(Sheriff MacLeod, D J Houston)
(Application RN SLC/174/08 – Order of 29 July 2009)
AGRICULTURAL HOLDINGS – APPLICATION UNDER SEC 39(2) OF AGRICUTURAL HOLDINGS (SCOTLAND) ACT 1991 SEEKING APPROVAL FOR INSTALLATION OF VOLUNTARY MILKING SYSTEM ON DAIRY FARM – WHETHER ADEQUATE NOTICE GIVEN IN TERMS OF SEC 38(1) – WHETHER PROPOSED IMPROVEMENTS REASONABLE AND DESIRABLE FOR EFFICIENT MANAGEMENT OF THE HOLDING – WHETHER PROVISIONS IN LEASE RESTRICTING USE OF PART OF SUBJECTS TO USE AS “LIVESTOCK AND CROPPING UNITS” EXCLUDED GRAZING AND CROPPING FOR DAIRY STOCK
The applicants were tenants of an agricultural holding comprising several units which they ran collectively as a dairy farm. Wishing to install a new milking system known as a “Voluntary Milking System” comprising a number of robotic milking units, they served notice on their landlords under sec 38(1) of the Agricultural Holdings (Scotland) Act 1991 of their intention to do so. The landlords having served a counter-notice under sec 39(1) of the Act, the applicants applied to the Court for an order under sec 39(2) approving the carrying out of the proposed improvement. In opposing the application the landlords argued (i) that the notice served under sec 38(1) was so deficient in its description of what work was to be done and how it was to be carried out as not to be a valid notice for the purposes of the Act, and (ii) that the improvements, which were estimated to cost in the region of £400,000, should not in any event be approved because they would lead to the over-capitalisation of the holding. As a subsidiary issue they also argued that a clause in the lease stating that two of the units making up the holding were to be used “as Livestock and cropping units” meant that these units could not be used for the grazing of dairy stock or the taking of crops as feed for dairy cattle elsewhere on the holding and that these units could not therefore be taken into account in assessing the carrying capacity of the holding for dairy purposes.
The Court held, (1) that although said notice was deficient in certain respects, most notably in failing to state the number of robotic units intended to be installed, the defects were not such as to render it invalid, (2) that the said terms of the lease did not preclude the use of the units referred to for the purposes of grazing dairy stock and growing crops as feed for dairy stock, and (3) that, although not the only means available to the applicants of improving the efficiency of their diary enterprise and although what was proposed was the more risky of the options open to the applicants from a financial point of view, what was proposed nevertheless satisfied the test of being reasonable and desirable on agricultural grounds for the efficient management of the holding (statement of the relevant test in these terms in Gill para 26.16 applied) and approval granted in terms of sec 39(2) of the 1991 Act.
The Note appended to the Court’s Order was in the following terms:-
 Robert and Mary Whiteford have been tenants, as general partners in the limited partnership of “Messrs R & M Whiteford”, of “the farms of Abbey, Holm, McWhanrick and Crossleys (part) … collectively known as ‘Abbey Farm’ and also the holdings of Broomrigg … and Carnsalloch, Kirkton in the Nithsdale District of the Dumfries and Galloway Region” since 11 November 1987 in terms of a written lease (production 11) executed on 15 November 1993 and 20 January and 18 February 1994. Their landlords are the Trustees for the Cowhill Trust and the lease is for a period of 40 years from 11 November 1987.
 In this application Mr and Mrs Whiteford, as general partners aforesaid, ask us to grant approval to the carrying out of certain proposed improvements in respect of which they served notice on the respondents under sec 38(1) of the Agricultural Holdings (Scotland) Act 1991 (“the Act” or “the 1991 Act”) on 28 July 2008. The respondents having served a counter-notice objecting to the proposed improvements on 26 August 2008, the consent of this Court is required, in terms of sec 39 of the Act, if the tenants are to be eligible for compensation for said improvements at waygoing.
 We heard proof on the application at Dumfries on 1 to 4 June 2009 when the applicants were represented by Mr Hamish Lean, solicitor, and the respondents by Mr Lewis Kermack, solicitor. Closing submissions were heard in Edinburgh on 9 June.
 The background to the application is as follows:-
(a) Mr and Mrs Whiteford came to Dumfries from Cumbria where Mr Whiteford had a family background in dairy farming.
(b) With the exception of a period of approximately three years (2001 to 2004) following the Foot and Mouth epidemic, the subjects have been operated almost entirely as a dairy farm.
(c) At present the applicants have a herd of 380 to 400 Holstein cows, between 330 and 350 of which are in milk at any one time. Despite a reference in the application (Statement of Fact 3) to the herd size being increased to 500 to 600 animals, Mr Whiteford disavowed any such intention in his evidence and stated categorically that it was not his intention to increase the size of the herd beyond its present number.
(d) Whilst the notice of 28 July 2008 is referred to for its whole terms, the improvements which we are asked to approve comprise, essentially, the extension of the present main cubicle shed so as to accommodate a Voluntary Milking System (“VMS”). The total projected cost, as stated in said notice, is £600,000.
(e) In terms of a covering letter which accompanied their formal notice of objection, the ground upon which the respondents object to these improvements is that they would “over equip the farm for dairy use relative to the extent to which it is specified and intended in the lease”.
(f) Clause First of the lease includes the following provision:-
“The farms and Lands of Abbey, Holm, McWhanrick and Crossleys (part) to the extent that they are comprised within the Lease shall be farmed as a Dairy, Livestock and Cropping unit only. Said holdings of Broomrigg and Carnsalloch shall be farmed as Livestock and cropping units.”
The Agricultural Holdings (Scotland) Act 1991, Part IV, secs 34(1), 36(1), 38(1), 39(1) and (2)
Barber v McDouall 1914 S.C. 844
Buccleuch Estates Ltd v Telfer Unreported, SLC/141/07. 15 February 2008;  CSIH 31, 18 March 2009
Fotheringham v Fothringham 1978 SLCR App. 144
Gibson v Milne 2006 1 SLCR 87
Hutcheson v Wolfe-Murray 1980 SLCR App. 112
MacKinnon v Arran Estate Trust 1988 SLCR 32
Morrison-Low v Howison 1961 S.L.T. (Sh. Ct.) 53
Renwick v Rogers (No 1) 1986 SLCR 126
Renwick v Rogers (No 2) 1988 S.L.T. (Land Ct.) 32
Strachan’s Trustee v Harding 1990 S.L.T. (Land Ct.) 6
Taylor v Burnett’s Trustees 1966 SLCR App. 139
The Hon. Lord Gill, The Law of Agricultural Holdings in Scotland, 3rd ed.
 For the applicants, evidence was given by Mr Whiteford, Mr Bryn Davies, an agricultural consultant who has been advising Mr and Mrs Whiteford in relation to their operation for some 5-6 years, Mr Chris Mather, a representative of Mather Dairy Utensils Company Limited who are the local agents for the manufacturers of the milking system which the applicants intend to install and Mr George Gourlay, a Chartered Surveyor who acted for the applicants in the negotiation of a rent review and ancillary matters in 2007-2008. For the respondents, evidence was given by Mr John Allen, an agricultural consultant with no prior involvement with the present parties and who was giving evidence as an independent expert, Mr Thomas Florey, a Chartered Surveyor and partner in the firm of Smiths Gore who are the estate managers for the respondents and Ms Kay McClelland, also a Chartered Surveyor with Messrs Smith Gore at Dumfries. We found all of the witnesses to be generally credible and reliable and there was no real conflict among them on any substantial matter of fact. The issues of relevance on which they were divided were (i) the merits of voluntary milking systems and (ii) the wisdom of the applicants investing is such a system at this point in time.
 In terms, respectively, of paragraphs 18 and 20 of part II of Schedule 5 to the 1991 Act, the enlargement of buildings and the provision of fixed dairying plant are improvements for which notice under sec 38 of the Act is required if they are to be eligible for compensation under Part IV of the Act at the tenant’s waygoing. It is possible that elements of the proposal may fall within paragraph 17.
 Sec 39 of the Act contains a mechanism for objection on the part of a landlord receiving such a notice. So far as relevant for our purposes it is in the following terms:-
“39— (1) Subject to subsections (2) to (4) below, compensation under this Part of this Act shall not be payable in respect of a new improvement specified in Part II of Schedule 5 to this Act if, within one month after receiving notice under section 38(3) of this Act from the tenant of his intention to carry out the improvement, the landlord gives notice in writing to the tenant that he objects to the carrying out of the improvement or to the manner in which the tenant proposes to carry it out.
(2) where notice of objection has been given under subsection (1) above, the tenant may apply to the Land Court for approval of the carrying out of the improvement, and on such application the Land Court may approve the carrying out of the improvement either —
(a) unconditionally, or
(b) upon such terms, as to reduction of the compensation which would otherwise be payable or as to other matters, as appears to them to be just,
or may withhold their approval.”
 Where either no Notice of Objection is served by the landlord or the Land Court gives approval in terms of sec 39(2) the result is that, subject to any conditions imposed by the Court in giving its approval, the tenant will be entitled, in terms of section 34(1), on quitting the holding at the termination of the tenancy to compensation from the landlord the amount of which is to be, in terms of section 36(1), “such sum as fairly represents the value of the improvement to an incoming tenant”.
 Before recounting the evidence bearing on the matters we have to decide, it is convenient to set out certain facts which, in addition to those set out above, are uncontroversial. These are as follows:-
(i) The various constituent parts of the holding are as shown on the plan which is production 30. The holding extends to approximately 248 ha (612 acres) in all (see the Minute of Agreement between the parties, production 13, Clause Third (4)). It is made up of six parts; “McWhanrick”, “Crossleys”, “Abbey” “Holm”, “Broomrigg” and “Carnsalloch” all as shown on said plan. It is bisected by the River Nith; Carnsalloch lying to the east of the river and the rest to the west. There is no bridge or other permanent connection linking Carnsalloch to the rest of the holding. The holding is farmed as a single unit run as a dairy farm.
(ii) The main steading is on “Abbey” and the layout of the various constituent buildings is as shown in production 39.
(iii) Mr and Mrs Whiteford farm to a very high standard. They are progressive, innovative and dedicated farmers and have invested heavily in their operation both in terms of finance and labour. They practise a high standard of animal husbandry with the result that they have a first rate herd of dairy cattle.
(iv) Over the years they have carried out various improvements, mostly with the tacit, rather than express, approval of the respondents for the purposes of Part IV of the Act. There have also been occasions when notices of improvement have been served which have been objected to by the respondents and/or not proceeded with by the applicants.
(v) We need not narrate all of these notices and improvements. The developments relevant to the present case can be traced back to the service, on 26 April 2005, of a sec 38 notice (production 2) in respect of the intended construction of, amongst other things, a Permastore Slurry Tank. It became clear from the evidence that this was with a view to the eventual conversion of the main livestock building (shown as “Cubicle Shed (1)” on production 39) to a cubicle building. The landlords agreed to pay forty per cent of the cost (production 7) and the work relating to the slurry tank went ahead.
(vi) A sec 38 notice in respect of said conversion duly followed on 13 February 2006 (production 4). This was a major piece of work, estimated to cost £100,000, which would create a cubicle shed with three rows of 100 cattle cubicles each, fitted with swing over-gates, mattresses, automatic scrapers and ventilation curtains. The work went ahead and it is in relation to the same building that the present improvements are proposed.
(vii) The improvements already carried out have allowed the applicants to gradually increase the size of their herd from about 250 cows in February 2006 to the present total of 380-400. Of these up to 300 are housed in Cubicle Shed (1) and the balance in other (cubicle) sheds. The applicants practise a system called “zero grazing” with the milking cows being kept inside all the time. The cattle are all milked through a herringbone milking parlour situated at the opposite end of the steading buildings from the main cubicle sheds (the relative layout is shown on production 39). The present installation, which has been in operation for many years, was an extension of a smaller parlour installed at an earlier time. In addition to the milking herd itself the applicants have about 300 young stock on the holding during the summer months, these beasts being wintered off the holding, although some return during winter for artificial insemination purposes.
(viii) On 28 July 2008 the applicants served the sec 38 notice with which this application is concerned (production 5). It refers, inter alia, to (a) the construction of a cattle shed measuring approximately 45’ by 90’ as an extension to the north end of the 2002 cattle shed, (b) the construction of a slurry channel, (c) the internal construction, within said extension, of various rooms as toilets, wash facilities, store rooms and an office, (d) the creation in the roof area of housing for electrical installations and equipment, water tanks and further storage rooms and (e) the installation of an unspecified number of voluntary milking systems with ancillary facilities. On 26 August 2008 Messrs Smith Gore, as agents for the landlords, served notice of objection as aforesaid.
(ix) On 14 January 2009 the applicants served notice of further proposed improvements (production 23) which involved doubling the size of the extension proposed in the notice of 28 July 2008. This doubling of size was said to be necessary to properly accommodate the proposed new milking system with its incidental services. However this notice has subsequently been departed from and does not form part of the proposal before us.
(x) The applicants presently employ five of a staff: a main herdsman, an assistant herdsman, a tractorman/stockman, a general farm labourer and a relief worker covering weekends and holidays.
(xi) An “all year round” calving pattern is practised. Both milking and dry cows are cubicle housed. The whole herd is put in calf to a dairy bull and the majority of the heifer calves are reared to calve at between two years and two years and three months of age. Approximately 25 per cent of these heifers come into the herd as replacements, with the balance being sold.
(xii) The calves are machine reared at Abbey. After weaning, their diet is based on straw as a roughage source. During the winter months they are kept on a holding near Dunscore with return to Abbey for artificial insemination as aforesaid and also for a two months accustomisation period prior to calving. The summer months are spent on the holding. At any one time there is an average of about 300 head of dairy followers either on the holding or being wintered away.
(xiii) Apart from areas of permanent grassland, generally situated in the low lying areas next to the River Nith and used for grazing of young stock, the whole unit is cropped to produce fodder comprising grass silage whole crop cereal silage and forage maize silage. More particularly, 120 acres of grass is cut three times a year for silage; 130 acres of wheat is grown, 80 acres of which become whole crop silage with the balance being left to ripen and be harvested in the normal way; between 120 and 130 acres of forage maize is grown all of which is ensiled; and the balance of the land is in grass for grazing dairy followers.
(xiv) Various feeds, both moist and dry are purchased and the system adopted is known as “Total Mixed Ration”. It is a system whereby measured quantities of home-grown forages, purchased moist feeds and purchased or home-grown cereals and other dry feeds are mixed by machine to be fed as a consistent diet to batches of cows according to their nutritional requirements. In addition, supplementary concentrated feeds are dispensed independently on a controlled basis via “out of parlour” feeders. Purchased moist feeds used include brewery, distillery, biofuel and sugar beet by-products. These feeds are palatable, nutritious and consistent and are at least as attractive on a least cost basis as home-grown feeds.
(xv) The holding is situated within the Nith Valley Nitrate Vulnerable Zone. It is therefore subject to regulations as to slurry disposal, the relevant regulations currently being The Action Programme for Nitrate Vulnerable Zone (Scotland) Regulations 2008 (SSI 2008 No. 298, hereinafter referred to as “the NVZ Regulations”). At present all slurry produced on the holding is applied to the holding, slurry being transported to Carnsalloch via an umbilical pipe and pump. By 1 January 2012, when a current derogation expires, the applicants will require to have storage for six months’ slurry on site.
 Turning to the evidence relevant to the matters we have to decide, Mr Whiteford explained that a voluntary milking system is one which effectively allows the cows to be milked when they want to be milked although frequency is in the overall control of the herdsman using a computer programme. Milking is by a robotic arm and Mr Whiteford claimed that once the cows become used to the system they use it more contentedly and subject to much less stress than in a conventional milking parlour.
 He intended to install four such milking stations. The machines he intended to use were made by DeLaval, an international company based in Sweden, and would be supplied and maintained by their local agents, Mathers Limited, Dumfries.
 One of the reasons for installing the new system was to reduce labour costs. The present workforce of five (see above) could be reduced by two.
 The average working day for the herdsmen and his assistants at present was 3.45 a.m. to 6.45 p.m. with breaks for an hour at breakfast and 1½ hours at lunch time. These were typical working hours for a dairy farm. They worked six days a week. It was very difficult to get and retain skilled staff. There was an acute shortage of such staff in the dairy world and workers could therefore pick and choose where they went. The result was that they would go only to the most efficient units which offered the best working conditions. That is what he wanted to achieve.
 Under the new system the working day would be from 6.00 a.m. to 5.00 p.m. Weekends would be more relaxed.
 Not only would they be able to cut back on labour costs but productivity would increase by about 10%. Four milking stations would milk about 260 to 280 cows a day in total. Because they had approximately 350 cows in milk at any one time, that meant that they would still go on using their existing, conventional, milking parlour for some of the cows.
 Denying that he was proposing to increase the size of his herd, Mr Whiteford explained that a limiting factor was the need to dispose of slurry in accordance with the NVZ Regulations.
 At present, slurry was spread over the whole of the holding. If you were able to export your slurry, you could keep more stock. Exporting slurry was quite a common practice in the dairy industry.
 In terms of providing fodder for the animals, the whole holding was cropped to provide grass, whole crop cereal and forage maize silage.
 The notice of improvements dated 15 January 2009 (production 23) had been withdrawn. That had been done because it was going to cause too many problems and delays. They needed to get the robotic system up and running. He said that they just did not know at the present time whether the works referred to in the Notice to 15 January 2009 would ever be carried out. However, those contained in the Notice of 28 July 2008 would definitely go ahead.
 In cross-examination Mr Whiteford confirmed that the holding was farmed as a single unit, without distinction as to use among its various parts. He did not accept that the proper construction of Clause First of the lease was that Broomrigg and Carnsalloch could not be used for dairy purposes. “Livestock” included dairy animals. The reason there was no reference to dairying in relation to Broomrigg and Carnsalloch was that there were no buildings on those two parts of the holding; they were just used for cropping and grazing young stock.
 He could carry on on the same scale as at present even without Carnsalloch and Broomrigg. He could keep the same number of cows. He would, however, have to buy in forage or forage and other replacement feeds. He could not go into the details of the cost implications of that; that would be explained by his expert witness, Mr Davies.
 It was possible to buy in the by-products of brewing and distilling. He seemed unsure as to whether alternative feeds could be got from a local source; he was already buying wet grains from Cameron Bridge Distillery. The cost was about £50 per tonne. He resisted further analysis of costs of various forms of feed from various sources, saying that he employed a professional nutritionist to advise in relation to these matters and he (Mr Davies) would be able to answer these questions.
 Slurry was presently transported to Carnsalloch by an umbilical system across the Nith. Without that, it would take about 15 minutes to transport slurry by tractor to Carnsalloch via the Dumfries bypass.
 The umbilical system involved a high pressure pipe of about 6 inch diameter and a pump. The alveus of the river was not part of the tenancy but if it became impossible to use the umbilical he would just transport the slurry by road. The present system pumped about 20,000 gallons of slurry per hour. A road tanker could carry about 3,000 gallons.
 Asked why no mention had been made of voluntary milking systems in a Minute of Agreement entered into with the respondents with effect from 11 November 2007, Mr Whiteford explained that voluntary milking systems had not been on the agenda during these negotiations. However, in 2007 he and his wife had travelled widely, both within the United Kingdom and to Sweden and Holland, and done a great deal of research into such systems and he insisted that the subject had been mentioned at a meeting at Cowhill before the notice of July 2008 had been served.
 Before deciding on voluntary milking systems they had planned a conventional parlour alongside Cubicle Shed (1). They had then had a serious look at their labour problems and decided to go for a VMS system.
 Questioned about the effects of the NVZ Regulations, Mr Whiteford explained that he had obtained derogation from the Regulations and the present position was that by November 2011 they would have to have six months slurry storage available. That would require a new slurry store.
 At present, when he used the umbilical to transport slurry to Carnsalloch, two SEPA officers were always present to monitor the operation.
 As to the number of cows which could be milked by one robotic station, that number was 60 or, at most, 70. So if they had a voluntary system comprising four robotic units in the cubicle building that system would milk 240 to 280 cows. A further 100 to 120 cows would be milked in the existing parlour.
 The present equipment could deal with the present number of milking cattle but the staff could not. In that context the introduction of voluntary milking did not amount to over-capitalisation of the holding.
 Asked by the court as to whether he was confident that the extension proposed in the notice of July 2008, as compared with an extension double that size as had been proposed in January 2009, would work totally satisfactorily, Mr Whiteford replied that in an ideal world the additional extension would bring great benefits but for the time being they had decided to go ahead on the basis of the present notice. That, on its own, would produce the benefits in terms of labour reduction and increased yield which he hoped to get from the VMS system. Additional benefits, not previously mentioned, would include better dry cow management, segregation of cows being made easier and management of stock generally being made easier. VMS systems also used less water so there would be less dirty water to dispose of via the slurry system.
 Mr Davies confirmed that he was the author of the report which is production 45.
 He is a director of Advanced Ruminant Nutrition Limited, a company he, along with a colleague, set up in November 2006. Its business is to advise not only on nutritional matters but matters relating to the efficiency and profitability of dairy farms. The company operates as a management consultancy, advising 60 dairy herds at various locations throughout the United Kingdom, pedigree Holstein breeders being dominant among their clients. The herds they deal with average 330 cows with the largest numbering 2200 cows.
 Mr Davies explained that more and more of the herds on which he advised were moving to 365 day housing, with the cattle being kept inside year round. About 40% of the herds he dealt with already fell into that category and the proportion was growing all the time. The trend was particularly noticeable among high yielding herds.
 This system gave the farmer greater control over the cow and her environment.
 Generally, it made sense to maximise the use of home grown forages combined with the efficient use of purchased feeds. Labour costs should be minimised. Labour was in any event getting harder to find.
 Mr and Mrs Whiteford had been clients of his business for five or six years. Mr Davies’ impression of them was that they were very good farmers with very high standards. Their historical records reflected their good practice and the condition of their cows told its own story.
 The introduction of four VMS machines would have a number of positive benefits. Firstly, there would be labour savings. Secondly, standing time while the cows waited to be milked would be reduced. At the moment they could be standing for four hours for a milking. If they were standing they were not eating and standing also exacerbated problems of lameness consequently increasing the stress on the cows. In a VMS system the cows tended to be milked more frequently, 3.2 or 3.3 times a day on average. That in itself allowed the cow to “express her genetic potential better”. Three milkings per day were better than two in terms of yield. The present yield of the Whitefords' herd could be increased by 15% by use of a VMS system. The present average annual yield was 9,471 litres per cow. The anticipated increase would put the Whitefords “right at the top of the tree” across the country. There were very few herds producing 10,000 to 11,000 litres of milk per cow per annum. The increase would have a very big effect on profitability.
 Cubicle Shed (1) would be served by the new system. It would deal with 250 cows. That would have a knock-on beneficial effect on the cows going through the existing parlour in as much as they would be standing for a shorter period of time.
 Interest in VMS systems was very high. Within his own client base, four herds had transferred to voluntary systems within the past four months and two others were thinking about it. Historically there had been issues with voluntary systems but these had now been resolved as a result of improved technology.
 As to the effect on the Whitefords’ operation of having only 240 acres available to them (the extent of the holding excluding Carnsalloch and Broomrigg), the witness did not think that that need have a large impact. Home grown forage supplies would still be at a substantial and acceptable level. There would have to be some change from forage to moist feed, of which there were numerous sources, notably breweries, in Scotland.
 Asked by the court as to whether there was a minimum level of home-grown forage needed, Mr Davies confirmed that there was: the lowest he was presently recommending was 5.5 kilograms of dry matter per cow per day. Even if only 240 acres of the holding was available it would still yield 7.3 kilograms of home-grown forage per cow. It would be perfectly possible to continue with the present enterprise even if the acreage available for cropping was substantially reduced.
 Asked about the effect of a reduction in acreage in relation to the NVZ Regulations, the witness explained that it was possible to export manure if acreage was available elsewhere. He had seven clients who did that. He had two clients who carried a similar level of stock to the Whitefords but on smaller areas of land. One client had 1,200 cattle on 360 acres and another 2,000 cattle on 710 acres, both in NVZ areas.
 But if Carnsalloch and Broomrigg were to be used for growing crops at all they would still need fertiliser: that need would not disappear simply because they were not used as a single unit with the rest of the holding.
 In cross-examination Mr Davies accepted that 365 day housing of the cattle and use of VMS systems were two separate things. He agreed that he was mainly involved with the former. He agreed too that the Whitefords have healthy cattle and good yields on their existing system.
 Although he accepted also that the present parlour was not constantly breaking down, he said that features of it made it less efficient than certain other possible arrangements would be. In particular the location of the parlour was not the most advantageous in relation to the cattle in Cubicle Shed (1). That could, he accepted, be remedied simply by moving the existing parlour.
 Mr Davies also accepted that dairy farming involved a highly volatile market with constant pressure on milk prices and individual farmers having no control over the market. He admitted that a significant number of farmers had left the industry. He agreed that a VMS unit cost about £80,000 although pointing out that that was for the first unit; additional units should be cheaper. There would also be the cost of a milk storage tank. He had never seen the final figure for what the Whitefords proposed but he imagined it to be of the order of £400,000 to £500,000.
 If one had a VMS system, although one would not need a second parlour, a second parlour would still give more flexibility for problematic cows. As well as cows whose teat placement was poor there were cows which just could not be trained to use VMS systems. However there was generally no problem in training newly calved heifers to use a voluntary system: they were generally easier to train than cows used to a conventional milking parlour because they had never been through a parlour of any kind.
 Asked whether an element of prudence was not advisable given the volatility of the present market, Mr Davies said that he would prefer to be “in milk” than in any manufacturing industry at the moment.
 As to the possibility that supermarkets would turn against milk from herds which were constantly housed, consumers’ concerns as to animal welfare were really matters of perception. As long as the industry continued to learn its lessons it could “keep ahead of the game”. Welfare arrangements were constantly being reviewed.
 Mr Kermack put to the witness figures attributed to Muirside Farm, which the respondents have in hand. These showed a yield of 10,500 litres per cow per annum on average, based on two milkings a day using a conventional milking parlour. Such yields could, he accepted, be achieved by such means.
 Questioned about his baseline figure of 5.5 kilograms of dry matter per day produced on the holding per animal he said even the reduced holding of 240 acres could sustain 500 cows on that basis. If only 240 acres was available for use and 300 young stock were kept on the farm rather than being wintered away the only effect that would have would be in terms of the need for more storage because the heifers were fed on straw, not forage. Young stock did not require home-grown forage.
 Asked about his expertise in relation to the effects of NVZ Regulations, Mr Davies explained that he had no particular expertise in that area.
 As to whether the present herd could continue to be serviced by the existing parlour, one was always looking to improve. The existing parlour was already working too many hours and it was not labour-efficient. One could go on using it but not necessarily to best profit. He always advised his clients to work on the principle that if they looked after the cows, the cows would look after them.
 In re-examination Mr Davies confirmed that the dairy market was highly volatile and that a number of farmers were leaving it. In his view it would be the most efficient who would survive.
 In answer to questions from the court, Mr Davies confirmed that he was aware of three herds in south-west Scotland which were housed for 365 days of the year. In another case three-quarters of the herd were housed and one quarter – the “low yielders” – were allowed out to grass.
 In his view the highest yielding herds were also the most profitable. Increased output and reduced costs were the two main drivers towards efficiency.
 VMS systems did not require a continuous human presence. In Holland and Denmark, where such systems are more common, VMS systems were linked to the operator’s mobile phone. An operator he had spoken to in Holland had had only two call outs in the previous year. So someone had to be on call but not necessarily in constant attendance.
 Asked by the court as to whether he was aware of discussion between the supermarkets and farmers operating 365 day housing systems, the witness said that the supermarkets were advocating this trend as an efficiency measure. The public had already in effect decided the welfare issue: “locally produced milk” was now almost “off the shelves”.
 In the context of the Whitefords’ operation, if the choice was between moving the present parlour to Cubicle Shed (1) or installing a VMS system in an extension to that shed he would opt for the VMS system so as to reduce labour costs. He would use the VMS system to service the high yielders while the low yielders would continue to go through the existing parlour.
 Mr Mather gave evidence that he was a director of Mather Dairy Utensils Company Limited, Dumfries.
 His company sold DeLaval voluntary milking systems. DeLaval were, he said, world leaders in this field and his company were their local agents in Dumfries and south west Scotland.
 Mr Mather explained what a voluntary milking system is. It is a system whereby the cows, of their own volition, enter the milking area when they are ready to be milked. Food is also used as an incentive for them to enter the system. Once fed they can get back into the main part of the cattle shed only by passing through a “smart door” which identifies each cow from its electronic transponder and checks whether it has a “milking permission”. That involves a check as to when the cow was last milked and whether, according to its pre-set milking programme, it is now ready to be milked again. If it is ready to be milked it is directed to a robotic milking machine or station; if not it is released back into the main shed. Production 26 is a DeLaval brochure about the system.
 Mr Mather had been to Abbey Farm and knew where it was proposed to site the new system. Under reference to production 39 he indicated that the VMS machines were to be at the top (north) end of Cubicle Shed (1) and the dairy building to the right (east) of that shed. He had discussed these proposals with the Whitefords, most recently some six to eight weeks ago. He explained, however, that the arrangements could be a “moveable feast”. He would give advice as to layout but the final decision would be a joint one. That said, there were not many alternatives at Abbey. The dairy could be situated elsewhere but the layout he had indicated was the best.
 As to the milking capacity of DeLaval voluntary systems, it was difficult to answer that in terms of the number of cows per day. They tended to work on litreage throughput. However, his estimate was that a single DeLaval robotic machine would milk about 70 low yielding cows in a day and 50 higher yielders. Having been told of earlier evidence that 280 to 300 cows were expected to be serviced by the intended arrangements at Abbey, he said that that would be “about the top side” of what the intended system could achieve. He would have thought about 260 cows was the number it could comfortably service.
 In cross-examination he frankly accepted from Mr Kermack that he had a financial interest in the outcome of this case: “If I don’t sell these four machines I don’t make money”. The cost was to be £120,000 per robot. However, because four were being supplied there would be a discount on the individual unit prices.
 In addition to supplying the robots his company supplied all the ancillary parts. A bulk storage tank for four robots would cost £25,000-£30,000, fitted. Mr Kermack put it to him that the need at Abbey could be met by a traditional “herringbone” parlour but that to say so would not suit his purposes, as a salesman of VMS systems. The witness replied that his company sold conventional parlours as well.
 Mr Mather then gave more detailed evidence as to the operation of the system which we need not repeat here. It is perhaps, however, relevant to note that each time the cow is milked there is automatic sampling for the presence of antibiotics and mastitis. The software could also be set up to carry out further sampling.
 Mr Kermack, who said that he had watched the DeLaval video of the system online, put it to the witness that a lot of things had to go right first time for the system to operate smoothly. The witness responded that any system will go wrong from time to time but whenever something went wrong in one of these systems the system automatically alerted both the farmer and Mr Mather’s company. As part of their contract for the supply of the machines they guaranteed that technical failures would be sorted within two hours. His company had every single part required for these machines in their shop in Dumfries and they were contactable 24 hours a day. He was very positive about VMS machines, believing them to “be the future”.
 Returning to the number of cows the system could milk in a day, he agreed with Mr Kermack that if Mr and Mrs Whiteford’s cows were within the top 10% of yielders in the UK that would mean that relatively fewer of their cows would be processed in a day. The witness replied that 9,400 litres per annum was a high but not dramatically high yield. They had supplied systems where the average annual yield was 13,000 litres.
 Asked to compare the cost of installation of a VMS system to a conventional system for a 300 cow herd, the witness thought the figures would be about £130,000 for a herringbone parlour, £250,000 for a rotary parlour and £400,000 for a VMS system. To service 300 milking cows would need five robots but he was talking about a herd of 300 out of which 240 or 250 would be milking at any one time and that would need only four. More specifically, in the present case, his company had sold (and that is how he put it) Mr and Mrs Whiteford a four robot system which would milk 250 to 260 cows. That was going to cost the Whitefords about £400,000. In their particular case that cost did include the bulk storage tank because Mathers were anxious to promote the sale of such systems in their area and were, therefore, offering good deals.
 Asked by Mr Kermack as to how flexible arrangements were at Abbey in terms of where the equipment and dairy could be located, he said that it was not really a moveable feast; it was a question of working round what was already there.
 As to the size of extension which was necessary to accommodate the proposed system, he could not remember the exact measurements but he thought one would need 40 to 60 feet or 15 to 20 metres of an extension to Cubicle Shed (1).
 He had first spoken to the Whitefords about VMS systems some time last year. Mr and Mrs Whiteford had gone over to Denmark to see systems in operation in May or June of last year. Asked whether the Whitefords had changed their minds in January 2009 he said that they had not changed their minds so far as robotics were concerned. Referred to production 23, the sec 38 notice of 12 January 2009 giving notice of the proposed additional extension, he said that the expert’s advice referred to in the second paragraph of the undernote would have come from himself and DeLaval. Although he had said that one needed 40 to 60 feet added on to the present building, one could get away with as little as 25 feet but that would mean that there would be no shedding facility.
 He had not had any discussions with the Whitefords about increasing the size of their herd. Asked whether dairy farming could go on at Abbey with the existing parlour he said it was just like deciding to change your car. Did you really need to change? Everyone wanted to move on to something better. In business one wanted to move forward. The present system could not cope properly with the number of cows involved because of labour problems.
 It was very unusual for cows to take a dislike to voluntary systems. Cows were usually much happier with such systems. They just walked in when they wanted to. They were less stressed. There was less buffeting and pushing while they were waiting to be milked.
 Asked, in re-examination, whether a 45 foot long extension was sufficient to house four robotic machines, Mr Mather said that it would house the machines but not the ancillary facilities. That would, however, be a workable system.
 As to why Mr and Mrs Whiteford should not opt for the cheaper option of a conventional system, Mr Mather said that one had to set against VMS construction costs the savings in labour and increased yield produced by voluntary systems in future years. Voluntary systems were selling: in the last three or four weeks his company had sold 24 robotic machines, several of which were already in operation. They anticipated that in a few years’ time 50% of their business would be in the sale of robotic systems.
 Mr Gourlay is a Fellow of the Royal Institution of Chartered Surveyors and a partner in the firm of Messrs GM Thomson Allied Surveyors in Dumfries. It is a rural practice and Mr Gourlay said he was thoroughly experienced in agricultural matters.
 In 2007 he had been instructed by Mr and Mrs Whiteford in relation to a rent review. Production 13 was a Minute of Agreement which had resulted from the subsequent negotiations. Negotiations had got under way in around February 2007 and had continued for a year or so.
 Production 51 was a letter from Mr Gourlay to Mr Whiteford dated 20 July 2007 summarising discussions with Mr Thomson Florey of Messrs Smiths Gore. The reference in the fourth paragraph to the Trustees (of the Cowhill Trust) having withdrawn their offer of a contribution towards the proposed parlour building was a reference to a proposed contribution of £100,000 referred to in a letter dated 15 January 2007 from Mr Florey to Mr and Mrs Whiteford in which Mr Florey had said that he had recommended to the Trustees that the sum of £100,000 be made available specifically for structural work required to re-locate the milking parlour so as to be adjacent to the large cubicle building in which most of the cows were kept. Mr Gourlay explained that the Trustees had contributed to earlier dairying improvements and had been considering further contributions to further improvements at the time of Mr Florey’s letter.
 In his discussions with Mr Florey in the course of said negotiations, Mr Florey had not disclosed any concern about over-capitalisation at Abbey. The Trustees had decided not to go ahead with the £100,000 contribution because the funds were required elsewhere. As the witness had put it in his letter of 20 July 2007 to Mr Whiteford (production 51, fourth paragraph):-
“I was told by Thomas Florey that this is in no way a reflection on you, but rather the result of an unforeseen and pressing demand for funds elsewhere. Your proposals are still approved of and your impact for the good of Abbey are welcome”.
 Referred to production 28, a report prepared by the respondents’ expert witness (see below), Mr Gourlay questioned the reference to limitation on the number of cattle which could be managed under the lease. In particular he could not understand why it was being said that there was a limit on the disposal of slurry. It was quite common practice in the local area for slurry to be exported, particularly because of NVZ requirements. He, personally, had knowledge of slurry being exported a distance of eight miles. Any farmer conducting an arable operation would find slurry valuable.
 In cross-examination Mr Gourlay accepted that he had been involved, in terms of the Minute of Agreement finally negotiated, in more than the negotiation of a rent review. The negotiations had also included the setting aside of the Post Lease Agreement and there had also been a query as to what the effective acreage of the farm was; movement of the river Nith having affected the usable area. These negotiations had been a general attempt to clear the air as between landlord and tenant. They had resolved historic matters.
 The number of cows being kept had not been a relevant consideration for him in terms of negotiating the rent review; he had done it on the basis of acreage and land classification.
 With reference to the Trustees’ suggested contribution of £100,000 to a proposed new milking parlour, he imagined that the type of parlour being contemplated had been a herringbone parlour but in any event it had not gone ahead.
 Asked about previous involvement on behalf of Mr and Mrs Whiteford, Mr Gourlay said that he had been involved in something to do with diversification payments some years ago. He was instructed by them on an entirely ad hoc basis.
 If the new building and parlour towards which the landlords had intended to contribute had gone ahead it was his understanding that one of the existing buildings would cease to be used. His understanding had been that the new parlour would be housed in the proposed new complex. Part of that had arisen out of concerns which had been raised by Mr Whiteford about effluent and dirty water caused by the cows having to walk to the existing milking parlour
 His understanding of the landlords’ position in relation to that proposal was that they had been in favour of it and that although, because of the need for investment elsewhere, no contribution to the cost was possible Mr Whiteford was nevertheless being encouraged to go ahead with it.
 As he had understood it, what the landlords had had in mind was a contribution to the super-structure of the proposed new building, by which he understood the roof, frame, walls, foundations and floors but not the parlour equipment.
 Asked by the court what his understanding of “over-capitalisation” in the farming context was, he said that he understood it in this case to mean investment in tenants’ improvements beyond what a hypothetical tenant might do. An outgoing tenant who had over-capitalised would get less compensation than he might expect.
 Dairy cattle being housed year-round was becoming increasingly common in south west Scotland. Half his clients now practised “zero grazing” but he had been familiar with the concept since starting work in Dumfries in 1972.
 The use of forage maize was also increasing locally. He did not know what proportion of farmers were growing it but he saw someone new doing it every year as more suitable varieties became available.
 The rent review negotiations had been based on Grade 2 and 3 dairy ground in the Dumfries area with adjustments for the particular location of the holding. That analysis had been applied to the whole of the farm. He could not remember whether he had made Mr Florey aware that that was how he was approaching matters but he was sure that they were both looking at it as a dairy farm. They had not argued about the applicable rates because they had been in agreement. Mr Florey had never said to him that parts of the holding were not for dairy use.
 Mr Allen gave evidence that he was the managing partner of Kite Consulting, an independent agricultural consultancy, providing technical and business advice to farmers. He had a B.Sc. (Hons) in agriculture, 30 years experience as a consultant, senior consultant, senior manager and director with the Agricultural Development and Advisory Service from 1979 to 2000 and had been with Kite Consultants since then. He was a founder member of the Institute of Agricultural Management, a member of the Royal Agricultural Society (RAS), the Royal Association of British Dairy Farmers and the British Institute of Agricultural Consultants. He had worked with farming businesses throughout his career, determining business strategies and investment in dairy premises and operations and specialising in financial aspects.
 Mr Allen spoke to production 28, a report he had prepared on the instructions of Messrs Turcan Connell, Solicitors, for the respondents.
 In the preparation of that report he had visited Abbey Farm on 19 May 2009. He had not previously been employed in any capacity by either party to this application.
 Mr Allen took us through his report. There is no need to repeat all he said.
 Asked by the court as to what he understood by “over-capitalisation” (a term which is used at pages 5 and 7 of his report), Mr Allen described it as a situation whereby a lot of money is spent upon a holding in circumstances such that it may not be possible to recover the investment by sale of the asset at some future date.
 He had worked with clients who had knowingly over-capitalised but in situations where it was their money and their holding. They had taken a “knowing” risk. It might be possible to re-coup the investment in terms of increased income generation over a period of years. There might be residual value in buildings remaining available for use for other purposes once the particular purpose of the investment had come to an end.
 He looked at over-capitalisation from the point of view of somebody else putting money into the proposed improvements. Over-capitalisation involved a mismatch of capital to the requirements and potential of the holding. It also involved the ideas of viability and sustainability. For example 150 cows on 250 acres was a safe investment. But he instanced a client of 20 years’ standing who had invested in a 350 cow dairy herd unit on 160 acres. If he was forced to sell, he might not recoup all of his money.
 As to sustainability, his understanding in the preparation of his report had been that Mr Whiteford believed he needed 500 to 600 cows to create a viable herd.
 The NVZ legislation placed certain limitations on stocking options going forward. In his report he had assumed that it was uneconomic to export slurry outwith a five mile radius. Access to land for the purpose of slurry spreading could be a problem and a limiting factor. His report had been based on current NVZ Regulations and worked out in four scenarios, one involving derogation from certain aspects of the Regulations until 2014, one involving no derogation, one involving use of the whole farm as at present and one excluding use of Carnsalloch. The results were as set out in his report. The conclusion reached was that because of NVZ restrictions, even assuming use of the whole holding, the current stocking level could be maintained only until 2014. So a legitimate case could be made out to the effect that the landlords should not capitalise anything above 180 cows and 90 followers.
 Referred to the existing arrangements at Abbey and, as part of that, to production 39, there were problems with the layout and Cubicle Shed (3) also posed challenges because it was an older building. A sensible way forward would be to concentrate the dairy herd in Cubicle Shed (1) and have the milking parlour and dairy there as well. A disadvantage of the present arrangements was the long distance the cows had to walk from Cubicle Shed (1) to be milked. He would have thought that it would be better to relocate the dairy to Cubicle Shed (1) but he was not an expert on that and professional advice should be taken.
 Asked by the court as to the effect on stocking levels of a reduced area of land being available, the witness said that provided Mr and Mrs Whiteford could go on using all of the farm they would be able to maintain their present level of 380 to 400 cows. If the young stock currently wintered away were kept at home all year round, that number would fall to 300 cows. If the young stock was away for half the year the number of cows they could keep would be half way between 300 and 380.
 Asked by the court about conclusion 6 at page 7 - that Abbey would become over-capitalised if the investment in new milking robots to accommodate 400 to 600 cows went ahead - the witness explained that in general his consultancy would not recommend VMS systems for herds of more than 250 cows. Beyond that level these systems tended to be uneconomic. What was being proposed here was a very high cost solution to the problem. The cost was up to £600,000 and further capitalisation would be required in the slurry system.
 If there were to be only 400 cows and followers there would still be an element of over-capitalisation. The proposed system would cope with roughly 270 cows and the remaining 70, or so, would have to go through the existing parlour. To him that was over-capitalisation. He was aware of farms where VMS systems had been installed but the parlour kept as they wished to keep animals which rejected the VMS systems.
 What he would do in order to meet present problems, consistently with the principles referred to in his report, would be to take advice on having one milking system capable of milking the whole herd and located in or adjacent to Cubicle Shed (1). That would require an extension of the present building although, alternatively Cubicle Shed (2) could be used to house 40 or so cows.
 On a scenario of having 300 cows plus followers, put to him by the court, the present system would not require much adjustment. But it was not a good arrangement long term and it would still be better in his judgement to concentrate everything in Cubicle Shed (1). It would be his recommendation to relocate the present milking system into Cubicle Shed (1): that would be a good outcome for both landlord and tenant.
 He had extensive experience of working with major milk purchasers such as Milk-Link and Wisemans and with retailers such as Waitrose, Asda, Sainsbury and Tesco. These retailers represented something like 60% to 70% of the milk market. “Zero grazing” was a very contentious area and one of great concern to his clients. The retailers were very concerned about public perception of zero grazing so producers presently operating zero grazing were probably going to have to provide areas of grazing for part of their herd for part of the year. It was a perceived, rather than actual, welfare problem.
 Consumer perception that they should see cows at grass created a risk to investment of the kind contemplated in this application. The dairy industry did not want to end up where the poultry industry had ended up with farmers having made big investments in sheds which were useless.
 Mr and Mrs Whiteford should probably consider having the in-calf heifers or dry cows at grass for a time. He was not suggesting that every cow should be sent out for part of each lactation but he had clients whose cows went out in late lactation and in good weather. His business had clients with 200 to 250 cows with access to grazing and producing average annual yields of 10,000 litres but that was difficult to achieve with 400 to 500 cows.
 Asked by Mr Kermack as to what he would advise Mr and Mrs Whiteford to do at this stage in their business he suggested that they take notice of NVZ requirements and the difficulty, as he saw it, of accessing additional land locally and that they should reduce the size of their herd slightly and improve their milking conditions. He thought 330 cows plus followers would give a good balance. That would still require the transporting of slurry to the other side of the River Nith. He would not advise the acquisition of four voluntary milking systems.
 In cross-examination he accepted that if the use of Carnsalloch was restricted to arable use, slurry would still be useful. So Carnsalloch could continue to be used for slurry from the dairy operations. He had taken that into consideration in his calculations.
 As to whether six miles was a wholly unreasonable distance over which to export slurry, he would have thought that it was well worthwhile looking at a slurry separation system, separating liquids from solids, the solids being stackable and transportable by trailer. That was a more economic way of transporting slurry over a long distance. Liquid slurry, under this arrangement, could be used on the near side of the river and the dry product on the far side. Such a separation system could be added to the existing system. In his judgement the fact that SEPA approved the present arrangement did not mean that it was a good long term solution.
 He was aware that the compensation payable to an outgoing tenant was the value of the improvements to the incoming tenant. He had tried to apply his mind to what that value would be. He thought that what was proposed was an expensive solution to the milking problem. If the investment in the voluntary system was written off over 10 years that would be fine but one did not know what might happen.
 Mr Allen confirmed his previous evidence that 330 cows plus followers was sustainable on Abbey if one had access to all of the land comprised in the present holding. But he had been instructed on the basis that the herd size was going to be increased to 500 or 600 cows. He had been told that the tenant was proposing to confine himself to the present level of stocking only that day. There would, in any event, be a risk to the landlord if he was asked to compensate for the value of an expensive milking system as might be the case if Mr and Mrs Whiteford left at short notice.
 Mr Allen accepted that if the current proposals went ahead they would result in a reduction of labour requirements in respect of the existing parlour because fewer cows would be milked there. However, in his experience the move to voluntary milking systems did not produce the expected savings in labour: the machines still needed close supervision. The quality of life for the farmer tended to be better but he thought the applicants would be lucky to reduce their labour requirements by one man, never mind two.
 Asked by Mr Lean about whether a number of his clients practised zero grazing, Mr Allen replied that they were trying to move away from continuous housing. Zero grazing had a lot of negative connotations for consumers. The way to deal with that was to have grazing for a portion of the herd for part of the year. That would not necessarily be the actual milking herd: it could remain indoors. It could be the dry cows and in-calf heifers which would be out at grass.
 In re-examination Mr Allen confirmed that his evidence as to the carrying capacity being sustainable at 330 cows plus followers was based on the young stock being on the farm only for summer grazing.
 Asked about the principal advantages of voluntary milking systems, Mr Allen said that in addition to improved lifestyle they potentially had stock management benefits. But the impetus for their use came largely from Holland where farms tended to be smaller, with dairy herd sizes of about 200 cows and labour very expensive. In that situation putting two robotic milkers in was a very attractive proposition. Once you had a herd of over 250 cows it become uneconomic to have a robotic system. That was based on economic cost: when you got to larger herd sizes it was far cheaper to put in a conventional parlour. For a herd of 200 cows one might put in two robots at a cost of £200,000 and reduce staffing by one. For 500 cows you would need eight robots at a cost of £800,000 but you might still lose only one man. At that level it became more economic to put in conventional parlours. It was wrong to assume that one would not be able to find the staff to operate a conventional parlour. To assume that the dairy industry had no good people in it was to malign the industry. Good conditions, good pay, sensible hours and good management would always attract labour.
 Having both a voluntary system and a conventional parlour was not an ideal solution. He did not know of anyone who had set out to create such a system although some people ended up with it by default.
 Asked by Mr Lean whether a hypothetical tenant of average experience would rule out taking over a VMS system, Mr Allen said that one could never rule anything out but in the near future, until these systems had proved themselves, investment was going to be at risk and an incoming tenant was unlikely to pay a premium for such a system.
 Mr Florey has been a member of the Royal Institution of Chartered Surveyors since 1981 and a Fellow since 1994. He is presently a partner in Messrs Smith Gore, Dumfries.
 Mr Florey gave evidence that he had had dealings with Abbey Farm and Mr and Mrs Whiteford. These dealings had been on behalf of the landlords, the Trustees of the Cowhill Trust. He had acted for them since 1999.
 The landlords, as well as having farms which were let out, had two farms in hand: Bellfield and Muirside. These were dairy farms.
 Mr Florey was referred to production 11, the lease of Abbey Farm, and to the last two sentences of Clause First. He took from those sentences that Carnsalloch and Broomrigg were not to be used for dairying. To him, dairying, although it involved livestock, was not “livestock farming”. To him that phrase meant beef and sheep farming but not dairying. Asked by the court what activities, in his view, would constitute dairying Mr Florey replied that the whole cycle of animal breeding for dairy purposes and the provision of food for those animals would be dairying.
 Referred to Clause Sixth of the lease and the apparent prohibition on the tenants adding to buildings and fixed equipment without the landlords’ written consent, the witness confirmed that they had had a series of requests for approval of improvements to buildings from Mr and Mrs Whiteford. He could not remember whether Mr and Mrs Whiteford had ever asked for written consent in respect of any of these.
 Referred to the plan attached to production 13, Mr Florey confirmed that the lease did not include the River Nith. Asked about the umbilical arrangement which presently allows slurry to be transported across the Nith, he confirmed that the landlords were worried about the possibility of major incidents if anything went wrong with that arrangement. He knew that SEPA and the Nith District Salmon Fishery Board, which he chaired, were also concerned about the risk posed to the river by the umbilical. A discharge could result in a major fish kill.
 He identified production 14 as a letter he had written to Mr and Mrs Whiteford on 14 August 2001. It dealt with a number of land management issues. The writing of the letter had followed discussions with Mr and Mrs Whiteford. The letter referred to an additional “young stock building” which Mr and Mrs Whiteford had been planning in the wake of the Foot and Mouth epidemic. The existing buildings were not modern and not suitable for young stock and there had therefore been a need for a modern and suitable building for young stock on the holding. A building had subsequently been erected. On production 39 he identified Cubicle Shed (1) as that building. The Estate had been prepared to pay £30,000 towards the cost of a building. That was confirmed by reference to production 15, a letter from Caroline Leggat, his assistant of the time, to Mr and Mrs Whiteford dated 5 September 2001.
 Production 16 was a notice to his firm, as agents for the landlords, of a planning application made by Mr Whiteford in respect of a “proposed beef and sheep building” at Abbey Farm. In terms of its stated purpose that application seemed to fit exactly with what he had been discussing with Mr and Mrs Whiteford.
 Production 17 was an internal Smiths Gore minute of a meeting between Caroline Leggat and Mr and Mrs Whiteford on 9 May 2002. Among various matters discussed was the proposed general purpose shed (item 4). It recorded that the Estate would contribute a figure of £35,000 to the total estimated cost of £130,000. That contribution had been paid. (The witness went on to give evidence about the building as built being much bigger and costing much more than had been originally indicated but since we do not regard that evidence as relevant to the issues we have to decide we need not repeat it here.)
 That shed had been built in 2002. Production 4 was a sec 38 notice dated 13 February 2006 for work to be carried out within that shed for, among other things, the construction of three rows of 100 cattle cubicles each. That had caused the Estate to have concerns about a possible big expansion of Mr and Mrs Whiteford’s dairy operations. The provision of cubicles was generally inconsistent with using the building for beef cattle although cubicles were sometimes used for suckler cows.
 Mr Kermack then took the witness through productions 18, 19 and 21 dealing with proposals by Mr and Mrs Whiteford to improve the dirty water, effluent and slurry management arrangements at Abbey Farm so as to comply with SEPA and NVZ Regulations. These productions show how the Estate co-operated with Mr and Mrs Whiteford in relation to these proposals, in particular in the construction of Permastore slurry tank, and eventual payment by the Estate of the sum of £21,928 as a contribution towards the cost.
 Evidence then turned to the notice of 28 July 2008. Mr Florey had taken it to mean that Mr and Mrs Whiteford were proposing to put up two buildings; a milk parlour and a bulk storage facility. The first building was to measure 45 feet by 90 feet. No size was specified for the second. A voluntary milking system was to be installed in an extension to the far end of Cubicle Shed (1).
 In response to this notice his firm had been instructed to send the notice of objection on behalf of the landlords dated 26 August 2008. That had been accompanied by a covering letter, of the same date, stating the ground of objection which was “that the improvements would overequip the farm for dairy use relative to the extent to which it is specified and intended in the lease”. At that time, the witness explained, the Estate had no knowledge of the Whitefords’ “intended game plan” or the eventual intended size of their herd. The Estate would not be happy with a contravention of the lease which would result in the whole farm being used simply for a dairy operation.
 The location of the present milking parlour was a matter of concern. It was a major concern for SEPA. It was desirable to re-site the parlour if possible in order to avoid the cows walking outside. Any new parlour would best be situated next to where most of the cattle cubicles were.
 Production 42 was a minute of a meeting he had had with Mr and Mrs Whiteford on 5 March 2009. The reference to Mr Whiteford saying that “the parlour required to be replaced now” was, as the witness understood it, a reference to the present parlour requiring to be replaced. The witness’s understanding was that this was to be in addition to the new voluntary milking system.
 In cross-examination Mr Lean explored with the witness his understanding of the last two sentences of Clause First. He accepted that these did not say, in terms, that part of the holding was not to be used for dairying but explained that it was the difference between the two sentences which made it clear that Carnsalloch and Broomrigg were not to be used for that purpose.
 Mr Lean also asked about the apparent restriction contained in Clause Sixth and put to the witness the notices comprising productions 2, 3, 4 and 5. The witness confirmed that the lack of written requests for written permission for additions to fixed equipment was not something which had particularly troubled him: the correspondence passing and the discussions ongoing at the relevant time would have told the tenant whether or not he had permission for what was intended.
 Mr Florey confirmed that one of their concerns on receipt of the notice of 28 July 2008 had been that the Whitefords might be proposing to increase the size of their herd to such an extent that it would be out of proportion to the land let for dairying purposes. Mr Lean put to him the evidence that we had heard that there was no intention of increasing the size of the herd and that the voluntary milking system would milk 280 cattle. Mr Florey said that that would not change the Estate’s position. Two milking parlours were not necessary for a herd of the present size and a conventional parlour could be installed as a much cheaper alternative to a voluntary milking system. He still stood by the recommendation to relocate the parlour and the Estate would contribute to that on the basis that the present milking parlour would be closed and made redundant. Under reference to production 51, Mr Gourlay’s letter to Mr Whiteford dated 20 July 2007, reporting on various matters, including the withdrawal of the Estate’s intended contribution to the proposed parlour building, Mr Florey confirmed that the reason for that had been a pressing need for funds elsewhere which meant that the Estate could not make this contribution. Mr Lean put it to the witness that the Estate had nevertheless approved of the proposals (that is to say those referred to in paragraph 10 of production 37) and that, given SEPA’s concerns about the present arrangements, Mr and Mrs Whiteford had little choice but to proceed with those improvements. In response Mr Florey made three points. Firstly, retaining the present parlour would not meet SEPA’s concerns. Secondly his clients may now be in a position to make a contribution towards the cost of such improvements. And thirdly, it was not necessary to spend £600,000 on a voluntary system when a smaller conventional parlour could be put in for much less.
 Referred to production 34 and the section dealing with the “fixed equipment”, Mr Florey agreed that a herd of 380 cows was consistent with what he says in production 34 about a considerable number of Scotland’s remaining dairy farmers now having herds in excess of 300 cows. It was not, however, fair to conclude from that that a herd of over 300 cows was consistent with the Trustees’ wishes for Abbey. His contention had always been that a herd of 250 to 300 cows was reasonable for the acreage of permitted dairy use and that a lot of the older buildings on Abbey should be made redundant or used for young stock. Although he had not said that in his letter of 15 January 2007 (production 34) it had been at the back of his mind.
 Asked in re-examination whether, in his opinion, a herd of 380 cows was a reasonable or unreasonable for Abbey, Mr Florey said that such a level would be “very much at the top end of where (he) would like to see it”.
 Mr Kermack referred to an averment in the applicants’ pleadings that “the notice of 13 February 2006 makes clear that the improvements were necessary to service an additional 300 dairy cows bringing the dairy herd to a total of 500 to 600 animals”. The witness made the point that the notice of 13 February 2006 says nothing about the eventual intended size of the herd but in his view 500 to 600 animals would be totally excessive.
 The court explored with the witness what the Estate’s position might be in relation to the possibility of extending Cubicle Shed (1) by approximately the same dimensions as in the notice of 28 July 2008 but housing a conventional milking parlour, the herd size remaining the same as at present. The witness thought such a proposal would be sensible. Subsequently referred, by Mr Lean, to his earlier evidence that 250 to 300 cows would be a sensible herd size, he thought 380 would be at the top end of what was reasonable.
 Ms McClelland gave evidence that she was a member of the Royal Institution of Chartered Surveyors working with Messrs Smith Gore in Dumfries. She had been the author of the notice of objection and covering letter both dated 26 August 2008 which had been their clients’ response to the sec 38 notice with which this application is concerned. What she had meant by concerns about over equipping the farm, in said covering letter, had been that the Estate felt that what was proposed was excessive in relation to the land let for dairy use.
 She could not recollect voluntary milking systems being discussed at a meeting on 4 July 2008 attended by Mr and Mrs Whiteford and Mr Percy Weatherall. She could not say one way or the other whether VMS systems had even been mentioned.
 Mr Lean submitted that, the tenant having proposed an improvement and the landlord having objected, the onus was on the landlord to demonstrate that the proposed improvement was not an appropriate improvement for the purposes of the Act. The onus was on the respondents to show why we should not approve what was proposed. But even if the onus was on the applicants, it had been discharged.
 The respondents’ position was, in essence, as set out in Answer 4 of their pleadings:-
“The ‘proposed improvements’ referred to by the Tenant are not appropriate to the holding in circumstances where dairy farming is a permitted activity on only certain parts of the holding comprising approximately 40% of the whole holding. As hereinbefore condescended upon, the Landlord should not be liable either to provide dairying equipment beyond that required to carry out dairying on approximately 250 acres, nor, equally, should the landlord be exposed to the liability to pay compensation for dairying equipment which is placed on the holding in excess of the requirement to carry out dairying operations on approximately 250 acres.”
 The respondents’ position was therefore based upon two parts of the holding, Carnsalloch and Broomrigg, not being available for dairy use as a result of the terms of the last sentences of Clause First of the lease. Clause First was unusual. It was unusual to provide for different types of farming on different parts of a holding. In Mr Lean’s submission the provisions relating to Broomrigg and Carnsalloch were permissive rather than restrictive. It was the use of the other parts of the holding which bore to be restrictive in as much as they were to be farmed “as a dairy, livestock and cropping unit only” (emphasis added) whereas all that was said about Broomrigg and Carnsalloch was that they were to be “farmed as livestock and cropping units”.
 As a matter of common sense, it was an odd proposition that a tenant was expected to conduct two different enterprises on the same unit. Had that really been the intention one would have expected it to have been an important consideration at the rent review. But Mr Gourlay’s evidence had been to the contrary. His evidence had been that these provisions of the lease had played no part in the rent review negotiations.
 With reference to the case of Taylor v Burnett’s Trustees, referred to in the respondents’ pleadings, the court had held, in approving some improvements and not others, that improvements should be restricted to such as were reasonably required to allow the tenant to carry on the type of farming specified in the lease. In that case the lease had been for use as a general purpose farm and the court had said that there was no liability for meliorations which may become necessary by law for any particular type or class of holding or farm.
 Taylor could be distinguished from the present case because in the present case the lease specifically allowed dairy farming.
 Turning to the meaning of the last two sentences of Clause First in the present lease, and particularly to the meaning of “livestock” in the final sentence, Mr Lean said that if we had been dealing with a separate lease of Broomrigg and Carnsalloch which had contained that provision, the landlord would not be liable for dairying improvements. That did not, however, mean that dairying on those units was excluded by the final sentence of Clause First. For dairying to be excluded there had to be specific, explicit exclusion.
 One often saw leases where the landlord had been anxious to avoid liability for improvements to support a particular kind of farming and that was clearly provided for in the lease in question. However, the present respondents had not done that. Not having inserted such provisions in the lease, the respondents were not now entitled to introduce such restrictions. In any event “livestock” included dairying.
 Asked by the court as to the meaning of “cropping” in the final sentence of Clause First and in particular whether the growing of forage maize would constitute cropping, Mr Lean said that it would. Similarly making silage was “cropping”. It did not matter whether something was simply cropped or combined: the act of growing it and cutting it was still “cropping”. So far as grass for grazing was concerned, a distinction had to be made between permanent grass and new grass planted for grazing or silage; the latter would be “cropping”.
 As to whether feeding a crop to dairy cows made the activity “dairying”, Mr Lean said that in the context of land worked as a single unit it would be part of a dairying enterprise. But if there are two separate farms and farmers, one involved in the growing of crops and the other involved in dairy farming, the growing of crops ultimately consumed on the dairy farm by dairy cattle would not be part of a dairying enterprise. If there was a distinction, as the respondents argued, between the penultimate and final sentences of Clause First that distinction should be applied consistently so that the growing of crops on Carnsalloch and Broomrigg was seen as a separate activity from the dairying carried on on the remainder of the holding.
 As to whether the rearing of dairy heifers was dairying, Mr Lean submitted that where the heifers were being reared as replacements for one’s dairy herd that would be dairying but otherwise not.
 For present purposes these distinctions may in any event be academic. Mr Allen’s evidence had been that the main limiting factor on the size of herd which could be kept at Abbey was the tenant’s ability to dispose of slurry. Mr Allen had accepted that, even if it was not to be used for dairying, Broomrigg at least could be used for the spreading of slurry for cropping purposes. Mr Allen’s calculations had been on the basis that only the ground at Carnsalloch would be unavailable for slurry spreading purposes and his reason for that exclusion had been practical: the difficulty of getting the slurry to Carnsalloch. But he had accepted that the applicability of his five mile radius limit was subject to a large number of variable factors, such as route and price of fuel. Mr Gourlay had said that he had clients who transported slurry as far as eight miles. Mr Whiteford’s evidence had been that if he was not able to use the umbilical he would transport the slurry to Carnsalloch by road, a distance of only six miles. If the court were to accept that that was a reasonable proposition then the whole of Carnsalloch became available for the purposes of slurry disposal.
 Even in the absence of the Carnsalloch ground, however, Mr Allen’s evidence had been that it would be possible to retain a herd of 330 milking cows plus followers. Mr Florey had said that he thought 250 to 300 cows would be appropriate and that 380 milking cows would be at the top end of what he considered reasonable. Therefore, even taking the restrictive view of the use which could be made of different parts of the holding, the evidence from the respondents’ own witnesses was that a dairy herd of similar size to that presently kept, was sustainable.
 Under reference to production 34 and the heading of “Fixed Equipment”, Mr Lean sought to demonstrate knowledge on the part of Mr Florey (1) that dairy cows were being housed in Cubicle Shed (1); (2) that the installation of 300 cubicles had been carried out for dairying purposes, and (3) that these cows, so housed, did not comprise the whole of the milking herd. In essence, therefore, the respondents’ position was that a milking herd of over 300 cows was appropriate on Abbey even if the use clause in the lease was to be interpreted in favour of the respondents.
 If the court were to accept that the present size of the enterprise was appropriate to Abbey, regardless of how the lease was to be interpreted, and that there was consensus between the applicants and the respondents that a new milking parlour required to be sited adjacent to Cubicle Shed (1), what then required to be determined by the court was whether the actual improvement proposed was “reasonable and desirable” for the efficient management of the holding.
 Asked by the court for his view on the competency, in terms of sec 39 of the Act, of the court approving certain parts of a notice but not others, Mr Lean submitted that in terms of section 39(2) the court could restrict the amount of any compensation eventually payable. The court therefore had the power to look at proposed improvements and state that part only of these improvements would be the subject of compensation. Reference was made to Gill at paragraph 26.16 and to section 36 of the Act, prescribing the measure of compensation payable. It would be preferable to identify the parts of the proposed improvements which were to be eligible for compensation rather than restrict the amount of compensation payable in some other, formulaic, way.
 However, on the facts of this case, a particular size of building was required for the housing of VMS machines and a different kind of building would be required for a traditional parlour. So this was a case of “all or nothing”. There was no point in approving only some of what was proposed.
 The test to be applied was whether the improvement was “reasonable and desirable for the efficient management of the holding”: Gill 21:16 and the cases cited there.
 The point in time at which that test had to be applied was the date on which the sec 38 notice was served.
 The respondents’ agents had recommended a substantial contribution to the building costs of an extension to hold a new milking parlour. The Trustees had not been able to make that contribution but had wished Mr and Mrs Whiteford well with their proposals. It ill became them now to dispute the manner in which those proposals were to be carried out.
 The benefits of a voluntary milking system had been described in the evidence of Mr Whiteford, Mr Davies and Mr Mather. These were labour efficiency, an increase in yield of up to 15%, better animal welfare and health, better management of dirty water and less time taken to put cows through the present parlour (because fewer cows would be milked there).
 Some time had been spent in the evidence on potential welfare issues but these issues related to the practice of “zero grazing” rather than the use of VMS systems. There were relatively easy management solutions to the welfare issues as had been discussed in the evidence.
 Asked by the court as to the standpoint from which the court was expected to assess the proposed improvements, Mr Lean said that there was an objective element to be applied. The phrase “for the efficient management of the holding” required the holding itself to be looked at. The value to an incoming tenant was a question for later, when compensation came to be measured. The court was therefore, not making any judgement as to the value of the improvement to an incoming tenant in due course but only deciding whether what was proposed was reasonable and desirable on agricultural grounds at present.
 Importantly, the proposals need not be the only reasonable and desirable way of effecting the improvement.
 Again in response to the court, Mr Lean resisted any suggestion that we had to consider any sort of potential negative effect on the landlord in the near future (in the event of an early waygoing). What the court had to decide was whether the improvements were reasonable and desirable at present and compensation was a matter to be dealt with later.
 Asked by the court as to whether we had to consider a theoretical possibility that an “improvement” might turn out to have a detrimental effect in the future in terms of the value of the holding, rent realisable or some other aspect of the landlord’s interest, Mr Lean said that there was an objective element to the test, but if the objective test was passed, a hypothetical incoming tenant was likely to consider that the improvement had value.
 With reference to the requirement, in Clause Sixth of the lease, for written consent on the part of the landlord to additions to fixed equipment, Mr Florey had said that he had not been concerned about not having been asked for written permission in the past. If the respondents were now to prevent the proposed improvements in that way they may be putting the applicants in a position where they were being forced to practise bad husbandry.
 As to the later improvement notice served, referring to more extensive works, in January 2009, we should accept Mr Whiteford’s evidence that it was not being proceeded with. If the applicants were to build a bigger extension than they had given notice of they would struggle to get compensation for such a building so nothing turned on the discrepancies between sizes given in the notice of improvement, on one hand, and the planning application, on the other.
 We should find that the proposed improvements were reasonable and desirable on agricultural grounds for the efficient management of the holding.
 Mr Lean, in conclusion, moved for the certification of Mr Bryn Davies as an expert witness.
 Mr Kermack moved us to withhold approval under sec 39(2).
 If we were not minded to do so, we should exercise our power in terms of said section, to grant consent conditionally upon a substantial reduction of the compensation which might otherwise be payable. In their pleadings the respondents had said that they should only be liable to pay compensation at 40 per cent of what would otherwise be payable but in light of it having become apparent that the applicants were seeking the voluntary milking system and incidental facilities in addition to, and not instead of, the existing milking parlour, a greater reduction may be appropriate, perhaps to 20 per cent.
 There was doubt as to the competency of the notice of 28 July 2008; as to whether it truly gave notice of the tenants’ intention as to both what the proposal was and the manner in which it was to be carried out. Reference was made to sec 38(1) of the Act. That was for the following reasons:-
(a) There was no disclosure of the number of voluntary milking systems proposed. Where each of those carried a cost of £120,000 that was a significant omission.
(b) The notice provided for “construction of a shed adjacent to the cattle shed” but when planning permission had been sought it was for extension of the present cattle shed.
(c) By the time of the notice of 12 January 2009 (production 23) “the shed adjacent to the cattle shed” was also to accommodate the toilet and office facilities, which had previously been intended for “the cattle shed”.
(d) The tenants’ planning application was for an extension 22% bigger than that proposed in the notice of 28 July 2008. The tenants had deliberately departed from the terms of the earlier notice. Mr Whiteford’s explanation for doing so was implausible.
(e) The evidence of Mr Mather that the design of the layout was a “moveable feast” and had not yet been finalised.
(f) The evidence of Mr Whiteford himself that he was not sure as to his final intentions, as between the notices of July 2008 and January 2009.
(g) The evidence of Mr Florey that Mr Whiteford had refused to tell him what the “end game” was to be. The notice of 28 July 2008 was, therefore, a point along the route rather than the destination. For all of the foregoing reasons the notice of 28 July 2008 was not a valid notice.
 As to the proper interpretation of the use provisions of Clause First of the lease, Mr Whiteford’s evidence that dairy farming was merely a type of livestock farming was inept and should not be accepted. These provisions of the lease, entered into in 1993 when Mr and Mrs Whiteford had already been in possession of the subjects for six years, reflected the proposal put forward by Mr Whiteford when he had offered for the lease.
 In interpreting Clause First one had to look at the whole of the lease. Reference was made to a provision at Clause Sixth in relation to milk quota for the farms of Abbey, Holm, McWhanrick and Crossleys (part). At Clause Seventh it was agreed that “in respect that the holdings of Abbey, Holm, McWhanrick and Crossleys (part) shall be used as a Dairy Farm it is agreed between the parties and understood that the existing buildings and fixed equipment are in a good and proper state of repair and are sufficient for the purposes of milk production and comply with existing legislation or regulations in regard to milk and dairies”. So, if one took the lease as a whole, as a contractual document negotiated from at least equal bargaining positions, it was clear that part of the farm was to be principally for dairy use and the other part was not for dairy use. Mr Whiteford’s argument that “Livestock” in the final sentence of Clause First included dairy cattle was implausible: it was neither the language of the industry nor the language of agricultural leases.
 The distinction between the two forms of farming could be tested by asking what was the purpose of putting the cow into calf. For a dairy cow it was to put the cow into lactation; for a beef cow it was for the production of a suckler calf. So, in Mr Kermack’s submission, where one had a dairy herd and young stock that entire operation was part of dairy farming and none of it was livestock farming.
 Silage, maize and whole crop cereal were all fodder crops and normally a farmer would only grow a fodder crop if he had need of it himself; he would not grow a fodder crop for sale. Although he referred to these as “crops”, Mr Kermack submitted that they were different from other crops. The old section 2 of the 1991 Act had made a distinction between cropping, on the one hand, and grazing and mowing, on the other. Historically they had always been different. “Cropping” in Clause First, therefore, did not include crops grown for fodder.
 Asked by the court what a combined cereal crop fed to a dairy herd would be, Mr Kermack said that would be cropping and gave as an example malting barley of insufficient quality for its intended purpose going instead for feed.
 Mr Kermack then turned to the provision in the lease prohibiting addition to or alteration of the buildings or fixed equipment without the written authority of the landlord. The absence of any objection on the part of the landlord in the past did not prevent reliance on this provision now. Similarly the absence of any objection on the part of the landlords to the transporting of slurry by umbilical across the river Nith would not bar the landlords from objecting now. Such transportation was an encroachment on the property of the landlords whether carried out by the present umbilical system or by any alternative structure built over the river. The evidence was that the present arrangements were a matter of concern to SEPA, to the landlords and to the District Salmon Fishery Board. The landlords would be entitled to put an end to it.
 The landlords were now free to insist upon written permission being obtained for additions or extensions to fixed equipment and buildings and to encroachments on their property because their failure to insist on their rights in the past had not resulted in prejudice to the applicants; Morrison-Low v Howison, Gibson v Milne. On the contrary there had been positive benefit to the applicants from the respondents’ forbearance to date.
 The test to be applied in deciding whether to approve an improvement, notice of which had been given in terms of section 38, was “whether the improvement was reasonably necessary on agricultural grounds for the efficient management of the holding”; Taylor v Burnett’s Trustees, Hutcheson v Wolfe Murray, Renwick v Rodger (No. 1), Renwick v Rodger (No. 2), MacKinnon v Arran Estate Trust, Fotheringham v Fothringham.
 In Taylor v Burnett’s Trustees a link had been made between the fixed equipment which a landlord was obliged to provide in terms of section 5 of the Act and improvements which were to be allowed in terms of section 52 and the Court had said:-
“It would be strange indeed if a tenant could under section 52 provide equipment (for which his landlord would ultimately be obliged to pay) which the landlord himself is not obliged to provide under section 5”. (Page 141)
 In applying the test there were three linking factors; the holding itself, the hypothetical incoming tenant, and the requirement to provide fixed equipment in terms of section 5. It also had to be remembered that what we were dealing with here was not whether the improvements should go ahead but whether the landlord should have to pay compensation for them in due course. That was where the hypothetical incoming tenant came into the picture.
 So far as consideration of the holding itself was concerned, the holding comprised the subjects let in the lease. It therefore had to be looked at as being subject to the constraints contained in the lease. In determining what was reasonably necessary on agricultural grounds for the efficient management of the holding the court required to take these matters into consideration.
 In the present case, dairy farming was a permitted activity only on Abbey, Holm, McWhanrick and Crossleys (part). Only arable and livestock farming was permitted on Broomrigg and Carnsalloch. The part of the holding on which dairy farming was permitted therefore extended to only about 240 acres. The holding with which we were concerned for the purposes of applying the test, therefore, was, in effect a holding of 240 acres.
 The test of efficient management should therefore be based on the size of dairy herd which was compatible with a holding of 240 acres and the carrying capacity should not be artificially enlarged by taking into account the ability to house the stock permanently indoors and to bring fodder in from elsewhere.
 Stock numbers for the purposes of applying the test should be based on all young stock being kept permanently on the holding rather than being wintered away. It should not be assumed that the hypothetical incoming tenant would be able to secure other land for the keeping of young stock and replacement heifers away from the holding.
 Consistently with this approach, the test should also be applied on the basis that the hypothetical incoming tenant required to dispose of all slurry on the holding and it should not be assumed that it would be possible to dispose of slurry off the holding.
 Again consistently with this approach, it should not be assumed that the hypothetical incoming tenant would be able to send slurry across the River Nith by the use of the present umbilical or any substitute therefore. The assumption would have to be that he would have to transport slurry to Carnsalloch by road.
 The hypothetical incoming tenant would be assumed to carry out some livestock farming on Broomrigg and Carnsalloch, in compliance with the terms of the lease, and also carry out some arable cropping.
 In determining his stocking numbers, the hypothetical incoming tenant would have regard to the ability of the holding to produce grass during the summer and would not wish to be reliant on fodder produced off the holding and bought in.
 The hypothetical incoming tenant would not necessarily wish to farm the holding intensively, as by a zero grazing regime, and would not want equipment which was of use only for such a regime.
 The hypothetical incoming tenant would have regard to the constraints on the holding, not only in terms of the lease and the physical geography of the holding but also by its inclusion in the Nith Valley Nitrate Vulnerable Zone. Account therefore had to be taken of the advice of Mr Allen, whose evidence had not been contradicted and was to the effect of the carrying capacity of the holding (without consideration of other ground being available or of the restrictions in the permitted use of Broomrigg and Carnsalloch) was 300 Holstein cows of the high yielding type, based on the retention of 25% replacement heifers.
 Given the extremely narrow margins for milk producers the hypothetical incoming tenant may well prefer to be rented only on conventional dairying equipment rather than a highly capitalised system. On Mr Mather’s evidence the cost of four VMS machines with associated facilities was in excess of £520,000 whereas the cost of a new herringbone parlour would be about £130,000 and a rotary parlour £250,000. The VMS machines would not replace the existing parlour but be in addition to it, whereas the alternatives of a new herringbone parlour or rotary parlour would replace the existing parlour and remove the difficulties regarding waste foul water in the yard. A rotary parlour would also probably achieve the desired result in terms of reduced manpower and man hours.
 In any intensive farming operation there was always the risk that public opinion might become hostile. It had happened in the poultry, pig and veal industries. Mr Allen had given evidence of his supermarket contacts disclosing deep concern about public perception and of his view that best practice involved all cows spending part of their cycle at pasture. The hypothetical incoming tenant would be at least alive to this risk and would be reluctant to become committed to an intensive dairying operation. The intensive farming option being pursued by the applicants was a high risk one, involving a very considerable capital investment and being dependent on unchanging public demand for cheap milk without concerns, rational or otherwise, as to how it was produced.
 Against the background of those factors, elaborated at greater length by reference to MacKinnon v Arran Estate, Strachan’s Trustee v Harding and, again, Taylor v Burnett’s Trustees, Mr Kermack’s submission was that the hypothetical incoming tenant in the present case would reduce the dairy herd to the order of 200 milking cows and would make do with a herringbone milking parlour.
 It might be the case that an incoming tenant would wish to retain a zero grazing system and might wish to retain a new milking parlour in Cubicle Shed (1), perhaps even a VMS system. However, no reasonable hypothetical incoming tenant would want to retain both the VMS system and the existing parlour; hence the respondents’ arguments regarding over-capitalisation.
 We should therefore withhold approval of the proposed purported improvements.
 If, on the other hand, we were minded to grant approval we should consider whether the proposed improvements would result in an element of over supply of fixed equipment, in which case a condition should be imposed reducing the amount of compensation payable in due course.
 What the court should not do was approve the proposed improvements on the view that if they resulted in over-capitalisation the landlord would be protected in any event because the value to an incoming tenant in that situation would be assessed accordingly. We should not proceed on that basis for the following reasons:-
(i) because sec 39 was an express statutory provision, available to landlords, which required the court to consider the question of future compensation at this stage.
(ii) because where, as here, the tenant had shown a willingness to commit a very high level of capital expenditure, the landlord was entitled to know, at this stage, how things were to stand as regards future compensation; and
(iii) because if a tenant was to over-extend himself by incurring too great capital expenditure that might, in due course, hazard his ability to pay his rent timeously. That might give the landlord little choice but to bring the tenancy to an end but, in that situation, the potential liability for a large amount of compensation for “a white elephant” would be a material consideration for the landlord. Again the landlord was entitled to know in advance where he stood as to liability for compensation.
 Mr Kermack moved us to reserve expenses and certify Mr Allen as an expert witness. His training and experience had been directly relevant to the issue before the court. He had not otherwise been instructed by either side. He had given his evidence in a clear and impartial way. He had no interest in the outcome of the case. He was truly independent.
 By contrast Mr Davies was not independent. He was otherwise employed by the applicants and could not be said not to have an interest. Mr Kermack opposed his certification as an expert witness.
 Finally, as to onus, it fell squarely on the tenant in the present case because on Mr Whiteford’s own evidence he was proposing to install a system which was three or four times more expensive than the alternative and it was therefore for the applicants to show that that system satisfied the test which we had to apply. The statutory provisions did not require the landlord to give any reason for his objection and only the tenant was allowed to bring the matter to the Court, so the onus was in terms of the statutory terms, was clearly on the tenant.
 Mr Lean made a brief response.
 In applying the test we should not confine ourselves to a self-sufficient holding as Mr Kermack had argued. The court had heard evidence of practice and the court was entitled to have regard to that evidence. Consideration of “the holding” could not be restricted to a self-sufficient holding. Otherwise how would matters stand in the case of a hill farm where hoggs were wintered away? That would not be an indication of the holding being over-stocked but an indication of good management.
 As to what size of herd Mr Allen would advise for a hypothetical incoming tenant, his evidence had been 330 cows plus followers.
 As to the point made by Mr Kermack about an incoming tenant not wanting to pay a high rent, it had to be remembered that the outgoing tenant would not have been rented on his own improvements. The two rents, therefore, would never be the same; the rent paid by the incoming tenant would always be higher.
 With reference to the linking of what is appropriate by way of improvement to what is required in the way of provision of fixed equipment by the landlord, it would not be legitimate to look simply at the standard applicable at the outset of the tenancy because the improvement may be proposed many years later and what was considered appropriate by way of provision of fixed equipment may have changed in the interim.
 As will be clear to parties from our selective recounting of the evidence, there were several matters explored at proof which we do not consider relevant to the task before us. These are:-
(i) much of the history of the applicants’ use of the holding including the purposes they had in mind when they first took on the lease and whether or not the respondents were aware of their increasing concentration on dairying ultimately to the exclusion of all else: that does not seem to us to be relevant to a proper consideration of what they now propose;
(ii) the improvement notice of 12 January 2009: the present application is concerned only with the notice of 28 July 2008 and the later notice has, in any event, been withdrawn;
(iii) discrepancies between the dimensions and certain other details of the proposed improvements as between the notice of 28 July 2008 and the applicants’ subsequent planning application: for reasons which appear more fully below we are concerned only with said notice; and
(iv) the prohibition of addition to buildings and fixed equipment contained in Clause Sixth of the lease: the effect of that prohibition is not something we are asked to decide in this application (although we take leave to doubt whether the whole statutory regime of compensation for tenants’ improvements can be set aside so simply - sec 53 of the 1991 Act may be relevant – besides which there may be considerations of the landlords being required to act reasonably in order to give the lease business efficacy).
 Our task is, instead, a very specific and focused one. It is whether to grant approval to the particular improvements contained in the notice of 28 July 2008.
 Accordingly we now turn to the matters which we consider to be properly in issue and to require decision by the court, the first of which is the validity of that notice.
 Mr Kermack argued that the notice of 28 July 2008 was not a competent notice for the purposes of sec 38(1).
 Sec 38(1) requires notice by the tenant to the landlord of the tenant’s intention to carry out the improvement “and of the manner in which he propose[s] to do so”. Mr Kermack argued that the present notice was defective because it failed to specify the number of robotic units to be installed. Given the cost of each individual unit, that was a significant omission. More generally, it was impossible to know what the applicants’ true intentions were in relation to this improvement. The picture was confused because of conflict between the terms of the present notice and both the later notice of 12 January 2009 and the application for planning permission lodged in March 2009. Moreover there had been evidence from Mr Mather of the location of the new milking facility being “a moveable feast” and from Mr Florey of Mr Whiteford’s reluctance to tell him what his “end-game” was.
 In assessing the validity of the notice we cannot look beyond its four corners. If, on the face of it, it gives notice to the landlords of an intention to carry out an improvement and of the manner of doing so that is enough, in our view, for a valid notice.
 The only one of Mr Kermack’s criticisms which is targeted at the terms of the notice is the failure in the notice to specify the number of robotic units involved. That criticism is certainly not without merit. It would obviously have been preferable to include that information. However we are not prepared to hold the notice incompetent on account of its omission. It is possible to regard it as a point of detail given that the notice makes clear the kind of milking system to be installed and that a very substantial investment is being contemplated. In that context does the actual number of machines – whether two, three, four or more – really matter? Looking at the terms of the notice it is clear what, in general terms, the tenants propose to do and how they propose to do it. We do not consider, therefore, that we are obliged to hold the notice incompetent – a step which would serve only to put parties back to square one – on the basis of that single omission. As it happens that omission was made good in the evidence: we now know that what is being contemplated is the installation of four robotic units and we have proceeded to a decision on the merits of the application on that basis. But our decision on the validity of the notice has nothing to do with that: that decision is taken solely on the terms of the notice and we adjudge these to satisfy the requirements of sec 38(1). We would observe, however, that it must be in the interests of a tenant giving notice under sec 38(1) to make it a full and detailed notice not only in order to avoid any challenge of the kind made by Mr Kermack in this case but in order to avoid the risk that some element of the improvement for which he hopes to be compensated at outgo is not in fact eligible for compensation because the landlord has not been given adequate notice of it.
 We return to the question of the number of robotic milking stations to be installed at paragraph  below but in the meantime, although we do not regard them as truly bearing on the competency of the notice, we should say something about the other matters raised by Mr Kermack under this heading. These were the discrepancies between the two notices and between the present notice and the planning application.
 We think the short answer to them is that this application is concerned only with the notice of 28 July 2008. That is all that is formally before the court and in his evidence Mr Whiteford predicated the applicants’ position on that notice. If we approve what is proposed in that notice and the applicants go on to build or install something else then, depending upon the degree of divergence between that of which notice has been given and the reality, they run the risk that what they build or install will not be eligible for compensation at the end of the day. If, on the other hand, they carry out the work conform to our approval the resulting improvements will be eligible for compensation at waygo. But all that matters for present purposes is that the applicants have tabled a particular proposal on which they are entitled to the adjudication of the court. We decide the question before us by applying the relevant test to that proposal in the context of the lease, the holding, the applicants’ farming operation and the circumstances of the wider dairy industry. Evidence as to other proposals or possibilities is irrelevant.
 Accordingly we hold that the notice of 28 July 2008 was a valid sec 38(1) notice.
 As is said in some of the cases we were referred to, the legislation gives no guidance as to how we are to exercise our discretion under sec 39(2). In the cases themselves the matter does not seem to have been the subject of much discussion or analysis nor of such full submissions as we have had, particularly from Mr Kermack, in this case.
 A variety of formulations for the test to be applied appears from the cases. Taking these chronologically, in Taylor v Burnett’s Trustees it was expressed as whether the improvements were “reasonably required to enable the tenant to carry on the type of farming specified in the lease”; in Fotheringham v Fothringham it was whether the improvement was “a reasonable project and desirable for the efficient management of the holding”; in Hutcheson v Wolfe-Murray “whether the actual improvement for which our approval is sought is reasonably necessary on agricultural grounds for the carrying on of the holding”; in Renwick v Rodger (No. 2) “whether the improvement is reasonably necessary on agricultural grounds for the efficient management of the holding”, a formulation which was also applied in MacKinnon v Arran Estate Trust. Lord Gill’s formulation of the test (at para 26.16) is “whether the proposed improvement is reasonable and desirable on agricultural grounds for the efficient management of the holding”.
 On the view that something is either necessary or it is not, we have difficulty with the “reasonably necessary” formulation. Absolute necessity is a very high test and will often be difficult to establish. Maybe that is why the cases refer to reasonable necessity. At all events none of the cases goes as far as to put the test in terms of unqualified necessity. On that basis we think it possible to reconcile the various formulations given in the cases but the formulation we prefer is that given in Gill, which is the learned author’s distillation of what is said in the cases, and that is the test we consider falls to be applied in this case.
 There was some discussion before us of the relationship between sec 39 of the Act, and the protection it seems to offer a landlord at this stage in the statutory procedure, and the protection which a landlord is in any event afforded under sec 36 at the tenant’s waygoing. There was also a suggestion in both the evidence (Mr Allen’s evidence as to what “over-capitalisation” might mean) and in Mr Kermack’s submissions that reasonableness has to be viewed from the perspective of the landlord. That might be on the view that what sec 39 is about is protecting the landlord’s position and that it is therefore necessary to try to anticipate what adverse results the improvement in question may have on his interests in the future. Or - and we think this is what Mr Kermack had in mind - looking at the proposal from the landlord’s point of view may be necessary as part of a test for over-capitalisation.
 In our view, however, that is to conflate two separate stages in the statutory scheme. It seems to us that at the present stage the court is acting as the gatekeeper to eligibility for compensation at the later stage. Sec 39 allows a landlord to object to an improvement as soon as it is mooted and to obtain from the court a ruling as to whether, if the proposal goes ahead, it is to be eligible for compensation at waygoing. Subject to the court’s power under sec 39(2) to impose terms as to the reduction of compensation at this stage, whether compensation is actually payable and, if so, how much, will only be known at waygoing when sec 36 falls to be applied. The purpose of secs 38 and 39 together is to allow parties to know whether an improvement is going to be eligible for compensation before it is carried out. In other words they are about letting parties know where they stand at the outset. A tenant who can show that his proposal is desirable and reasonable on agricultural grounds for the efficient management of the holding goes forward knowing that he will be entitled to be compensated for it if the improvement has value to a tenant succeeding him. A landlord who can show that the proposal is not reasonable and desirable is not left in uncertainty as to possible liability for compensation further down the line. The degree of protection offered to the landlord at this stage, therefore, is the right to have “improvements” which do not satisfy the test excluded from further consideration under Part IV of the Act at a later stage.
 Before we go on to consider the application of the test identified above we would say three things about it. Firstly, it is an objective test and we must not decide this case on the basis of what we think the applicants should be doing or what we would do in their place. Secondly, in a given situation there may be a range of possibilities all of which are reasonable. Some may be at the more adventurous end of the spectrum, others at the safe end and yet others somewhere in between, but all reasonable. Thirdly, and this is not necessary to our decision in this case, we think that where there are a number of ways of proceeding, each of which is reasonable on its own terms but one of which seems clearly preferable to what the tenant is proposing, the court would be entitled to refuse approval on the basis that in all the circumstances, including the availability of a better alternative, to go ahead with the proposed improvement would not be reasonable in the interests of the efficient management of the holding.
 Mr Kermack argued that for the purposes of applying the test we had to confine ourselves to the holding exclusive of Broomrigg and Carnsalloch. That was because what was being proposed was a dairying improvement and the lease prohibited the use of these units for anything to do with dairying.
 There is no doubt in our mind that the last two sentences of Clause First of the lease are intended to create a distinction in the permitted uses of Abbey, Holm, McWhanrick and Crossleys, on the one hand, and Broomrigg and Carnsalloch on the other, with the former enjoying wider permitted usage – despite the use of the word “only” at the end of the penultimate sentence – than the latter. It makes no sense to read these sentences as permitting co-extensive usages. The difference is in the absence of a reference to “dairy” or “dairying” in the final sentence. Although it would make more sense for the word “only” to come at the end of the final, rather than penultimate, sentence a fair reading of the sentences together can only mean that dairying is not permitted on Broomrigg and Carnsalloch.
 How far does that take us? Is the prohibition to be construed narrowly – so as to exclude only the keeping of milking cows and a milking parlour – or widely, so as to exclude anything connected to the keeping of a dairy herd – such as grazing of young dairy stock and the production of fodder for a dairy herd kept elsewhere?
 In considering that question we have been guided by two principles.
 The first is that we should assume that, since the purpose of the phrasing used was to limit the use which could be made of Broomrigg and Carnsalloch, the words used must be taken as delineating the extent of the limitation. We should, therefore, be careful not to extend the scope of the restriction by adding words which are not there.
 Bearing that in mind, we note that the permitted use of these parts of the holding is “as Livestock and cropping units”. By inference, therefore, what is prohibited is use as dairy units. The prohibition is not, therefore, of use of the subjects for anything to do with dairying, however remotely, but of use of the subjects as dairy units.
 The second principle is that the existence of the distinction requires us to look at the use made of these parts of the holding in isolation. The question is not, therefore, whether Broomrigg and Carnsalloch are being used as parts of a dairy unit – to which the answer is undoubtedly “yes” – but whether they are being used as dairy units.
 Against that background it seems to us that the question whether the applicants are using Broomrigg and Carnsalloch as dairy units has to be answered in the negative. Use of an area of land as a dairy unit connotes the use of the land primarily for milk production. To have a milking herd permanently accommodated on these parts of the holding would be to operate them as dairy units, even if the milking was done elsewhere, but to use the land for grazing of young or dry dairy stock and for fodder production for milking cows kept on another holding would, in our view, not be.
 The matter can be tested in this way. If Broomrigg and Carnsalloch were occupied by someone other than the applicants and that person used the land (i) to graze some of the applicants’ stock under some grazing or agistment agreement and (ii) to grow crops, including silage, which were then sold to the applicants for use at Abbey, would that person be using the land as a dairy unit? We think clearly not - he would not be producing any milk - and if that is right the answer does not change because Broomrigg and Carnsalloch are occupied by the same person as occupies Abbey etc.
 In our opinion, therefore, the reference to “Livestock” in the final sentence of Clause First does not have the effect of prohibiting the presence of dairy animals on these parts of the holding but only the use of these parts of the holding as dairy units.
 This interpretation does not deprive the prohibition of content. If, for example, the applicants were to give a sec 38(1) notice proposing the erection of buildings for dairy use on either of these units not only could we not approve such proposals (Taylor v Burnett’s Trustees) but the respondents could take steps to prevent them being carried out on the basis that to do so would be in breach of the lease.
 The application of the test we have identified, although not easy on the facts of this case, is, we think, in general terms a more straightforward matter than Mr Kermack made it appear. He sought to build linkages between the lease, the holding, the equipment a landlord is required to provide in terms of sec 5 of the Act and the value of the improvement to a hypothetical incoming tenant. Although interesting, we have not found these suggested linkages helpful because they seem to us to have the result of burying the essential simplicity of the test we have to apply under a weight of unnecessary and unwarranted accretions. What follows is largely an exercise in stripping these away.
 We entirely accept that the terms of the lease are relevant and, in particular, that only such improvements as are consistent with the usages permitted by the lease may be approved; Taylor v Burnett’s Trustees. The point made by the court in that case about sec 5 was that “[i]t would be strange indeed if a tenant could under section 52 provide equipment (for which his landlord would ultimately be obliged to pay) which the landlord himself is not obliged to provide under section 5” (page 141). That point was made in the context of the tenant seeking approval for improvements which were not consonant with the permitted purposes of the lease. That has no application in the present case where there is no conflict between the permitted purposes in terms of the lease and the intended improvement. We are however mindful of this court’s recent decision in Buccleuch Estates Ltd v Telfer (upheld on appeal to the Court of Session) to the effect that a landlord is not obliged, under sec 5, to modernise or upgrade equipment after the commencement of the tenancy.
 In his submissions as to the level of fixed equipment a hypothetical incoming tenant would expect to be provided by the landlord, Mr Kermack suggested that sec 5 applied in a wider sense; as a test of whether a holding was over equipped. Approaching the matter in that way, the question would be whether a landlord of a holding such as Abbey embarking upon a 1991 Act lease now would be obliged to provide a voluntary milking system as part of his obligation under sec 5(2)(a) to “provide such buildings and other fixed equipment as will enable an occupier reasonably skilled in husbandry to maintain efficient production as respects both (i) the kind of produce specified in the lease, … and (ii) the quality and quantity thereof”. The answer to that would clearly be “no”. On the evidence, production of milk can at the moment be maintained with tolerable efficiency at Abbey with the present equipment albeit that there is little dispute that the existing parlour is now far from ideal from a number of viewpoints.
 But that approach is not justified, far less required, by the statutory scheme which governs these matters. In that scheme the landlord’s obligation to provide fixed equipment and the tenant’s right, in certain circumstances, to be compensated for improvements are two separate sets of provisions. It is in the nature of improvements that they will result in something better than the landlord has provided or is obliged to provide and Part IV of the Act does not impose, explicitly or implicitly, a cap on compensatable improvements by reference to what the landlord is obliged to provide under sec 5. Instead the landlord’s protection takes another form; the two-stage protection of secs 39 and 36 to which we have already referred. There is no injustice to a landlord in this. A landlord will only have to pay compensation if the improvement is of value to an incoming tenant and if that is the case, all other things being equal, the landlord will be entitled to look for a higher rent from the incoming tenant. Sec 5 cannot therefore be used to fetter what can be done under sec 39(2) except to the extent of what was said by the Court in Tayloras read in the context of that case.
 Likewise the discretion we have to exercise at this stage is not to be fettered by reference to what may happen at outgo. Mr Kermack’s submission was that overcapitalisation may result in a situation in which the lease was less attractive to an incoming tenant because a high rent was being looked for to reflect that capitalisation.
That would clearly be prejudicial to the respondents’ interests and the risk of it happening was, in Mr Kermack’s submission, a reason for not granting approval at this stage. However an improvement which is reasonable and desirable on agricultural grounds for the efficient management of the holding is, by its very nature, likely to make the tenancy of the holding more attractive to an incoming tenant rather than less. If some intervening change of circumstance renders what looks like a reasonable and desirable development now a disadvantage at waygoing that is just a risk which is inherent in the statutory scheme but the landlord is protected to the extent of not having to pay compensation for the “improvement”. On that scenario it would be the tenant who would suffer potential capital loss.
 Mr Kermack also submitted that in applying the test of efficient management of the holding we had to limit ourselves to the holding as comprised by the 240 acres of Abbey, Holm, McWhanrick and Crossleys and limit stock numbers to the level which could be accommodated year-round on that part of the holding without resort to housing stock permanently indoors or to importing fodder from elsewhere. In other words we had to contemplate a self-sufficient unit and must not assume the availability of land elsewhere for wintering or the possibility of being able to buy in fodder or equivalent feeds. We had to approach matters in that way, according to Mr Kermack, because it could not be assumed that an incoming tenant would want to practise zero grazing or that land would be available to him for away wintering or that he would have available a supply of imported fodder or would want to be dependant upon such a supply if he had it.
 We have already dealt with the use which can be made of Broomrigg and Carnsalloch. So far as the rest of this submission is concerned we think it is flawed. Firstly, the point in time from which it views the matteris wrong. We have to view the matter as at the present, not as at waygo. Secondly, the matter does not fall to be approached from the point of view of an incoming tenant: that is simply not part of the test we have to apply. And thirdly we do not see why, for the purposes of applying the test, such a straitjacket should be applied to the way in which the holding is to be managed. Viewed from the present point in time, the restrictions which Mr Kermack propounds are artificial. The fact is that the applicants have access to land for wintering and that they can and do buy in supplementary feed. There is no reason to think that these arrangements cannot be sustained. Away wintering and the buying in of feed are common practices. They are also, in the appropriate context, sound practices. Indeed, as Mr Lean observed, they may even be required in the interests of good husbandry. As practised by the applicants in this case they are unexceptional and unobjectionable practices. Accordingly, although the situation would be different where the tenant’s farming practices were clearly unsustainable or such as would not be pursued by a prudent tenant, we are of the view that the correct approach for us to take in this case is to apply the test to the holding as presently farmed by the applicants.
 We do not mean to suggest that the size of the holding and its capacity for such things as grazing and production of feed are not relevant. The test we have to apply itself directs us to have regard to the characteristics of the holding: it is a test as to the efficient management not of the tenant’s business but of the holding.Thus the size and productive potential of the holding would have to be such as to provide a sound basis for the operation being performed or contemplated before we would be likely to approve of expenditure of the order being contemplated here. To take an extreme example, a large herd permanently housed on a holding without grazing for young or dry stock and incapable of producing enough fodder to contribute significantly to the feeding of the herd is unlikely to be regarded as a sound basis for large capital expenditure. (More fundamentally, there may be a question in a situation such as that as to whether one was talking in any meaningful sense about the management of the holding, as distinct from the tenant’s business, at all.) But it is quite another thing to say that a holding must be entirely self-sufficient in terms of grazing and feed. The judicious use of additional grazing and bought feed in accordance with good practice ought not to be a bar to approval of even high cost projects such as this one. Likewise, so far as slurry disposal is concerned, we ought not to exclude the possibility of land being available for the export of slurry should that become necessary.
 Another matter on which Mr Kermack touched in submissions was the effect such a large investment may have on the applicants’ ability to pay rent. We agree that an investment which was so extravagant that it would peril the ability of a tenant to remain in business would not be reasonable or desirable for the efficient management of the holding. But on the evidence we do not appear to be in that territory here. There was no direct evidence led of the applicants’ financial position. There was no evidence led as to how they propose to finance this improvement. But they appear to have farmed successfully at Abbey for 22 years now and Mr Whiteford in his evidence impressed us as a responsible farmer and businessman. He is someone who has been prepared to take reasonable risks in progressing his business but he is not, we think, a reckless individual. That impression was corroborated by what we saw at inspection: the state of the holding and the condition of the applicants’ stock were eloquent of good management. There is therefore no evidential basis for holding that the applicants’ business cannot cope with this scale of investment.
 We have also had to consider what place there is for arguments about “over-capitalisation” in applying this test. Although we heard some discussion as to what “over-capitalisation” might mean, we note that the respondents’ objection in terms of the covering letter which accompanied their formal notice of objection is in the simpler language of the farm being over-equipped if this investment goes ahead. In the context of a test as to whether an improvement is reasonable and desirable on agricultural grounds for the efficient management of the holding that is, we think, a more comprehensible and easily applied concept than one which involves projections as to whether the requisite investment is likely to be recouped.
 So, shorn of these accretions, the question remains whether the improvements listed in the notice of 28 July 2008 are reasonable and desirable on agricultural grounds for the efficient management of the holding.
 In answering that question the area of dispute between the parties is fairly confined. The respondents agree that it would be reasonable and desirable to have the milking parlour located in an extension of Cubicle Shed (1). That is where most of the milking cows are accommodated and to have the milking parlour there would shorten their walk, reduce the amount of time they would be on their feet and reduce the dirty water problems caused by their present walk to milking. What the parties disagree about is (a) the use of a robotic voluntary milking system rather than a conventional herringbone or rotary parlour and (b) the desirability of maintaining a second parlour in addition to such a new system.
 We heard evidence that the sensible thing for the applicants to do would be to put a new system capable of milking the entire herd into an extension of Cubicle Shed (1) and close the present parlour. That would solve most, if not all, of the problems caused by the location of the present parlour.
 That could be done with either a VMS system or a rotary system (there seemed to be doubt as to whether a herringbone parlour would be suitable given the size of the herd). But if it was to be done by a VMS system it would have to be one of at least five robotic units rather than four. What the applicants propose, of course, is to confine themselves to four robots and milk the rest of the cows through the existing parlour.
 From one point of view that seems a somewhat untidy solution. Although it would alleviate the problems associated with the present parlour it would not by any means eradicate them. It would mean that milk would have to be collected from two locations rather than one. So if one’s only priority was the eradication of present problems what the applicants propose would not be the best solution.
 But that is not the applicants’ sole priority. The greater priorities for them are to increase yield and reduce labour costs. They say these aims can only be achieved by the introduction of voluntary milking and what they propose is, in effect, a splitting of their herd into high and low yielders with the high yielders using the automatic system and the low yielders continuing to go through the present parlour. Presumably they do not think the introduction of voluntary milking for the whole herd would be cost-effective although we heard no direct evidence on that. Apart from these considerations, retaining the present parlour would have the advantage that it would be available for milking of cows which, for whatever reason, cannot be milked automatically and also possibly for emergency use in the event of failure of the voluntary system. Viewed from these perspectives the retention of the present parlour makes some sense and would not constitute over-equipping of the holding because it would be necessary for the milking of the whole herd.
 As to the competing advantages and disadvantages of the two types of system, we were not favoured with the evidence of a witness who was both well qualified to speak to the matter and independent. Mr Davies was probably the best qualified witness in this area because as a nutritionist the focus of his work is on how to maximise the yield of dairy cattle. But he was not truly independent in that he has an ongoing relationship with the applicants in the role of a consultant (although it is by no means necessarily the case that his evidence cannot be relied on because of that, as we discuss in the following paragraph). Mr Allen was truly independent but he did not seem to us to be as well qualified in this area as Mr Davies. In any event neither produced hard, objective, empirical data as to the performance of voluntary milking systems as against conventional systems. Perhaps no such data exists but the result was that their evidence tended to be impressionistic and anecdotal albeit supported by cogent reasoning. Mr Mather, although his company sells both voluntary and conventional systems, had an obvious vested interest in the completion of the sale of four VMS machines which he seems to have provisionally arranged with the applicants.
 Working on the basis of such evidence as we have, our conclusion is that VMS systems can be expected to deliver the advantages of reduced labour costs and higher yield for which the applicants are looking. Mr Allen doubted both propositions but he was not in a position to produce evidence to the contrary. The evidence of Mr Whiteford, based on the applicants’ researches including visits to Sweden and Holland, and of Mr Davies, based on the experience of some of his other clients, was to the effect that both reduced labour costs and higher yields – up to 15% higher on Mr Davies’ evidence – could reasonably be expected from the introduction of VMS milking. Although Mr Davies was not an independent expert it is reasonable to assume that he has the best interests of his clients at heart and that he attempts to give them good, sound advice. It is not in his interests that his clients go out of business as a result of having over-extended themselves. Accordingly, we hold it proven for the purposes of this case that introduction of a VMS system of four robotic units is likely to reduce labour costs by reducing the work force by at least one man and is likely to lead to an increase of milk yield of up to 15%. (Although it was obvious from the evidence that there are herds – including the respondents’ own herd at Muirside – which produce substantially higher yields than the applicants’ herd without the use of VMS milking, no one gave evidence as to how these levels are achieved or, more particularly, how the applicants might achieve them without the use of voluntary milking.)
 But we have to consider the “downside” as well as the advantages. The downside comprises a combination of the high cost of installation of such a system and certain risks and limiting factors to which the applicants’ business is said to be vulnerable. Some of these risks and limiting factors will, of course, apply whatever milking system is installed but the respondents’ argument is that it does not make sense to invest £600,000 in a VMS system as compared with £300,000 in a rotary parlour in the face of such risks and limitations.
 The apprehended risks are: (i) that consumer concerns may soon make “zero grazing” a thing of the past; (ii) that the NVZ Regulations may lead to the herd having to be reduced in size after 2014; and (iii) that SEPA concerns, or the occurrence of a spillage, may bring an end to the transportation of slurry to Carnsalloch by the present umbilical.
 So far as the first is concerned not even Mr Allen, who was very well placed to speak to supermarket concerns in this area, was saying that milking cows would have to be put to grass during lactation. What he was saying was that each animal would have to spend time out at grass in the course of its life and milking cycle. So young animals, in-calf heifers and dry cows could be out at grass in good weather. That of course already happens with young animals at Abbey and we do not see that it would be a problem for the applicants to adapt their present system so as to, for instance, let dry cows out to grass in the summer months. Accordingly, on the evidence, we consider that the applicants would be able to change their practices to satisfy market concerns should these concerns develop in the direction Mr Allen described.
 So far as the second is concerned, the only authoritative evidence as to the effect of the NVZ Regulations in relation to the holding came from Mr Allen. Part of his evidence was that at present it is permissible to apply 250 Kg of nitrogen per hectare per annum. On that basis, and assuming that all of the holding is available for slurry spreading purposes, 400 cows and 200 followers could be kept at Abbey. The derogation which makes that possible is, however, not expected to be available beyond 2014 when the limit of 170 Kg of nitrogen per hectare per annum contained in the Regulations will become enforceable. On Mr Allen’s evidence the result of that will be to reduce the stock carrying capacity of Abbey to 300 cows and 150 followers. (For all of which see Table 1 of Mr Allen’s report, production 28).
 So far as the situation after 2014 is concerned, it appears that the applicants will have to contemplate either reducing the herd size to around 300 cows and followers or exporting slurry if they can find available land. We heard no evidence as to what they intend specifically in relation to this problem. Likewise there was no evidence as to what the effect of a cut in herd size of these proportions on the applicants’ business might be, so we are unable to quantify the risk, if any, that such a reduction would pose to the survival of the business. The result is something of an evidential vacuum. So far as reducing the size of the herd, if that proved necessary, is concerned, it occurs to us that there may be advantage to the applicants in taking that course in that it could allow them to discontinue the use of the existing parlour as a second milking facility and have their operation concentrated exclusively around Cubicle Shed (1). Alternatively, Mr Gourlay gave evidence that in general terms there was no problem with the disposal of slurry in the area, although, of course, come the end of derogation there will be more demand for land onto which slurry can be exported. On the whole matter we are not persuaded that the expiry of this derogation poses such a threat to the continuity and profitability of the applicants’ business that we should refuse approval of the proposed improvements on the ground of this uncertainty alone.
 It is also convenient to say at this stage that although Mr Whiteford did not give us the impression of having given much serious thought to possible slurry disposal problems after 2014, he was aware of the expiry in 2012 of another derogation from the NVZ Regulations in terms of which he will require to have six months’ slurry storage available by 1 January 2012. He is clearly planning on meeting that requirement and that aspect of the effect of the NVZ Regulations causes us no concern.
 So far as losing the use of the umbilical is concerned, although the alternative of transportation by road is much more troublesome and no doubt more costly, the evidence was that it was viable. Six miles is only a mile longer than Mr Allen’s suggested radius and Mr Allen recognised that the applicability of this limit was subject to a number of variables. Mr Gourlay gave evidence of slurry transportation over longer distances in the Dumfries area. Mr Whiteford gave evidence of his ability and determination to transport the slurry by road if he had to, although, as with his approach to NVZ questions generally, we got the impression that not much detailed thought or analysis has yet been applied to the problem. (In fairness to Mr Whiteford, however, he did say in evidence that he takes advice from the Scottish Agricultural College on slurry disposal which show proper appreciation on his part of the importance of compliance with the Regulations.) Mr Allen himself suggested a partial solution in the form of installation of a slurry separation system which would result in the liquids being used on the Abbey side of the Nith and the more easily transportable dry, or solid, material being used on Carnsalloch. On the evidence, therefore, we accept that the transportation of slurry to Carnsalloch by road would be a viable alternative to the present arrangements should an alternative be necessary.
 It appears to us, therefore, that these challenges are not insurmountable nor indeed of such a magnitude as to make the proposed investment unreasonable or undesirable.
 In the whole circumstances, therefore, we are persuaded that what is proposed by the applicants is both reasonable and desirable on agricultural grounds for the efficient management of the holding. It is desirable because action is undoubtedly required to deal with the difficulties created by the existing facilities which, perhaps sooner rather than later, could pose a serious threat to the physical viability of the whole dairy enterprise at Abbey. Although the proposal must be regarded as much the more adventurous of two possible courses of action, both are reasonable because both have potential for improvement in productivity, in its widest sense, and for ensuring financial viability in a difficult market.
 As we have seen, the notice of 28 July 2008 was attacked for lack of specification as to the number of milking stations to be installed. Although we have not regarded that as a sufficient basis for holding the notice invalid, that lack of specification has in fact been made good in the evidence. From the evidence it is clear that what the applicants seek is approval for the installation of four robotic units. We have thought it right, therefore, in the interests of clarity and certainty for the parties going forward, to expressly limit our approval to four such units and the terms of our order reflect that. The possibility of the existing parlour becoming redundant and the requirement for its capacity to be replaced are for the future.
 We have considered whether conditions should be attached to our approval. In their pleadings the respondents suggest we restrict any compensation ultimately payable to 40% of what would otherwise be payable. That is on the basis that “it would be unjust to permit the Tenant to be compensated for dairy equipment sufficient for a farm of 620 acres when dairying is a permitted activity on less than 250 acres of the holding”. Given the view we have taken of the permitted use of Broomrigg and Carnsalloch, however, that restriction on acreage does not arise. In submissions Mr Kermack suggested that we should cap compensation at an even lower level because it was now apparent that the applicants were seeking to retain the present parlour as well as the new facility. However, having come to the conclusion that what is proposed is reasonable and desirable we do not see any basis for the imposition of conditions as to compensation or anything else. We have therefore granted approval unconditionally.
 Following our usual practice, we have reserved expenses, including Mr Kermack’s motion for certification of Mr Allen as an expert witness, and invited motions thereon within 21 days.
For Applicants: Mr H Lean, Solicitor, Aberdeen
For Respondents: Mr L Kermack, Solicitor, Edinburgh