by William Matheson
18th November 1977
Transactions of the Gaelic Society of Inverness, Vol. LI (1978-80), 1981, pp. 68-80
[NOTE: Click on the Footnote Numbers to read the footnotes; Click on the number again to return to the main body of text.]
It would appear that histories of the MacLeods were not written until later than similar works relating to neighboring clans such as the MacDonalds and the MacKenzies. The Bannatyne Manuscript, in the muniment room of Dunvegan Castle, dates only from about 1830, though in incorporates some material of earlier provenance.  Current ideas about the ancestry of the clan derive largely from this work, and there is need to examine the question anew, making use of other sources that have become available in more recent times.
The MacLeods have for long claimed lineal descent from the Norse kings of the Isle of Man. But there is some disagreement on the details. Just over three centuries ago the first Earl of Cromartie included in his MS history of the MacKenzies a genealogy of the Macleods of Lewis, from whom he was descended on the distaff side; and according to his account, Leod, the eponymous ancestor of the clan, was a son of Harald, son of Godred Don,  who usurped the throne of Man in 1249.  But in both Douglass Baronage  and the Bannatyne MS  there is a different story: Harald is discarded, and Leod is made out to be a son of Olaf the Black, king of Man, who died in 1237.  These two versions of early MacLeod genealogy are unsupported by any evidence; and, not only so, but they are flatly contradicted by the other sources referred to above -- sources that are preferable by reason of their earlier date, , and also because, being written in Gaelic, they are likely to represent authentic native tradition, unaffected by extraneous influences.
There is, first of all, the Kilbride MS, no longer extant, which is to be dated c. 1550.  It gave Leods pedigree as Leod mac Oloig mic Oib mic Oilmoir mic Iamhar, and so on back to Iamhar Athacliath (Ivar of Dublin).  A century later the Irish genealogist Duald MacFirbis writes: Leod mac Gillemoir mic Raice mic Olbair snoice mic Gillemoire.  And about the same time Cathal MacVurich, poet and historian to MacDonald of Clanranald, sends a translation of Thomas à Kempiss Imitation of Christ to Sir Norman MacLeod of Berneray, accompanied by an address in which Sir Normans ancestors are traced back for fourteen generations, the last five generations reading as follows (with the first name here put into the nominative case): Leod mac Olbuir mic Raisi mic Olbuir Snáithe mic Aonghusa. 
At first sight, this evidence seems difficult of interpretation. The three lists of names as they stand do not wholly coincide. Yet, even as they stand, they make one assertion possible, namely, that there is no trace whatever in any of them of descent from either Harald Godredson or Olaf the Black. These names are just not there. But they should nevertheless be borne in mind; for there is still a need to consider the possibility of descent in a female line.
In the meantime, however, let us look more closely at these Gaelic genealogies, taking first the question of Leods great-great-grandfather. Of him it is clear that they tell us nothing, giving, as they do, different names -- Iamhar, Gillemuire and Aonghus -- that cancel each other out, and merely indicate that, when the genealogies were compiled, there was a significant lack of reliable information as to his identity.
The case is different with regard to Leods great-grandfather. Allowing for vagaries of spelling, and remembering that in Gaelic orthography b and m when lenited (bh, mh) represent the same sound, there is complete agreement. it is also clear what the name is: none other than Olbhar, that of one who is frequently mentioned as an ancestor in praise-poetry relating to the MacLeods -- Olbhar in the literary language,  and Olghair in vernacular Scottish Gaelic.  Not only so, but in such poetry we find it stated that he was a Norseman;  so that it is safe to equate Gaelic Olbhar with Norse Olvir. And it will be argued in what follows that this equation is a vital clue to the true ancestry of the MacLeods.
The Orkneyinga Saga tells of certain happenings in Caithness and Sutherland during the years 1135-1139, as a result of which one Olvir, known as Rósta or the Unruly, fled to the Hebrides, to be heard of in his former haunts by the Helmsdale river no more.  And further reference to him should presumably be sought in the Isles, and the primary purpose of this paper is to show the grounds for thinking that it is this same Olvir who turns up in the Gaelic genealogies as Olbhar, great-grandfather of Leod, eponymous ancestor of the Clan MacLeod.
What are these grounds? There is, first, the fact that Olvir fled to the Hebrides and did not return from there. Then, again, there is the date of his flight -- c. 1130. It is difficult to find dates on the Hebridean side of the equation; but we know that Malcolm shown in the genealogies as grandson of Leod, had a charter of lands in Glenelg from David II, apparently not long after the latters return to Scotland from France in 1341.  We do not know Malcolms age at the time, but if we strike an average and suppose that he was in middle life, he could have been born c. 1290, and his grandfather Leod at least sixty years earlier. It is therefore evident that the floruit of Leods great-grandfather Olbhar could be placed as far back as c. 1139, making him contemporary with Olvir Rósta, who appeared in the Hebrides at that time. And this leads to the third consideration, which suggests that he was, not just a contemporary, but the same person. Leod, Gaelic Leòd, Norse Liótr, is in Scotland a name peculiar to the MacLeods, though almost unknown among them except as that of their eponymous ancestor. It is also to be remarked that the name is rare, indeed exceedingly rare, even in Scandinavia and Iceland. So it is of great significance for the purpose of the present enquiry that this was the name of Olvir Róstas maternal grandfather, known as Liótre Nidingr or the Worthless.  And there is a fourth consideration. The Gaelic genealogies show some agreement back to Leods great-grandfather Olbhar. Their total disagreement with regard to the previous generation is an indication of ignorance; and this is exactly what we should expect, given that Olbhar appeared in the Hebrides as a newcomer of Norse extraction whose antecedents would be understandably obscure from the Gaelic genealogists point of view.
That the above account reveals the true ancestry of the MacLeods is a view reinforced by a study of their long standing claim to descent from the kings of Man. The claim has to be taken seriously if only because of frequent references in eulogy and elegy relating to the clan. One ancestor to whom reference is made is Magnus, whom Neil MacVurich, in his elegy to Sir Norman MacLeod of Berneray, identifies as Maghnus ó mhúr Manainn  (Magnus from the house of Man). This could be Magnus, king of Man, son of Olaf the Black. If so, it is curious that Olaf Himself never rates a mention by the poets. This was possibly because he may have been the eponymous ancestor of the Clan MacAulay (Clann Amhlaidh) of Lewis, who, in any case, had pre-empted the name, thus causing it to be shunned in poetry celebrating the MacLeods.
But there is a larger question. Given, as the poets aver, that the MacLeods were descended from the royal house of Man, the evidence of the Gaelic genealogies indicates that this, if true, could only have been in a female line. How, then, did a different view of the matter come to be proposed? Deliberate misrepresentation for reason of social prestige is a possibility not to be ruled out; but, in fact, what happened can be accounted for otherwise. The poets and shennachies were certain of two things: that the MacLeods were descended from the kings of Man, and that they had an early ancestor called Olbhar. Then a new element entered their calculations in the form of a literary work from outside their own tradition. This was the famous Britannia of William Camden, first published in 1586,  copies of which reached the Highlands at an early date. Thus, for example, the Earl of Cromartie  and Hugh MacDonald, the Sleat historian,  both writing in the reign of Charles II, refer to Camden as an author with whom they were acquainted; and so also does their contemporary the Rev. James Fraser in the course of writing the history of his own clan.  Camdens work contained an abridgement of the Chronicle of Man, and it may be imagined that the MacLeods in particular would look to this source for more light on their own early history. Unfortunately, the information gained from Camden led, not to illumination, but to serious misapprehension. All the indications are that, having met with the name of Olaf the Black as Olavus in the Chronicle of Man,  they mistakenly identified Olavus with Olbhar or Olghair, so often mentioned as an ancestor of the MacLeods by the poets and shennachies of the Isles. There is ample confirmation of the error, for example in the fact that when the name Olghair was resuscitated after a lapse of centuries, as in the case of Olaus MacLeod, tacksman of Varkasaig in Glendale, the usage of Olghair in Gaelic but Olaus in English;  and Olaus, also found in the Bannatyne MS, is a slightly modified form of the Chronicle of Mans Olavus. Here, then, we discover that the claim to direct lineal descent from Olaf the Black can be accounted for as the result of an erroneous equation between Latin Olavus (Norse Óláfr) and Gaelic Olbhar or Olghair: Erroneous, because Olbhar or Olghair is the Gaelic form, not of Norse Oláfr, but of Norse Olvir, as found in the Orkneyinga Saga; while Norse Oláfr (Latin Olavus or Olaus), on the other hand, becomes in Gaelic, not Olbhar or Olghair, but Amhlaoibh, and nowadays, to indicate current pronunciation, Amhlaidh. It may be added that this cardinal error can be traced back to within little more than a generation after the appearance of Britannia; for we find the chief of the MacLeods styled John McOlaus of Dunvegane in a document dated at Edinburgh, 11th August, 1630. 
Other names in the Gaelic genealogies provide further evidence on the ancestry of the MacLeods. It is fairly obvious that, not only Leods great-grandfather, but also his father, was called Olbhar. It would be in accordance with custom for him to be named after his grandfather, especially if he was the eldest son. MacFirbiss Gillemuire can be accounted for by supposing that a form such as Oilmoir in the Kilbride MS was mistranscribed as Gilmoir, followed by subsequent modification of spelling. In this connection it may be added that the Rev. James Fraser, already mentioned, describes Leod as the sone of Oliverius Norwegie,  where Oliverius is a Latinised form of Gaelic Olbhar.
But the name that deserves special attention is that of Leods grandfather. In the Gaelic genealogies it is noticeable that the variations as between one manuscript and another are to be found exclusively in the names of Norse derivation. The reason is obvious. The writers could transcribe the Gaelic names, with which they were familiar, with comparative confidence and accuracy. But the unfamiliar Norse names left them somewhat baffled. Thus, Oib in the Kilbride MS is a most improbable form,a nd should no doubt be put down to mistranscription. The forms in the other two manuscripts, Raice and Raisi, are very close, though unrecognisable as they stand. However, if MacFirbiss Raice is emended to Paice, we have a very significant name indeed; for Páice is the Gaelic form of the Norse personal name Bálki. It occurs in place-names: Tobhta Phàic on the west side of Lewis at Borve;  Gil Mhic Phàic on the eastern shore of Loch Seaforth in the Park district of Lewis;  and in Skye Dùn Phàic near Kilmore in Sleat.  But more to the present purpose is the fact that, early in the thirteenth century, Pàll son of Bálki, or Paul Balkason, was deputy-governor of Skye (vice-comes de Ski) in the Norse kingdom of the Isles, appearing as such on record in 1223.  He has also been recognised as the Pàll son of Bálki by name, who, according to the saga of Hakon Hakonson, was slain in 1231.  The same source mentions as son of his, Bálki by name, who was a grown warrior by that date.  Chronologically, therefore, Paul Balkasons father Bálki is to be assigned to the same period as Páice in the emended Gaelic genealogies, and, as will appear in what follows, there is reason to suppose that he is to be identified as the grandfather of Leod, eponymous ancestor of the MacLeods, who b the same token must have been a nephew of Paul Balkason.
A later piece of evidence on this point turns up in the Fernaig Manuscript. In or soon after 1693, the writer of the manuscript, Duncan Macrae of Inverinate in Kintail, composed a song directed against the Blind Harper, Roderick Morison, in which he refers to Roderick, chief of the MacLeods, as éighre (i.e. Oíghre) Shìol Phàic,  meaning that he was the heir of the MacLeods, whom he designates the progeny of Pàic.  There is nothing very unusual in such variations of nomenclature. Various Highland clans, and even individual members of clans, in the days before registration standardised usage, are found identified at different times by different surnames, because more than one person in the direct line of ancestry could serve as the eponymous.  And so, Duncan Macraes choice of nomenclature is evidence that, if Leòd was one name in the direct line of ancestry of the MacLeods, Pàice was another.
Of course, all this was unacceptable to the MacLeods once the idea gained ground that they were descended from the kings of Man in unbroken male succession. Not that the existence of Pàice was unknown or denied. Like that of Olaf the Black, the name of Paul balkason, in the form Pol filius Boke, was accessible to them in the pages of Camden. The author of the Bannatyne MS calls him a man of great power and authority, and that is pretty well a translation of the description vir strenuus et potens in the Chronicle of Man.  But it is also clear that there was some traditional knowledge  of his name and that of his father, as is shown by the following passage in the same source: Paul had a natural son whose descendants for several generations held the lands of Bernera and several other places in Harris of the MacLeods. In the course of time they fell into decay and a few peasants only now remain of a race once numerous and powerful. They are called Clan Vic Phaick and are considered a fierce vindictive tribe who prided themselves on their descent from Paul.  Some things in this statement need not be taken at face value. The imputation of bastardy was a typical gambit in a situation where one branch of a clan was -- or had bee -- in competition with another.  And in this case the admission that Clann Mhic Phàice were once numerous and powerful is of some significance. As to falling into decay, while it is true that they are no longer numerous and powerful, it is interesting to find that they are by no means extinct. There have been down to the present day in the island of Berneray families with Mac Phàice as surname in the speech of the local community, though rather unaccountably known for official purposes as Mackillop.  Their authentic traditional surname is presumably to be traced back to Paul Balkasons father Bàlki, or to his son of the same name. To this may be added the probability that the MacPhails, originally of the Sand district in North Uist and of Carloway in Lewis, took their name from Paul himself. 
The MacLeod historians were ready to concede that Paul Balkason was closely connected with the clan near its beginnings. But, once connected to the idea of direct lineal descent from Olaf the Black, the connection had to be played down, and so the Bannatyne MS represents Paul as only a foster-relative. We are told that he was Leods foster-father, and that he bequeathed his lands to him because he himself had no legitimate issue.  A most unlikely story; foster-relationship did not work like that. And, besides, legitimacy or the lack of it was not a great issue in the Hebrides at that time or for several centuries thereafter. There are many examples of sons succeeding to their fathers possessions after being legitimised, or even while still illegitimate in the eyes of the Church.
It is possible, however, that Paul Balkason acted in a capacity rather similar to that of foster-father, namely, as guardian or tutor to Leod, if, as the Bannatyne MS has it, the latter was a minor at the time of his fathers death. This was a relationship between uncle and nephew that circumstances often made necessary; and the conjecture is attractive in the present instances because it would explain why it is Paul Balkason who figures so prominently in contemporary sources, and not Olbhar (Olvir), assuming that that was the name of Leods father. It would also absolve the MacLeod historians from the charge of deliberate misrepresentation, for in oral tradition the roles of foster-father and tutor could very easily be confused.
Further evidence regarding the ancestry of the MacLeods may be enbedded in the tradition of a dynastic marriage that took place early in their history. According to the Bannatyne MS, Leod, the eponymous ancestor, married a daughter of Mac Rhaild Armin (mac Arailt Armainn),  that is to say, the son of Harald, with the title àrmann, a Norse-derived word meaning a steward,  which in the Hebrides came to denote a member of the local aristocracy. The account states that he possessed the lands of Minginish, Durinish and Bracadale, and that he belonged to a Norman (by which the author means Norse) family who had built a fort on the site now occupied by Dunvegan Castle. In due course, so we are told, the fort, together with the lands just mentioned, passed into the hands of his son-in-law Leod.
It is interesting to find that this is not the only account of an ancestor of the MacLeods marrying into the house of a man named Harald. The MacFirbis genealogy, referring to Leods alleged great-great-grandfather Gillemuir, has the following: Ealga fholtalainn ingean Arailt mic Semmair (rect Sen Imair)  righ Lochlan mathair an Gillemuire sin;  i.e., Helga of the beautiful hair, daughter of Harald, son of Ivar the Old, king of Norway, was the mother of that Guillemuire -- a statement that calls for comment in the light of what has gone before.
Modern folklore studies recognise the existence of variants, and it looks very much as if what we have here are two variants of a tradition relating to the same marriage. The much earlier of the two is likely to be nearer the truth in indicating that it was not Leod who was a party to this marriage, but an ancestor of his several generations farther back. Not the father of his great-great-grandfather, however, of whom, as we have seen, the genealogists had no real knowledge. The most plausible conjecture is that it was the formidable Olvir Rósta who married Helga, heiress of Harald, and by so doing succeeded his father-in-law in possesson of Dunvegan and adjacent lands in Skye. As for Leod, it can be readily understood, once he was established as eponymous ancestor, how his name would come to be substituted in the story. Almost inevitably, there would be some confusion between the first of the race to appear in the Isles, on the one hand and the eponymous ancestor of the clan, on the other.
But what of the connection with the royal house of Man? There is, first, the fact, for what it is worth, that the MacLeods quartered, with their own coat of arms, the arms of the ancient kingdom of Man; though significantly, this feature dates only from the seventeenth century, and is absent from earlier representations of their arms.  Apart from that , unfortunately all we have to go on is the frequent mention by the poets and shennachies of such a royal connection, and of Magnus in the Isle of Man as an ancestor of the MacLeods, leaving us merely with the conjecture that Leod may have married a daughter, or sister, of Magnus Olafson King of Man, the latter the more feasible possibility on chronological grounds. Such a union with the blood royal would do much to enhance his prestige and help to ensure thant this should be the name commemorated in that of the Clan MacLeod.
It is unsatisfactory that the evidence on this aspect of the early genealogy of the MacLeods is so meagre and inconclusive. But two statements about their ancestry can be made with reasonable confidence, one negative and the other positive. On the one hand, the MacLeods cannot be lineally descended from Olaf the Black. His name (in Gaelic Amhlaoibh) is conspicously absent from the old Gaelic genealogies, and was never mentioned until the seventeenth century,  who, as a result of reading Camdens Britannia, Olaf (in Latin Olavus) was wrongly taken to be the same as Olbhar, a name that does appear in these genealogies. On the other hand, there is good evidence, of the circumstantial kind, that this Olbhar, always acknowledged by the MacLeods as an ancestor of reknown, is to be identified as Olvir Rósta, described in the saga as the tallest of men, and strong of limb, exceedingly overbearing, and a great fighter;  a man in the heroic mould, long to be celebrated by poets as fit progenitor of a warrior race.
1. For a transcript of the Bannatyne MS I am indebted to Mr. Alick Morrison.
2. I.F. Grant, The MacLeods, 21.
3. Sir William Fraser, Earls of Cromartie II. 510-511.
4. The Chronicles of Man and the Sudreys (1874) (ed. Munch and Goss), I. 102.
5. Douglas of Glenbervie, The Baronage of Scotland, 374.
6. See also R.C. MacLeod, The MacLeods of Dunvegan, 25-26.
7.The Chronicle of Man I. 94.
8. The Gaelic genealogies written in the sixteenth and seventeenth centuries represent an older manuscirpt tradition; mistranscriptions bear witness to copying from earlier sources.
9. MacKinnon, Catalogue of Gaelic Manuscripts, 219.
10. Skene, Celtic Scotland III. 460.
11. Skene, op. cit, 482.
12. RIA MS E i 3; Eigse X, 270. Captain F.W.L. Thomas thought the epithet snoice meant the Hewer (Proceedings of the society of Antiquaries of Scotland XIV (1879-80), 364n); if so, it would be the genitive case of the verbal noun of snaidh, ie. snaidhte. But the explanation is unconvincing, especially in view of the fact that Cathal MacVurich has an accent on the first vowel -- snàithe.
13. Reliquiae Celticae II, 266, 274; Scottish Gaelic Studies VIII. 30, 38, 42; Féil-sgribhinn Eóin Mhic Nèill, 172; and see references in Bardachd Ghàidhlig (3rd ed.), 305.
14. Watson, Gaelic Songs of Mary MacLeod, II. 521, 698, 791, 875, 1148; Rev. James Fraser, Chronicles of the Frasers: the Wardlaw Manuscript (Scottish History Society), 40.
15. E.g., Mary MacLeod sings of sliochd Olghair is Ochraidh/ o bhaile na Bierbhe (seed of Olvir and Ochraidh from the city of Bergen) (Watson, op. cit., 1. 698). Is Ochraidh to be accounted for as a mishearing by her of Crú Olbhair diomad orchra in Cathal macVurichs elegy to John MacLeod of Dunvegan, who died in 1649? (Féil-sgribhinn Eóin Mhic Néill, loc. cit.)
16. Taylor, The Orkneyinga Saga, 263-264.
17. The Book of Dunvegan (ed. R. C. MacLeod) I. 275; Robertson, Index of Charters, 48, 99, 100.
18. Taylor, op. cit., 214.
19. Reliquiae Celticae II. 270. Mary MacLeod also describes the MacLeods as sliochd solta bh air freumh Mhànuis (a puissant race of Magnus stock) and asserts that the chief of the clan was de shloinneadh nan rìghrean/leis na chìosaicheadh Manainn (from a line of kings who laid Man under tribute)(Watson, op. cit., II. 367, 695-6).
20. Britannia, Sive florentissimorum Angliae, Scotiae, Hiberniae, et Insularum adjacentium ex intima antiquitate Chorographica descriptio . . . L(ondini), 1586. The work rapidly passed through several editions, with a translation into English in 1610.
21. Sir William Fraser, op. cit., II. 510.
22. Highland papers (Scottish History Society) I. 8.
23. Rev. James Fraser, op. cit., 485.
24. Chronicles of Man I. 82 ff.
25. Mackenzie, History of the MacLeods, 265; Mackinnon and Morrison, The MacLeods: the Genealogy of a Clan -- Section Two, 114; information from the writers uncle-by-marriage, the late Donald Morrison, a native of Glendale, and others in Skye. It may also be noted that the Rev. Donald MacDonald, minister of Barvas, Lewis, refers to a site in the district of Ness called Caistel Olgre (i.e. Caisteal Olghair), with the translation Olaus his Castle (Old Statistical Account XIX. 270).
26. Ante, Vol. XXXVIII, 398.
27. Rev. James Fraser, op. cit., 40.
28. Ante, Vol XXXVI. 370.
29. Ordnance Survey One-Inch Map, Sheet 18.
30. Old Statistical Account XVI. 538.
31. The Chronicle of Man I. 86, 189. Cf. note 39, post.
32. Anderson, Early Sources of Scottish History II. 478.
33. Ibid., 474.
34. Làmh-sgrìobhann Mhic Rath (ed. Macfarlane), 265, 322; (Matheson, The Blind Harper, 176, 107ff.
35. Matheson, op. cit., 107.
36. Cf. ante, Vol. L. 62-63.
37. Op. cit., I. 86.
38. There were memories of Pàice (described as Fear Caisteal Eilean Chaluim Chille) in Kilmuir, Skye, early this century. See a collection of traditons, knocked together with an extraordinary disregard for chronolgy, by I.N.M., in an Sgenlaiche I. 44053. Here Pàic is erroneously rendered as Pàire. These traditions, in their original disparate state, were collected by the late Dr. D. J. MacLeod when teaching school at Kilmuir.
39. The passage shows a realisation that Paul filius Boke in the Chronicles of Man was the same as the Pal mac Phàice known in Gaelic oral tradition. In fact, Paul is indentified as Paul Mac Bok or Phaick, and called Sheriff of Skye, a rendering of vice-comes de Ski that seems to have been taken from the edition with translation of the Chronicle the the Rev. James Johnstone in his Antiquitates Celto-Normannicae, published in Copenhagen in 1786.
40. Cf. ante, Vol. XXXIX-XL. 210.
41. Mackillop means son of Philip. If Mac Phaice were to be spelt Mac Phàilce, as it could be in view of derivation, then one might note the consonantal correspondence between P-L-C and P-L-P, that between C and P not unknown in some other contexts.
42. See Matheson, The Blind Harper, 109; and ante, Vol. XLVIII. 423-427.
43. See also R. C. MacLeod, op. cit., 25.
44. See also ibid., 26.
45. Taylor, op. cit., 400.
46. I owe this redistribution of minims to Mr. W. D. H. Sellar.
47. Skene. op. cit., III. 482.
48. I. F. Grant, op. Cit., 626-627.
49. Reference as in note 27, ante. In addition, it may be noted that Olaus McLeoid matriculated at Glasgow University in 1682 (Muniment Alme Universityatis Glasguensis III. 140). Not further identified, but he must have been of the Skye MacLeods. It may also be remarked that John Morison of Bragar, in his Description of the Lewis, calls Leod, ancestor of the MacLeods Claudius the sone of Oliipheous . . . the King of Noruway his, where Olipheous is presumably to be equted with Olavus and hence with Olaf (Macfarlanes Geographical Collections II.214).
50. Taylor, op. cit., 217.
The elegy to Sir Norman MacLeod (see ante, and Note 19) is in another source attributed to Donnchadh O Muirgheasáin (Mackinnon, Catalogue of Gaelic Manuscripts, 280-281).