The appellants are four national cellular (mobile) telecoms operators (“H3G”, “Vodafone”, “Orange” and “T-Mobile”). These appeals referred to the Tribunal under Section 1(3A) of the Lands Tribunal Act 1949 relate to entries across Scotland of the appellants’ interests at radio transmission mast (or similar, for example rooftop) sites at the 2005 Revaluation. The respondent is the Assessor who undertook lead responsibility for such assessments in Scotland.
 A feature of the UK system of mobile telephone networks is that in many cases telecoms operators operate at the sites as “sharers” under sharing agreements with “hosts” who are tenants under leases of the sites from the landowners. The hosts who hold such leases are either other similar operators or infrastructure providers. They may have no sharer on a particular mast site, or there may be more than one sharer. The landowner generally receives payment, in addition to the basic agreed rent payable by the host, of a percentage (the “payaway”) of each sharer’s payment received by his tenant as host.
 The appeals relate to both host sites and sharer sites where telecoms operators are occupiers of rateable subjects.
 The appellants principally appeal against two aspects of the Assessor’s valuations. Firstly, while there is no dispute that the sharers are rateable occupiers of cabins or cabinets to house their equipment on the ground and also of the cables attached to the masts, so that the sharers are correctly entered on the roll as well as the hosts, the Assessor contends that their rights, under the sharing agreements, to place their cables and aerials on the masts (“sharers’ rights”) are themselves heritable and rateable subjects if, as is usually the case, the sharers occupy accommodation on the ground for their equipment. The appellants do not accept that.
 Secondly, the Assessor argues that the payments by the sharers under the sharing agreements provide a better measure of the rateable value of both host subjects and (if it is correct that sharers’ rights are rateable) sharer subjects (“sharer payment” basis) than the method contended for by the appellants (and used elsewhere in the UK), based on analysis of site rentals with the addition of standardised decapitalised site development and rateable plant and machinery cost rates (“site rental” basis).
 Happily, parties were generally agreed on categories of sites and values under six different scenarios, summarised below and added as an appendix to this Opinion, possible in the light of the Tribunal’s decisions on these issues. This followed very considerable analysis of the many site rents, sharer payments and payaway arrangements, etc. for different categories of site. There was also agreement on the amounts of additions for the plant and machinery, including the masts and the accommodation on the ground (‘cabins’, or ‘cabinets’, which we shall generally refer to together as as ‘cabin(et)s’), again under the various scenarios possible. Notwithstanding that agreement, however, the Assessor endeavoured to establish that the cost figures drawn from the Valuation Office Agency cost guide agreed for England and Wales, were actually significantly understated, a contention which he used, not to increase values under the site rental basis, but as part of his argument that the site rental basis should not be used at all (there being no need for separate valuation of the masts when valuing on the sharer payment basis).
 Although there are many hundreds of such telecoms sites, Article 2 of the Non-Domestic Rating (Telecommunications and Canals) (Scotland) Order 1995 requires such subjects within each valuation area and occupied by the same person to be the subject of only one entry. A further issue between parties (logically, the first issue) was whether, as a matter of interpretation, that Order required each companies’ sites within each area to be treated unum quid, as the Assessor contended, or merely authorised single, ‘cumulo’ entries for each company in each valuation area as a matter of administrative convenience, as the appellants contended. This was not, however, an issue of such importance as might be thought. As the matter was argued before us, the categorisation of types of site would vary according to the answer but the Assessor would still reach a value for each entry at each site and add them up. Neither the Assessor nor the appellants ultimately contended for any overall adjustment, either upwards or downwards, of the totals arrived at. We were told that looking at the position across the board in Scotland, the overall results would not differ significantly according to the decision on this issue, but parties, and the Assessor in particular, understandably wished a ruling on it. The differing results in that situation were, again, generally agreed.
 The hearing technically covered only 8 appeals, in respect of the valuations of each of the appellants’ subjects in two valuation areas, Glasgow and Perth and Kinross, but the outcome is intended to enable the values under appeal in all 128 appeals against entries in all 32 valuation areas in Scotland to be determined. Each side chose a number of so-called ‘test sites’ (not restricted to these two areas) to be used in evidence and submissions as illustrative of the situations giving rise to the issues, although parties referred additionally to some other mast sites, including some in England and Wales.
 In summary, the Tribunal has upheld the appellants’ submissions on the principal issues and decided that the values listed in “Scenario 6” for 13 categories of sites (“Cumulo based on Site Rents – sharer rights not heritable”) are to be applied, but with the addition, in respect of the payaway, of £1900 (“greenfield” sites), £2,400 (“rooftop” sites) or £6,000 (“Base Station Controller” (“BSC”) sites) for each sharer, to ‘host site’ values. In the case of “rooftop sharer” sites, the Plant and Machinery values should be £200 (cabinet) and £500 (cabin). The arithmetical outcomes in each of the appeals, i.e. the resultant totals of the values of each appellant’s sites in each area, are to be agreed by the parties.
 Our decisions on the particular issues arising are summarised as follows.
 (i) The 1995 Order prescribes cumulo entries for each company in each valuation area, but does not prescribe unum quid valuation, with the result that each rateable subject at each site requires to be valued as a separate unit (although, as is not in dispute, the sites are, as a matter of practicality where there is such a multiplicity of sites, categorised with differing values for each category).
 (ii) The sharers’ rights, i.e. the rights, at sites rented by the hosts from the landowners, to affix aerials to the masts and lead cables from the aerials down the mast to the cabin(et)s housing their equipment on the ground, are not heritable or rateable.
 (iii) The appellants’ method of valuation, the site rental basis, is to be preferred to the Assessor’s proposed method, the sharer payment basis.
 (iv) The amount of the payaway in respect of each sharer should be added to the host’s assessment in respect of each site under appeal. This is a percentage agreed on the basis of analysis at 27.5%. The additions under Scenario 6, the scenario applicable on the basis of our decisions (“Cumulo based on site rents – sharer rights not heritable”) are agreed at: “Greenfield sharer” - £1900; “Rooftop sharer” - £2400; “Base Station Controller share” - £6,000). In the only other scenario to which this decision could apply, viz. Scenario 5 (“Unum Quid based on site rents – sharer rights not heritable”), which would apply if we had found in favour of unum quid valuation, the figure is agreed at £2,000 for each sharer at each category of ‘Macro’ site.
 (v) Had we decided issue (iii) in favour of the Assessor’s proposed method of valuation, we would have deducted £500 from the agreed sharer payment figures, to reflect non-rateable items included in the sharer payments.
 (vi) Issue (i) does not, as we understand it, require any consideration of the rule against ‘double counting’. We do not consider that our decisions under (ii) to (iv) above involve any double counting. If, however, we had decided issues (ii) and (iii) in favour of the Assessor, we would have further decided that the resultant valuation of the host and each sharer’s subjects on the sharer payment basis, with the values at each shared mast site increasing by multiples with each sharer, would involve impermissible double counting. In that situation, in Scenarios 1 and 2, we would, in the case of each sharer site, substitute the amount of the payaway for the value proposed by the Assessor, i.e. the host would be valued on the full sharer payment basis but each sharer’s valuation would be reduced to the relevant payaway figure plus the agreed plant and machinery value. If we had decided only issue (ii) in favour of the Assessor, so that either Scenario 3 or Scenario 4 applied (“Site Rents - Sharer rights heritable”), the issue would not have arisen as under those scenarios the Assessor himself limited the values of sharer sites to the amount of the payaway.
 (vii) Finally, on a very minor issue, we have decided to fix the Plant and Machinery values of “Rooftop sharer” sites under Scenario 6 at £200 (cabinet) and £500 (cabin), the figures (albeit lower than the appellants’) contended for by the Assessor.
 128 appeals by the 4 appellants, i.e. appeals against the valuation roll entries, in respect of the 2005 Revaluation, for all their mobile telephone mast subjects in Scotland, were referred to the Tribunal under Section 1(3A) of the Lands Tribunal Act 1949. After various procedural hearings at which questions as to the provision of information and as to the form in which the appeals would be heard, it was decided that the Tribunal would hear 8 appeals together, viz. each appellants’ appeals against their entries in Glasgow and in Perth and Kinross, and, additionally that a total of 24 so-called ‘test case sites’ (not restricted to these two areas) would be identified, to assist consideration of the issues arising. Evidence was heard at an oral hearing from 10 to 21 December 2012, with a further hearing for submissions on 15 and 16 January 2013. The appellants were jointly represented by Mr Christopher Haddow QC, instructed by GVA Grimley Ltd (H3G), James Barr (Vodafone and T-Mobile) and Storeys Edward Symmons (Orange). The respective Assessors were represented by Mr Ralph Smith QC and Mr Graham Dunlop, Advocate, instructed by the Assessor for Renfrewshire Valuation Joint Board, who, although not formally nominated as the responsible Assessor in relation to the 2005 Revaluation, helpfully undertook representation of the various assessors involved. Evidence on behalf of the appellants was given by Colin Sharp MBA, FRICS (general, and on behalf of H3G), Timothy John Duff MRICS (Orange), Craig Donaghy BSc, MRICS (Vodafone) and Steven Dalton BSc, FRICS (T-Mobile). Evidence was given on behalf of the Assessor by Ian Thornton-Kemsley (“Mr T-K”) MRICS, ACIArb, FAAV and Alasdair MacTaggart RD, BSc, FRICS, IRRV, the Renfrewshire Assessor. Both sides lodged substantial documentary productions as well as expert reports to which the witnesses spoke. The appellants’ productions are referred to as H3G (Mr Sharp) Vod (Mr Donaghy), Orange (Mr Duff) and T-Mobile (Mr Dalton). The Assessor’s productions are referred to as T-K (Appendices to Mr T-K’s report) and Assr. Written submissions were also exchanged in advance of the hearing on submissions.
 The parties had helpfully reached a large measure of detailed agreement, for the purposes of the appeals, on various analyses and values, and had agreed a list of tabulated values under 6 different scenarios applicable according to the Tribunal’s decisions on the main issues. The 6 scenarios with tabulated categorisations and values (H3G21, revised 26.11.2012 to take account of final agreements) are attached as an Appendix hereto and here summarised. Generally, the categorisation of the individual sites and the figures contended for by both sides are set out side by side, showing the figures agreed, in relation to disposal under the particular scenario held applicable, of each category of site across all 132 appeals. Scenarios 1 and 2 are the Assessor’s primary and fall back approaches, with Scenarios 4 and 6 the appellants’ principal and fallback approaches. Parties’ further positions on issues (iv), (v) and (vii) above are also noted.
 Scenario 1 – “Unum Quid based on Sharer Payments”. Valuation is on the unum quid basis. Sharers’ rights are included in sharers’ valuations as heritable and rateable. All occupiers on a mast site, whether hosts or sharers, are valued on the sharer payment basis. A broad average is applied, with only 5 separate categories of site (‘Cable Only’, ‘Pico’, All Streetworks’, ‘All Macro sites with cabinet’ and ‘All Macro sites with cabin’) because, as it is to be based on a unum quid, the analysis is carried out over the whole of a local government area. The resultant basic figures are agreed, except that the appellants claim (issue (v) above) that an adjustment for non-rateable items included in sharer payments is required except in the first two categories. The adjustment claimed, a reduction of £1900, is not an agreed figure.
 Scenario 2 – “Cumulo based on Sharer Payments”. This scenario applies if it is found that a unum quid approach is wrong but sharers’ rights are included and valuation is on the sharer payment basis. Assessors contend in this event that a more detailed analysis is necessary because the approach has to be one of aggregation and it is not appropriate to look at an overall local government area, so different types of site must be considered. The number of categories rises to 8, with ‘Macro’ sites categorised as ‘Greenfield’, ‘Rooftop’ or ‘Base Station Controller’, producing a wider range of values. The same issue about adjustment for non-rateable items remains as in Scenario 1.
 Scenario 3 – “Unum Quid based on site rents – sharer rights are heritable”. Valuation on the unum quid basis, with sharers’ rights heritable, but valuation on the site rental basis. Sharers are valued by reference to the additional site rent flowing through to the landlord from the payaway. In both cases, values for cables and cabin(et)s are added. There are 7 categories and all individual elements of value are agreed between the parties.
 Scenario 4 “Cumulo based on site rents - sharer rights are heritable”. The number of categories expands to 13 to reflect a non unum quid approach but otherwise on the same basis as 3. Again all values agreed.
 Scenario 5 – “Unum Quid based on site rents - sharer rights not heritable”. Valuation on the unum quid basis. No entries appear for sharers (apart from the small sums for the cabin(et)s and the cables). The hosts are valued on the site rental basis. All figures are agreed, except that the Assessor would add £2,000 (the agreed payaway figure) to the host’s site value for each sharer using the mast (issue (iv) above).
 Scenario 6 - Cumulo based on site rents - sharer rights not heritable. Again, a greater number of different categories but no entries for sharers except for cabin(et)s and cables. Hosts valued on the site rental basis.
All figures are agreed subject again to the Assessor’s contention for addition of payaway figures (in this scenario, £1900 greenfield, £2400 rooftop and £6,000 BSC, all again agreed figures), and, in relation to rooftop sharers’ plant and machinery, a minor divergence with the appellants’ figures in fact slightly higher.
Lands Valuation (Scotland) Act 1854, Section 42
Valuation and Rating (Scotland) Act 1956, Sections 6, 6A, 8A
Telecommunications Act 1984, Sched 2, Paras 1-6, 9, 21
Non-Domestic Rating (Railways, Telecommunications and Canals) Regulations 1994, SI 3123
Non-Domestic Rating (Telecommunications and Canals) (Scotland) Order 1995, SI 239, Article 2
Valuation for Rating (Plant and Machinery) (Scotland) Regulations 2000, SI 58
Non-Domestic Rating (Telecommunications Apparatus) (England) Regulations 2000, SI 2421
Dowall v Miln (1874) 1 R 1180
Brand’s Trs v Brand Jnr’s Tr (1876) 3 R (HL) 16
Parish Council of Edinburgh v Assessor for Edinburgh (1906) 8 F521
Bo’ness Town Council v Linlithgowshire Assessor 1907 SC 774
IRC v Clay  3 KB 466
Glass v IRC 1915 SC 449
Great Western and Metropolitan Railway Companies v Kensington Assessment Committee  1 AC 23
Assessor for Glasgow v Gilmartin 1920 SC 488
Assessor for Dunbartonshire v Baird & Co Ltd 1926 SC 479
David Allen & Sons Billposting Ltd v Assessor for Clydebank 1936 SC 318
John Menzies Ltd v Assessor for Glasgow 1937 SC 288
Perth Magistrates v Assessor for Perth and Kinross 1937 SC 549
LNER Co v Assessor for Public Undertakings 1938 SC 751
Assessor for Renfrewshire v Old Consort Co 1960 SC 226
Associated Portland Cement Manufacturers Ltd v Assessor for Dundee 1965 SC 260
Central Fire Area Joint Committee v Renfrewshire Assessor  RA 242
ITA v Assessor for Lanarkshire 1968 SC 249
Scottish Malt Distillers v Assessor for Banff 1973 SC 63
Assessor for Aberdeenshire v Pye Telecommunications Ltd 1973 SC 157
Scottish Gas v Assessor for Fife, LVAC, 14.4.1977, unreported
Automobile Association v Assessor for Lanark, LVAC, 30.5.1975,unreported
Civil Aviation Authority v Assessor for Orkney and Shetland 1978 SC 141
Lowland Cereals Ltd v Assessor for Lothian 1980 SLT 133
Street v Mountford  AC 809
Assessor for Central v Samuel Properties (Developments) Ltd, LVAC, 14.6.1985, unreported
Assessor for Lothian v Blue Circle Industries Ltd 1986 SLT 537
BP Petroleum Developments Ltd v Assessor for Lothian 1989 SLT (Lands Tr) 2
Assessor for Lothian v Lowland Leisure Ltd 1990 SLT 353
Civil Aviation Authority v Assessor for Strathclyde 1990 SLT 378
Brador Properties Ltd v British Telecommunications plc 1992 SC 12
Whitbread plc v Assessor for Lothian 1996 SC 374
B & Q plc & Ors v Assessor for Central, LTS, 23.11.1998, unreported
Coventry and Solihull Waste Disposal Co Ltd v Russell VO  1 WLR 2093
Ravenseft Properties Ltd v Assessor for Strathclyde 1991 SC 266
Shell UK Exploration and Production Ltd v Grampian Assessor  RA 295
Cabletel Surrey and Hampshire Ltd v Brookwood Cemetery Ltd  EWCA Civ 720
Orange PCS Ltd v Bradford VO  EWCA Civ 155
Armour on Valuation for Rating (5th Ed’n)
Rankine on Leases (3rd Ed’n)
Hunter, Landlord and Tenant (4th Ed’n)
 The facts which we found admitted or proved are summarised as follows.
 Mobile telephony is achieved by transmitting radio waves over cellular networks requiring installations of aerials, cables and related equipment at appropriately located sites across the country. Radio transmission masts have been used for a variety of purposes, for example radio and TV broadcasting, air traffic control, etc., for many years. Vodafone and O2 started to roll out cellular networks in the mid-1980s, followed by Orange and T-Mobile in the early 1990s. In March 2000, 5 third generation (3G) licences were granted, to those 4 companies and Hutchison 3G (“H3G”), with the latter launching its network in March 2003 having built it up in a very short time. Another similar network, Airwave, provided a service to the Emergency Services. Neither 02 nor Airwave maintained appeal against their entries in the 2005 Revaluation.
 Cell planning involves identifying search areas, rather wider in rural than in built up locations, and relevant factors in the choice of site include topography, accessibility, availability of service, in particular electricity, network (generally BT) link and planning. Mast-sharing is common and is encouraged. Generally the sites required were taken on lease from the owners of land, roof tops of pre-existing buildings, etc., either by one of these operators or by an infrastructure company. Such tenants have the right to place and maintain masts, cabin(et)s and other equipment on the sites, and also to admit other operators (sharers) to shared use of the masts by attaching their own aerials. The tenants are referred to in these appeals as ‘hosts’ whether or not they have admitted any sharers. Some infrastructure companies have been acquired by a company called Arqiva, and a number of sites are described for convenience as “Arqiva sites” although occupied at the material date by companies with different names.
 The physical sites and masts, etc., take a variety of forms and may be categorised for the purposes of these appeals as ‘Greenfield’, ‘Base Station Controller’, ‘Rooftop’, ‘Streetworks’, ‘Micro’, ‘Pico’ and ‘Cable Only’.
 ‘Greenfield’ are sites on which a mast, pole or other structure is erected for the purpose of accommodating operators’ cables and aerials or antennae, typically on agricultural land but sometimes also in urban settings such as in factory yards. The typical basic greenfield site, known as a ‘base station’, is a fenced compound with a mast around 15 metres high and one or more cabins or cabinets, i.e. small buildings to house operators’ ground radio transmission equipment linked to the aerials by cables.
 ‘Base Station Controllers’ (BSCs), a term used also to include ‘high transmission sites’, are a rather smaller number of higher masts (usually, but not necessarily, in a greenfield situation), which are involved in one way or another in switching (or handing over) the radio signals from one sector and/or base station to another so as to enable continuous cellular coverage. This normally involves dishes rather than antennae.
 ‘Rooftop’ sites are installations located on the roofs of buildings and other structures, e.g. water towers, grain silos, floodlighting towers, etc. and because of their height require relatively short poles which are either bolted to the fabric of the building or free-standing but weighted down, and to which the antennae are attached. Rents paid to the land or buildings owner for rooftop sites tend to be higher than for greenfield sites because the need to build a mast is avoided.
 ‘Streetworks’ sites take the form of monopoles erected beside public roads and connected to nearby equipment cabinets.
 ‘Micros’ are very small, and ‘Picos’ even smaller, antennae and equipment, with an appearance similar to burglar alarms or even smaller, attached to the walls of buildings or perhaps poles or street furniture, picos being often within buildings to assist coverage inside.
 ‘Cable Only’ sites occur where the ground equipment is located in someone else’s property and the appellants therefore do not themselves occupy any cabin or cabinet of their own.
 An alternative categorisation, used by the Assessor in his proposed unum quid valuations, is simply between ‘Macro’ and ‘Micro’, according to (very broadly) the number and capacity of the aerials and equipment of the particular company at the particular site.
 The cabins or cabinets and the cables running to the aerials, are, like the masts, poles, etc., agreed to be rateable and categorised as ‘plant and machinery’. The Assessor has, however, not included the values of aerials or antennae in the assessments.
 The valuations of all micro, pico and ‘Cable Only’ sites included in these appeals are now agreed, and a difference of view as to whether streetwork sites could be ‘macro’ also resolved for the purposes of these appeals.
 The 5 mobile operators, i.e. the appellants and O2, and also infrastructure companies and some other operators requiring radio transmission facilities, commonly take conventional leases of sites of each of the above types (except ‘Streetworks’) from landowners or building owners (or possibly tenants of buildings), usually for terms of 10 or 20 years.
 The Agreement between QS developments Limited and Hutchison 3G UK Limited of a site at Balshagray Avenue Glasgow and the Minute of Agreement and Lease between Gordon Stewart Murray etc. and Racal Vodafone (predecessor company to Vodafone) in respect of a site at Undercraig Farm near Langbank in Renfrewshire are reasonably typical examples of site leases. (H3G 3 and Vod 1.8). The sites are identified, they are let to the tenants subject to the tenants’ paying rent and complying with the terms of the leases or agreement and they are given permission to erect their telecommunication equipment. Plans are attached identifying the locations and sites and in the first example include details of the tenant’s proposed works. The tenants are in effect on ‘repairing and insuring’ terms. Rent review may be by “OMV” (Open Market Value), “RPI” (Retail Price Index or similar) or “higher of” (greater of OMV or RPI). 636 site rents, being either new lettings or reviews by OMV, during 2002 and 2003, were analysed, producing agreed host site rents as follows:-
Greenfield - £4,000
Rooftop - £7,100
Base Station Controller - £9,500
Apart from the additions for tenants’ rateable plant and machinery (masts, cables and cabin(et)s), the rents do not require adjustment to the statutory hypothesis.
 The tenants under site leases are generally free to enter into sharing agreements, although some leases may restrict the number of sharers so that a tenant wishing to take on more sharers would have to re-negotiate. The site leases generally provide for ‘Payaway’, i.e. an additional payment by the host to the landlord of a percentage (agreed on analysis at 27.5%) of site sharing fees paid by each sharer. Payaways are not included in the above figures. 27.5% of the sharer’s fee equates to 47.5% of the agreed greenfield site rent and 34% of the agreed rooftop site rent.
 Rooftops might be the subject of several separate leases of different areas to operators or of one lease to an infrastructure company who could then enter into sharing agreements with one or more operators. Arqiva have an overall lease of a large number of BT rooftop sites used in this way.
 There is no site rental of streetworks sites, because telecoms operators have statutory powers to erect structures on land falling within public roads free of charge. No lease or licence is executed. The occupiers of such sites are, however, liable to be assessed. Agreed hypothetical site rents of £2,500 per annum for streetworks sites have been arrived at by analysis of a small number of analogous sites, including sites where the land had subsequently been discovered to be private so that rents had to be negotiated.
 Site rents do not take account of the potential or “hope” of an increase because of the opportunity to share a mast. It is only when a sharer co-locates on a mast that an increase in ground rent, the payaway, arises.
 There is no effect on the level of the commencing rent whether the review is to open market, on the basis of RPI or the higher of these.
 It is agreed for the purposes of these appeals that location has no influence on the level of site rent within the same classes.
 Very approximately 25% of the site entries in these appeals relate to occupation by sharers, who are admittedly rateable occupiers of at least their cabin(et)s (unless, as in a small number of cases, their equipment on the ground is stored on racks within the host’s cabin(et)). Sharing agreements may be between mobile operators, the host who has the site lease and has erected and maintains the mast, and one or more sharers, or alternatively between some form of infrastructure company or other company which, again, has the site lease and has erected and (usually) maintains the mast, and one or more mobile operator ‘sharers’.
 Sharer agreements exist between the 5 main operators and with infrastructure providers such as Arqiva. Infrastructure providers actively seek to attract mobile operators to their facilities. In general there is no competition between the operators for individual sites.
 The contractual sharing arrangements are normally derived from three documents following a common pattern: ‘MSSA’s’, ‘rate cards’ and ‘site specific agreements'. The various documents are described differently from company to company, ranging from lease to agreement and to licence. The MSSA agrees the principle of sharing and sets out the procedure to be operated where sharing is sought. It covers issues of non-radio interference, compliance with safety requirements, etc. It lists the services to be provided by the host such as insurance, statutory inspections etc. There is commonly a draft or pro forma licence or agreement referred to in the MSSA. This becomes the site specific agreement. The host has the right to require the sharer to remove or to re-position cabin(et)s, aerials or other equipment if commercially required. Such rights are in practice only used where there is radio interference between the different operators on the mast. The financial terms are set out in the rate cards. The operators attach great commercial confidentiality to the rate cards, which are normally not available in the marketplace.
 The procedure requires a sharer to issue a notice setting out a list of the equipment proposed for the site, its intended bearing and its desired height above ground level. There is commonly included a general right to have included the site for a cabin(et) up to a maximum footprint, normally 20 sq. metres. Scottish National Enabling Agreement between Vodafone Limited and T Mobile (UK) limited (H3G 1) and National Enabling Agreement between Vodafone Limited and Orange Personal Communications Services Limited (Vod 2.2) are examples of MSSA’s between mobile operators. The agreements between Crown Castle UK Limited and Hutchison 3G Limited (H3G 2), National Transcommunications Ltd and Hutchison Telecommunications Ltd. (Orange 7) and National Enabling Agreement between National Transcommunications Limited and Vodafone limited (T-K 25) are examples of MSSA’s between infrastructure providers and mobile operators. The infrastructure providers tend to be more proactive in providing an inventory of their sites, the additional capacity available and agreeing to undertake securing any planning consent etc. Up front fees are required by the infrastructure providers to cover their costs of providing information and to the checking of suitability but these are normally offsettable against future payments.
 T-K 23 contains ratecards between Orange and O2, between O2 and T Mobile, Airwave and O2, O2 and H3G as well as O2 and Vodafone. There are rate cards by infrastructure providers which may be specific to a mobile operator or can be, in effect, a statement of charges. All of these agreements are not set out in legal terms and are on company paper only or on unheaded paper. They comprise simple lists of differing sizes and configurations of equipment such as micro wave dishes with levels of annual payment against each category. The larger the dimension of the dish and the greater the number of items in a configuration the larger the sum in the rate card. None of the rate cards are signed. The rates vary significantly between different operators. There is no common or standard level of payment. They were established at different times. They are reviewed regularly, normally annually, using RPI indexation. New sharing arrangements are based on the established rate card prevailing at the time which may have been initially set some time earlier and simply updated by RPI.
 Site specific agreements are entered into based on the procedures in the MSSA’s and at the payment levels established by the ratecard in existence. The rights granted by the site specific agreements, typically, allow the sharer to ‘instal, inspect, maintain, operate, repair and renew’ on the host mast the agreed equipment. The location of the cabin(et) (if any) and aerials is normally specified on detailed technical drawings appended to the agreement. T-K 25 contains a pro forma site sharing licence. H3G 9 is a copy of a simple site specific Tower Share Agreement. Orange 14 is an agreement in respect of the same tower but between the infrastructure provider and Orange (same landlord company but changed name). T-K 27 includes a licence between O2 and Orange for a specific site at Achterarder.
 The extent of equipment required by an operator on a host’s mast can vary significantly over short periods of time. Orange 4 is a copy of a licence between NATS and Orange in respect of a BSC on the outskirts of Dundee. In T-K 29 there are 11 Variation of Licence documents recording an alteration of Orange equipment on that tower, with consequential variations of the licence fee, in the period February 2000 to November 2003. The fee changed from £31,406 in 2000 to £42,820 in 2003.
 All telecoms operators have sharing agreements with each other based on reciprocal rights. Identical payments apply for a similar extent of equipment between the two parties to a reciprocal agreement.
 ‘Early access agreements’ allow the mobile operators to take entry to a site before completion of any formal agreement. This enables them to establish a presence, which assists in terms of the ‘Code’, as well as to commence their construction works. H3G 7 is an early access agreement between an infrastructure provider, Spyder Facilities Limited, and H3G. Infrastructure companies offer a “turnkey” service whereby they will, in effect, act as project managers in establishing a site to the operators’ requirements. Thereafter a formal agreement will be entered into.
 The services provided by a host or infrastructure provider generally comprise:-
Mast and site maintenance,
Insurance of the mast,
Annual inspection of mast and any required remedial works,
Site control records. A record requires to be kept of visitors to the site.
If no separately metered electricity supply can be arranged the supply will come from the host and will be re-charged.
All fees for rent review negotiations with the site landlord.
A sharer avoids the cost of these services which are, in effect, included in the payment made by the sharer.
 Following receipt of information on sharer payments the Assessor analysed the information in respect of each operator. On the basis of his unum quid approach, involving 5 only categories of property he determined a level of £7,450 as being appropriate for all sites other than pico and cable only. This level has been agreed by the appellants in the event of the Assessor’s contentions in regard to unum quid and sharers’ payments being the correct basis of valuation prevailing (Scenario 1) On his fall-back position which has a greater number of categories this figure reduces to £6,800 for greenfield sites and streetworks but increases to £9,000 for rooftops and to £22,000 for BSCs (Scenario 2).
 A practice of ‘Recharge’, whereby the sharer has to refund the amount of the payaway to the host, thus increasing the payment made by the sharer, takes place, to an unknown extent, and has been left out of account.
 The “Code”, as it is referred to in the industry, is set out in the Telecommunications Act 1984 (which legislated for the introduction of a cellular telephone system in the UK), as amended by the Communications Act 2003. It grants to licensed operators inter alia powers to acquire rights of occupation in order to erect a telecommunication mast, on payment of an appropriate consideration which can be a rent. It also allows operators to retain facilities at termination against otherwise legitimate rights of landlords to terminate or resist renewing existing rights. Awareness of the Code is part of the backdrop to negotiations between operators and land or building owners or lessees and allows operators to threaten to revert to the Code if negotiations are not progressing satisfactorily.
 Very substantial capital payments were made to the government for the 3G licences issued in 2000. The bursting of the “dot com bubble” in 2000 badly affected the share price of the operators and the payments made for the licences accentuated their financial constraint. This curtailed the ability of the operators to expand their coverage
 Over the years site rental levels for greenfield and rooftop sites have shown a gradual but consistent growth. Prior to the year 2000 greenfield rents were averaging £3,500 and by 2005 they had increased to an average of £4,500. A similar growth position was evident with rooftop sites albeit at a higher level. The market can be described, notwithstanding its recency, as mature and not showing evidence of significant volatility. The rents paid by H3G were not significantly higher in this period notwithstanding that they were a new entrant to the market and under pressure to accelerate their site acquisition programme. The amount of rent involved in greenfield situations as a sum of money is relatively insignificant and especially in the early days many landowners were not represented by professional advisers. By 2003 there was more recognition of the possible opportunities and a few advisers became prominent in representing landlords.
 The sharing market is entirely comprised of the existing operators and the infrastructure providers.
 The costs involved in greenfield site establishment include site preparation, laying of access roads, laying of a concrete base, fence enclosures, provision of an electrical connection from the nearest distribution point, the erection of a mast of the required height and a cabin or cabinet. All cell sites require to “communicate” with their adjacent sites and to have a “backhaul” connection to a BT or equivalent main telephone service. Rooftops need a grillage or equivalent for the support of the antennae and, if no on site internal space is available, a cabin or cabinet. They too require an electrical and backhaul connection.
 By 2003 there were companies specialising in the supply of towers on an “off the shelf” basis for a range of heights as well as contractors specialising in the building of complete base stations. Likewise companies specialised in the installation of power and equipment. Cabins and cabinets were effectively mass produced.
 Many valuations for rating include elements (mostly plant and machinery) which require to be valued on a contractor’s or decapitalised cost basis because there is no rental evidence available. These elements are extensive in range and can be of an unusual or unique character. Many are outwith a rating surveyors’ available knowledge. Consistency of valuation across the country is increasingly regarded as a prerequisite at any revaluation. To assist in achieving consistency both the Scottish Assessors Association (SAA) and the Valuation Office Agency (VOA) (the body responsible for revaluations in England and Wales) have regularly prepared papers and guidance notes for the use of their staff. By the 2000 Revaluation these guidance notes were available to members of the public and professional practitioners. A charge was made for them.
 In the case of the VOA a specialist team is responsible for the preparation of the guidance notes relating to plant and machinery. In Scotland individual assessors take responsibility for preparing different parts of their guides. Although each assessor is responsible for the preparation of the Valuation Roll for his area there is considerable liaison between individual assessors. There is also considerable liaison between the VOA and the SAA. This has grown over the years.
 The relevant part of the VOA guide for the 2005 Revaluation is headed:-
“Practice Note 1 2005: Revaluation 2005:
Communication stations- (Mast and Aerial sites) –
The document states: “The Mast Working Party (MWP) was initially created in May 1998, consisting of representatives of the VOA, the industry and their rating advisors… The MWP was reconvened for the 2005 list (English equivalent of valuation roll) … The MWP has agreed the costs to be applied for the 2005 list valuation.”
 The VOA cost guide is recorded as being stored centrally to allow it to be constantly reviewed and updated. Caseworkers are advised always to refer to this master copy rather than any earlier copy as it may have become inaccurate.
 No copy of the SAA cost guide is available before 2010. The SAA web site for the 2010 revaluation states: “The Rating Cost guide is produced jointly by the SAA and the VOA and contains guidance on cost levels solely for use in Contractors Basis Valuations in relation to the 2010 Revaluation”. It is available for purchase from the Assessor’s office in Edinburgh. In addition to the cost guide the Scottish Assessor’s Association issued a Practice Note for the valuation of telecommunication Subjects (Wireless). Although it referred to contractor’s principle valuations no cost analysis or cost information was included.
 A very limited number of costs from Scotland were included in the VOA cost guide of 2005
 Development costs vary from site to site depending on a very wide variety of factors. Among these the principal ones will be physical circumstances, such as site conditions, remoteness, closeness of main access point, the geographical location, any requirements of the town planning regime or the landlord, whether removal of an older installation is required and whether the site forms part of an operation facility such as a transport yard thereby increasing safety requirements. The size of the contract, any repetitiveness of design and the ability to utilise standard or ‘off the shelf’ equipment can significantly affect overall costs.
 Information on costs can be gained from a range of sources. The most reliable and truest source is a final measurement or equivalent document setting out the actual costs incurred ( used as the basis for payment for the work done). Quotations, informal indications, general references and off- hand remarks provide nothing more than a very general background of the possible general level of cost. Mr T-K obtained cost information while acting for landlords in the leasing of sites and from agents acting for the operators and from contractors. He has never acted for a mobile operator. Information on costs arises particularly where the mast requires to be rentalised. At Clarendon Farm, Linlithgow (an Arqiva site) the lease provided for the mast to be rentalised. Arqiva provided costs in 2008 including the then current costs of a 15m mast. These costs exceeded the standard cost levels in the VOA guide. No information is available regarding the particular circumstances at this site, in particular access arrangements, ground conditions, availability of electrical supply etc. Mr T-K also investigated costs at Torwoodlee near Galashiels (Orange) in connection with a rent review arbitration. (The arbiter decided the mast was not to be rentalised). Mr T-K referred to several other transactions where he had obtained cost information. In all cases the information related to part of a project only. In several cases it was information from a colleague in another office. Often it was simply a reference by an opposing agent that a certain cost figure had been incurred. As an example he was advised that “in negotiations with Orange their agent informed him that the cost to decommission, dismantle and construct the new mast was £102,000”.
 Further information was obtained from the annual returns to assessors in respect of sites in Fife. A 1995 cost of £78,925 for a 22.5m mast was given by O2 and the same company gave a cost of £67,350 for a 15m mast in 1996. Cellnet advised the Fife assessor in 1999 of costs of £56,000 and £62,000 respectively for two different sites. The company advised that due to “bulk purchasing” it is not possible to break the figures down to individual costs.
 What were described as “generic costs” are contained in a letter from One2One (subsequently T- Mobile). These generic costs, as provided, are very much less than any of the other costs obtained by Mr T-K. For a simple 15m mast including transportation and erection the cost is between £12,500 and £13,000 including cabin, delivery and concrete base. The date of the letter is 1998 but it does not state the date of the costs. Mast costs increase in accordance with the height. A 30 m mast will be almost double the cost of a 15 m one. Neither the appellants nor the Assessor adjust for such differences unless the mast is in a different category, e.g. a base station controller. An electricity supply can be very expensive depending on the length of connection and any special requirements for the laying of the cable, e.g. underground.
 MFD is a company in the market for the construction of telecom masts. They were asked by the Assessor to quote for a tower on a hypothetical basis. They quoted a figure of £93,450 in 2009. It was caveated on a number of assumptions. MFD are not well known or a main supplier. There is no evidence of them having developed many sites. Any general indication of cost at a non-specific location and to a non-particular specification would require to be adjusted for such matters as location, ground conditions etc.
 Assessors repeatedly asked the operators for accurate cost information. They were rebuffed by the Companies. Generally they were told the information was variously historic, i.e. the site had been developed many years previously, the papers were archived and no longer available or was simply no longer available.
 The recency of the industry’s establishment, its subsequent growth, the number of individual sites, many of them small and not clearly identifiable on site all affected how the valuations were carried out by regional assessors over the years. Although the 1995 Order was in existence it played no real part in the valuations prepared under the 2000 Rating Revaluation. Each property was valued as a separate unit based on the rents paid for Greenfield sites and a contractors basis for the masts and plant and equipment.
 In the period immediately prior to the preparation for the 2005 Revaluation Assessors became aware of the practice of site sharing and that rental or equivalent payments passed between the various companies. At any General Revaluation there is both a requirement and an opportunity for assessors to consider afresh general or particular approaches to all of the different categories of property in the relevant valuation rolls. Past practice is of limited significance if there is evidence justifying a change.
 Assessors made repeated formal approaches to the operators seeking appropriate information on site sharing arrangements and costs. The response was very limited. A more direct approach to senior executives was tried. It was similarly unsuccessful. The deadline for issuing valuation notices for the 2005 Revaluation was missed. Information was eventually provided by a single network provider.
 Valuation Notices were issued in March 2006 with an effective date of 1 April 2005 (using the provisions of the Local Government (Scotland) Act 1975). Single entries for each operator in each local authority area were made, in compliance with the basic provision of the 1995 Order. These entries simply aggregated the values for each separate property. Appeals were lodged by the appellants.
 Following receipt of the appeals and the issuing of an interlocutor in the present proceedings more comprehensive information was made available to Assessors. Assessors reconsidered the terms of the 1995 Order in the light of the information secured in regard to sharing arrangements. They recast their valuations of the appeal subjects to secure what they considered was a unum quid approach. To do this they adopted the sharer payments as the best evidence of what the rent achievable in the open market might be for a whole network, the unum quid. The individual sites were all re-assessed and if appropriate, higher figures were applied based on the evidence as interpreted from the sharer payments. Assessors recognise that the assessments stated in the issued notices could not be increased. However any change in the values of the individual properties could be made subject to the cap of the issued notice.
 The regulations in England are the Non-Domestic Rating (Telecommunications Apparatus) (England) Regulations 2000. These require that there should be a single entry in the List for each individual property in the name of the host. In effect the hosts will be assessed for the whole property including any sharer’s interest. Certain hosts are excluded from assessment by statute but are nevertheless included in the lists in respect of any sharers on their mast. A similar arrangement is in place for Wales. The value applied in respect of the sharers is the amount of additional rent arising from the sharing of the mast flowing through to the site landlord (the “payaway”).
 In Northern Ireland valuations for rating follow an approach close to the position in Scotland prior to the changes brought in by Assessors in 2006, namely a site value based on site rents with masts and any other rateable equipment valued on a contractors basis.
 The valuations in the roll against which the appellants have appealed are defended by Assessors on the basis of the changed approach based on an analysis of sharer payments applied to each sharer on the mast and with a value derived from these payments applied to the host as well. The resultant figures for each of the appellant company within a local authority area are then aggregated. There is then effectively a cap at the figure originally entered in the roll when first published with an effective date of April 2005 should the revised approach result in a figure higher than the roll entry.
 Mr Sharp, engaged on behalf of H3G and also the principal witness for all the appellants, has extensive specialist experience, over some 25 years, in relation to business rates issues and also rent reviews and lease renewals for telecoms properties. He has advised many of the companies involved at various times. He first summarised the matters in dispute and then described mobile telephony networks, the different types of sites and the different arrangements for use of masts and other sites. Mr Sharp was careful in his evidence to avoid expressing legal opinions, although his primary position clearly proceeded on the basis that ‘sharers’ rights’ were not rateable and that cumulo valuation, for administrative convenience, was appropriate He clearly had a degree of technical knowledge which enabled him to describe the subjects and their operation sufficiently for the purposes of these appeals.
 Mr Sharp drew the distinction between site leases, which the host establishes, and sharing by second and subsequent operators who instal their own equipment. He introduced and spoke to examples of the differing types of agreement. H3G 3 was a reasonably typical site lease, standard template leases often being used if accepted. As far as the operators were concerned, site sharing was not a transaction in land, but was a business arrangement. He spoke to variations of MSSA’s and site specific licences, with the licence fees determined from the pricing menus or rate cards. He spoke to features of the sharing agreements which appeared to him to distinguish the rights granted to sharers as licensees from real rights acquired by hosts under site leases. Referring to an example of a rate card, he identified 4 features. Firstly, while there was provision for different configurations of antennae, all included for cabin(et)s “up to 20m 2”, i.e. the amount of the footprint was neither defined nor determinative of the payment. Secondly, looking at the rates for different configurations of antennae, the distinctions would not be apparent to chartered surveyors and were driven by business reasons and not transactions in land. Thirdly, fees were reviewed periodically to RPI: even a new site share agreement would have an old rent which had been indexed, not a freely negotiated rent at the time. Fourthly, the sharer’s fee was reduced where it had to bear the cost, for example, of extending or strengthening a tower, again making analysis of site share fees unreliable as anything other than measuring the parties’ business relationship. Shared base station controllers were legally and conceptually no different from shared greenfield sites. At rooftops, shared installations were physically indistinguishable from host sites, but a sharer’s entitlement to share a rooftop should be treated in the same way as a greenfield shared site.
 Mr Sharp described the telecoms market in 2003. Stock prices had fallen sharply shortly after the grant of the 3G licences in 2000. The four incumbents tended to upgrade existing sites and acquire limited infill sites. H3G had to rush to build up a network and had to “throw money at the problem,” with an inflationary impact on deals which all the operators had to make at that time.
 Mr Sharp next explained his approach to the “evidence pool”. With most sites acquired by the hosts on a leasehold basis and the operator negotiating at arm’s length with the land or building owner, the site lease rents were good, and plentiful, evidence for rating purposes. This was a very substantial basket of evidence, both in absolute terms and in proportion to the number of sites. By contrast, payments for shared sites resulted from overarching agreements regulating parties’ relationships and negotiated in many cases years before the tone date; the price was dictated primarily by the number and size of items of equipment, these being the non-rateable antennae and dishes; the agreements were license agreements conferring personal rights, not leases of lands and heritages; the fee was not driven by the value of the land; the host undertook certain services; and the sharer avoided numerous items of cost that it would have to incur were it running its own greenfield host site. Numerous adjustments would be needed to equate to rateable value. Parties were agreed that, having selected the appropriate pool of evidence, averaging the entire pool was an appropriate methodology. Site lease rents reviewed to RPI had been excluded. The Assessor appeared to have “cherry-picked” unusually large sites, including a disproportionate number of BSC sites, as his ‘test’ sites.
 Mr Sharp elaborated on the contrast between arms’ length open market rents, to which it was only necessary to add the decapitalised cost of the rateable infrastructure, and the site share or licence fees. The latter needed adjustment. Mr Sharp found a means within the evidence pool of demonstrating the difference, by comparing rooftop site share fees with rooftop lease rentals, because, he said, both the rooftop tenant and the rooftop sharer had to instal both the rateable infrastructure and the non-rateable telecoms equipment. His analysis showed a difference of £1,900, which was therefore, he said, shown to be the value of additional items not according with the rating hyprothesis. That figure fitted a comparison between greenfield site share fees (£6,800) and greenfield lease rentals (£4,000) when the ratebale value of a typical 15 metre tower (£900) was also deducted. The most reliable basis of valuation of a host site was the leased rent plus decapitalised cost of infrastructure, with the large majority of the value so arrived at (80-95%) being comprised in the rent itself. Although this was ‘hybrid’, that could be said also of the ‘sharer payment’ basis, because that also involved decapitalised Plant and Machinery values for the cabin(et) and cables. Use of unadjusted site share fees was wholly inappropriate as requiring significant adjustments and a subjective and circuitous route to arrive at the same answer. If site share fees had any relevance, it was as demonstrating that once adjusted they supported use of the more direct rental comparison. Mr Sharp provided another cross-check from the rent paid (£5,500) by Orange to a football club for the use of floodlights for telecoms equipment (Clifton Hill Stadium, Assr 5.14): the landlord maintained the tower, contrary to the rating hypothesis (10%, i.e. £550) and deduction of the value of a typical tower (£900) again brought an adjusted rental (£4050) into line with the greenfield site lease rental of £4,000. He had not considered whether the landlords had been professionally represented, but they did have lawyers. There was another example (Easter Bleaton Farm, Assr 5.15) of a rent of an existing structure at a lower level of value where the landlord was not providing telecoms related services. These were two examples from the ‘test cases’, there being, he said, other examples.
 Mr Sharp said that he and the industry had been consistent in the view that the ‘site rental’ basis was appropriate. He also referred to correspondence between 2006 and 2010 (Assr 9.1,4 and 6) in which the appellants had offered calculations, on a ‘bottom up’ basis, attempting to quantify, on a rough and ready basis, the value of various services in respect of which adjustment of site sharing fees would be necessary. He listed the services referred to. The figure arrived at had been £1569 per site. Operators never quantified the value of such services.
 Mr Sharp then set out the appellants’ primary valuations (Scenario 6) on the basis of his opinion. On the face of it, this appeared to be a contractor’s test using the estimated replacement cost of the rateable infrastructure, but it differed, he said, from a conventional contractor’s test valuation in that the rental value of the land was derived direct from rents. Until the 2005 revaluation, all parts of the UK had used this basis, and England and Wales, and Northern Ireland, had continued to do so. Mr Sharp said that the 2005 assessments of Arqiva had been agreed on this basis, the figures having accorded with the appellants’ method and the two Glasgow site values being otherwise inexplicable. The method followed how these sites were established in the real world, with the large majority of the value derived from analysis of open market rents. Taking the totality of Scottish rental evidence (425 greenfield sites and 181 non-shared rooftop sites) was appropriate.
 It was Mr Sharp’s opinion that the Valuation Office Agency cost guide for the 2005 revaluation in England and Wales should be adopted for these Scottish appeals because it was developed by specialist plant and machinery valuers within the VOA contemporaneously, drawing on cost information available at the time from a number of sources. The figures had been discussed with agents for the operators. It was not possible in 2012 to establish actual costs in 2003 as operators did not maintain records going that far back. The Scottish Assessors Association stated on its website, in relation to the 2010 revaluation, that it adopted the VOA cost guide. That cost guide was used in settling values in respect of the Arqiva telecoms cumulos. The Assessor’s suggestion that the figure were too low arose, typically, from evidence of some special situations, such as tower strengthening works.
 Mr Sharp was unable to explain one small difference under Scenario 6, in relation to rooftop sharers, where the Assessor had £200 for Plant and Machinery (cabinets) and £500 (cabins), as against the appellants’ £300 and £600.
 Mr Sharp spoke to the appellants’ “fall-back” valuation, Scenario 4 (“Cumulo based on site rents – site sharers heritable”). On the ‘site rents’ basis, the appellants (as well as the Assessor) took the 27.5% payaway figure as representing the rental value of the sharers’ interests if these had to be included. Mr Sharp accepted that the addition of a sharer did “on many occasions” create an extra element of value for the underlying landowner, so that if instructed to include a land value for this interest, he would take these percentages of the agreed analysed sharers’ payments, i.e. £1,900 (greenfield), £2,400 (rooftop) and £6,000 (BSC). He defended the adoption of one third (rather than all) of the mast value in greenfield sharers’ values under this scenario, on the view that the sharers’ rights were restricted in comparison with the host and were intuitively of less value: when trying to find the real value, that corresponded with the amount the land was enhanced by.
 In relation to the valuations for each council area, Mr Sharp had proceeded on the basis that the 1995 Order provided for single entries only as a matter of administrative convenience.
 Finally in his principal evidence, Mr Sharp made a number of points in rebuttal of the Assessor’s approach. He criticised the Assessor’s unum quid position as being in his opinion an afterthought, not mentioned even in the SAA Practice Note for the 2010 revaluation (H3G 19): it appeared to be “a red herring justification for the most rough and ready of valuation methodologies.” It also ran contrary to the methodology of averaging across the whole of the country. The issue had, he said, not been raised in negotiation of the Arqiva assessments.
 Using the site share licence fees to value sites including host sites had, in Mr Sharp’s view, a number of pernicious effects. Site sharers were only about 25% of all the sites. Gross fees had not been adjusted. The methodology multiplied the value of the underlying land in a direct 1:1 relationship to the number of sharers, contrary to the evidence, and valued the mast multiple times. It failed to take account of the hypothetical tenant’s options when considering a streetworks site, the different rights at streetworks sites and the restricted possible identity of hypothetiocal landlords of streetworks sites. It would give rise to significantly greater rent liabilities on operators in comparison with the current real world, where operators had a mix of shared sites at high cost, host sites at somewhat less cost and rent free streetworks sites: the total amount available for rent (which he calculated at £4502 per site, less even than the appellants’ primary valuation of a greenfield hostsite) was fixed as the rating hypothesis does not permit extra revenue to be imported.
 In relation to costs, the Assessor had, said Mr Sharp, come very late to his attempt to show that costs were higher than the VOA cost guide, which, again, had been referred to in the Practice Note and on the Assessor’s website. Mr Sharp was particularly critical of the MFD International costing (Assr 7), and said that Mr Thornton-Kemsley’s productions on this did not contain any specific detailed investigation. He had not personally been involved in discussion of the VOA guide but understood it to be very much based on tenders.
 Mr Sharp said that there had been no payaway additions to the Arqiva values. Mr Sharp accepted that in the real world the landlord received these payments. Asked why they should not be reflected, Mr Sharp said that in the real world there was a sequence of title, i.e. landowner, host and sharer, whereas in the rating world anything rateable occupied by the sharer would be carved out of the host’s occupation, they both being hypothetical tenants of one hypothetical landlord (it being too much to think about two hypothetical landlords). To the extent that the sharer had heritable rights, the landlord would receive rent in addition to that from the host, i.e. he would receive both rents, but not if the rights were not heritable. He compared the situation with an office building which was sub-let and the tenant had his own assessment, but if it was not a sub-tenancy, the payments made were not in assessment.
 Asked about ‘shared occupation’, Mr Sharp said that in his view in that case there would be a joint occupier, i.e. a single hypothetical tenant.
 In cross-examination (insofar as not covered above), Mr Sharp agreed that the cabin(et)s were fixed. He agreed that the transmission parts commenced in the cabin, although there was both receiving and transmitting. He agreed that mast-sharing was commonplace in the industry. There was an obligation on sharers to ensure their equipment did not interfere with others’. He accepted that operators had their own expertise, but said that on a shared site the host had to ensure compliance among the sharers, although this was probably done at the outset. He agreed that sharing of a streetworks site was possible but it had not been done during the course of the 2005 Valuation Roll. He agreed that masts were located where they were required, sometimes in very precise locations, and a site in one particular place might save two or three others. He accepted, under reference to one site-specific licence (Orange 14), that there was essentially a letting of space, some sort of right under/over land. Sharers either got access to a power supply or supplied their own power, most having their own, separately metered, supply. Plans would normally specify where the sharer’s equipment was to be placed, following approval of the location showed on application drawings. He accepted that a rate card produced (T-K Appx 23) crossed between ‘licence’ words and ‘lease’ words. He accepted that a landlord might, or might not, restrict user. There were all kinds of possible explanations for substantial differences between sharers’ fees. He accepted that where a sharer’s fee used in the analysis had been reduced because of the cost incurred in extending or strengthening a tower that would make it too low. Variations in RPI indexed payments might be either higher or lower than OMV variations. He said that use of RPI was more symptomatic of business than land values. He considered that it would not necessarily be more attractive to operators to co-locate on rooftops rather than sharing as under the Arqiva-BT situation. He agreed that it was not possible on sight to tell whose antennae were the host’s and whose the sharer’s.
 In relation to rooftop shares, Mr Sharp did not accept that the BT rooftops would be more valuable to operators: they did not necessarily use BT, and there would be BT service in the street. Mr Sharp was asked a number of questions about aspects of BT rooftop shares, particularly with H3G as the sharer, within the basket of 18 used in his comparison of rooftop rentals and rooftop sharer payments.
 Mr Sharp said that the VOA cost guide figure for building a mast (£18,000) was what operators were building masts for at that time. Costs incurred included connecting up the technology, mostly non-rateable. Figures such as the figure of £28,500 in an H3G MSSA (H3G 2, p 79) had not been shown to VOA, who had looked at contract prices for construction works, rather than charges to site sharers. It would be necessary to ask what proportion related to construction. It was better to go to the constructors.
 Asked about figures showing the cost of renting an equipped site, Mr Sharp said that if there were above average figures, there must be others below. He agreed that a figure of £8,465 might be a standard charge by one company, “Spyder”, for a site equipped with a monopole, but that was a site sharing site He produced, and was allowed to lodge, an arbitration award (H3G 26) in relation to one rent of an equipped site, relied on by Mr Thornton-Kemsley (Clarendon Farm, Linlithgow), showing peculiarities in the circumstances and valuation. Mr Sharp did not accept the suggestion that the benefit to a tenant of an already equipped site, reflected in sharing fees, was not included in the appellants’ valuations.
 Mr Sharp was asked a number of questions about the services supplied to sharers without payment. He agreed that only the appellants could provide evidence about these figures. He denied that he had ignored the body of evidence of the level of sharers’ payments. He agreed that Arqiva had maintained that sharers should be separately assessed, although not necessarily on the value of sharers’ rights. He also accepted that the appellants had initially agreed that sharers’ rights were rateable.
 After Mr Sharp’s evidence was concluded, the Tribunal was informed that parties had agreed that six of the rooftop sites used by Mr Sharp in his calculation based on rooftop shares were not in fact shared sites and that re-calculation of the average of the remaining 12 rooftop shares produced a figure of £10,100 (compared with Mr Sharp’s original £9,000), thus increasing the differential on which he had founded from £1,900 to around £3,000.
 Each of the appellants’ witnesses spoke to reports in relation to each appellants’ ‘test case’ sites. They also set out the each of the appellants’ valuations in their two appeals.
 Mr Sharp himself reported in relation to H3G sites. He spoke to H3G’s primary valuations (Scenario 6), i.e. the total of their sites in the two appeals, viz. Glasgow (£317,300, as against the Assessor’s primary figure of £362,775) and Perth and Kinross, where H3G only had 7 sites, all sharers (£1,700, c.f. £40,250). Valuations, generally agreed with the Assessor, under each of the scenarios, were produced.
 Mr Sharp also said that all the H3G ‘test case’ sites had been selected by the Assessor, the appellants having been content with putting forward a number of Vodafone sites. Mr Sharp said that hardly any of them were typical, and went on to identify what he referred to as unusual features and/or reasons why he believed these sites to have been selected by the Assessor.
 The other three witnesses for the appellants adopted Mr Sharp’s opinion evidence and approach and gave similar evidence in relation to each company’s ‘test case’ sites, primary values in the two appeals and agreed values under the various scenarios, all following the same basis.
 Mr Donaghy spoke to Vodafone’s 10 ‘test case’ sites; their primary valuations for Glasgow (£389,700, as against the Assessor’s primary figure, £374,900) and Perth and Kinross (£280,300, c.f. £381,500); and the agreed valuations under each of the scenarios.
 In relation to a ‘Pico’ site within Glasgow Airport (St Andrews Drive, Paisely – Vod 3, a ‘Host Pico’ site in respect of which there is no dispute), Mr Donaghy explained that a national ‘access market fee’ of a minimum of £600,000 related to Vodafone being allowed to pick up signals before incoming passengers reached the normal access area. In relation to one ‘Greenfield Share’ site (Easthills, Lochwinnoch – Vod 6), he said that the national agreement (between NTL, now Arqiva, and Vodafone) allowed Vodafone to share the host’s cabin, i.e. it was left up to the parties to agree whether or not the sharers took their own cabin(et).
 Mr Duff spoke to Orange’s 6 ‘test case’ sites; their primary valuations for Glasgow (£397,000, as against the Assessor’s primary valuation of £316,000) and Perth and Kinross (£235,000, c.f. £354,000); and the agreed valuations under each of the scenarios.
 Mr Duff referred to provisions of the Agreement under which Orange erected an antenna and dish on a footbal club’s floodlighting tower, with a cabinet in a small compound at the base. (Clifton Hill Stadium, Coatbridge – Orange 1), at a rent of £5,500. This was a specific agreement, not under any form of national umbrella. It was treated as a lease and not a sharing agreement and categorised as a ‘Rooftop Host’. Mr Duff did not know whether the landlords had been represented. He also spoke to arrangements at a 50m mast treated as a shared Base Station Controller (Craigowl, Tealing – Orange 3, 4): the 1995 licence had not been disclosed, but there had been numerous Variations between 2000 and 2003 where alterations to the permitted installations of non-rateable plant had led to changes in the amount of the licence fee, but in Mr Duff’s opinion the fundamental right was not changing. Orange had their own cabin, inside one of several large buildings at the base of the mast, there apparently being several other, non-telecoms, sharers at the mast. He would expect Orange to have a key but perhaps to have to ask permission to take access. Mr Duff did not know whether Orange had the final say as to the siting of the equipment. At a leased rooftop site, i.e. a ‘Rooftop Host’ (Gleneagles Hotel – Orange 2), Orange were not permitted to share, but sharing was never an issue in such rooftop lease situations. Mr Duff agreed that, logically, that would have a depressing effect on value, but sharers were only a limited part of the basket of rents. He said that where there was permission to share, and a Payaway provision, that would amount to a business reward. At East Yonderton, Bridge of Weir (Orange, 6-12), Orange were paying less than another sharer, T-Mobile. He understood that Orange had a micro-cell there, which might explain that.
 Mr Dalton spoke to T-Mobile’s 2 ‘test case’ sites; their primary valuations for Glasgow (£305,200, as against the ssesso’r £485,050) and Perth and Kinross (£120,600, c.f. £313,100) and the agreed valuations under each scenario.
 Mr Dalton spoke to provisions in a redacted licence fee facilities chart in respect of a ‘Greenfield Sharer’ site (Dewshill Farm, Shotts – T-M 2). He was not, however, able to explain whether or not any services were separately charged, a matter he had not been asked to look into.
 Mr T-K is a consultant with considerable experience advising and acting on behalf of landowners in relation to new radio mast leases and rent reviews and arbitrations in relation to mast sites. He also has some experience of negotiating site sharing agreements. He has limited experience, and did not profess any specific expertise, in rating issues. He had been asked to provide an overview of the site rentals for the varying types of site and the payaway market during the period from 2000 to 2005. He had considered the ‘test case’ sites and also other transactions in the market place. He had analysed his own extensive database of UK-wide transactions with which he had been involved or which had been quoted to him by agents acting for operators, but he had not had access to the returns available to Assessors.
 Mr T-K described the radio mast market. He referred to site sharing agreements as involving landlords who were either telecoms operators or infrastructure providers. Competition for sites was very rare in this market. Mr T-K reviewed the development of demand, with steady increases in subscriber numbers and an anticipation of ability to offer mass market consumer 3G handsets by the end of 2005. There was at the time an increasing demand for base station sites. Masts were normally developed by operators for their own use, with sharing facilities as a secondary function, rather than built speculatively, although infrastructure companies sometimes acquired leases and developed sites for specific operators.
 In relation to the valuation approach, Mr T-K referred to definitions of ‘market rental value’. Acting ‘knowledgeably’ involved being represented. He considered that a landowner in a radio mast negotiation, unlike a shopkeeper, was not necessarily aware of the market. On one occasion, he had found that only 8 out of 21 landlords who had replied to his enquiry in relation to 49 sites quoted to him were represented. New lettings of equipped sites would provide the best comparable evidence, followed by rent reviews where the parties were properly advised, but there were disadvantages in over-reliance on rent reviews, particularly in this market where landowners generally had only one mast and were unwilling to risk arbitration. The tenants had access to a wealth of evidence and in his experience achieved rental settlements below market. Little weight should be placed on transactions where no evidence was provided to indicate that parties acted in knowledge of the open market. He would rank arbitration awards below agreed rents, as it would not be known how the parties had argued: it would be more of an opinion on the cases put than a valuation. Mr T-K had analysed rents within his database from 2000 to 2003 but considered rents up to March 2004 also relevant as illustrating the trend. In relation to the addition for fixtures including the masts, Mr T-K was surprised at the appellants’ indication that build costs were no longer available. He said that he was aware of the general level of costs from negotiations and information provided by tenant’s agents. He had drawn his comparable evidence from a wide variety of operators and, there being little evidence of regional variation, from outwith the immediate areas and from English transactions.
 In a section on rental trends, Mr T-K said that he had carefully analysed his database. There had been a gradual growth in rental values at greenfield sites from an average of £3632 in 2000 to a peak of £5,406 in 2009. There was a general trend of site, including rooftop, rentals upwards and above the straight line trend of RPI. He did not agree that H3G transactions should be excluded: they were acquiring new leases, the prime market evidence; and analysis did not support the assertion that they were significantly out of line with the market.
 Mr T-K spoke to his analysis of average rentals for greenfield mast sites. Neither the area of the site, nor of the cabin, made much difference, but height was a significant driver and there was a trend of higher rents for sites of higher masts. BSCs tended to command higher rents.
 Mr T-K spoke to evidence which he had collated in relation to build costs. He seriously doubted the use of the VOA cost guide as a means of assessing construction costs of radio masts. He said that he had always been taught to take a transactional approach, so that on a contractor’s approach he would prefer to look for evidence of actual build costs. Cost guides might leave out vital factors, such as the cost of obtaining an electricity supply. He did not know how the VOA had come up with figures or whether they had been given the type of transactional evidence which he had. He indicated that in the real world the landlord would expect a return of at least 7%, compared to the 5% decapitalisation rate used in the Guide. He accepted that he had only started to look at costs at a relatively late stage. He had referred to costs as a check to his equipped site rental valuations. He had started going back to look for vouching of his knowledge of construction costs. He said that the V.O.A. would have the same difficulty which he had, viz that the information was in the hands of the operators and not landlords or Assessors.
 Mr T-K referred to three rents, between May 2004 and May 2009, two review agreements and one arbitration award, of sites with equipped masts taken into account, at £9,000, £11,300 and £8,500.
 He gave his opinion, based on his consideration of comparable evidence (including his database), of the rental values of a number of the ‘test case’ sites, including:-
(i) East Fulwood Farm, Inchinnan (Assr 18) - Vodafone ‘Greenfield Host’ with a 15m mast. The actual site rent was low. Mr T-K would value this site on a ‘site rental’ basis at £4,200 plus £3,000 P & M including build costs (£60,000 @ 5%), i.e. £7,200 (c.f. appellants’ primary valuation - £4,000 + £1,300; Assessor’s primary valuation, on ‘sharer’s payment’ basis - £7,450 + £500 (cabin)).
(ii) Abercorn Street, Paisley (Assr 17) - Vodafone ‘Greenfield Host’ with a 20m monopole mast. Again, the actual site rent was low. Mr T-K would value this site on a ‘site rental’ basis at £5,200 plus £3,250 (£65,000), i.e. £8,450 (appellants - £4,000 + £1,300; Assessor - £7,450 + £500).
(iii) Undercraig Farm, Langbank (Assr 16) - Vodafone ‘Greenfield Host’ with a 55m lattice mast. Mr T-K would value this site on a ‘site rental’ basis at £6,500 plus £4,000 (£80,000), i.e. £10,500 (appellants - £4,000 + £1,300; Assessor - £7,450 + £500). Mr T-K also referred to 6 rents, between 1999 and very recent (“lease not yet completed”), including two new leases, 3 reviews and one “quoted” at review, of sites in England equipped with higher masts, at figures varying from £8,500 to £15,000.
(iv) Edinburgh Road, Perth (Assr 19) – Orange ‘Streetworks’ with a 15m monopole. Mr T-K would value this site on a ‘site rental’ basis at £3,750 plus £2,500 (£50,000), i.e. £6,250 (appellants - £2,500 + £100; Assessor - £7,450 + £500).
(v) Balshagray Avenue, Glasgow (Assr 20) – H3G ‘Greenfield Host’ with monopole, situated beside the road – Mr T-K would value same as (iv), £6,250 (appellants - £4,000 + £1,100 (cabinet); Assessor - £7,450 + £200).
(vi) Old Inns, Stirling Road, Cumbernauld (Assr 1) – H3G ‘Greenfield Sharer’, again a monopole situated beside the road, but this time erected by Spyder Facilities specifically for site-sharing with H3G, as an equipped site at a payment of £8,465. Mr T-K referred to the relevant ‘Enabling Agreement’ and ‘Lease’ (T-K Appx 12), as indicative of a lease of an equipped site, i.e. prime evidence. He would value this, on the basis of that rent, at £8,465 plus £480 (£9,600) P & M (cabinet, etc), i.e. £8,900 (appellants - £200; Assessor - £7,450 + £200). In Mr TK’s opinion, this transaction illustrated the artificial nature of decapitalising construction costs at 5% and using this to derive a rental value for equipped sites. (Mr T-K argued similarly in relation to Dewshill Farm, Shotts (Assr 3), where the site had been specifically identified by the operator albeit leased by (now) Arqiva and made the subject of ‘site sharing’ with T-Mobile).
 Mr T-K went on to consider rooftop sites. He pointed out that these (or equivalents) were to be found in rural as well as urban areas. In comparison with greenfield sites, the essential infrastructure – a building at height – was provided, potentially saving capital expenditure. There was less need to certify safety. Procuring power/utilities, and also “backhaul”/fibre connections (for switching or handing over), was likely to be less onerous. Although the re-negotiation of MSSA’s and Rate Cards might be much more complex than negotiating rent at one site, this was divided over many sites. There would still be a need to insure equipment and to provide indemnity. In cross-examination, he agreed that the infrastructure saving was very substantial and that there was also a saving in security. Access, however, required to be taken through another’s building. Mr T-K’s analysis of average rentals produced figures ranging from about £5,100 to £8,300 over the period from 2000 to 2004. He again produced a variety of figures (on which we have made findings above) on build costs, again varying considerably from the appellants’ valuations using the VOA cost guide, arriving at an average rooftop construction cost of £30,000, decapitalised at 5% to £1,500.
 Again, Mr T-K opined, on the basis of comparable evidence of open market transactions, on the values of rooftop ‘test case’ sites. His analysis of rents differentiated between city rooftops and town rooftops, which he considered were fairly stated at £8,500 and £7,000 respectively plus Plant and Machinery. (Neither the appellants nor the Assessor contended for any town/city/rural differentiation). He considered 5 ‘test case’ rooftop sites:-
(i) Gilmour Street, Paisley (Assr 12) – Vodafone ‘Rooftop Host’ with cabin, the subject of a ‘bare’ site lease. Mr T-K would value this site at £7,000 plus £1,420 P. & M. (the appellants’ valuation of these subjects being £7,100 plus £600; Assessor - £7,450 + £500).
(ii) Easter Bleaton Farm, Bridge of Cally (Assr 15) – antennae on concrete grain silo, a Vodafone ‘Rooftop Host’, with cabin. On the basis of two similar agreements in which he had been involved in 2003, Mr T-K valued this site at £7,300, apparently at a rent of £5,500 and on the basis of decapitalised costs of £36,000 (appellants - £7,100 + £600; Assessor - £7,450 + £500). He accepted in cross-examination that the actual rent was at a much lower level and that that might reflect a less valuable site to the operator, with less than the normal number of cell sites.
(iii) Westerlea, Crieff Hydro Hotel (Assr 23) – antennae within natural treetop, a Vodafone ‘Greenfield Host’, with cabin. Mr T-K would value this as a rooftop at £7,300, on a similar basis to (ii) above (appellants - £4,000 + £1,300; Assessor - £7,450 + £500).
(iv) Clifton Hill Stadium, Coatbridge (Assr 14) – antennae attached to stadium floodlighting tower, an Orange ‘Rooftop Host’, with cabinet. On the basis of 7 similar agreements, Mr T-K valued this site at £6,000 plus decapitalised construction costs of £1,000 (appellants - £7,100 + £300; Assessor - £7,450 + £200). In reaching his valuations, Mr T-K considered it noteworthy that his investigation had suggested that the landlords in 6 of these 7 cases had not been professionally represented.
(v) Gleneagles Hotel, Auchterarder (Assr 13) – antennae concealed as flagpoles on hotel roof, an Orange ‘Rooftop Host’ (with other appellants also there or elsewhere on the building, all with, apparently, bare site leases). Mr T-K offered no view on the value of this site (appellants - £7,100 + £300; Assessor - £7,450 + £200), but considered that the appellants’ construction cost (£6,000) was understated.
 Mr T-K’s arrived at final, standard rateable value figures for rooftops, of £9,920 (£8,500 plus P & M) for city sites, and £8,220 (£6,800 plus P & M) for town/rural sites.
 It should be mentioned that we have not included all the individual site valuations produced by Mr T-K, but have hopefully included enough to explain his position relative to areas in dispute.
 On site-sharing, Mr T-K accepted Mr Sharp’s estimate that about 25% of occupancies were on that basis. Planning policy guidelines and Codes of Practice encouraged that. It also saved capital and running costs. Certain operators appeared to describe their agreements as ‘agreements’ with the language of licences rather than leases, but exclusive possession for a term at consideration was apparently being granted. He considered that space on land or a building was being let for use by the sharer, who tended to have a cabin(et) or cage of their own or the right to instal one. These housed the sensitive equipment necessary for the operation of the network, both processing and transmitting calls. The site was modified by the installation of foundations and the cabin had obvious boundaries. The right to use the mast by hanging equipment from it would seem to be an essential pertinent of the space and structure occupied by the sharer, the air space involved being generally identified in the agreements by drawings or schedules. There would normally be a clause entitling the host to turn off the sharer’s equipment, at least if the host himself was operating equipment. The cabin(et) was normally accepted by the industry as being in the sole occupation of the sharer. The host’s rights did not seem to amount to predominant control.
 Mr T-K said he had considered the argument that there were matters included in the sharer rentals which should be excluded for rating purposes, and set out his consideration of the differences from greenfield site rentals. He accepted some of the specific individual cost items mentioned, and on a “very generous” basis, particularly in relation to questions about estate management and rent reviews, arrived at a figure of £827 per annum. Electricity did not form part of site share rentals. It was difficult to determine the value of maintenance of mast-type equipment on rooftops. He understood that a ‘sinking fund’ allowance would be inappropriate, and also developed an argument that as the 5% decapitalisation rate under-stated the return a landlord would require in the real world, any appropriate deduction should be applied to gross open market value and not any lower figure. He considered that the comparison between rooftop sharing payments and rooftop rents might not be comparing like with like, and there might be ‘double counting’ on this argument. He also referred to the practice of ‘recharge’ as a payment which would not, under ‘Payaway’ terms, be declared to the landlord.
 In Mr T-K’s opinion, the ability to site share was a valuable concession on the part of any landlord. There were also clear examples of ‘all-inclusive’ rentals where site sharing rights had been granted as part of the base rent. Such an effect on the landlord’s interest occurred whether or not the site sharing was by lease or by licence. Rents in his database did not include the payaway elements. When valuing a greenfield site for review purposes he would include payaway only where there was actual sharing, i.e. not for the potential for sharing.
 Mr T-K reviewed market evidence of site share payments (or rentals), including at ‘test case’ sites. On the basis of standard rights in the order of 6 antennae and 4 dishes with a cabin(et), he derived a site share fee of £9,250. He also cited one case (Cloburn Quarry, Lanarkshire) in which he had been involved in reviewing rentals paid by Orange and Vodafone for rights to instal equipment on one 15m mast, the agreements stipulating 5 yearly rent reviews to open market on an upwards only basis, with unrestricted equipment rights. That case, he said, illustrated the close relationship between transactions for equipment on landlord’s structures and site sharing. It was reasonable to assume capital costs of £14,000 for a cabinet and £26,300 for a cabin, decapitalised (at 5%) at £700 and £1,315 respectively, supporting rateable values in excess of the Assessor’s figures, allowing for some adjustment (or even the whole claimed adjustment of £1,900) for the fact that additional infrastructure had to be provided by the sharer and for the fact that the sharer was not responsible for the maintenance of the mast. In Mr T-K’s opinion, the Assessor, having started from a figure lower than £9,250 and added fairly nominal figures for the cabin(et)s in comparison with actual construction costs for site sharing, may have understated the value of sharing.
 Mr T-K supported the Assessor’s use of site share rentals obtained from the open market as better reflecting the valuation of equipped sites than the appellants’ approach. The rental value of an equipped site in the hands of an operator was unlikely to be less than the site share payment. This reflected better “the differential between a shared and equipped site with the operator in the latter being responsible for maintenance, etc.” In many instances the approach produced a figure less than his opinion as to value.
 On the basis that the rating hypothesis involved valuation of the network, Mr T-K believed that access to established networks of sites across the country represented by the unum quid would attract a significant premium. The Crown Castle – Hutchison agreement, in relation to BT sites, supported such an approach. He also understood that factors driving the market at the time – speed of establishment, or ability to compete – should not be excluded.
 Having given some evidence in relation to the analysis of ‘rooftop shares’ (where, as recorded above, Mr Sharp’s initial figure of £9,000 required to be altered to £10,514), Mr T-K said that this was a very small basket, very sensitive to individual rentals.
 In cross-examination, Mr T-K agreed that he was not in a position to attempt any ‘bottom-up’ valuation of services received by site sharers, and also that the revised arithmetic on Mr Sharp’s exercise based on rooftops suggested a higher figure. He agreed that it was dangerous to use a small basket of evidence and he had only taken four equipped sites. He agreed that having identified differences between city and town/country rooftop rentals, in his opinion a standard figure across Scotland was inappropriate. He agreed that new rents were preferable to reviews, but he had taken a review in which he had been involved after the valuation date in order to demonstrate the approach taken. His UK-wide database included rooftops and the equipped sites referred to, but the vast majority were not site sharer payments.
 Mr T-K was challenged on the suggestion that rents were not reliable because the landlords were not knowledgeable. He did not point to anything in the information provided to show any difference. Farmers, he said, did not necessarily always read references in the press to value levels. He suggested that similar rent levels with football club landlords might reflect operators’ offers. He agreed that he had not always classified sites on the basis agreed between the parties. He referred to sites not in fact shared because the ‘host’ was not an operator and there was only one operator ‘sharing’, but accepted that that would apply to all Arqiva sites with only one operator: no operator there could object to other sharers.
 In answer to questions by the Tribunal, Mr T-K agreed that, in relation to sharing agreements, he had never advised sharers in terms of what rent to offer, but rather as to what the rate card should be. He referred to other types of occupation of masts, e.g. for broadcasting, and said that he would, in valuing all types, consider revenue from other types of occupation. In relation to equipped sites, he would look for comparables, and would be going for a much higher proportion of a site share payment, but would always carry out a ‘reality check’ from greenfield site rentals and decapitalised costs (which he would never decapitalise at as little as 5%).
 In re-examination, Mr T-K said that operators paid site sharing fees on what they were entitled to, and might have reserved apertures for which they were paying. If a host took on two sharers using the same level of equipment, there would be no reason why the second should pay any different from the first (for example, Spyder at Stirling Road, Cumbernauld, if they had the right to bring on another operator). Operators would always have rights of access, the only question being how they were exercised. Mr T-K considered that if the operators had considered V.O.A.’s cost figures advantageous, they would not have volunteered costs information.
 Mr MacTaggart is a very experienced rating surveyor who has been the Assessor for Renfrewshire Valuation Joint Board since 2007 and was previously a Divisional Assessor. He is the Designated Assessor in respect of Fixed-Line Telecommunications under the Non-Domestic Rating (Valuation of Utilities) (Scotland) Order 2005. He is Secretary of the Scottish Assessors Association. He undertook lead responsibility for these appeals in around 2007.
 Mr MacTaggart explained how the entries came to be made one year late in March 2006. He said that the information returned remained far from complete or satisfactory when the entries were made. Each entry was described as ‘Telecommunications Network (Part)’, it being clear in his opinion that the 1995 Order directed valuation as a unum quid in each valuation area. Mr MacTaggart gave a brief overview of the networks, pointing out that while sites may be rural they nonetheless occupied strategic positions designed to maximise coverage, and many ‘greenfield’ sites were in fact urban or semi-urban: he would resist the inference that ‘greenfield’ was somehow less valuable. Each entry was the Assessor’s estimate of the value of interconnected sites as a single subject being a network. Operators saw the value of established sites as significantly greater than the cost of acquiring the bare site and erecting the necessary structures.
 Mr MacTaggart described the division of occupation between Arqiva or other host company and sharer operators who installed their own cabin(et)s. He referred to information declared on behalf of Arqiva. Arqiva values would not reflect operators’ equipment and so were residual. Arqiva, O2 and Airwave telecomms network subjects had all also been valued on the ‘sharers payments’ basis. Where the service was entirely provided by Arqiva, or the telecoms company put its equipment into the Arqiva building, there would be no separate entry. Other, larger types of transmission subjects, where there was no evidence of lets, were valued on the hybrid basis of site value plus equipment value, without consideration of sharer payments, with the proviso that if a telephone operator occupied space on such a mast, the appropriate unit value would be included in the unum quid valuation for that operator in that area. The Assessor had always taken the view that where a sharer had his own equipment and accommodation, there was sufficient separate occupation.
 Mr MacTaggart referred to the regulations in England and Wales, criticising their general effect as being to put the site host in rateable occupation, irrespective of the actual extent of rateable occupation, with only the host’s interest valued and that only increased for each sharer by the amount of the ‘Payaway’, thus missing the ‘sharer’ value. The valuation of BT’s ‘unbundled loop’ was harmonised, subject to the “absurdity” that there was no balancing increase in value despite reduction in BT’s ‘receipts and expenditure’ valuation resulting from loss of income from unbundled loops. There was no regulation for England and Wales comparable with the 1995 Scottish Order, so that each and every occupation had to be valued individually, the Scottish regulations providing a pragmatic way of reducing the number of entries.
 Valuation as a unum quid involved, in Mr MacTaggart’s opinion, valuing the network as a single entity, the subjects being used as, and only functioning properly, as an interconnected network. It was not appropriate to value the subjects as an aggregation of individual sites as if they existed in isolation and if this was done in practice it would be necessary to consider whether the end result was appropriate or whether the value as unum quid might be less or more. Here, however, the Assessor had made no end adjustment as it was understood that the rental evidence reflected the volume of sites. Given the available rental evidence, it was not proposed to value by means of the Receipts and Expenditure method. The total rateable value on the Assessor’s method was in the region of £30-31 million, or £40 million including O2.
 Mr MacTaggart turned to consider the component elements of the valuation. A value had to be established for the heritable parts of each company’s infrastructure listed in the 1995 Order. An agreement, for example, to ‘instal the Conducting Media to and through the site over a route to be agreed’ was synonymous with a wayleave or servitude and thus fell within the Order. It should attract a value but not double counting. The type of installation described in the ‘Pye’ case differed significantly in that there was a managed service where Pye hosted and managed the service and the ‘operators’ installed their own transmission equipment but shared the receiving equipment provided, the cable being only heritable by accession to the mast which belonged to the host and not the sharer. Here, the right to use the mast was a pertinent of the lands and heritages occupied by the sharer, i.e. the cabin(et) and the cable was deemed heritable by statute. The technology where a central mast provided wide area coverage was not compatible with the technology used in mobile telephony. ‘Pye’ was similar to the arrangement where Arqiva, etc. managed terrestrial broadcast facilities for BBC and independent broadcasters or, rarely, in relation to mobile operators who had no separate occupation of cabins. The host remained in rateable occupation of any residue, so would be responsible for the balance of overall value if the operator/sharer was deemed not to be in rateable occupation. The Assessor’s understanding of the distribution of occupation was supported in correspondence (Assr 9.1,2) and in the various sharing agreements, e.g. Assr 6.1-8, partic 6.3, 6.2,and 8.9, which appeared to be granting rights to locate/instal and remain, in short to enjoy occupation of the site for the purposes of mobile telephony.
 Mr MacTaggart held the buildings and equipment cabin(et)s as heritable, being firmly fixed to substantial founds and plinths and attached to the masts by cables, etc. They were buildings separate from the equipment, and were not ‘moveable’. Paramount occupation remained with the sharer. It was accepted that there were a variety of scenarios ranging from full occupation rights to situations where operators were only in rateable occupation of the cable as the equipment was installed in the host’s building in shared equipment racks. However, Mr MacTaggart considered that the sharer had sufficient permanence, possession, control and actual use of the buildings, cabins or cabinets to justify rateable occupation and that the buildings or cabin(et)s were so fixed and essential to the occupation of the mast as to be heritable by accession with “an additional heritable right in the form of a wayleave or servitude”.
 Mr MacTaggart explained how assessors had come to take a fresh approach in the 2005 revaluation. It had become clear in the period between 1998 and 2003 that interests other than bare site rents existed and that they had considerable value. There was a lucrative “industry” for the provision and letting of developed masts or development of masts “to order”. The subjects were used and occupied as a network. There was no value to an operator in having just one mast. Large scale agreements between hosts and operators indicated a “network effect” on value. Substantial additional payments were being made for “Network Access” (Crown Castle/H3G agreement - Assr 6.2) or “Access to Market” (BAA/Vodafone - Assr 6.8). The Regulations required the assessor to consider a hypothetical network within a single council area and value that network in its actual state and as it was occupied and used. There being no evidence of the let of an entire network, an alternative body of evidence had to be found and this was the arms-length payments by operators to unrelated mast hosts. These were not agreements between competitors in the same market. What was being valued was the pertinent or heritable rights attaching to occupation of a series of existing masts or structures for the purposes of operating a cellular communications service, i.e. what a hypothetical operator would reasonably pay in rent to occupy what already existed in order to access the established sites necessary to provide a seamless network. The limited evidence available did not warrant any adjustment beyond the averaging reflected in the analysis. Account was taken of the different types of occupation – Macro/Micro/Pico. “Rate-card” agreements as to proportionate values were not dissimilar although the Assessor’s approach had a broader brush: market levels of value might reflect the bulk acquision of the sharer’s interest in multiple sites. The net rents reflected any deductions for occupation of many sites at once. Consideration was given to ‘sharer payments’, ‘ground/site rents’ or ‘bare rooftop rents’, the first of these being considered the best evidence of rents paid to occupy space on established masts/structures within an established network infrastructure; or, failing that, evidence in respect of lets of established masts. The site rental basis was inappropriate as not valuing the established site and as resorting to contractor’s valuation – a method of ‘last resort’ – despite a significant body of rental evidence being available. To value merely by considering the ground rent plus a proportion of any sharer rent plus any rateable plant and machinery was analogous to valuing an office block with reference to a ground rent plus a proportion of rents received for offices within the block plus some decapitalised plant and machinery. That did not properly reflect actual rateable occupation or the true value of the various rateable occupations.
 Responding to Mr Sharp’s identification of particular features of the sharing agreements, Mr MacTaggart pointed out that the fees were based on the amount of equipment, including the value of equipment housing, as part of a unified whole. The size of the equipment was of little consequence compared to the ability to attach to the mast. There was no evidence, he said, of the practical effect on valuation of linking sharing fees to RPI. If an ingoing tenant had to make a contribution, e.g. towards improving the mast, that might have to be discounted but there was no way of knowing that and it would not work in favour of the Assessor. The most comprehensive evidence available in 2006 related to agreements between infrastructure providers and operators.
 Mr MacTaggart mentioned the practice of “recharge”, something which the Assessor had not been able to establish but which would mean his values were under-stated.
 In Mr MacTaggart’s opinion, the sharer agreements, for example NGW/H3G (Assr 6.3) indicated contracts which conferred the benefits normally enjoyed under a lease. The typical ‘licence’ should be taken to have the status of a lease. In any event, they could be considered as wayleaves, which were rateable.
 Mr MacTaggart then spoke to his analysis of the rental information in respect of sharer payments and site rents. A significant body of evidence had been considered. The site rent information was not ignored, but did provide a useful balance or check measure. Comparison of “bare site” rents for rooftops indicated that a site at the required height stood comparison with payments made to share a greenfield mast. It was understood that certain amounts might properly be deducted to allow the rent to meet the statutory definition, although there was confusion between the appellants’ references to ‘non-rent’ items (of which no detailed evidence had been produced) and their references to ‘overage’. Any differences would have been minor. The true net rents were equivalent to the Arqiva rents. There would be a supplementary charge to cover the main services provided, such as electricity. The amount paid to instal sharer equipment was reasonably consistent across the various installations, which could be broadly categorised as: Macro sites with cabin(et); Macro Streetworks with cabin(et); Pico Cells; and Cable only sites.
 Analysis supported the level of value on a ‘broad brush’ approach of sharers and hosts at all macro sites of £7,450 (including at BSC’s because this was a value for individual occupation) – the Assessor’s primary valuation on the unum quid basis (Scenario 1). This approach was supported by rents paid to occupy existing structures as a whole. The correlation between normal rooftop rents and rooftop sharer rents was clear. Tenant’s plant, i.e. the cabin(et) was added, at £500 (£200). The remainder of the rateable plant was deemed included in the rental paid. The only entries to which the Assessor did not apply these values were the Micro/Pico sites where values were agreed at £2,100 plus £100 P & M, i.e. £2,200.
 Mr MacTaggart then spoke to the rents/sharer fees and the Assessor’s primary valuations at each of the ‘test case’ sites, in each case contrasting both the ‘VOA’ approach, i.e. his understanding of the result in England and Wales, and the appellants’ approach (Assr 5.1-24).
 Mr MacTaggart addressed the £1,900 deduction sought by the appellants from the value of £7,450 on the ‘sharer’s payment’ basis. If this was for the ‘convenience factor’, i.e. the sharer not having to develop the site, it was illogical and unjustified: to the contrary, the landlord/host was receiving compensation in the rent for his expenditure/risk. Mr MacTaggart would agree that certain outgoings, if properly established, should be deducted in order to arrive at net annual values, but these should come off the gross rent and not the analysed net position. Mr T-K’s figure of £827 should be back-dated to 2003 (at 0.8, i.e. £661). Further, on the same basis as scenarios in which a factor of 3 had been taken to allow for sharers, these costs should be spread, so a figure of £220 would be accepted as reasonable. In short, Mr MacTaggart accepted the principle; did not accept that any figure was made out; but if it was, the deduction should be £220. The adjustment required to Mr Sharp’s comparison of rooftop rents with rooftop sharer payments indicated the sensitivity of that exercise.
 On Scenario 2, where valuation was on a cumulo rather than a unum quid basis, but still on the sharer payment basis, each site type would require to be valued according to its own level of value, i.e. without the “networking” effect, producing the Assessor’s figures (agreed subject to the claimed deduction of up to £1,900), viz. Streetworks and Greenfield - £6,800; Rooftop - £9,000; and BSC - £22,000), all plus the same agreed figures for P & M. This would not result in any change to the total applied Rateable Values although the balance between council areas would be distorted (with some values having to be adjusted as exceeding the original entries). Mr MacTaggart said that a cumulo using sharer payments should result in a higher value as it would have to be assumed that each site existed in isolation, but there was no direct evidence for either discounts or premiums for individual lets.
 Mr MacTaggart set out the Assessor’s position on each of the four scenarios (3 to 6) on the site rentals basis. It was not accepted that this method accurately reflected the net annual values of the units of occupation, but to assist the Tribunal the Assessor had agreed the levels of valuation on that approach. This approach was the application of Contractor’s Valuation in defiance of the body of rental evidence for shared sites and lettings of equipped masts. It took no account of the difference in rent for a site at the required height above ground level. It also accepted that the VOA cost guide was the most appropriate source of cost information superseding actual costs as led in evidence. There was a significant disparity between the appellants’ approach and actual rental income receivable by the host. The hypothetical landlord would be undervaluing his interests significantly. The appellants had not analysed developed site rentals or checked against rents of sites which already provided the necessary height.
 Mr MacTaggart spoke to obtaining cost estimates from experienced quantity surveyors, and to examples of actual cost returns made to Assessors, showing that the VOA guide undervalued the cost of development. That guide would only have made use of the information available, which may not have been complete. It was a ‘guide’, which was why the Contractor’s principle had a final stage of standing back and looking at the result. He had seen nothing to support the cost guide figures. Generally, the appellants, having valued sharers’ interest at, e.g. £200 and £500 (Scenarios 5 and 6), had not explained how the balance of value was distributed among the occupiers at a site and had made no attempt to ‘stand back and look’. For example, at Easthills, Lochwinnoch, the rental value of sharers’ occupation was £33,700, a host’s value would be agreed to be £8,300, but the appellants’ values totalled £2,000.
 Mr MacTaggart said that his approach did not double count, but reflected the actual occupation of the mast or structure with a residual value on the host if there was any residual occupation. The number of sharers reflected the quality/strength of the mast: the more expensive the mast, the more the income. If anything, the masts were undervalued where sharing potential had not been achieved. By contrast, the appellants had attempted to replicate the situation in England which was “artificially formed by regulations”. This was in conflict with the normal rules of rateable occupation simply to arrive at a harmonised value. The Assessor’s preferred approach was the best and most appropriate measure of value for the array of masts occupied to form a mobile telecommunications network, which was an array of developed sites whether provided by hosts or “owned and occupied” by the operator. It stood comparison with the ‘bare’ site at the necessary height above ground level and with equipped sites re-let as they stand.
 In cross-examination, Mr MacTaggart agreed that a basket of 6 rents would be even less reliable than the reduced basket of rooftop sharer payments. Various forms of wording in the agreements, suggested as highlighting the difference between leases and licences were put to him. He said that there was a variety of wording. The word ‘lease’ had been used in other sharing agreements.
 Mr MacTaggart said that he would sometimes consider whether parties had been professionally advised and consider whether a rent was outside the general pattern, this being quite common with secondary or tertiary shop rents. He agreed that whereas Mr T-K had differing analyses for masts at different heights, the appellants had higher figures for BSC’s/THS’s. He had become aware of equipped site rents at a late stage, after the disclosure of documents. He agreed that the Old Inns, Cumbernauld site had been characterised as a sharer, but there was sole use there; similarly with the Canal Social Club, Kirkintilloch. These and Mr T-K’s two examples were not a basket but a background check. He felt it was difficult to conclude that the difference between the two ‘Spyder’ rents was down to different rate cards. He agreed that the Tender Price Index had regional variations and had moved from 94% at the last revaluation to 102% in the current one.
 On costs, Mr MacTaggart had reservations about the amount of information used for the cost guide, in relation to the amount of machinery connected with telecoms. He had not been involved in making up the guide. He accepted that information from Scotland had been included but was very limited. He accepted that various items were not unique to telecoms sites, but not, in his opinion, including foundations. He said that if actual costs were used, the appellants’ contractor’s part would be more than 50% of value. He agreed that the VOA figures has been used at the main transmission sites in 2005 as well as in 2000, but he had not been involved in enquiries before those values were settled. He said that the Assessor had continually “faced a brick wall” in relation to information about costs: this industry did not like to share information. He agreed that he had initially instructed Mr T-K only in relation to levels of value. The question whether the cost guide was appropriate had been raised some 10 years previously, when it had begun to be understood on the limited evidence available that use of the cost guide was not bridging the gap with payments for the site as it stood. He had not known what telecom-specific information VOA had. The decision to move to the ‘sharer payment’ basis had been made in 2005. The Assessor had attempted to obtain more costs but had had very limited information. Rent actually being paid for the equipped site showed the difference. The advantages of telephone exchange rooftops were as good an explanation of higher payments as the suggestion as to services supplied by hosts. He defended dividing the figure for services by three: just as at a tenement, a joint owner would not pay the whole.
 Mr MacTaggart insisted that he was valuing the whole network, having anlaysed the whole evidence in Scotland. To the extent that less might be allowed for quantum the allowance should perhaps be less in Scotland than in the rents from the national agreements.
 Mr MacTaggart agreed that some of the large broadcasting sites not valued in the same way had many ‘sharers’. It was a matter of fact and degree. There were entries at only £100 where equipment racks within the host’s building were shared. Mr MacTaggart did not know what other services were being supplied in that situation, but this was not rateable occupation in relation to use of the mast because all that was happening was that the ‘sharer’s’ signal was being routed through the mast. He did not know by how much the valuations of non-telecoms operator hosts, e.g. at Craigowl, had been increased on account of income from such sharers with £100 entries, but would assume that payaways would be included. He agreed that what was important was whether there was any payaway, but pointed to the possibility which had now become apparent that negotiations between the host and the sharer took account of payaway where there was agreement to ‘recharge’, i.e. reimburse the host for the payaway.
 Mr MacTaggart considered that if, hypothetically, two operators jointly took the site from the landowner (as opposed to one, the host, taking the site and the other sharing) and if they each had separate cabin facilities, etc, there would still be two entries, doubling the value of one. If, however, they shared racks in a single building, it would be necessary to decide who was the host; the other would only have a £100 entry; the host’s value would be increased; and the base level of value might be higher. He agreed that on his preferred approach the pertinent interest would have the same value as that of the host: this was not double counting but discreet rateable occupation (unless the host was not in rateable occupation). He agreed that sharers had less physical ground, although they had access across the hosts’ ground. He accepted that the sharer paid according to the number of aerials, not in comparison with property values. However, such equipment only had that value if they also had the equipment housed on the ground, so there was value in land: properties could be measured in various ways, and he was using the same measure on both sides.
 In answer to questions by the Tribunal, Mr MacTaggart agreed that it was only the 1995 Regulations which led to his view on unum quid. He could not say what had led to this provision. He considered that his approach was in accord with the statutory hypothesis, as sharer agreements between two parties related to occupation of lands and heritages and sharers occupied at least some part of the site. The payments were in respect of real rights. He had preferred to look primarily at the Arqiva information because evidence of payments between parties who might one day be licensors and the next licensees might be less valuable. Rent review to RPI was becoming more common. Differences between payments made by operators at the same site gave pause for thought, but there might be an analogy with, e.g., Boots at a shopping centre. H3G did not pay proportionally more but tended to push the others up. The Plant and Machinery Order left questions in relation to items, such as aerials, which were neither included nor excluded. Apart from the Order, the question was whether items were heritable by accession. Attaching to something, e.g. putting flags on, or searchlights, might involve a specific right but not be sufficient.
 In re-examination, Mr MacTaggart said that although examples of rents from landowners of equipped sites were very limited, their levels could be contrasted with the lowest rents on the appellants’ basis, viz. £4,000 plus £1,100, following analysis of hundreds, sometimes thousands, of site rents not including payaways. On Mr T-K’s evidence as to costs, the proportion of P & M to site rental might be more than 40%. On the suggestion of double counting, although the host had a bit more land, the sharer’s value based on the sharer payments was no less and no more. The method of averaging across the board resolved the issue of differing values, e.g. between macros and micros. There was no evidence of RPI going up by more than OMV before 2003, but if there was such a difference at a particular time, operators would not be forced to take a site.
 Mr Haddow elaborated on a written outline submission.
 Mr Haddow said that this argument was based on the 1995 Order but a unum quid approach had not been adopted in the 1995 or 2000 revaluations or, at least in terms of consequences, in the preparation of the 2005 revaluation. The appellants contended that the provision was no more than one introduced for administrative convenience and to avoid the need for multiple entries, as happened for subjects like, e.g., bus stops, and by contrast with provisions in relation to utilities. Separate entries for each rating authority indicated a historical approach. Neither the Order nor Section 6A(1)(a) provided that separate lands and heritages were to be combined to form a single unit of valuation. They were only to appear in one entry. The Assessor’s approach to valuation was really inconsistent with his contention, the only suggested effect being to maintain that it was not necessary to differentiate between several categories of mast, an averaging approach which was a misconceived ‘part-network’ approach, with averages taken from the evidence across the country. If the appellants’ ‘site rental’ approach was preferred, an opinion on this issue could be reserved. The guidance in Coventry and Solihull v Russell on reference to explanatory notes was accepted, but there was no ambiguity here.
 Mr Haddow first submitted that one of the main aims of the revaluation system was to avoid double counting. This was a long recognised principle – see Assessor for Dunbartonshire v Baird & Co, per, L. Hunter at 482, L. Sands at 483 and L. Morison. With mobile phone masts the value to the sharer was not in the cabin and cables but in the ability to use the mast. What had to be avoided was the value – or any part of the value – of rateable subjects appearing more than once in the roll. The mast was properly in the valuation of the host’s interest, whether considered heritable in its own right or as rateable plant (Mr Haddow considered it to be both), and that value should not reappear elsewhere in the roll. The principle against double counting was not limited to sewers. It also extended to car parks (Armour, 20-05B; Ravenseft Properties v Strathclyde Assessor, per, L. Clyde at 269, L. Coulsfield at 271; B & Q v Assessor for Central) and common parts of covered shopping malls (Assessor for Central v Samuel Properties - now Section 8A of 1956 Act). Whatever the result of applying other principles, any decision which would give rise to double counting must give way to the principle against that. The English and Northern Irish approaches both met that approach (increases for ‘Payaway’ being passed through the tenant host to the host’s landlord).
 Mr Haddow next analysed and assessed the Scottish mast-sharing cases:-
Central Fire Area Joint Committee v Renfrewshire Assessor, per, L. Hunter at 244-5
ITA v Assessor for Lanarkshire, particularly, per, L. Fraser (dissenting) at 255-8, L. Avonside at 259-261 and L. Thomson at 263.
Assessor for Aberdeenshire v Pye Telecommunications Ltd, particularly per, L. Fraser at 163-5 and L. Avonside at 168-9
AA v Assessor for Lanark
Scottish Gas v Assessor for Fife
 On wayleaves, only L. Avonside in ITA had considered the right in question to involve a wayleave and had subsequently stepped back from that view. Parish Council of Edinburgh v Assessor for Edinburgh was not a case on which to found in relation to wayleave.In the Central Fire Area, AA and Scottish Gas cases, where the appellants did not have cabins or buildings, the court had no difficulty in rejecting the argument that the right was heritable. The ITA case was confused, as Lords Fraser and Thomson had referred more to pertinents. Regard had to be had to the particular terms on which the right was granted.
 Mr Haddow then focused on the application of the law in the current appeals, starting with a consideration of what was being valued. In the case of host sites without sharers, that was straightforward. In the case of sharers on a host site, the host had the whole site and erected its mast. He retained, and did not confer, occupancy of the site on the sharer. It was not even a shared right of occupancy. In terms of the licence agreements, the principal objective was for the sharer to obtain the right to utilise the host’s lands and heritages in the sharer’s mobile phone business, but without taking full occupation and control of the site and mast. The host’s lease “behaves like a lease”, whereas the first item in the license agreement, even where the sharer had his own cabin(et), was a right to attach aerials to the mast. It was important to separate the value of any personal obligation or consideration other than the right to occupy ‘lands and heritages’. Mr Haddow did accept, apparently on the basis of accession, that sharers did have rateable occupation of a heritable subject, the cabin(et) site, within the host’s ‘premises’ under the Valuation for Rating (Plant and Machinery) (Scotland) Regulations 2000, hence their acceptance of ‘cable only’ entries. The host, however, could allow other sharers; and sharers could not simply find their own space.
 The right to attach the aerial to the mast could not of itself be entered in the roll as an incorporeal heritable right, otherwise the cable only entries would be increased by the value of that right. The majority of the court in ITA had, against the background of the lease there, seen it as a part or pertinent of the cabin and its site. The alternative, where the intention of the parties majored not on an area of land but on the right to attach aerials to the mast, was that this was a personal right. The general use of licence agreements, in contrast to the leases taken by the hosts, pointed that way. This submission started with a consideration of the nature of occupation necessary to make it rateable and then proceeded by looking at the circumstances in which sharers acquired their interest as occupiers of discrete areas of land and their right to attach aerials to the host’s mast and run cables between the two. Not every regular payment under an agreement relating to lands and heritage could properly be looked at as in respect of occupation of the heritable subjects. C.f. John Menzies v Glasgow Assessor, particularly per, L. Fleming at 297-9, Perth Magistrates v Assessor for Perth & Kinross, particularly per, L. Fleming at 558-9, Armour at 19-03, Bo’ness T.C. v Linlithgowshire Assessor, per, L. Dundas at 779 and Lothian Assessor v Lowland Leisure Ltd, particularly per, L. Clyde at 356G-H, 357-8, L. Prosser at 358-9 and L. Coulsfield at 360. There was a warning in the latter case to keep to the facts of ITA or, at least, be slow to extend that decision beyond its particular facts. There was no dispositive clause (c.f. the ‘demise’ in ITA) in the general run of licences (although it was accepted that the words used could not be determinative). Were it not for ITA, there would be no doubt that the licence rights were personal rights to use the host’s lands and heritages, i.e. incorporeal rights no more appropriate for entry on the roll than the rights of Lowland Leisure, John Menzies or the Air Training School, who all had subordinate but associated rights to use areas of land defined to such an extent as to amount to areas in separate occupation.The sharer’s primary, and valuable, right was the non-exclusive right to attach aerials to the host’s mast. The ability to have a cabin(et) was ancillary. ‘Payaways’ were anomalous, not a normal feature of leases. On the evidence, the mere possibility of payaways did not increase site rents. The payaway was a percentage of the sharer’s payment which, on the appellant’s analysis, was not a payment for the right to be the rateable occupier. If it added to the host’s value (though only where there were actually sharers) it should not also be reflected in the sharer’s assessment, the sharer’s payment being a payment for a personal right. To add it to a sharer’s valuation was to introduce double counting. Morrison v Central Assessor was an example of the treatment of turnover-based “top-up” rent.
 Mr Haddow submitted that this was not an issue of law but purely one of valuation. Both the appellants and the Assessor had adopted hybrid valuations, the appellants’ valuations being about 80-95% site rents, the balance being VOA cost guide and the Assessor’s normal case (‘macro’ site with cabin) having costs greater than 5% of the valuation. Mr Haddow noted, under reference to the 2010 SAA Practice Note, the practice of hybrid valuation on the ‘site rental’ basis for ‘main transmission sites’ and, at least on the appellants’ understanding, in Arqiva entries. A TGV broadcasting mast valued in that way was no different from a mast used by a mobile operator. The real strength of using site rents was that these were the rents actually negotiated and paid to the landowner. The hypothetical landlord was not, typically, a mobile operator or an infrastructure company. Mr T-K’s criticism of the one-sided nature of the negotiation ignored the obvious information networks available. Bodies such as football clubs and hotels were likely to have business persons as board members. Mr T-K had in any event not really disagreed with the site rent levels adopted by the appellants. The agreed site rent levels were accurate and derived from a large number of disparate landowners and the various hosts. It would be absurd to conclude that they were unreliable.
 The cost figures were a subsidiary element. Following the analysis and averaging, they were not necessarily exact in respect of each site. The VOA cost guide had been accepted up to this revaluation; continued to be so in England and Wales and Northern Ireland; and was still currently listed as approved by SAA. It was based on information supplied to VOA in a thorough exercise, and discussed and agreed with operators’ representatives. Many of the items costed were not unique to mobile phone operators. Aerials were not in valuation. The case against use of the cost guide was weak. The Assessors themselves appeared to have done no work on this. The MFD International report suffered from some serious disadvantages. The very recent remit to Mr T-K had not led to the provision of any alternative cost figures for general application.The appellants distinguished between ‘ordinary’ mast sites and BSC/THS sites. The cost guide might understate the full costs at some sites because they assumed electricity provision entering the site, which Mr T-K had shown to be incorrect in some examples, but there was no indication of that being a common feature in the ‘test case’ sites. In relation to rooftop sites, the suggested financial advantage of telephone exchange roofs involved a departure from reality. The criticisms might, on the evidence, be based on a few extremely atypical sites. Not all of Mr T-K’s evidence demonstrated figures different from the cost guide ones. C.f. Shell UK Exploration and Production Ltd v Grampian Assessor, at 296-7, 335-6 and 347 and following. The attack on the strength of rental derived evidence and use of the cost guide had done nothing to destroy the appellants’ method of valuation.
 Turning to the sharer payments basis, Mr Haddow said that the typical sharer’s payment was made to another mobile phone operator, although there were other situations, including that in which an infrastructure provider had provided the mast and site for a mobile phone operator who then occupied the site alone. In that example, the site provider has the tenants’ rights to the site subject only to the rights granted to the mobile operator and reserving the right to licence other users. The provider was then occupying the site for its own purposes as an infrastructure provider, the mobile operator’s right being still subsidiary. The provider, like other hosts, will be paying the site rent and not a rent based on the sharer payment. The basis on which site share payments were based was not the open market, but rather a set of payments negotiated en bloc. In the case of mobile operator hosts, that was a universal agreement reciprocal as between those two parties, but there was no universal agreement among all the operators. The agreements were business contracts rather than dealings in land. Payments were not dictated by land values and decapitalised costs but varied with the amount and type of equipment for which the sharer was licensed. It was highly significant that one of the main determinants was the number and type of (non-rateable) aerials. The area for the cabin had a subsidiary role in assessing the licence fee. The benefit to the sharer was the personal contractual right to transmit and receive from the host’s mast. The amount might alter frequently although the value of the site to the hypothetical tenant would not have altered. The sharer fees often had a historical basis with even new agreements often being based on RPI increases of earlier ones, which was not in accordance with the rating hypothesis. There was also a difficulty as to what the payments included or excluded, the information provided being described as net payments. This might take discounts into account, but might also include payaway recharges. The latter might be included in the agreed license fee analysis. Both the assessor and Mr T-K acknowledged the need to adjust sharer payments which would include the benefit of spared expenditure, of which the most obvious element was the cost of the mast. Assessment of savings in expenditure, timing or convenience had been next to impossible to calculate. The appellants’ attempt at ‘bottom up’ assessment was not particularly successful, although it was misconceived to divide mast costs by three as each sharer was saved that cost. The attempted ‘top down’ assessment now produced a figure of £3,000 (although only the £1,900 claimed could be allowed), but that exercise perhaps mainly showed that sharer payments were not a reliable measure of land values. If sharer payments were to be the measure of valuation, the value of the same site would increase in multiples according to the number of sharers. Further, the host’s valuation would be in excess of what he paid in rent and decapitalised costs: the value would not be being determined by reference to land values. Even if payaways were taken into account as part of the site value, the increase was not of that nature.
 Mr Smith similarly spoke to an outline submission.
 Mr Smith submitted that the effect of Article 2 of the Non-Domestic Rating (Telecommunications and Canals) (Scotland) Order 1995, made under Section 6A of the Valuation and Rating (Scotland) Act 1956, is that the telecommunication subjects within the local government area fall to be valued as a single unit of valuation. Each of the conditions (a), (b) and (c) applied. The items listed in (c) were parts of a telecommunications network. A single entry in the roll was required. The legislation suggested that the intention was more than mere administrative convenience. “For all purposes” in Section 6A(1) meant just that. The Explanatory Note to the Order referred to provision for treatment “as a single valuation unit”. C.f. Coventry and Solihull Waste Disposal Co Limited v Russell (V0), per, L.Hope at 2103, on looking to the explanatory note in secondary legislation. This interpretation was consistent with the position in England (Article 4 of the Non-Domestic Rating (Railways, Telecommunications and Canals) Regulations 1994. Accordingly, for the 2005 Revaluation, assessments had been made in respect of ‘Telecommunications Network (Part)’ for all valuation areas. The practical differences between a cumulo and a unum quid approach might not be very significant in this case, although the cumulo approach might have undervalued the subjects as there may have been discounts for multiple sites under the various agreements. The Assessor had not found any basis for an upwards ‘network effect’.
 Mr Smith first referred to the appellants’ concessions that the cabin(et)s and the cables were rateable and that the sharers were in rateable occupation. These concessions were well founded. Cabin(et)s were ‘lands and heritages’ by accession and in any event were ‘buildings’ under Section 42 of the 1854 Act: Armour 9-13, referring to the first rule in Brand’s Trustees at 20; Dowall; David Allen & Sons v Assessor for Clydebank (advertisement hoardings); Assessor for Glasgow v Gilmartin, per, L. Salvesen at 493; Lowland Cereals v Assessor for Lothian, per, L. Robertson at 141 and L. Avonside at 139; Lothian Assessor v Blue Circle Cement, per, L. Robertson at 548B. The cable was heritable in terms of Class 3(e) of the Plant and Machinery Regulations, and would also be heritable by accession (c.f. Pye).
 Mr Smith then looked at the nature of occupation, identifying the four significant factors (Armour 14-03 and following; Renfrewshire Assessor v Old Consort Co). Title was not conclusive. A ‘mere right of privilege’ did not give possession, but construction of the cabin(et)s would constitute ‘works upon the soil’. Wayleaves might be classed as heritable for valuation purposes, even where they were a personal right of way (Armour 8-21). Mr Smith considered the MSSA’s/EA’s and the individual agreements/licences. They provided expressly for ‘provision of space’ on the mast and on the site; or alternatively, the right to instal, operate, maintain, repair and renew the cabin(et)s, cables, etc. in certain locations or ‘positions’, necessarily implying work carried out, and items placed, on the land. All the agreements provided for the establishment of the location of the equipment.
 The distinction between Pye and ITA was occupation of something heritable, and the lease was the point of distinction. The agreements involved the right to use land or heritable structures, and as such were leases recognised under Scots law. C.f. Brador Properties v BT, per, L.J.-C. Ross at 19-22; Rankine on Leases, at 1,2,89,208,211; Hunter on Landlord and tenant at 212. The words used were not conclusive either way. There could be a lease despite numerous reservations (David Allen at 327-8; Street v Mountford, per, L. Templeman at 819F). The sharer’s possession was to a degree exclusive and permanent and was protected from interference by the host: ‘emergency’ provisions, not unusual in other types of lease, did not substantially interfere with enjoyment of the sharer’s use of the property. There were fixed terms with reasonable notice periods. In practice the host did not have access to the cabin(et); the sharer had a key (c.f. Brador at 22); and it was unusual for the sharer to be required to relocate. The actual possession in practice, rather than the names given to the rights, was relevant: LNER v Assessor for Public Undertakings per, L. Fleming at 759. There was a finite amount of space on masts. There were arrangements to prevent electrical interference between operators. Further, the ‘Code’ rights underlined the permanency, an operator asked to leave being able to remain until the Code allowed his removal (Schedule 2 to Telecommunications Act 1984, paras 1-7, 9, 21). The occupier’s or landlord’s rights could be adequately compensated by money. Heritability depended on entitlement to be present for more than the whim of the landlord. The agreements were not consistent in using language appropriate to a mere ‘licence’ (e.g. H3G 6, 7, T-K Appx 23 – Orange/O2 Ratecard). The effect of the agreements was that the sharer appellants had leases, or subleases at the very least, to use land, and had rateable occupation of the cabin(et)s, cables and the right to use the mast.
 The pertinent right to use the mast had, said Mr Smith, been recognised by the majority of the court in ITA. The mast should not be viewed in isolation but was an essential part of the television station (L. Avonside at 259-60). See also L. Thomson at 263 and L.Fraser at 256-258, the latter’s dissent clearly having been influenced by an issue of double counting which occurred because of the ‘supplementary roll’ anomaly that the value of the subjects in the ordinary roll could not be altered. In Pye, the sharers did not have rateable occupation of any land or buildings and ITA was distinguished on that basis. The aerials and cables only became heritable by accession and were attached to the mast in accordance with the right to use the mast, which right was the only basis upon which the cables became heritable. So the right was not a ‘pertinent’ of anything heritable. (L. Fraser at 165-6). The aerials and cables were heritable by accession, but the radio equipment in the hut was moveable, i.e. the sharers had paramount occupation of the cables and aerials despite Pye’s extensive rights over the facility. Critically in the present case, the sharers have what can be regarded as leases relating to the cabin(et)s, so that the right to share the mast can be considered a pertinent of those heritable subjects. Also, the cables are heritable by statute and do not depend on accession for heritability. The distinguishing features in Pye, it was submitted, were absent. A pertinent did not have itself to be heritable, merely to pertain to something heritable. Mr Smith suggested that what pertains could be in the nature of a wayleave. Use of the mast, as well as the aerials and cabin(et)s, were essential to enjoyment of the heritable feature, as in ITA, by contrast with the right to use the canal in Lowland Leisure – see per, L. Clyde at 357A-B, L. Prosser at 359B and L. Coulsfield at 360-1. On the approach in the latter case, the right to use the mast can properly be regarded as a pertinent right. Without the pertinent right, the ratepayer’s occupation of the subjects would be valueless. The sharers all had their independent occupation of bits of the mast. There a finite number of sharers and a certain amount of exclusivity, all unlike, e.g. the aerodrome which was ‘open to all comers’. It did not matter that only a small amount of land was involved – David Allen.
 Mr Smith then considered wayleaves, which, he said were recognised as being rateable: Armour 8-11, 21-23. A “personal right of way” might be classed as heritable in this context. In one case, the Assessor’s argument that pits were of great value as affording the only means by which minerals from other mines could be brought to the surface, giving a value to those other mines as wayleaves, was sustained. The possibility of wayleaves in certain cases had been recognised in both Pye and Lowland Leisure, fortifying L. Avonside’s view in ITA at 259. Further, Para 2(c) in the 1995 Order included “servitudes or wayleaves”. ‘Code’ rights, e.g. under para 2(1), had been referred to as “wayleaves”: Cabletel v Brookwood at para 44. This gave the right a strong degree of permanence. L. Fraser in ITA was considering common law, not statutory, wayleave. This right could properly be described as a wayleave for rating purposes.
 Mr Smith submitted that the Assessor had established in evidence that telecoms subjects, including ‘host’ subjects, could be properly valued by a comparative assessment of share payments which were in respect of the critical right to use the subjects. Other ratepayers had been valued on that basis (although there was dispute about the basis of the Arqiva assessments, which could be rationalised either way). Mr T-K had pointed out that the closest evidence to a hypothetical tenancy was the letting of an equipped site, and his analysis demonstrated that share payments were the closest way of doing that. From the limited evidence of equipped sites the rental would be £8,500 per annum compared with the agreed share payment of £7,450. The discovery that the Old Inns, Cumbernauld site was in fact equipped was important. There was a correlation between rooftop and share transactions since in both the tenant did not have the capital outlay of building a support structure, and there was a good basket of rooftop site rents which correlated well with share payments -£7,100 (scenario 4 – rooftop host) and £7,450 (Scenario 1 – share payment). Against Mr Sharp’s criticism that the sharer/host was a ‘business’ one, it was the use of heritage which attracted the value, with factors such as the location or the type of equipment taken into account in assessing rental. Variations on account of operator requirements would be ironed out over a network. There was nothing unusual in the rent depending on the nature and extent of the use as opposed to the extent of space, and the amount of space taken would anyway incrementally affect the rent. Reviews to RPI did not render the share payment figures unreliable, OMV having increased by more over the relevant period. Operators could go elsewhere if ratecard regimes did not reflect market value. If share payments had been reduced to take account of capital payments by the tenant, that would tend to reduce the analysed rentals, so any error would work in favour of the ratepayers.
 Mr Smith did not accept that there was double counting. He distinguished the situation in ITA, which depended on the findings as to the “true”, or “whole beneficial” value of the mast. Here, it would be unrealistic to ignore the turnover of the mast. Where sharers existed, they should be taken into account. Provision of shared sites was an industry in itself. The high turnovers which often resulted from multiple sharers had been graphically illustrated. The subject should be valued as it stands and as used and occupied: Great Western and Metropolitan Railways v Kensington Assessment Committee, at 54. Further, on the basis that the rental value of an equipped site in the hands of an operator was unlikely to be less than the share payment, use of the share payment to value the host site was appropriate, even where the mast was not in fact shared. The share payment established the value for the sharer’s use of the subject and for the occupying host. On the Assessor’s approach, nothing was added to the occupying host’s value for the fact that there may also be sharers.
 Mr Smith accepted that it might be necessary to adjust the share payments to bring them into the statutory hypothetical terms. However, the only evidence of potentially relevant outgoings came from H3G (Assr 9.6), and their relevance was doubtful. Some expenses would have been incurred by the host in its own interests and regardless of the presence of a sharer. Other costs, while arguably of benefit to the sharer were not items normally passed down to a sub-tenant and would not constitute “a normal obligation of a landlord under a lease” - c.f. Armour 19-15. Even if the items were relevant, the facts did not entirely support the appellants. Most sharers obtained their own electricity supply. Rent review savings had to be offset by ratecard negotiations, and enabling agreements had to be set up. On a worst case scenario, the deductions totalled £827, which required to be divided by three for the average number of sharers per shared site (as with the mast value in Scenario 4), i.e. £276. Then the figures had to be indexed back (by RPI, appropriate for these types of costs) to 2003 (£220). That would reduce £7,450 in Scenario 1 to £7,230 (c.f. rooftop hosts in Scenario 4 - £7,100). There was no exact science in this, but the appellants had only themselves to blame for any unfairness for providing such sparse information on matters within their knowledge: Whitbread plc v Assessor for Lothian at 376D-G; Associated Portland Cement v Assessor for Dundee at 271. The Assessor had called for that information.
 Mr Smith then addressed the appellants’ alternative argument for a deduction of £1900 on account of “additional site management services” and “avoided costs”. The starting point of the difference between a rooftop share of £9,000 and a rooftop site rental of £7,100 had turned out to be severely flawed, a basket of only 18 sites having been reduced to 12 with an average share fee of £10,100, producing figures well beyond the figures in Assr 9.6 and the largest ‘srevice’ being the inspection of masts of which there none on rooftops. A basket of 12 sites was too small and unreliable; and there were particularities in H3G’s situation and in relation to BT rooftops, which would be inherently attractive. Also, some direct site rents for BT rooftops were in fact more than the H3G payments. The attempt to test the deduction by taking £1900 plus the ‘RV of a typical 15m tower’ (£900) off a greenfield site share fee depended on two unreliable assumptions. A further ‘cross-check’ from the Clifton Hill Stadium was also problematic, as that site should be seen as a rooftop and the managing director himself had negotiated the site rent. Reference was made to the definition of ‘market value’ implying knowledge of the market (IRC v Clay at 472 ; Glass v IRC at 465): comparisons in which one party was unrepresented should be given limited weight: this applied both to Mr Sharp taking one particular figure and also in the case of greenfield rents, where there was a significant proportion of unrepresented landlords.
 Mr Smith then turned from the sharer payment basis to the site rental, or hybrid, approach, which in his submission should be rejected. It was relatively more dependent on the contractor’s principle as it required the establishment of site development costs. The proportion of costs to the site rent was, in the ‘real world’, a figure closer to 40%, a significant part of the total, than Mr Sharp’s 80%. The contractor’s method was a ‘last resort’ method. C.f. Civil Aviation Authority v Assessor for Orkney and Shetland, per, L.Clyde at 148, and Pye at 169, 171. Moreover, it could be affirmed that reliance on the contractor’s principle produced unsatisfactory results because of the use of the VOA cost guide figures, on the evidence of Mr T-K and the MFD International figures. MFD were mast specialists. The cost guide very significantly underestimated the cost of construction of a mast, the actual costs of site works, foundations and fencing being apparently consistently underestimated by more than £30,000 on the basis of returns to the Assessor (Assr 5) and £27,000 (MFD), figures suggesting that Mr T-K’s actual cost examples were not isolated worst cases. The guide made a best case assumption for electrical supplies, which could be very expensive. Mr T-K had suggested that build costs could differ across the UK, but limited information had come from Scotland for the preparation of the guide. Further, the statutory decapitalisation rate would not be the rate used in the market and a higher rate, perhaps 9%, would be used. The fact that a hypothetical landlord would want a higher return tended to coincide with Mr MacTaggart’s ‘convenience factor’, the uplift a sharer was prepared to pay for the convenience of going into an existing site. The cost guide figures had been shown to be wrong or at least highly suspect. The appellants could not avail themselves of dicta in Shell UK . Further, the starting point in a contractor’s approach was actual, not estimated, costs: BP v Assessor for Lothian. The appellants criticised the Assessor’s attempts to produce actual costs but were themselves the parties who would be expected to have the original information and had not submitted any figures even as a check. The Assessor’s exercise in submitting evidence of actual costs should be given weight. Mr MacTaggart had explained the limited circumstances in which the Assessors had used the cost guide in a hybrid approach, in the case of broadcast sites, particularly large ones, where the buildings tended to be somewhat varied and exceptional and where the valuation had also involved looking at sharer payments.
 Asked whether the Assessor suggested any particular amount by which the appellants’ cost figures were understated, Mr Smith, referring to Mr T-K’s figures alongside the highest agreed figures (Scenario 3) said it was “more than twice”.
 Mr Smith said that any result of the contractor’s principle should be cross checked on account of its “notorious unreliability” (Civil Aviation Authority v Assessor for Strathclyde, per, L. Clyde at 390E). The appellants had not done so. Their approach did not cross check with the prime evidence of the market in equipped sites. Nor did the hybrid approach begin to coincide with a rooftop host figure of £7,400, a major anomaly given the relationship between equipped sites and rooftop sites. A hybrid approach using actual costs would produce markedly higher figures than the appellants’.
 In relation to streetworks, Mr Smith pointed out that it was necessary to look at market comparisons which were not influenced by the Code right to instal such equipment without charge. The hybrid method was unsatisfactory for this further reason. There were only 7 cases of rents, where a landowner had had claimed that equipment had been placed on private land: the figures might include an element of settling a disputed claim. Sharer payments, however, could appropriately be used.
 An additional point arising in relation to sharers’ entries under Scenarios 3 and 4, where use of the mast was valued at “1/3 mast value” (£200, within Plant and Machinery figures). That did not truly reflect the value of the right to share the use of the mast. It was a problem of the contractor’s method that a sum had to be added for wayleaves which were in addition to other physical property – c.f. Edinburgh Parish Council. Again, the true value was best shown by the share payment. This was another reason for not adopting the hybrid approach.
 Mr Smith submitted that in Scenarios 5 and 6, with the value of sharers’ rights excluded, the uplift to the host value required to be taken into account by adding not less than the payaway amount to the host’s entry.
 Finally, the recent discovery that payaway ‘recharge’ had not been declared by Vodafone, with Mr T-K’s evidence that the practice might be more widespread, meant that sharers’ payments might be 130% of the declared figure and suggested the possibility that the agreed figures underestimated the true scale of the share payments.
 These appeals raise two substantial major issues and a number of minor ones. We think it appropriate to mention some preliminary matters and thereafter consider matters in the following order: unum quid or cumulo; whether ‘sharers’ rights’ heritable- the first major issue; method of valuation – the second major issue; addition of ‘Payaway’ to host values; deduction from ‘sharer payments’; double counting; and plant and machinery values at rooftop ‘sharer’ sites.
 We are much obliged to parties and their representatives for the extent of preliminary work done to achieve agreements on very nearly all the various values arising under the 6 possible scenarios. Although we have nevertheless required to set out the experts’ competing opinions on various valuation issues, this has made our task easier.
 However, we do require to draw attention to the appellant companies’ failures to provide information on a number of matters. Generally, while we have not enquired fully into the rights and wrongs of this matter, it appears that the companies’ reluctance over very long periods to provide information reasonably sought by Scottish Assessors made the latters’ tasks in this difficult area much more difficult than it need have been. It appears to have been the companies themselves, rather than their agents, who were refusing to provide information. They appear to have spent a long time obstructing the Assessor’s legitimate attempts to uncover the nature of agreements relating to use of the masts. Their attitude appears to us to have extended far beyond the reasonable concern for commercial confidentiality which they no doubt had in relation to some information. We express the hope that a more co-operative attitude will prevail in the future.
 More specifically, in relation to particular points which remain in issue at the hearing, we find it surprising, to say the least, that it is suggested that the appellants were unable to access any costs information. We also noted their almost complete failure to provide vouching of the deductions sought from sharer payments, if that basis of valuation was used, in order to meet the statutory hypothesis.
 We found the evidence of both Mr Sharp and Mr MacTaggart generally clear and reliable. In particular, we entirely accept Mr MacTaggart’s explanation of the Assessor’s substantial changes in approach in this revaluation. He is entitled to take a fresh approach and we attach no weight to the fact, in itself, that this is a new approach. The mere fact that that approach would apparently produce a substantial divergence between Scotland and the rest of the UK in these matters, while it might not seem very desirable, also has no weight. There might be room for concern that as this new approach would not apply to all radio transmission masts it might seem to be introducing a new valuation category, but our principal task is of course to address the correct valuation of the subjects of appeal.
 We must, however, express a degree of reservation about the complete reliability of Mr Thornton-Kemsley’s evidence. He is clearly a very knowledgeable and experienced expert, who included voluminous information in his Report, but on a number of occasions his definitive assertions were not fully borne out on closer scrutiny. We were not convinced of the reliability of the enormous database on which he drew and which seems to have included a certain amount of almost anecdotal information. He seemed to us too ready with opinions, including in the area of valuation for rating in which, as he himself noted at the outset of his report, he does not have experience and was asked not to opine about. We would have had difficulty in accepting many of the specific value figures to which he spoke, although we have carefully considered and appreciate the force of some of his more general points, taken along with Mr MacTaggart’s evidence.
 This apparently fundamental issue turns out not to have a major bearing on at least the overall result of these appeals in Scotland, and was the subject of relatively brief submissions on each of the two occasions on which it was aired before us.
 There is no doubt that the subjects of these appeals (to the extent that they are found to be ‘lands and heritages’) fall under Article 2 of the 1995 Order, being subjects of the specified description which “would (apart from this Order) be treated as justifying separate entries in the valuation roll”, are to be “treated as justifying only one such entry” for each company in each valuation area”. The enabling power is in Section 6A of the Valuation and Rating Act 1956 and is similarly expressed. The Assessor places reliance for his interpretation on the Explanatory Note to the Order, including in particular:-
“This Order makes provision for the treatment as a single valuation unit for non-domestic rating purposes of certain property in Scotland which would otherwise be treated as several such units.”
We of course accept, as Mr Haddow did, the guidance of Lord Hope of Craighead in Coventry and Solihull Waste Disposal Co Limited v Russell (VO), at 2103:-
“In my opinion an explanatory note may be referred to as an aid to construction where the statutory instrument to which it is attached is ambiguous.”
 Neither counsel looked for assistance to the wording of any other Scottish regulation in relation to the valuation of utilities or utility undertakings, other than the similarly worded Article 3 of the 1995 Order in relation to canal subjects. Mr Smith contended that the Assessor’s interpretation was consistent with the position in England and Wales under Article 4 of the Non-Domestic Rating (Railways, Telecommunications and Canals) Regulations 1994, but the wording of that provision is not the same and we heard about a more complicated (in some ways) system of entries of mobile telecoms subjects there. We do not feel that the wording of any other provision helps in the interpretation of Article 2.
 This Order clearly does not prescribe ‘formula’ valuation. Indeed it replaces it. It also differs from the more modern provisions in certain later orders which, while using similar wording, place responsibility for some types of utility subjects (including fixed line telecoms subjects), on one particular Assessor. The idea of a part network of some utilities being a single unit in each valuation area, while others, the subjects of different regulations, would be a single unit across Scotland, seems inconsistent in comparison with, on the one hand, cumulo entries of single subjects as a matter of convenience and, on the other, formula valuation on a national (in practice, UK) basis.
 Be all that as it may, in our opinion the wording of Article 2 is unambiguous and refers to the administrative matter of entries rather than the substantive matter of valuation. It provides for one entry, not valuation as one unit. It is the single entry which is ‘justified’, and not anything more.
 If it were necessary or permissible to refer to the Explanatory Note, that would not in our view point, or certainly not point clearly, to unum quid valuation, when it is remembered that the reference to ‘single valuation unit’ refers to “only one such entry”, i.e. one entry “in the valuation roll”. A ‘value’ is simply a figure, which might be arrived at in either way, in the roll, and ‘valuation’, in the context of entry of a value in the roll, is similar. So the Order, using the language of the Explanatory Note, “makes provision for the treatment” of this property by putting only one entry in the roll.
 The rights in issue here are the rights granted to mobile telecoms operators to attach their aerials to masts or other structures occupied by other parties who have taken ‘ordinary’ leases of land or parts of buildings and (usually) erected the masts or other structures. The series of agreements under which these rights are conferred typically also include rights (which are agreed to be heritable and rateable) to attach cables connecting the aerials to the operators’ wireless equipment and also rights to erect and occupy cabin(et)s housing that equipment. In a small number of cases, although the agreements are apparently similar in relation to sharers’ rights the sharer does not erect his own cabin but shares space for his equipment within the host’s building so that he is not in rateable occupation of any building on the ground. It is not suggested, as far as this revaluation is concerned, that either the aerials or the wireless equipment are rateable. The cables and the cabin(et)s are agreed to be heritable without reference to the Plant and Machinery Order, although Mr Smith does, as we understand it, place some reliance on the fact that the cables are also ‘lands and heritages’ under that Order .
 We refer to our findings about the series of agreements involved. Although there is a variety of agreements, both national and site-specific, among the various companies involved, parties are agreed that this issue should be resolved in the same way in all “sharer” cases, except that the Assessor does not contend that sharers’ rights are heritable or rateable in the small number of cases in which the operator does not take his own cabin(et). We were also not asked to resolve any differences of opinion as to whether any particular subjects were ‘host’ or ‘sharer’ subjects.
 The Lands Valuation Appeal Court considered issues like this in 5 Scottish ‘mast’ cases between 1967 and 1977, as well as in cases involving other types of subjects.
 The unsuccessful attempt in the first such case, Central Fire Area Joint Committee v Renfrewshire Assessor,to justify an entry of a right to attach aerial equipment to a mast apparently lacked any real foundation in fact, and little can in our view be taken from that decision. It is necessary, however, to consider the next two cases, ITA and Pye, at some length.
 In ITA, ITA owned and occupied a television transmitting station including the mast with a designed capacity to carry three aerials. It agreed with the other company, BBC, to ‘demise’ firstly some of the land, secondly the right to use in common with it and anyone else the mast, access road, utilities, etc., and thirdly the right to lay and use cables, feeders, etc. ITA’s subjects were in the roll at what was found to be full value on a contractor’s valuation. ITA and BBC each had aerials, i.e. two-thirds of the mast’s capacity was used. The Assessor entered the BBC subjects in the roll as ‘TV Station and Wayleave’. The Assessor did not treat the BBC as being in exclusive occupation of any part of the mast, but did include the right to use the mast in his valuation. As the BBC entry was on the ‘Supplementary Roll’ (as the system then was), the ITA’s entry would remain in the roll at full value. It was held, Lord Fraser dissenting, that the right to use the mast was rateable and properly included in the BBC entry, although the majority did reduce this part of the valuation from the Assessor’s figure of 60% of the value of the mast to one third of that value. Lords Avonside and Thomson held that the right to share the use of the mast was a pertinent of the BBC’s land and buildings; and Lord Avonside also held that it was separately heritable as a wayleave. It is fair to say that the issue of ‘double assessment’, or double counting, was of concern both to Lord Fraser and the majority, although the latter accepted that the result they reached appeared to involve partial double assessment, which they said was unavoidable for the year in question under the supplementary roll procedure. It appears to be implicit in their consideration of the value that they considered that ITA as the occupiers of the mast should bear two-thirds of the value (unless and until a third party acquired rights similar to the BBC’s) and BBC one third. The value of the BBC’s land, buildings and plant (not in dispute) was £1376; of the cables (confirmed by the court as correctly entered) £573 and the right to share the mast reduced from £1686 to £937.
 Lord Avonside said, at page 259, that ITA let the subjects, including this right, to BBC, with the result that the Assessor found at Blackhill:-
“… buildings and land occupied by the BBC and a BBC cable attached to the mast and connected to an aerial at the top, by means of which transmission was made … It is not disputed that the BBC are in rateable occupation of lands and buildings on the site. Their right to share the use of the mast, which they exercise, can, I consider, be properly regarded as a pertinent of these lands and buildings, or, if it must be looked at in isolation, as itself a land and heritage, under the definition of the 1854 Act, which, it is recognised, has an extensive and broad application and, for example, covers a variety of rights which in other aspects of the law would be considered personal in nature. Looked at in another way, I see no objection to describing the right as a “wayleave” and as such of rateable value.”
 Lord Thomson said, at page 263:-
“I think there is much force in the assessor’s contention that in valuing the subjects occupied by them he must have regard to their right to hang an aerial on the mast, a right without which the occupation of the subjects would be wholly sterile. The whole and sole purpose of the mast is to carry aerials and, that being so, I doubt if BBC’s right can properly be described as a wayleave. I think it is better regarded as a “pertinent” of the station occupied by them.”
 In Pye, decided by the same three judges, one company leased the ground and erected the mast and a hut. Apart from their own aerial, they allowed 10 other users, under a standard form of agreement, to hang aerials and lead them by cable to their transmitting equipment which was situated in the hut. Pye provided receiving equipment. On this occasion, the court did not accept that the rights of the other users, other than their actual occupation of their cables and aerials (aerials being then agreed to be rateable), should be reflected in the roll. Lord Fraser, at pages 165-6, pointed to “an obvious distinction” insofar as BBC had had a lease of land and buildings which were “unquestionably heritable” and:-
“their right to hang cables from the mast could naturally be regarded as a pertinent of those heritable subjects, whereas in the present case the only physical subjects which are heritable and occupied by the appellants are the aerials and cables, and they only become heritable by accession and because they are attached to the mast in accordance with the very rights which we are considering. In my opinion, the rights of the appellants other than Pye are not pertinents of their heritable aerials and cables but are on the contrary the basis upon which the aerials and cables become heritable. I would therefore exclude any reference to wayleaves or other rights from the roll.”
 It is perhaps fair to say that the opinions of Lords Avonside and Thomson in Pye, in relation to this issue, may not go beyond upholding the Committee’s reasoning, although Lord Avonside, at pages 168-9, might be taken as having withdrawn his earlier view that this right could amount to a separate wayleave, or at least not been prepared to apply it again.
 Two subsequent cases, Automobile Association v Assessor for Lanark and Scottish Gas v Assessor for Fife, provide further examples in which the appellants did not occupy any land and buildings so that it was wrong to describe the right to use the mast as a pertinent of heritable subjects. In the AA case, Lord Thomson added to that reasoning:-
“Nor can I accept as sound, Counsel for the Assessor’s submission that occupation by the Appellants of the aerials and cables involves occupation by them of part of the mast itself. Not only would this in the ordinary case result in double valuation for the whole mast is valued as in the rateable occupation of the owner, but in my judgment the exercise of the right to fix aerials and cables to the mast cannot result in rateable occupation of part of the mast as a separate subject of valuation.”
In these two cases as well as in Pye, the Assessor had valued the subjects, which he wished to describe as something more than simply an aerial and cables, by comparison with rents paid to the main operators who were the tenants and occupiers of the mast, i.e. in the terminology of this case, ‘hosts’, for shared use of the mast involving the right to attach cables. In Scottish Gas, the court made a number of observations, including the following:-
“The proper method to be adopted by an Assessor in valuing wireless installations similar to the subjects of appeal, has been laid down in a number of previous decisions of this Court. From these decisions it is clear that, in the normal case, there should be an entry for the “Unmanned Radio Station” showing the main operators as tenants and occupiers, and that, where the main operators allow others to attach aerials and cables to their mast, separate entries of “Transmitting Aerial and Cables” should be made in respect of the aerials and cables of each of the other users.
When dealing with the Appellants’ installations at the two sites in question in the Supplementary Valuation Roll for the year ending Whitsunday 1974, however, the Assessor has not confined himself to making an entry in respect of the transmitting aerials and cables attached to the South of Scotland Electricity Board’s masts and used by the Appellants. He has made entries in respect of each site for “Wireless Installation”, and has based his valuation on the rent which the Appellants pay to SSEB for “the common use of the road, the common use of the compound and the hut and the aerials and cables affixed to the mast”
At the end of the day, Mr Henderson’s argument depended on establishing that there were leases in favour of the Appellants. His contention was that the Minute of Agreement relative to the Knockhill site and the Offer to Let relative to the Downfield site were leases, and that the Assessor was well-founded in basing his valuations on the rents payable under these leases.
In our opinion, the argument for the Assessor is unsound. The Assessor should have adopted the method of valuation which was approved in the cases cited above … It is plain from the decisions cited above that such aerials and cables are heritable by accession to the masts and should enter the Roll … In our opinion, however, the Assessor was not entitled to take into account the “rent” payable by Scottish Gas in respect of the rights afforded to them under the Minute of Agreement and the Offer to Let. The rights conferred on Scottish Gas under these two documents are not rights of a heritable character at all. What SSEB purported to let to the Appellants under the Minute of Agreement was “accommodation at the Board’s Radio Station … for the installation of one VHF Transmitter/Receiver set and one UHF Transmitter/Receiver set, together with rights to the Gas Board to affix to the Board’s radio mast on the site the appropriate “aerial equipment”. Similar language appears in the Offer to Let. … The Minute of Agreement and the Offer to Let also confer on the Appellants right of access to the mast, but the compound, hut and mast are all in the occupation of SSEB, and such rights as the Appellants were given do not constitute heritable subjects in the occupation of the Appellants and should not have been taken into account by the Assessor in valuing the Appellants’ installations. The situation is different from that which existed in ITA, where the right to share in the use of the mast was regarded as a pertinent of the lands and buildings occupied by the ratepayer. In the present case, there are no lands and buildings occupied by the appellants of which the rights conferred upon them by the Minute of Agreement and Offer to Let could be the pertinents.
Counsel for the Assessor laid stress upon the language used in the Minute of Agreement and Offer to Let, which includes the words “let” and “rent” which would certainly suggest a lease but in this context “Regard is to be had rather to the substance of the rights conferred than to the names given to these rights” (LNER v Assessor for Public Undertakings 1938 SC 751, per Lord Fleming at 759). The rights conferred upon the Appellants under the Minute of Agreement and Offer to Let did not give the Appellants any exclusive right to occupy any specific part of the road, compound, hut or mast. The road, compound, hut and mast were all in the occupation of SSEB, and the use of the facilities conferred on the Appellants did not constitute a heritable right. What was conferred on the Appellants was a licence to make use of certain facilities, and such rights should not have entered the Valuation Roll and should not have been taken into account when valuing the aerials and cables.”
It is clear from that last paragraph, as well as other authorities to which we were referred, that the terminology used in these agreements cannot be decisive. We require to look at the substance of the rights granted. To the extent that Mr Haddow placed particular reliance on use in the agreements of expressions such as ‘licence’, we agree with Mr Smith that this cannot be decisive.
 In several of these cases, consideration of the differing methods of valuation – contractor’s, based on the cost of, in particular, the mast, or comparative, based on ‘sharer’s payments’ - in a sense helps to focus the issue as between corporeal and incorporeal rights. At all events, we think it must be taken from these cases that the sharer’s rights, viewed in isolation, do not themselves involve any rateable occupation of the mast. They do not comprise a separate incorporeal right which is itself heritable or rateable. Lord Avonside’s minority view in ITA that they could amount to a wayleave has not found favour and the balance of authority appears to be against it. The right to share use of the mast appears wider than a right to lead pipes, cables or whatever across land or subjects of a different character. (See also Assessor for Lothian v Lowland Leisure Ltd, per, L. Clyde at 357B-C, L. Prosser at 358K-L and L. Coulsfield at 359L-360A.) We do not consider that the reference to “wayleaves” either in the 1995 Order, or in the course of consideration of the “Code” rights, adds to that argument.
 On the other hand, the absence, except in ITA, of land and buildings in the rateable occupation of the ‘sharer’ was clearly important and arguably the ratio of all the decisions. The right to share use of the mast amounted to a ‘pertinent’, at least under the 1854 Act, in the one case in which it was enjoyed along with rateable occupation of other land and buildings. So while there is a degree of similarity between the Assessor’s approach of using the comparative principle to bring sharer’s rights into valuation, an approach seemingly rejected in the normal type of case in the series of decisions, including Pye, referred to, in the present case the Assessor points to this distinction and argues that the rights are rateable as pertinents.
 Is it, however, sufficient to identify buildings in the separate occupation of the sharer in terms of the agreements, or does one still have to consider the true nature of the right?
 It seems to us that the other decisions highlighted by Mr Haddow clearly point to the need, even where the ratepayers are in occupation of land or buildings, still to ask whether the rights under consideration in the particular case are really pertinents of the land and buildings. A monopoly right conferred along with what was held to be a lease of station bookstalls was not a heritable right, part of the heritable subject or a pertinent of the bookstall, and the value of that right could not be included in the value of the bookstall (John Menzies Ltd v Glasgow Assessor). Annual payments by a company carrying on a reserve training school for pilots for the use, along with others who wished to use it, of an airport landing-ground and for exclusive occupation of parts of the terminal building and hangars, were not rents conclusive of the annual value: it was impossible to regard the landing ground as accessory to the other subjects:-
“But I have come to the conclusion that the appellants’ right to use these premises is merely an accessory of their right to use the landing ground, and that it must be treated as being of the same legal character.” (Perth Magistrates v Assessor for Perth & Kinross, per, L. Fleming at 559).
Mr Haddow also referred to Armour at 19-03 for the propositions that not every regular payment under an agreement relating to lands and heritages can properly be looked at as evidence of the annual value of the subjects, and whether in any particular case an agreement stipulates for rent is a matter of the construction of the particular agreement.
 In Assessor for Lothian v Lowland Leisure Ltd, the canal board and the operator of a restaurant barge had an agreement giving the operator exclusive possession of a jetty and surrounding land and also a non-exclusive permission to operate restaurant barges on the canal. The Assessor argued, founding on ITA, that either the right to operate the barges was a land and heritage in its own right or, in any event, it was a pertinent of the jetty and surrounding land. Both arguments were rejected. The court was careful to distinguish the circumstances of ITA, but some of their general observations on the pertinents argument appear to us instructive. Lord Clyde said at 358F:-
“The mere fact that both grants are contained in the same agreement does not seem to me sufficient to make the one a pertinent of the other. Further so far as the terms of the trading agreement are concerned it is not evident to me that the permission to operate the boat is accessory to the use of the land. On the contrary it might well be thought that in so far as the two grants are related the right which is given to use the land for embarking or disembarking passengers or cargo is accessory to the right to operate the vessel.”
 Lord Prosser said at 359D-E:-
“There is no doubt a wide variety of rights which may be seen as belonging or pertaining to heritable subjects. In addition to what are normally called “pertinents” which are carried by a disposition of named or described subjects without being mentioned, it may well be that the word can also be used to describe rights which are less inherently part of the subjects, and have been acquired as identifiably separate adjuncts thereto. In such cases, it does not appear to me that exclusive occupation of the main subjects will entail exclusive occupation of the adjuncts, and quite apart from the fundamental question of whether such rights belong with the land, rather than belonging to its occupier, there is thus in my opinion a special problem in relation to such pertinents, which would not arise in relation to rights which pertain by implication to the subjects.
A lease between landlord and tenant may well include other personal rights granted by the former to the latter, and I see no jurisprudential need, or indeed a need in common sense, for a merging of these personal rights with the rights relating to the property.
Even if an adjoined right plainly ran with the lands, and could thus properly be seen as an explicit pertinent of those lands, there remains the issue of exclusive occupancy which I have mentioned. I do not wish, on the basis of the present case, to express any general and concluded view as to the occurrence of such adjoined pertinents with occupancy which is not exclusive. But it seems to me that any such adjoined pertinents must be heritable rights capable of occupancy.”
His Lordship went on to decry use of the word “occupancy” in “a somewhat metaphorical sense”.
 Lord Coulsfield doubted whether the right to fix an aerial to the mast in ITA could be a pertinent in the conveyancing sense, and continued, at 360K-361A:-
“In that case, there was considerable divergence of opinion among the members of the court and I would hesitate to extract from it any general rule as to what rights may be regarded as pertinents for valuation purposes. Even if, however, the case must be taken as indicating that in valuation the word may bear an extended sense and that rights which are not pertinents in the conveyancing sense may form part of heritable subjects for the purposes of valuation, it does not, in my view, follow that all rights which the occupier of subjects enjoys and which enhance the value of his occupation of those subjects can be regarded as pertinents, even if these rights are enjoyed under the lease or other agreement by virtue of which he occupies the principal subjects. Lord Thomson’s justification for regarding the right to fix an aerial to the mast as a pertinent was that without the use of the mast, the BBC’s occupation of the TV station would be “wholly sterile” and Lord Avonside regarded the use of the mast as essential to the station. In view of that, the ITA case was in my view exceptional and the facts of the present case are not comparable with those which were held to justify the application of the description ‘pertinent’ in the ITA case.”
 We take from these cases that the mere fact that the sharer’s agreements result in rateable occupation of land and buildings as well as conferring the right to share use of the mast is not necessarily sufficient. It does indeed seem odd that this issue should turn on whether the sharer takes a cabin(et) of his own or puts his equipment on a rack in the host’s building.
 We also have a distinct impression that the agreements here are similar to those characterised by the court in Scottish Gas as in the nature of licences rather than leases and rather different from BBC’s agreement. On the other hand, the right to share the use of the mast may, as in ITA, be seen as essential to the enjoyment of the cabin(et)s, which would, as Lord Thomson put it, be sterile without it. However, we cannot decide this issue without a deeper look at the ‘occupation’ identified in ITA and asking ourselves whether or not the position in these cases is materially similar.
 We accept that the sharers’ rights are of sufficient permanency for rateable occupation. That is not the issue. The question is really about the nature of possession of heritable subjects.
 It seems to us that the puzzle about ITA is that while the Assessor did not regard BBC as having exclusive occupation of any part of the mast, the majority who found for a pertinent (and perhaps also Lord Fraser) were actually considering the mast of which the BBC had shared use as a part of their broadcasting station. Lord Avonside said at 259:-
“As to the mast, the appellants assert that, since it is agreed that the BBC were not regarded as being in exclusive occupation of any part of its structure, there existed no rateable occupation. In my opinion this contention is ill-founded and depends for any attraction it may have on regarding the mast as an isolated and independent entity. But what is being valued is a television station, of which the mast is only part, although no doubt an essential part … That the use of the mast is shared between the ITA and the BBC seems to me to be of no moment. No doubt a common entrance or passage shared by two houses is not a subject of separate valuation in rating one house. It is, however, illogical to argue from that either that it has no value or that the right to share the use of a heritable subject cannot be rated.”
Lord Thomson, at 263, accepted that the result was ‘double valuation’ in the sense that the mast had already been entered at full value, which also seems to imply that BBC were sharing actual occupation of the mast. So there appears to have been this physical element in the rateable occupation, as opposed to the “metaphorical” occupation referred to by Lord Prosser, making it possible to characterise the right as a heritable pertinent even if not in a strict conveyancing sense. Is the position the same in the present case, as the Assessor argues?
 It seems to us that, while there are conceptual similarities, there are physical differences and also significant differences in the nature of the agreements. The conceptual similarities are, firstly, that once the site-specific agreement has taken effect, there is an identified location, or locations, at which the sharer may locate an aerial, or aerials, of the specified kind. In this respect, we acknowledge that the rights are more closely related to the heritable subjects than were the rights to the airport landing ground, the railway station or the canal. Secondly, it is clear, as Mr Smith submitted, that the cabin(et)s have the character of buildings.
 Physically, however, the ITA mast had an 80 foot slot designed for the BBC aerial and a definite capacity, enabling it to be said that the BBC were being given a particular share of the mast and thus a degree of control over the extent of use although not over the identity of other sharers. Although the fact that only a small amount of the mast is involved does not matter (David Allen), attaching various shapes and sizes of aerials, and cables, to different positions on the mast, with the host able to invite others to do likewise, is less easy to see in terms of exclusive occupation. As regards the subjects to which the right is said to be pertinent, the comparison with ITA appears to be between a TV broadcasting station in a substantial building and a small equipment storage hut or cabin. Unfortunately, the full extent of the BBC building is not set out in the findings in the law report, but the value, £1376 at 1966 values, compares with £500 for the larger cabins at 2003 values. The building was around one half of the value of a 300 metre mast, compared with not more than a tenth of the value of standard, say 15 to 30 metre, masts (including the land) in the current case. The BBC aerial and cables were maintained by BBC engineers (although we are not told whether, as one might infer, they had a regular presence at the building). The view that a specified one third share of the mast was a part of a substantial subject on the ground, and therefore a pertinent of it, seems more understandable.
 One ground on which Mr Smith distinguished Pye was that the cables in that case (to which the appellants wished the sharing rights to be pertinent) were rateable only because they were heritable by accession to the mast, whereas the cables now are rateable on a statutory basis. However, if that made a difference, the Assessor might have been expected to include sharer’s rights in the ‘Cable only’ cases. It also seems to us that Lord Fraser was going further and was identifying that the right to attach an aerial and cables was the whole basis of occupation of the cables, i.e. the point has the same significance even if it is not necessary to rely on heritability by accession. Like the premises at the airport and the jetty and ground beside the canal, occupation of the cables stemmed from the right to use the mast, so that the latter was not a pertinent of the former. Similar reasoning seems to us to apply to the equipment storage cabin(et)s; the cabin(et)s are really accessory to the right to use the mast. Put another way, each may be said to be essential to each other, but that does not make each a pertinent of the other. Again, viewing the BBC TV broadcasting station differently seems understandable.
 As far as comparison of the agreements is concerned, the ITA agreement, following a preamble which refers to sharing the use and occupation of the TV transmission station, ‘demises’ firstly, in conventional lease terms, the land and buildings, secondly, the right to use in common various specified items including the mast, and, thirdly, the right to lay and use cables, etc. through ITA’s land. There is no indication that ITA are to provide additional services. The sharer agreements in this case are, as we have seen, part of national arrangements allowing the parties from time to time to enter into site-specific agreements, with payments determined under a national rate-card, primarily by the value of the non-rateable aerials, subject to changes from time to time in the equipment allowed and its precise location. The right to locate aerials on the mast is generally stated first and followed by rights (including the right to erect a cabin(et)) which seem really subsidiary.
 We have come to the conclusion that, notwithstanding the sharers’ occupation of land and buildings, their rights to use the mast fall short of heritable rights and are to be seen as personal rights which cannot be characterised as pertinents of the cabin(et)s. The sharers seem to us to be on the same side of the line as in Pye, Scottish Gas and the other similar cases. Their rights are not in the same position as those of the BBC in ITA. They should not be included in the values of the sharers’ subjects.
 There is a degree of overlap between the issue whether the sharers’ rights are heritable and the issue between the appellants’ ‘site rental’ method of valuation and the Assessor’s ‘sharer payment’ basis. As we have seen, some of the court’s consideration in the earlier cases reflects that. It might seem that once it is determined that the sharers’ rights are not heritable, the sharer payments might not be an appropriate measure of the hypothetical rental value of the ‘host’ sites. Mr MacTaggart, however, confirmed in evidence that he would still seek to base host site values on sharers’ payments under Scenario 1 or Scenario 2, because, he argues, they are a better measure of the rental value of sites equipped with masts. We did not understand the appellants to suggest any legal or other reason why this position could not be maintained, and in our view Mr MacTaggart is entitled to do so. As Mr Haddow put it, this is not an issue of law but purely one of valuation.
 That said, in this chapter the nature of the sharers’ payments, and whether they can truly be seen as a reliable measure of rent under a hypothetical tenancy of the lands and heritages, i.e. the sites equipped with masts, requires to be considered.
 If the Assessor’s preferred method is accepted, there is no separate valuation of the mast, but there are small additional agreed values for other rateable plant and machinery, including the cabin(et). There was no scenario under which the method of valuation would differ as between host sites and (if rateable) sharer sites, except that if the Assessor, having won on rateability, lost on this issue (Scenarios 3 and 4) he would limit his valuation of the site rent of the sharer sites to the amount of the payaway (again with the addition of agreed values of plant and machinery). If he won on both issues, he would not so limit his valuations.
 The Assessor claims to have found a large body of evidence – the sharers’ payments - which he says shows the value put on occupation as a mobile telecoms operator of sites equipped with masts. This, he says, provides a body of comparative evidence which, in relation to these subjects, provides a better measure of value than the appellants’ method. The appellants’ method includes a significant ‘contractors’ principle’ element, that being the ‘principle of last resort’. The Assessor accepts in principle that while the analysed level of sharers’ payments is agreed, there might have to be some downward adjustment to reflect non-rateable items for which the sharers pay, although he disputes whether the appellants have established any, or certainly all, of the figure of £1900 claimed.
 The Assessor’s characterisation of the appellants’ approach as ‘hybrid’ is no doubt accurate, as it does involve adding a contractor’s value of the masts, plant and machinery to the comparative principle valuation of the site. It seems to us, however, that there is nothing particularly unusual or objectionable in principle about such a method. Many types of subjects are valued on the basis of rental evidence to which the decapitalised costs of rateable items not included in the rents is added, using the statutory decapitalisation percentage (which applies to values of parts as well as of whole subjects). The system of arriving at a standardised cost guide to be applied in such valuations, as well as in cases where there is no or insufficient rental evidence so that a ‘pure’ contractor’s principle has to be used, is well established. Classic criticisms of the contractor’s principle as a principle of ‘last resort’ were particularly directed at ‘pure’ rather than such hybrid valuations. These things said, we accept that it is appropriate for the valuer to consider whether, if a body of rental evidence does not reflect the full hypothetical rental value, there is reliable rental evidence which does include such additional items and therefore better reflects that value. Conversely, however, any such evidence which might go beyond or outside hypothetical rental value required close consideration.
 Before looking at the Assessor’s approach, we consider the site rental basis and the Assessor’s criticism of it.
 Mr T-K and the Assessor claim to have identified points at which the appellants’ values are in themselves unreliable and too low. Our consideration of the suggestion that the appellants’ values are made too low is made more difficult by the parties’ agreement on levels of site rents under Scenarios 3 to 6. We must remember that these are agreements for the purpose of deciding these appeals and not necessarily agreements on the true value levels. On the other hand, evidence that, for example, there has been analysis of hundreds of ‘Open Market Value’ rents, i.e. initial rents or rents where the stipulated review was to OMV, requires us to ask ourselves how much can be taken from Mr T-K’s and Mr MacTaggart’s opinions and the relatively considerably smaller number of rentals used to advance particular points. Further, it was not suggested that the 25 test case sites provided a representative sample, although they did help in illustrating the issues.
 On the material before us, we reject Mr T-K’s opinion that the analysed level of site rents is too low because of a degree of lack of knowledge or representation on the landlords’ side. The small number of individual cases, with some anecdotal evidence, fell far short of establishing such a position. We accept that market value implies knowledge of the market, and that such knowledge may be inferred from the fact of professional representation and cannot necessarily be assumed in its absence. In the case, however of rates derived from analyses of hundreds of rents, any argument that such rates do not, because of this factor, represent open market value would in our opinion require very much fuller information and reasoning. The one concrete figure which we were given – 8 out of 21 – does not in our view go anywhere close to establishing this point. Nor was there any suggestion of any figure by which the analysed rents were ‘below market’. We do accept that there may be a need to look more closely at this, as well as other particular aspects, when individual, or perhaps a small number of, rents are looked at for particular purposes.
 While all the plant and machinery values under the different scenarios are (with one very minor exception) agreed for the purposes of disposing of the appeals, the Assessor nevertheless led evidence that the values of the masts based on the VOA cost guide are under-stated. This is part of his case for preferring his approach which does not require separate valuation of the masts. This issue of site development costs is in a different position, because Mr T-K’s evidence is opposed, not by a large body of transparent evidence but by the appellants’ reliance, without any analysis or supporting information, on the V.O.A. cost guide figures. Mr T-K’s evidence – as he himself would acknowledge – is highly problematic in some respects, although it does include some cost information provided by ratepayers on earlier occasions. Although Mr Smith, when asked, suggested that the evidence showed mast values “more than twice” the cost guide figures, the Assessor did not claim any specific figure by which the appellants’ values of Plant and Machinery including the masts should be increased. Rather, we were asked to use this as another factor pointing away from the appellants’ valuation method.
 Mr Haddow derived some support from the Tribunal’s reasoning in Shell UK Exploration and Production Ltd v Grampian Assessor, at pages 347 to 350. That was another case in which an assessor challenged ratepayers’ reliance on a V.O.A. cost guide figure, but the circumstances were not the same and we should note that that dispute was presented on the basis that no issue of law or principle was involved. The approach taken, however, is worth noting:-
“It was plainly intended to be relied upon as a guide to unit costs. For the assessor it was stressed that it was no more than a guide …
It would of course be wrong to give the cost guide figures a definitive status. However, we are entitled to have regard to their agreed status as material disseminated by the Valuation Office Agency and the Scottish Assessors Association and prepared by representatives of these bodies for the very purpose for which the ratepayers seek to rely on them. It was not suggested that there had been any correction or any admission of error …
We consider that until the figures are shown to be wrong, they are entitled to be accepted as representing the best efforts of responsible experts to determine an accurate unit figure.”
Having considered whether the assessor had advanced a good reason for rejecting the cost guide figure, the Tribunal accepted that regard must in any event be had to all the evidence. If, however, the available evidence left doubt, the cost guide approach, which tended to promote consistency as between ratepayers, should be preferred. We find this a helpful approach. We consider that we must accept that the guide, having been produced by expert plant valuers applying their mind properly to the task, gives acceptable answers which can be applied consistently and fairly among the various subjects, unless and until it is demonstrated otherwise. We also bear in mind that some of the items are not exclusive to these subjects.
 Mr Smith submitted that the cost guide figures had been shown to be wrong, “or at least highly suspect”. In our opinion, firstly, the Assessor has shown one particular flaw in the cost guide approach, in relation to the supply of electricity, but has not demonstrated how much difference that would make in the present case; and, secondly, the evidence does leave a more general doubt as to whether the cost guide makes adequate provision for all the site development costs. We have some sympathy for the position which the Assessor found himself in, having sought cost figures from the appellants and obtained none. To our mind, however, the Assessor has not gone far enough, in areas other than electricity supply, in demonstrating where the cost guide is wrong. He might have produced more evidence of the extent of information made available to the V.O.A. by mast operators, or of exactly what development costs were included in the items listed in the guide extracts produced to us.
 Our very limited findings in fact on costs show the extent of findings we have felt able to make on Mr T-K’s evidence on this issue. Much of his evidence appeared to us anecdotal, incomplete or so far different in time as not really to assist. The uncertain accuracy or knowledge of what is included in figures quoted or referred to in a different context substantially discounts the suitability of that information for valuation purposes. There were very serious gaps and flaws in the cost information gathered by Mr T-K which in our view rendered it of very limited use at a Rating Revaluation. A much more extensive information base coupled with a detailed and vouched for analysis would be required.
 While we would have been more comfortable if the appellants had really engaged this evidence and supplied information which we have difficulty in believing is all no longer available if there was a will to locate it, we do think it dangerous to place much reliance on substantially later, or earlier, costs, despite some of this coming from earlier – very much earlier – returns by ratepayers. We could place little confidence, in the absence of more detail, in evidence of relocation costs. In relation to electricity supply, while we can imagine that this could at some locations be a costly item, we have no idea how often this would be required or how much effect it would have when averaged over all the sites, many of which might have the supply on site or at least close by. We also did not feel that the MFD estimates were adequately explained or tested.
 The information in the VOA guide is generally a more reliable basis for establishing cost information for the purpose of a General Revaluation for Rating. It has the added advantage of being prepared for this specific purpose and thus includes only rateable elements. Use of this guide ensures uniformity between different areas and operators. There was, however, enough to suggest that this matter might merit closer enquiry into the extent of material made available to V.O.A. and S.A.A., and also into actual costs incurred around the material date. The latter would require to be sufficiently detailed to consider whether and to what extent appropriate adjustment was required.
 At best for the Assessor, on the available evidence in this case, and following an appropriate approach to the cost guide, the cost figures relied on by the appellants are very slightly understated.
 We appreciate that the Assessor’s argument is not solely based on the intrinsic nature of the sharer payments, because he refers to additional ways of showing that the values arrived on this basis can safely be relied on. However, the sharer payments are the essential central component, and this makes it necessary to look first at the nature of these payments and their suitability for measuring the hypothetical rental value of equipped mast sites. If they are, as the Assessor contends, essentially payments for the same use of masts as hosts obtain, the argument that the hypothetical rent payable by the host will not be any less has a logical attraction, as does the suggestion that the difference between the appellants’ values and the Assessor’s values represents what the tenant will pay for the convenience of renting an equipped site.
 In our opinion, however, agreeing in general with the appellants’ position, there are very significant differences between the sharer payments and the rent which would be payable to the landowner for an equipped site. The sharer primarily pays the going rate for the right to hang aerials, according not to the particular location or the capacity of the mast to take the aerials but to the specification of his aerial, in accordance with en bloc nationally agreed rates negotiated often some considerable time before and updated by RPI. Some of the ‘test case’ sites, e.g. Lochwinnoch and Maxwell Road, Glasgow, illustrate very large variations in sharers’ payments, apparently based on different rates for different strengths of equipment, on the same mast. These variations seemed to be of a different order from variations in the site rentals. Of course these would average out, as any list of figures will, but that does not help to show that they are reliable indicators of the level of rent for the same right. The market in which these agreements are made is characterised by reciprocal agreements between companies, even if, as we accept, payments under non-reciprocal agreements analyse lower than reciprocal. The sharer does not have the tenant’s responsibility for the mast. The extent of additional services provided by the host as a result of retaining responsibility for the site and mast itself casts doubt on the use of these payments. Without being over-critical of the Assessor, who was dependent for any detail on this from the appellants themselves, we also have to note the uncertainty as to the value of these services as another factor against the reliability of the payments as a measure of rent under a hypothetical tenancy.
 The evidence does appear to show that sharers pay a premium over the rents paid by hosts (at least where the host has not taken on any sharers), although the position about that appears to us slightly more complicated than was suggested. The appellants’ values are agreed to be on the basis that the site rent does not reflect the potential for sharers, but a sharer’s payment only arises where there is a sharer! To take a straightforward case, if one host and one sharer are using one mast, the host’s rent is increased by the payaway, which in fact brings the parties’ valuations, at least in that situation, approximately into line even without adjustment to the statutory basis (Scenario 2 – Greenfield with Cabinet - £7,000, c.f Scenario 6 - £5,100, plus £1,900 if we agree with the Assessor that the payaway would be added to the host value – as will be seen, we do). Even, however, where there is a premium, as for example in the case of a host who has no sharers, in our view that premium might reflect the difference in responsibility undertaken for the site and mast, or the value to the sharer (as opposed to the cost to the host) of the additional services, or both. To say that it reflects the convenience of not having to erect one’s own mast (something which we must say we have not come across in our experience in orthodox valuations with a contractor’s element, for example retail subjects let as shells) is in these circumstances speculation which we do not think can properly be made.
 The Assessor sought support for the valuations brought out on his approach from two other lines of evidence. Firstly, Mr T-K spoke to three rents of equipped sites, with effective dates between May 2004 and May 2009, all at review, two agreed and a third, at Clarendon Farm, Linlithgow, an arbitration award. Given the agreed analysis of site rents and our conclusion in relation to costs, we are unable to accept that these three individual comparisons, with very little information, either upset the appellants’ values or support the Assessor’s. We allowed the respondents to produce the arbitration award, showing a very curious valuation method (in the arbiter’s words, “not necessarily one which I would apply in other cases”) and recording that both parties had provided “varying valuation concoctions, some not standard by any means and flawed in certain respects”. We were also not given any indication of the height of that mast. Mr T-K also referred to another situation which he considered to involve rent for an equipped site. These were agreements with an infrastructure company called Spyder, which apparently located sites in the vicinity of petrol stations and built masts, as it were, to order for the telecoms operators, the agreements following a slightly different structure and admittedly containing more ‘lease’ than ‘licence’ terminology. However, apart again from the very small sample, these appear to have been ‘sharer’ subjects in which the payment levels apparently derived from rate cards so that they would not provide any independent support for the ‘sharer’s payment’ basis. (We do appreciate that that situation, where apparently a monopole mast was built with just one company in mind, does appear in some respects closer to a tenancy, or rather sub-tenancy, of these subjects.)
 The other line of evidence which was used to support the Assessor’s approach was the rental levels under rooftop leases, partly compared to greenfield rents and partly compared to rooftop sharer payments. This is the situation in which the cost of erecting a mast is avoided because of the height of the building or other structure (although some form of support for the aerials is still required). Here, there was analysis of a good number of leases, bringing out a considerably higher level than the appellants’ greenfield host site valuations, more comparable with the analysis of sharer payments. So there is some apparent support for the contention that sharer’s payments reflected the value of equipped sites. Again, however, that is speculative, and there may be other explanations related to the different situations. A further matter which arose in evidence was that, while the various agreed figures do not contain any locational variations, i.e. parties were agreed for the purposes of the hearing that locational differences could be ignored, Mr T-K gave evidence that city centre rooftop values were higher, smaller town centres slightly less and rural rooftop situations (of which there are a few) lower again.
 Overall, while the last point does offer the Assessor some limited support, we are unable to accept his ‘sharer’s payment’ basis of valuation. In short, it ignores the real differences in the nature of the payments and involves speculation which, on the evidence, we cannot accept, as to the reason for higher levels of payments. The appellants’ method does have a slight question mark over it in relation to the costs of erecting the masts but that does not in our view tip the balance in favour of the Assessor’s approach. The evidence provided the basis for a small (although unquantified) increase in the appellants’ Plant and Machinery values, but we were not asked to do that (and of course the appellants were not required to meet any such submission).
 The above decisions take us to Scenario 6 –“Cumulo based on Site rents – sharer rights not heritable”. The Assessor then argues that the agreed payaway figure requires to be added to host site values in respect of each and every sharer (but not, as we understand it, where there is no sharer).
 We have little hesitation in accepting that submission. The accepted factual position is, firstly, that the agreed site rental figures do not reflect the potential or hope value in relation to possible sharing; and, secondly, that the market in fact pays the landlord a percentage of the sharer payment, agreed at 27.5%, i.e. £1900 at greenfield subjects, £2,400 at rooftops and £6,000 at base station controller sites. We did not understand the appellants to oppose the principle of increasing host site values where there were sharers, although they did not accept any addition to sharers’ valuations in respect of ‘Payaway’ (something for which we did not understand the Assessor to contend).
 Mr Haddow did, somewhat faintly, refer to the possibility of adding only a proportion of the payaway, by analogy with a decision of the Tribunal in Morrison v Central Scotland Assessor that only 50% of ‘top-up turnover’ rents should be used in the rental analysis. That decision, however, turned on its own particular circumstances, which do not appear to us to be analogous. There is no suggestion of any problem with the calculation of the figure of 27.5% in this case and the payments are not being used to arrive at basic value levels.
 This method of reflecting this additional value is perhaps slightly unusual, but follows this particular rental market. Presumably, in practice, the incidence of sharing will be considered at revaluation rather than, as might possibly appear to be the theoretical implication, as material changes of circumstances each time a sharer arrives at or departs from a site.
 We should express a view on this issue although it does not arise on the basis of our decision on the method of valuation. Here, the appellants contend for downward adjustment of valuations of streetworks and macro subjects, i.e. greenfield, rooftop and base station controller sites, on the sharer payment basis to reflect services provided by the host outwith the terms of the statutory hypothetical tenancy. The Assessor accepts the principle but argues that the appellants have failed to prove any figure to be deducted or in any event that any deduction should be considerably less than the claim of £1900. We were not asked to distinguish among the different categories of subjects in respect of which the claim was made.
 On the assumption here that the Assessor has established his method of valuation, it is clearly for the appellants to establish the amount of any deduction. On this matter also, Mr Haddow’s submissions did not show much faith in the appellants’ evidence. Their first attempt, on a ‘top down’ basis, by reference to the difference between the level of rooftop rentals and rooftop sharers’ payments appears to us to suffer from a similar problem to one which we identified above in relation to the Assessor’s case. It is obviously possible that that difference reflects to some extent the services provided, although the extent of these is distinctly less clear in the case of rooftops, particularly where, as Mr Smith pointed out, there is no shared mast. But it could also reflect more general differences in the nature of the rights and responsibilities undertaken, and there again seems to us to be no way of differentiating these. Then, as recorded above, it transpired that the appellants’ relatively small basket of rooftop sharer payments was too full and taking six cases out made a striking difference to the result. In these circumstances, we could not hold the figure of £1900, or any figure, established on that basis.
 The ‘bottom up’ approach seems more appropriate in principle, but here, in marked contrast to the very extensive work which appears to have been done, on both sides, to analyse the various rentals and sharer payments, the evidence led by the appellants was flimsy in the extreme – essentially unvouched - and has not enabled us to make any specific findings. We do not think it necessary to consider it in detail, as Mr Haddow acknowledged that it was “not particularly successful” and did not advance any particular figure on this basis. We do, however, consider that there was evidence, albeit in the form of rough estimates, from which we should find some figure established. Mr Sharp was at £1569, Mr T-K, “on a generous basis”, £827, and Mr MacTaggart, back-dating Mr T-K’s figure and dividing it by three, £220. That division in three perhaps raises a question of principle, although it might also seem to be an additional pointer to the unreality of the exercise. Mr Haddow suggested that it was quite unwarranted, as the question was the value to the tenant. But if it is correct, factually, to base this calculation on an assumption that there might be an operator host and two operator sharers, it is not clear to us why the sharers should be taken as having agreed between them to add to the rents twice the value to each of them of services supplied at, or substantially at, the same cost to the host.
 In the circumstances, we have decided that if we had accepted the Assessor’s valuation method, we would have deducted £500 from each of the values in Scenarios 1 and 2 in relation to which the appellants made this claim.
 The existence of the important rule against double counting is clear and not disputed. Although Mr Haddow put it at the beginning of his submissions, it seemed to us appropriate to consider it after arriving at results on the various other issues, in order to see whether those results, or any of them, involve impermissible double counting and, if so, to what extent. We should also, again, express our view as to whether, if we had decided any of issues (i), (ii) and (iii) differently, any double counting would have been involved.
 The rule, as described by Mr Haddow without, we think, any contradiction, prohibits the full value, or any part of the value, of heritable subjects, appearing more than once in the valuation roll, in this case in either the entry for the host or the entry for the sharer. The rule may be illustrated by the situations, considered in a number of more modern cases, in relation to shopping malls and shopping centre or retail park car parks. The issue arose where the rents of the shops included the benefit of the mall or car park. A prior issue, to be determined as a matter of law, was whether the mall or car park should be in the roll at all as a separate subject or was simply a part or pertinent of the shops. If it was a separate subject, was there nevertheless double counting? The answer was that there could be double counting if the value of the mall or car park was included in the rents of the shops and the rents had not been appropriately adjusted in arriving at the rateable values of the shops. The fact that there were separate rateable subjects would not preclude double counting. (c.f., in relation to car parks, Ravenseft Properties v Strathclyde Assessor; B. & Q. v Central Assessor)
 The issue arises here of course because of the sharer payments and the subject matter of the sharing agreements, which involve sharing the use of the mast.
 The issue between unum quid and cumulo does not appear to have any bearing on questions of double counting.
 On the assumption, agreed in this case, that the initial or basic site rent does not include any element for the hope or potential of attracting sharing, it appears clear to us that the rental value of the mast, assessed on the site rental basis, is increased when a sharer, or an additional sharer, is taken on, because the rent payable to the land or buildings owner is then increased by the amount of the payaway. The full value of the subjects has increased. If as we have decided the sharer’s rights are not regarded as heritable (Scenarios 5 and 6), no part of the value has appeared more than once in the roll. There is no double counting.
 If valuation is still on the site rental basis, but the sharer’s rights were heritable and so properly included in value (Scenarios 3 and 4), there is no further increase in the value. If any more than the payaway were added, double counting might have to be considered, but under these scenarios the Assessor only sought to include the amount of the payaway in the sharers’ entries and to limit the values of sharers’ rights to that. There would be no double counting.
 However, double counting would in our view have to be seriously considered if there were sharers, the sharers’ rights were heritable and included in value, and the Assessor’s valuation method is preferred. In that situation (Scenarios 1 and 2) the Assessor seeks to enter the full value, i.e the full rent for the equipped site, of both the host and the sharer’s subjects. He argues that this reflects the actual situation of two (or more) occupiers, each getting the same use and therefore value from the mast. The mast, he says, has produced twice the income and has twice the value (or some higher multiple). The two or more rateable occupiers have separate interests, each of the same, full value.
 We have not found this an easy question, but in our opinion the rule against double counting would have to be applied in this situation. Looking separately at the host’s and the sharer’s (or sharers’) subjects, each can be seen, on this valuation approach, as having the full value. However, it seems to us that even on the analysis employed to find the sharer’s rights heritable, these rights are closely related to the value of the mast. The value is derived from use of the mast. Addition of a sharer has increased the value, but not by more than the amount of the payaway. The finding of a separate – it may be said additional – heritable right has not increased the value of the property at that site by any more than that. Just as the Assessor did not seek the full value of the sharer’s rights in Scenarios 3 and 4, the double counting rule requires that the full value should not be entered here. The host’s value should be entered at the Assessor’s figure (less £500), and the sharer’s rights should be valued, as in Scenarios 3 and 4, at the payaway figure, in order to avoid double counting.
 We consider that under that scenario – sharers’ rights heritable, valuation on sharer’s payment basis – that value, i.e. the ‘Payaway’ figure, would fall to be added to each sharer’s, rather than to the host’s, valuation. It might be thought that as the payaway is an addition to the rent paid by the host, it should go on that value, but under this scenario the host’s valuation would not be based on the site rent. In effect, the double counting rule would apply to reduce the sharer’s rights part of the sharer’s valuation to the amount of the payaway. This is also consistent with the position under Scenarios 3 and 4.
 We find the consideration of the value of the BBC’s right to use the mast in ITA instructive here, although the valuation method was different. As we have indicated above, their Lordships clearly did not consider that the addition of BBC increased the value of the mast, which in that case already included its full potential, although the majority felt that the 1/3 of the value had to be added for the period of the Supplementary Roll. There were in that case no rentals under consideration. Here, we have the advantage of rentals, showing the increased revenue which the Assessor wishes to include, but we do not have any indication that the rental value of the mast subjects is increased by any more than the amount of the ‘Payaway’. The site owner receives the ‘Payaway’, no more. Hypothetically, of course, he has provided the mast as well as the site, but we cannot see that as making a difference. It is the tenant who is receiving the sharer’s payment or payments. The value of his occupation is already in the Roll. Conceptually, no doubt, there is one or more subsidiary interest, but this, like the BBC’s interest, is an interest in the use of the mast. To value the sharer’s, or sharers’, interest at 100% is to include more than the full value of the mast. It would, to our mind, be a little bit like adding the rent paid by a sub-tenant to the full rent for the subjects unlet.
 We have to resolve a small difference between the parties in the values for Plant and Machinery (i.e. cables and cabins or cabinets) at rooftop sharer sites under Scenario 6. The appellants’ figures, £300 (cabinet) and £600 (cabin) are in fact higher than the Assessor’s, £200 and £500. There were in fact through the various scenarios and categories of site subtle variations in the agreed plant and machinery values. Quite understandably, the way in which these values were worked through was not explained to us, and the only thing we were told about this small difference was that the appellants’ figures were in line with the agreed figures for plant and machinery under Scenario 6 at rooftop host sites. That might seem to justify the higher figures, but we note that the corresponding figures at greenfield sharer sites are £200 and £500. The appellants may be correct but as the Assessor did not contend for the higher figures we think the correct course is to fix the Assessor’s lower figures.
 Although the parties produced full tabulations of the values of the sites under each of the scenarios, and therefore of various possible values to be fixed in the 8 appeals which we heard, our decision on issue (iv) requires some adjustment to be made. A further brief hearing will be fixed at which parties can report the final outcome in each appeal and formal orders made. The position in the remaining 120 appeals will also be considered then. It is to be hoped that these can be resolved on the same basis at the same time. There may be no need for representation at the hearing if matters are all agreed.