Database of British hills: about the list

Introduction

The purpose of the database is twofold:
  1. To facilitate personal record keeping
  2. To record accurate data on every hill belonging to the most popular lists.
When we brought out the first version of the database in 2001, our primary focus was on classification and the provision of a user-friendly tool for record-keeping. Many lists of hills are available on the Web but not all are kept up to date. Furthermore, many hills appear in two, three or even four different lists. Anyone wishing to amalgamate the lists so that they need only record each ascent once will soon find, as we did, that this is a time-consuming and far from straightforward task. We were not the first to collate the lists into a single database, but by putting our version in the public domain we hoped to save other hillgoers the effort of compiling their own.

We were originally content, for the most part, to use data from published lists. Then in v5 we introduced 10-figure grid references from vetted sources. In subsequent versions we overhauled the Wainwright Outlying Fells and the Deweys and added much data from levelling surveys. As we have a substantial body of summit data that is more accurate than in published tables, and in many cases more accurate than available from large scale maps, our current practice is to align the 6-figure grid references with our own data.

Some hill lists allow changes to the version currently in print to be adopted at any time. We have kept abreast of these via the Marilyn News Centre, The Angry Corrie magazine, the RHB e-group and The Nuttalls site.

The database adopts the convention that no summit can have more than one record. Many hills occur in more than one list, which may give slightly different heights or grid references. Where list authors appear to have chosen different locations for the summit, we have recorded this in the database. In cases where the locations could be regarded as separate summits we have listed the hills separately. Sometimes this is a subjective decision; we hope the majority of users will agree with our choices.

We have limited the data fields to information on name, location, height, ground observations and classification. Those wishing to record additional details can adapt the database to meet their own needs. For reasons why we do not provide drop, please see Enhancements.

If you find any errors or wish to query any of the data, please email the authors at the address on the home page.

Lists

As everyone knows, list making started in Scotland with Sir Hugh Munro in 1891. The Munros were joined by the Corbetts and Donalds and became well known thanks to their publication in Munro's Tables. It took a surprisingly long time for a definitive list of Scottish hills in the range 2000-2500 feet to appear, but in the 1997 edition of Munro's Tables the SMC adopted the list of Grahams which Alan Dawson had first published in 1992. Of Dawson's other Scottish lists, the Murdos and the New Donalds were motivated by a desire to bring objectivity to the classification of Munro Tops and Donalds. Breaking new ground was Corbett Tops and Corbetteers (1999). Included in this database, Corbett Tops are subsidiary summits of Munros and Corbetts within the Corbett height range. The most recent publication from this stable is Graham Tops and Grahamists (2004). This last list subsumes the New Donalds.

Completions of the Munros are recognised by the SMC. Completions of Corbetts, Grahams and Donalds are recorded by the editor of TACit Press.

The Nuttalls and the Hewitts (Hills in England, Wales and Ireland over Two Thousand feet high) are the most accurate and up-to-date of a succession of publications listing the 2000-foot summits of England and Wales. Some of the others are of historical interest but for practical use most people will have little reason to look beyond one of these two lists. However we have received some requests for Buxton & Lewis (1986) and since some walkers are evidently still using this list we have included it. Those interested in the earlier lists of Simpson, Moss, Bridge and Wright are referred to Moss. The only significant differences between Moss's table and ours come from our decision to isolate the different locations given by Buxton & Lewis for Pen Pumlumon Llygad-bychan and Wether Hill.

Completions of the English and Welsh 2000-foot hills are recognised by the Long Distance Walkers Association. It might be thought unreasonable to treat England and Wales as one country, but only The Nuttalls will record completions of one list without the other.

The Deweys effectively extend the Hewitts down to 500m. They provide a worthwhile objective for those who have completed the 2000ft hills of England and Wales.

The Marilyns are unique in having a criterion on drop but none on height. Published in The Relative Hills of Britain in 1992, they have a growing band of devotees which includes the most obsessive peak baggers on the UK hillwalking scene. The completion of the 1554 Marilyns is a formidable challenge which because of the inaccessibility of five of the St Kilda's summits has yet to be achieved, but a tally of 600 is sufficient to join the Marilyn Hall of Fame.

There are no qualification criteria for Wainwrights, which were almost certainly not conceived as a list, but a tradition of climbing them has developed and completions are recognised by the LDWA.

Keen baggers may be interested in a US led initiative to list hills worldwide with a drop of at least 2000ft (609.6m). Anyone who has climbed 100 of these is eligible to join the Century Club. For information see the Topographical Prominence website or join the prominence discussion group. There are 90 qualifying hills in Britain (formerly 91, but Ben Loyal is now regarded as falling short) and 20 in Ireland; members of the RHB e-group can download Rob Woodall's list from the Files section. If there is sufficient interest we may include it in a future release of the database.

Definitions

Marilyns
British hills of any height with a drop of at least 150 metres on all sides. The geographical area includes the Isle of Man and the islands of St Kilda.

A Marilyn Twin Peak is a summit of equal height to another Marilyn where the drop between the two is less than 150m. The 2006 RHB update sheet lists 8 "official" twins: An Stuc, Creag an Dhail Bheag, Buidhe Bheinn, Meall nan Damh, Sidhean a'Choin Bhain, Cnoc Coir a'Phuill, Cruachan Dearg and Middleton Hill. The consensus amongst baggers seems to be that for these hills at least, both summits should be climbed in order to claim the Marilyn. The summits listed in the second table of the RHB update do not make a strong case for being given a separate identity and are given as examples (not a complete list) of hills with two or more nearby points of apparently equal height.

When we created the database in 2001, we included the (then) five hills in the second list as it was not clear to us (or to some other readers) that they did not qualify as Twin Peaks. However it became apparent that there are potentially many other hills in this category. Four more appeared in the 2006 update sheet, although the location of the summit of Ward's Stone has since been identified unambiguously by surveying, and Wendover Woods is an area of flat ground whose summit location is in doubt rather than a hill having two summits with a perceptible drop between them. Alan Dawson subsequently clarified that only hills in the first group have the status of Twin Peak. We have therefore not added any more hills of the second category to the database, but have noted the existence of alternative high points in the Comments field. To delete the original five "examples" would break the hill number sequence and create difficulties for users updating their records, so we have retained four of them as Twin Peaks with a suitable comment and demoted 2190 Gwastedyn Hill to Deleted Marilyn after surveying showed it to be lower than hill 2189. It is left to the walker to decide whether all such points should be visited; on some hills there are many candidates for the highest point and the exercise could degenerate into pedantry.

Cruachan Dearg, Meall nan Damh and Sidhean a'Choin Bhain are also joint Grahams. The SMC lists Buidhe Bheinn as a separate Corbett, illogically in the opinion of many. Creag an Dhail Bheag does not receive the same status but is mentioned in the current edition of the SMC guide to the Corbetts.

Munros
Scottish hills at least 3000 feet in height regarded by the SMC as distinct and separate mountains, based on a list originally published in 1891. Subsidiary summits meeting the height criterion are designated Munro Tops.

Murdos
Scottish hills at least 3000 feet in height with a drop of at least 30 metres on all sides. All Murdos except Beinn a'Chroin are also Munro Tops but some Munro Tops fail to qualify as Murdos.

Corbetts
Scottish hills between 2500 and 2999 feet high with a drop of at least 500 feet (152.4m) on all sides.

Corbett Tops
Subsidiary summits of Munros and Corbetts between 2500 and 2999 feet high with a drop of at least 30 metres on all sides.

Grahams
Scottish hills between 2000 and 2499 feet high with a drop of at least 150 metres on all sides.

Donalds
Hills in the Scottish Lowlands at least 2000 feet high. 'Tops' are all elevations with a drop of at least 100 feet (30.48m) on all sides and elevations of sufficient topographical merit with a drop of between 50 and 100 feet. Certain of these are designated 'Hills' according to a complex formula based on both distance and drop.

A related list is Dawson's New Donalds, not given here, in which the qualifying criterion is simplified to 30 metres of drop. The New Donalds are a subset of the Graham Tops. Anyone who has completed the Donalds and Donald Tops will have visited all the New Donalds.

Hewitts
Hills in England and Wales at least 2000 feet high with a drop of at least 30 metres on all sides.

Nuttalls
Hills in England and Wales at least 2000 feet high with a drop of at least 15 metres on all sides. The list includes many minor summits not qualifying as Hewitts. The most notable of these is Pillar Rock whose ascent by the easiest route is a Moderate rock climb or Grade 3 scramble.

Deweys
Hills in England, Wales and the Isle of Man at least 500m high with a drop of at least 30m on all sides. The list was published with no upper bound on height, but in practice the name is applied to hills below 610m (2000ft) high, as hills over 610m are usually called Hewitts. This convention is followed in the database.

Wainwrights
The 214 hills listed in volumes 1-7 of Wainwright's A Pictorial Guide to the Lakeland Fells.

Wainwright Outlying Fells
Hills listed in The Outlying Fells of Lakeland. The best known source of data is New Combined Indexes to A. Wainwright's Pictorial Guides, John M Turner, Second Edition (1984), Lingdales Press. This list has many inaccuracies. Not least, it is short of two summits explicitly mentioned in Wainwright's book (Caermote Hill, for which the author gives the location of St John's Hill; and the southern summit of Newton Fell) and it also omits the 12 nameless summits. In version 6 we overhauled the entire list using the latest 1:25000 maps and where necessary, older metric and imperial maps. The assistance of correspondents in this onerous task is gratefully acknowledged.

We mention that the task of producing a definitive list of summits is far from straightforward because of the many occasions when the location of Wainwright's summit is not coincident with the highest point in the vicinity. Often this is due to Wainwright's preference for designating a recognisable feature such as a cairn, but in in other cases it stems from inaccuracies in the OS maps of the time. Where matters of interpretation come into play we have mentioned this in the Comments field. A few summits would benefit from a site visit for confirmation of the exact location.

The list at the back of Wainwright's book contains 110 named fells and summits. Close inspection shows seven of them to refer to other hills in the list, while Newton Fell has two summits. Thus: The addition of the 12 nameless summits brings the total number of Wainwright Outlying Fells to 116, 14 more than in Turner's list.

Deleted Tops

Sometimes a hill which once appeared on a list is later removed. The most common cause is remapping, though many Munro Tops were deleted on subjective grounds. We include categories for Deleted Munro Tops (which includes deleted Munros not retained as Tops), Deleted Corbetts, Deleted Donald Tops (which includes deleted Donalds), Deleted Marilyns and Deleted Nuttalls. In the absence of an accepted definition of what constitutes a deleted top, we have adopted the following convention: All deletions at any time in a list's history are included. Thus Deleted Munro Tops includes all deletions from 1921 onwards. Details of specific changes to Munros and Munro Tops through the different editions of Munro's Tables are given in The Munros: 1891-1997.

Subs

A "Sub" is a hill that falls short of meeting a list's classification threshold by a specified amount. The database includes SubMarilyns, SubGrahams, SubMurdos and SubHewitts. Dawson originally defined Subs as hills that miss out by less than 10 metres in height or 10 metres in drop. This definition is adopted in the TACit booklets on the Murdos, Grahams and Hewitts and in most private lists circulating on the web. Recently Dawson has changed the definition of Sub, limiting it to hills that only miss out on drop. The rationale is that lists are now available for all height ranges with 30+ metres of drop (see Marhofn 153) so that a Sub that misses out on height will always be a member of another named list. Hence a "Sub" category is usually only needed for hills with 20-29m of drop.

There may be some logic in this change, but it represents a departure from the original idea of a Sub being a "near miss". This was always an attractive concept, and the term "Double Sub" continues to be used for hills that marginally miss out on both height and drop. A drawback of Dawson's revised definition is that readers wanting to know which hills to climb to be certain of bagging all hills that might in reality qualify for the parent list (see below), or to survey borderline cases on the ground, cannot simply look at a table of Subs. As we believe these are important reasons for being interested in Subs, we have retained the original definition in the database. In any case, the new definition of Sub has yet to appear in the published lists of the Sub categories included in the database.

Prior to v9 we did not include Double Subs because the relevant categories of hill with less than 30m of drop appeared not to have been fully researched. This is no longer the case, although there is a possibility that more remain to be discovered. Accordingly, we now list DoubleSubMurdos, DoubleSubHewitts and DoubleSubGrahams. Rather than create three new categories in the Queries, each of which would include only a few hills (and a single DoubleSubGraham!), we have included the hills in the Sub categories and distinguished them by means of a superscript in the Classification field, e.g. a DoubleSubHewitt is denoted h².

We have not included SubCorbetts, Sub Corbett Tops or SubDeweys. SubCorbetts are listed in the TACit tables and will be included in a future revision if there is demand. If retaining the original definition of Sub, it is tempting to extend the "subness" on height down to 750m, as suggested by Dawson. We do not have a complete listing of SubCTs or SubDeweys.

The statistical error associated with heights on OS maps means that some Subs have a non-negligible probability of qualifying for the main list. Serious baggers who wish to legitimately claim ascent of all hills meeting the list criteria will need to climb some Subs. As a rough guide, you should climb those hills within 3m of the qualifying height and 4m of drop. For further information see Allowing for measurement error. It would be wrong, incidentally, to assume that Double Subs have an insignifcant probability of qualifaction. A change in summit height will induce a corresponding change in drop, so errors in height and drop are correlated. Birks Fell was a DoubleSubHewitt before its promotion.

SubMarilyns, SubHewitts etc. that are deleted during the lifetime of this database will be retained as unclassified hills. Starting with version 10, we are adding unclassified hills that have been surveyed as falling short of qualification for list status. This will help to ensure that the survey results are not "lost" in the event of these hills being queried in the future.

Description of fields

Notes on the contents of the database fields follow. The following abbreviations are used when referring to sources:
RHB=The Relative Hills of Britain; TACit=TACit Tables. Significant changes to Marilyns since RHB's publication in 1992 are given in update sheets and can be found on the Marilyn News Centre. At the time of writing the Grahams list on the TAC website dates from the 1995 book and not from the 2nd edition dated 1999.

Number
A unique hill identifier to assist with revision and help users raise queries with the authors. The hill number will not be changed during the lifetime of the database. To upgrade your copy, sort the old and new releases by hill number and paste your personal ascent records from one to the other.

Name
The name by which the hill generally appears in lists. In the case of Munro Tops, Deleted Munro Tops and Corbett Tops, the name of the parent Munro or Corbett is generally given first followed by the name of the Top as it appears in the published list. For consistency, the same convention is followed for Sub Murdos which lie close to a Munro, and for a handful of other hills where the summit has a different name from the hill.

Alternative names by which a hill is known are given in square brackets.

Section
The RHB Section number. Regions 1-17 correspond with those in Munro's Tables, enlarged to include lesser hills. In Corbett Tops and Corbetteers (1999) sections 5, 7 and 8 were split for the first time into West (A) and East (B) sections. Section 26 was subsequently split for the Graham Tops booklet. To preserve the approximate hill order in RHB and the earlier TACit tables, these subdivisions are disregarded in all Queries except those relating to Corbetts and Corbett Tops. This has been done by use of the _Section field which uses the original numbering scheme. Note also that Sections 10A and 10B in Munro's Tables do not correspond to 10A and 10B in RHB.

Subsequent to the publication of RHB, the boundary between Sections 1 and 26 was moved to follow the course of the Highland Boundary Fault. This resulted in some hills being moved from 1B to 26B.

Hills duplicated in more than one section of the RHB/TACit tables, or which could be put in more than one section, have been classified as follows:

Black Mountain (2241, Wales)
Formerly listed in RHB/TACit as belonging to both England and Wales. With effect from May 2007, Black Mountain is deemed to be in Wales only (32A) for the purposes of lists and databases, and such evidence as exists points that way. This removes the anomaly of the hill being double counted in the Marilyn totals in the Microsoft Access version of the database. Black Mountain was always assigned to Wales in the Nuttalls' list.

Hills on the England-Scotland border
Assigned to Section 33 with the exception of Cairn Hill West Top in 28B. This Donald Top does not appear in English lists as the drop before ascending to Cairn Hill is only 5m. The summit is probably in Scotland.

_Section
A digital version of Section with regions 5, 7, 8 and 26 not subdivided, to assist sorting. Not shown in the queries.

Area
Nuttalls and Donalds area names are used for all hills belonging to those lists. This facilitates comparison with the original lists and will also serve for sorting Wainwrights by volume. For Wainwright Outlying Fells we have extended the areas defined in the Pictorial Guides by continuing the Windermere boundary southwards along the River Leven to Greenodd, and from Bassenthwaite Lake north-west along the River Derwent. In England and Wales, the Nuttall and Wainwright names have been used for all other hills falling within those areas. There are a handful of lesser hills situated between the English Lakes and Dales National Parks, and some Deweys in Wales, where one can make a case for a different area name from the one we have chosen.

A particular problem lies in the boundary between the Arans and Berwyns for the hills south of Bala from Rhiwaedog-uwch-afon (3421) in the north to Mynydd Maes-glas (3424) in the south. The easiest solution would be either to put them all in the Arans, as Dewey does, or all in the Berwyns. However in the Nuttalls' book, Moel y Cerrig Duon (2116) belongs to the Arans and Foel y Geifr (2115) and Foel Goch (2123) to the Berwyns. Geographically this is not very logical, but the Nuttalls clearly did so because Moel y Cerrig Duon is conveniently included in the same walk as the hills west of the road summit. Our current solution is to assign those hills south of Moel y Cerrig Duon and south-west of Lake Vyrnwy to the Arans, and those north of Moel y Cerrig Duon and to the north-east of Lake Vyrnwy in the Berwyns, with the exception of Moel Eunant (3412) which is a satellite of Moel y Cerrig Duon. We feel this is the best we can do without breaking the alignment with the Nuttalls' book.

To divide the Arenigs from the Moelwyns we chose to make the boundary Festiniog-B4391-B4407. There are other options but none are demonstrably better. The Moelwyns (as defined by the Nuttalls) span two RHB regions, 30B and 30D. Dewey's division of central Wales is a good one, but to adopt it would mean breaking the link with the Nuttalls' book for hills in their "Central Wales" chapter. We have used Dewey's area Hiraethog for the four hills within that region as it does not contain any 2000ft summits.

For hills outside the areas mentioned above, the RHB Section name is used except where a commonly used regional name exists, e.g. Forest of Bowland.

Height and Grid Reference
The majority of data for England and Wales come from 1:10000 metric maps, as used by Dawson in the TACit Tables. The main exceptions are Deweys, sub-2000ft Wainwrights that lack any other classification and Nuttalls that do not qualify as Hewitts or SubHewitts. Other published lists may give different data. One reason why grid references can vary is that authors have not measured them in the same way. By convention, the OS Grid Reference is the address of the 100m square in which the feature lies. This is given by the co-ordinates of the south-west corner of the square. For example, the trig point of Great Shunner Fell is located at SD84864 97288 so the correct 6-figure Grid Reference is SD848972. The TACit tables comply with this convention but some list authors round to the nearest 100m instead; in the above example they would give SD849973. To remeasure all doubtful summits on large scale maps ourselves would be a time-consuming task of uncertain value, considering that spot heights and trigs are often some distance from the summit anyway. It is better to regard the 6-figure Grid Reference of a hill as approximate if no 10-figure Grid Reference is given.

Data for Scotland are mostly from 1:10000 or 1:25000 mapping if from Dawson and from 1:50000 or 1:25000 maps for a handful of hills taken from SMC publications (we are aware of the errors in Munro's Tables). Some pre-1981 Deleted Munro Tops not otherwise classified are from our own map readings; any ambiguities regarding the intended summit are noted in The Munros: 1891-1997. As the OS are continually revising their maps it is inevitable that some details will differ from the most recent maps, but a particular problem for those seeking definitive data has been the discrepancies between spot heights on 1:10000 metric maps, or their digital equivalent, and those on recent 1:25000 Explorer maps. Until comparatively recently, spot heights on 1:50000 maps were generally metric conversions of imperial heights from the First Series survey. These are gradually being replaced, but many still remain.

In v7 we overhauled the 6-figure grid references to align them with the 10-figure grid references where we have them. This was done after making a small adjustment to the GPS measurement to correct for systematic error (see below). We also refrained from changing some existing 6-figure grid references where the change might be attributable to statistical error. We have repeated this process in subsequent revisions.

For Wainwrights the author sometimes gives a summit location that is not the highest point of the fell. This is particularly true of the Outlying Fells. Our policy is to take the location intended by Wainwright. Any doubtful cases are mentioned in the Comments field.

Metric heights are converted to feet using a factor of 3.28084.

10 Figure Grid Reference
Except where otherwise indicated, all ten-figure grid references were obtained on the ground using Garmin GPS units. The precision of most measurements (as given by the instrument) was also recorded at the time of measurement and correlated well with independent measurements made by the authors. In an experiment in which 100 measurements were made in a fixed location, the precision of eastings was significantly greater than the precision of northings; we do not know whether this was a quirk of the chosen location. For eastings, 95% of measurements were within ±3m of the mean and 99% within ±4m; for northings, 95% were within ±4.5m and 99% within ±5.5m. For consistency with this level of precision we have rounded all measurements to the nearest 5 metres, except those which come from OS data (see below).

We have included some ten-figure grid references published by Ordnance Survey of trig points on hill summits listed in the database. A comparison of 246 OS measurements with our GPS measurements has shown up systematic variation between the two data sets. GPS eastings range from being 7m higher than OS eastings in the westernmost parts of Scotland to 1m lower in the east. GPS northings vary from being 14m lower than OS northings in Northern Scotland to 9m higher in SW England. We are grateful to Darren Parker who had himself discovered this error and has researched its cause. We reproduce Darren's explanation below.
The latitude and longitude shown on Ordnance Survey maps are determined with respect to the OSGB36 (Ordnance Survey Great Britain 1936) datum. This datum is based upon a ground survey performed between 1936 and 1953 and uses the ellipsoid defined by Sir George Airy in 1830. The latitude and longitude can be converted to planar coordinates using a Transverse Mercator projection (once the origin is defined) to give the National Grid references we use. Since the advent of GPS the method of defining the National Grid has changed. It is now defined using the latitude and longitude determined with respect to the ETRS89 datum (which is based upon the WGS84 datum and uses the GRS80 ellipsoid) which are then converted using a transformation known as OSTN02 to give the give the grid reference with respect to OSGB36. The transformation OSTN02 is not a simple transformation defined by equations alone, but because of distortions in the OSGB36 grid, it also uses values of slight shifts in northing and easting values. The grids of northing and easting shifts between ETRS89 and OSGB36 cover Britain with a grid resolution of one kilometre. The shifts of a particular point are then interpolated from this grid. The transformation OSTN02 can be obtained from www.ordnancesurvey.co.uk/oswebsite/gps - the software is called Grid InQuest.

Thus the National Grid is now defined by ETRS89 and the OSTN02 transformation. A good guide to the subject is A guide to coordinate systems in Great Britain. Also available is an Excel spreadsheet with many useful functions for converting from one datum to another.

A GPS unit determines the latitude and longitude of its position in the WGS84 datum (which is almost identical to the ETRS89 datum). In order to display this position as a British National Grid reference the GPS unit must perform a transformation. Unfortunately, the transformation equations stored in the unit are not as accurate as the OSTN02 transformation. Garmin units use a transformation known as a Molodensky transformation (the equations and required parameters can be found in "Department of Defense World Geodetic System 1984 - Its Definition and Relationship with Local Geodetic Systems, NIMA TR8350.2, 3rd Edition, Amendment 1, 3 Jan 2000"). This leads to the discrepancies highlighted above. The transformations used by other manufacturers have not been investigated.

A number of strategies are available to remove the errors introduced by the Molodensky transformation. GPS waypoints can be downloaded to a computer using free or commercial software and the WGS84 latitude and longitude extracted. These can then be converted to British National Grid references using Grid InQuest (which has a batch mode facility). Alternatively, if the British National Grid references have been copied from the GPS screen, these can be converted back into WGS84 latitude and longitude values using the Molodensky transformation equations directly or by using the free program Geotrans. As before, these can then be converted to British National Grid references using Grid InQuest.

When entering position coordinates of a location into a GPS, one has to bear in mind the source of the coordinates. If a grid reference has been read from the screen of a GPS unit, then entering the grid reference into another GPS unit using its own input screen will result in a point in the correct location (even though the grid reference may be incorrect). However, if the position coordinates are uploaded using computer software, the coordinates should be in the form of WGS84 latitude and longitude values.

So, if you use the OS ten-figure grid references given in this database in your GPS be mindful that they will not take you quite as close to the trig point as the GPS measured ones, especially if you are in the Scottish Highlands or the far south. Since we presume users will be using the grid references in this way, we only include OS values for those trig points for which we do not have GPS measurements (even though the OS values are "correct").

Because 6-figure grid references are usually used with maps, it was desirable to correct the 10-figure GPS measurements for errors introduced by the Molodensky transformation before truncating them to derive the figure for the Grid Reference field. As described above, this can be done by successively transforming the data to WGS84 and British National Grid using Geotrans and Grid InQuest. In fact, prior to becoming aware of this software, we used our data on the 242 trig points to develop an empirical model that would estimate the correction in one pass. The model gives a satisfactory fit to the data and the residual variation (prediction error) agrees very well with the random component of the GPS error estimated from the repeatability experiment described above. The residual variation for northings also exceeded the residual variation for eastings, though the difference was smaller than in the reproducibility experiment. The correction factor estimated by the model will not be as accurate as the more rigorous method mentioned above, but it is easily programmed in Excel and the inaccuracy is small compared to the random error in the measurement.

Almost all values in the database refer to summit cairns, trig points or the obviously highest point in the vicinity of the six-figure grid reference. A few hills, e.g. Beinn Donachain NN198316, have poorly defined summits where several knolls vie for the title and we have chosen that point which we believe conforms most closely to the position given by the published six-figure reference. Other hills, e.g. Hail Storm Hill SD835193, have a huge summit plateau where the concept of a ten-figure grid reference is meaningless. Again in these cases we have chosen a point within the 100m square given by the six-figure reference that appears to have the highest elevation. In some cases authors believed that other features outside the 100m square defined by the six-figure grid reference might represent the true position of the summit. Few of these data are entered in the database, although all are recorded in the spreadsheet from which the database entries are derived. Anyone interested in seeing the complete spreadsheet may receive a copy from the authors upon request.

It is our aim to obtain GPS-measured ten-figure grid references for as many hills as possible in the database and we would be grateful if you are interested in helping. Each entry in the spreadsheet is credited to the person who made it. Please record: name of hill, ten-figure grid reference, height as given by GPS, precision of GPS at time of measurement, what the measurement refers to (e.g. summit cairn), your name and the date on which the measurement was made. Your GPS should also be set up according to the manufacturer's instructions using British Grid as the position format and ordnance survey GB as the map datum. Try to remember to switch on your GPS a few minutes before reaching the summit and try to leave the unit for at least five minutes to settle before taking a reading. We welcome your input.

Some 10-figure grid references are published on the web but comparatively few are from ground observations. Measurements from maps cannot provide the same degree of accuracy, and frequently fail to identify the highest point.

Our thanks are due to Ian Baines, David Baird, Alex Barbour, John Barnard, Bert Barnett, Jim Bloomer, David Claymore, Jim Coombes, Richard Cooper, Peter Cottam, Stephen Dawson, Dennis Foster, George Gradwell, Peter & Liz Hastie, Ian Henderson, Idwal Jones, Paul Kington, Sandra Morrison, David Purchase, Laurence Rudkin, Iain Rudkin, Toby Thurston, Paul Ward and David White for their invaluable contributions.

Feature
The summit feature to which the 10 figure grid reference refers. The summit area may be adorned with several objects (trig point, cairn, wind-shelter, fence etc.) and the resolution of the GPS is usually sufficient to be able to distinguish the positions of these features. Note that a cairn or trig point is not always at the highest point of the hill, which may itself be featureless. Consequently, the Feature field may contain the entry 'no feature' even though a cairn, trig point or other feature is in the vicinity. This is true particularly for hills that have been surveyed.

Observations
This field contains information that supplements the Feature field. For example, it may describe the summit relative to other more obvious features, or it may give a grid reference for an alternative summit position. We have included this field only in the Hills table and queries 43-45 in the belief that the level of detail it contains is more than many users will want. Please email us if you would wish to see this field included in all queries in future releases of the database.

Survey
This field records whether the summit position has been determined by levelling instruments. The basic tool we use is an Abney level which has a resolution of about 50cm of height per 100m of distance. This is sufficient to enable the true summit position of most hills to be determined. For hills where greater resolution is required, a Leica Runner 20 Automatic level has been employed. This instrument has a x20 telescope and gives resolution of about 1cm of height per 100m of distance. More recently we have purchased a Leica NA730 Automatic level which boasts a x30 telescope and correspondingly higher resolution of about 0.5cm of height per 100m of distance. An Abney level is a small device weighing about 150g which easily packs into a rucksack. A Leica Runner weighs about 1500g, occupies significant space in a daysack and also requires a sturdy tripod. We are grateful to John Barnard for researching these instruments and recommending models for purchase. Apart from the surveys by CMCR of Foinaven and Beinn Dearg, all other survey entries in the database have been made by John Barnard, Graham Jackson, Jim Bloomer and Myrddyn Phillips.

The entries in the Survey field are 'Abney level', 'Leica Runner 20', 'Leica NA730' or, for cases where an instrument is not required to determine the summit position, 'obvious summit'. A blank field denotes that the hill has not been surveyed.

Classification

M Marilyn D Donald
m Sub Marilyn E Donald Top
x Deleted Marilyn e Deleted Donald Top
P Marilyn Twin Peak R Murdo
U Munro r Sub Murdo
T Munro Top r² Double Sub Murdo
t Deleted Munro Top H Hewitt
C Corbett h Sub Hewitt
c Deleted Corbett h² Double Sub Hewitt
A Corbett Top of Munro N Nuttall
B Corbett Top of Corbett n Deleted Nuttall
G Graham W Wainwright
g Sub Graham O Wainwright Outlying Fell
g² Double Sub Graham L Buxton & Lewis top otherwise unclassified
5 Dewey

Revision
The date of the last change to the record, excluding those that relate only to 10 figure grid references and surveys. Shown in queries 43-45, 48 and the Hills table.

Comments
Significant revisions, alternative summit locations not from our own surveys, and other explanatory notes. Minor changes to the 6-figure GR are generally not commented.

Streetmap
Link to a 3 x 3km OS 1:50000 map on www.streetmap.co.uk. Changing z=3 to z=4 in the URL gets a 5 x 5km map. There was a period in the past when this link gave a general map, after Streetmap had included a variable parameter in the code to discourage deep linking. If this happens again, click on one of the grey arrows below the bottom right corner of the map.

OS Get-a-map
Link to a 2 x 2km OS 1:25000 map on http://getamap.ordnancesurvey.co.uk.

xcoord, ycoord
Absolute grid reference (eastings, northings) in metres relative to the Ordnance Survey National Grid origin. Required by some GIS software e.g. ArcView and MapInfo. The 10-figure grid reference is used if available. In the Microsoft Access version, these fields are produced by a formula in queries 43-45.

Recent changes and issues

The following hills are mentioned because of ambiguity, uncertainty, or changes that have not yet found their way into published lists. They are listed in order of Section number.

Beinn a'Chroin (Murdo) (2925, 1C) and Beinn a'Chroin West Top (37, 1C)
Research by Richard Webb suggested that the West Top was higher than the Munro (hill 36) as long ago as 1983. The latest 1:50000 and 1:25000 maps both show a 942m spot height at NN387185. This is the new position of the Murdo, the previously published location (1995) being coincident with the Munro Top. Munro's Tables gives the West Top as the cairned point at NN385185. Its height of 938m reportedly originates from the First Series 1:50000 map and it lies within a 930m ring contour on the new map. The names of both Beinn a'Chroin Murdos have changed to reflect the reversal in height, and we expect the SMC will follow suit when the Tables are next revised. To avoid confusion we separated the summits in v4 of the database, retaining the Munro Top as hill 37 with the old height and assigning the Murdo to hill 2925.

Meall Chomraidh (140, 2A)
New Marilyn promoted in May 2005. See Marhofn 133.

Beinn a'Chlachair East Top (359, 4B)
SubMurdo discovered since publication of the TACit booklet. The latest Explorer map has restored the original height of 977m, giving a drop of 29m.

Sron Bealach Beithe (2924, 4B)
New SubMurdo reported in Oct 2002 following publication of the 1:25000 Explorer map.

An Dun (399/400, 5)
The summit of the Marilyn has moved 300m along the summit ridge. We have assigned the new location to the Corbett and retained the old grid ref as a Deletion in both categories.

Gairbeinn - Corrieyairack Hill (632, 9B)
There has been some controversy concerning the location and height of the summit of this former Corbett. The Corbett Tops booklet states that the new summit height is taken from the 1:10000 map and is 1m higher than the old top. For the OS's reply to a query see Marhofn 106.

Beinn a'Chaorainn (663/665, 9C)
Older 1:50000 maps show the South Top at 1050m, higher than the Munro/Marilyn at 1049m. Following enquiries in August 2001, the OS admitted that "something odd seems to have happened" and have reheighted both peaks.

The Saddle North Top [1921: Sgurr na Creige] (706, 10A)
The intended summit is uncertain. It is described in all editions of Munro's Tables as being half a mile north of the summit and about 3100ft in height (the first edition says over 3000ft). In the 1932 SMC Guidebook for The Western Highlands, though, Stob na Creige is described as being about 2850ft in height. A visit showed Stob na Creige to be a short undulating ridge with three small tops, the drop between each of these being no more than 5m or so. The furthest from the Munro is 877m or 2877ft and is 1.1km or 0.68 mile along the ridge from the summit. Its grid reference is NG 93358 14128 and is the closest to the location given in previous releases of the database. The next hump is 930m or 3051ft at 0.7km or 0.44 mile from the summit, and the next after that is 938m or 3077ft at NG 93570 13715, 600m or 0.37 mile from the summit. There are two other significant bumps closer to the summit. We give the 930m feature in the database as it seems to best fit the description.

Buidhe Bheinn (713/715, 10A)
Munro's Tables assigns the Corbett to the west top at NG956087 and quotes a height of 885m. The 2002, 1997 and pre-1990 Landranger maps all spot NG956087 at 879m whereas the summit at NG963090 lies within a 880m contour. The 2002 edition of the SMC guide to the Corbetts has corrected the location, which it correctly refers to as a half Corbett since it is a twin of 716. Accordingly, we have relocated the Corbett to hill 713 and demoted hill 715 to Deleted Corbett.

Sgurr nan Eugallt (744/745, 10B)
The summit was moved in May 2001 following information from the OS.

Meall nan Damh (3319, 12A)
New Marilyn Twin Peak / twin Graham announced in May 2003. Twinned with Carn na Coinnich (hill 914), 4km away. Both summits should be climbed.

Beinn Dearg (970, 13A)
One of two 914m hills (the other being Foinaven) surveyed by CMCR Ltd. on the initiative of The Munro Society. The definitive height is 913.7m, which equates to the old imperial height of 2998ft.

Beinn Dearg North Top (3267,15A)
New Corbett Top discovered since the publication of Corbett Tops and Corbetteers.

Beinn Tharsuinn - Sidhean a'Choin Bhain (3660,15B)
New Marilyn Twin Peak / twin Graham announced in April 2006.

Foinaven (1124, 16B)
There has been speculation for many years that the height of this hill might reach 3000ft. Older 1:50000 maps show a spot height of 908m at NC316507, a conversion of the old imperial height. It was eventually replaced by the 914m figure from the metric survey, though the spot is at NC315507 on the 1:10000 metric map. Anecdotally, the 1:25000 map briefly gave the height as 915m, but the current Explorer gives 914m at NC316507. In response to an enquiry from the SMC in 1990, the OS confirmed the height as 914m but quoted a range of 913.8-915.2m including measurement error. The midpoint of this range is 914.5m, which would suggest that the hill was more likely to be a Munro than a Corbett. The confusion was compounded by the spot height on the 1:25000 map being within the small 910m contour to the east of the ridge, rather than at the cairn inside the much larger 910m contour to the west which ground observation suggested was about 3m higher.

In 2007 The Munro Society commissioned a private survey of Foinaven. The summit height, ratified by the OS, was reported as 911.0m.

Beinn Ceannabeinne (3336,16B)
New SubMarilyn reported by Alan Dawson in the rhb e-group in August 2005. The drop was estimated as 145m from altimeter measurements.

Knight's Peak (1261, 17B)
Speculation that this peak does not reach 3000ft abounded soon after publication of the 1997 Munro's Tables, supported by Harveys' surveyed height of ca. 911m and a figure of 2994ft on the 1965 OS 1:10500 sheet. In August 2001 the OS disclosed that it holds an unpublished air survey height of 912m for Knight's Peak. A spot height of 912m also appears in Land-Form PROFILE, the OS 1:10000 digital height product. For a number of years, therefore, it seemed that Knight's Peak did not merit the status of Munro Top.

Then in August 2006 Ken Stewart obtained a new definitive height of 914.95 ± 0.5m from the OS, derived from high order photogrammetry and GPS. On enquiring about the methodology, the OS replied "the photo model was controlled using sub 0.1m accuracy GPS (i.e. points on the ground were fixed that could be identified on the imagery - GPS was not taken to the summit). The accuracy of the imagery heighting using this method is quoted as +/- 0.5m for the Z (height) value". See also TAC69.

Suidhe Chatain (3681, 19C)
New SubMarilyn reported in June 2008.

Uamh Bheag (1644, 26)
RHB and TACit have not designated the East Top at NN696119, also 664m, as a Twin Peak since it looks lower and less significant than the main summit. Uamh Bheag was listed in Section 1 in RHB but moved to Section 26 and then to 26B in TACit.

Meikle Millyea (1693/1694, 27B)
The true summit of the Donald is at a point 400m SSW of the location given in Munro's Tables and is spotted 3m higher on the Harveys 1:40000 map. We have retained the old location as a Deleted Donald.

Castell y Gwynt (3662, 30B) and Carnedd y Filiast North Top (2001, 30B)
New Nuttalls adopted on 2 September 2007. Castell y Gwynt is a new 3000er. Details of the surveys can be found in the Files section of the RHB e-group.

Cadair Bronwen NE Top (2093, 30E)
We had seen reports that the bwlch to summit height might be as little as 7m rather than the 15m required for a Nuttall. Using staff and automatic level, John Barnard and Graham Jackson obtained 'uphill' and 'downhill' estimates of 10.6m and 10.7m for the drop. The biggest component of the measurement error is the location of the exact position of the bwlch at SJ 8355 3495, but the authors were entirely confident that the drop is not greater than 11.5m. The hill was demoted on 19 August 2007 with a ceremony reported by the Nuttalls.

Mynydd Ceiswyn (3431, 30F), Domen-ddu (3466, 31B)
Added to the original list of 500ft summits (with Great Yarlside) by Michael Dewey, but later challenged in an rhb usergroup message. Mynydd Ceiswyn and Great Yarlside have since been shown by accurate levelling techniques to lack the required drop. The changes have been accepted by Michael Dewey. The survey reports can be read in the Files section of the RHB e-group. We hope to resolve the status of Domen-ddu in 2008, along with 3575 Watch Hill which has also been queried. The two latter hills are classified as DoubleSubHewitts by other authors.

Moelfre (2217, 31C)
Listed in TACit (1997) as a SubMarilyn but later deleted on grounds of insufficient drop. The hill now lacks any classification.

Honister Crag (3320, 34B)
The first new English Nuttall to be added since the original publication. The drop is 65ft.

Baystones (2607, 34C)
Demoted from Marilyn status in 1995, this hill was reinstated in May 2001 following new OS data. The summit surveyed at 488m is on the other side of the wall from the 487m spot height at NY403051 given in RHB, but the positions are so close that it seems unduly pedantic to list the former location as a Deleted Marilyn. Wainwright gives the new location as the summit of Wansfell.

nameless (Naddle Horseshoe - 2) [Wallow Crag] (3329, 34C)
Previously we gave the location indicated on Wainwright's sketchmap. This does not match the route description, which in fact follows the ridge to Wallow Crag. It is possible that the 2½" map used by Wainwright was confusing, as with Great Yarside. There can be no doubt of the correct summit as the sketch of the cairn on p.227 of The Outlying Fells of Lakeland matches the photograph on p.202 of Birkett's The Complete Lakeland Fells (the original photo is in colour but the scan has been reproduced in B/W to match Wainwright's sketch). The 410m contour ring at NY497148 is at best 412 metres high and is nothing more than a heathery mound that has clearly never had a cairn on it. Birkett's location is verified in a recent photograph of Wallow Crag kindly provided by George Gradwell, who discovered the error.

Great Yarlside (3661, 34C)
Added to the original list by Michael Dewey, but demoted after a levelling survey demonstrated that the hill lacks the required 30m of ascent. The survey report can be read in the Files section of the RHB e-group. The hill now lacks any classification. The Wainwright Outlying Fell is hill 2575, close to a 19th century circular trig station. The height was mistakenly given on OS maps as 1986ft (605m) at one time, probably a transcription error as it was previously 1936ft, leading Wainwright to choose that location.

Arnside Knott (3321, 34D)
New Marilyn reported in May 2005. Formerly unlisted. For details see Marhofn 133.

Burnhope Seat (2714, 35A)
The Nuttalls regard the summit as being the trig point at NY788375 (746m). They reported having visited the other top, which is not spotted on the 1:10000 map.

Birks Fell (2799, 35B)
This hill has a history. Early lists of the English 2000s included Birks Fell on the basis of the 610m spot height at SD918763 on 1:50000 Landranger and earlier imperial maps. Later metric maps at 1:10000 and 1:25000 scale gave a 608m spot height at SD916764 instead. This caused the hill to be dropped from subsequent lists, and for the Marilyn to be moved to Horse Head Moor. The situation was confused by the continued appearance of the 610m spot on 1:50000 maps (known to be a metric conversion of an older imperial height) and a stream of visitors opining that the new 608m spot was not at the highest point. The OS were contacted but they merely confirmed 608m as the highest recorded point on the metric contour document.

The impasse was resolved when John Barnard and Graham Jackson surveyed the hill using precision levelling equipment. A follow-up survey, in which they obtained photographic evidence, confirmed their finding that the hill was definitely above 2000 feet, their estimate being 610.4 ± 0.2m or between 2002 and 2003ft. Detailed reports of their two surveys can be read in the Files section of the RHB e-group. This information was relayed to interested parties. Subsequently the Nuttalls obtained a revised figure of 2001ft at SD919764 from the OS (from a 1920 levelling survey) which they indicated will appear on the next update of the 1:25000 Explorer. The OS have now (19 Dec 2007) indicated that the new 610m spot will be shown at SD 9186 7637, a little to the south west of the cairn and in agreement with the 1:50000 map.

Thorpe Fell Top (2803, 35B) and Cracoe Fell (3682, 35B)
Speculation that Cracoe Fell was higher than Thorpe Fell Top was finally put to rest following two line surveys by John Barnard and Graham Jackson in May 2008. Cracoe Fell is 1.8m higher. Survey details can be read in the Files section of the RHB e-group.

Axe Edge (3631, 36)
This hill was on Dewey's original list but more recent mapping had challenged its status. A line survey on 9 May 2008 by John Barnard and Graham Jackson estimated the drop as 26.64 ± 0.05m.

Milk Hill (2872, 39)
Briefly acquired Marilyn status in 1997 where it was erroneously listed in Section 42.

Botley Hill (2910/3686, 42)
Several people have suggested that there is higher ground within the large 265m contour to the west of the trig point. The 1" New Popular Edition map gives a spot height of 882m and on the strength of this evidence Alan Dawson announced a relocation in June 2008. However there may be some doubt as to whether this height refers to natural ground or to a covered reservoir marked on the map, the site of the more recent water tower. It is hoped to line survey the hill in the fairly near future.

The Access database

Tables View
This contains the Hills table in which the data are stored. The individual fields contained in the table are described in Description of fields.

Queries View
The queries are of four main types:
  1. Queries 1 - 43 that extract a list of hills from the database, e.g. Munros, Nuttalls, Deweys. Within the query, hills are usually sorted first on section and then on height. Not all fields are currently shown
  2. Queries 44 - 49 that sort the whole database on a specific criterion, for example revision date
  3. Unnumbered queries giving the hills within a hill list that have been climbed by the user
  4. Hidden queries used in the construction of Totals, which is found in Report View.
When selected, these queries open and display the information appropriate to their title.

Query 50 returns hills within a section that have not been climbed by the user and it operates slightly differently. Upon opening this query, the user is presented with a dialog box, which prompts for the Section desired. So if the user wants to know how many hills in Section 1A he or she has not climbed, then 1A must be entered into the dialog box. Note that if 1 is entered then all sections containing 1 as the first character will be returned, that is all of Section 1 and all of Sections 10 to 19. There are many ways in which this query could be set up; please email the authors if you feel the output would be more useful in another form.

The General query is provided for database maintenance by the authors. It includes all the fields in the Hills table plus four calculated fields: xcoord, ycoord, Streetmap URL calc, OS URL calc. In order to display the latter two fields as hyperlinks, their contents are copied into the corresponding fields of the Hills table prior to each release.

Form View
This contains one form: Hill lookup.
This view is useful for presenting all the key information about a specific hill. Hill lookup allows the user to see at a glance the lists it appears in, survey information, comments, links to Streetmap and OS Get-a-Map, in addition to height, grid reference, section, map and date climbed. A Find button enables a hill to be located. Place the cursor in the Name field before clicking the Find button, or in the Number field if the unique hill number is known. The button also permits * to be used as a wild card. For searches by name it is advisable to set Match to "Any Part of Field" to allow for alternative hill names. Bear in mind that common hill names may have many records.

Take care not to type in the form itself or you will overwrite the data!

Report View
This contains one report: Totals.
Totals allows the user to see at a glance how many hills within a list have been climbed. All major lists are collected together within a single report: Marilyns & SubMarilyns, Munros & Munro Tops, Corbetts & Corbett Tops, Grahams & SubGrahams, Murdos & SubMurdos, Donalds & Donald Tops, Hewitts & SubHewitts, Nuttalls, Wainwrights, Outlying Fells and Deweys. In addition the Marilyns are subdivided into Scotland, Wales, England and Isle of Man.

Recreating the original lists

Microsoft Access users will find queries for all the individual lists, including some sub-categories. Users of the Excel or text versions of the database can recreate the original lists by sorting on the relevant category field followed by Area (for Donalds, Nuttalls and Wainwrights), _Section followed by Section (Corbetts and Corbett Tops), or _Section (for everything else). In Excel, the easiest way of selecting hills belonging to a particular category is to use the Autofilter facility in the Data menu. This is already set up in the file.

We have not provided a mechanism for ordering the hills within each Area or Section to match the original printed list as it would take much effort to produce and maintain. For Marilyns and Grahams, the hill order in RHB/TACit can be reproduced approximately by further sorting in order of descending height.

User feedback and database enhancements

We welcome users' suggestions for enhancing the database, and of course reports of errors.

We are particularly keen to receive 10 figure GPS measurements from readers. Please see 10 Figure Grid References for the information we need.

We have had several requests to give the drop on each hill. It would have been easiest to do this at the time of the original compilation, but the reasons why we did not were as follows: Having said all that, the drop is a fundamental property of a hill and there is clearly a demand for the information. We will consider adding drop in a future release.

Acknowledgements

We would like to thank Alan Dawson for granting us permission to use his data; Anthony Hartry for providing 1:25000 OS map numbers; Rob Woodall for making his revision of the Deweys available on the rhb e-group; Simon Edwardes and Geoff Crowder for providing data and assisting with queries; Gordon Adshead for screening the grid and map references of the first version; and all those correspondents who expressed their support and suggested improvements.

We are grateful to TACit Press for putting its lists on the Web, which considerably reduced the effort involved in the original compilation.

We thank a number of correspondents for prompting some of the changes in the current and previous versions by identifying errors or assisting us with research.

Chris Crocker and Graham Jackson
Updated 30 June 2008

Return to Downloads

Home