Iterum ad aliam partem dicitur insula
|
Magantia |
North Uist? |
10918 |
|
Anas |
South Uist? |
10918 |
|
Cana |
Canna |
10919 |
|
Atina |
Eigg? |
10919 |
|
Elete |
Loch Ailort? |
10920 |
|
Daroeda |
Muck? |
10920 |
|
Esse |
Coll? |
10921 |
|
Grandena |
Tiree? |
10921 |
|
Maiona |
Mull |
10922 |
|
Longis |
Loch Linnhe |
10922 |
|
Eirimon |
Desert Island? |
10923 |
|
Exosades ubi et gemme nascuntur |
Barra, Sandray, Pabbay, Mingulay and Berneray? |
10923 |
Again, very few of these names can be identified with any certainty, but all appear to relate to the west coast between Skye and Kintyre. Why the Cosmographer saw this group as being in ‘another part (of the Ocean)’ is not at all clear, but may derive from the inaccurate location of islands on a severely distorted map. It is worth noting that Ptolemy places Σκητις far to the ‘east’ (i.e. north) of islands such as Μαλαιος, Mull. It may be that the Cosmographer’s source used the same data and separated Skye and the islands to its north from those in this group, placed closer to Ireland.
Cana perhaps survives as Canna, as recognised by Richmond and Crawford (1949, 27). Atina may be an error for *Ituna, although it is unlikely to be the River Eden in Cumbria; it may be the name of Eigg, Rhum having perhaps been named in the previous section. The two names which precede Cana, *Magontia and Anas, are perhaps North and South Uist, far to the west, as they cannot lie to the north.
<Elete> (for *Fl Ete) may be the Ειτιος ποταμου εκβολαι of Ptolemy (II.3, 1), in which case it is perhaps Loch Ailort, a name with no ancient forms. <Maiona> is probably an error for Malaea, Mull. Between these, the names *Daruueda, <Esse> and *Grandina may represent Muck, Coll and Tiree respectively. <Esse> may be related to the insula Coloso of Adomnan’s Vita Columbae, identified with Coll (Watson 1926, 84). Watson found Adomnan’s -s- puzzling, but it could be a reduced -ss-, the original name perhaps being something like *Colossa or, more likely in view of the Cosmographer’s truncated form, *Colissa. <Longis> is then the same as the Λογγου ποταμου εβολαι of Ptolemy (II.3, 1), Loch Linnhe, while <Eirimon> is explained by Rivet and Smith (1979, 197) as a transcribed rather than translated ερημος, ‘deserted’.
The final name, <Exosades> is a problem, and Rivet and Smith (1979, 363) recognise the impossibility of deciding between the alternatives they list. Sir Ifor Williams (in Richmond & Crawford 1949, 34) suggested that the name should be *Esocades, ‘salmon islands’, which is reasonable enough; Dillemann (1979, 72) thought it a garbled form of (H)ebudes; Schnetz (according to Rivet & Smith 1979, 363) emended Electrides, explaining the reference to gemmae as being to amber, and identifying them with the islands of that name in Pliny (Historia Naturalis iv.103), located in Scandinavia; Rivet and Smith also propose a muddling with Orcades, the Orkneys, which follow. Neither Dillemann’s nor Rivet and Smith’s proposals seem likely; Schnetz’s is ingenious, but we have no definite examples of Greek or Latin names for islands in the region. This leaves us with Williams’s emendation, *Esocades; it is possible that as the Cosmographer appears to have begun this section in the Outer Hebrides, he is here listing the group of small islands which form the southern end of the archipelago, Barra, Sanday, Pabbay, Mingulay and Berneray.