Updated January 2013:
While the FTDNA test results posted here are up-to-date, results from other test labs are available. Google “Growing with America Fox” and click on “Growing with America FoxDNA” for further test data and interpretation.
Initial efforts to trace the origins of the Fox surname via DNA testing of the Y-chromosome began in Britain at the University of Leicester in 2002 as part of a general survey of British surnames by Mark Jobling and Turi King. Only two Fox men were tested. Further testing of other Foxes was begun in the United States in 2004, when the Fox Surname Project was established with Familytreedna (FTDNA) as the testing laboratory. Both Y-DNA (male line) and mtDNA (female line) and Family Finder (all ancestors but limited range) tests are offered but the emphasis is on Y-DNA testing of the Fox surname. The Y-Chromosome is ideal for tracking the Fox line since only males have this chromosome. The only problem is that only males in the direct male line (i.e. with the Fox surname unless there has been an adoption) can be tested. The other tests are open to males and females but do not test the Fox surname directly.
Technical Background – Y-DNA testing
DNA in the Y-Chromosome has many microsatellites where the nucleotides (adenine, guanine, cytocine and thymine) form repeat patterns, eg GATAGATAGATAGATAGATA. These markers are known as "short tandem repeats" or STRs. They are hereditary and the number of repeats is normally passed on from father to son without change. Cheek scraping samples are submitted to the laboratory where specific primers are used to isolate sections of DNA containing these microsatellites. These are then amplified with polymerase enzyme and subjected to separation by electrophoresis. The number of repeats at a given marker shows up as a peak height on the resulting electropherogram. The resulting set of numbers - number of repeats at each marker tested - is called one’s "haplotype." This is the result one sees when the laboratory submits their report. You will then be given access to your results and their interpretation on your personal web page at FTDNA.
A change in length of a marker is a random event that occurs primarily due to recombination effects as the chromosomes rearrange during the conception of a child (the process of meiosis.) This is termed a transmission event. Sons are separated by 2 transmission events, first cousins by 4. Thus, divide transmission events by 2 to obtain generations back to the most recent common ancestor (MRCA).
Changes in the number of repeats occur – on the average for FTDNA’s 37 marker test - once in every 300 transmission events, which makes these markers valuable for genealogical purposes. FTDNA has offered tests of 12, 25, 37, 67 markers and now offers a 111-marker test. According to MacDonald’s MRCA calculator
, (using Walsh's infinite alleles formula) the average mutation rates for these tests are as follows:
12 markers - 0.0020 (one in every 500 transmission events)
25 markers - 0.0023 (one in every 435 transmission events)
37 markers - 0.0033 (one in every 303 transmission events)
67 markers - 0.0028 (one in every 357 transmission events)
Preliminary information on the 111 marker test indicates a rate similar to the 67 marker test.
In contrast the 40-some markers used by Ancestry.com have a mutation rate of 0.0021
Choice of Test
Generally speaking, the 37-marker Y-DNA test – with its high mutation rate - is the best choice for a new recruit with limited resources. With 37 markers tested, a change might be expected - on the average – once in 303/37=8.2 transmission events or 4.1 generations. Very roughly, if two men having the Fox surname have 33 out of 37 matching markers they can be considered related within a genealogical time frame of 20 generations or 600 years. (FTDNA’s TiP analysis would estimate more like 400 years.)
A 67-marker or 111-marker test will provide further confirmation in cases where a relationship is questionable and may well help distinguish between different branches of the same family. They also offer additional information on stable markers useful for deep ancestry studies of Haplogroup subclades or single founder clusters. If affordable, these tests are certainly recommended. 12 or 25 markers can sometimes conclusively disprove a relationship but cannot reasonably confirm one unless there is a firm paper trail. Additional markers are available as special tests.
Often a close match may be found with an individual having a different surname. These matches should be treated with great caution, looking for possible cases of adoption or name change, examining the geographical proximity of the ancestors of both lines and, perhaps, deciding on additional testing. It is possible to have close matches across surnames that go back before the adoption of surnames and, if the two are in different Haplogroup subclades, this is a foregone certainty.
Genetic Distance (GD) is often used as a measure of deviation between two haplotypes. It is the sum of absolute numerical differences between haplotypes for all markers tested. The number of markers tested should always be specified along with the GD. FTDNA can search their entire tested database for matches by GD and number of markers tested but results are only shown where the GD is indicative of a possible genealogical relationship. (You must agree to this search in your personal preferences.)
Since two-step changes can occur in a single transmission step some 5% of the time, simply counting the number of matching markers is also useful and is used in the infinite alleles formula for predicting the number of generations to the most recent common ancestor (MRCA). Note that such predictions must be hedged in probabilities and that multi-valued markers require special care in interpreting deviations.
A word of caution about FTDNA’s TiP prediction of generations back to the MRCA: While this does attempt to account for measured father/son mutation rates in individual markers, the effect of deviations at different markers is really not that significant. The average mutation rate for all markers being compared is the critical factor and FTDNA’s mutation rates are thought by most experts to be considerably on the high side, thus predicting too low a generation difference. Keeping this in mind, a TiP analysis can be a useful tool.
The exact function of these STR markers is not yet known and they have no known medical function but recent research shows they have some sort of regulatory function on the genes. While there is no medical information in these numbers, the absence of a certain few markers near a fertility gene could indicate sterility - something that would certainly already be known.
The results do provide a partial means of personal identification and, for this reason, our haplotype tables list only the FTDNA kit number and the most distant known male line ancestor. Within the project, however, the administrators feel free to disclose identities, particularly when a close match occurs.
SNP Testing for the Haplogroup Subclade
Another type of mutation can also occur, but much less frequently – so much less that they are known (mistakenly perhaps) as unique events. This is a change from one nucleotide to another at a specific site on the Y-Chromosome, say a change from C to T. This is called a "single nucleotide polymorphism" or SNP. The chances of this occurring are once in 10,000 to 15,000 transmission events at a specific site but there are so many places on the Y-Chromosome where this can occur that sons are predictably different than their father. These are private SNPs unless they happen in a founder situation, where they are passed down for many generations and become public SNPs.
Certain specific SNP events have been identified over the course of history that have proven useful in tracing the migrations of mankind over time. These public SNPs identify ones "Haplogroup” and the subclade of that Haplogroup to which one belongs. They fall into a sequential tree-like structure with branches and twigs. Often one’s Haplogroup and primary subclades can be predicted from one’s STR haplotype but, to be absolutely certain, SNP testing needs to be done. The Haplotree section of ones’ personal web page provides further information and an opportunity to purchase deep clade SNP testing once one’s primary Haplogroup is ascertained. FTDNA will either make such an estimate or provide a backbone Haplogroup test when they can’t make an estimate. Quite a few Fox Project members have done deep clade testing at FTDNA or have results from other laboratories such as Ethnoancestry.
New SNPs identifying important subclades of the main Haplogroups are continuously being identified. Some of these newer subclades are as recent as 2,000 years old. This is already of some help for genealogical work since two very closely matching haplotypes in different subclades cannot be related within a time frame less than the age of the subclade. This would be an extremely rare occurrence but could happen due to back mutations.
With the advent of genome testing by 23and Me and DecodeMe as well as the FTY Project at FTDNA, additional information on Y-Chromosome SNPs has become available and amateurs and professionals have been mining this information for new SNPs of importance. As a result, the main Haplogroups (R and I, in particular), have been split into smaller and smaller subclades and Haplogroup designations are getting longer and longer. FTDNA has recently adopted a new revision of their Haplogroup Tree and this is shown on one’s Haplotree page. This is not entirely in agreement with the 2011 ISOGG Tree
, which generally includes more SNPs.
There is a general tendency nowadays to use the limiting SNP as the Haplogroup designator and this, for the most part, has been adopted by the Fox Project. Haplogroup R-S26/L1, for example, would be the subclade of Haplogroup R which tests positive (or derived) for the SNP called S26, which is identical to the SNP called L1 by FTDNA. In contrast, the long name for this subclade is currently R1b1a2a1a1a3 in both 2011 ISOGG and FTDNA terminology. The two are not always in this good agreement.
In the Y-Results Tables for the Fox Project, the results are grouped by Haplogroup Designation and – within Haplogroups – by Ancestral Family Groupings. All members of a Family Grouping will lie in the same Haplogroup subclade. Note that the tables show the Haplogroup in green only when SNP tested by FTDNA. There is also a separate table showing the results of SNP testing at FTDNA.
Fox Families Currently in the Project
While not always possible, the optimum way to use the Fox Project is by testing a hypothesis based on genealogical research. If two Foxes claim the same Fox ancestor, their Y-DNA haplotypes should match within probable limits. The first thing any interested Fox researcher should do is to look at the Y-Results tables, which are organized by Haplogroup and, where applicable, by known family groups. Wherever this has been supplied, the most distant known male line ancestor (MDKA), his location and date of birth are shown. Outside of the known family groupings, results are ordered by genetic distance to emphasize deep relationships.
Next, go to Ysearch.org and search for the surname Fox. The administrators endeavor to get every member into the Ysearch database, where their test results, most distant ancestor and, often, their pedigrees (prior to 1905) are made available to researchers. If you can find a match to your own genealogy research, you are a good prospect for the Fox project. Bear in mind, however, that many family trees available on the internet today have errors – errors that are being uncovered by DNA testing.
As the project builds in numbers, an increasing number of family connections are being made and the Y-Results tables are divided into family groupings in alphabetical order by Haplogroup and then by the known ancestors. The following is a summary of these groupings. For more detail see the Results Section.
In the Y-Results Tables
Three members have this estimated Haplogroup - Mostly German Origin. Only one has extended to 37 markers.
Two members tested Haplogroup G2, one is estimated G2c plus another from the SMGF database estimated G2. Two have upgraded to 67 markers and have joined special projects. One member has been tested G-M201 - German and Ukrainian origin.
Tested Haplogroup I-P19, estimated I-P37.2 from 12-marker result – one member German ancestry.
Haplogroup I-L39 Richard Fox of Virginia Descendants
Two men of Fox ancestry who match on 37 out of 37 markers. One from Virginia and one from Mexico via New Orleans. Latter has tested at 67 markers and has also tested Haplogroup I-L39 (or I2b2).
Haplogroup I-L533 Johann Fuchs of Germany Descendants
Two first cousins one of whom (114418) has been tested Haplogroup I-L533, a new subclade of I-L38/L39 or I2b2. FTDNA has him listed as Haplogroup Q-L272 or Q1b1a, which is in error.
Haplogroup I-M223 Probably From Lancashire England
Three Foxes who match closely enough to be related genealogically. Two from Lancashire and one from NC. Possibly originally from Dublin, Ireland, and of Viking origin.
Haplogroup I-M223 Christian Fuchs Descendants
Two known Fox cousins of German origin via Berks County, PA
Haplogroup I-M223 Other
Four unmatched Foxes of German and Irish origin.
Haplogroup I-M253 Elijah Fox of NC and TN Descendants
Four descendants of Elijah Fox’s three sons: Ransom, Absolom and John plus matching descendants of Allen Fox of NC and James Fox of Virginia.
Haplogroup I-M253 Jacob Fuchs of Germany Descendants
Five descendants of Jacob Fuchs who came to Bucks County, PA, in 1739
Haplogroup I-M253 Other
Four Foxes of German and British descent who have yet to find matches within the project – plus a Burris and a Roy admitted to the project because of possible Fox connections.
Haplogroup I-P109 Thomas Fox of Concord, MA, Group
One of our most prolific groups, consisting of six matching descendants of Thomas Fox of Concord, MA, born in England in 1618 plus four matching Foxes tested by FTDNA. We also have three matching results from the SMGF database who trace back to Thomas Fox and another matching Fox result who was tested by Relative Genetics. Two have been tested Haplogroup I-P109, a subclade of I-M253 that may indicate Scandinavian origin, and Thomas Fox's baptismal record has been tied to St. Olave's Church in London on Silver Street. Several men named Calkins/Caulkins also appear to be related. Apparent matches with other surnames are common in Haplogroup I-M253 but these are close enough to appear real.
Two non-matching results of Polish and Russian ancestry
Four non-matching Foxes one of whom (97877) has been tested Haplogroup J-M67*, or J2a4b*, and a second (110488) tested J-M172 (the backbone test for J2). The other two are of German ancestry, estimated as Haplogroup J2. These matched at 12 markers but diverged as more markers were added.
Two men, one a Fuchs descendant from the Ukraine, tested Haplogroup Q (backbone) and 25 markers. The other, of known American Indian male line descent, who tested Haplogroup Q-M346*, or Q1a3*, confirming his ancestry. Twelve markers only.
Haplogroup R1a1a Glastonbury, CT, and Dorset, England
This group is a real challenge. Descendants of Richard Fox of Glastonbury, MA, and of Vaniah Fox, who was married in Glastonbury, match a descendant of Harry Francis Fox of Dorset, England, almost exactly at 37 markers. Two have been tested Haplogroup R-M198, or R1a1a and one, (N55006) has been extended to 67 STR markers. They are closely matched at 67 markers by an adoptee surnamed Rude.
Haplogroup R1a1a Other
Two Foxes of German descent who do not match other Foxes closely – one tested at 37 markers the other at 12.
Haplogroup R-L1/S26 Null at DYS 439
This matching group consists of:
(1) Four Foxes descending from Justinian Fox of Philadelphia and his father, Edward Fox of Plymouth, England.
(2) Two Foxes descending from Francis Fox of Wilshire, England, plus a third Francis Fox descendant tested at 17 markers by Jobling and King and not listed on the Y-Results tables. Francis Fox is said to have been from the same family as Sir Stephen Fox.
(3) Three Foxes descending from Matthew Fox of Abbeville, SC, and his father, John Fox.
(4) A matching Clark ancestor from Lycoming County, PA, who was adopted.
(5) A less closely matching descendant of Thomas Dudley Fox of Vermont and Canada.
(6) A distant connection tracing back to John Voss of Germany.
All tested by FTDNA were null at DYS 439 and thus members of Haplogroup R-L1/S26 or R1b1a2a1a1a3. One (14179) was tested R-S26 by Ethnoancestry, which is the same as R-L1 at FTDNA. This Haplogroup is consistent with a late arrival to Britain. The Y-Results tables do not show this Haplogroup since S26 was not tested by FTDNA.
Haplogroup R-L2 and Subclades (tested) No Matches in Project
Including three members of this R-U106 Subclade, one of whom (50481) suspects royal Stuart ancestry, one (28579) who was also tested by Jobling and King and one (201967) who does not know his ancestral surname. This Haplogroup is also consistent with a late arrival in Britain.
Haplogroup R-L21 (probable) Andrew Fox of VA and TN Descendants.
Fourth cousins, once removed, these Foxes differ at 3 markers in the 26 through 37 marker panel. The most recent common ancestor (MRCA) was Andrew Fox’s son Jacob Fox, who married Elizabeth Broyles. The Haplogroup assignment is from Neal Fox’s clustering analysis.
Haplogroup R-L21 (tested) No Matches in Project
Three Foxes with ancestry from Ireland, Lancashire, England, and Saarland, Germany, plus a Bonham with possible Fox connections. The Fox with German ancestry suspects a reverse migration from Britain to Germany.
Haplogroup R-L21 subclade DF13 William Fox, Loudoun County, VA, Descendants
A group of three Foxes, three with records tracing back to William Fox, b 1710, in Loudoun County, VA, plus another matching William Fox descendant tested by Relative Genetics and not listed in the Y-Results tables. Two have tested 67 markers and one (59674) has tested Haplogroup R-L21 or R1b1a2a1a1b4, with negative results for M37, M222 and P66 downstream. They are closely matched at 67 markers by a descendant of Enos Fox, 1814, KY. Two Hugh Fox (1745, VA) descendant match them on 32 markers out of 37 with only 5 single step deviations and has been tentatively included in this group. Another Hugh Fox descendant matches at 12 markers. This Haplogroup determination is consistent with an early arrival in Britain. Three have been tested out to 67 markers
Haplogroup R-L47 Henry Fox/ Anne West Group
Five Foxes tracing back to Henry Fox, III, generally accepted as the grandson of Henry Fox and Anne West, plus an African American who matches them closely at 67 markers and is also tested Haplogroup R-L47 or R1b1a2a1a1a4a. This group can be traced back to Henry Fox, b 1521 in Missenden, Buckinghamshire, England. One researcher has proposed a line of descent from Robert de Vaux, a Norman invader. These men are part of a founder cluster that contains Norman names but there is, as yet, no direct evidence of a Vaux/Fox match. The Haplogroup assignment is consistent with a late arrival to Britain.
Haplogroup R-L47 Gloucestershire Ancestry
Four closely matching results with ancestry from Alabama, Georgia and Gloucestershire, England, plus a Smith, with Fox connections, also from Bristol, Gloucestershire. The Haplogroup is estimated to be R-L47 by the project administrators.
Haplogroup R-M269 & Subclades Descendants of Francis Fox of NC and SC.
Descendants of two different sons of Francis Fox, Jr.’s, son, William Moses Fox, 1801, Wilkes County, NC and his wife, Morning Ayres, these third cousins match on 37 markers out of 37.
Haplogroup R-M269 & Subclades Descendants of John B. Fox, 1745, of Orange Co., VA
One tested at 37 markers, the other at 12, they match exactly at 12 markers. Fourth cousins, once removed, the most recent common ancestor was John B. Fox, who married Ann Barber.
Haplogroup R-M269 Subclade Henry Fox 1745 Pfalz, Germany Descendants
Two matching cousins plus a third who is probably related to them back in Germany.
One is doing the Geno 2.0 test which should determine his subclade.
Haplogroup R-M269 Probably Pennsylvania Ancestry
Four closely matching Foxes, two descending from John Fox of Virginia, whose parents have in turn been identified as William Fox, 1777, Pennsylvania and his wife Sarah, another descending from Levi Fox, Sr.,1802, Pennsylvania, and, finally, one descending from Taylor Fox of Indiana, tested at 12 markers. The latter has been included because of research indicating a connection to the Levi Fox descendant, who has been tested at 67 markers. The Haplogroup, R-M269 or R1b1a2, is estimated – this is the basic Western Atlantic Modal Haplotype and the subclade has yet to be determined.
Haplogroup R-M269 Origin Southern US
There are so many unmatched Foxes in Haplogroup R-M269 and Subclades with Southern American Ancestry that it was deemed advisable to segregate them geographically. For the most recent known ancestor (MRKA) of these project members, simply refer to the Y-Results tables
Haplogroup R-M269 and Subclades (Other)
We have yet to find matches within the project for this large group of Foxes, some of whom have done extensive additional testing. Ancestry (outside the US) goes back to mainly to the UK and Germany but we have one from Romania, one from France and one quite unexpected result from the Ukraine.
Haplogroup R-M269 Vanfossen
A large group of men of Dutch ancestry most of whom descend from Arnold Van Vossen of Germantown, PA, and his descendant Peter Vanfossen of Chester County, PA. Several have changed their name to Fox. One of these has been tested M269+, S21-, S26-, S28- and S29- so they are all also in Haplogroup R-M269. All but 5 of them are close matches to each other. They maintain a separate VanFossen Project
at World Families where more detail is available.
Haplogroup R Subclade Determinations
See Y-DNA SNP Tables
A number of the Foxes in the Haplogroup R groupings have done additional SNP testing and genealogical research and have joined specific Haplogroup Projects, with some very interesting findings. If the origin is British, the R-U106 Subclades were late arrivals to Great Britain whereas R-P312 and R-L21 represent very early arrivals::
14179 was tested S21+ and S26+. These are the same as U106+ and L1+ putting him and his null439 relatives in Haplogroup R1b1a2a1a1a3.
26383 and 48275 are Haplogroup R-L47 or R1b1a2a1a1a4a. This is the Henry Fox/Anne West grouping.
93372 is Haplogroup R-L48 or R1b1a2a1a1a4 and negative for R-L47. Descendant of Matthias Fox, Saar, Germany.
123131 is Haplogroup R-U106 and negative for L1, P107 and U198, yet to be tested for R-L48. Descendant of Hyram Fox of Michigan.
R-U106 and R-U152 Subclades:
28579 is Haplogroup R-L20 or R1b1a2a1a1b3c1, a subclade of R-L2. Descendant of Thomas Fox, 1608, Norfolk, England.
50481 is Haplogroup R-L2 or R1b1a2a1a1b3c. A probable descendant of Charles II, based on a paper trail and Y-DNA haplotype matches, he is L20 negative but has yet to test for L196.
201967 is Haplogroup R-L2*, negative for L20 and L196. Admitted to the project because of Fox connections, he does not know his paternal ancestry.
38430, 65065 and 92311 are Haplogroup R-L21 or R1b1a2a1a1b4 and 38430 is tested Haplogroup R-L21*, meaning negative for all downstream SNPs. 92311 is negative for M37, M222 and P66 downstream of L21. Descendants of George Fox of Lancashire, Michael Fox of Ireland and a Bonham who has Fox connections
58674 is Haplogroup R-L21 and negative for M37, M222, P66 and L193 downstream. This is a member of the William Fox, 1710, Loudoun County, VA, Grouping.
Upstream of R-U106 and R-P312:
38215, with no Fox Project matches, has done additional testing and genealogical research and is Haplogroup R-M310* or R1b1a2a1a1*, a precursor U106 and P312, the more prevalent downstream subclades. English, but possibly with deep ancestry in Italy with the name of Forz. He has joined the R1b1b2/ht35 Project.
See the Y-DNA SNP Table for all SNP results determined by FTDNA, including other Haplogroups.
Summary of Y-DNA Testing
We have representatives of most of the major European Haplogroups represented in the Fox Project and, within the most common Haplogroup R, we have representatives of both of the major subheadings (R-U106 and R-P312) as well as one who is negative for both. You have to conclude that the Fox surname was adopted many times by many people of differing backgrounds as well as being the Americanized version of similar names such as Fuchs, Forz, Van Vossen and, possibly, Vaux.
Anyone with the surname Fox or its variants is welcome to join the project, including transfers from the Genographic Study, and we now have enough members that there is a good chance of establishing a connection. There are at least eight separate families represented in the Fox project that trace back to the VA, NC, TN, KY areas so Foxes with early ancestors from the Southern USA can reasonably expect to find future matches. We have at least seven members of German descent and also have several members with Irish ancestry. We are anxious to get more members of known Irish or German descent.
Mitochondrial DNA Testing
Many members (both men and women) have ordered Mitochondrial DNA testing in order to establish their maternal ancestry. Mitochondrial DNA surrounds the cell nucleus and is passed from the mother to her children of either sex. It is possible to use mtDNA testing to support a maternal paper trail but there are several problems. First, the straight maternal line is often not known back more than a few generations and, of course, the surname changes with each generation. Second, the mutation rate is much slower than in STR testing of the Y-Chromosome and, to establish a close connection, some people (including a number of Fox project members) have ordered the full mtDNA sequence. For both these reasons mtDNA testing has proven most valuable in studying the migration of populations. It is less useful for genealogical purposes, though some success stories have been told. For this reason, the administrators leave it up to the individual member to track down his or her own matches.
Nevertheless, a tabulation of mtDNA results is available on this website and can help to verify a paper trail. The results presented show where the sequence of nucleotides differs from one particular reference sample - the Cambridge Reference Sequence - in two regions of the mtDNA molecule - Hypervariable Regions 1 and 2 (HVR1 and HVR2.) Full genome mtDNA sequencing results for a number of project members are considered confidential.
Family Finder Testing
FTDNA’s version of autosomal chromosome testing for genealogical purposes is called Family Finder and is discussed elsewhere. For a project where the background of the project members are as diverse as are the Foxes in this project, it would be foolhardy to try to correlate results. Seven project members have elected this option so far and are following up on their own results.
The group administrators would like to acknowledge the assistance of other project members who have spent hours analyzing data and making contacts. Clay Fox has recruited members of the Vanfossen clan into the project and Dr. Lewis Fox has been recruiting Foxes with New Jersey ancestry. Genealogists John William Fox and Paul Fox have also been of great help in reviewing the paper trail and in soliciting prospective members. Terry Fox, John Gary Fox, Jon Myron Fox and Roland Fox have also been avid researchers and Roland was the one who got us Foxes into DNA testing in the first place. Kevin Archer Fox has also been very helpful in finding recruits and assessing results as has Kevin Daniel, a descendant of William Fox of Loudoun County, VA.
James Neal Fox is project co-administrator and has done extensive studies that group Haplogroup R results into related clusters. He also helped set up the Haplogroup L1/S26 (null39) Project at FTDNA. John Henry Fox has provided access to the Steadman Genealogy Books and, in general, project members have been most cooperative in supplying family background information. When this is done, the project administrator sees that their results are uploaded to the Ysearch database (though this is presently impossible.) All of us are unpaid volunteers.