Project Conclusions






What do the numbers really mean? How can we be certain that the conclusions we draw are true and accurate?

Every human male carries a Y chromosome—this is what determines that a child eventually develops into a man rather than a woman. Scattered over the Y chromosome are individual genes and groups of genes. Some of these have no obvious function, and are loosely termed “junk DNA.” At certain places on the Y chromosome (we call them “loci”; one of them a “locus”), these junk DNA segments can repeat themselves over and over again, sometimes as many as twenty times or more. These can be counted physically by lab technicians. The counts of these gene repetitions at specific locations on the Y chromosome make up the series of numbers reported back to us by Family Tree DNA. These gene repetitions are themselves inherited virtually unchanged in the direct male line over dozens—even hundreds—of generations.


Why examine these particular places on the Y chromosome? Experience has taught us that some sites there are more stable than others, and change or mutate more or less frequently over time; they tend to provide a better picture of what’s happening than other locations might. They give us a stable benchmark by which we can measure the inheritance of otherwise meaningless characteristics through a direct-male descent. They can be used to demonstrate that an individual is related in the male line to someone else of the same surname.


FT DNA tests Y-chromosome inheritance at four levels of sophistication: 12 markers, 25 markers, 37 markers, and 67 markers. In the higher-level tests, more places on the Y chromosome are examined for junk DNA repetitions. If we test more places on the Y chromosome, we get better results. Most Burgesses immigrated to North America from the British Isles, and so some members of the Burgess Surname DNA Project share very similar Y-chromosome markers. In general, the closer the match, the more likely it is that two males with the same surname share a common genetic heritage. And, in fact, this is exactly what happens.


However, someone can have a random match at 12 or even 25 markers that really isn’t a match at all. So how do you tell if a Burgess man who matches another Burgess male in 11 out of 12 markers or 23 out of 25 markers is actually related to him? The answer is: you can’t, not for certain, without enhancing the record to 37 or 67 markers—or unless you have some other evidence to weigh in the balance, such as knowing from standard genealogical research that the match is predicted. A 37-marker test refines the results by giving us another 12 points of comparison; the 67-marker test adds another 30. To look at it another way, testing for 25 markers as opposed to 12 markers reduces the average elapsed time until you reach the Most Common Recent Ancestor (MCRA) by half, from about 600 years to about 300 years. The 37-marker test brings the MCRA down to about 150 years, while 67 markers reduce the elapsed time to about 100 years.


We lump males with similar Y-chromosome numbers into “haplogroups.” This is just a convenient way to organize male descents. Most of us (about 85%) have the haplotype “R1b1” and its subdivisions, which were among the most recent of the groups to develop, and reflect the underlying populations of France, Spain, Southern Germany, and the British Isles. These aren’t the only groups deriving from Britain, of course—haplogroup “I” and its subdivisions and “E” are also fairly common—but R1b1 and its derivatives are by far the majority.


Haplotypes are actually determined, however, not by the Y-chromosome numbers, but by Single Nucleotide Polymorphisms (SNPs) that are unique to specific haplogroups. These have to be tested for separately. A simple test, the so-called Backbone test, looks for the overall group to which an individual belongs (I, R1b, E1b, etc.). Deep clade tests, on the other hand, look for further subdivisions of the major haplogroups—these involve testing for multiple SNPs, each reflecting a further subdivision of the overall group. Thus, we can go from Haplotype R to R1 to R1b to R1b1 to R1b1b2 to R1b1b2h, as one example. Only one such Deep Clade test needs to be performed for each Burgess family group; all other members of that group with similar Y-chromosome numbers must share exactly the same haplotype.


When I started the Burgess Surname DNA Project in August of 2004, I expected that a handful of the known major Burgess lines would dominate the results, and this is exactly what’s happened—not only with our project, but with all of the others sponsored by FT DNA. We have a few Burgess families that seem to pop up constantly, just because there so many of them represented in the general population. These groups are very old and well established. They’re the ones who had ten sons back in the 1700s. But I didn’t expect to find so many unique Burgess families whose markers don’t match anyone in the Project—and sometimes no one else in the entire FT DNA database. How can this happen?


Surnames have been constantly recreated throughout history, sometimes being adopted by immigrants from other countries or by individuals with similar names. Natural births and hidden adoptions also occur in every family of any antiquity; these have the effect of creating new genetic signatures for all subsequent descendants of that particular male. Such families tend to be smaller than the older lines (because they have more recent origins), with relatively few present-day descendants.


So in every major DNA surname project, we see three kinds of results: several large family clusters, a number of middle-sized groups, and a seemingly endless array of much smaller families that stand on their own—at least until they match someone else of the same name.


As the Project has grown, each of its major lines has also developed appendages; these include branches that show a break in the chain of descent somewhere in the past, and families that are related genetically to the original ancestor of that group, but can’t as yet be tied to it through conventional genealogical research. We’ve also seen a rise in the number of “ambiguous” results, where the number sets by themselves cannot prove or disprove a link between otherwise unconnected lines.


So what can we actually demonstrate? What good is the Project if we can’t always seem to find the links we want?


We’ve done remarkably well. The basic genealogical signatures of most the major Burgess lines in North America have now been established. The Burgess families of Great Britain and its former colonies are less well delineated; we need more samples from British Burgesses to reach the critical point where a majority of the major family groups there can be identified.


The Project has also disproved a number of supposed connections between the major Burgess family groups. Thus, Col. William Burgess’s numbers, while confirming the basic genetic signature of his family, have so far matched those of no one but his known descendants. The top half dozen Burgess family groups still stand alone, with only the large families of William Burgess of King George Co., VA, and William Burgess of Bedford Co., VA, amalgamating into one overall family group. However, we have been unable so far to identify the British origins of the major North American Burgess lines. We simply do not yet have enough samples from the British Isles to find the cognates that are certain to exist there.


The Project has also been able to sort out the genetic origins of several Burgesses families living in adjoining geographical areas. Thus, the two Burgess lines that settled in early Scott Co., KY, have now been placed with their respective family groups—and this is but one example of many. We will see more such matches, identifications, and separations as the Project grows.


Now let’s examine some of our findings:


William Burgess of Virginia


This is a huge family, currently the largest in the Project. The first linked group is that of William Burgess of Richmond (later King George) Co., Virginia, for which we have test results from all five branches.


Their numbers also exactly match those of the male descendants of another major Burgess group, the line of William Burgess of Bedford Co., Virginia, for which we have test results from all four branches. Together, these two family groups have a common Burgess ancestor, name unknown, who predates the known research in both lines, putting the Most Common Recent Ancestor back at least to 1650. Curiously, the genome of the Bedford family seems much more stable than that of the King George line.


We also have a third connected family descending from William Burgess of Albemarle Co., Virginia. He cannot be any of the William Burgesses who are grandsons of William Burgess of Bedford Co., but he may be descended from one of several undelineated families in the King George Co. line.


We also have a connection with the line of Joseph A. Davidson, whose direct-male descendant matches 36 out of 37 markers with one of the individuals in the Bedford Co. family. Davidson married a Burgess granddaughter out of the Bedford Co. line in the 1820s, but since Y-chromosome markers can’t be inherited through a female line, the only possible conclusion is that Davidson was himself a natural son of one of the male Burgesses of this family, and that he married a first cousin.


We have a good genetic flag in the three combined family groups: the markers 11, 13 at Loci 385a-385b, together with the value of 14 or 15 at Locus 392, is a scarce combination occurring in only 3.04% of all of the families in the Family Tree DNA database.


This is also the hallmark of the descendants of the Irish Chief Niall of the Nine Hostages (d. about 450 A.D.), and these families are very likely offshoots of Niall’s immense line, which accounts for 21.5% of all present-day males in Northern Ireland.


Thomas Burgess of Massachusetts


The Thomas Burgess line of Sandwich, Barnstable Co., Massachusetts is the oldest and perhaps the largest of the Burgess families in North America, and the second largest in the Burgess DNA Project, having been established here in the early 1600s. It utterly dominates the Burgess lines of New England and northeastern Canada. We have numerous test results from three of the four branches of this family and its possible affiliates, and they show that Thomas is the common ancestor.  These include descendants of Albert Burgess Sr. of Tolland Co., Connecticut, Benjamin Burgess of Saint John Co., New Brunswick, Canada, Josiah Burgess of Monroe Co., New York, and Thomas Burgess of York Co., Ontario, Canada.


The Thomas line may also be distantly connected to the family of John Burgess of Cheshire Co., England, but the numbers are inconclusive, and the relationship, if it exists, dates back more than 500-600 years.


This family has a relatively uncommon repetition of the markers 14, 13, 30 at Loci 389/1, 392, and 389/2, a scarce combination that occurs in only 14% of the families in the Family Tree DNA database. This can be used as a DNA “flag” for these lines. This also make a relationship between the John Burgess line and the Thomas Burgess family somewhat more likely than not.


Edward and William Burgess


The third largest line in the Project consists of the combined families of Edward Burgess of Prince George’s Co., Maryland and Pittsylvania Co., Virginia, and William Burgess of Montgomery Co., Maryland, who may well be brothers. These lines were long thought to be connected—both to each other and to the line of Col. William Burgess of Anne Arundel Co., Maryland. One of William of Montgomery’s sons married one of Edward’s daughters, and both families are associated with other up through the mid-1800s, suggesting that the two men were brothers. Multiple test results received from both lines have indeed matched each other, confirming that they have a common Burgess ancestor, name unknown—but they fail to match any descendant of Col. William’s line.


This family also has numerous affiliated lines that have not yet been attached either to Edward or William, all of which seem ultimately to derive from Maryland. The most interesting of these, the family of Edward Burgess of Fayette Co., Pennsylvania, cannot chronologically be fitted under either of the above families, and so has to be parallel to it. It also displays the most genetic diversity of any of these connected families. The other connected families include descendants of Austin “Auzy” Burgess of Laclede Co., Missouri, James Burgess of Jackson Co., Tennessee, and Samuel H. Burgess of Calhoun Co., Florida.


Again, we have a reliable genetic flag to use for these folks: the repetition of the markers 11, 16 at Loci 385a- 385b, together with the value of 10 at Locus 391, is a rare combination, occurring in just .7% of all of the families in the Family Tree DNA database.


Thomas Burgess of  Pittsylvania Co., Virginia


Thomas Burgess Sr. of Orange Co., Virginia, and Pittsylvania Co., Virginia, was the founder of the fourth largest family in the Burgess DNA Project, with a very large and well-established Burgess line. His family structure has been heavily reworked in this update. However, there remains one unattached but affiliated line—the family of John Burgess of Union Co., South Carolina. They have a common Burgess ancestor, name uncertain.


The branches of the Thomas line share a very common 12-marker genetic signature called the Western Atlantic Modal Haplotype (WAMH), although the 37-marker results are relatively scarce. They’re probably descended from a pre-Christian era Celtic chieftain.


Ralph Burgess of Cheshire Co., England


Ralph Burges of Cheshire Co., England, who was born about 1555, is the progenitor of the fifth largest family in the Project. His descendant has a close match with a descendant of John Burgess (I) of Cheshire Co., England. This very large Burgess line, which is probably one of the major British Burgess families, also matches with several descendants of William Burgess of Middlesex Co., Massachusetts. The American branch of this family is more distantly related to its parent, judging by the increased variability in its Y-chromosome numbers.


William Burgess of Anne Arundel Co., Maryland


Col. William Burgess of Anne Arundel Co., Maryland, was the most prominent individual named Burgess in colonial America, and by far the wealthiest. It’s not surprising, therefore, that many genealogists want to tie their own Burgess families into his line. This hasn’t happened, alas. We have several test results out of this family, deriving from both Col. William’s eldest son, Capt. Edward Burgess, and also from his youngest son, Dr. Charles Burgess, and they all match each other—but no one else.


Here again we have a useful genetic flag: the repetition of the markers 20 and 24 at Loci 458 and 447 are scarce values for these locations.


Two Burgar Families of Northern Scotland


An example of how a name can evolve to become “Burgess” is shown by the Burgess line of the Shetland Islands north of Scotland. The common progenitor here was surnamed “Burgar,” and there are still branches of this family retaining that name. But a number of the lines deriving from this group are now called “Burgess.” Test results from several individuals in this line all match each other—but no other Burgess family.


A second quartet of Burgars and Burgesses from the Orkney Islands, a neighboring archipelago, match either other but not the first group. The difference is that the first group demonstrates a Celtic heritage (Haplogroup R1b1), while the Orkney families have a possible Viking background (Haplogroup I).


Joel Burgess of Laurens Co., South Carolina


The importance of finding participants with known descents from the major early Burgess lines was demonstrated when a proven offshoot of the Joel Burgess family of Laurens Co., South Carolina, joined the Project. His numbers immediately matched those of three other individuals, descendants of Thomas Burgess Sr. of Scott Co., Kentucky, John R. Burgess of Clay Co., Alabama, and Alphonso Davis of Wayne Co., Illinois. Since the Joel Burgess family can be traced several generations further back in time, this is likely the paramount line. These individuals likely have a common Burgess ancestor whose name is unknown. The numbers suggest that the John R. Burgess family is an offshoot of the Joel Burgess line, while the Thomas Burgess and Alphonso Davis lines seem more closely paired to one another than to Joel.


Once again, we have a useful genetic signature for these families with the value of “14” at Locus 388, a rare number at that location for members of Haplogroup R1b1.


James Burgess of Sussex Co., England


Several descendants of James Burgess of Sussex Co., England, John Burgess of New Haven Co., Connecticut, and Joseph Burgess of Bergen Co., New Jersey, have matched; they have a common male ancestor, name unknown. The numbers indicate that the two American lines are more closely related to each other than they are to their British cousins.


Adam and Joseph Drury Burgess


Several descendants of Adam Burgess of Dumfriesshire Co., Scotland, and Joseph Burgess (also called Drury Burgess) of Sumter Co., South Carolina (originally from County Antrim, Ireland) have matched, indicating that they have a male common ancestor, name unknown. The Scottish line, which has deeper roots, is more likely to be the paramount family of this group.


Francis and Richard Burgess


Several descendants of Francis Burgess of Berkshire Co., England, and Richard Burgess of Hertfordshire Co., England, have matched, indicating that they have a common male ancestor, name unknown. This was the first random match in the Project between two British Burgess families.


John and David Burgess


John Burgess of Rutherford Co., North Carolina and David Burgess of Hardeman Co., Tennessee lived at about the same time in the early 1800s, but had no known connection to each other beyond geographic proximity, residing in counties on either side of the border between North and South Carolina. Test results from these families indicate that these two men have a common male ancestor, name unknown.


The genetic flag here is the marker 15 at Locus 393, a very scarce value for this location, occurring in only 2.54% of the families in the Family Tree DNA database.


Peter Burgess of Cheshire Co., England


Peter Burgess of Cheshire Co., England, has known male descendants in both England and Canada. Test results have confirmed a relationship between two branches of this line, but have failed to match any other Burgess line. Recent information suggests that the progenitor of this family was a Hughes.


Again, we have a useful genetic flag here with the repetition the markers 11, 15 at Loci 385a-385b, together with the value of 10 at Locus 391, which is a very scarce combination occurring in only 2.72% of the families in the Family Tree DNA database.


Emanuel B. Burgess of South Carolina


Test results received from two individuals descended from Emanuel B. Burgess of Union and Spartanburg Cos., South Carolina, have matched each other, proving the connection between two half-brothers in this line. However, they have yet to connect to any other Burgess family.


Keziah Burgess of Virginia & the Hale Family


Keziah Burgess of Bedford Co., Virginia, has long been a puzzle for researchers. DNA testing has now established that all three of her sons had different fathers. However, the numbers from the descendant of her first son, Armstead Lewis Burgess Sr., match those of numerous descendants of George Hale of Northumberland Co., Virginia, known to be ancestor of the Bedford Co. Hale family, providing a likely progenitor for this line. In addition, we have matching records for a Haas family and a Chilcote family, indicating that these lines are probably also related to the Hales. The overall number set is very scarce, suggesting that the proposed connection is real.


Achilles Burgess of North Carolina


Test results from several descendants of Achilles Burgess of Wake Co., North Carolina match each other; one of these lines, that of Thomas Burgess of Elbert Co., Georgia, was previously unattributed to this family.  This is a very rare Haplogroup, with only about twenty examples in the entire FT DNA database.


John Burgess of Henry Co., Virginia


John Burgess of Henry Co., Virginia represents an old Southern family dating back almost three hundred years. His numbers have matched with descendants of John W. Jamerson, previously known to be an offshoot of this line. The numbers have also matched the markers of a descendant of Green Burgess of Williamson Co., Tennessee, who was born in North Carolina. Green Burgess and two Johns share a common Burgess ancestor, name unknown.


John Burgess of Humphreys Co., Tennessee


The John Burgess Sr. line of Humphreys Co., Tennessee has matches from the descendants of two of his sons, Alexander Burgess and John Burgess Jr., thus confirming the genetic signature of this family. However, the Auzy Burgess line of nearby Dickson Co. has now been linked to the Edward and William Burgess group listed above.


William Burgess of Franklin Co., Alabama


William Burgess of Franklin Co., Alabama, fathered six or seven sons, including one, Joshua Burgess, whose mother’s name is unknown, but who is mentioned in the accounting of William’s estate. Two branches of this family, including that of Joshua, have now matched, thus establishing a genetic link between these lines (some researchers had thought that they were unconnected).


William Burgess of Lancashire Co., England


Two descendants of William Burgess of Lancashire Co., England, have matched each other, thereby confirming the numbers in this family.


James Burgess of Fayette Co., West Virginia


James Burgess of Fayette Co., (West) Virginia originated in Bedford Co., Virginia, and was long thought to belong to William Burgess of Bedford family. If he does, however, the connection can only be through a female line. We now have numbers from descendants of two branches of this family, and both match each other—but no one else.


Jonathan Burgess of Jackson Co., Alabama


Jonathan Burgess of Jackson Co., Alabama moved there successively from South Carolina and Georgia. Two of the descendants of his son, Starling G. Burgess, have matched, thereby validating the numbers in his line.


William Burgess of Marion Co., Tennessee


Descendants from two grandsons of William Burgess of Marion Co., Tennessee have now matched, providing the genetic evidence to link these lines. This family has a relatively scarce set of numbers deriving from Haplotype J2b.


William Burgess of Wayne Co., Indiana


Several descendants of William Burgess of Wayne Co., Indiana, have now matched, thereby affirming the base DNA numbers for this family.


Jacques Bourgeois of Canada


A number of descendants of Jacques Bourgeois of France and Nova Scotia, Canada have matched, thereby affirming the numbers for this family.


Burgess-Gilbert Connection


Two test results from the family of Timothy B. Gilbert of Norfolk Co., Virginia, one of whose sons changed his name to Burgess, have failed to match each other, and we don’t know as yet which sample represents the true genetic heritage of this family. The descendant of John Timothy Burgess has a very rare value (10) at Locus 19, occurring in only one out of 20,000 samples.


Burgess-Mellard Connection


The numbers of John Burgess of Hart Co., Georgia have matched those of a descendant of William Mellard of Dorchester Co., South Carolina. Since the genetic signature of these families is very unusual, these two seemingly unrelated lines probably have a common male ancestor, name unknown. Both lines likely derive from Colonial South Carolina.


Burgess-McGee Connection


Descendants from Richard Edward Burgess of Marshall Co., Alabama, and William McGee of Los Angeles Co., California, have matched at 66 out of 67 markers; they have a common male ancestor, name unknown. The numbers also suggest an affiliation with a much larger Hancock family (see the Hancock DNA Project at the website for Family Tree DNA).


Burgess-McManus Connection


A Burgess whose name and ancestry are unknown (but who may be descended from Jason Burgess of Fayette Co., Kentucky) has exactly matched at 37 markers with a descendant of the huge McManus clan of Ireland, strongly suggesting that the McManus line is the parent family of this Burgess group.


Unattached Burgess Lines


Roughly one-quarter of the samples in the Burgess Surname DNA Project display unique haplotypes, without a corresponding second sample from another cousin which might validate their base numbers or connect to other families. An “unknown __ line” means that the present-day descendant of this family, the individual whose DNA was tested, is not a member of the Burgess DNA Project and his ancestry is unknown:


Thomas Barger of Berks Co., Pennsylvania

Eberhard Berges of Westphalen, Germany

Unknown Bergess line

Unknown Bergis line

Johann Adam Borges of Hessen-Cassel, Germany

Unknown Borges line of the Açores, Portugal

Unknown Borges line

Ramón Bórquez of Sonora, México

William Burge of Somerset Co., England

Benjamin Burgess of Monroe Co., Michigan

Benjamin Burgess of Rock Co., Wisconsin

Charley Burgess of Screven Co., Georgia

Ephraim Burgess of Halifax Co., North Carolina

George Burgess of Allegheny Co., Pennsylvania

George Burgess of Cheshire Co., England

George Burgess of Northampton Co., North Carolina

George Burgess of Worcestershire Co., England (results pending)

Isaac Burgess of Cheshire Co., England

James Burgess of Gloucestershire Co., England

John Burgess of Aberdeenshire Co., Scotland

John Burgess (II) of Cheshire Co., England

John Burgess of Delaware Co., Indiana

John Burgess of Hertfordshire Co., England

John Burgess of Middlesex Co., England

Joseph William Burgess Sr. of London, England

Josiah Burgess of Telfair Co., Georgia

Richard Burgess of Tasmania, Australia

Sebray Burgess of Rockingham Co., Virginia

Trinity Burgess of Union Co., South Carolina

William Burgess of Columbiana Co., Ohio

William Burgess (I) of Cornwall Co., England                 

William Burgess (II) of Cornwall Co., England

William Moses Burgess of Adair Co., Kentucky

Adolfo Burgos of Barros, Puerto Rico

John Conrad Burgy of Lucas Co., Ohio, of Swiss origin

Unknown Burquez line of México

John Burris of Highland Co., Ohio

Andrew Kapel of Spokane Co., Washington

Simão Mendes Borges of Açores, Portugal


Some of these lines will eventually find genetic partners as more results are added to the Burgess Surname DNA Project.




The following individuals contributed their DNA to the Project, but have since passed. We thank them for their willingness to share, and honor their memories:


Daniel Keith Burgess (1956-2008)

(and his website,

Donald Leroy Burgess (1927-2007)

Eugene Carl Burgess (1912-2005)

Frank Joseph Kapel (1913-2006)

James Alan Burgess (1930-2005)

James Harold Burgess (1932-2006)

Lorene Elizabeth Burgess (1937-2011)

Orville Eugene Burgess (1922-2006)

Samuel Maurice Burgess (1925-2005)

William Lee Burgess (1933-2008)

John Nathan Mellard (1918-2010) 



The following individuals have contributed either their time or money or expertise to the Burgess Surname DNA Project, far above and beyond their original participation. Thanks to all of these good folks:


Sherrie Boone; Bill Burgar; Anthony & Lori Burgess; Charles O. Burgess; Clinton Burgess; Coy V. Burgess (thrice); Daniel K. Burgess (twice); David A. Burgess; Dean Burgess; Debbie Burgess; Eric S. Burgess in honor of the Cheshire Group (thrice); G. Ralph Burgess; Dr. J. Hugh Burgess (twice); James E. Burgess; Jim A. Burgess; John C. Burgess; Joseph L. Burgess; Keith E. Burgess; the late Lorene E. Burgess; Milo W. Burgess Jr. in memory of Milo W. Burgess, Sr.; G. Ralph Burgess; Dr. Richard R. Burgess; Dr. Russell E. Burgess; Dr. Scott A. Burgess; Seth T. Burgess; William L. Burgess; Venita Burgess Colley; Dr. David Faux; Richard Gardner; Shelia Gibson; Bennett Greenspan and the other good folks at Family Tree DNA; the late Mildred Burgess Guernsey (thrice); Elsie Burgess McVey; Susan Mortensen; Marion Nichols; Nancy Ozolins; Carolyn & Mark Shilts; Eleanor Gray White.




To contribute funds to the Burgess DNA Project, please go to www.familytree.con/contribution.html, and follow the instructions. Be sure to put the name of our project in the appropriate slot. Every dime contributed to the fund is used to support testing and test enhancements for those who otherwise would be unable to participate.



Additional information