and shrubs regenerate and fill the vacant space left in the canopy. Such species are more favourable to the roosting of crows than conifers. The island catches fire, perhaps because of the combustible nature of broom and gorse, which also have regenerated in the lighter conditions of the new woodland. The island becomes known as Eilean Loisgte. The writings of two travellers shed some light on the matter. Thomas Pennant, who journeyed through the Highlands in 1769, suggests that the island was wooded with ‘firs’, the vernacular for Scots Pine, whereas the geologist John MacCulloch, over fifty years later, mentions a scattering of trees and thickets of mixed species over the islands.
Pont’s cartography is not conventional in the modern sense. He gives the general disposition of the land, but not with geometric accuracy. Some of his sketches of mountains seem like rough perspective drawings, and the spontaneous quality of their draftsmanship suggests that they were drawn in the field. This contrasts to William Roy’s work in the 18th century, where the mass and void of mountainous areas is depicted symbolically by monochrome pen and ink hatching.
Before William Roy’s survey, Blaeu’s Atlas, based on Pont’s work, was the best map available. Yet a month after the Battle of Culloden, Captain Frederick Scott, writing from the prominent landmark of Castle Stalker in Appin, noted ‘this Place is not marked on any of our Maps’. He also found out that place-names differed between those shown on his charts and those used by local people in their daily lives. The 1745 rising demonstrated the need for an accurate map of the country. Cartographic accuracy, or the ‘quantifying spirit’, became an ideal for the Age of Reason.
William Roy was commissioned in 1747 to undertake a military survey of the Scottish mainland, beginning in the Highlands. Roy used the military roads constructed by Generals Wade and Caulfield after the 1715 rising, as a way of organising his work. Unlike previous maps, which were often little more than an amalgam of previous charts, Roy’s survey used actual measurement and traverse survey along the length of the new road network.
After summers in the field and winters collating the results of their work in Edinburgh Castle, in 1755, Roy’s team produced two versions, a sketch and a fair copy. The latter comprised 84 brown linen map rolls giving 38 folding sheets. The complete map measured 20 by 30 feet. This was an impressive achievement accomplished in so short a time. However, as the work was not related to longitude or latitude and the maps were not aligned to true north, mistakes, after many traverses and successive offset measurements, accumulated steadily. The east end of Loch Leven in Fife is shown 20 miles south of its true position. The maps also lack features which had no strategic military value. Informed guesses were made about the nature of the landscapes remote from the military roads. Most of the islands were also omitted. Roy himself thought the survey was more of a ‘magnificent military sketch than a very accurate map of the country’. Place-names were not recorded systematically, and where they were, this was not done with any apparent knowledge of Gaelic. Some phonetic renderings are shown, which do have a certain value, as they reflect local pronunciations of the time.
Roy’s ‘magnificent … sketch’ was stored away in a cupboard and never seen until its rediscovery in the early 19th century. Maps were state secrets. Yet in this seminal work lay the origins of OS, formally established in 1791. Their mapping of Scotland continued fitfully from 1819 and at various scales until its completion at the end of that century. In Ireland matters were conducted more swiftly, motivated by the aim of maximising tax revenue.
As part of their work in Ireland, the OS established rules for collecting, recording and mapping place-names. These were subsequently applied to the Scottish Highlands. The following extract comes from ‘Instructions for the Interior Mapping of Ireland’ drafted by Thomas Colby, OS Superintendent in 1825.
The names of each place is to be inserted as it is commonly spelt, in the first column of the name book: and the various modes of spelling it used in the books, writings ... are to be inserted in the second column, with the authority placed in the third column …
(Withers 2000 535)
Rules were also laid down for choosing ‘authorities’ to authenticate the place-names collected.
For names generally the following are the best individual authorities and should be taken in the order given: Owners of property; estate agents; clergymen, postmasters and schoolmasters ... Small farmers and cottagers are not to be depended on, even for the names of the places they occupy …’
(ibid 535)
This exclusive approach, reliant on informants from the landowning and professional classes, meant that those who worked the land and were closest to it were marginalised in the formal recording of place-names. As a result, local variants in naming were ignored. Subtleties of the spoken language were mistakenly represented as fixed and constant. What was not standard became accepted as standard. Parts of the map also remained blank for no other reason than an understandable reluctance of the informants to give freely of information, which the Government might use to their disadvantage. In some cases, settlements of native Irish were omitted. Those who worked the landscape were either mapped off, or mis-mapped onto a paper landscape. Problems of representation were more complex still, since those collecting information had no influence over its final transcription in OS’s Southampton HQ, giving further scope for misrepresentation.
In the Scottish Highlands, mapping recommenced in earnest on Lewis in 1846, for no other reason than that the landowner, Sir James Matheson, wanted an update on the resources of his estate. Rules developed in Ireland about gathering place-names were extended to Lewis. Alexander Carmichael was one of the ‘authorities’ commissioned to ascertain the authenticity of names. He was a Gaelic-speaking historian and folklorist from Lismore and thus an exemplary ‘authority’. Today he would be called an ethnologist. Carmichael wrote about what happened to his painstaking work, once it was beyond his control.
I have gone to the locality and in every instance corrected the place-name from the living voice on the spot. From these corrections I have written out each name in correct Gaelic and have revised and re-revised my own work. I have adhered strictly to the local sound and pronounciation of every word. Well then fancy my mortification when Cap. MacPherson tells me that he means to adopt neither Norse nor Gaelic theory in spelling but to give the name in phonetic spelling.
(ibid 547)
In a subsequent letter he comments further on the fate of his translations:
... I found that many ... place-names which I was at so much pains and expense in collecting were entire left out ... that some names on the old maps were left unaltered in form thus lending the meaning different. I took the liberty of drawing the attention of the Dir G of the OS to these alterations and the reply was that the names were omitted to save expence that old names were left out as they were obviously incorrect and that the final mode of spelling rested with the Inspector General.
(ibid 547)
We must treat maps with caution. They reflect the society of the mapmaker and the way the landscape and its inhabitants have been interrogated. Nevertheless, over most of the Highlands, now bereft of native speakers and without further knowledge coming to light, today’s mapped record is arguably as good as it gets. This may remain true even after consulting original name books and manuscripts of earlier maps and charters. For this study, it has not been possible to substantiate the deep authenticity of the many thousands of names listed. Places mapped by OS represent a surface layer. But it is the surface layer which we use.
Later OS maps include information about landform, augmenting spot heights with contours. These connect points of equal elevation. Before leaving this chapter it is worth recounting how an early use of contours was employed during the ‘Schiehallion experiment’. After Newton’s discourse on the universal theory of gravitation, there was much scientific debate about the shape and mass of the earth.
