform to production 27. She concluded her submissions by inviting us to determine the boundaries of croft 113 in terms of productions 5 and 27.
For the respondents
 Mr Fraser dealt with the history of matters. The extract from the proceedings before the Napier Commission, production 30, told us that Borve had been formed in the 1820s, originally with 14 crofts but with 36 crofting families. By the time of the 1890 fair rent application, however, there were 33 crofts in existence. There had been a change of numbering between 1890 and 1916 but the extent of croft 88 in 1916 was the same as it had been in 1890. It appeared that a very detailed measurement had been prepared in 1890 but nobody had thought to draw a plan.
 This was no ordinary boundary application. Instead it involved an attempt to dismember a croft as it had been occupied by the MacNeil family for a period of over 50 years. If the applicant was correct something like 4.5 acres of the original 10+ acres would be part of 113 and we would be left with the anomaly of a linear croft which was no longer a linear croft because it was divided into four segments (by the three indentations) with no obvious way of making a connection between them.
 Returning to the history of matters, it was seen (from production 33), that the stock held on 113 at that time was very small; a horse a cow and four sheep.
 The 1959 extract from the Register of Crofts (part of production 35) stated the total extent of croft 88 to be 6 acres 3 roods and 12 poles. Mr Fraser submitted that this was a mistake and that what had happened was that the whole area as referred to in production 32 had not been carried forward, because when one looked at the Ground Officer’s report of 23 March 1959 (also part of production 35), following inspection, the extent of croft 88 was said to be 12½ acres. The Ground Officer was not an inexperienced crofter or a landlord who knew little about the land; he was somebody who knew what he was talking about.
 Mr Fraser next referred to the 1985 application to fix a fair rent, production 34, although we think there was some confusion in his submission as to whether the extent showing her croft 88 was in hectares, as he seemed to submit, or in acres, as it actually is. The production shows the croft as extending to 2.75 acres arable and 4.075 acres of outrun. Mr Fraser thought that the 1985 application overstated the extent of the croft but, if his submission in relation to the Ground Officer’s 1959 measurement is correct, it in fact understates it, at 6.825 acres rather than 12.5 acres.
 Dealing with the application for consent to divide the croft, Mr Fraser submitted that what had probably happened had been that the Commission had probably corrected those acreages by reference to its own records. In any event the total shown of 5.206 ha shown for the two crofts, 88/1 and 88/2, equated to 12.86 acres which was not terribly far from what the Ground Officer had claimed in 1959. The extent of the three indents claimed as part of croft 113 was 4.52 acres.
 Mr Fraser then referred to production 42A, the 1880 OS map. The boundary shown as the east boundary of croft 88 was a hard boundary. The court would see it for itself at inspection. The IDP fence and the fence which it had replaced had both followed the line of that boundary. The fence which the IDP fence had replaced was already an old fence when Mrs MacNeil had come to the croft in 1959. It had also been a continuous fence without a gate or gap.
 It was not for Mrs Galbraith to explain why croft 113 seemed to have become smaller than it had been in the past but it was for her to defend croft 88 as it had been possessed by her family for some 60 years. Mrs MacLellan was right to say that occupation could not confer rights of tenancy but occupation was very strong evidence as to what the boundary could be. Croft 88 was now as it had always been. It was possessed by Mrs Galbraith and her grandmother as their predecessors had possessed it. The onus was on someone challenging that evidence. What challenge was there here? Were it not for the mysterious map (production 27A) we would not be here today. There would be no challenge because the challenge was predicated on that map. The map was without provenance. There was no information as to who had prepared it and why. It contained no measurements, was schematic only and until recently had been unknown within the township. His clients were absolutely clear that it was inaccurate in relation to crofts 88/1 and 88/2 and Mr MacLean’s evidence showed it to be inaccurate in respect of several other crofts. The applicant had entirely failed to discharge the onus upon her and the application was without foundation. Mrs Galbraith’s various regulatory applications had been made entirely in good faith, honestly and openly. The first three of them had been unchallenged. Mr Fraser appreciated that the circumstances of the MacFadyen family at that time had to be taken into account but the purpose of advertisement was to bring applications to the attention of people and, in any event, those things quickly became “the talk of steamie”. Yet, for whatever reason, no challenge to those applications had been brought. We should find the boundary between crofts 113 and 88 to be a continuous straight line boundary and hold that the three indents did not form part of croft 113.
 We take each chapter of evidence separately, beginning with the documentary evidence.
(a) Production 27
 The applicant’s case is predicated to a very large extent on the map which forms production 27 together with the key to that map, production 28, so it makes sense to begin with it.
 We accept the evidence of Mr MacFadyen and Mrs MacIntyre that there has been a copy of that map in their family home for many years. But that on its own means very little: only that the map was in existence and in the MacFadyen family’s possession since perhaps the early 70s (which would have been Mr MacFadyen’s earliest recollection). As to where it had come from, who had prepared it, when and for what purpose, it tells us nothing. Mrs MacIntyre said that we would have to ask her mother but Mrs MacFadyen did not give evidence.
 We know that the copy of the map which has been lodged in support of the application came from the Scottish Government Rural Payments and Inspections Directorate (as “the Department” came to be called) and was sent to Mrs MacFadyen under cover of Mr Muncaster’s letter of 15 February 2012. In that letter it is referred to as “some information that has come to light” but how and where Mr Muncaster discovered it is not explained in his letter and has not been explored in evidence before us. We find that puzzling because it is such an important matter from the applicant’s point of view. If a party seeks to rely on a document it is necessary to set up that document by leading appropriate evidence as to its provenance and status. The provenance of this map is not explained and, so far as its status is concerned, it does not seem to us that Mr Muncaster, on behalf of the landlords, is positively asserting these to be the correct croft boundaries, standing (a) the terms of production 22, which don’t go that far, and (b) the terms of his later letter to Mrs Galbraith, production 39 dated 18 January 2013, in which he discusses various other possibilities by way of solution to the boundary problems.
 In passing, it is also curious that Mr Muncaster described the map as “some information that has come to light” if he was already aware, from his meeting with Mrs MacFadyen and Mrs MacIntyre, that Mrs MacFadyen was familiar with it. But we need attach no great weight to that.
 Given that paucity of evidence as to the map’s provenance, what can be made of it?
 Firstly, it is a fairly painstaking piece of work. Whilst it does not contain measurements and, therefore, does not purport to show precise croft boundaries, whoever prepared it has gone to a great deal of trouble. It must, therefore, have been produced for a purpose.
 Secondly, it seems to us that its most obvious purpose would be to show how the land was occupied in 1957, although why that was being looked at, and by whom, is unknown. We think it unlikely that it represents someone’s idea for the rearrangement of the township. If starting from scratch, one would hardly come up with such a complex matrix of occupation.
 Thirdly, however, in considering whether it probably shows how the land was occupied in 1957, we have to take into account the extent to which it is at odds with current occupancy. It would require a lot of informal swapping of land to have taken place to reconcile the map with present de facto boundaries. Mr MacLean was able to point to numerous discrepancies between the map and present usage and said that informal swapping of land in the township was rare.
 Fourthly, as to its status as a “township map” widely known about and historically regarded as authoritative, his evidence was that he had never seen it before, nor had anyone, including tenants older than himself whom he had asked about it. Mr MacLean’s evidence on this deserves great weight since his is now serving his third spell as Township Clerk. If anyone could be expected to know about the existence of a township map he could.
 Therefore, although on the face of it the most likely purpose of the map is to show 1957 occupancy, we are unable to make a finding in fact to that effect for the foregoing reasons. But even if we were able to make such a finding in fact, how far would it take us? That the land was occupied in that way in 1957 would not necessarily mean that the boundaries shown are the true historical boundaries. Indeed, if there has been a history of informal swapping of areas of land in Borve, as seems to have been the case, then it almost certainly does not show the original (as at 1886) croft boundaries.
 The result of all of that is that we cannot put any reliance on this production as representing the true boundaries of croft 113. What it does do, of course, is provide a basis for the MacFadyen family’s belief (including the late Roderick MacFadyen’s belief) that the true boundaries of crofts 88, 113 and 95 are as claimed in this application. The trouble is that, due to lack of other evidence, that basis is not very strong.
(b) Other documentary evidence
(i) The 1880 OS map
 The earliest documentary evidence is the copy of the OS 1880 map showing Borve, production 42A. It shows a series of long, linear units, running side by side, their boundaries running parallel with each other and extending to both sides of the public road. It was accepted by Mr MacFadyen that the lines identified by Mr Fraser on his projected image of this production overlaid onto production 16, the aerial photograph of Borve, were the western and eastern boundaries of croft 88 (subject, of course, to the indentations said to be part of 113). We are not concerned with the western one; it is the eastern one which marches with croft 113. It is a continuous line on both sides of the road. The superimposition of production 42A onto the aerial photograph shows that boundary to be a dyke which is still visible today, as we saw for ourselves at inspection. There is no indication on this map of any indentations having been carved out of what is now croft 88 to form part of 113. Although the dyke is now broken down, neither is there any indication of any openings having existed along its length. Nor was it suggested in evidence that there ever had been. Production 42A is, therefore, powerful evidence of the crofts having been de facto separated along that dyke in 1878, when the map was surveyed.
(ii) Excerpt from Napier Commission Evidence
 Production 30, comprising two pages of transcript from an undated session of the Napier Commission taking evidence in Barra, was not referred to in the evidence, only by Mr Fraser in submissions. Nevertheless its content is uncontroversial and its only evidential value is the evidence of the witness, Mr Michael Buchanan, as to the township of Borve having been formed some 60 years previously, enabling us to date it to the 1820s; that it had at one time operated as a runrig system; that it had then been arranged into, first, 14 and then 18 crofts; that these crofts were occupied by some 36 crofting families as well as an unknown number of cottars; and that the township had seen an influx of 19 “men” (which may be a reference to crofters with families) from other townships as a result of clearances.
(iii) Schedule of Fair Rents prepared by Crofters Commission in 1890
 By 1890 Borve consists of 33 crofts, perhaps due the influx of people from other townships spoken to by Alexander Buchanan in front of the Napier Commission. Mr Fraser referred Mr MacFadyen to the entry for Alexander Buchanan, tenant of what was then ¾ 8 Borve, with a total holding of 10 acres, 2 roods and 4 poles. Mr MacFadyen seemed to accept that this was a reference to what has become 88 Borve by the date of the next production.
 Although it was not brought out in evidence this production (which is an extract from the Crofters Commission’s Annual Report for 1890) also contains an entry, number 460, for a Neil MacFadyen at ½ 9 and ¼ 8 Borve with a total holding of 10 acres 1 rood 36 poles. Given the name MacFadyen and the (presumed) proximity of ¼8 which he tenanted to the ¾8 which Alexander Buchanan tenanted taken together with the renumbering shown in the next production, we may assume that this entry is for what became 113 Borve.
(iv) The 1916 Land Court Fair Rents application
 The relevant documentation is at production 32. It gives the total acreage of 113, now tenanted by Donald MacFadyen, as 10 acres 1 rood and 36 poles, the same as that shown for Neil MacFadyen’s holding in 1890; that of 95 as 7 acres 3 roods and 15 poles and that of 88, now tenanted by a James MacDonald in place of Alexander Buchanan, as the same as shown in 1890.
(v) The 1925 Land Court application by Donald MacFadyen for a revaluation of croft 113
 Production 33 shows Mr Donald MacFadyen, grandfather of the late Roderick MacFadyen and great-grandfather of Iain Roderick MacFadyen applying to this court for a reduction of his rent due to flooding of the burn running through his holding and deterioration of the peat moss. He also asks for his souming of 1 horse, 1 cow and 4 sheep to be doubled. His application is refused and the rent and souming remain as they were. Nothing is said about the extent of the croft but the small souming and the fact that it was not changed (although the court gives no reasoning for this) does not suggest a large croft.
(vi) Correspondence from Department of Agriculture for Scotland in 1959-1960
 This is production 35. It comprises a number of letters relating to the proposed assignation of 88 Borve by Roderick Macdonald (not, surprisingly, Michael MacNeil) to Angus MacNeil. The first is a letter of 23 March 1959 from the Department to a Mr Rasche and the Crofters Commission. It refers to an inspection having been carried out on 18 March of that year. It says the croft extends to 12½ acres arable and outrun although an extract from the Registers of Crofts Return made by the landlord’s agents to the Crofters Commission shows a total of 6 acres 3 roods and 12 poles. That return states the souming as 2 horses, 2 cows and followers. By this time Mr MacNeil was already the tenant of 95 Borve, which is said to extend to 6.5 acres arable and outrun with the same shares in machair and hill. The application was granted by the Commission on 22 April 1959.
(vii) The 1985 application by MacNeil of Barra to fix fair rents
 This is production 34. Croft 88 is given an acreage of 6.825, number 95 an acreage of 7.1 (both tenanted by Angus MacNeil) and 113, now tenanted by Roderick MacFadyen, is said to be 6.931 acres.
(viii) Mrs MacNeill’s and Mrs Galbraith’s regulatory applications
 Production 36 is the application for Crofters Commission consent to the assignation by Mrs MacNeil of the tenancy of 88 to Michelle MacNeil, as Mrs Galbraith then was. It includes a copy of a document giving the holding’s Scottish Office Farm Code number and an area of over 6 hectares but it is not clear what this includes.
 Mrs Galbraith’s application for consent to divide the croft is production 37. It gives the extent of the existing croft as 4.8 ha and of the two new units as 2.88 ha (88/2) and 1.92 ha (88/1). Mrs Galbraith was uncertain as to where she had got these figures. However, the documentation includes a “Notification Form – Agreement with Landlord” which states the revised extent of 88/1 to be 1.658 ha and that of 88/2 to be 3.548 ha, a total for the original croft of 5.206 ha or 12.86 acres.
 Production 38 is Mrs Galbraith’s application for Commission consent to the assignation of 88/2 Borve back to her grandmother. It does not state any extent for the croft.
(ix) Current entries in the Register of Crofts
 Productions 12 (also production 23) and 13 (also production 25) are the current Crofting Commission Register of Crofts entries for, respectively, 113 Borve, stating an area of 2.87 ha (7.09 acres), and 88/1, stating an area of 1.522 ha (3.76 acres). Production 24 is the current entry for croft 95 and gives an area of 2.87 ha, the same as 113.
 The productions include, at production 40, a copy of a letter dated 8 October 2013 from Ms Susan Walker, Convener of the Crofting Commission, to Niall MacPherson and Ian Muncaster at the Department’s Balivanich office, which gives mapped hectorages for 88/1 Borve of 1.159 ha (2.86 acres), for 88/2 of 3.499 ha (8.65 acres), for 113 of 2.365 ha (5.84 acres) and for 95 of 2.263 ha (5.59 acres).
Conclusions from this documentary evidence
 It is difficult to make much of these figures, given their fluctuation over the years, but one thing which stands out is the reduction of the size of croft 113. It diminishes from 10 acres 1 rood 36 poles in 1890 to 7.09 acres in the Register of Crofts or 5.84 acres as plotted by Commission staff for the purposes of Mrs Walker’s letter, which plotting exercise seems to have included ground at the bottom of croft 95 as part of 113.
 Whatever the explanation for that reduction, it does not seem to be that part of the croft has been added to 88. That is because the two crofts were originally of similar size and to transfer almost five acres (the approximate total area of the indentations) from 113 to 88 would have resulted in 88 now being of the order of 15 acres, which, on any view, it is not.
 The absence of a map showing where the original 10 or so acres of 113 were is critical. The result of our analysis is, therefore, inconclusive and of no assistance to us in the matter of fixing the present boundaries of 113.
(c) The evidence of Mr MacFadyen and Mrs MacIntyre
 Mr MacFadyen gave evidence that his father had shown him the boundaries of 113 and that they were as shown on production 27A. Again we accept that evidence. There is no reason to disbelieve Mr MacFadyen on that. It is consistent with his evidence about his family’s possession of the map. He could be wrong about both of course, in which case he would be lying because this is not something about which someone could make an honest mistake. His evidence about the family’s possession of the map is supported by his sister but there is also another consideration which supports his veracity on this matter. That is that we think it unlikely that anyone who never had an inkling that this land was part of his family’s croft – to whom that came as complete news with Mr Muncaster’s letter of 15 February 2012 – would object to Mrs Galbraith building her house on land which has been in her family’s possession for at least 30 years. If there was no understanding on the part of the MacFadyen family that it was theirs and no expectation that they would ever have any right or reason to occupy it, it is difficult to understand why Mr MacFadyen would imperil a friendship between the two families which goes back several generations for the windfall gain apparently disclosed, although by no means guaranteed, by the 1957 map. In other words we find his opposition to Mrs Galbraith’s proposed housebuilding more understandable in the context of a long held bona fide belief that the land on which the Nissen hut stands was part of his family’s croft. We therefore hold that this has been his – and Mrs MacIntyre’s – genuinely held belief for a long period of time, going back to their childhood.
(d) Occupation and usage
 It is clear from the evidence that the two indentations south of the road have been occupied and used solely by the successive tenants of croft 88, to the exclusion of use by the tenants of 113, since at least 1959, when Mrs MacNeil came to 88 Borve. Furthermore they have been used as integral parts of 88, with open access to the rest of 88. Not only that but Mr and Mrs MacNeil used the part of 113 south of the road for their cattle. Except at the hill end of the crofts, where there is no boundary, a continuous dyke and, later, continuous fences run along the whole eastern extremity of croft 88 south of the road. The dyke is shown on the 1880 OS map, so has been there since the passing of the Crofters Holdings (Scotland) Act 1886.
 The position with the indent north of the road, where the Nissen hut stands, may be marginally less clear, since the hut was erected only in or around 1985 (hence our reference to 30 years at para 150 above). But it too has been separated from 113 by a continuous dyke or fence.
 In summary, therefore, all of the evidence of historic occupation and use (save, perhaps for the 1957 map) is unequivocally in favour of the land on each side of the dyke/fences to which we have referred belonging to the two separate crofts. No one remembers the indentations having been worked as part of 113 and indeed it is difficult to see how they could have been, given the existence of the dyke. This is compelling evidence against the indentations – which, in truth, never seem to have had an existence or use independent of 88 – being part of 113.
(e) The area to the north of croft 95
 It is agreed that the late Roderick MacFadyen used the land to the north of croft 95, from the fence along the north side of the Abhainn Mhòr to the Craigston boundary, for sheep. There is no fence between it and that part of 113 which is north of the road, so his sheep had access to that whole area. The respondents say this is all part of 113 but the applicant and her son deny that, saying it is part of 95 and that the only basis on which the MacFadyen family have occupied it is the agreement between Roderick MacFadyen and Angus MacNeil by virtue of which Mr MacNeil occupied the Nissen hut area.
 Looking at production 42A, the boundary of what appears to be 95 extends all the way to the Craigston (Balnacraig as it appears on the map) boundary. Looking at production 27, for what it is worth, although the area in question has been shaded green by the applicant or her agent prior to lodging it, the document in its original form shows no distinction between this area and (the rest of) croft 95. For these reasons, therefore, we hold that this area is part of 95.
 That means that we accept Mr MacFadyen’s evidence that it was occupied by the MacFadyen family not as a matter of right but on the basis of Angus Macneil’s consent. That, of course, supports the MacFadyen side of the argument. They say that they were allowed to use it in return for the MacNeils using the Nissen hut area. But that is not necessarily so. Although Mr MacFadyen gave evidence that Mr Donald John MacNeil of croft 107 at one time used the part of 113 south of the road, we also heard (from Mr MacFadyen and Mrs MacIntyre) that Michael and, later, Angus MacNeil had use of that part, which would be a more than adequate exchange for the area at the north of croft 95. What appears to have happened, in fact, is that Neil and, later, Roderick MacFadyen were content to confine their crofting operations to the area north of the road, including the northmost part of 95, leaving everything south of the road to the MacNeil’s, subject to whatever use Donald John MacNeil had of it and for how long.
 This has been a very difficult case to decide. On the one hand it is plain that there was an understanding between the MacNeil and MacFadyen families to use parts of each other’s crofts. They were not alone in that in Borve. Donald Joseph MacLean, although he described the practice as very rare, gave the example of an exchange he himself is involved in and we do not doubt that there are others. But the people who could speak to the relevant agreement in this case are no longer with us, so the information is second hand. Mrs MacNeil’s time in Borve goes back to 1960, before Ian Roderick MacFadyen was born, and she denies all knowledge of any arrangement whereby Neil or Roderick MacFadyen allowed the MacNeils to use land in return for milk.
 In that uncertain state of the evidence regarding agreements, with the witnesses contradicting each other, we have to rely on the objective evidence in the case, that is to say the historical documentation and evidence of land use. In terms of historical documentation, the 1880 OS map, with its pattern of continuous parallel lines separating the crofts, is persuasive and corroborated by the stone dykes which run along these lines to this day. That, as we have already said, is powerful evidence as to how matters were at the dawn of crofting as a distinct legal regime. It is corroborated by the use actually made of the land, with there being no evidence that the MacFadyens and their predecessors have ever occupied any of the indentations shown on production 5. Furthermore, there is no logical justification for such a division. This is not an area in which land quality varies so much that it has to be parcelled out in small patches in an effort to give everyone a share of the better land. Instead, as we saw at inspection, it is much of a muchness. So we see no reason why croft 88 should be divided into four unconnected and unconnectable parts.
 We have therefore found that the boundaries of croft 113 are as shown on the map attached to this order.
 Following our usual practice, we have allowed 21 days for motions and submissions on expenses.
Effect of this decision on appeal against decrofting
This case arose out of Mrs MacFadyen’s appeal against the Crofting Commission’s decision of 9 December 2013 granting a decrofting direction in respect of Mrs Galbraith’s proposed house site. The basis of that appeal (RN SLC/9/14) was that the decrofted land was part of 113 Borve rather than 88/1. The decision in the present case is to contrary effect. The court therefore intends to pronounce an award dismissing said appeal. That order will be intimated shortly.