A Rocky Problem: getting the rocks of the northern oceans on the map.

Carte Reduit des Iles Britanniques, sheet 4, by Jacques-Nicolas Bellin, 1757.

Carte Reduit des Iles Britanniques, sheet 4, by Jacques-Nicolas Bellin, 1757.

There are various small rocks lying to the north of Mainland Scotland and Lewis, and to the west of the Orkney Islands. The most prominent of these is North Rona, which may have been the site of a Christian Oratory as early as the 8th century. Nearby is Sulisker (Sula Sgeir), which may again have had early Christian settlements, and was certainly visited by men from Lewis in the mid-16th century. They caught young gannets there for food and feathers.

To the east of North Rona, lying approximately due north of the Kyle of Tongue, Sule Skerry can be found. It was thought to be the home of the Selkie - creatures that could transform themselves from seal to human form.

These, then were 'islands' that had been known about for many centuries. Getting them marked accurately on the map was not, however, a straightforward business.

The early maps of the north of Scotland were full of inaccuracies, none more so than those of Tommaso Porcacchi, found in his L'Isole Piu Famose del Mondo published in 1572.

A map engraved by Girolamo Porro for Porcacchi's L'Isole Piu Famose del Mondo, 1572. The islands are stewn about most carelessly, with Hirta (part of the St Kilda group) marked to the north.

A map engraved by Girolamo Porro for Porcacchi's L'Isole Piu Famose del Mondo, 1572. The islands are stewn about most carelessly, with Hirta (part of the St Kilda group) marked to the north.

One looks in vain there for islands as small as Sule Skerry. So too on Ruscelli's slightly earlier (1561) map of Scandinavia, which is dominated by the island of Thyle, which was a complete invention.

Girolamo Ruscelli's map of Scandinavia, which first appeared in 1561.

Girolamo Ruscelli's map of Scandinavia, which first appeared in 1561.

Gerard Mercator was altogether more precise: his map of Scotiae Regnum, first issued in 1595, marks both Rona, and what he calls Pfouil Skarre, Sule Skerry, though it is positioned too far to the west, and Rona, too far south.

Scotiae Regnum, by Gerard Mercator, 1595.

Scotiae Regnum, by Gerard Mercator, 1595.

This information was copied by other mapmakers throughout the 17th century, as can be seen on this one by Jan Jansson....

Scotia Septentrionalia, a map by Jan Jansson published in 1628.

Scotia Septentrionalia, a map by Jan Jansson published in 1628.

....and, later in the century, this superbly decorative map by P. Coronelli, published in 1692.

A map by P. Coronelli of Scotia, parte settentrionale, published in 1692.

A map by P. Coronelli of Scotia, parte settentrionale, published in 1692.

Detail from the above map.

Detail from the above map.

One of the first maps to mark Sula Sgeir was published in 1703, by Martin martin, in his Description of the Western Islands of Scotland. Calling it Sulisker, he unfortunately positioned the island to the east of North Rona, rather than to the west as it should be, and another error was passed on to later maps of the 18th century.

A detail from Martin Martin's map of the Hebdrides, this from a 1716 edition. Rona, with what is called Sulisker to its east, can be seen at the very top.

A detail from Martin Martin's map of the Hebdrides, this from a 1716 edition. Rona, with what is called Sulisker to its east, can be seen at the very top.

George Buchanan's History of Scotland was one of the first books to describe in some detail Sulisker. "Sixteen miles to the west [of North Rona] is Sulisker, a mile long, which bears no herb, not even fern...Hither the inhabitants of Lewis come in boats....[and] return loaded with fowls and their feathers." Given that Buchanan had made a special mention of Sulisker, it is surprising that Adair's map does not show the rock. Instead, he adopts the Mercator model, and shows Rona and Sule Skerry, though the latter is even further to the west. Is there some confusion in the cartographer's mind between the Sule Skerry of Mercator, and the Sulisker of Martin?

Detail of Adair's map found (only) in the 1627 edition of Buchanan's History of Scotland.

Detail of Adair's map found (only) in the 1627 edition of Buchanan's History of Scotland.

There are plenty of examples of maps which perpetuate Martin's error throughout the first half of the 18th century. Maps by Thomas  Kitchin...

The map of Scotland from T. Kitchin's Geographia Scotiae, 1756.

The map of Scotland from T. Kitchin's Geographia Scotiae, 1756.

....and Emanuel Bowen. On his map of 1747, he actually notes that these islands are sometimes absent, with a note:

"A late author affirms that these islands are in none but de Lisles's Maps,whereas they are in Moll's County Maps done 50 years ago, and in a map sold by Mr Garret a hundred years ago. His old friend Buchanan wrote of them above a century past, and Martin has them in his West islands, and calls that Solisker which lies east, being a rock about half a mile in circuit: so that instead of publishing a useful hint, this gentleman has said nothing to the purpose."

Detail from A New and Accurate Map of Scotland by Emanuel Bowen, 1747.

Detail from A New and Accurate Map of Scotland by Emanuel Bowen, 1747.

The 'late author' Bowen here is referring to is John Elphinstone, who made the above claim on his 1746 map. "Every navigator that has been on these seas knows the existence of them [Rona and Sulisker] and give them the same situation they have here." Unfortunately, on his map 'here' he marks them the wrong way round, with Sulisker (which he calls Barra I.) to Rona's east. (See the National Library of Scotland's fine online catalogue for Elphinstone's map).

Unfortunately, Bowen too persists with Martin's error and places Sulisker to the east of Rona.

"A New and Accurate Map of Scotland..." Emanuel Bowen, 1747.

By 1765, Thomas Kitchen had marked the three 'rocks' in their correct positions, but this map was by no means the first. 

|An \accurate Map of Scotland by Thomas Kitchin, 1765.

|An \accurate Map of Scotland by Thomas Kitchin, 1765.

Baba, or Sulisker, is now correctly placed to the west of North Rona. Why Baba? Possibly because the two islands Rona and Sulisker were part of the Barvas estate on Lewis, and remained so until 2016.

Detail from Kitchin's 1765 map.

Detail from Kitchin's 1765 map.

In fact, their correct placing seems to have begun in 1750. James Dorret's map, which is recognised as an important step forward in the mapping of Scotland just manages to squeeze Rona and Sulisker into the upper margin, and positions them accurately. It had taken nearly 70 years since the publication of Buchanan's History, in which he stated quite clearly that Sulisker lay to the west of Rona, to place it correctly on the maps of Scotland.

A detail from Dorret's map of Scotland first issued in 1750, this a reduced copy published in 1751. Rona and Sulisker are squeezed into the upper border. Dorret's 'New Survey' is recognised as being an important contribution to Scottish cartography.

A detail from Dorret's map of Scotland first issued in 1750, this a reduced copy published in 1751. Rona and Sulisker are squeezed into the upper border. Dorret's 'New Survey' is recognised as being an important contribution to Scottish cartography.

But, in fact, I suspect the man who had sorted the problem out once and for all was Murdoch Mackenzie, whose magnificent set of maps titled Orcades was also published in 1750. It included 3 maps of Lewis. On the chart of what he calls the North Part of Long-Island [Lewis], there is found, clearly marked "Sulisker or Bara", placed to the south-west of Rona:

Detail from

Detail from "The Lewis, or North Part of Long-Island, surveyed by Murdoch Mackenzie." 1750.

I suspect that Dorret had learned of the correct positioning of these rocks from Mackenzie, who had been working on the surveys of the Northern Isles for some time. In addition to positioning correctly the two rocks, Mackenzie's map of the Long-Island displays a number of profiles of Rona and Sulisker:

Detail from Mackenzie's chart of North Lewis, showing profiles of Rona and Sulisker.

Detail from Mackenzie's chart of North Lewis, showing profiles of Rona and Sulisker.

More profiles of Sulisker and Rona on Mackenzie's chart of North Lewis, 1750.

More profiles of Sulisker and Rona on Mackenzie's chart of North Lewis, 1750.

Seven years later, Jacques-Nicolas Bellin published his superb 4-sheet Carte Reduite des Iles Britanniques. Clearly the correction was established in the minds of all serious cartographers by 1757.

Detail from Bellin's

Detail from Bellin's "Carte Reduits des Iles Britanniques", 1757, showing the rocky islands accurately placed.

Anomalies still continued. Lieutenant Campbell's "New and Correct Map of Scotland" published in 1795 does not show Sule Skerry, but it does mark  a "Rock discovered 1785 by Capt. Maughan seen at 3/4 Ebb." 

Detail from Lt. Campbell's 1795 New and Correct Map of Scotland.

Detail from Lt. Campbell's 1795 New and Correct Map of Scotland.

Campbell's map has "Sulisker or Baba I." correctly placed to the west of Rona. The new rock is too far west to be Sule Skerry, and it is surprising that Campbell has failed to include the latter when there is so much other detail.

Other maps from this period offer slight variations on the name of Sulisker. For example, Cary (normally a very reliable and accurate cartographer) calls it 'Raba I.' on his 1789 Map of the Northern Part of Scotland.

John Cary's Map of the North Part of Scotland, published in 1789.

John Cary's Map of the North Part of Scotland, published in 1789.

Another map published in 1826 by George Virtue calls it Bura.

A map of Scotland published by george Virtue in 1826.

A map of Scotland published by george Virtue in 1826.

At least calling Sulisker Baba, or whatever, avoids the confusion caused by the similarity of the names of the two rocks, Sulisker and Sule Skerry. 

Dr Jacob King, an expert on Gaelic names informs me that he has a record of only one Gaelic name for Sulisker, Sùlasgair, and two names for Sule Skerry; A’ Chleit ‘the rocky outcrop’ and Cleit Shiùrisgil ‘the rocky out crop of Siùrisgil'. He thinks all these names are derived from the Norse for Gannet Rock, which explains the possible confusion between the two rocky islands.

And what of my favourite map of the North Coast, that of 1744 by the Rev. Alexander Bryce? His coast does not extend far enough to the west to include Rona and Sulisker, but there, hidden amongst a quantity of text is Sule Skerry.

Detail from A Map of the North Coast of Britain by Alexander Bryce, 1744.

Detail from A Map of the North Coast of Britain by Alexander Bryce, 1744.

A tribute once again to his accuracy.

Detail from Bryce's map, 1744.

Detail from Bryce's map, 1744.