The Ancestry Of Somerled: Godfrey MacFergus, Colla Uais, Conn of the Hundred Battles and Neill of the Nine Hostages

Clan Donald has always been known as sons of Somerled. Since the early 1400s Clan Donald historians, including the famous 1904 three volume treatise by the Rev. Archibald and Angus MacDonald, "Clan Donald" [MacDonald 1904], have either asserted or concluded after discussion that Somerled was Gaelic in his paternal line. They usually trace his descent from Colla Uais, a high king of Ireland whom the Irish histories indicate reigned approximately 330AD after conquering Ulster. The ancient Irish histories asserted that both Colla and Neill of the Nine Hostages, who was the ancestor of the O’Neills, were each of common paternal descent from Conn of the Hundred Battles, high king of Ireland circa 175AD. Clan Donald poetry from at least the 1400s asserted this ancestry. “Children of Conn be hardy in adversity in battle ...” was the commencement of the speech used to incite Clan Donald troops before the great battle at Harlaw in 1411. Norman MacDonald, immediate past president of Clan Donald Society of Edinburgh and probably the top Clan Donald historian in the world, supported the Gaelic descent hypothesis. An additional historical note supporting the Gaelic descent of Somerled was that contemporary Norse historians treated him as descended from the Dalriadic family of Argyll.

Some modern Irish historians have asserted that Conn was not a historical person but merely a sun god figure and that the three Collas were simply a creation of the Irish great love of triads. Other modern Irish historians have asserted that the alleged common descent of Colla and the O’Neills was simply O’Neill propaganda and that the battles which destroyed the power of the earlier Ulster king were simply fought by mercenaries hired by the O’Neills. These historians also emphasize the subordinate nature of the kingdom of Oriel and its members, the Airghialla, to the O’Neills. To add to the historic confusion, Archibald and Angus MacDonald had chosen to combine the Colla Uais line with the line of Cairbre Riata, whose descendants Fergus, Angus and Lorne, have historically been believed to have founded the Kingdom of Scottish Dalriada in the late 400AD period. These concerns still assume a Gaelic ancestry for Somerled.

Though there are variations, a typical line is as follows, adapted from Sellar [Sellar 1966], Schelgel [Schlegel 2000], and A & A MacDonald:

Colla Uais (~330)
Godfrey (~830)
[Maine] (not present in A & A MacDonald)
Mearrdha (Meargaige in Sellar)
Somerled (~1150)

Long before genetic evidence appeared there were hints that this all-Gaelic descent was amiss.

There are not enough generations listed in the historically accepted Somerled descent. These lie generally in the period between Colla and Godfrey MacFergus. There were also a number of hints that Somerled was Norse in his father’s line as well as the very heavily Norse pedigree of his last and most famous wife Ragnhilda. “Somerled” is a Norse name which means “summer sailing,” i.e. pillage. The early stories about Somerled emphasized his Viking-like cunning. Even though he led Gaelic tribes, our ships were classic Viking galleys, albeit adding a rudder as a new technical innovation. Branches of Clan Donald such as Glengarry used a major Norse symbol, Odin’s raven, as their symbol. Clan Donald prominently displayed the Norse galley in its heraldry. Sir Ian Montcreiffe, the now deceased holder of one of the two major genealogical offices in Scotland, had proposed a chart suggesting Norse paternal descent for Somerled, based in part on the heraldic device of the Norse galley. James McDonald, a former U.S.A. Clan Genealogist, prepared a chart for Clan Donald U.S.A. suggesting Norse paternal descent and hypothesizing the generation in which the male line of the Argyll house had failed and the crossover to the female line occurred. As National Historian Mark also leaned toward the Norse descent hypothesis. One recent excellent book by Ronald Williams, "The Lords of the Isles: The Clan Donald and the Early Kingdom of the Scots", pp. 111-123 (1997) [Williams 1997] suggested a different death in battle of one of the Dalriadic kings of Argyll as the potential crossover point for Norse lineage.

Although there is great historical debate over the paternal ancestry of Godfrey mac Fergus, there is far greater concurrence on the generations between Godfrey and Somerled. W. H. Sellar, a leading academic historian, had closely examined Somerled’s historic genealogy and concluded that it was probably accurate from Godfrey MacFergus onward. The only difference between Sellar and the following is that Sellar inserts a Maine after Godfrey. The Reverends Archibald and Angus MacDonald, in their 1904 book on the genealogy of the Clan Donald [MacDonald 1904], provided the following description:

XVII Godfrey, whose daughter was the wife of Kenneth MacAlpin, and who was known in his day as Toshach of the Isles. The son and successor of Godfrey was
XVIII Nialgus, or according to some, Neill, his son was
XIX Suibne, according to Dean Munro Swyffine. His son was
XX Mearrdha, Latinized Marcus, and Hailes in his Annals states that Kenneth, King of the Scots; Malcolm, King of the Cambri; and Marcus, King of the Isles, entered into a bond of treaty for mutual assistance and defense in the year 973. This shows that Lords of the Isles existed before Somerled’s time. The son of Mearrdha was
XXI Solaim, Solan, or Sella, whose son and heir in the Lordship of Argyll and the Isles was
XXII Gilledomnan. It was during the lifetime of this chief that the Western Isles of Scotland were completely subjugated by the piratical Norsemen. His daughter married Harold Gillies, King of Norway. Gilliedomnan was succeeded by
XXIII Gillebride or Gilbert, who is mentioned by the oldest Highland genealogist as “rig eilean Shidir,” that is, King of the Sudereys or Southern Isles. His daughter was the wife of Wymund MacHeth, Earl of Moray. He was called Gillebride na h-Uamh, from the fact that during a certain period of his depressed fortunes he lived in a cave in the district of Morvern. From Gillebride are said to have descended – besides the Clan Donald and Clan Dougall, etc. – the Maclachlans, MacEwin of Otter, and others. His son was Somerled rex insularum, or, as he is known in Highland tradition, Somhairle Mor MacGillebhride
The assertion that King Kenneth McAlpin married Godfrey’s daughter is quite consistent with typical treaty terms of the period and consistent with significant services by Godfrey on Kenneth’s behalf in obtaining the crown. Kenneth “granted” Godfrey the isles which Kenneth undoubtedly could not control without Godfrey’s combined viking and Oriel ships, especially in the face of the ongoing viking onslaught.


Many historians have been baffled by the name Godfrey MacFergus, who the Irish annals say was a Toiseach of Oriel who came up from Ireland in 836AD to assist Kenneth MacAlphin in his wars against the Picts. Those wars unified the Picts and Scots under a single dalriadic line king. See Sellar [Sellar 1966] and "Somerled And the Emergence of Gaelic Scotland" by John Marsden [Marsden 2000]. Godfrey (Gutfrith) was clearly a very Norse name born at a point in time within a few decades after the Vikings had commenced burning, raping and pillaging in Ireland around 795AD. It was not the sort of name which should be popular among a Christian people whose churches were being looted and monks murdered.

At this point the genetic DNA evidence comes into play. As explained here, the DNA signature of the descendants of Somerled is decidedly non-Gaelic. The Somerled signature is very distinct from the dominant Gaelic ones. Furthermore, the signature itself is prototypically Norse, being very common in Norway (and Iceland) , and uncommon in other European populations (including Denmark, source of other Vikings.)

Mark hypothesizes that the Norse genetic signature found in our chiefs and other Clan Donald descendants of Somerled suggests that Godfrey was a Norse (not Danish) viking who arrived with a fleet of ships in Ireland near Derry. As separately occurred with Rollo in his establishment of Normandy, this viking came to a prompt treaty with a Fergus chief of the MacCarthend MacUais. Those treaties often involved marrying into the family of the local chief or king. Marriage also was regularly accompanied by an agreement by the viking to become a Christian. There are other recorded instances such as the baptism and fosterage of Hacon of Norway under the sponsorship of a king of the English Saxons. Whether by marriage to the daughter of Fergus, by terms of a treaty, by baptism or a combination of all three, Godfrey became Toiseach of these descendants of Colla Uais. Although the term Toiseach normally referred to a member of the Derbfine (male descendants of a common great-grandfather in the line of the existing chief), these were very dangerous times for the MacUais. The O’Neill high kings (to which Oriel and its Airghialla were subkings) were expanding rapidly and required new lands for the many sons (up to thirty) being born to a single O’Neill chief. The MacCarthends lived closest to the O’Neills and were directly in the path of O’Neill expansion. The particular branch of the O’Neills were the Cineal Eoghan whose lands pronounced “Tirowen” for whom County Tyrone is currently named.

We can not be sure how long Godfrey stayed in Ireland until the opportunity to seize lands in Scotland by supporting Kenneth MacAlpin presented itself. This was a strategic opportunity for Godfrey’s Vikings and the MacCarthend to seek more hospitable lands in Lorne (and probably the Isle of Coll which had excellent harbors). The combined fleet of Godfrey and the MacCarthend must have been of substantial assistance to Kenneth.

Historians have long noted that in 839AD the pictish nobility was decimated in a key battle with a “foreign army.” Before we knew the genetics, our Clan Donald and other historians consistently had looked through the blinders that Godfrey was somehow a descendant in the male line from Colla. Although it is possible that Kenneth MacAlpin had hired multiple groups of viking and/or Irish mercenaries to decimate the picts, it appears probable that the ancestors of our Clan Donald chiefs were instrumental in placing Kenneth MacAlpin on the throne of Scotland. In the years succeeding the initial destruction of the pictish nobility in 839, Professor Alfred Smyth in "Warlords and Holy Men" [Smyth 1984] has asserted that Godfrey’s fighting band was additionally used to wipe up remaining pockets of pictish resistance.

By the time that Godfrey died in 853, he was no longer simply Toiseach of Oriel but Righ Innse Gall, King of the Islands of the Foreigners. Interestingly, although we have no specific history concerning the cause of Godfrey’s death, it occurred the same year in which Olaf the White, King of Norway, came through the Hebrides to conquer Dublin. Godfrey may well have died because he was in Olaf’s attack path. For the following 300 years, a variety of groups were fighting for control of the highlands and islands.

In any event this explanation, which materially changes our understanding of ancient Clan Donald history helps explain why the leaders of a very kinship and ancestry driven gaelic society were willing to follow Somerled and his MacDonald and MacDougal descendants even though Somerled’s father, Gillebride, was simply a Thane of Argyle who had been driven to living in a cave by invading norse.

The MacInnes and MacGillivray of Morvern are historically believed to have been the first groups to follow Somerled. The MacInnes claim that they are descended from the sons of Gillebride, father of Somerled. Following this line is unfortunately made more difficult because John, Lord of the Isles, chose to slaughter the MacInnes chief and his sons due to an alleged slight concerning Amy MacRaurie’s housekeeping skills. From the MacInnes DNA study 3 out of about 50 named M(a)cInnes appear to be Somerled line relatives with the principal difference that they are 15 at 458 instead of 16 for the descendants of John, Lord of the Isles. We have at several Somerled descendants named McKeen in the project with a tradition of descent from Glencoe. Those individuals also have 15 at 458 supporting the finding that the mutation from 15 to 16 first occurred at John, Lord of the Isles. There are also a series of MacNeills of Barra which are Somerled lines with 15 at 458. Those results would be consistent with at least Gillebride already being of Norse paternal descent. Mark believes it is a further argument supporting Godfrey as the originator of the Norse paternal line, especially since three out of the MacInnes comparables to Clan Donald are I-M307 vikings and potentially consistent with the notion that Godfrey brought vikings with him from Ireland. That is obviously not the only way that a viking line would appear in Morvern but provides one potential explanation. Mark believes that the explanation for the story in which Somerled was selected to replace the MacInnes chief who had been killed by vikings, taken together with the Somerled like signatures among the MacInnes is that the MacInnes sought Somerled because he was the younger brother or cousin in the paternal line of their deceased chief. The salmon fishing portion of the story would be consistent with Somerled seeking to obtain the occult wisdom of the salmon which goes back to the Finn MacCool stories of Ireland. The MacInnes have a number of Atlantic Modal Haplotype signatures which are not particularly close to the Clan Donalds in our sample and at least one O’Neill/potential Colla as part of their mix.

Don Schlegel, a researcher tracing the history of his mother’s MacDonell Irish ancestry, has done a very persuasive analysis of the Colla Irish lines between 300AD and approximately 700AD. A branch of Colla Uais descendants, the MacCarthend MacUais who lived on the shores of Loch Foyle just east of current Londonderry and close to the center of O’Neill power, were probably the source of the Fergus which lead to Godfrey “MacFergus.” Fergus was a clear ancestor of Somerled from the early 800s, but not in a “pure” paternal line.

The portion of Schlegel’s book "The Ancestors of McDonalds of Somerset" [Schelgel 2000] describing Colla Uais and the Fergus branch of the MacCarthend MacUais is reproduced here. Please note that Schlegal still believes that the paternal norse descent of our chiefs is simply a non-paternity event which occurred after Godfrey MacFergus. Mark disagrees. The most usual candidate among people disagreeing with Mark for the "earliest reliable R1a" in the line is Gilledomnan.