These are 13 appeals against net annual values ("NAV’s”) entered in the Renfrewshire Valuation Roll and one NAV entered in the Glasgow Roll for the 2010 revaluation. The subjects are clinics, surgeries and health centres. The assessors have made the entries on the basis of the comparative principle, based upon office rents appropriate to locality. The appellant ratepayers contend that the entries should be made on the contractor’s basis.
 The parties have agreed a schedule of appeals and agreed values in joint production (“JP”) 19 as follows:
|Case No.||Description||Address||Valuation Office||Issued Values||Comparative Method||Contractors Method||Effective Date|
|1||Health Centre||20 Kennishead Road, Thornliebank||Renfrewshire||£48,000||£48,000||£54,500||1 April 2010|
|2||Clinic||56 Busby Road, Clarkston||Renfrewshire||£54,000||£54,000||£27,500||1 April 2010|
|3||Office||38/40 Seres Road, Clarkston||Renfrewshire||£28,500||£25,000||£9,250||1 April 2010|
|4||Health Centre||203 Main Street, Barrhead||Renfrewshire||£84,000||£84,000||£57,500||1 April 2010|
|5||Clinic||5 Boglestone, Port Glasgow||Renfrewshire||£37,000||£29,000||£26,500||1 April 2010|
|6||Health Centre||37 Bay Street, Port Glasgow||Renfrewshire||£160,000||£130,000||£85,750||1 April 2010|
|7||Health Centre||20 Duncan Street, Greenock||Renfrewshire||£400,000||£300,000||£146,000||1 April 2010|
|8||Health Centre||181 Shore Street, Gourock||Renfrewshire||£70,000||£54,000||£29,750||1 April 2010|
|9||Surgery||The Tannahill Centre, 76 Blackstoun Road, Paisley||Renfrewshire||£10,900||£10,900||£9,500||1 April 2010|
|10||Clinic||The Tannahill Centre, 76 Blackstoun Road, Paisley||Renfrewshire||£35,000||£35,000||£22,000||1 April 2010|
|11||Clinic||8 New Sneddon Street, Paisley||Renfrewshire||£67,000||£40,000||£19,600||1 April 2010|
|12||Health Centre||103 Paisley Road, Renfrew||Renfrewshire||£90,000||£74,500||£54,750||1 April 2010|
|13||Office||10 Ferry Road, Renfrew||Renfrewshire||£450,000||£390,000||£229,000||1 April 2010|
|14||Clinic||15 Quarry Street, Johnstone||Renfrewshire||£20,600||£13,500||£7,600||1 April 2010|
|15||Day Centre||15 Quarry Street, Johnstone||Renfrewshire||£21,700||£16,000||£8,300||1 April 2010|
|16||Health Centre||60 Quarry Street, Johnstone||Renfrewshire||£80,000||£60,000||£50,500||1 April 2010|
|17||Clinic||7 Bargarran Square, Erskine||Renfrewshire||£25,000||£20,500||£13,100||1 April 2010|
|18||Clinic||6 Sandyford Place, Glasgow||Glasgow||£169,800||£169,800||£62,000||1 April 2010|
 It can be seen that in many of the above entries the comparative method produces a figure less than the issued value. This is because parties have now agreed values to be produced by this method, so that on any view the appeals will succeed to some extent. Case 1 has been withdrawn, since the contractor’s method has produced a higher figure than the issued value, and it is accepted it is not possible for the issued value to be increased on an appeal such as the present. Case 3 is described as an “office” in the roll, but parties accept it is to be treated as a healthcare subject like the other cases. Cases 7, 8 and 13 have been agreed by the assessor to fall under the contractor’s principle. The assessor will be amending the Roll to the appropriate figure and accordingly no order is required from us in respect of those three cases. Fourteen cases remain to be determined.
 The hearing took place on 4 June 2014 and two succeeding days. The appellant ratepayers were represented by Mr Christopher Haddow QC who led evidence from Ms Jenny Mitchell MRICS of GVA James Barr, building surveyor; Mr Stephen Robertson MRICS of GVA James Barr valuation surveyor and Mr Tom Davidson MRICS, GVA James Barr rating surveyor. The assessors were represented by Mr Steven Stuart QC who led evidence from Ms Jacqueline Dell MRICS, senior valuer in the Renfrewshire section of Renfrewshire Valuation Joint Board and Ms Ashley Weir, chartered surveyor, valuer at the Glasgow City Assessor’s Office. We carried out site visits on 19th June to appeal subjects 5, 6, 10, 14, 15, 17 and 18. We also visited settled appeal subjects case 7 and Renfrewshire Assessor’s comparison 104.
 The appeals were referred by the Tribunal to itself following a successful appeal by the appellants against a refusal to refer by Valuation Appeal Committee. The Tribunal accepted that the case raised a fundamental or general issue likely to be used as a precedent in other cases namely whether local office rents provided appropriate comparison for this type of healthcare subjects. In its decision of 25 September 2013 the Tribunal noted that a potential complicating factor was the impact of the Primary Medical Services (Premises Development Grants, Improvement Grants and Premises Costs) Directions 2004. We understood that under the Directions certain “notional” and actual rents are reimbursed to certain occupiers of healthcare premises by the NHS. The Directions were said to produce artificial figures. The assessors undertook not to rely upon such notional rents, or actual rents inasmuch as these were reimbursed under the Directions, since the potential for reimbursement could be said to have tainted the fixing of the actual rents to be reimbursed. The undertaking was thus clarified at the main hearing. We were also informed that there remain appeals referred to the Tribunal concerning 2010 revaluations in Lothian, where the Directions are in issue. Those appeals await the outcome of certain proceedings before the Lands Chamber in England, where, we were informed, there was a hearing underway. The Directions were touched upon in the evidence, but were not discussed in detail.
Local Government (Scotland) Act 1966, section 15.
Valuation for Rating (Decapitalisation Rate) (Scotland) Regulations 2005, regulations 2 – 4; “the 2005 regulations”.
Town and Country Planning (Use Classes) (Scotland) Order 1997.
Callander Abbots & Dobbie Forbes Ltd v Assessor for Stirlingshire 1962 SLT (N) 58
Edinburgh Merchant Co v Lothian Assessor 1981 SC 377
Fife Assessor v National Trust of Scotland  RA 501
Glasgow City Assessor v Monti Marino (Glasgow) Ltd 2013 SC 124
Leith Docks Commissioners v Assessor for Edinburgh 1960 SLT 43
North British & Mercantile Insurance Co v Assessor for Edinburgh 1908 SC 601
Post Office v Assessor for Fife Region 1981 SC 214
Rolls Royce plc v Assessor for Renfrewshire VJB 2013 SC 131
Spudulike Group Ltd v Tayside VJB Assessor  RA 91
Stirlingshire Assessor v Myles & Binnie 1962 SC 530
Woodrow v Lothian Region Assessor 2002 SC 530
Armour, Valuation for Rating, 18-05, 18-09, 18-10, 18-11, 18-12, 19-19, 19-28, 19-39, 19-40, 19-56, 20-35.
 We find the facts to be as follows. All areas stated are net internal areas (prior to any area reduction for application of the comparative basis), unless otherwise stated.
 Appeal case 2 is entered in the Roll as a “Clinic” at 56 Busby Road, Clarkston. It is in East Renfrewshire. It is a purpose built health centre (1967/68) with dedicated car park. It is owned and occupied by the appellants. It is on two floors. Its area is 641 sqm. It is in a busy commercial location amongst shops, offices, a train station and houses. It provides a long list of medical services as well as chiropody and physiotherapy. The ground floor contains a mixture of treatment type rooms and the first floor is office accommodation including accommodation occupied by health visitors and school nursing services. The treatment rooms have vinyl floors and sinks.
 Case 3 is entered in the Roll as an “Office” at 38/40 Seres Road, Clarkston in East Renfrewshire. There is a shop at the front and purpose built doctors surgery or clinic to the rear built circa 1985, and car park at the rear. The shop section is let and there is a passing rent for the shop. The surgery at the rear is owner occupied by the appellants. On Seres Road there are two parades of local shops, a fire station, a church, a medical practice and houses with a train station close by. The internal area is 206 sqm. The property provides treatment from a mental health team. There is a waiting area, interview rooms and offices.
 Case 4 is entered in the Roll as “Health Centre” and was the former health centre at Main Street, Barrhead, East Renfrewshire. It closed in April 2011 and has subsequently been demolished. At the material time it was owner occupied by the appellants. It was a purpose built health centre circa 1980 with dedicated car park. It is within a prime commercial location within Barrhead. The street has a shopping centre, council offices and a sports centre. Asda are constructing a new shop on the site and a new health centre has been built close by. It was on a ground floor. Its area was 1189 sqm. It located two or more doctors’ surgeries and the Community Health Partnership. The services provided in addition to GP services were physiotherapy, speech and language services, podiatry and mental health care. There were offices used by school nurses, district nurses and health visitors. The premises appear to have consisted of the order of 70 rooms.
 Case 5 is entered in the Roll as “Clinic” at Boglestone, Port Glasgow in the Inverclyde area. It is a modern, single-storey purpose build clinic circa 2002 with dedicated car park. It is owned and occupied by the appellants. The immediate surroundings include a leisure centre, a doctor’s surgery, a parade of shops, a school and a nursing home. It is within a predominantly residential area. Its internal area is 371 sqm. A wide range of clinical services are provided. There is a large waiting area and over 30 rooms including a 40 sqm room.
 Case 6 is entered in the Roll as a “Health Centre” at Bay Street, Port Glasgow, Inverclyde. It is a purpose built health centre built 1977, extended 2001 with dedicated car park. It is owner occupied by the appellants. It is a two-storey building located in a central part of Port Glasgow. There is a library, swimming pool, shops, offices and houses nearby. The health centre is adjacent to the bus terminus. Numerous clinical services are provided there. The floor area totals 1,867 sqm of which the majority is on the ground floor. There are over 90 rooms including consulting rooms, treatment rooms and offices. There are four doctor’s surgeries, and the Community Health Partnership and social work department are also located on the subjects.
 Case 9 is entered in the Roll as “Surgery”, the Tannahill Centre, 76 Blackstoun Road, Paisley. The surgery is within a purpose build centre (circa 1995) with shared car park and dedicated doctor’s car spaces. The Roll has recently been corrected so that the occupiers are Drs Latif and Mahood. The subjects are owned by the Tannahill Centre Ltd. There is a lease to the doctors as tenants produced at Renfrewshire Assessor (“RA”) 1 tab 9. The Tannahill Centre within which the surgery is located is a multi-use community facility located in the Ferguslie Park area of Paisley. Much regeneration work has taken place in this formerly deprived area. The Tannahill Centre is at the hub of the community and the centre contains a library, some shops, offices and the clinic which is case 10. Nearby there is a police station, offices, a church and houses. Its area is 160 sqm. There is a reception, waiting room and four consulting rooms and ancillary accommodation.
 Case 10 is entered in the Roll as “Clinic”, the Tannahill Centre, Blackstoun Road, Paisley. This is the clinic within the same building as case 9. It is occupied by the appellants. The property is rented with the passing rent of £1 if asked. The net area is 424 sqm. over two floors The clinic provides various services including child health and dental services. There are dentist rooms, a room used for the baby clinic, a meeting room and administration rooms holding records for schools within the area on the ground floor. On the first floor there is office accommodation. There are over 20 rooms in all.
 Case 11 is entered in the Roll as “Clinic” at 8 New Sneddon Street, Paisley. It is owner occupied by the appellants. It was purpose build circa 1960. It is on ground, first, second and third floors totalling 494 sqm. The property is within Paisley Town Centre. There are nearby shops, restaurants, a train station, car parks, a bank, a car sales garage and offices. It is a central commercial location. Various clinical services are provided. The ground floor mainly comprises rooms relating to chiropody and the first floor is mainly consultation rooms. The second and third floors are used as office accommodation. There is also a treatment room and a lab-type area.
 Case 12 is entered in the role as “Health Centre” at 103 Paisley Road, Renfrew. It is a purpose built clinic circa 1980 with dedicated car park. It is owner occupied by the appellants. It is on the periphery of Renfrew Town Centre. The immediate surroundings comprise a fire station, a library, a public house, shops and houses. It is occupied by the Community Health Partnership which provides a child health service and a looked after children service. It houses offices of the rehabilitation and enablement services team. It provides speech and language services, children’s physiotherapy and occupational therapy services. The net area is 1,114 sqm, on ground and first floors. There are various treatment rooms and offices. Many of the rooms are fitted out like offices since the treatment provided is for children and individuals with addiction problems and the aim is to create a relaxed environment.
 Case 14 is entered as “Clinic” in the Roll at 15 Quarry Street, Johnstone. It is situated in the first floor of a purpose built clinic circa 1930, with dedicated car park. It is owned and occupied by the appellants. It is a traditional stone building. It is located in the periphery of Johnstone town centre. The immediate surroundings comprise a police station, a church, a public house, the town hall, a dentist, shops and flattered accommodation. The property was refurbished circa 1996. Its net area is 195 sqm. It has various offices, a meeting room, a board room and ancillary accommodation. At the material time it was used to provide psychiatry and psychology services both on site and as a base for such services to be provided at patients’ homes. The psychiatry and psychology team moved out subsequently. The subjects are now used as back office accommodation and do not provide any treatment or receive patients on the premises. There was no significant change in the fit out of the building between the two uses.
 Case 15 is entered as “Day Centre” at 15 Quarry Street, Johnstone. It is the ground floor of the same building in case 14. At the relevant time it was occupied by the Archdiocese of Glasgow Social Services and was used as a day centre. The Archdiocese vacated the property and on 31 August 2011 the NHS moved a dietician team to the ground floor. It consists of various offices extending to 208 sqm. It is owner occupied by the appellants. Internally it is similar in appearance to case 14.
 Case 16 is entered as a “Health Centre” in the Roll at 60 Quarry Street, Johnstone. The building is a purpose built health centre circa 1992 with dedicated car park owned and occupied by the appellants. It is a two‑storey building. It is also located in the periphery of the centre of Johnstone close to the above two subjects. The ground floor consists of a doctor’s surgery and the Community Health Partnership which provides services such as retinopathy, chiropody, speech and language, physiotherapy and family planning services. The first floor provides back office accommodation for staff including district nurses. The net area is 860 sqm. There are about 50 rooms including a large waiting area. There are treatment rooms and offices.
 Case 17 is entered in the Roll as a “Clinic” at 7 Bargarran Square, Erskine. It is a purpose built clinic circa 1980 with shared car park. It is owned by Renfrewshire Council and is let to the appellants. It is single-storey and housed within a larger building owned by the council. The clinic is within the Bargarran Shopping Centre in Erskine and the immediate surroundings comprise a community centre, a school, a public house, various shops and, predominantly, local housing. There is a bus service. The subjects house a doctor’s surgery and the Community Health Partnership which hosts services such as chiropody, speech and language therapy, psychology and district nurses. There are treatment and consulting rooms, offices and a main waiting area. There are about 16 rooms. The net area comprises 294. sqm.
 Lease documentation was produced for case 17 under RA 1 tab 17. The original lease was between Renfrewshire Council and the Scottish Ministers for a period of 10 years from 28 May 1999 with an initial rent of £24,500. There was a review date at the fifth anniversary, at which the rent increased to £25,000 from 28 May 2004 in terms of a minute of agreement. Paragraph 1.3 b of the schedule provides that one of the assumptions at rent review is that the premises are available to let by a willing landlord to a willing tenant. Paragraph 4 provides that the revised rent shall be the open market rent at the review date. The documents also include a report by an officer to a committee of the council dated 20 April 2011. This indicates there has been provisional agreement between Scottish Ministers and the council to extend the lease for a further 10 years from 28 May 2010. It mentions the rent being “reviewed” to £31,500 as from 28 May 2010. It would appear the lease had been continuing on tacit relocation until 28 May 2010. The report also mentions that the district valuer was acting on behalf of the Scottish Ministers in the extension negotiations.
 Case 18 is entered in the Roll as “Clinic” at 2 Sandyford Place, Glasgow. It is owned and occupied by the appellants. It comprises former townhouse dwellings at 2-6 Sandyford Place built around 1860, converted in the 1930s to form an annex to the Glasgow Eye Infirmary and again in 1999/2000 to form a family planning and sexual health centre. It is a listed building to the west of Glasgow city centre. There are basement, ground, first, second and various mezzanine floors. The internal area is agreed at 1,781 sqm. Quite how many rooms there are used for consulting and examination was not made entirely clear, but could be of the order of 40, and in all the building comprises over 70 rooms including offices, a library, laboratory, waiting/reception areas, crèches and a staff training room. The building is laid out in a somewhat diffuse manner. Many of the first and second floor rooms have lightweight partition walls. It is considered to be a city centre location.
 One factor which all the subjects have in common is that at the material time they were predominately used for patient care in which patients will attend the subjects for consultation and/or treatment. The extent to which there is ancillary office use, and/or additional office use unrelated to patient care on site, varies. Certain subjects such as the surgery case 9 are dedicated in their entirety to onsite patient care with offices ancillary to that. Others to a lesser degree. Case 14 had just over half its use devoted to onsite patient care and consultation, which appears to have been the smallest proportion of all the cases.
 Much of the foregoing has been taken from JP 1- 18 which provides agreed factual information. The joint productions also provide agreed factual details for the now withdrawn or settled cases 1, 7,8 and 13. We do not set out the details here, but where we mention these cases we have adopted the agreed facts.
 The above subjects are typically characterised in physical detail by somewhat “hard” finishes. There are numerous cellular consultation rooms, soundproofed partitioning or blockwork around the consultation rooms, wash hand basins in the consultation rooms, large waiting areas, wide corridors and doors and clinical rooms finished with vinyl floors and walls often hipped at skirting board. There are no hard and fast rules but there is something of an exception in subjects used for treatment of children and for treatment of addiction/ mental health type issues where a softer finish to the treatment rooms is used. This is the case for cases 3, 12, 14 and 15.
 Scottish Assessors Association (“SAA”) Practice Note 21 for the 2010 revaluation is headed “Valuation of Clinics and Health Centres”. It provides as follows:
1.1 This Practice Note is intended to provide guidance on the valuation of Clinics, Health Centres and Surgeries etc which are occupied by local Health Boards, Medical Practices and the like. There is a wide range of subject types and sizes covered by this Practice Note ranging from the large multi-functional Health Centres to be found in cities, towns and rural areas to Medical Centres and Surgeries operated by General Practice groups which can be located in virtually any building type or location.
1.2 Generally, there is little to distinguish between Health Centres and Surgeries, save for the fact that in any given location Health Centres may be relatively larger than Surgeries and will contain additional accommodation supporting activities such as dentistry, chiropody, speech therapy, physiotherapy, etc.
2.0 Basis of Valuation
2.1 Comparative Approach
In situations where rental evidence exists for clinics, surgeries or health centres it will be appropriate to value such subjects by reference to the prevailing level of rents.
Measurement will be on the basis of Net Internal Area
2.2 Contractor's Basis
The following types of Health Centres/Clinics/Surgeries would normally be expected to be valued on the Contractor's Basis. It is unlikely that suitable rental evidence will exist for the following types of property and therefore the Contractor's Basis may be used.
(a) Purpose built (or purpose adapted) specialist units, eg, Dialysis Units or Diagnostic Centres.
(b) Health Centres converted from other specialist buildings, eg, hospitals or schools.
(c) Health Centres located adjacent to and operated in tandem with Hospitals (general/cottage).”
 Practice Note 21 for the 2005 Revaluation had similar passages to those quoted above. However paragraph 2.1 (Comparative Approach) contained the following additional passages:
“Such rental transactions may indicate a level of value commensurate with local evidence for offices but conversely may set their own level of being higher or lower than the local office market.
Caution should therefore be exercised in blindly applying local office rates/m2 for this class of subject.”
Paragraph 2.2 in addition to the three examples (a), (b) and (c) given above for the contractor’s basis also mentions:
“(b) Large purpose built Health Centres exclusively or almost exclusively occupied by a trust/health authority
(e) Purpose built Health Centres in rural areas serving several small communities
(f) Purpose built Health Centres located in housing estates.”
 SAA Practice Note 21 for the 2010 Revaluation relates to valuation of offices. It applies to:
“the valuation of offices and premises which are, by reason of their situation and physical characteristics, similar to offices.”
The Note provides for the comparative principle to be used based on local rental evidence.
 The appeal subjects had all been valued by means of the contractor’s principle for the 2005 Valuation. However, both the Glasgow and Renfrewshire Assessors took a fresh start for 2010. They adopted a different approach for the appeal subjects and valued them by means of the comparative basis by reference to office rentals.
 Parties have agreed the NAV’s which are produced by means of the comparative and contractor’s methods respectively, in terms of JP 19. While the figures are agreed for the purpose of the exercise, parties maintain their competing positions as to which method should be used. Although the appellants agreed figures for local office rents on this basis, they expressed the opinion there would be no market for the subjects as offices. It can be seen from JP 19 that the NAV’s are almost always significantly higher in the case of comparative valuations as opposed to contractor’s valuations. This can be partly explained by the fact that at the tone date, 1 April 2008, a reasonable landlord investing in property would have expected a net initial yield of 10%; ie a rent providing an initial return on capital at about this level per annum. This would be one of the considerations for rents struck on the open market at this time. On the other hand the 2005 regulations provide for a decapitalisation rate, where lands and heritage are being valued in accordance with the contractors basis, of 3.33% in the case of healthcare property. This is the assumed cost of borrowing money for the hypothetical construction of such subjects.
 Valuations for a further eight health centres and clinics occupied by the appellants were agreed with the Renfrewshire Assessor on the basis of the contractor’s method. These are detailed in Appellants Production (“AP”) 15 comprising appellants’ comparison list numbered 1-8 and corresponding details in AP16 . Numbers 9, 10, 11, 12 and 13 are community centres and libraries situated in the same building as appeal cases 9/10, 17, 9/10, 12, and 17 respectively. The community centres and libraries were all agreed and valued on the comparative principle. Thirty-five health centres and clinics occupied by the appellants were agreed with the Glasgow City Assessor on the contractor’s principle. These are numbered 14-50 in AP 15. The appellants further produce comparisons 51 – 105 consisting of health centres, clinics and one surgery all valued on the contractor’s basis in other valuation areas. The appellants detail their comparisons in AP 16. The contents of AP 15 and 16 were not factually disputed. It would be impractical to set out the extensive detail here, but we have adopted the information as part of our findings inasmuch as the details are mentioned later.
 The Renfrewshire Assessor provided tables of comparisons at RA 17 and 23. Numbers 1-352 are in East Renfrewshire, Renfrewshire and Inverclyde and the remainder 353-380 are in other valuation areas. With the exception of 9 examples (62, 63, 123,124, 203-7 highlighted in green on RA 17 where contractor’s method was used) all were based upon the comparative principle as offices. The details of the assessor’s comparisons as contained in AP 2A were uncontentious. Again we do not attempt to narrate them here, but adopt the information as part of our findings inasmuch as the details are mentioned later.
 The comparisons relied upon by the Glasgow City Assessor are listed in Glasgow City Assessor Production (“GCA”) 2. There are 18 “non-office” subjects in the west end for which valuations were agreed on the comparative basis. These include nurseries, clinics, surgeries, and a club. Some of these properties have an actual rent. Comparison subjects 19 - 37 all provide actual office rents in the same locality together with rateable values as in fact applied on the comparative basis. Details regarding these were produced. GCA 3 is a list of subjects entered in the Roll as clinics or health centres within Glasgow valued upon the comparative principle, with supporting details. GCA 4 is a list with supporting details of four of NHS clinics and health centres in Glasgow in which the comparative method was used and appeals with agents withdrawn. GCA 5 is a list of 50 clinics and health centres in Glasgow valued on the contractor’s principle. GCA 6 details four subjects entered in the Roll for Glasgow as offices and contains supporting detail. The factual information as far as contained in the foregoing productions was undisputed. Again we do not attempt to narrate the details here, but adopt the information as part of our findings in fact inasmuch as it is discussed later.
 Ms Mitchell visited each of the appeal subjects. She advised on the estimated costs of converting each from the existing layout use, which she considered to comprise predominately clinical rooms and ancillary space, to open plan office accommodation of a mid-level standard.
 She split the buildings into three groups and carried out a detailed analysis of one property within each range; namely small (GIA less than 500m2) – Erskine Health Centre (case 17); medium (GIA greater than 500m2, less than 1,000 m2) – Gourock Health Centre (case 8); and large (GIA greater than 1,000 m2) – Greenock Health Centre (case 7).
 She assumed the majority of partition walls forming clinical rooms are formed in masonry, it being understood this was to reduce levels of sound transmission between rooms. She assumed that 70% of existing partition walls could be removed without affecting structural integrity. She assumed that all consulting rooms are provided with sinks which require to be stripped out and associated services disconnected, and that existing ad hoc wc provision is stripped out and new toilet facilities will be provided. Various other assumptions were made including that existing floor coverings, wall linings, ceilings and so on will be stripped out and renewed to a mid-level standard. The cost estimates summarised on page 9 of her precognition were extrapolated to construction rates which are used in the study by Mr Robertson.
 Ms Mitchell’s exercise did not seek to quantify a lower quality conversion to offices whereby the offices would retain the cellular characteristics of the subjects without creation of open plan space.
 Mr Robertson considered that with the exception of the Sandyford Clinic, the health centres and clinics are generally located in areas regarded as secondary or tertiary office markets, characterised often by weak office demand. Most of the speculative office supply in these locations is by way of conversion of existing buildings to provide low cost accommodation.
 General Practitioners have been encouraged to form group practices. The grouping together of GPs dictates the layout of many of the appeal subjects which feature large reception areas of which there are often ancillary waiting rooms for the individual practices. The GP’s consulting rooms and ancillary services are clustered in pods off the main waiting areas. The consulting rooms are in a linear arrangement off the central corridor, which are cellular in nature. In his opinion this layout suits the health service but is poorly configured for use as mainstream offices which tend to have small waiting and reception areas. In particular, modern offices are increasingly open plan rather than cellular in nature.
 Generally his view of the appeal subjects was, with certain exceptions, that they contain a large amount of space in areas with limited office demand and that the location, size, layout and specification of the premises were ill-suited for mainstream office use. In many of the properties there would be a risk of unusable voids for which empty business rates would still have to be paid (apart from case 18 being exempt as a listed building). His calculations showed that the cost of converting a small, medium or large clinic or health centre to speculative offices was likely to be commercially unviable. He accepted that low cost accommodation as might be used by small or medium businesses were not generally open plan. The demand for such offices however tended to be for a requirement less than 200 sqm GEA. If a more modest conversion was carried out so as to retain the cellular nature of the building, this would mean the rent would have to be reduced below the assumptions he had used . He had not quantified or calculated the result of such an exercise, but in his view if the proposal was a “cheap” conversion the project would still be unviable for offices.
 In Mr Davidson’s opinion the specifications of the appeal subjects including numerous cellular consultancy rooms and “hard” physical features made them dissimilar to offices.
 The conclusion to be taken from the advice of Ms Mitchell and Mr Robertson was that the cost of adaptation would outweigh any potential rental return and so there would be limited or no demand from general office users for the appeal subjects in their actual state. Thus in his opinion there would be an absence of a willing tenant for the appeal subjects in their actual state for use as offices, or to a landlord as an office development project.
 In his opinion all the appeal subjects were used predominately for patient care at the material time. This is the basis on which they fell into the category of “health centres, clinics and surgeries.” In his view regarding Renfrewshire Assessor comparison 104, namely Crown House, King Street, Greenock, only about 10% of the building was used for patient care. As regards comparison 115 (Cathcart Street, Greenock) the figure was more like 7%. We should point out here that our impression of Crown House was that more than 10% of it was used for patient purposes since a large part of the ground floor comprised consulting rooms in a secure area, and we had no reason to think the characteristics had changed since the “as at” date of 1 January 2010. Nevertheless we accept the majority of the building was in office use.
 He pointed out that the new approach of assessors significantly increased the values where the comparative method was used. There was no new evidence to support the changed approach. The assessor’s position was unclear since the eight health centres and clinics (AP15/ 1-8) had been agreed on the contractor’s method, as well as the three appeal subjects which have now been agreed. He noted that AP 15 no’s 9-13 were libraries and community centres located within three buildings where 4 appeal subjects, namely 9, 10, 12 and 17 are located. This appears inconsistent because the libraries and community centres have all been agreed on the contractor’s basis whereas the health properties within the same buildings have not.
 The appellants’ comparisons AP 15 no’s 6, 34, 38, 40, 42, 44, 45, 49 and 61 are health centres or clinics including clinics with Sandyford “hubs” providing the same service as the Sandyford Clinic. The nine comparisons, with seven being in the City of Glasgow have all been valued and agreed with assessors on the contractor’s method.
 In the case of all GP and dental surgeries, in which the subjects are occupied by the respective doctors and dentists, these would all be covered by the 2004 Directions. In the case of an owner occupier, there would be a notional rent paid by the NHS. In the case of an actual rent paid by the practitioner, this would be reimbursed to him by the NHS. The figures which the Directions provided were artificial and not suitable for assessing a true rent. The rent in question could have been influenced by the extent of potential reimbursement under the Directions.
 In the assessor comparisons comprising GPs’ and dental surgeries which had been valued without appeal on the basis of office rents, Mr Davidson considered these should not be given weight. The rates were also reimbursed under the Directions, and accordingly the ratepayers would have no incentive to appeal the valuations. This was especially in cases of value under £10,000 where there is rating exemption. Only in one case in Ayrshire AP 16/105 involving the occupation of a rather prestigious building was he aware of the health board seeking to check the rent.
 The passing rent for the surgery appeal subjects number 9 was not a clean rent. The NHS had been involved in the Tannahill development. The passing rent for the adjacent clinic, part of the same building, was £1 per annum. The rent for the surgery was a rent rate of £153 as opposed to the office rate of £70 which suggested some form of repayment of capital. The lease indicated an involvement of the district valuer acting for the tenant who would be involved in cases under the 2004 Directions.
 Turning to case 17 it appeared from the relevant report of the landlord that the district valuer was acting for the Scottish Ministers who were tenants in the review/ extension negotiations. In Mr Davidson’s opinion, the rents referred to were not clean evidence of open market value. He did not know specifically why the district valuer was involved in the new lease.
 The assessor’s evidence of actual rents for health subjects could be criticised. The table at RA 16 disclosed that case 3 involved the lease of a shop whose rent was thus irrelevant. Comparisons 43, 58, 271 and 341 were dental surgeries and thus presumably involved the 2004 Directions. Comparisons 5, 104, 118, 142 and 209 were offices, albeit used for health purposes. There were no true comparisons.
 The interior of the Sandyford Clinic with numerous consulting rooms and corridors did not resemble the interior of large offices in Glasgow as shown in AP 19. He did not accept that rental evidence for offices in the west end of Glasgow could be used. The Sandyford had an internal area of 1,781 sqm whereas the largest offices and non-offices in the area were much smaller.
 She first considered the area of East Renfrewshire, comprising appeal cases 2, 3 and 4. Case 2 contained a substantial amount of office accommodation, case 3 indicated typical office type accommodation and case 4 contained offices. Within RA 2A, RA 17 and RA 23 she referred in her precognition to numerous comparisons comprising various surgeries and offices. She referred to RA 4 and 5 being maps showing the location of the comparisons in relation to the appeal subjects.
 For the Inverclyde area consisting of cases 5 and 6 she considered that the vast majority of the building of case 5 was used as office accommodation, and accepted that case 6 was the largest of the subjects. She referred to numerous comparisons within RA 2A comprising various healthcare subjects and offices. She referred to RA 6 and 7 which are comparison maps of the area.
 For the area of Paisley there are cases 9, 10 and 11. She referred to numerous comparisons within RA 2A. These were mostly offices and surgeries with the exception of 291 which is a health centre, 292 which is a large health centre and 293 which is a hospice. 293 has an actual rent and all were valued on the comparative basis for offices. The comparisons are shown in RA 9 which is a map of the Paisley area.
 Ms Dell referred to cases 12, 14, 15, 16 and 17 in the context of Renfrew, Johnstone and Erskine. Case 12 was “softly” fitted out since the treatment was for children and individuals with addiction problems. In her opinion case 14 did not change in terms of its fit out between the clinical use and subsequent office use once the psychiatry and psychology team started to use the subjects as back office accommodation. In her opinion case 15 comprised typical office accommodation. Case 17 was significant in that it was a health centre rented by Renfrewshire Council to the appellants. The lease was good market evidence. There was no reason to think the District Valuer’s presence in negotiations on behalf of the tenant would favour the landlord. She referred to numerous comparisons within RA 2A, RA 17 and RA 23. All but 251, 256 and 271 are either offices, GP or dental surgeries. 251 is a purpose built health centre but is occupied by GPs. 256 was a surgery but is now used as back office accommodation for the GPs. 257 is a similar example. The comparisons are marked on maps RA 10, 11 and 12.
 Ms Dell referred to the use class order RA 14 which is the Town and Country Planning (Use Classes) (Scotland) Order 1997. The appeal subjects are found within Use Class 2 which is defined as financial professional and other services “where the services are provided principally to visiting members of the public”. She pointed out that if the NHS moved out of any of the appeal subjects there would be nothing to stop a lawyer or surveyor moving into that subject.
 Ms Dell considered RA16 showing a table of rented subjects. Case 3 could be disregarded because it comprised a shop. For the others it could be seen that the actual rent rate is more than the applied rate. The table includes offices where there was healthcare use of one sort or another, surgeries and two clinics. The applied rate is the appropriate office comparison rate for each area in question at the tone date. There was consistency between healthcare and office rents. RA 17 and 23 were also referred to more generally.
 There was explanation why the eight health related subjects in Renfrewshire were valued using the contractors method. Using the assessors equivalent numbering to A15/1-8, RA 17/ 206 and 207 are in the middle of local authority housing estates. 204 is large (1,147 sqm GIA) in a location in Linwood which would not generally support this amount of office space. Similar reasons were given for comparisons 203, 205, 123 and 124. 207 is an ornate grade-A listed building, is dissimilar to offices, and is ongoing redevelopment.
 Ms Dell drew attention to the SAA practice notes. As there was rental evidence PN 21 could be followed and the comparative basis used. The subjects were “similar” to offices and thus the same principle could be used under PN 22. She also confirmed that where there had been differences between appeal subjects and comparative evidence, allowances have been granted mainly in the form of quantum allowances to reflect differences in size. In her opinion the comparative method is the appropriate means to value the subjects.
 She had not seen inside RA assessor comparisons 104 and 115. She accepted for case 9, the Tannahill Centre surgery, that the rent was out of kilter with the office rate.
 Ms Weir referred to non-office subjects and office rents and agreed values in GCA 2 in the same locality of the Sandyford Clinic. The non-office subjects comprising clinic, surgery, day nursery, and club rents and values stood comparison with those of commercial offices. GCA 3 comprising clinics and health centre subjects in Glasgow were valued on the comparative basis. This supported the assessor’s approach. GCA 4 and 5 referred to certain appeal outcomes in which values were agreed or appeals withdrawn. In particular these included 8 Woodside Crescent and 2 Woodside Terrace, both clinics close to the Sandyford Clinic. The internal areas are 287 sqm and 527 sqm respectively. These were valued on the comparative principle. 8 Woodside Crescent in particular has been sold and converted to form commercial offices. The agents withdrew appeals for both these subjects. There was a cosmetic surgery at Elderslie Street nearby which had a passing rent comparable to office levels. There are two other healthcare subjects occupied by the appellants at 118 Napiershall Street and 1882 Dumbarton Road which it was accepted were correctly valued on the comparative principles. GCA 5 showed subjects which had been valued on the contractor’s principle. These, 50 in number, clinics and health services were located within suburban areas to provide healthcare services to local communities.
 She also referred to 77 Nelson Mandela Place occupied by the appellants and used as a breast screening clinic. She referred to 8 Nelson Mandela Place which is occupied by the Scottish National Blood Transfusion Service. In each case the actual use of the subjects is of a clinical nature since the public go there, and the subjects are large. Both are described in the Roll as “Office” and have been valued on the comparative principle. Appeals have been withdrawn or agreed for these subjects. The blood transfusion subjects at 8 Nelson Mandela Place have large open plan areas and ancillary offices. She also referred to an ”office” at 1 Sandyford Place which in fact is a cosmetic surgery clinic, 80 sqm, valued on the comparative principle.
 She believed that not all the internal walls at the Sandyford Clinic were solid. She had looked at architects plans and the second floor appeared to have demountable partitions. Thus it could be converted to open plan space. When the building had been converted to the Eye Hospital structural walls from the old townhouse had been removed so in her opinion the property was more adaptable to modern use than other townhouses. At 1,741 sqm internal area it was not a large subject by Glasgow City Centre office standards. A quantum allowance would be appropriate and had been agreed at 4%. Her explanation of the internal walls appeared correct from what was observed at the site visit.
 In her opinion the comparative principle of valuation by reference to local rental evidence had been established and should be used for the Sandyford Clinic.
 The appellants’ position was that valuation should be set out in SAA Practice Note 21. In the absence of clinic, surgery or health centre rental evidence, the subjects should be valued on the contractor’s principle. From the history and the pleadings, the case was about determining if the subjects are properly valued by the use of office rents. The subjects were not occupied and used as offices and their use is distinct from those of offices. The assessors admit the subjects are not being used as commercial offices, but contend that the appeal subjects are “quite similar in age, character, construction and sized office accommodation in their respective localities”. The case the appellants had to meet in the pleadings was simple, namely whether the subjects are sufficiently similar to offices so that office rents could be used.
 Under reference to Armour 18-05 onwards the actual state (or rebus sic stantibus) rule should be applied. Reference was made to Armour 18-09, 18-10, 18-11 and 18-12, Stirlingshire Assessor v Myles & Binnie, Edinburgh Merchant Co v Lothian Assessor, Spudulike Group Ltd v Tayside VJB Assessor, Woodrow v Lothian Region Assessor and City of Glasgow Assessor v Monti Marino (Glasgow) Ltd. These cases support the view that the rule is qualified only to the extent that categories of subjects to be compared should not be minutely subdivided; eg that shops should not be subdivided to bakers or grocers shops which were equally comparable. If appropriate comparison values are not available it was necessary to apply the contractor’s principle.
 It is accepted that all NHS hospitals are to be valued on the contractor’s principle. The assessor had conceded that case 7 namely 20 Duncan Street was tantamount to a hospital, lacking only overnight hospital use. The method is routinely followed for libraries and community centres which in certain cases were adjacent to appeal subjects.
 The fact the appellants had agreed figures for a comparative valuation did not mean there was agreement as to the method of valuation. It was noteworthy that in case 1 the assessor’s comparative valuation was lower than the appellant’s contractor’s principle valuation.
 The evidence of Mr Robertson and Ms Mitchell demonstrated that for the appeal subjects in their various locations and size there was no market for them as offices. There was no likelihood conversion costs of adaptation would be incurred to make them more marketable . The Sandyford Clinic would require an extensive strip out of most internal walls to produce the layout of a modern office such as that demonstrated in two city centre offices. It was not feasible to adapt it for modern use.
 Some of the assessor’s comparisons had been agreed by agents on office values. This is with particular reference to assessor’s comparisons 104 and 115. However this was because these subjects were used as offices. They were not subjects “constructed or adapted wholly or mainly … for the reception of treatment of persons suffering from any illness, injury or infirmity” in terms of the 2005 regulations. In such examples the clinical use was no more than 10%, in contrast with all the subjects of appeal where clinical use was greater than 50% and normally substantially greater.
 In examples of surgeries whose use had changed to office (since the factual “as at” date of 1 January 2010) namely assessor comparisons 256 and 257, these were owner occupied and no longer used for their intended purpose. If not used the occupier would be liable for the unoccupied rate.
 There were extremely limited rents for the subjects of appeal. They were not open market arm’s length transactions and extremely limited in number. Many were listed as offices in any event. Others were dentist’s surgeries and therefore required to be disregarded on account of the assessor’s undertaking. Comparison 293 did not appear to be a typical clinic or health centre.
 The assessors accepted lands must be valued according to their actual physical state and use without regard to the potential for physical adaptation provided the use is beneficial and not subject to arbitrary restrictions. There was no question of arbitrary restriction in the present appeal cases. Reference was made to Glasgow City Assessor v Monti Marino per Lord President at  and Spudulike Group Ltd v Tayside Valuation Joint Board Assessor.
 The subjects should be valued on the comparative basis. The contractor’s principle is unquestionably a method of last resort “ where no proper comparison values are available”: Post Office v Assessor for Fife Region per Lord Avonside at p223. Its shortcomings are well known: Armour 19-39.
 Proper comparables should relate to subjects comparable in use, size, character and situation. It may be very difficult to find exact comparisons and absent open market rentals around tone date the comparative approach can also take the form of reference to agreed valuations either in the absence of or in addition to rental evidence. The use of comparative valuations was supported in Armour 19-19 as opposed to the contractor’s principle, Rolls Royce Plc v Assessor Renfrewshire VJB Lord Hodge para , Callendar Abbotts and Dobbie Forbes v Stirlingshire Assessor. If there are comparable subjects standing at a value in the roll then if the comparative principle is to be thrown overboard it must be for specific reasons on facts found: Leith Docks Commissioners v Assessor for Edinburgh 1968 Lord Sornp46; and Armour 19-40.
 It was recognised that comparisons could be relied upon even if not directly comparable. Reference was made to Fife Assessor v National Trust for Scotland, North British & Mercantile Insurance Company v Assessor for Edinburgh and Armour 20‑35. In Spudulike it was observed that particularly where there was difficulty in categorisation and a shortage of directly comparable evidence, a valuer applying valuation skill and judgement may look to some extent of categories outside the chosen category for guidance.
 Counsel divided the subjects into two categories as far as the Renfrewshire was concerned. Firstly there were doctor’s surgeries which also provided a variety of clinical or other services in addition to GP core services. These were described in the roll as health centres and a clinic and comprised cases 4, 6, 10, 16 and 17. Case 9 was a surgery. The other category comprised subjects whose use was as clinics or for the provision of health care or health related subjects other than core GP services. These were mainly described as “clinic” but in one case was an office and another a health centre. These are cases 2, 3, 5, 11, 12 and 14. Case 15 is described as a day centre.
 In relation to both groups the subjects have been valued on the comparative method on the basis of office rents reflected in the office rate applied to the respective areas.
 For the subjects in which there were core GP services, there were assessor comparisons 8 (Clarkston), 9 (Newton Mearns), 69 (Port Glasgow), 85-87 (Greenock), 251 and 256 (Johnstone), 284 and 285 (Erskine), 137, 291, 292, 294-296, 305 and 307 (Paisley). In respect of clinics and other health related services there were assessor comparisons 5 (Newton Mearns), 48 (Clarkston), 51 (Barrhead), 104, 115, 118 (Greenock),209 (Renfrew) and 142 and 293 (Paisley). These although mainly entered in the roll as “office” were still used for the provision of health related services. Some of these had actual rents and in any event were settled valuations.
 The assessor’s approach did not involve a failure to value according to existing use. The subjects were not being valued in respect of use as offices but in respect of use as surgery, health centre or clinic on the basis of established office space valuations to comparisons used for the same or similar purposes.
 The assessor had produced evidence of actual rents in RA 16. The evidence was modest but related to health related services in an office setting. The rental evidence so far as it goes was supportive of the rates applied to the appeal subjects. There was no issue in relation to the office rates in fact used as being referable to offices. Office rates could be used for healthcare use in an office setting in terms of SAA Practice Note 21.
 Where the assessor had used the contractor’s principle this was where by reason of location, size and characteristics there were no suitable comparisons. This was a matter of degree. The assessor had been entitled to start again and adopt a different approach in a revaluation.
 There was no purpose to be served in contrasting values produced by the comparative basis as opposed to the contractor’s basis. Nor was it necessary to carry out the exercise as to the cost of converting subjects to mainstream office accommodation. If the subjects were already properly valued on the comparative basis by reason of the extent of comparative valuations fixed by the application of office rates reflecting office rentals, it was unnecessary to consider any perceived lack of demand or the viability of conversion to open plan offices.
 In any event the appeal subjects all contained office accommodation, many had the appearance of an office to some extent and there was no obvious barrier to office use.
 Turning to the Glasgow appeal it was pointed out there was a significant number of rents in respect of subjects in the vicinity for various purposes. There were offices and non-office commercial uses in the locality, including clinics, day surgery and day nursery. There was a good comparison between office rent and non-office rent, supported by comparative valuations. This could be seen from GCA 2. Moreover in terms of GCA 3 there were 20 clinics in Glasgow valued on the comparative principle. There were also subjects in the roll described as “office” as used for clinical purposes. These included the blood donor centre occupied by the Scottish Blood Transfusion Centre at 77 Nelson Mandela Place and a breast screening clinic occupied by the appellant at 8 Nelson Mandela Place. These were valued on the comparative basis with appeals agreed or withdrawn by agents. Although the subjects are larger than other office and commercial subjects in the West end of Glasgow, it is not larger by city centre standards and a quantum allowance had been allowed.
 It is first necessary to consider the borderline between “health centres, clinics and surgeries” on one hand and “offices” on the other. At one end of the spectrum offices are used for healthcare administration by the appellants and are, in fact, valued as such on the roll. At the other end there are large health centres which, if they provided beds for overnight stay, would be classified as hospitals. In that event there would be no dispute the contractor’s principle applied. Plainly the distinction would not be significant if the assessors are correct that “health centres” etc fall to be valued as offices in any event. On that basis the fact a “health centre” may have offices attached to it to a greater or lesser extent would be immaterial for valuation purposes since the whole subjects would be valued as offices on the comparative basis. But as we discuss the distinction does appear to be relevant in determining whether certain comparisons can properly be taken into account and in assessing the weight to be given to them.
 We consider possible statutory definitions in play. The 2005 regulations are not strictly relevant in that they only apply once it has been determined that the contractor’s basis of valuation is to apply. But they do contain a useful definition of “healthcare property” in regulation 2(3) being:-
“‘Healthcare property’ means any lands and heritages constructed or adapted wholly or mainly … (a) for the reception or treatment of persons suffering from any illness, injury or infirmity … and used for such a purpose”.
There was no dispute that the appeal subjects fell within class 2 of the Town and Country Planning (Use Classes) (Scotland) Order 1997. The schedule defines buildings within this class as:-
“Use for the provision of … (b) professional services or (c) any other services … which it is appropriate to provide in a shopping area and where the services are provided principally to visiting members of the public.”
It is not suggested the latter definition is definitive.
 It seems to us that both the above definitions have a point of distinction where the main use of the subjects is to provide a service to visiting patients. Offices and other areas within the subjects may be ancillary to that use. But where offices within a “health centre” etc are used for healthcare purposes not associated with the visiting patients, to such an extent that the majority of the subjects cannot be said to be devoted to “clinical” use, then we accept the subjects fall to be regarded as offices. In this connection Mr Davidson accepted that all the appeal subjects were in clinical use to an extent greater than 50%, albeit to different degrees. He had visited all the appeal subjects and we agree with his assessment.
 We agree with the assessor’s submission that the contractor’s method is one of last resort. This is not disputed. It is only resorted to where no proper comparison values are available: Post Office v Assessor for Fife, Lord Avonside at p223; Armour 19 -39. The distinction between the comparative and contractor’s approach is particularly acute in the instant appeals, since in all but the first case (now withdrawn) the contractor’s approach produces a significantly lower valuation.
 The issue is whether office rents and valuations are appropriate comparisons for health centres, clinics and surgeries. This subsumes the question whether we are bound by the “actual state” rule which, it was argued, did not allow us to look at separate categories from the category of the appeal subjects. If the appellants are correct that the exception to the actual state rule goes no further than for example, valuing one shop by comparison to other shops although the businesses carried out in the various shops are different, then we would be unable look at office rents. However, it seems to us the authorities on this point are in context of the general application of the comparative approach, not the preliminary question of whether it is necessary to resort to the contractor’s approach at all.
 We prefer the assessors’ legal analysis on this point. In particular we are persuaded by the following passage in Armour at 20-35, making reference to Assessor for Stirlingshire v Miles and North British & Mercantile Insurance Co v Assessor for Edinburgh:-
“In practice, offices which have the physical characteristics of shops, and are located among shops, tend to command the same level of rent as neighbouring shops. Where there are no comparable office premises it may be legitimate for the assessor to use other premises as his basis for comparison.”
 This approach is fortified in Spudulike Group Ltd v Assessor for Tayside Valuation Joint Board at para . The Tribunal said this:-
“We think, however, that the dicta, albeit obiter, of their Lordships, and particularly Lord Clyde in Textile World, supra, are also pertinent. It was emphasised that over rigidity of categories is to be avoided. The purpose is always to ascertain the fair value of the subjects. It seems to us to follow, particularly in a case where there is difficulty in categorising and a shortage of directly comparable evidence, that the valuer applying valuation, skill and judgement may, as well as applying that judgement to categorisation, look to some extent at values of subjects outside the chosen category for guidance. Again the Lands Tribunal in Williams, supra, recognised this ( RA at 192, para 198). We are also reminded in Textile World that the rents of the actual subjects may be very important: where a bona fide rent which is at a different level from other rents is encountered, the valuer has to exercise skill and judgement to decide, on the basis of all the significant evidence, whether the difference is significant or merely the sort of variation sometimes referred to as the “higgling of the market”.
 So the extent to which comparison evidence may be looked to from other categories comes down to the application of skill and judgement by the valuer. In the present case, we think this appears to be a somewhat fact sensitive exercise. As we discuss in the case of Renfrewshire, we consider that the correct view lies somewhere between the extreme positions adopted by parties.
 There are actual rents in the cases of appeal subjects 3, 9 and 17 summarised in RA 16. Number 3 relates to a shop rent for part of the subjects and is agreed is not relevant. Case number 9 is a surgery occupied by two GPs and the evidence suggested this was not a “clean” rent. In the light of the assessor’s undertaking, we do not take it into account.
 Case 17 is significant. The actual rent analyses to £90 per square metre (internal area 294 sqm). The agreed office rate is £75 per sqm appropriate to the location. The only possible inference this is not a “clean” rent comes from the fact the district valuer acted for Scottish Ministers when there was a provisional agreement to extend the lease as from 28 May 2010. The original lease provides for rent review on the basis of the open market rent. The original rent was £24,500 in 1999. As at the review date the rent from 28 May 2004 the rent was £25,000. For the extended lease this was provisionally agreed at £31,500 from 28 May 2010. There was no explanation for the apparent “jump” occurring in 2010 which did not occur in 2004. The appellants could not say why the district valuer was involved, and absent evidence from the appellants themselves on this issue we cannot infer any reason as to his involvement other than to protect the taxpayer or why the landlord would be unduly benefitted. There were no circumstances about the transaction to suggest that the rent might have been other than on an open market basis. So although we must apply some caution it can be concluded the actual rent is not out of line with the tone applied by the Renfrewshire assessor. The clinic in question is in a predominantly residential area and is part of a district centre in Erskine.
 RA 16 lists ten further actual rents in addition to the above three. Comparisons 43, 58, 131, 271 and 341 are all dental surgeries and fall to be disregarded in the light of the assessor’s undertaking. Comparisons 5, 104, 118 and 209 are all described in the roll as “offices”. They all provide a certain amount of clinical use involving interface with members of the public at consultation. Number 104, for example, contains consultation rooms within a secure area, but also has large open plan office areas on the first floor. As senior counsel for the appellants put it, it is not surprising that “offices” have been valued as offices. However, at this stage we note the following details: number 5: modern, 430 sqm in central Newton Mearns, child health and support use ; number 104 – modern, 1,652sqm in central Greenock, used by community mental health team; number 118 - 79 sqm in central Greenock, used by Scottish Association for Mental Health; and number 209 – 426 sqm central Renfrew, used by social work service relating to substance abuse. This suggests there is a rental market for offices where the subjects have an element of clinical use, and there may be a fine line before the subjects fall into the category of healthcare subjects.
 That leaves comparison 293 comprising a “clinic” which is a small hospice comprising a support centre north of Paisley town centre of 74 sqm, with an analysed passing rent at £125/sqm in which the appropriate office rate figure is agreed at £70. Agents were involved in settling the valuation.
 The assessor’s comparisons consisting of entries on the roll based on office values consist of numbers RA 2A and RA 17/ 8, 9, 69, 85, 86, 87, 137, 251, 256, 257, 284, 285, 291, 292, 294, 295, 296, 305 and 307. All are GP surgeries except 251, 291 and 292 which are entered as health centres, but in terms of the roll all are occupied by GPs. The majority of these are less than 400sqm net, and all but two (291 and 251) are smaller than 550 sqm net. All have been valued on a comparative basis as offices.
We note the details thus:
8. Surgery, Clarkston, 447 sqm, modern, GP and additional clinical services.
9. Surgery, Newton Mearns, 363 sqm, modern, GP and additional clinical services.
69. Surgery, Boglestone, Port Glasgow, 249 sqm, modern, significant office areas, GP and additional clinical services.
85. Surgery, Greenock, 529 sqm, modern, fairly central, GP and additional clinical services.
86. Surgery, Greenock, 456 sqm, modern, fairly central, GP and additional clinical services.
87. Surgery, Greenock, 433 sqm, modern, fairly central, GP and additional clinical services.
137. Surgery, Paisley West End, 507 sqm, modern, GP and additional clinical services.
251. Health Centre, Johnstone, 2,555 sqm, modern, fairly central, GP and additional clinical services.
256. Surgery, Johnstone, 302 sqm, modern, GP and additional clinical services, practice has relocated and subjects now used as back office without significant alteration.
257. Surgery, Johnstone, 237 sqm, fairly modern, now used as offices.
284. Surgery, Erskine, 275 sqm, modern, located at shopping centre.
285. Surgery, Erskine, 314 sqm, modern, GP and additional clinical services.
291. Health Centre, Paisley Town Centre, 992 sqm, modern, GP and additional clinical services.
292. Health Centre, Paisley South, 273 sqm, modern, GP and additional services.
294. Surgery, Paisley Centre, 422 sqm, modern, GP and additional clinical services.
295. Surgery, Paisley Centre, 345 sqm, same building and use as 294.
296. Surgery, Paisley Centre, 343 sqm, same building and use as 294 and 295.
305. Surgery, Paisley, Town Centre, 405 sqm, fairly modern, GP and additional clinical services.
307. Surgery, Paisley North, 263 sqm, fairly modern, GP and additional clinical services.
 The appellants refer to valuation comparisons AP 16 - 17/ 1-8, all valued on the contractor’s basis. All are purpose built health centres owned and occupied by the appellants. We note the characteristics thus:
1. Health Centre, Paisley, Central, 1,595 sqm GEA, (say 1,060 sqm internal) circa 1927.
2. Health Centre, Greenock, 553 sqm GEA, (say 370 sqm internal) circa 1960.
3. Health Centre, Greenock, 936 sqm GEA, (say 624 sqm internal)circa 1965.
4. Health Centre, Paisley, 574 sqm GEA (say 380 sqm internal) circa 1967.
5. Health Centre, Paisley, 574 sqm GEA ,(say 380 sqm internal) circa 1967.
6. Health Centre, Barrhead, central, 6,917 GEA ,(say 4,600 sqm internal) built 2011.
7. Health Centre, Linwood, 1,147 sqm GEA, (say 765 sqm internal) built 1975, ext 1996.
8. Health Centre, Bishopton, 676 sqm GEA, (say 450 sqm internal) built 1978.
In addition there are the settled appeal cases on the contractor’s basis:
Case 7. Health Centre, Greenock, 4,587 sqm GEA (agreed 3,079 sqm internal area), purpose built 1979 – 1986.
Case 8. Health Centre, Gourock, 810 sqm GEA (agreed 565 sqm internal area), built 1979.
Case 13. Health Centre, Renfrew, 5,739 sqm GEA (agreed 3,415 sqm internal) built 2010.
 We have applied a rough and ready calculation of taking 66% of the gross external area (generally used by the appellants) to arrive at net internal area (generally used by the respondents.) It can be seen that the majority of the appellants’ “contractor’s” examples exceed 400 sqm net, with four being below 550 sqm internal (2,4,5 and case 8) and three below 400 sqm internal (2,4 and 5). There are several large examples, including two in central locations (1 and 6). The smaller appellant’s examples of contractor’s approach (2,4,5 and case 8) are all in outlying areas where it can be inferred there would be weak office demand. On the other hand the assessor’s comparisons are smaller, mostly below 550 sqm. Of these the larger subjects, say over about 300sqm, are usually in more central locations where it can be inferred office demand is stronger.
 There are three particular issues with the valuation comparisons. Firstly, on the basis of Mr Davidson’s evidence, the GPs would be reimbursed their rates from the NHS, and in practical terms there was little incentive for them to negotiate or take the assessments to appeal. For that reason we require to exercise significant caution with regard to the assessor’s comparisons at  above. However, as there is a significant number of them we do not think we can ignore them altogether. Secondly, we have some hesitation in considering other cases where the contractor’s basis has been used in the sense of these being “comparisons”. But, we are prepared to do so on the limited basis that these may provide examples where the assessor has conceded there is no proper comparison evidence in the true sense or no office market in the particular locality. Thirdly, it is by no means easy to identify an “established” valuation roll which would be “upset” by our favouring one valuation basis over another, since it is apparent that both contractor’s and comparison bases have been used for similar subjects. But that said, subject to certain exceptions, it is just possible to identify a pattern from the figures which tends to show the smaller subjects to have been fixed on the comparative basis, and the larger ones on the contractor’s basis, and that in borderline cases the location is significant in terms of likely office demand.
 The more notable office comparisons comprise assessor’s comparisons 48 and 51. Number 48 is an office in Clarkston of 171 sqm providing counselling by social workers. Number 51 provides a support service to young people in Barrhead from an office of 283 sqm in size. These offices are not dissimilar in layout to surgeries.
 We have mentioned at  and  assessor’s comparisons 256 and 257. These are two former GP surgeries which have converted to office use at 256 sqm and 302 sqm respectively. Both are in Johnstone but not particularly central. Although these are owner occupied offices, used in conjunction with surgeries elsewhere, it does show it is possible to convert from surgery use economically at least in cases of this size.
 We also note that appeal cases 14 and 15 (195 sqm and 208 sqm) are in the same building at 15 Quarry Street, Johnston, on the periphery of the town centre. Case 14 is the top floor and is no longer a clinic but is now used as offices with little or no effort having being required to convert office. Since the design of the ground floor (case number 15) is very similar, we would think it equally possible to convert it to an office. This supports the view it is relatively straightforward to convert clinics previously used for mental health treatment to offices.
 In order to rely upon office comparisons, it seems to us necessary for the skilful valuer to identify some basis on which to do so. We conceive it would be enough to find a meaningful pattern correlating office rents and/or valuations and health centre etc. rents and/ or valuations. A pattern might simply be identified which sufficiently establishes there is a correlation, so it can be shown the market behaves in a particular way, even although there is no clear cut explanation. In this case there is appeal case 17 and to some extent assessor’s comparison 293, from which it is possible to infer that clinic rents are at or more than office rent levels, at least in subjects up to 300 sqm. There is also, albeit with considerable caution, and subject to exceptions, an identifiable pattern having emerged on the roll as discussed at  and . That is, where subjects are above about 300 sqm but less than 550 sqm they tend to have been valued on the comparative basis, where they are reasonably centrally located. On the other hand subjects within this size range have been valued on the contractor’s basis where they are not fairly centrally located.
 Beyond that pattern evidence, as far as it goes, we think it necessary for the exercise to address why an office rent would otherwise be comparable to a health centre etc. rent before relying on office rents as proper comparisons. The explanation would have to be that in the case of the health centre etc., it could be readily or commercially converted to office use so as to be available to the office market. As we have indicated Mr Davidson considered there was little such market for office properties below 200sqm gross. We agree in principle that the size of the property is a major factor in relation to the office market. However, it appears to us that the appellants’ study of conversion costs and viability has been too pessimistic in the light of the evidence. In particular cases 14 and 15 and assessor comparisons 256 and 257 tend to show that conversion from clinic to office is viable in cases at least up to 300 sqm net. The latter two cases are not located at particularly central locations.
 In the foregoing circumstances we consider that office comparison values are generally appropriate in subjects up to 300 sqm internal area. If subjects are reasonably centrally located, or perhaps if there are other circumstances such as ease of conversion, we think the evidence justifies use of the comparison basis in cases of size up to 550 sqm. On this basis cases 3,9,14, 15, and 17 (all below 300sqm) and case 11 (494 sqm but in central location) would fall to be determined accordingly.
 On the other hand, the larger subjects would be to all intents and purposes unmarketable as offices on the open market. We accept Mr Robertson’s evidence to this effect and the conversion costs would clearly be an issue. Only if located in an area of significant office demand or if readily convertible would it be possible to find a tenant willing to pay a commercial rent. We do not think there is proper comparison evidence in respect of the larger sizes and at the locations we have indicated. So we would resort to the contractor’s approach in cases 2, 4,6, 12 and 16 which are all above 550 sqm, and Cases 5 (371 sqm) and 10 (424 sqm) which are between 300 and 550 sqm but are not centrally located.
 The Sandyford Clinic is characterised as a large (1,781 sqm) purpose converted clinic from townhouses in a city centre location. It has a large number of rooms – as many as approximately 40 consulting rooms alone – spread out over various corridors on three floors and two mezzanine floors. The first and second floors are mainly partition walls on account of load bearing walls having been removed in the past. The location is in a part of Glasgow where there are many offices and, as Ms Weir put it, the building is not large by Glasgow standards.
 Although we do not think the building is particularly attractive for office use in its present condition, we do not think it is so unattractive as to be unable to command a rent for that purpose. It follows that comparisons with office rents are not inappropriate. The appellant’s feasibility study concentrated on Renfrewshire conversions at a “midlevel standard” of office quality, rather than a “low level” standard in central Glasgow. It is clear from the rental evidence there is significant demand for office space in this part of Glasgow.
 We think this conclusion can be supported by three significant factors so as not to require us to resort to the contractor’s principle. The first is that a nearby clinic has recently been converted into an office. This is Glasgow Assessor comparison 17 in GCA 2, a former NHS clinic nearby at 8 Woodside Crescent. Although this building’s internal floor space is only 237 sqm, the type of building being a former townhouse is similar.
 Secondly, there is evidence that clinics and office rents are similar. One good example the assessor was able to produce in GA2 was comparison 4 being a cosmetic surgery clinic at 171 Elderslie Street in the same area. The analysed rent is £160 as opposed to the office valuation rate of £140, for a building of 157sqm.
 Thirdly, there are two large “offices” in the city centre in the roll where there is clinical use, and the comparative principle has been used. These are 8 Nelson Mandela Place occupied by the Scottish National Blood Transfusion Service, having an internal area of 760sqm. There is also 77 Nelson Mandela Place occupied by the Breast Screening Centre comprising 1407 sqm. The former has two large open plan areas in it, but nevertheless supports the view that large traditional buildings with clinical use can be similar in construction to offices. There is a particular similarity in that Sandyford Clinic has a number of light partition walls which could be converted to open plan in the upper floors thus making it more attractive for office use. In these circumstances we are persuaded that it is unnecessary to resort to the contractor’s principle in contrast to cases located in more outlying areas. We agree that the comparative basis should be used, no doubt subject to due allowance for the condition and size of the property.
 We anticipate that as a result of this decision there will be larger subjects valued on the contractor’s basis which will have a smaller NAV than smaller subjects valued on the comparative basis. We think this is inevitable given the considerable disparity produced by the respective methods. This apparent irregularity however cuts both ways; as Ms Weir explained when the Sandyford Clinic was valued previously on the contractor’s approach there were smaller healthcare subjects being valued higher on the comparative approach. Ultimately the use of the contractor’s principle occurs where there is no market evidence for the letting of the subjects in question; that in turn implies there is limited market demand for the subjects and theoretically therefore low rents.
 We therefore allow the appeals to the following extent. Cases 2,4,5,6,10,12 and 16 should be valued under the contractor’s principle on the figures shown on JP 19. Cases 3,9,11,14,15,17 and 18 should remain valued on the comparative basis, but with the agreed reduced values as shown on JP 19.