Κορδα/Corda

Attested:  (1) Corda at position 171 in the Ravenna Cosmography.   (2) Ptolemy 2,3,8 Κορδα, a πολις of the (Σ)ελγοουαι.

Where:  (1) At or near Kirkcudbright, where no Roman site is known, but it is reasonable to guess that an early port existed.  Located by its position in the Cosmography's sequence of names heading west along the top of the Solway Firth.  At Castledykes on the riverside at NX67715088 there used to be a huge medieval castle, occupying what would have been a prime location for a Roman fort.  Major native hillforts were nearby at Drummore and Carse Moat.
(2) At or near Castledykes Roman fort at NS92864426, on the upper reaches of the river Clyde, in “a meeting place of the main roads in southern Scotland”, and near the hillfort Cairngryffe Hill.  The error inherent in Ptolemy's coordinates amounts to tens of kilometres, so other sites are possible.

Name Origin:  Latin corda ‘cropping late in the season’, thanks to the generally mild (maritime) climate of this area.  (Of unknown deep etymology.)  Note the proximity to Ιηνα ‘warm’ (Wigtown Bay).  Delamarre (2017:37) pointed out that Corda was a common personal name in the ancient world, meaning ‘late-born’.

Notes:  A less likely explanation is that Latin corda ‘intestines’ fits the shape of Kirkcudbright Bay, just as the English word gut has a dual meaning: both intestines and also small creek or “narrow coastal body of water, channel or strait, usually one that is subject to strong tidal currents”.  This raises three awkward questions.  Is it safe to ignore the sense development of corda to cord, and hence to the string of a musical instrument or rope to bind a slave?  Did corda evolve from the same deep origin as gut?  PIE dictionaries currently distinguish *gherə ‘bowels’ leading to Greek χορδη versus gheu(d)- ‘to pour’ leading to Greek χυδην ‘poured out’.  And how far back in history can that usage of gut be traced?  There are plenty of related words, including gutter, gowt, French égout and goutte, but (at least in Germanic areas) they seem to have collided with words for gate (as in Kattegat, for example).  Gothic giotan ‘to pour’ existed in AD 350. Sailing directions (whether purely verbal, written down, or carved into wood) have long been a key tool of mariners entering unfamiliar harbours, and plenty of Roman naval supply vessels would have travelled from the Solway Firth, past Little Ross island, up long narrow Kirkcudbright Bay, and then into the bendy river Dee, on their way to the large fort and camps at Lucotion (Glenlochar).  The parallel with human anatomy (haemorrhoid, rectum, and colon, in shapes if not in scale) which is apparent on a modern map of site (1), would not have escaped a Roman sailor, with plenty of time to talk in salty language.  Carse Gut is not far away, as a branch off the estuary of the river Nith.  See also about Cordonovis, which might have been in Kirkcudbright Bay.  The name Kirkcudbright came from Cuthbert, where cuþ- became English couth ‘known’ but was confusable with guþ- ‘battle’.
Site (2) poses a different problem.  It lies a little downstream on the river Clyde from the main watershed of southern Scotland, in the Biggar - Romannobridge area, leading to the river Tweed.  A possible explanation for a Κορδα in that location may lie in some words by Selkirk (1995:251-2), wondering if some watercourses cutting across that watershed, now represented only by tiny streams and marshes, might have been artificially cut in Roman times.  Selkirk managed to annoy “establishment” archaeologists, so we post below a scan of the key paragraphs in which he tried to encourage amateurs to investigate the concentration of Roman forts and camps around the Tweed.  There are still uncertainties in that area about locating Venutio and Duabsisis.