The lost island of Saint Columba

Colonsay Cross

Sculptured cross from Riskbuie Chapel, Colonsay. Illustration from Allen & Anderson The Early Christian Monuments of Scotland (1903).

According to the vita or ‘Life’ of Saint Columba written by Adomnán at the end of the seventh century, the monastery on Iona had a number of satellites on various islands and coastlands around Argyll. One of these was on an island called Hinba and seems to have been the chief daughter-house of Iona. Adomnán tells us that it was founded by Columba himself and comprised not only a monastery but also a separate hermitage. Frustratingly, the precise location of Hinba is not made clear, so we are left to wonder which of the numerous Hebridean islands it was.

People have been suggesting possible identifications for Hinba for a long time, ever since modern historians first began to study Adomnán’s Vita Columbae. The obvious starting-point is to rule out those islands which are clearly identifiable in Adomnán’s narrative, such as Skye, Islay, Tiree, Eigg, Mull and of course Iona itself. None of these was Hinba, so the search is immediately narrowed. It also seems clear that Hinba lay at no great distance from Iona, for Columba was able to visit the satellite monastery quite easily and frequently. His uncle Ernán, who served as prior on Hinba, was able to undertake the sea-voyage to Iona when very elderly and in poor health.


The hermitage on Hinba was situated near what Adomnán calls Muirbulc Mar, ‘Great Sea-Bay’. As with some other places in Vita Columbae he gives the name entirely in Gaelic – his own native language – rather than rendering it into a Latinised form. Muirbulc Mar must have been a prominent feature, so any island without a large bay can effectively be ruled out in our search for Hinba. For example, the small island of Eileach an Naoimh, ‘Rocky isle of the Saints’, in the Garvelloch archipelago has been suggested as a possible candidate for Hinba but it doesn’t have a prominent sea-bay. Also, Hinba is a Gaelic name, so it is very unlikely that it would be given an additional or alternative Gaelic one. Indeed, it is far more likely that it today bears a name of Norse origin, as do many of the Hebridean islands.

The eminent place-name scholar William Watson proposed that Hinba derives from inbe, a Gaelic word meaning ‘incision’. In this context, the ‘incision’ would presumably be the great sea-bay of Muirbulc Mar. If Watson’s derivation is correct, the bay must have appeared to slice through the island, as if the sea had bitten a big chunk out of the coastline.

Only two candidates seem to tick all the boxes: Jura, which has a large sea-bay called Loch Tarbert; and the single island which is formed by Colonsay and Oronsay when the sea-bay between them is at low tide. Jura and Colonsay/Oronsay have Viking names, and we don’t know what they were called in Adomnán’s time. Jura has an early church dedicated to Columba; Oronsay has a medieval priory with a Columba dedication and an old tradition of having been founded by the saint. In favour of Colonsay and Oronsay is the observation that they are closer to Iona.

The upshot is that the puzzle of Hinba remains unsolved. This mysterious island, so important in the early history of the Columban familia or network of monasteries, seems to float beyond our reach. My own view is that it is now the single island formed by Colonsay and Oronsay at low tide, and that Oronsay Priory stands on the site of Columba’s monastery.

Oronsay Priory

Oronsay Priory

* * * * * * *


References to Hinba in Adomnán’s Vita Columbae occur at: Book 1, chapters 21 & 45; Book 2, chapter 24; Book 3, chapters 5, 17, 18 & 23. The Latin edition I use is the one edited by Alan and Marjorie Anderson in 1961 (revised in 1991). For an English version I use the Andersons’ translation and the one by Richard Sharpe for Penguin Classics (1995).

I discuss Hinba on pp.109-11 of my book on Saint Columba.

A useful summary of the various Hinba theories can be found on pp.91-102 of Alan Macquarrie’s The Saints of Scotland (Edinburgh, 1997).

Hinba is the island where Saint Columba narrowly escaped being murdered. The story is told in my blogpost Columba and the Pirates.

* * * * * * *

27 comments on “The lost island of Saint Columba

  1. Tim: this is not strictly to do with St Columba or Hinba but I wonder if you have read ‘The Ancient Paths’ by Graham Robb published quite recently by Picador? Perhaps you already know each other? Robb lives on the English-Scottish border. I have a feeling that your work and his may be complementary.

  2. Fantastic synthesis. I’d be inclined towards Jura, myself. Watson was dead on, I reckon. Old Irish ‘inbe/inbech’ means ‘notched’ and is attested in Cormacs Glossary. The inlet into Loch Tarbert is very ‘notch-like’…


    • Tim says:

      Thanks, Terry. I must admit to leaning slightly towards Jura whenever I look at the shape of Loch Tarbert, which does appear more notch-like than An Traigh between Colonsay and Oronsay. Also, as Watson noted, the loch is surrounded by several hermit-friendly caves, one of which seems to be known as ‘The Cave of the Iona Folk’. But what keeps me on board with Colonsay/Oronsay is their proximity to Iona and to the direct sea-route between Iona and Ireland, although this is based on an (unfounded) assumption that these factors were important to Columba as well.

  3. Jo Woolf says:

    I would instinctively opt for Colonsay/Oronsay, but I admit that the Jura option is very persuasive too! I have a book, ‘The Lands of the Lordship’ by Domhnall MacEacharna, which deals mainly with the place names of Islay but he gives his opinion that Hinba was Colonsay: “The Columban monastery stood on Kiloran Bay and the ruins were plundered for the building of Colonsay House which now stands on the site.” Apparently there was also a reference to ‘Hinbina insula’ (Little Hinb or Hinba), where there was a penitentiary, and he believes that this was Oronsay.

  4. Thanks Tim/Jo, Fascinating stuff. If only they were further apart! 😉

    Never heard about the caves, or Little Hinba. Was there any antiquarian attention given them before modern arch in any of the places?

    • Tim says:

      Jo – I’ve made a note to track down The Lands of the Lordship which sounds like a useful book.

      Terry – A quick search of Canmore hasn’t yielded much on these sites. Watson says the Jura caves were used by ‘gentlemen who were hunting’, while Canmore mentions herring fishermen. I can’t see much for Kiloran on Colonsay either, but the place-name suggests a link with St Oran of Iona

  5. Andy C says:

    Just started reading John Lorne Campbell’s “Canna, The Story of a Hebridean Island” (I haven’t finished it yet) and he makes claim that Canna could be Hinba. Does this theory have any supporters?

    • Tim says:

      Thanks for mentioning this, Andy. Yes, Canna is favoured by some of those who see Hinba as lying north, not south, of Iona. It does have a suitably large muirbolc or sea-bay, so it can’t be ruled out of the search. As far back as 1857, Reeves considered Canna one possibility (among others) in his edition of Vita Columbae. John Lorne Campbell lived on Canna for many years, as owner of the island, so his support for the theory is understandable.

  6. […] The lost island of Saint Columba […]

  7. George Campbell says:

    If you go to Kindle you can get my book “The First and Lost Iona.” All the answers are there with explanations.

  8. George Campbell says:

    Hinba is simply the Latinised version of Achinbady, once the non- Gaelic sounds are removed from each end

    • George Campbell says:

      Sorry, I should have said the non Latin sounds. The failure to identify Hinba is the direct result of an age old translation error to which I was alerted by Chisholm Batten’s “Charters of Beauly Priory.” He said that, before you translate anything, you must first identify whether or not the author was writing in his native tongue. If he was, then translate straight into English. If he wasn’t, then you MUST translate first into his native tongue before you translate into English. In this case, Adomnan’s native tongue was not Latin, but Irish Gaelic, so the word he would have translated into Latin as insula was the Irish innis. This doesn’t just mean island, but also meadow, water meadow, or what we in Scotland call an inch (from the Irish innis). Achinbady is the medieval name for Fort Augustus which was built originally on an inch and has a muirbolc mor (Borlum Bay) beside it which in turn is overlooked by the hill Mourvalgan, which linguists say is a later spelling of muirbolc mor.

  9. George Campbell says:

    Mourvalgan now appears on the map as Murligan and its origin is traced back to muirbolc mor at
    P.80 in Watson’s Celtic Place-names of Scotland.

    • Tim says:

      Thank you George for sending this information. Your theory is certainly a radical alternative to the conventional assumption that Hinba was an island.

  10. George Campbell says:

    No problem, Tim. I wouldn’t call it a theory, though, a theory being defined as something abstract constructed to explain something real in the absence of proof. Proof requires evidence and all I’ve done is to present the evidence. What others make of the evidence is up to them and I’d be very interested to hear any other interpretation of it. When I was writing the book, I was staggered by the amount of evidence which seems to have escaped attention for centuries. For example, the Book of Pluscarden’s Latin Index states clearly that Columba had two monasteries called Insula Iona. In his second book, Adamnan is at pains to point out at the start that in his first book the term Insula Iona is monastic slang for Mother House. So some of the stories in Vita are from when the Mother House was on Loch Ness side and some from when it was in the west. Thus, it would have been possible (and I’ve done it) to run from Iona to the chapel by the Awe and back within the hour. The ruins of the chapel are still there near the River Oich (formerly the Awe then Awich then Oich) close to the hamlet of Auchterawe. Give me a map of Inverness-shire and I will point out to you every single one of the so-called “lost place-names of Iona.”

  11. […] The lost island of Saint Columba […]

  12. George Campbell says:

    Is nobody going to offer even a jot?

  13. cully says:

    My guess is that Hinba is Eileach an Naomih in the Garvellachs.There are two deep narrow bays on the eastern side and would offer a sheltered landing place. The most northerly has a natural well called Columba’s Well..The monastery and bee hive cells are just above.The route from Colinsay to Iona is through the Torran Rocks, not the safest of passages unless in calm weather. My hunch is that Columba would have travelled the short distance by land to Loch Buie (from Bude) and from there, the Isles of the Seas are only a five mile crossing. St Bude was a friend of Columba and had his own place in Loch Buie. All speculation of course. Imagination is a great thing!

    • Tim says:

      Thanks for this, Cully. It’s useful to be reminded of the case for Eileach an Naoimh, not least because it was proposed by the great W.F. Skene himself. William Watson in CPNS wondered if Eileach an Naoimh ought rather to be identified with the otherwise unlocated Ailech Brenainn, an island associated with St Brendan. Like you said, the whole thing is a matter of speculation – and a continuing source of lively debate.

    • George Campbell says:

      Sorry, but I’ve explained why this is nonsense. Hinba is the Latinised Achinbady. Lack of imagination lets you down.

      • Tim says:

        For me, and I suspect for the majority of historians and archaeologists too, it would certainly require a huge leap of imagination to believe that Columba’s principal monastery lay anywhere other than on the Hebridean isle of Iona.

        Achinbady is surely one of the many Scottish place-names bearing the Gaelic prefix ‘Achadh na’ (‘Field of …’) like the examples quoted by Watson in CPNS. e.g. Achingall in Lothian.

        In any case, I tend to be wary of any theory whose author claims it to be The Great Truth rather than just an alternative interpretation of the data. A more circumspect, less strident approach to the topic under discussion tends to work better at this blog.

        • George Campbell says:

          When you say “surely” it means you have done no research. When you say “suspect” it means you don’t know. I never mentioned Achingall, nor Achnasheen, Achnacarry, Achnagairn etc. I said Achinbady and found that there was indeed one other hamlet of that name owned by the long extinct Mortimer family in Banffshire. Have you found any others? May I suggest you try reading to advance your knowledge? I recommend “De Lociis Sanctis” by Adomnan, wherein he confirms that the term “insula Iona” is a title, meaning “mother house” (as opposed to “insula Hinba,” which is a place). For your next read, try the Latin volume of The Book of Pluscarden, where you will find that Columba had two monasteries called Iona. You migtht also want to look at a map of Fort Augustus, where you will see that the huge bay next to it has a hill overlooking it called Murligan Hill. According to Watson, Murligan is a derivative of muirbolc mor.

          I was wrong to refer to imagination, as it is not my imagination which I followed, but the evidence. You were wrong to refer to my work as being a theory, as a theory is only what is produced in the absence of evidence.

          • Tim says:

            No, I used the phrase “I suspect” in the sense of “presumably”. I could press this point further by stating that there is no doubt whatsoever that the current academic consensus is overwhelmingly in favour of a Hebridean context for places like Hinba and Iona. But I assume you know this already.

            In using “surely” I was merely expressing appropriate caution on my suggestion that Achinbady may be a name derived from Gaelic achadh. Caution is always useful when dealing with early medieval history and is a position that I recommend you consider adopting. You’re absolutely right – I don’t know what the true derivation of Achinbady is. But neither do you, in spite of your unwavering belief in the certainty of your argument.

            Adomnan’s Vita of Columba makes such a compelling case for a Hebridean context that I feel no need to revisit DLS on this point. Nor do I see any need to consult the Book of Pluscarden. This fifteenth-century work, written 900 years after Columba’s arrival on Iona, is hardly a reliable voice on sixth-century geography. Like the chronicles of Fordun and Bower, upon which it relies heavily, Pluscarden is late medieval pseudo-history. It cannot be trusted on earlier periods. This is not an opinion I have invented but is rather the prevailing view among present-day historians.

            So your stack of “evidence” is indeed just a theory, regardless of however you choose to define the term. It is as much an unproven theory as any of my own radical ideas on a number of topics relating to Columba’s contemporaries among the Northern Britons. The difference is that I am prepared to accept that I might be wrong.

            In your area of research there really isn’t much to be gained from trying to convince people that you have a theory backed up by “proof” (this being rather thin on the ground in early medieval studies). It is just a theory (regardless of how stridently you vouch for it) until such time as the academic community decides to accept it as the new consensus.

            • George Campbell says:

              Sorry, but “academic consensus” is not evidence and evidence is not a theory. If you want invent your own meanings for words and close your eyes and mind to the evidence, that is entirely up to you. The academics to whom you refer are patently not linguists or they would have noticed the error in definitively translating “insula” as “island.” The first step in any the translation of any language is to identify whether or not it was the native tongue of the speaker. If so, then one can translate directly into English. If not, then one must first translate into the speaker’s native tongue and only then into English. In this case, Adomnan was not a native Latin speaker, but an Irish Gael, and so the word he would have translated into Latin as “insula” was “innis.” which doesn’t just mean island, but also water meadow or inch, thus undermining the fixation that Hinba was on an island. But you will doubtless ignore that fact, just as you will happily ignore the fact that EVERY one of the so-called lost place-names of Iona can be found around Fort Augustus. I have merely presented the evidence, and there is plenty of it. Fourteen years on, I am still waiting for an alternative explanation.

  14. George Campbell says:

    The speculation is over, having been replaced by my evidence, which a representative of one university has confirmed is now accepted. William Watson attended Inverness Academy with my grandfather, with whom he corresponded for the rest of their days, so I’m quite happy for both their sakes to have brought this issue to an end.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s