Ork 1750 3

The map by J & C Walker (1834) was on a relatively small scale but showed a close error correlation with several of the earlier 19th century maps. Despite this, place name evidence failed to indicate a close relationship with any of these maps.

In the late 1820s, J. Lothian published a map of Orkney in his “County Atlas of Scotland” but the version analysed here was dated 1835. Its links to Mackenzie have been weakened (error correlation 78%) and its outline appears to rely more on Thomson’s 1822 map (correlation 92%). Despite this only 45% of place names selected were spelt the same as Thomson’s and in fact the highest agreement with any of the preceding maps was only 49 percent. It appears that Lothian either had a free interpretation of spelling or had a separate as yet unidentified source of information.

In 1847, A & C Black produced an atlas stated by Moir (1983 p. 56, 3) to be a reissue of Lothian’s maps with addition of a number of names. The error correlation coefficient between the Lothian and Black maps of Orkney was 93% and 91% of the place names common to both maps were spelt identically.

A similar close inheritance of one map from another was apparent with the maps of Orkney by W. Blackwood 1838 and A. Fullarton 1855. Their error correlation was 93% (Table 8) and 89% of place names in common were spelt identically (Table 9). The origin of Blackwood’s map is not apparent from the study of error correlations but 75% of place names were spelt the same as on the map by Thomson.

The map of Orkney by T. Brown in his Atlas of Scotland 1807 gave an error correlation with Mackenzie of 95% again strong evidence that its outline was based on the Mackenzie map. Despite this, only two thirds of the selected locations in common had the same spelling of their place names. Surprisingly, he followed the example of Kitchen & Barber and Conder and based his longitude depiction on Edinburgh.

The next major map of Orkney was that published by J. Thomson in 1822 . This was on twice the scale of those by Arrowsmith, Brown and Scott and contained even more names than that by Mackenzie. Thomson adopted a system of submitting his draft maps to local dignitaries for comment and the map was published as attested to by those persons who are listed in its cartouche. Nevertheless, the map’s errors still show strong parallels with Mackenzie’s (86%) and a 98% correlation with Aberdeen’s (Table 8). In contrast, the comparison of place names showed fewer spelt the same as Mackenzie’s ( 59%) and only 51% the same as Aberdeen (Table 9). Since by this time some 70 years had elapsed from the publication of these earlier maps, alteration in place names is perhaps not surprising in an age before these became subject to standardization. Furthermore, reference to local authorities would be expected to result in an updating to current usage and hasten the demise of older versions current in the middle of the previous century.