Introduction

The Three Collas lived in Ireland in the 4th century A.D. Their descendants have been kings, lords, chiefs, and saints. Their history survived through oral tradition and eventually written histories. Recent Y-chromosome DNA tests of men with surnames mentioned in these histories confirm their common ancestry. The names include Biggins/Beggan, Boylan, Calkins/Colcan, Carroll, Connolly, Devine, Duffy, Hart, Higgins, Hughes, Kelly, MacDougall, McAuley, McClain, McDonald, McGinnis, McGuire, McKenna, McMahon, Neal, Newell, Paden, Pate, Roberts, Roderick, etc. Most Irish names are found in multiple septs, so not all people with these names are descendants of the Three Collas. And, some names of people with Clan Colla DNA are not included in the Clan Colla histories due to adoption and other reasons.

We have two people with Colla DNA who can trace their ancestry back to the Three Collas: a McDonald and a McMahon.

DNA has confirmed ancient Irish history, but it has also shown that the history is wrong in some areas. For example, the Three Collas were said to be related to Niall of the Nine Hostages. The DNA of the descendants of the Collas, however, does not match the DNA of the descendants of Niall.

Table of Contents     

An Unusual Test Result. In July 2008, with some skepticism, I had my Y-chromosone DNA tested for 67 genealogical markers. Results showed that I was of Atlantic European ancestry and matched up well with people named Biggins or Beggan. That was interesting but expected. Results also showed a good match with people named Maguire, Carroll, McDonald, McKenna, and McMahon. Furthermore, we all had an unusual null value for DNA marker 425. I described the results on my Web page: Biggins/Beggan Irish Roots.

My Web page related a conversation I had with Gerard Beggan, whom I had met in September 2007 at his home in Carrickmacross, County Monaghan. Gerard told me that in 1969 Rev. Peadar Livingstone (1932-1987) told him that Beggan was a branch of the Maguire family. I had found Gerard's name on the Web site of Al Beagan. Father Livingstone was a renowned scholar in both the Irish language and local history. He wrote comprehensive histories of two counties in Ireland, The Fermanagh Story in 1969, and The Monaghan Story in 1980.

Enniskillen Castle built by Hugh Maguire prior to his death in 1428. On the River Erne in Enniskillen, County Fermanagh. In the 17th century, it became an English garrison fort and later served as part of a military barracks.

Table of Contents     

An Email from Josiah McGuire. In March 2009, I received an email from Josiah McGuire. He was getting reports that he was matching up with me and found my Web page. He said, "I think it is really quite amazing and very interesting that Peadar Livingstone thought the Beggans had their origins from the Maguires, and here, of all things, are some fairly close matches that may confirm this."

Josiah's email went on: "I have been watching and studying the 425 nulls between the Carrolls, McMahons, McKennas, and McGuires since I had my markers upgraded to 67 markers in 2006. I suspected that we probably shared a common ancestor, but very few researchers took my comments very seriously. These surnames and several others who I also have matches with are said to descend from "Colla da Chrioch" as stated in the Irish Pedigrees by O'Hart."

Participation in the Biggins DNA project has turned out to be worth far more than I had anticipated. It confirms that the names Biggins, Beaghen, Beggan, and Little are based on the Irish word for small, beag, as mentioned in Irish surname books by Patrick Woulfe (1923) and Edward MacLysaght (1969). And it confirms what Professor Peadar Livingstone had told Gerard Beggan in the 1970s--that Beggans are related to Maguires.

But most importantly, the DNA project established a connection with ancient Irish history. We Beggans were no longer just a humble people with a name based on the Irish word for little. With big names like Carroll, McMahon, McKenna, and Maguire, we were descended from the Three Collas who lived in the 4th century and established the ancient kingdom of Oriel in Monaghan and Fermanagh and later Hy-Maine in Galway and Roscommon.

In February 2008, the Carroll DNA Project posted this news: "We think that a group of our participants may have hit the Jackpot! They may be related to the O'Carroll Princes of Oriel (Monaghan and Louth). This Kingdom located in the North of Ireland was founded by the three Collas Brothers around the year 327 AD. Wow! Will keep you posted on this as we get more news."

Clan Donald has been collecting DNA since 1999. Their Web site has now established a Magenta subgroup that they say is parallel to "the McMahons of Fermanagh (one of the territories of ancient Oriel founded by the Collas who allegedly conquered Ulster around 330 AD). . . . It is distinguished by a null value for marker DYS425."

In December 2008, the McQuillan Clan Association announced "DNA Project Sheds New Light on McQuillan Connections." Specifically, they said that "our first "cluster" shows recent shared ancestry for Monaghan & Fermanagh McQuillans. The close match between these three McQuillans reveals the first localized McQuillan haplotype cluster to emerge in our study. This cluster also shows a relationship with some Cullens, and perhaps with the McMahons of Monaghan."

Table of Contents     

By Four Colla Descendants. This study of the Y-chromosome DNA of the Three Collas is the product of four descendants of the Three Collas--Colla da Crioch to be specific.

• Josiah McGuire. Josiah McGuire had his DNA tested in 2004. By early 2005 he recognized that he matched several other people with Colla surnames and began studying those matches. After upgrading to 67 markers in 2006, Josiah learned that he had a null value for marker 425, as did the other Colla surnames. In 2007 Josiah computed modal values for the DNA of the Three Collas and put it on Ysearch with the user ID of DURRQ. In June 2009, he started the Clan Colla Null 425 Project at FTDNA to attract Clan Colla descendants, encourage upgrades to the 67-marker test, and promote Clan Colla research. He is also the administrator of the Mag-Uidhir Clan Project. Josiah's DNA is kit #23171. His genetic distance from the Colla Modal DNA is 9. He traces his family back to the eponymous Josiah McGuire, who was born in 1794 in the Commonwealth of Virginia. Josiah lives in Indiana, USA. His email address is: j_j_mcguire@msn.com
  
  
• Peter Biggins. Peter had his DNA tested at FTDNA in 2008. When he realized with Josiah's help in May 2009 that he was descended from the Three Collas, he added this page on the DNA of the Three Collas to his genealogy website called PetersPioneers. Peter is an administrator of the Biggins Project and the Ely Carroll Project at FTDNA as well as the Clan Colla Null 425 Project. His DNA is kit #127469. His genetic distance from the Colla Modal DNA is 6. Peter has traced his family back to Patrick Biggins (Beggan), who was born in 1807 in Ireland, most likely Drumgill, County Cavan. Peter lives in Connecticut, USA, where he a director of the Middlesex Genealogical Society. His email address is: pabiggin@optonline.net
  
  
• Patrick McMahon. Patrick Ciaran McMahon had his DNA test in 2009. He is an administrator of the Clan Colla Null 425 Project. In addition to being a fellow Colla descendant, Patrick spent a good part of his career working as a geneticist and lives in Ireland. Patrick's DNA is kit #145687. His genetic distance from the Colla Modal DNA is 4. Patrick has an advanced degree in genetics from Trinity College Dublin. He has traced his family back to Faolan MacMathghamhna (Felim/Phelan MacMahon), who lived in County Monaghan in the early 12th century (see 49 Generations: Colla to McMahon). Patrick also has done an in-depth analysis of McMahon DNA and Colla and Its Wider Diaspora. Patrick lives in Gorey, Ireland. His email address is: mcmahongorey@utvinternet.com
  
  
• Thomas Roderick. Thomas H. Roderick, PhD, was one of the original members of the international Human Genome Organization in 1988 and had his DNA tested at FTDNA in 2003. He is administrator of the Roderick-Rhydderch Family Project at FTDNA as well as the Clan Colla Null 425 Project. In addition to being a fellow Colla descendant, Tom is a research geneticist and an Associate Editor of the Journal of Genetic Genealogy. He is on the Emeritus Staff at The Jackson Laboratory. Tom's DNA is kit #8551. His genetic distance from the Colla Modal DNA is 8. He has a PhD in genetics from the University of California, Berkeley. Tom is a member of the National Genealogical Society and has traced his family back to Rhydderch Evan who was born circa 1700 in Llantrisant, Glamorgan, Wales. Tom lives in Bar Harbor, Maine, USA. His email address is: roderick@gwi.net

  Table of Contents     

FTDNA Clan Colla DNA Project

Two of the three swab kits in the FTDNA Kit. For more on how it's done, see Dave Dorsey.

If you are a male with the name of a Colla descendant, you may benefit from participating in the Clan Colla 425 null project at Family Tree DNA. FTDNA has the largest DNA database in the field.

Josiah McGuire started the Clan Colla 425 Null Project in 2009 and serves as administrator of the project along with Peter Biggins, Patrick Mcmahon, and Thomas Roderick. The project is designed to attract Clan Colla descendants, encourage upgrades to the 67-marker test, and promote Clan Colla research.

You can participate in our Colla project as well as a project specifically set up for your surname. There is no additional cost for being part of two projects.

By testing the Y-chromosome DNA, males can determine the origin of their paternal line. Note that the Y-chromosome DNA strictly checks the paternal line, with no influence of any females along that line. Females do not receive the Y-chromosome, and therefore females cannot be tested for the paternal line. If you are a female and would like to know about your paternal line, you would have to find a brother or a male relative from that line willing to be tested.

You sign up online for FTDNA and they deduct the cost from your credit card. They send you in the mail a kit containing three scrapers that you use to swab the inside of your cheeks in four-hour intervals. You return the scrapers in receptacles and mailer provided in the kit. You get final results on line two months later.

If you decide to have your DNA tested, you should choose the 67 markers. The lesser tests of 12, 25, or 37 markers do not include markers 511, 425, 413a, and 481, which are key to verifying a match with Colla descendants.

Most names have multiple origins. For example, there are Monaghan McMahons (Colla descendants) and Clare McMahons (not Colla descendants). For this reason, your results may show that your DNA does not match the Colla DNA, which will lead you in a different ancestry direction.

Predicting a Clan Colla Match. If you have a Colla name, you can use the values of key markers in the table to the right to predict your chances of matching the DNA of Colla descendants. If you have done the 12-marker test, you have a better chance of matching Clan Colla if you have a value of 15 or more for Marker 385b and 13 or more for 439. If you have neither marker, you stand practically no chance of being a Clan Coll descendant. If you have both, you have a better chance of being descended from the Three Collas than if you have only one.

If you have done the 25-marker test, you have a better chance of matching Clan Colla if you have a value of 15 or more for Marker 385b, 13 or more for 439, and 28 or less for 449. If you have none of the three markers, you stand practically no chance of being a Clan Coll descendant. If you have two or three, you have a better chance of being descended from the Three Collas than if you have only one.

If you have done the 37-marker test, you have a better chance of matching Clan Colla if you have a value of 15 or more for Marker 385b, 13 or more for 439, 28 or less for 449, and 18 or more for 570. If you have none of the four markers, you stand practically no chance of being a Clan Coll descendant. If you have three or four, you have a better chance of being descended from the Three Collas than if you have only one or two.

If you have a Colla surname, you can join the Colla project before receiving test results for all 67 markers.

Following is a chart showing the Y-DNA of the current participants in the Clan Colla 425 null project.

Table of Contents     

Colla Surnames

Ireland was one of the first European countries to adopt hereditary surnames. At the end of his chapter on the Middle Ages in his 1969 book The Fermanagh Story, Rev. Peadar Livingstone (1932-1987) says on pages 23 and 24 that "Irish people in this era seem to have been obsessed with names. Long pedigrees are drawn up, giving the origins of most common families. As might be expected, most of the Fermanagh families trace themselves back to an Oriel origin. This, for the most part, is genuine enough. However, since an Oriel line ruled the country, it must have been popular to have Oriel origins. Some of the earlier Leinster Fir Manach must have been tempted to invent an Oriel connection where it did not exist." With that caveat in mind, we will take a look at the surnames of Clan Colla. See also Multiple-Sept Surnames.

Surnames from Hy-Maine have been removed from O'Hart's list of Colla surnames because DNA has shown them to have a different DNA (see Hy Maine Modal):

• O Kelly, princes and lords of Hy-Maine, a territory in Galway and Roscommon. See O'Hart, page 684.
• Madden, lords of Siol Anmcha or Silancha, which ancient territory comprised the present barony of Longford, in the county Galway, and the parish of Lusmagh on the other (Leinster) side of the river Shannon, near Banagher, in the King's County. See O'Hart, pages 568 and 572.
• Traynor. See O'Hart, page 572-573.
• Naghtan or Norton, chiefs in Hy-Maine See O'Hart, page 603
• Hoolahan, chiefs of Siol Anmchada in Hy-Maine. Also Oulahan, a branch of Hoolahan, and variants Holland and Holligan. See O'Hart, page 487
• Leahy, chiefs in Hy-Maine. See O'Hart, Volume II, page 577

Several Colla names not listed by O'Hart have been accepted on the authority of Peadar Livingstone or Edward MacLysaght:

• McAuley, a variant of McCawley, which is listed by Peadar Livingston in The Fermanagh Story (exerpt), page 436, as a descendant of Maguire.
• Biggins/Beggan/Beaghen/Little/Bigham, which was said by Peadar Livingstone to be descended from Maguire.
• Collins, a variant of Callan, which MacLysaght shows as an Oriel sept from Armagh/Monaghan.
• O'Donoghue, a variant of Donaghy, which is listed by Peadar Livingston in The Fermanagh Story (exerpt), page 426, as a descendant of Maguire.

John O'Hart (1824-1902) said he himself was a descendant of the Three Collas. He is generation No. 125 on page 679. As indicated in his footnote on page 678, one of John's relatives was John Hart (1713-1779) who signed the Declaration of Independence for New Jersey on July 4, 1776. There are several Harts who have Clan Colla DNA: kit Nos. 48620, 166797, and 182999. At least one of these Harts, 166797, goes back to James Hart, born in 1835 in County Cavan, Ireland. However, there are several Harts who claim descendance from Hart the Signer who do not have Colla DNA: kit Nos. 48743, 173068, and 196006. At least one of these Harts goes back to Joseph Hart, born in 1688 in Newtown, New York. Information on John Hart the Signer can be found at: Descendants of the Signers of the Declaration of Independence, by members Grace Keiper Staller and Thornton C. Lockwood; Independence Hall Association, by Thomas E. Kindig; Glen Valis, a descendant of John Hart.

Table of Contents     

Annals of the Four Masters. The Annals of the Four Masters were composed from 1632 to 1636 in the Franciscan Monastery of Donegal chiefly by Michael O'Clery (1580-1643). In 1845, Irish historiographer Owen Connellan translated the book from Irish to English. In May 2007, Google digitized a copy of the translation from the Library at Oxford University: Annals of the Four Masters. A long footnote that starts on page 2 describes the ancient kingdom of Oriel and includes the descendants of the Three Collas. "The posterity of the three Collas, called clan Colla, founded many powerful clans and noble families in Ulster and other parts of Ireland.

• From Colla Uais were descended the MacDonnells, earls of Antrim in Ireland, and lords of the Isles in Scotland; also the MacRorys, a great clan in the Hebrides, and also many families of that name in Ulster, anglicised to Rogers.
• From Colla da Chrich, were descended the MacMahons, princes of Monaghan, lords of Ferney, and barons of Dartree, at Conagh, where they had their chief seat. The MacMahons were sometimes styled princes of Orgiall. An interesting account of the Mac Mahons, of Monaghan, is given by sir John Davis, who wrote in the reign of James the First. It may be observed that several of the MacMahons in former times changed the name to Mathews. The other chief clans of Monaghan were the MacKennas, chiefs of Truagh; the MacCabes; the MacNeneys, anglicised to Bird; the MacArdells; MacCassidys; O'Duffys, and O'Corrys; the O'Cosgras, MacCuskers or MacOscars, changed to Cosgraves, who possessed, according to O'Dugan, a territory called Fearra Rois, which comprised the district about Carrickmacross in Monaghan, with the parish of Clonkeen, adjoining, in the county of Louth; the Boylans of Dartree; the MacGilMichaels, changed to Mitchell; the MacDonnells; the O'Connellys, and others.
• From Colla-da-Chrich were also descended the MacGuires, lords of Fermanagh, and barons of Enniskillen; the O'Flanagans of Fermanagh; the O'Hanlons, chiefs of Hy-Meith-Tire, now the barony of Orior in Armagh, who held the office of hereditary regal standard-bearers of Ulster; the MacCathans or MacCanns of Clan Breasail, in Armagh; the O'Kellys, princes of Hy Maine, in the counties of Galway and Roscommon; and the O'Madagans or O'Maddens, chiefs of Siol Anmchadha or Silanchia, now the barony of Longford, in the county of Galway."

Table of Contents     

Don Schlegel's Colla Family Trees. Jim McMahon's Web site, Clan McMahon of the Kingdom of Oriel, has several family trees for the Three Collas. The family trees were created by Donald M. Schlegel.

1. na tri Colla shows the three brothers and the tribes and family names descended from them.
   • Colla Uais: MacDugal, MacRory, MacDonald of Scotland, O'Flynn, O'Donnellean
   • Colla da Crioch: Maguire, MacMahon, MacCafferty, MacDonnell of Clan Kelly, MacManus, O'Connolly, O'Hanratty, O'Cooney, O'Neillan, O'Hanlon, Kearney
   • Colla Menn: O'Heenan, MacAllan, O'Lannan
2. Ui Chrimthainn, a particular tribe descended from Chrimthann Liath, the great great grandson of Muredach Colla da Crioch, the first King of Airghialla. You can see on this chart that Eochaid, son of Crimthainn Liath was King of Airghialla and a contemporary of Saint Patrick.
   • Daimin: O'Hart, MacDonnell of Clan Kelly, O'Kelly, O'Keran
   • Cormac: Maguire, McManus, McCaffery
   • Nadsluag: O'Carroll of Oriel, McMahon
3. Clan Nadsluaig from whom descended the MacMahons and the O’Carrolls. This chart shows Clan Nadsluaig descending from Fergus, son of Nadsluag, grandson of Eochaid. As you follow this chart, you can see that the Lords of Fernmaig were of this line, including Mathgamna, from whom came the name MacMahon. From the line of Mathgamna come both the families of O’Carroll, known as the Kings of Oriel at the height of its power just prior to the coming of the Normans, and the MacMahons, the latter descending from Niall, brother of Murrough O’Carroll the last O’Carroll King of Oriel. (According to "The MacMahon Pedigree: A Medieval Forgery," by Katharine Simms, in Regions and Rulers in Ireland, c.1100-1650, David Edwards, Editor, 2004, the McMahons did not descend from O'Carrolls but co-existed with them in the same time frame. See also: 49 Generations: Colla to McMahon.)

Table of Contents     

Maguires. The Maguires of County Fermanagh, descendants of Colla da Crioch, spawned many other Colla surnames in the period 1250 to 1350 AD.

• Edward MacLysaght (1887-1986) in his Irish Families, Their Names, Arms and Origins mentions nine surnames descended from the Maguires: MacAwley, McCaffrey, O'Corrigan, Corry, Devine/Davin, Fitzpatrick, O'Hanraghty/Enright, MacAlilly/Lilly, and McManus.
• Rev. Peadar Livingstone (1932-1987) in his 1969 book The Fermanagh Story (exerpt) includes 15 surnames descended from Maguire: MacAuley, Breen/McBrien, McCaffery, Corry, McCusker, Donaghy/Donahoe, McElroy/Gilroy, Fitzpatrick, Gilleece/McAleese, McGrath, McHugh/McGee, Lilly, McLaughlin (possibly), McMahon, McManus, Martin (possibly), Murphy. Father Livingstone told Gerard Beggan in Clones in 1969 that the Beggan surname also is descended from Maguires. Gerard lives in Carrickmacross, County Monaghan.
• Maguire Princely Pedigree, by Jim Maguire, presents a chart of descendants of the Maguires, including 12 surnames descended from Maguire: MacAuley, Breen/McBrien, McCaffery, Corry, Gilleece/McAleese, McGrath, McHugh/McGee, McLaughlin, Lilly, McMahon, McManus, Murphy.

1969 map of surnames in County Fermanagh from Chapter 31, "Fermanagh Families," in The Fermanagh Story by Rev. Peadar Livingstone, Cumann Seanchais Chlochair, 1969.

1979 map of surnames in County Monaghan from page 51 of The Monaghan Story by Rev. Peadar Livingstone, Clogher Historical Society, 1979.

RootsWeb. Another excellent source for information on the Three Collas is the RootsWeb article on the Kingdom of Airghialla by Dennis Walsh.

Table of Contents     

Clan Colla Saints

In his Irish Pedigrees; or, the Origin and Stem of the Irish Nation, published in 1892 (fifth edition), Volume II, page 581, John O'Hart says The Four Masters record 39 saints as descended from the Three Collas: 19 from Colla-da-Chrioch, 16 from Colla Uais, and 4 from Colla Meann.

The following were the 19 saints descended from Colla-da-Chrioch.

1. St. Begg (1st August)
2. St. Brughach (1st Nov.)
3. St. Curcach, virgin
4. St. Daimhin (or Damin), abbot of Devenish Abbey (see page 189), on Devenish Island, Lough Erne.
5. St. Defraoch, virgin.
6. St. Donart.
7. St. Duroch, virgin.
8. St. Enna of Aaron (21st Mar.)
9. St. Baodan (5th Feb.)
10. St. Fergus (29th March)
11. St. Fiachra (2nd May)
12. St. Flann Feabhla (20th April)
13. St. Lochin, virgin
14. St. Loman of Loughgill (4th Feb.)
15. St. Maeldoid (13th May)
16. St. Mochaomog.
17. St. Muredach (15th May)
18. St. Neassa, virgin
19. St. Tegan (9th Sept.)

Saint Berchan. Saint Berchan lived in the 5th century and is listed on page 17 of the The Martyrology of Donegal: A Calendar of the Saints of Ireland, translated by John O'Donovan in 1864. He lived on Inish-Rochla, an island in Lough Erne near Enniskillen in County Fermanagh, Ireland. He was five generatiions down from Colla da Crioch, one of the Three Collas. His feast day is November 24.

Saint Cinnia. Saint Cinnia lived in the 5th century. She was a princess of Ulster, Ireland. Saint Cinnia was converted to Christianity by Saint Patrick. When she entered a convent, Saint Patrick gave her the veil. She was descended from Colla da Crioch, one of the Three Collas. Her feast day is February 1.

Saint Tigernach. Saint Tigernach was said to have been the godchild of Saint Brigid, and educated in Scotland. He may have been a monk at Clones as well as a bishop of Clogher in County Monaghan, but accounts are not too clear. He also is called Tierney and Tierry. Saint Tigernach died in 549. He was descended from Colla da Crioch, one of the Three Collas. His feast day is April 4.

Saint Cairnech. Saint Cairnech was born after the middle of the 5th century. His brothers were St. Berchan and St. Ronan. His monastery was probably at Cruachan Ligean on Loughh Foyle, near Lifford. He died about 530. He was descended from Colla da Crioch, one of the Three Collas. His feast day is March 28.

Table of Contents     

Clan Colla DNA Study

Summary Report

This story and study of the DNA of the Three Collas was started in May 2009. A rudimentary modal Colla DNA was established from a group of people with Colla surnames such as Biggins, Carroll, Maguire, McMahon, McDonald, and McDaniel, who had similar DNA. This was facilitated by the fact that these people all had an unusual feature in their DNA: one of the 67 markers, number 425, had a null value. It was relatively easy to search among the DNA of these people for the ones that had the null value for marker 425. Not all of the people with these surnames had the null value but the ones that did all seemed to match up fairly closely. The ones that did not have the null value did not match up closely. The next step was to search for all the people, regardless of surname, who matched up fairly closely with the developing modal DNA. This has resulted in a list of people with Colla DNA--some with Colla surnames and some without. There were plausible reasons for the existence of this latter group: adoption, name change, genealogical omissions, and distant cousins.

A Reference Group was put together of people who do not have the Colla DNA. This group helps to understand how close the people are who have the Colla DNA. It also helps make predictions about closeness to Colla DNA based on markers other than the null 425.

The ancient genealogies and DNA test results seem to be confirming each other. A pattern or "signature" DNA has emerged for Colla descendants identified long ago by John O'Hart and his predecessors. And the ancient genealogies have given us a clue as to which people have Colla DNA.

Table of Contents     

DNA Testing. By testing the Y-chromosome DNA, males can determine the origin of their paternal line. Note that the Y DNA strictly checks the paternal line, with no influence of any females along that line. Females do not receive the Y-chromosome, and therefore females cannot be tested for the paternal line. If you are a female and would like to know about your paternal line, you would need to have a male relative from that line to be tested.

The test results evaluated here all came from Family Tree DNA. Only 67 or 111 markers tested are included here because those tests include the 48th marker, 425, which is crucial to the analysis. The cost of this test varies from time to time, but the average is approximately $250. More advanced "deep-clade testing" provides more specific information about origin but is not essential to this study.

Table of Contents     

Colla Modal DNA. The genetic distance for participants in all groups is computed from a Colla Modal DNA. This study began with a preliminary Colla modal DNA for relatively small number of people who had Colla names and the null value for marker 425. This modal DNA evolved into a modal DNA that was essentially the same as one established by Josiah McGuire in June 2009, based on data from the Colla DNA Project, under the user ID of DURRQ at Ysearch. This DURRQ Colla Modal DNA is now the one used in this study. Since June 2009 the database of people with Colla DNA has expanded and the modal has been recomputed. Each time, the modal has remained the same.

There are several over-represented families included in the database now. To assure that these were not skewing the data, the five largest surname groups were removed, and the modal was recomputed without them. The resulting modal, however, remained the same.

Table of Contents     

Genetic Distance. The next step was to compute the genetic distance from Colla Modal DNA for each person in the study. Genetic distance occurs because of mutations from one generation to another. If two people are identical in all markers except they are off in one marker by 1 point, the genetic distance would be 1. If they were off at 2 different markers by 1 point in each marker, then the genetic distance of those two samples would be 2. If they are off by 2 points at one marker and 1 point in a second marker, then the genetic distance would be 3. The value for marker 425 when it is not 0 is usually 12, but the genetic difference between 0 and 12 is treated as 1. Based on FTDNA practice, genetic distance for certain markers or marker groups is limited to 1. This method of computing genetic distance is called the hybrid mutation model. FTDNA revised its method of computing genetic distance in 2011. Our study used the revised method.

Actual calculations were made using the FTDNA 111 Mode BETA version of the McGee Utility. Input to the utility consists of the DNA of the participants from the Source Data. Output consists of a Genetic Distance Matrix which shows the genetic distance of each person from the Clan Colla modal as well as the genetic distances among all participants. The genetic distances were then added to the Source Data.

The last step was to average the genetic distances for each group and various subgroups.

Table of Contents     

Distribution of people by genetic distance from Colla Modal DNA. For example, it shows that 40 people have a genetic distance of 4 from the Colla Modal.

Distribution of genetic distances in DNA among participants. For example, it shows that 232 people have 2,161 genetic distances of 7 among themselves.

Scattergram of genetic distance from Colla modal and number of matches up to -7 for 232 people. For example, there is one person who has a genetic distance of 4 from the Colla modal and has 27 matches with genetic distance of 7 or less. Another with a GD of 4 has 105 matches.

Summary of Study Results. As of June 2011, there are 232 people included in the Colla Group. They have been included because their DNA has been found to match fairly closely with Colla Modal DNA. Of these 232, 148 are included in the Clan Colla 425 null project at Family Tree DNA. The rest have been by searching surname studies at FTDNA and through the use of Ysearch.

The DNA of Clan Colla is part of a haplogroup called L21 (also called R1b1a2a1a1b4). Of the 232 participants, 55 have done "deep-clade testing" for the L21 SNP (single-nucleotide polymorphism). All have tested positive. Clan Colla respresents a small part of the L21 haplogroup--perhaps 3%. Other groups have been identified within L21 besides Clan Colla. The major ones are: Northwest Irish (Niall of the Nine Hostages), Irish Type II (South Irish), Irish Type III (Brian Boru), Airghialla 2, Ely Carroll, Dalriada (Scots). For more information on these groups, see L21 DNA.

Most of the participants, 80%, have surnames that the ancient genealogies say are descended from the Three Collas. The surnames are related to both Colla Uais and Colla da Crioch. There are none related to Colla Menn, but relatively few surnames are attributed to him.

Many people do not know where their patronymic ancestor came from, which is not uncommon. Only a small number of those tested live in Ireland or Scotland. Most live in America. Many of those have resorted to DNA testing for the very reason that they do not know where there ancestors came from when the emigrated to America.

The Colla Group includes some people with non-Irish sounding names. It includes some people who are related to each other. It includes surnames where there is only one representative.

The genetic distance between the Colla Group and Colla Modal DNA ranges between 1 and 11, and averages 6. Of the 232 people, 219 or 94% of the group have a genetic distance of 3 to 9.

There are 26,796 possible comparisons among the 232 people: n*(n-1)/2, where n is the number of people. The genetic distances for these pairs range between 0 and 21. The average is 9.9. The total of such genetic distances up to 7 is 4,937, or 18% of the total possible matches. So, the matches that Colla people see are probably all fellow Collas, but only 18% of the total. If FTDNA were to raise the limit to 10, people would see 57% of the total possible matches, but they might also see some non-Colla matches.

FTDNA shows each participant his 67-marker matches up to a genetic distance of 7 on his homepage. And it allows participants to restrict the showing of their matches to the people in their surname project. The theoretical 67-marker match experience (within a genetic distance of 7) varies considerably by individual, from 1 to 145. The average is 43. The person with 145 theoretical matches has only 94 showing on his homepage at FTDNA because some participants restrict display of their matches to the people in their surname project. Another person has a theoretical 33 but actually sees only 25 because of the restrictors. Yet another person has a theoretical 55 but actually sees only 45 because of the restrictors.

For more detailed information on the Colla Group, see Colla Group Detail and Source Data. Source Data is an Excel spreadsheet that includes the following for 232 testers:

• Brother-assigned Colla brother
• Group-the subgroup to which the person has been assigned out of a total of 38. See Colla Group Detail
• Surname
• Kit-the tester's kit number at FTDNA or in some cases a Ysearch ID or an ID unique to a surname project
• Haplogroup-assigned by FTDNA depending on when they were tested and whether they have done "deep clade testing", such as R1b1, R1b1a2, R1b1a2a1a1b4
• Ancestor-most distant ancestor specified by the tester
• Location-where the most distant ancestor lived
• Genetic Distance (GD)-see Genetic Distance
• Values for 67 Y-chromosome markers

Table of Contents     

Historical Surnames. Within the Colla group, 146 have been assigned to Muredach Colla da Crioch and 54 have been assigned to Carrell Colla Uais. None have been assigned to Aedh Colla Menn. Assignment to a brother is based on the tester's surname appearing in O'Hart's Irish Pedigrees. O'Hart indicates which brother the ancient histories say the name is descended from. In the case of McDonald, however, the name is descended from both Crioch and Uais.

Those assigned to a brother have 35 different Colla surnames. Only 30 percent of O'Hart's list of Colla Surnames are included in the study. There are a number of good reasons.

• Many Colla names are uncommon.
• Only a small number of people have had their DNA tested thus far.
• Many who have had their DNA tested have tested only 12, 25, or 37 markers rather than the 67 required for this study.
• Some people have been tested by an organization other than Family Tree DNA.
• Some people may have lost their name because an ancestor changed his name or was adopted. These people would show up in the Unassigned Group.
• O'Hart probably included some surnames of people who were not really Colla descendants.

• There was a name change by an ancestor.
• They or an ancestor were adopted. Seven are known to be adopted or have an adopted ancestor.
• The historical lists of Colla descendants were incomplete.
• The name is on an historical list of Colla descendants but we have not found it yet.
• Their ancestors were descended from cousins of the Three Collas; and their surnames, therefore, evolved differently.
• In early Irish history there was the concept of “fostering,” where two powerful tribal leaders would place their infant son with the other family to seal a defensive alliance. It is likely that some of these sons took on the tribal name of the family with whom they were placed.

First sketch of the deoxyribonucleic acid double-helix pattern in 1953 by Francis Crick.

Table of Contents     

Null Value for Marker 425. All Colla participants by study design have taken the 67-marker test conducted by FTDNA. One of the markers in the 67-marker test that is not in lesser tests is marker 425. A null value for marker 425 separates the Colla Group from the Reference Group, but it only accounts for a genetic distance of 1. The remaining distance between the mean of 6 for the Colla group and 23 for Reference Group is due to differences in other markers.

A special test, called the DYF371X test, is offered for those who have the null value for marker 425, and a Null 425 DNA project. The DYF371X test has been done for 35 Clan Colla descendants: 34 have subvalues of 10c-12c-13c-14c and one has subvalues of 10c-12c-14c-14c.

As indicated by the Null 425 DNA project, there are people with the null value for marker 425 other than Colla people. These people do not match the Clan Colla modal. Only one of them has the L21 SNP.

There are 7 people who match the Colla Group but do not have a null value for marker 425. Their genetic distances from the Clan Colla modal range from 3 to 10. It is very unlikely that a null value would ever be reversed. So, these 7 people may be descended from ancestors who existed sometime before the null value first occurred. See No Null 425 Detail.

Table of Contents     

Colla Versus Reference DNA. Each DNA marker of the 232 Clan Colla participants was compared the corresponding DNA marker of a reference group of 2,710 other people. The people in the reference group have the same L-21 haplotype as the Colla people. The reference group people were obtained from the database maintained by Mike Walsh as of June 2011. Mike is administrator of the R-L21 project and recognized as the major source of L21 data.

The percentage distribution of marker values was determined for each group. The eight markers with the largest differences in the distribution of values were 385b, 439, 449, 570, 511, 425, 413a, and 534. For the source data and frequencies for all 67 markers for the Colla and the reference group, see: Reference Source Data.

Table of Contents     

Colla Subgroups. To better understand the DNA that has been accumulated, the Colla group has been broken down into 38 subgroups. The groupings are based on:

• Surname
• Family genealogy
• Genetic distance among group members
• Uniqueness of certain markers

Patrick McMahon is using network software provided by Fluxus to show how the subgroups of the Colla Group members relate to each other.

There is a Fluxus User Guide that explains the technical details of the network.

Results will be posted here as they become available.

In addition to being a fellow Colla descendant, Patrick has an advanced degree in genetics from Trinity College Dublin and spent a good part of his career working as a geneticist. Patrick lives in Ireland and has traced his family back to Faolan MacMathghamhna (Felim/Phelan MacMahon), who lived in the early 12th century.

Table of Contents     

L21 and DF21 SNPs

All Clan Colla participants who have done "deep-clade testing" for the L21 SNP (single-nucleotide polymorphism) have tested positive. L21 was discovered in October 2008. People with the L21 SNP are said to be members of the R1b1a2a1a1b4 haplogroup. (L21 is known as S145 in some testing organizations.)

Clan Colla particpants have tested negative for all of the recognized downstream SNPs: M37, M222, P66, L96, L144, L159.2, L193, L226, P314.2.

In August and September 2011, however, several Clan Colla participants tested positive for a new SNP called DF21. This further narrows the haplogroup for Clan Colla descendants. All Clan Colla descendants are expected to have the DF21 SNP. Other groups also are expected to have the DF21 SNP--perhaps 10 percent of all those with the L21 SNP. The DF21 SNP is estimated to be 2,500 to 3,000 years old, compared with 4,000 for the L21 SNP.

Clan Colla participants are urged to join the DF21 project at FTDNA and order the DF21 SNP test. The DF21 project has been set up by David Reynolds for people who have tested positive for the DF21 SNP or are interested in ordering the test.

The DF21 SNP is not yet part of deep clade testing and it may be a few months before FTDNA creates a new haplogroup designation. To order the DF21 test from your FTDNA home page, go to ORDER ADVANCED TESTS (not ORDER ADVANCED SNP TESTS nor ORDER DEEP CLADE TEST). The cost is $29.

A R-L21 project and R-L21 Walk Through the Y project have been set up for people whose DNA haplogroup is L21. This type includes other subgroups as well as Clan Colla. The purpose of the project is to find a special identifying a SNP for subgroups of L21. SNPs are very difficult to find. This effort is being directed and managed by Dr. Thomas Krahn, Technical Laboratory Manager of FTDNA's Genomics Research Center in Houston, Texas.

A Discussion Group has been set up for the R-L21 project at Yahoo Groups. Included is a spreadsheet compiled by Michael Walsh that identifies Clan Colla members. The discussion includes the DF21 SNP.

Several subgroups within L21 have been identified and are outlined below. Based on very rough estimates, these groups constitute 45-50% of L21.

The genetic distances among the modal DNAs of the groups listed above range between 10 and 26 as shown in the following table.

Patrick McMahon analyzed the geographic distribution of the members of the L21 project at FTDNA in January 2011 and made the following observation.

Assuming today's testers are a random sample, these results support the views put forward by many that the L21 SNP occurred somewhere north of the Alps (about 4,000 years ago) and the L21 population drifted Northwest over time concentrating in the western fringes of the British Isles mainly in Ireland. North of the Alps would most likely be Germany or France where the original (presumed) high numbers would over time be replaced by further waves of migrants or simply driven North by more advanced civilisations.

How they made their way to Ireland is open to conjecture. The shortest sea journey then (3,000 to 4,000 years ago) as now would be from France to Southern England. However, they could have made their way directly to Ireland from Brittany (or via Cornwall or Wales). The figures support the view that there was no significant migration towards the Eastern parts of Europe and only minor ones to Scandinavia and Spain with the main thrust through Northern France to Britain and Ireland. Archaeologists have termed these peoples (and the proto-Collas are part of this population) as 'Bronze-age' Britons.

For a good presentation of the origins, age, spread, and ethnic association of Europeans see The Peopling of Europe and Eupedia. Following is a haplogroup timeline taken from Eupedia. "ybp" is years before the present.

Table of Contents     

Multiple-Sept Surnames

In December 2010 a study was done of the different septs represented within surnames found in the Colla study. The following table shows the distribution among various R1b septs of people with ten of the more common Colla surnames. Only four of the ten surnames have a majority in Clan Colla: Calkins, Roderick, McKenna, and Biggins. The other six surnames have a minority in Clan Colla.

The descendants of the Three Collas share a common pattern of DNA markers, and many have surnames mentioned in ancient genealogies as descending from the Three Collas.

The people with Colla DNA, however, do not have to be the majority of those with their surname. In fact, the majority of people with Colla DNA are a minority of those with their surname. And, therefore, most people with Colla surnames do not have Colla DNA. The reason is that most Irish surnames appear in mutiple septs. For example, as shown above, there are McDonalds descended from Clan Colla, but there are also McDonalds descended from Niall of the Nine Hostages (Northwest Irish) and Dalriada Scots. There are many McDonalds with a R1b haplotype for whom a sept has not been identified. There are 61 McDonalds with a R1a haplotype descended from Norse invaders. There are McGuires descended from Clan Colla, but there are also many Maguires descended from a sept of unknown origin in the Fermanagh area, called Airghialla 2.

There are several explanations for this phenomenon of multiple-sept surnames.

• The same surname developed independently in different geographic areas.
• At the time surnames came into being around a thousand years ago, many surnames were based on relatively common given names.
• A male was adopted by a clan other than the one he was born in.
• A male married a woman from another clan and took her surname, perhaps because it was a more respected name.
• A male changed his surname when he was ennobled or otherwise came into possession of territory, perhaps adopting the name of a respected prior holder of that territory.
• A male was a subordinate (vassal, servant, slave, etc) of a member of another clan and took his master's surname when he became free.
• A male took the surname of another clan without any connection to the clan, simply because it was a respected name.

Table of Contents     

Two McGuire Septs

In August 2009, Joseph Donohoe reported on the DNA of descendants of the Three Collas in his Breifne Clans DNA Report 5, Subgroup O1, posted at Donohoe Clan. He independently came up with essentially the same group as the Colla Group above, which he calls Subgroup O1 (also called Airghialla 1). As part of his study, Joseph established a modal DNA for his Subgroup O1 under the user ID of WHYAA at Ysearch. It is the same as the DURRQ modal used here, but he calls it an Airghialla rather than a Colla Modal because he is "not fully persuaded yet of the validity or applicability of the Colla tradition, particularly in view of the great number of traditionally Colla surnames not represented here."

To test the validity of Subgroup O1 (also called Airghialla 1), Joseph came up with a second group called Subgroup P1 (also called Airghialla 2). As part of his study, Joseph established a modal DNA for his Subgroup P1 under the user ID of 9U5BW at Ysearch.

In comparing Subgroups O1 and P1 on page 184 of the report, Joseph says that Subgroup O1 "appears to have been prominent in the South Tyrone – North Monaghan area from the sixth century, if not earlier," while Subgroup P1 "rose to historical prominence later . . . in the ninth century." He concludes that Subgroup O1 "would appear to be the best candidate" to represent the DNA of the Three Collas. His Subgroup P1 is now included in our Reference Group.

Joseph himself is not Airghialla 1 or 2. He is a descendant of Niall of the Nine Hostages, also called R-L222 or Northwest Irish.

Brad McGuire started an Airghialla Mag Uidhir DNA project in October 2010.

There is a Charles Robert McGuire (kit #21228), whose ancestry has been traced back to the the McGuire "Junior Line" from Tempo, County Fermanagh. His DNA is clearly Airghialla 2 rather than 1. There are several sources that refer to this pedigree.

• Maguire Princely Pedigree.
• The Fermanagh Story (exerpt) (1969) by Rev. Peadar Livingstone (1932-1987) - page 116
• Captain Cohonny: Constantine Maguire of Tempo 1777-1834, by W. A. Maguire, Ulster Historical Foundation.

How do we explain the two different Airghialla 1 and 2 McGuires? One of the two Airghialla McGuire groups must have received their name through adoption or some other way at some point in the history of the McGuires.

• An Airghialla 2 McGuire may have adopted an Airghialla 1 McGuire sometime before or after Donn Mor.
• An Airghialla 1 McGuire may have adopted an Airghialla 2 McGuire sometime before Donn Mor.

Chapter 1 is about Early Fermanagh. It does not say that there were two Maguire clans, but it identifies two possible sources of the McGuires: the Fir Manach of Old Leinster and the Three Collas.

• Three Collas. "Who were these people of Oriel or Airghialla? This is one of the many questions . . . that we cannot answer yet. . . . Possibly they lived here as a subject race to the Ultaigh before the Ui Neill eventually 'liberated' them. They may have been the descendants of hostages (Airghialla) captured by the Ui Neill and used by them to 'plant' their new conquest. . . . Later genealogists make them descend from Cormac Mac Airt's grandsons, the Three Collas. . . . Many modern scholars deny that the Collas existed at all." - page 5
• Leinster Fir Manach. "If the legend can be believed, they reached the Upper Lough Erne country long before [400 A.D.]. They hailed from Leinster. According to one story they killed Eanna, the king's son, and had to leave. They . . . . came north-west, entered the country from the east and settled along the north shore of the lake. It is most probable that Lisnaskea was the centre of their operations. . . . Soon we will see the Oriel families come and take control of it." - page 6

Chapters 3 and 4 cover the Middle Ages (500-1300) and the Maguire Years (1300-1589). Here Livingstone expresses uncertainty about the origin of the Maguires several times:

• "Irish people in this era seem to have been obsessed with names. Long pedigrees are drawn up, giving the origins of most common families. As might be expected, most of the Fermanagh families trace themselves back to an Oriel origin. This, for the most part, is genuine enough. However, since an Oriel line ruled the country, it must have been popular to have Oriel origins. Some of the earlier Leinster Fir Manach must have been tempted to invent an Oriel connection where it did not exist." - pages 23-24
• "we . . . are doubtful about Fermanagh's greatest family, the Maguires, who come into our picture towards the end of the thirteenth century. Genealogists give them Oriel ancestors. But were they really of Ulster stock?" - page 24
• "where the Maguires come from or from what origins we do not know with certainty" - page 25
• "Possibly they were of old Leinster Fermanagh stock, now again asserting themselves after the Oriel ascendancy." - note 65 (chapter 3)
• "we do not know who the Maguires really were or where they came from. Faulty-looking genealogies give them an Oriel pedigree. Possibly they were descended from the old Leinster settlers" - page 26

What would Peadar Livingstone say if he knew what we know today about the DNA of the Three Collas and people named McGuire and McManus? I think he would say that the McGuires with Airghialla 1 DNA have Oriel roots and are descended from the Three Collas. And he would say that the people with Airghialla 2 DNA are descended from the McGuires perhaps back to Donn Mor McGuire but have other roots such as Leinster Fir Manach or possibly Ultaigh. Other than that, he would have to say "we do not know who the earlier McGuires really were or where they came from."

Table of Contents     

Two Somerled Pedigrees

The Clan Donald DNA project is one of the oldest and largest surname projects at FTDNA. The project administrator is Mark MacDonald, National Historian for Clan Donald U.S.A. The project co-administrator and webmaster is professor J. Douglas McDonald of the University of Illinois.

There are two subgroups in the Clan Donald DNA project with different DNA who descend from different sons of Angus Mor MacDonald, a great grandson of Somerled.

• The R1b Red and Pale Violet subgroups, who have Clan Colla DNA. The latter subgroup includes Frank Everett McDonald (Clan Donald code &NHHMM). Frank's pedigree, 43 Generations: Colla to McDonald, goes back to Somerled through Alasdair Og, son of Angus Mor who was a great grandson of Somerled. Alasdair Og allied himself with Edward I, the king of England, against Robert the Bruce, the titular king of Scotland, in the War of Scottish Independence.
• The R1a Red subgroup, who have Norse DNA. This subgroup includes several McDonald Chiefs and Chieftans (Clan Donald codes &CXYIE, &5XOBA, &EDC4L, and &BATPB). The Chiefs and Chieftans have pedigrees from Burke's Peerage that go back to Somerled through Angus Og, son of Angus Mor who was a great grandson of Somerled. Angus Og deposed his brother Alasdair Og in 1308. Angus Og had allied himself with Robert the Bruce, the titular king of Scotland, who was battling Edward I, the king of England, during the War of Scottish Independence.

The Clan Donald DNA results for its R1b Magenta subgroup state that "signatures parallel to this group can be found among the McMahons of Fermanagh (one of the territories of ancient Oriel founded by the Collas who allegedly conquered Ulster around 330 AD)." Furthermore, their R1b Pale violet subgroup results state that "this group is probably a subset of the Magenta group, but we are not including them together since these people have not tested for DYS425. This group appears to all be descendants of Lt. Brian McDonald, chief line of Leinster and Ulster in Ireland who emigrated to Brandywine Creek Delaware in the late 1600s."

Two opposing positions on the ancestry of Somerled have been taken by Donald Schlegel and Clan Donald:

• Donald Schlegel takes the position that Somerled was Clan Colla, and a Norse descendant adopted the McDonald name sometime later.
• Clan Donald takes the position that Somerled was Norse and that a Colla descendant adopted the McDonald name some time later.

Clan Donald graciously admits that Donald Schlegel has done a "very persuasive analysis" of the Colla Irish lines in his book, The Ancestors of McDonalds of Somerset, and the portion of that book describing the Colla ancestry of Somerled is reproduced on the Clan Donald site.

The oldest son of Somerled was Dugall, Lord of Lorne, from whom MacDougalls are said to descend. The MacDougall DNA project includes two Clan Colla descendants and no Norse descendants. One of the Clan Colla dscendants, David MacDougall (FTDNA kit #21971), has an ancestor, Iain MacDhubhghaill, who is from Lorne according to an 1890 history of Christmas Island.

Table of Contents     

43 Generations: Colla to McDonald

Our study of the DNA of the Three Collas, who lived in the 4th century, is based on present-day Y-chromosome DNA and surnames from ancient pedigrees. It is not based on genealogies which show a paper trail from the 4th century to today. One of the McDonalds in this study, however, can trace his ancestry back to one of the Three Collas, Colla Uais, who lived in the 4th century. He is Frank Everett McDonald, Jr., a retired dairy farmer from the Blue Ridge Mountains of Virginia. He was born there in 1926. His FTDNA kit is #133546. The genetic distance of 67 markers of his Y-chromosome DNA from the Colla Modal DNA is 8. He participates in the Clan Colla 425 Null DNA Project. The genealogy and DNA of Frank E. McDonald, Jr., were brought to our attention by Vaden McDonald, who is unrelated.

The Generations. Frank has 43 generations back to Colla Uais, as shown on the right. Included among the 43 generations are the following:

• 1. Carrell Colla Uais, brother of Muredach Colla da Chrioch and Aedh Colla Menn
• 12. Goffrad: left by Kenneth MacAlpin in the west of Scotland to try to hold the isles against the Northmen. He was styled toiseach, (prince) of Insi-Gall (all of the Hebrides); died in 853
• 20. Somerled (from Gaelic Sòmhairlidh): Thane of Argyll, the patriarch of the McDonalds, early in the 12th century acquired the Western Islands, by his marriage with Effrica, daughter of Olavus, King of Man; and assumed the designation of King of the Isles, which his successors held, independent of the Scottish Kings, for three generations; died in 1164
• 24. Alasdair Og MacDonnell: ancestor of all the MacDomhnaill Gallóglach families; died 1308
• 27. Charles Thurlough Mor McDonald: born in Antrim; acquired lands being known as the Clan Donnell Country, including Tynekill Castle at the base of the boundary of the mountains of Leix and Wicklow; died 1435
• 36. Lieutenant Brian McDonald, who served in an Irish Volunteer Regiment in the cause of King James II. Emigrated to America in 1684
• 38. Bryan McDonald III, born 1732 in Mill Creek Hundred, New Castle Co., Delaware; married Susannah Ogle in 1752 in Wilmington, Delaware; died 1777 in Buffalo Creek, Botetourt County, Virginia
• 43. Frank Everett McDonald, Jr.; born in Virginia in 1926; married Betty Jean Verna in 1951; FTDNA kit #133546

Generation 27. Tighearna Coille (Forest Manor), County Laois. Also known as Tinnakill or Tynekill, this tower house, built c.1450 by John Carrogh (Thirlough Oge) McDonald, served as the “Caput” (Seat) of the MacDonnells of Leinster for more than two hundred years. See The MacDonnells of Leinster.

Generation 38. Bryan McDonald House, Botetourt County, Virginia. Constructed in 1766, the house is associated with one of the early families to settle in Botetourt County. The McDonald House is the earliest known extant building in the 27-county western region of Virginia. A brick addition was added to the house around 1840. Today’s house sits on 5.9 acres, which is substantially less than its original acreage, however, the house still retains the feeling of its original setting within the landscape. Approved June 2011 for listing on the Virginia Landmarks Register. See Virginia Department of Historic Resources .

• Generations 1 to 20: The Ancestors of the McDonalds of Somerset (exerpt from Clan Donald DNA Project), by Donald M. Schlegel, 1998. Sources consulted by Schlegel include: Book of Ballymote (ca 1400), Book of Lecan (Early 1400s), Nat. Lib. Scot. MS 72.1.1 (1467), Monro (1549), Harleian MS 1425/190-191 (ca 1620), Annals of Clan Macnoise (1627), Geoffrey Keating (ca 1634), O'Clery
  (mid 1600s), and Book of Clanranald (ca 1715). Also see The Three Collas - An Alternative Explanation.
• Generations 20 to 36: The MacDonnell Family of Leinster, 1100 AD to 1707, Irish Midlands Ancestry, compiled by the Family History Centre, Bury Quay, Tullamore, County Offaly, Ireland
• Generation 36: Contributions to the Early History of Bryan McDonald and Family, 1879 , by Frank V. McDonald
• Generation 36 to 40: Kegley's Virginia Frontier, 1938 (or selected pages)
• Generations 25 to 43: Genealogy of Frank Everett McDonald, Jr. by Vaden McDonald, 2007.
• Generation 39 to 43: 1820-1930 U. S. Censuses:
  • Buffalo Creek, Botetourt County, Virginia: 1820 for generations 39 (widow Ruth) and 40
  • McDonalds Mill, Montgomery County, Virginia: 1830 for generations 39 (widow Ruth) and 40
  • McDonalds Mill, Montgomery County, Virginia: 1840-1, 1840-2, 1850, 1860, 1870, and 1880 for generations 40 and 41
  • Salem, Roanoke County, Virginia: 1880 for generation 41
  • Chamblissburg, Bedford County, Virginia: 1900, 1910, and 1920 for generations 41 and 42
  • Roanoke Creek, Big Lick District, Virginia: 1930 for generations 42 and 43

Other Sources. See also

• The MacDonnells of Leinster
• Clan Donald Heritage
• Clan Donald, by Donald J. McDonald of Castleton, 1978
• The Clan Donald, by Angus and Archibald MacDonald: Volume I, 1896; Volume II, 1900; Volume III, 1904
• McDonald Map

Patrick McMahon, a co-author of this Web page, has identified a Colla Subgroup of 8 Colla descendants who have similar DNA: Frank McDonald, three other McDonald, three McDaniel, and an Edwards. The subgroup has values of 14 for Markers 437 and 446. The modal values for Clan Colla are 15 and 13.

The genetic distances among the 8 people range from 1 to 7 and average 3.9.

Table of Contents

49 Generations: Colla to McMahon

Our study of the DNA of the Three Collas, who lived in the 4th century, is based on present-day Y-chromosome DNA and surnames from ancient pedigrees. It is not based on genealogies which show a paper trail from the 4th century to today. One of the McMahons in this study, however, can trace his ancestry back to one of the Three Collas, Colla da Crioch, who lived in the 4th century. He is Patrick Ciaran McMahon, a co-author of this Web page and a co-author with Eugene McMahon of The McMahons of Trohanny, a Private Publication. His FTDNA kit is #145687. The genetic distance of 67 markers of his Y-chromosome DNA from the Colla Modal DNA is 4. He participates in the Clan Colla 425 Null DNA Project. Patrick has an advanced degree in genetics from Trinity College Dublin and lives in Gorey, Ireland.

Patrick has 49 generations back to Colla da Chrioch, as shown on the right. Included among the 49 generations are the following:

• 1. Muredach Colla da Chrioch, brother of Carrell Colla Uais and Aedh Colla Menn
• 5. Cairbre an Dam Airge, d. 513
• 22. Faolan MacMathghamhna (Felim/Phelan MacMahon), early 12th century
• 49. Patrick Ciaran McMahon; born in Ireland; FTDNA kit #145687

• Generations 1 to 5: The Clougher Record, "Reweaving the Tapestry of Ancient Ulster," by Donald M. Schlegel, Volume XVII, No. 3, 2002, pages 689-749
• Generations 5 to 22: Irish Pedigrees; or, the Origin and Stem of the Irish Nation, Volume I, 1892, by John O'Hart: MacMahon (No. 1) for generations 5 to 22 (91 to 108 on pages 549-550)
• Generations 22 to 42 are derived from genealogical charts reproduced in The McMahons of Trohanny by Patrick and Eugene McMahon, Private Publication, February 2008, pages 155 & 157-158, augmented by recent confirmatory information from Katharine Simms (Senior Lecturer in Medieval History, Trinity College Dublin) in her article "The MacMahon Pedigree: a Medieval Forgery" in Regions and Rulers in Ireland, 1100-1650, edited by David Edwards, 2004, pages 27–36.
• Generations 42 to 49: The McMahons of Trohanny by Patrick and Eugene McMahon, Private Publication, February 2008.

In his McMahon DNA, Patrick McMahon analyzes the DNA of other McMahons who have tested their DNA, indicating that his ancestry may be helpful to other McMahons in understanding their ancestry.

Table of Contents     

Origin of the Three Collas

1. Cousins of Niall. The traditional explanation is that the Three Collas are descended from the kings of Ireland and are cousins, twice removed of Niall of the Nine Hostages. The Three Collas were the sons of Eochaid Dublein, who was the younger son of Cairbre Lificar, the 117th King of Ireland. The older son, Fiacha Sraibtine, was the 120th King of Ireland. The Three Collas waged war against their uncle Fiacha Sraibhtine and slew him in the battle of Dubhcomar, AD 322. Colla Uais then ascended the throne as the 121st King of Ireland. In AD 326, he was deposed by Muredach Tirech, son of Fiacha Sraibtine. Muredach Tirech then banished to Scotland the Three Collas and their principal chiefs, to the number of three hundred; but through the influence of the King of Alba, and the mediation of the Druids, they were afterwards pardoned by the Irish King, who cordially invited them to return to Ireland, and received them into great favor.      ...more
2. Romanized Trinovantes. In 1998, Donald M. Schlegel, suggested in his article "The Origin of the Three Collas and the Fall of Emain" in the Clogher Record that the Three Collas were Romanized Britons from the Trinovantes, a celtic tribe from Colchester, the oldest recorded Roman town in England. They received military training from the Romans and eventually went to Ireland as mercenaries in the service of the King of Ireland. The Three Collas were not descended from the kings of Ireland and were not related to Niall of the Nine Hostages. This is consistent with recent DNA findings.      ...more
3. Tribal Drift. In 2010, Patrick McMahon, a geneticist and a contributor to this study, cited a westward and north westward movement of bronze-age Celts through Europe, including the L21 line which started north of the Alps about 4,000 years ago. He estimates that L21 Celts started to populate Ireland about 3,000 years ago. By the time the Romans came to Britain, these Celtic people would have subsumed earlier cultures and diverged from one another over the 30 or so generations. There arose in one L21 tribe, Clan Colla, the relatively rare and stable null mutation at marker 425 shortly before the Roman invasion of Britain. In addition to names representing Wales, there is an equally strong contingent of Scottish names. This, in conjunction with the occurrence of the null mutation at the beginning of the first milennium, is strongly suggestive that the Colla tribe was well established and had branched in northwest Britain before coming to Ireland. The three Colla brothers arrived in Ireland about A.D. 300, allegedly as mercenaries to the High King. Based on his analysis of the DNA data, Patrick sees no confirmation that specific surnames descended from specific brothers. See his McMahon DNA.      ...more

Stone of Destiny, where Kings of Ireland were crowned in a fort on the Hill of Tara, County Meath.

Cousins of Niall. In 1892, genealogist John O'Hart (1824-1902) published a two-volume book entitled Irish Pedigrees; or, the Origin and Stem of the Irish Nation. Copies of these volumes at the University of Michigan were digitized by Google in June 2008: Volume I and Volume II. On page 575 of Volume II, O'Hart says the Three Collas invaded Ulster, conquered the country, and there formed for themselves and their posterity, the Kingdom of Orgiall (latinized Orgallia), sometimes called Oriel, and Uriel.

The Three Collas were the sons of Eochaid Dublein, who was the younger son of Cairbre Lificar, the 117th King of Ireland. The older son, Fiacha Sraibtine, was the 120th King of Ireland. The Three Collas waged war against their uncle Fiacha Sraibhtine and slew him in the battle of Dubhcomar, AD 322. Colla Uais then ascended the throne as the 121st King of Ireland. In AD 326, he was deposed by Muredach Tirech, son of Fiacha Sraibtine. Muredach Tirech then banished to Scotland the Three Collas and their principal chiefs, to the number of three hundred; but through the influence of the King of Alba, and the mediation of the Druids, they were afterwards pardoned by the Irish King, who cordially invited them to return to Ireland, and received them into great favor.

The Three Collas are a separate line from Niall of the Nine Hostages. Both are descended from Conn of the Hundred Battles and the Milesian Kings.

An Irish priest, Geoffrey Keating (1569-1644), had this to say in Book I, Section XLVII (pages 358 and 359) of The History of Ireland, published in 1629-31 and translated into English by David Comyn and Patrick S. Dinneen:

It is at Cairbre Lithfeachair that the Oirghialla—that is, the family of the Collas—separate in their pedigree from the clanna Neill and the Connachtaigh. And Fiachaidh Sraibhthine son of Cairbre Lithfeachair was grandfather of Eochaidh Muighmheadhon son of Muireadhach Tireach, son of Fiachaidh Sraibhthine; and it is from this Muireadhach that the clanna Neill and the men of Connaught are descended. Eochaidh Doimhlean son of Cairbre Lithfeachair was brother to Fiachaidh Sraibhthine; and this Eochaidh had three sons, to wit, the three Collas, and from these are descended the Ui Mac Uais, the Ui Criomhthainn, and the Modhornaigh. The real names of the three Collas referred to were Cairioll, Muireadhach, and Aodh.

In 1946, Thomas O'Rahilly took the position in his book Early Irish History and Mythology that the Three Collas did not exist. They were simply Eoghan, Conall, and Enda, the three sons of Niall of the Nine Hostages who conquered northwest Ireland.

For more information on the traditional origins of the Three Collas, see the Edward Cartin website and the Alec Conn website.

Table of Contents     

Romanized Trinovantes. Donald Schlegel has proposed an alternate explanation of the origin of the Three Collas. He starts by saying that the Collas are perhaps the only instance in prehistoric or early historic Ireland of three brothers having each a personal name, a name incommon, and an epithet. The implication is that such a naming convention must have been imported, and the obvious source is the Roman Empire. He suggests they were not descended from Irish Kings but instead were Romanized Britons, originating in the Celtic tribe named Trinovantes from Colchester, the oldest recorded Roman town in England. They received military training from the Romans and eventually went to Ireland as mercenaries in the service of the King of Ireland. Don presented this theory in the 1998 Clogher Record. It is one of the many articles he has had published in the Clogher Record, a local history journal published annually since 1953 by the Clogher Historical Society at St. Macartan's College in the townland of Mullaghmurphy on the outskirts of the town of Monaghan, County Monaghan.

See The Clogher Record, "The Origin of the Three Collas and the Fall of Emain," by Donald M. Schlegel, Volume XVI, No. 2, 1998, pp. 159-181. Also see The Clogher Record, "Reweaving the Tapestry of Ancient Ulster," by Donald M. Schlegel, Volume XVII, No. 3, 2002, pp. 689-749.

The first part of this alternative explanation is consistent with DNA results. Descendants of the Three Collas have a unique DNA which is significantly different from the DNA of descendants of Niall of the Nine Hostages. So, it seems pretty clear that the Three Collas were not cousins of Muredach Tirech, 122nd King of Ireland and grandfather of Niall of the Nine Hostages. See Modal DNA of Clan Colla Versus Niall of the Nine Hostages and L21 DNA.

The second part of this alternative explanation has not been verified yet by DNA. We have not found a family that matches Clan Colla DNA and traces itself back to the area around Colchester. There are, however, two families that match Clan Colla DNA and trace themselves back to towns that had Roman settlements in Wales on the west of England.

• The Calkins family matches Clan Colla and traces itself back to Chester, England, on the northern border with Wales near Deva, a town occupied by the Cornovii, a Celtic tribe, and where the Roman city of Viroconium Cornoviorum was from 70 AD to 410 AD.
• The Roderick family matches Clan Colla and traces itself back to southern Wales near Caerleon, a town occupied by the Silures, a Celtic tribe, and where the Roman city of Isca Silurum was from 75 AD to 410 AD.

As shown in the study above, there are a number of people, like the Calkins and Roderick families, with DNA similar to the descendants of the Three Collas who do not have the surnames of the descendants. These people could be descendants of cousins of the Three Collas who remained in Briton or went elsewhere, and evolved their own unique surnames. Some of these people may be descendants of the Three Collas who were adopted by a non-Collas or simply had their names changed over time. And, the historical lists of Colla descendants may well be incomplete.

Table of Contents     

Tribal Drift. The archeological and geneology evidence indicates a westward and north westward movement of bronze-age Celts through Europe. Among these would be the L21+ line which started north of the Alps about 4,000 years ago. Judging by todays distributions of L21+, the heaviest concentrations are found in Ireland and the Celtic fringes of Britain. It is estimated they could have started to populate Ireland about 3,000 years ago. By the time the Romans came to Britain, these Celtic people would have subsumed earlier cultures and diverged from one another over the 30 or so generations. They probably developed into tribal groups as they did in Ireland which the Romans, for their administrative convenience, gave them names such as the Trinovantes, Cornovii etc. Genetically, they would have similarities and differences as exhibited between Irish tribes. There arose in one such L21+ tribe, the relatively rare and stable null mutation at DYS 425, which our working hypothesis claims is the key identifier of Colla DNA. A trawl through the L21+ Project showed that only about 5% had this Colla DNA.

The DNA evidence would appear to indicate that the ancestors of the Colla brothers were part of a gradual westward migration through Britain to Ireland in either pre-Roman or Roman times. The current study would position the null mutation as having occurred shortly before the Roman invasion of Britain. Interestingly, a present-day Roderick family who have Clan Colla DNA and traces itself back to southern Wales near Caerleon, a town occupied by the Silures just north of the Bristol Channel, where the Roman city of Isca Silurum thrived from 75 A.D. to 410 A.D. Prior to that, the local Celtic people, loosely defined as Silures, mounted a fierce resistance c. 48 A.D. to the Roman conquest. It is therefore possible that the ancestors of the Colla brothers formed part of a minor Celtic exodus to Ireland following the collapse of that resistance in the first century A.D. A similar argument could be made for yet another family, the Calkins, who are also of Clan Colla and come from Cheshire/North Wales, being proto Colla and originating in another Roman Town, Chester. Equally, they could have come from any part of Britain, propelled westward by the Romans following failed uprisings such as Bodicea’s in East Anglia. This family might bear a derivative of the original tribal name, Colla Kinsmen or abbreviated to Colla-kin.

The most colourful explanation of Colla origins has been put forward by Don Schlegel in the Clogher Record. He states that the Collas are perhaps the only instance in prehistoric or early historical Ireland of three brothers having each a personal name, a name incommon, and an epithet. The implication is that such a naming convention must have been imported, and the obvious source is the Roman Empire. He suggests they were not descended from Irish Kings but instead were Romanized Britons, originating in the Celtic tribe named Trinovantes (who fought with Bodicea) from Colchester, the oldest recorded Roman town in England. They received military training from the Romans and eventually went to Ireland as mercenaries in the service of the King of Ireland.

In addition to the above well represented names (among testees) representing Wales, there is an equally strong contingent of Scottish names. This, in conjunction with the occurrence of the null mutation at the beginning of the first milennium, is strongly suggestive that the Colla tribe was well established and had branched in NW Britain before coming to Ireland.

It is well documented that the three Colla brothers arrived in Ireland about A.D. 300, allegedly as mercenaries to the High King. As such it has to be assumed they were trained soldiers of some sort and would have been accompanied by a band of warriors (otherwise they would never have established themselves in a hostile environment). A further assumption might be that the band of warriors was composed of their 425 null bearing kinsmen, kinsmen without the null and co-opted non-kinsmen.

Whatever the starting ratios were and how these fluctuated over time, we know that among todays descendants, only about 25-30% have the 425 null. It is not possible to track back from today’s ratios to establish the starting ones as too many survival factors would have been involved1. If as alluded to in this study, there was an additional earlier 425 null mutation, their descendants formed a much lower proportion of the Colla population. In addition, during the course of their warlike activities they would have enslaved/subsumed their defeated enemies who could have been part of an earlier indigenous population thus accounting for the few from haplogroups E & I with a Colla name.

Following their successful campaigns in Ulster, the Colla tribe would have continued to diverge both genetically (markers other than 425) and geographically but those with the null would continue to retain it. At a much later date (c. A.D. 950), surnames were gradually adopted by the various Colla branches and depending on where they were living and to which Colla chieftan they owed fielty to, they would have been granted that name irrespective of their DNA. Thus, those living in different parts of Oriel became Carrolls, McKennas, McMahons, McGuires etc. It is difficult to understand how these tribal branches (now clans) had more or less similar proportions of null to non-null.

Table of Contents