The Ancestry of Leod

Andrew P. MacLeod

(Clan MacLeod Magazine, No. 91, November, 2000)

[NOTE: Click on the Footnote Numbers to read the footnotes; Click on the number again to return to the main body of text.]

The forthcoming Volume LX of the Transactions of the Gaelic Society of Inverness will include a paper by W. D(avid) H. Sellar entitled “The Ancestry of the MacLeods Reconsidered”.[1] The title of Mr. Sellar’s paper harks back to an earlier one, by the late Rev. William Matheson (1910-1995), entitled “The Ancestry of the MacLeods”, which was published in 1981.[2] Mr. Sellar’s paper contradicts that of the Rev. Matheson at several points and it was out of respect for his colleague that Mr. Sellar withheld going public with his views until after the Rev. Matheson’s death. However, although their views differ at various points, these two scholars are in broad agreement over Leod’s ancestry, to wit, that in the male line Leod was not descended from the Kings of Mann & the Isles but was the great-grandson of a man named Olvir. This may surprise many clans folk because the ‘official’ MacLeod view is that Leod was the son of King Olaf the Black, one of the last of the dynasty. The main purpose of this article is to prepare its Clan MacLeod readers for a complete reappraisal of their origins by explaining why, in the field of early clan genealogy, the two most eminent scholars of recent times should agree that the theory of descent from Olaf the Black is quite simply untenable.

The earliest known historical records pertaining to the MacLeods are two royal charters of King David II (reigned 1329-1371), one to Malcolm son of Tormod MacLeod for two parts of Glenelg and the other to Torcall MacLeod for the four penny land of Assynt, both of which are thought to date to c. 1343.[3][4] but it does not name Leod. The last recognized King of this dynasty, Magnus Olafsson, died in 1265[5] and the last known male of the family, Godred, son of Magnus, lost a battle against the Scots at Ronaldsway in 1275[6] and fled to Wales.[7] By the end of the thirteenth century, legal claims to the Isle of Mann were being pursued on behalf of daughters of the family,[8] implying that the (legitimate) male line from Olaf the Black was then extinct. In short, there is no historical reason to believe that Leod was the son of Olaf the Black.

The earliest evidence yet known for the belief that the MacLeods descend from Olaf the Black belongs to the time of Iain Mor MacLeod of Dunvegan (Chief 1626-1649), who is styled “John McOlaus of Dunvegane” in a document dated at Edinburgh, 11th August, 1630”,[9] the “McOlaus” representing “MacOlaf”. It seems that Iain Mor’s son Iain Breac (Chief 1664-1693) is the first recorded MacLeod of Dunvegan to have included the Arms of the Kings of Mann (the three legs in armour “flexed and conjoined in triangle at the upper part of the thigh”) with his own Arms.[10] Given that we know of Arms for the MacLeods of Dunvegan back to the time of Iain Borb (on record 1420-1463)[11] this all gives the impression that in the seventeenth century the MacLeods’ belief in their descent from Olaf the Black was not of long standing.

How then did that belief come about? It must have been based on something, but given that there is no known historical basis for it, then that something can have been only a tradition, or an error, or both. It is noteworthy then that it is only in a Scots or an English guise that the link to Olaf is made. Gaelic tradition about the origin of the MacLeods almost never mentions the name Olaf.[12][13] In vernacular Gaelic the name became ‘Olgar’,[14] as in the poetry of Mairi nighean Alasdair Ruaidh, e.g. “Olgharach thù” (‘one of Olgar’s race art thou’) and “de shliochd Olghair nan Iann” (‘Of the race of Olgar of sword-blades’).[15]

The name Olvir also appears in Gaelic genealogies of the MacLeods, of which there are six known. One of them, the genealogy of Norman MacLeod of Dunvegan (d. 1706), goes no farther back than Leod so is irrelevant here. Another, in Duald MacFirbis’s collection of genealogies made in 1650 (the manuscript of which is in University College, Dublin),[16] is also irrelevant: although said to be a genealogy of the MacLeods it is rather a genealogy for the MacCabes.[17] A relevant genealogy in the MacFirbis collection is that of Christina MacLeod, appended to MacFirbis’s genealogy of the MacLeans (Christina having been the ancestress of the MacLaines of Lochbuie). Christina, sister of Torcall MacLeod of Lewis, was the daughter of Murchadh son of Tormod son of Leod and her lineage is taken back through a female generation to Norse royalty, to wit, “Ealga fholt-alainn ingean Arailt mic Semmair rig Lochlan (“Ealga of the beautiful locks daughter of Harald son of Semmair, king of Lochlan”).[18] This genealogy will be discussed later in this article. The remaining genealogies are of Sir Tormod MacLeod of Berneray (d. 1705),[19][20] and of an unknown sixth-generation descendant of Leod’s. This last genealogy (the ‘Kilbride MS’ genealogy of c. 1540) has been lost and its subject is unknown because all six descendants of Leod “have been carefully erased from the MS.”[21]

For the three generations before Leod the four relevant genealogies are broadly in agreement with each other. The Kilbride MS gives Leod’s three immediate ancestors as Oloir, oib and Oilmoir; Ruaidri’s genealogy gives “Olbair, Raoige, and Olbair Snaige; Sir Tormod’s genealogy gives Olbuir, Raise and Olbuir Snaithe, and Christina’s genealogy gives Gillemuire, Raice and Olbair Snoice. Their greatest similarity is in the name of Leod’s great-grandfather. Noting that the Gaelic letters ‘b’ and ‘m’ sound the same when they are softened (to ‘bh’ and ‘mh’) and making allowance for differences in spelling, Matheson concluded that the name of Leod’s great-grandfather was in all cases the same as the name of the ancestor mentioned in Gaelic poetry, i.e., Olvir.[22][23] Matheson also concluded that Leod’s father was named Olvir too. He suggested that a form like the Kilbride MS’s Oilmoir, being of an unfamiliar Norse name, was misread and miscopied by Gaelic scribes to give the familiar Gaelic name Gillemuire of Christina’s genealogy.[24]

Matheson suggested that, as acceptable Gaelic tradition on the origin of the MacLeods was limited to the three generations before Leod, going back to an ancestor named Olvir, and that beyond that there was only a vague claim to royal Norse ancestry, the MacLeods would eagerly adopt information from outside sources and that this was supplied in 1586 when William Camden published the first edition of his Britannia, which gave an account of the Kings of Mann & the Isles. Matheson suggested that the MacLeod genealogists, finding the royal name ‘Olavus’ (Olaf) in Camden, made an understandable error and took it to be the same as their ancestral name Olvir,[25] i.e. the Gaelic genealogies’ Olvir grandson of Olvir was matched with Camden’s Olaf grandson of Olaf, thus making Leod the son of Olaf the Black. In short, the ‘Olaf the Black theory’ of the origin of the MacLeods (which currently can be traced back no farther than the record of “John McOlaus” in 1630) can be explained away as a simple case of mistaken identity.

Looking at Leod’s grandfather in the genealogies, Matheson dismissed Oib as an outright error and suggested that the other forms were mistakes for Paice, the Gaelic form of the Norse name Bálki, and that Leod’s grandfather was the father of Paul Balkason,[26] an historical ally of Olaf the Black [27] who was remembered in MacLeod tradition as the foster-father of Leod. [28] Matheson made an uncle and nephew out of Paul and Leod and drew further evidence for his theory from a song by Duncan Macrae of Inverinate in which Roderick MacLeod of Dunvegan (Chief 1693-1699) is referred to as ‘éighre Shìol Phàic’, the ‘heir of the seed of Bálki’. [29]

However, Matheson’s main identification was of Leod’s great-grandfather, Olvir, with Olvir Rósta, who is mentioned in the Orkneyinga Saga as having fled to the Hebrides, c. 1139.[30] This had already been suggested by Capt. F. W. L. Thomas,[31] but he did not develop his suggestion and it was left to Matheson to propose the male line link to the MacLeods and to put the theory on a firmer footing. Matheson basically had (in addition to the name Olvir) three reasons for this identification: --

1. Career - Olvir Rósta fled from his estate in Helmsdale to the Hebrides and never returned. The differences in the Gaelic genealogies before the MacLeods’ Olvir suggest his ancestors were unknown to the Gaelic genealogists and that therefore he was a newcomer to the Hebrides.

2. Date - Olvir Rósta arrived in the Hebrides c. 1139. It is possible therefore that he was the great-great-great-grandfather of Malcolm MacLeod of Glenelg, on record in the 1340’s.

3. The rare name Leod was common to the families of both Olvirs - Olvir Rósta’s maternal grandfather was Liótr (i.e. Leod) Nidingr. The MacLeods’ Olvir was the great-grandfather of the eponymous Leod.

Matheson made two other suggestions. First, he suggested that the Ealga, daughter of Harald son of “Semmair” (which he corrected to ‘Sen Imair’, i.e., ‘Old Ivar’, at David Sellar’s suggestion[32]). at the head of Christina’s genealogy, and the daughter of Mac Rhaild Armin, who in later MacLeod tradition was Leod’s wife, were variants of the same tradition, and probably referred to the wife of Olvir Rósta.[33] Second, having disconnected the MacLeods from the dynasty of the Kings of Mann & the Isles he still had to account for the Gaelic tradition claiming that the clan was of royal origin, so he very halfheartedly suggested that Leod may have married a sister of Magnus Olafsson, King of Mann.[34] We shall see reason to reject both of these suggestions.

Matheson’s paper drew several criticisms from Alick Morrison in the first chapter (entitled “The Origin of Leod”) of his book The Chiefs of Clan MacLeod, published in 1986.

Alick Morrison bravely tried to retain the Olaf the Black ancestry by claiming that the names Olvir and Olaf were effectively the same name,[35] but this is merely a repetition of the seventeenth century error that caused the problem in the first place. The Norse name ‘Olaf’, Óláfr, stems, via Áleifr, from Anleifr, the nasal quality of the first syllable appearing in the old English form Anlaf and the Old Irish forms Amlaip, Amlaim and Amlaib, and surviving into modern Gaelic as amhlaidh.[36] The name itself means ‘ancestor-inheritance’ or ‘ancestor-remains’.[37] ‘Olvir’, Old Norse Olvér, has a different history, being cognate with Old English Alewih and Alweo and German Alawig and Alahwih.[38] It would also seem to be cognate with the French name Olivier, whence (Norman-)English Oliver. Whatever, Olvir is clearly a different name from Olaf.

Three of the Gaelic genealogies give Leod’s great-grandfather an eke-name (Snaithe, Snaige, and snoice), the meaning of which is obscure. Captain Thomas suggested that snoice meant ‘the Hewer’,[39] Matheson offered no suggestion, and Morrison suggested snaithe meant ‘White’.[40] Morrison further suggested that because of their different eke-names, ‘Olvar the White’ was not the same as ‘Olvir Rosta’ (i.e., Olvir the Turbulent). This argument is based upon two assumptions.

The first assumption is that the words have been accurately translated. “Rósta” seems to have been a rare word that became obsolete. It has been translated in several ways (e.g., ‘Strife’,[41] ‘the Turbulent’,[42] ‘the Unruly’[43] and ‘Brawl’[44]) and though these have similar meanings one would like to know what evidence there was for those translations. Likewise, the translation of Snaithe, Snaige and snoice is open to discussion: two widely different meanings (both assuming a Gaelic origin) have been mentioned above, and David Sellar, suggesting a Norse origin, proposes a third meaning.[45] The late Aubrey Halford-MacLeod pointed out to this author[46] that the obsolete Gaelic word “snatha” (which is a possible nominative form of the genitive snaithe) is given a secondary meaning of “Grief, trouble”[47] and thus could conceivable overlap in meaning with the Norse “Rósta”. Until the words can be translated accurately then arguments based upon their meaning will be weak.

The second assumption, common to both Morrison and Sellar, is that a man can have only one eke-name. In theory it is quite plausible that a man can be known to different people in different places at different times by different eke-names, as would be the case if the MacLeods’ Olvir and Olvir Rósta were the same man: “Rósta” would have been given to him by his enemies in the Orkney area before he fled to the Hebrides, whereas Snaithe/Snaige/snoice, or whatever the word is,[48] would have been given to him by his friends and descendants in the Hebrides after he had fled thither. However, one need not rely on theory for a double eke-name as there are known examples from the Hebrides.

The founder of the last dynasty of the Norse Kings of Mann & the Isles was Godred Crovan (d. 1095),[49] who is so named in the Manx chronicle.[50][51] Likewise, an Irish poem to the same Reginald names Godred as “Gofraidh Méraig”[52] and the Annals of Ulster twice name him as “Goffraigh Méranach”.[53] The “mearch”, “Méraig” and Méranach” eke-names seem to derive from Gaelic mér, ‘finger’, which led Brian Ó Cuiv to to suggest that “Crovan” “may be from crobh-bhán ‘of the white hand.’:[54] Godred Crovan was eventually succeeded by his son Olaf (k. 1152), who is named “Bitling” (‘little bit’ or ‘Tit-Bit’) in the Orkneyinga Saga,[55] and “Dheirg” (‘the Red’) in the Book of Clanranald.[56] This Olaf Bitling/Dheirg was the King of the Hebrides when Olvir Rósta fled there, thus giving a definite example of a double eke-name both at the same time and in the same place as Olvir.

From the foregoing it should be clear that differing eke-names are not necessarily a bar to identity. The exception, of course, is if the two eke-names are in direct contradiction to each other. Thus if Alick Morrison is right to explain the eke-name of Leod’s great-grandfather as meaning ‘the White’ then this is nota bar to identity with ‘Olvir the Turbulent’; but it is a bar to identity with ‘Olaf the Red’, who was Leod’s great-grandfather under the Olaf-theory.

Morrison was on surer ground in his criticism of Matheson’s belief that the name of Leod’s grandfather was Bálki. Morrison rightly pointed out[57] that all three acceptable name-forms of Leod’s grandfather in the Gaelic genealogies, namely, Raisi, Raoige and Raice, have ‘R’ as their initial letter, and that this agreement rules out Matheson’s amendment to ‘P’. Morrison could also have pointed out that whereas the three name-forms are all in the genitive case, Matheson’ Páice is a nominative form. For a fair comparison between the names, Raisi, Raoige and Raice should be compared with the genitive form of Páice, which on Matheson’s own showing[58] is Phàic. From this comparison it is clear that the two names are different. This means that Leod’s grandfather was not named Bálki and so Paul Balkason was not Leod’s paternal uncle. Therefore, if Duncan Macrae of Inverinate could refer to Roderick MacLeod of Dunvegan as ‘éighre Shìol Phàic’, the ‘heir of the seed of Bálki’, then it was not because of male-line descent.

What then is the name of Leod’s grandfather as given in the acceptable Gaelic genealogies? David Sellar suggests that it may be the Gaelic name Raingce.[59]Raise, Raoige and Raice than is Phàic, but it too suffers from need all three acceptable name-forms to be amended, this time by the addition of the letter ‘n’. There is, however, another candidate for the name. In a four-step process one can strip away the Gaelicizing elements to reveal a known Norse name.

(a) Reduce the genitive forms to their Gaelic nominative case. A common way in Gaelic to make a genitive case from a nominative one is to insert the letter ‘i’ before the last consonant cluster. Thus Leod gives Leoid, Tormod gives T(h)ormoid, Torcall gives Torcaill and Aralt gives Arailt. Raise, Raoige and Raice all have the letter ‘i’ in the stated place so it is a fair assumption that their respective nominative forms should be Rasi, Raoge and Race.

(b) Ascertain the basic structure of the name. The only discrepancy here is Raoge’s diphthong ‘ao’, but we occasionally find this diphthong representing a single Norse ‘a’,[60] so we may agree with Rasi and Race that the structure is basically consonant-vowel-consonant-vowel (CVCV) and the name is therefore disyllabic.

(c) Assign values to the structure. Clearly, the first two letters (C1 and V1) are ‘R’ and ‘A’. The second vowel (V2) is either ‘e’ or ‘i’, and because Gelic often turned Norse terminal ‘i’ into ‘e’,[61] then it is almost likely that V2=”I’. This just leaves C2, the various forms of which are ‘s’, ‘g’ and ‘c’. Of these, the ‘g’ is the link between the other two: the link with ‘c’ is through pronunciation (hence Helgi becomes Ailche,[62] and Sigurd becomes Siucraid[63]) and the link with ‘s’ is through spelling (Gaelic ‘g’ and capital “S’ have a similar shape). This suggests that C2=’G’, giving the extrapolated name ‘RAGI’.

(d) Match the extrapolation with a known name. In this case the match is perfect: ‘Ragi’ was a (very rare) Norse name.[64]

This author thinks it most likely that the name of Leod’s grandfather in the Gaelic genealogies is Ragi, but, whatever it is it cannot be Godred, which is what it would need to be if Leod were the son of Olaf the Black.

David Sellar agrees with Alick Morrison in rejecting Matheson’s Bálki as the name of Leod’s grandfather. He also rejects Matheson’s Olvir as the name of Leod’s father, arguing that Gillemuire of Christina’s genealogy is right and that the other genealogies have been led into error by the prominence of the name Olvir in the MacLeod tradition.[65] The point is well made and it is hard to decide which view, Matheson’s or Sellar’s is right.

David Sellar also uses the differing eke-names as his main reason to reject Matheson’s identity of Leod’s great-grandfather with Olvir Rósta. His second reason for rejecting the identity concerns the differing genealogies of the two Olvirs. In Olvir Rósta’s male line the Orkneyinga Saga names only his father, Thorljot,[66] and there is no reason to doubt its accuracy in this. Sellar treats Christina’s genealogy as being completely accurate and so he accepts that Olbair Snoice’s father was named Gillemuire. Certainly, the name Thorljot does not occur in any of the MacLeod genealogies, and this is a major stumbling block for those who would believe that Leod’s great-grandfather was Olvir Rósta. However, whether or not Olvir Rósta was the great-grandfather of Leod is still a moot point and it will not be discussed any further here. Of more relevance here is Sellar’s confidence in Christina’s genealogy for in this we find his major advance, from Matheson’s paper, in the field of MacLeod ancestry.

Matheson believed that the genealogies displayed an ignorance of the generations before Leod’s great-grandfather, their differences canceling each other out; but that assumes that they are of equal value to each other, and that is not the case. Quite simply, Christina’s is the genealogy of greatest value: Christina is the oldest of the subjects, so her genealogy should have been the earliest one in origin; unlike Ruaidri’s and the Kilbride genealogies it is not so long as to be unreal; it was deemed sufficiently prestigious to have been added as an appendix to the MacLean genealogy; and it is specifically traced back to Norse royalty through an ancestress of Christina’s. It is the description of this ancestress as “Ealga fholt-alainn ingean Arailt mic Semmair rig Lochlan”[67]

This author is in complete agreement with this identification. Reginald’s Welsh genealogy goes on back, making Old Ivar the son of Olaf son of Sigtrygg,[68][69] The most important Norse leader slain at Clontarf was Sigurd, Jarl of Orkney, who also ruled over Mann & the Isles. The power-vacuum occasioned by Sigurd’s death seems to have enabled the Dublin Norse to control the Isle of Mann.[70][71] Given that Dublin (under King Sigtrygg, who ruled until 1035) had during the 1020’s extended its power over Mann, it is reasonable to suppose that it tried to push into the Hebrides too, and that it accomplished this under King Sigtrygg’s brother, Old Ivar, “rig Lochlan”, which can now translated as ‘King of the Hebrides’.

The renewed power of the Orkneys under Jarl Thorfinn the Mighty would explain why Old Ivar’s son, Harald, is described as of ‘Ysland’,[72] which seems to mean ‘Iceland’.[73] It would seem that Old Ivar’s family fled from the Hebrides to escape Jarl Thorfinn’s westward campaign and took refuge in Iceland, thus explaining why Harald did not bear the title of ‘King’.[74] It would be Harald’s son, Godred Crovan, who would reestablish the royal dignity of the family in the Isles. Godred Crovan and his sister Ealga were children of Harald the Black of Iceland, son of Old Ivar, King in the Hebrides, son of Olaf Cuaran (d. 981), King of Dublin & York, son of Sigtrygg Caoch (d. 927), King of Dublin and York, grandson of Ivar (d. 873), King of Dublin. The death of this last Ivar in 873 is recorded in the Annals of Ulster, which describe him as “rex Nordmannorum totius Hiberniae et Britanniae”, that is, ‘king of all the Scandinavians in Ireland and Britain’.[75] This is the line through which the MacLeods could claim their royal Norse ancestry, and they claimed it not through descent from Olaf the Black but through descent from Ealga (i.e. ‘Helga’), sister of Godred Crovan.

In summary, the ‘official’ MacLeod ancestry must be rejected because: (a) there is absolutely no historical evidence to support it; (b) what historical evidence there is is against it (female line claimants to Mann); (c) Gaelic tradition on the subject directly contradicts it (Olvir not Olaf); (d) it can be traced back no farther than the early seventeenth century; (e) it can be explained as a mismatching of Gaelic tradition with the account of the Manx kings in Camden’s Britannia; and (f) Gaelic tradition can be matched with the genealogy of King Reginald (d. 1229) in such a way as not only to explain the MacLeod claims but also to fit in with and to add to known history.

Whether Leod’s father was named Olvir or Gillemuire and whether or not Leod’s great-grandfather Olvir was Olvir Rósta, the MacLeods are of the Sliochd Olbhuir and their royal Norse ancestry comes not from Olaf the Black but from Olvir’s ancestress Helga, sister of Godred Crovan.

Leod himself does not appear in contemporary record. The Manx chronicle names four sons of Olaf the Black, Instead, there often appears in Gaelic sources the name Olvir, thus in elegies on Iain Mór MacLeod of Dunvegan (d. 1649) and on Sir Ruaidhrí Mór MacLeod of Dunvegan (d. 1626) we find “Crú Olbhuir” (‘stock of Olvir’), “aicme Olbhuirsi” (‘race of Olvir’), “í Olbhair” and “ó Olbhair” (‘descendant of Olvir’), and “síl Olbhair” (‘seed of Olvir’). of Ruaidri MacLeod of Lewis This conclusion of Matheson’s has been generally accepted. Yet in the genealogy of his great-grandson Reginald, preserved in a Welsh genealogical collection, he is named “Gwrthryt mearch”. This is certainly a better fit for which has proved this genealogy’s value, because, after correcting the genealogy’s “Semmair” to “Sen imair”, i.e., ‘Old Ivar’, David Sellar was able to identify the two earliest generations, ‘Harald son of Old Ivar’, with the father and grandfather of Godred Crovan as given in the above-mentioned Welsh genealogy of Reginald, thus making the MacLeods descend from Godred’s sister, Ealga (Helga). which last two generations are identifiable as Olaf Cuaran (d. 981) and his father Sigtrygg Caoch (d. 927), both of whom were Kings of Dublin and York. In addition, Old Ivar, the “rig Lochlan” of Christina’s genealogy, becomes the brother of Sigtrygg Silkiskegg (Silkenbeard), the King who remained safely inside Dublin during the battle of Clontarf (1014) and died in 1042. Jarl Sigurd’s son Thorfinn the Mighty (d. c. 1064/5) eventually reestablished Orkney control over the Hebrides, but he was only five years old when his father was killed and he had first to wrest control of the Orkneys from his elder half-brothers. It would appear that it was only after the death of Jarl Brusi, c. 1031, that Thorfinn was able to do more than merely raid the Hebrides. The saga records that when, c. 1035, Jarl Brusi’s son, Jarl Rognvald, came to the Orkneys, Jarl Thorfinn received him diplomatically because he needed help in the great quarrels he was having with the Irish and the Hebrideans.

1. The author here takes the opportunity to record his many thanks to Mr. Sellar both for an advance copy of his as yet unpublished paper and for permission to refer to it in this article. The lecture upon which the paper is based was delivered to the Gaelic Society of Inverness on Friday 30th January 1998.

2. Being pp. 68-80 in Transactions of the Gaelic Society of Inverness (henceforth TGSI) Volume LI (1978-80). The lecture upon which the paper was based was delivered on the 18th November 1977.

3. John Maitland Thomson (ed.) (1912) Registrum Magni Sigilli Regum Scotorum -- The Register of the Great Seal of Scotland A.D. 1306-1424, pp. 569-570. An old index lists the charters with four others, two of which are dated 12 June 1343.

4. George Broderick (1973) Chronicle of the Kings of Mann and the Isles, pp. 34-46 & 75-81: Harald, Reginald, Magnus, Godred.

5. George Broderick (1973) Chronicle of the Kings of Mann and the Isles, pp. 46 & 81.

6. P. A. Munch (1860) The Chronicle of Man and the Sudreys, revised and translated by the Right Rev. Dr. Goss (1874), p. 232 & footnote ‘c’ on pp. 232-233, quoting the Chronicle of Lanercost.

7. George Broderick, “Irish and Welsh Strands in the Genealogy of Godred Crovan”, being pp. 32-38 in The Journal of the Manx Museum Vol. VIII No. 89 (1980), p. 36 & endnote 38 on p. 38, quoting “Continuator of William of Newburgh”.

8. P. A. Munch/Dr. Goss (1860/1874) The Chronicle of Man and the Sudreys, p. 233, notes that in 1293, one Affreca claimed to be the legal heiress of Magnus Olafsson. Sir George MacKenzie, first Earl of Cromartie, in his manuscript ‘History of the Family of MacKenzie’ refers to a claim on behalf of Mary, daughter of Reginald Olafsson (in William Fraser (1876) The Earls of Cromartie Their Kindred, Country, and Correspondence Vol. II, p. 510).

9. William Matheson, “The Ancestry of the MacLeods”, p. 72.

10. I. F. Grant (1959, 2nd edition 1981) The MacLeods: The History of a Clan, Appendix II “Some Notes on the Heraldry of the MacLeods’ (by C. I. Fraser of Reelig, Albany Herald), pp. 635-636.

11. Jean Munro & R. W. Munro (1986) Acts of the Lords of the Isles 1336-1403, pp. 266-267; the Scottish section of the Armorial de Berry is dated to c. 1445 -- I. F. Grant (1059/1981) The MacLeods: The History of a Clan, pp. 634-635.

12. The only occurrence of the name Olaf in an acceptable Gaelic genealogy of the MacLeods actually belongs to a Nicolson Lineage that has been grafted onto the end of the MacLeod one. One other occurrence of the name is in a genealogy that, despite its title, is rather of the MacCabes than of the MacLeods.

13. J. Carmichael Watson, “Cathal Mac Muireadhaigh Cecinit”, being pp. 167-179 in Reverend John Ryan (Eóin Ua Riain), ed., (1941) Féil-sgrighian Eóin Mhic Néill, . p. 172, stanza 29, line 1 & stanza 31, line 2. This elegy is in the Royal Irish Academy (henceforth RIA] MS. E i. 3. John MacDonald, ‘An Elegy for Ruaidhri Mór, being pp. 27-52 in Scottish Gaelic Studies Vol. Viii, Part 1 (1055), pp. 30, 38 & 42. This elegy is in RIA MS. 23.N.12. p. 9.

14. This equation between classical Olbhur and vernacular Olgar was made clear by William J. Watson (1918/1932) Bardachd Ghaidhlig (second edition) Specimens of Gaelic Poetry 1550-1900, p. xxxi: “Olghar is with the classic poets Olbhur, which is likely to be nearer the original form, representing the Norse name Olver”, though earlier, Alexander MacBain had equated the names “Ollghair”, “Olvir” and ‘Olbair’ on p. 420 of his edition (1902) of William F. Skene (1836( The Highlanders of Scotland.

15. J. Carmichael Watson, ed., (1934/1982) Gaelic Songs of Mary MacLeod, lines 521 (p. 44) & 791 (p. 66).

16. W. D. H. Sellar, “The Ancestry of the MacLeods Reconsidered”, in TGSI Vol. LX, forthcoming, endnote 48.

17. It does have one or two MacLeod touches to it but it has long been recognised as historically worthless for the MacLeods. William F. Skene, “The MacLeods of Scotland”, being pp. 317-320 in Ulster Journal of Archaeology, 1st Series, IX (1861-1862), p. 319, described it as “more like a jeu d’esprit of some Senachaidhe than a pedigree seriously inteded to be taken as authentic.”

18. William F. Skene in “The MacLeods of Scotland”, p. 320, and (2nd edition, 1890) Celtic Scotland: A Hsstory of Ancient Alban Volume III. Land and People, p. 482.

19. On page 12 of RIA MS.E.i.e, i.e., the page after the elegy on Iain Mór MacLeod of Dunvegan, already mentioned.

20. On page 48 of RIA MS.23.H.22 and in RIA MS 23.G.4 (see W. D. H. Sellar, “the Ancestry of the MacLeods Reconsidered”, in TGSI Vol. LX,

21. William F. Skene in The Iona Club (1847) Collectanea de Rebus Albanicis, consisting of Original Papers and Documents relating to the history of the Highlands and Island s of Scotland, p. 362, endnote 6, the genealogy having been printed on p. 361. Skene printed the genealogy twice more: “The MacLEods of Scotland”, p. 318, and (2nd edition, 1890) Celtic Scotland Volume III, pp. 460-461. Unfortunately, he published it slightly differently on each occasion.

22. William Matheson, “The Ancestry of the MacLeods”, p. 69.

23. Even Matheson’s main critic was prepared to accept this - Alick Morrison (1986) The Chiefs of Clan MacLeod, pp. 9-10.

24. William Matheson, “The Ancestry of the MacLeods”, p. 72.

25. William Matheson, “The Ancestry of the MacLeods”, pp. 71-72.

26. William Matheson, “The Ancestry of the MacLeods”, p. 73.

27. George Broderick (1973) Chronicle of the Kings of Mann and the Isles, pp. 30 & 73.

28. The Rev. Canon R. C. MacLeod of MacLeod (1927) The MacLeods of Dunvegan, pp. 4 & 24.

29. William Matheson, “The Ancestry of the MacLeods”, p. 74.

30. William Matheson, “The Ancestry of the MacLeods”, p. 70.

31. Capt. F. W. L. Thomas, “Traditions of the MacAulays of Lewis”, biing pp. 363-431 in Proceedings of the Society of Antiquaries of Scotland 1879-1880 Vol. XIV (1880), p. 364 and footnote 2 thereto.

32. William Matheson, “The Ancestry of the MacLeods”, p. 76 & endnote 46 on p. 80.

33. William Matheson, “The Ancestry of the MacLeods”, pp. 76-77.

34. William Matheson, “The Ancestry of the MacLeods”, p. 77.

35. Alick Morrison (1986) The Chiefs of Clan MacLeod, pp. 10-12.

36. Alfred P. Smyth (1077) Scandinavian Kings in the British Isles 850-880, p. 280; Michèle Brown (1985) The New Book of First Names, pp. 490 & 491; William Matheson, “The Ancestry of the MacLeods”, p. 72.

37. Michèle Brown (1985) The New Book of First Names, pp. 490 & 491.

38. R. W. Chambers (1912) Widsith: A Study in Old English Heroic Legend, pp. 202 & 203.

39. Capt. F. W. L. Thomas, “Traditions of the MacAulays of Lewis”, p. 364, footnote 2.

40. Alick Morrison (1986) The Chiefs of Clan MacLeod, p. 10.

41. Joseph Anderson (ed.), Jon A. Hjaltalin & Gilbert Goudie (trans.) (1873/1981) The Orkneyinga Saga, p. 69.

42 . Capt. F. W. L. Thomas, “Traditions of the MacAulays of Lewis”, p. 364, footnote 2; Rev. Canon R. C. MacLeod of MacLeod (1927) The MacLeods of Dunvegan, p. 12; and Alick Morrison (1986) The Chiefs of Clan MacLeod, p. 10. Thomas cites “Anderson’s “Ork. Saga.”, which suggests the 1873 edition of the previous reference, but the 1981 edition is a facsimile of that.

43. William Matheson, “The Ancestry of the MacLeods”, p. 70, citing (p. 78, endnote 16) “Taylor, The Orkneyinga Saga, 263-264.

44. Hermann Pálsson & Paul Edwards (trans.) (1978/1981) Orkneying Saga: The History of the Earls of Orkney, pp. 91, et al.

45. W. D. H. Sellar, “The Ancestry of the MacLeods Reconsidered”, in TGSI Vol. LX, forthcoming.

46. During the last week of September 1987.

47. See, e.g. Edward Dwelly (1901-1911; 11th ed. 1994) The Illustrated Gaelic-English Dictionary, p. 864.

48. If one applies to these word-form the four steps used below on the name of Leod’s grandfather, one ends up with a possible Norse “snagi”, which David Sellar notes, in his forthcoming paper, is an Old Icelandic word meaning “clothespeg” -- not quite the heric ideal for the eponymous ancestor of Mairi nighean Alasdair Ruaidh’s “sliochd Olghair”!

49. George Broderick, “Irish and Welsh Strands in the Genealogy of Godred Crovan”, p. 32 & endnote 3 on pp. 36-37.

50. George Broderick (1973) Chronicle of the Kings of Mann and the Isles, pp. 6 & 61. godred is four tiems named Crovan in the chronicle: thee first time is under the year 1047 (recé 1066), under which entry the chronicle also notices the conquesto of England by “Willelmus bastard” (p. 61), ‘William the Bastard’ (p. 6), who of course is the same person as ‘William the Conqueror’, the first Norman King of England.

51. P. C. Bartrum (1966) Early Welsh Genealogical Tracts, p. 99. This genealogy is of great importance to the MacLeods.

52. Brina Ó Cuiv, “A poem in Praise of Raghnall, King of Man”, being pp. 283-301 in Éigse, VIII (1957), p. 292.

53. W. H. Hennessy (1887) Annals of Ulster Vol. II, pp. 52/53 & 54/55 for the years 1094 & 1095.

54. Brian Ó Cuiv, “A Poem in Praise of Raghnall, King of Man’, footnote 6 on pp. 283-284.

55. Joseph Anderson (ed.) Jon A. Hjaltalin & Giilbert Goudie (trans.) (1873/1981) The Orkneying Saga, pp. 69, 181 & 222; Herman Pálsson & Paul Edwards (trans.) (1978/1981) Orkneying Saga: The History of the Earls of Orkney, pp. 97, 208 & 238.

56. Alexander MacBain & Rev. John Kennedy, edd. (1894) Reliquiae Celticae Texts, Papers, and Studies in Gaelic Literature and Philology left by the late Alexander Cameron, LL. D., p. 156.

57. Alick Morrison (1986) The Chiefs of Clan MacLeod, p. 10.

58. William Matheson, “The Ancestry of the MacLeods”, pp. 73-74.

59. W. D. H. Sellar, “The Ancestry of the MacLeods Reconsidered”, in TGSI Vol. LX, forthcoming.

60. Thus in the Nicolson section of Ruadri’s genealogy the name Olaf is rpresented by “Amhlaoiby” and exactly the same spelling occurs in Duald MacFirbis’s ‘MacLeod’ (MacCabe) genealogy.

61. Thus Bálki became Páice (William Matheson, “The Ancestry of the MacLeods”, p. 73) and Helgi became Ailche (Alfred P. Smyth (1979) Scaninavian York and Dublin, Voolume II pp. 20 & 356; Ailche is a genitive form).

62. See previous endnote.

63. James Henthorn Todd (1867) Cogach Gaedhel re Gallaibh -- the war of the Gaedhil with the Gaill, pp. 152 & 194, in which Sigurd Hlodversson, Jarl of Orkney, appears as Siucraid mac Lotair, “iarla Insi Orc”.

64. Thorarin Ragi’s brother was the second Law-Speaker of Iceland (950-969) and took his name from his brother, Ragi of Laugardale. They were sons of Olaf Hjalti andare mentioned in Njal’s Saga and in Egil’s Saga.

65. W. D. H. Sellar, "The Ancestry of the MacLeods Reconsidered", in TGSI, Vol. LX, forthcoming.

66. Joseph Anderson (ed.), Jon A. Hjaltalin & Gilbert Goudie (trans.) (1873/1981) the Orkneyinga Saga, p. 69: “Thorliot”. Hermann Pálsson & Paul Edwards (trans.) (1978/1981) Orkneyinga Saga: The History of the Earls of Orkney, p. 97: “Thorljot”.

67. William F. Skene, “The MacLeods of Scotland”, p. 320.

68. P. C. Bartrum (1966) Early Welsh Genealogical Tracts, p. 99: Ifor gamle m. Afloyd m. Swtrig.”

69. James Henthorn Todd (1867) Cogadh GAedhel re Gallaibh -- The War of the Gaedhil with the Gaill, pp. 188-193, 278 & 288-290.

70. George Broderick, “Irish and Welsh Strands in the Genealogy of Godred Crovan”, p. 33.

71. Joseph Anderson (ed.), Jon A. Hjaltalin & Gilbert Goudie (trans.) (1873/1981) The Orkneyinga Saga, p. 26, with background from p. 4 onwards; see also pp. cxxvii & cxxxii. Hermann Pálsson & Paul Edwards (trans.) ( 1978/1981) Orkneying Saga: The History of the Earls of Orkney, p. 59, with background from p. 38 onwards.

72. George Broderick (1973) Chronicle of the Kings of Mann and the Isles, pp. 6 & 61.

73. W. D. H. Sellar, “The Ancestry of the MacLeods Reconsidered”, in TGSI Vol LX, forthcoming; Alick Morrison, “The Kingdomof Man and the Isles: 839-1266”, being pp. 425-481 in TGSI Volume LVIII 1993-94 (1995), pp. 452-453.

74. Alick Morrison (“The Kingdom of Man and the Isles: 839-1266”, pp. 452-453) also believes that the family took refuge in Iceland after being expelled by Jarl Thorfinn, but he treats Harald as first having been a king.

75. Alfred P. Smyth (1977) Scandinavian Kings in the British Isles 850-880,, p. 234. forthcoming). It is this genealogy that has the Nicolson lineage appended to it, presumably because the MacLeods of Lewis acquired their lands from the Nicolsons.