In the December 2006 Newsletter you ask about the end-points of the Ridgeway Base. The short answer is that this ran from White Horse Hill (SU 30083 86375) to Liddington Castle (SU 20982 79752). There are trig points marking the end points of the line, which is 11,260 metres (plus a few centimetres) long.

At that time (1935) the Ordnance Survey (OS) needed a new baseline in central southern England. The previous one north of Salisbury was no longer satisfactory as the end points were not intervisible (i.e. you could not see one from the other). The requirements for the new baseline included intervisible end points about that distance apart (most baselines were between 8 and 12 km), with open and fairly quiet country in between. (No M4 in those days!) No doubt the presence of the bridleway itself was an advantage, although of course the actual measuring would have been in an absolutely straight line, not along the Ridgeway Path. It may have been a bonus, though it would not have been necessary, that the baseline was not very far from 2° W, which is the 'origin' of the National Grid (the grid line at 2° W is the only one which points exactly true north).

I should explain that for two centuries or more, mapping of large areas depended on 'triangulation'. If all three angles of a triangle are measured accurately (which is, in principle, easy to do if all three corners are intervisible), and if the length of one side is known, then the lengths of the other two sides can be calculated precisely. Even if no length is known, the triangulation determines the ratio of all the lengths. So by triangulating (measuring all the angles) from the trig points across the entire country, all distances, directions and locations can be established if the length of just one side of one triangle is known. Thus it is vital to measure that one side as accurately as possible: that measured side is known as the Base or Baseline. (In practice the OS would measure at least one more Base, far removed from the primary one, to provide a check on the calculations.) At a local level, locations were found by reference to nearby trig points and bench marks.

For those who recall their spherical trigonometry, I just add that the angles of each individual triangle will sum to 180°, because any one triangle is flat. However the sum of all the angles recorded at any one point will be very slightly less than 360°, because the earth is spherical (or nearly so), not flat. The difference is trivial at any one spot, but (if not allowed for in the calculations) would be catastrophic over an area as large as the United Kingdom.

The Ridgeway Base was formally measured in November and December 1937, and was remeasured in October 1951: the two measures differed by only 6.6mm! The Liddington Castle end point, at 277m, was the highest of all the bases used by the OS.

Readers may be interested to know that the baseline was divided into 18 sections (the length varying somewhat with the terrain). Each section was then measured with tapes which were 24m long but only 3mm wide. Three or more readings of each 24m stretch (known as a 'bay') were taken and they had to agree to within 0.2mm before the surveyors moved to the next bay. How this was achieved in practice I cannot even begin to explain.

The Ridgeway Base did have some disadvantages. Several sections had steeper slopes than were regarded as ideal for a base; also at 249825 a ravine about 100m wide and 15m deep had to be 'bridged' and measured without prejudice to the overall accuracy. Odstone Barn at 280849 was exactly on the line of the base; this obstacle was dealt with by threading the tapes through holes cut in the walls of the barn! Finally the ditches and ramparts of the hill forts at each end presented problems especially when readings were being made in windy weather. Even so, the resulting distance was estimated to be accurate to within 1 centimetre.

Incidentally the Lossiemouth Base mentioned in your note had first been used in 1909 as a check on earlier work. (Until then there had been no base in Scotland, the previous 'second' primary base being near Lough Foyle in Ireland. Of course Ireland was mapped at a large scale several decades before Scotland.) But after remeasurement in 1938, it was abandoned. At only 7.17 km it was short, the ground was rough, and because it was close to and parallel with the Moray Firth it could only be 'locked in' to the triangulation from one side. Instead a new base was established in Caithness and measured in 1952. At 24.8 km this is by far the longest of any OS base. It is a testimony to the accuracy of the surveyors in those pre-GPS days that the length of the Caithness Base as determined by the triangulation all the way from the Ridgeway differed from the measured length by less than half a metre!

David Purchase

References
1  The Bases of the Ordnance Survey by Peter Haigh, Sheetlines no. 62, pp. 45 - 56 (2001). [Sheetlines is the journal of The Charles Close Society for the study of Ordnance Survey Maps.]
2 History of the Retriangulation of Great Britain by the Ordnance Survey (HMSO 1967). [Any reader fortunate enough to see a copy of this book will be rewarded by several photographs of the measurements of the Ridgeway Base, including the antics in the ravine, which to modern eyes are rather entertaining.]