Roy’s map is a national treasure and an important part of Scotland’s heritage. It is a document of significance to the Scottish Enlightenment and also a key part of the Ordnance Survey’s early history. It iis well preserved in the map room of the British Library where it is presented in all the original forms and is also now available on-line where great care has been taken to avoid any distortion so that it can be fully appreciated as a document of its time. Roy’s surveying colleague and friend, Hugh Debbeig, may have prophesied the true value: ‘...the greatest work of the sort ever performed by British Subjects and perhaps for the Representation of the Country not to be equaled (sic) in the World.’ However Roy himself was more critical:‘rather... a magnificent military sketch than a very accurate map of the Country, ....limited by Instruments of the common or inferior kind.’ In 1920 Sir Charles Close, Director General of Ordnance Survey also described the map as ‘little more than an elaborate compass sketch.’ However there were some very significant innovations: a genuine and comprehensive attempt was made to take measurements across an entire country; it was exclusively the outcome of a formally commissioned survey; the product was a map of a scale and detail unmatched for a further forty years, when the Cassini family at last produced their map of France. At the time, Roy's map was firmly at the leading-edge of the developing science of cartography.
The map describes a landscape in which little had changed since medieval times; however the terrain was about to become unrecognisable. Within forty years the agricultural revolution would impose enclosure and run-rig would disappear; the industrial revolution would take hold, emptying the countryside and expanding the cities; Georgian architecture would remodel urban layout; Turnpikes would re-fashion the road network. If Roy's map had been created any earlier there would not have been the skills or knowledge to make a record of such detail and precision; had it been made later it would have recorded a different picture. Curiously, the developing science of mapping was part of a revolution that itself would alter what had to be mapped. Roy's survey 'jumped the gun' and gave us a unique, detailed record of a lost landscape. This is the substance of its importance as an historical resource.
Georeferencing techniques permit a range of resizing controls and adjusment techniques to match old maps to modern equivalents; however in the case of the online version of Roy's map it seems that this treatment was limited to resizing along with a small amount of orientation and adjustment at the seams. In consequence a reasonable assessment of the accuracy of the old map could be made: it seems that on average a discrepancy of approximately 5 miles is evident at the greatest distance from the resize control points . This amounts to a distance-error of around 2%, at the limits of the map, when compared to the length of Scotland.
Despite this, the map is of a different caliber from its 17th century predecessors and the detail and contouring at local level certainly suggests a convincing improvement in usefulness. It is, of course, significantly less accurate than Cassini's Atlas of France, which was not completed until 1789, and cannot compare with the first edition OS maps which arrived 50 years after that. Unlike Cassini, Roy had not yet developed the process of ground-triangulation. He was using non-telescopic theodolites and simple measuring chains and the implied errors would be compounded over distance. There was no reference to latitude or longitude. The greatest shortfall appears to be absence of account of magnetic variation which alone would account for a 4 mile discrepancy over the distances covered. In practical terms the survey was completed over eight summers by a maximum complement of six survey parties: logistically improbable unless simple sketching was employed as much as careful measurement.
Whittington and Gibson have tested reliability by zoning in on a dozen small areas of the map for comparative analysis. They emphasise that much of the detail is much clearer on the protractions than in the fair copy. Some cities and town maps had been plagiarised: clearly no new survey had been done, in particular, on Aberdeen, although parts of Edinburgh are also suspect. Settlements seem more reliably depicted in location, though this is less so at house-block level and many settlements are missing. Placenames were often a guess by Scotsmen from the south, challenged by not speaking Gaelic; few of the residents to whom they would speak would be literate, so the surveyors’ spelling was imaginative. Field boundaries are little more than sketchwork and farm cultivation area is consistently exaggerated. They noted that Roy’s maps were the first to minimise the importance of great estates and fine castles. Overall it was a military map for troop movement and very fit for purpose : more accurate on rivers, lochs and roads and less so on property boundaries, fields and habitation. Whittington judges that the level of detail was in excess of that which would be expected for military use at the time. A contemporary comment was that ’on Roy’s Map the Ground looks as difficult to traverse as it really is.’