ANCESTRY BY DNA ANALYSIS Dannie Lee Craycraft Edited By Robert Lynn Craycroft And Stephen Patrick Cracroft-Brennan Copyright 2002, 2003, 2007, 2012 INTRODUCTION Very recently, the subject of using DNA testing to do "reconstruction" genealogy has been in the news. From the basic genealogy work that I have done on the first two books on the various Cracroft lines (includes all spellings), DNA testing had an obvious appeal. I will use the England surname Cracroft in this work to include all the various spellings of our surname: Cracraft, Craycraft, Craycroft and Creacraft. I work with a small group of dedicated Cracroft genealogists on the Internet. I especially wish to thank Robert Lynn Craycroft for his assistance on this report. Our group has tremendous exposure on the Internet hosting several websites. We have been working as a group for several years now. No one can do a Cracroft search on the Internet without finding one of our member's postings or websites. With this group as a base, we selected a number of male volunteers to participate in a DNA study to help verify our "documented" family genealogy. There were some surprises! Our belief has always been that regardless of the spelling of our surname, we all descend from the England Cracroft line. From the work of a highly respected and well-known heraldry authority, Stephan Patrick Cracroft-Brennan HonFHS FSA, Ragemar circa 1050 is the oldest identified Cracroft ancestor. Patrick has done extensive work documenting the Cracroft genealogy. The desire to identify and discover our roots is an increasingly popular American pursuit. We came here from England for many reasons: religion, freedom, and a chance to have a new beginning. However, some 200 plus years later, Americans of all heritages are striving to uncover their roots. We Craycrafts and Craycrofts are no exception. Dannie Lee Craycraft THE PROGRAM AND ISSUES The basic goal of our DNA program was to verify that all the various spellings of our surname in this country actually descend from a common English ancestor. We have proven this in part. The lines tested were: Cracraft, Craycraft, Craycroft and finally Cracroft. John Cracroft, with wife Ann, is listed in "The Early Settlers of Maryland, An Index to Names of Immigrants Compiled from Records of Land Patents, 1633-1680, in the Hall of Records, Annapolis, Maryland" by Gust Skordas. His surname there is spelled Creacroft. His line here in America is mostly spelled Craycroft. However, we have not tested all the descendents of each of these lines. I am sure that there are descendents of these lines who would refute our conclusion. Also, we have not tested any Creacraft, Crecraft, Creacroft or Creacraft lines. To understand the test program and its results, one must understand the basics of DNA testing and what it can tell us. For this explanation, I have used, with the permission of Bennett Greenspan, the work done by FamilyTreeDNA. This is the DNA testing lab we used to perform the testing for us. However, before getting into the technical discussion, I should explain circumstances that might affect the results of the DNA testing: DNA testing can only be done on the male descendent line for purposes of establishing a common ancestor between two males. By this I mean the Y Chromosome DNA that is passed only from father to son is used to determine if two males share a common ancestor. It cannot be done on the female line. There are several things that can cause problems when testing the male descendent line. First, at times a non-Cracroft individual may be adopted into the family line and assume the family surname. This means he is not a blood relation. Second, a female Cracroft may have a male child out of wedlock and the male child assumes the family name. Although technically this male child is a Cracroft, he has broken the male descendent Y Chromosome DNA chain since he descended from a female Cracroft. There is a third possibility, which Patrick Cracroft-Brennan thinks highly unlikely because of the name originating from ownership of the manor of Cracroft, of the person being tested coming from a family of the same name but which has no connection to Ragemar. When surnames first originated in the 13th century spelling was very mutable. It was a descendent of Ragemar, Lord of Welle, Claxby, Withern, etc., Stephen de Cracroft, Lord of the Manor of Cracroft, who first adopted the Cracroft family name in about the 11th century. Stephen is known to have been living in November, 1256. There is no reason to think that Cracroft, Craycroft and Craycraft are separate families and not, as they plainly are, phonetic variants of the same name. Records exist in England that show the surname Craycraft existed in England as a totally distinctive surname used by members of the same family as recent as the 1600's. We have been in communication with English Craycrafts this past year (2006). Where or how the use of Craycraft as a surname by members of the family in England came into practice is unknown. Patrick Cracroft-Brennan is carrying out research in this area. He is of the opinion that the large number of Craycrafts/Craycrofts that one comes across in Kent are probably descended from the Craycrafts/Craycrofts that were working in the City of London in the 17th and 18th centuries. In their turn some of these might have descended from his own Gainsborough line and others from the Winthorpe and Burgh branches of the family. WHAT IS DNA TESTING The subject of DNA is highly complex. I will attempt to present information here to help you with the basics. If it gets too complex, don't be concerned. The most important concept is that fathers pass on their genetic code to their sons and their sons to their sons and this goes on for hundreds of generations. Because of this trait, we can determine if a modern day Craycraft is related to our most distant ancestor, Ragemar. I will only be discussing Y-Chromosome DNA (deoxyribonucleic acid) testing. The mother does not pass the Y-Chromosome to her children. That is why only males can participate in such a program. Variations in the Y-Chromosome allele passed from father to son do occur but the occurrence is estimated at 0.2% per generation or once in every 500 generations (about 10,000 years). FamilyTreeDNA, our testing lab, tests 12 alleles (markers on the DNA chain) for basic testing. These alleles are sometimes referred to as "systems", loci, "microsatellite loci" or markers. The 12 markers tested are: DYS 385a, 385b, 388, 389-1, 389-2, 390, 391,392, 393, 394, 426 and 439. Marker DYS394 is sometimes referred to as DYS19. A computer is used to read the DNA scans. It measures the size of the allele result and assigns a weighted number to it. These are the numbers that are compared between individuals being tested. If, for instance, the DYS 388 allele result is off by one between two individuals and all the other numbers are equal, then it is called an 11/12 match. It means that the two individuals are related but the Most Recent Common Ancestor (MRCA) is further back in time. PARTICIPANTS IN THE DNA RESEARCH PROGRAM We wish thank the following individuals representing the indicated family lines for participating in this research program. Close to a hundred emails and letters were sent out here and in England. Dozens of messages were posted to numerous genealogy bulletin boards on the Internet. In some cases we even offered free testing. There were few positive responses. Those currently researching our family surname and our future descendents will always be grateful for their assistance. The following lines were tested as part of our project: Line 1 - Peter Kenneth Cracroft Peter Kenneth26 Cracroft (Hugh John Bernard25, Hugh24, Bernard23, Walter Sewell22, William21, Robert20, Robert 19, Robert 18, Robert 17, John16, Francis15, John14, Richard13, John12, Robert11, John10, Robert9, John8 de Cracroft, Robert7, William6, Stephen5, Walter4, Humphrey fitz3 Walter, Walter2 fitz Ragemar, Ragemar1 de Welles) (Ragemar line) Line 2 - Dannie Lee Craycraft Dannie Lee9 Craycraft (Donald Brant8, Charles Abner7, William Harrison6, John5, John4, Charles3, Samuel2, Joseph 1, Unknown Cracroft ) (author's line) Line 3 - Kelly Dean Cracraft Kelly Dean8 Cracraft (Don Clifford7, Francis Richard6, Eli Leonard5, David4, Joseph3, Joseph2, Joseph ?1 ) ( Holly Dianne Cracraft line) Line 4 - Robert Lynn Craycroft Robert Lynn7 Craycroft (William Clifton6, William Clifton 5, William Tecumcah4, William Benjamin3, Edward Scott2, Benjamin Joseph1 ) Line 5 – (Private) Craycroft (Private)6 Craycroft (Robert Newton5, John Henry4, Benjamin3, Benjamin 2, Benjamin Joseph1 ) Line 6 - Matthew Charles Craycraft Matthew Charles6 Craycraft (Charles Theodore5, Chester Reuben4, Charles Reuben3, Reuben (Ruben)2, Charles C.1 ) (Dianne Hume, Denise Miller and Jeff Craycraft line) Line 7 - Glen Woodrow Craycraft Glen Woodrow8 Craycraft (Oliver Muir7, Richard6, Enoch5, William4, William3 Jr., William2, Joseph 1 ) (Nancy Stull's line) Line 8 - Timothy Craycraft Timothy8 Craycraft (Wilbur7, John Edgar6, Zedekiah5, William4, William3 Jr., William2, Joseph1 ) (Nancy Stull's line) Line 9 - Rev. Bernard Leo Craycroft Bernard Leo9 Craycroft, Rev. (George Hix8, Thomas Francis7, Charles Alexius6, Thomas Alexius5, Thomas4, Ignatius3, Ignatius2, John1 ) (John & Ann Craycroft line) Line 10 – John Shaver Craycroft John Shaver6 Craycroft (Charles Burr5, Harry Judge4, John Wesley3, Benjamin2, Benjamin Joseph1 ) Line 11 – Justin David Cracroft Justin David30 Cracroft (David Storrs29, Ralph Paul28, Ralph27, George William Bone26, Thomas John25, James24, Thomas23 Luke22, Luke21, Luke20, Charles19, Charles18, George17, Thomas16, Robert15, William14, Richard13, John12, Robert11, John10, Robert9, John8 de Cracroft, Robert7 de Cracroft, William6 de Cracroft, Stephen5 de Cracroft, Walter4 fitz Humphrey, Humphrey3 fitz Walter, Walter2 fitz Ragemar, Ragemar1 de Welles) Line 12 – Larry Frantz Cracraft Larry Frantz10 Cracraft (Harlin Stanley9, James Marshall8, George Washington7, John T.6, Joseph5, Joseph4, William3, Joseph ??2, Unknown1 Cracroft) FAMILY LINES TESTED Three distinct family lines and a fourth line of an undetermined ancestor were sampled with Peter Kenneth Cracroft, representing direct descent from Ragemar, as the control sample. Reverend Leo Craycroft is a documented descendant of John and Ann Craycroft, the earliest known immigrant to America. Dannie Lee Craycraft, Kelly Dean Craycraft, Matthew Charles Craycraft, and Timothy Craycraft are all documented descendants of Joseph and Nancy (Ann) Stanton Craycraft or Craycroft. Glen Woodrow Craycraft is in the same line but is in a broken line of descent. Robert Lynn Craycroft and (Private) Craycroft are descended from Benjamin Joseph Craycroft. Based on oral tradition as yet unproven it is possible that Benjamin Joseph was also a descendant of John and Ann. Research continues on this aspect. According to the few documents available, a Joseph Cracraft or Cracroft came here circa 1702 from Lincolnshire, England. He came here with his wife Ann or Nancy Ann Stanton. He may have even married her here in America. We don't know. He had six children: Joseph Jr., William, Samuel, Charles, Sarah and Thomas. Where this narrative becomes less certain is the true origin of Joseph and Nancy Ann Stanton. It is possible that our Joseph line actually may descend from one of John and Ann's descendants. However, this uncertainty does not interfere with our DNA project. What is well documented is that there was a John Craycroft (ship manifest spelling: Creacraft or Creacroft) who arrived here circa 1665 with his wife Ann from Lincolnshire, England. We have recently traced this John (see iv below) back to the following line in England. I must state, however, that we do not have conclusive proof on this connection to Robert's son John. Circumstantial evidence in the USA looks very good, but, as Patrick Cracroft-Brennan has said, there are no surviving records in England that mention any son of Robert and Martha Cracroft emigrating to Maryland . Descendants of Robert Cracroft ROBERT17 CRACROFT (JOHN16, FRANCIS15, JOHN12, ROBERT11, JOHN10, ROBERT9, JOHN8, ROBERT7, WILLIAM6, STEPHEN5, WALTER4, HUMPHREY3, HUMPHREY2, RAGEMAR1). Robert was born in Whisby, Lincolnshire, and died 1667. He married Elizabeth and then Martha Amcotts on February 28, 1608/09 in Aisthorpe, Lincolnshire, daughter of Sir Richard Amcotts) Children of Robert Cracroft: (i) Robert b.1610 d. before 8/23/1647. Married Margaret Remington (ii) Mary b. before 3/28/1611. d. aft 1667. Married John Shelton (iii) Mary b. about 1614 (iv) John b. about 1616 d. 1697. Married Ann Unknown (this is not verified from records). (v) Elizabeth b. before 3/17/1615/16. Married William Bedell (vi) Richard b. before 11/3/1617 d. about 1684. Married Susannah Unknown (vii) Jane b. before 11/3/1623. Married Francis Fland (viii) Martha b. bef 1/8/1624/25. Married Edward Barker (ix) Thomas b. before 4/8/1628 d. about 1680. Married Mary Whelpdale (x) Henry b. before 8/15/1631 The line of Benjamin Joseph Craycroft coming forward to the present is extremely well documented. The problem arises in that his ancestors have not been identified. There is a fictional family history that gives a spotty record of his ancestors dating back to 1297 but this has been proven to almost completely false. But as stated earlier there is a possibility that Benjamin Joseph is a descendant of John and Ann Craycroft of Maryland. Here are the results of the DNA testing. 393 390 19 391 385a 385b 426 388 439 389|1 392 389|2 458 459a 459b 455 454 447 437 448 449 464a 464b 464c 464d Name Peter K. Cracroft 13 24 13 11 11 14 12 12 11 13 13 29 17 9 10 11 11 25 15 19 29 15 15 16 17 Rev. B. Leo Craycroft 13 24 13 11 11 14 12 12 12 13 13 29 17 9 10 11 11 25 15 19 29 15 15 16 17 John Shaver Craycroft 13 24 13 11 11 14 12 12 12 13 13 29 17 9 10 11 11 25 15 19 29 15 16 16 17 Dannie Lee Craycraft 13 24 13 11 11 14 12 12 12 13 13 29 17 9 10 11 11 25 15 19 29 15 15 17 17 Timothy Craycraft 13 24 13 11 11 14 12 12 12 13 13 29 17 9 10 11 11 25 14 19 29 15 15 17 17 (Private) Craycroft 13 24 13 11 11 14 12 12 12 13 13 29 Kelly D. (c/o Holly Pace) Cracraft 13 24 13 11 11 14 12 12 12 13 13 29 Matthew Craycraft 13 24 13 11 11 14 12 12 12 13 13 29 Glen W. Craycraft 13 22 14 10 14 14 11 14 11 12 11 27 Justin David Cracroft 13 25 15 11 13 17 11 13 11 12 11 28 17 8 10 9 12 24 15 21 28 14 14 15 16 Robert Lynn Craycroft 13 24 14 10 11 15 12 14 12 13 13 29 16 9 10 11 11 25 15 19 29 14 15 17 17 Larry Frantz Cracraft 13 24 13 11 11 14 12 12 12 13 13 29 17 9 10 11 11 25 15 19 29 15 15 17 17 These results can be interpreted in this manner. Peter, being the control sample, is in the 26th generation following Ragemar. Rev. Leo Craycroft, nine generations removed from John Craycroft (the original immigrant), is also the closest match to Peter's sample. Knowing that Peter K. Cracroft and Rev. B. Leo Craycroft could not have had a common ancestor in the last 8 generations, their 25 marker comparison, with one variation in results, shows that the probability that they shared a common ancestor within the last: 8 - 12 generations is 46.35% 16 generations is 73.03% 20 generations is 86.98% 24 generations is 93.89% 28 generations is 97.19% 32 generations is 98.73% So it is there is about a 94% probability that Peter and Leo share a common ancestor within the last 24 generations. The next results to look at are those of John Shaver Craycroft. John's descendancy from John and Ann Craycroft is not known at this time, but given the fact that John and Rev. Leo have only 1 marker difference at 464b it is likely that John, like Rev. Leo, descends from John and Ann. Comparing John's results to Peter and given that they have no common ancestors in at least the last 10 generations, with two variations in results these are the probabilities that they shared a common ancestor within the last: 18 generations is 60.1% 22 generations is 77.18% 26 generations is 87.49% 30 generations is 93.37% Looking at the differences between Dannie and Peter we see that Dan's earliest known ancestor is presumed to be Joseph Craycroft. Joseph was born approximately two generations after John, the immigrant, so this would seem to indicate that there couldn't be a common ancestor within at least 10 generations. Assuming that Peter K. Cracroft and Dannie L. Craycraft could not have had a common ancestor in the last 10 generations, their 25 marker comparison, with two variations in results, shows that the probability that they shared a common ancestor within the last: 10 - 14 generations is 34.19% 18 generations is 60.1% 22 generations is 77.18% 26 generations is 87.49% 30 generations is 93.37% 34 generations is 96.57% As with Dannie, the same reasoning would apply to Larry Frantz Cracraft as they have identical results in the 25 maker testing. Timothy is the eighth generation from Joseph so he would be 10 generations from any possible common ancestor with Peter. Knowing that Peter K. Cracroft and Timothy Craycraft could not have had a common ancestor in the last 10 generations, their 25 marker comparison, with three variations in results, shows that the probability that they shared a common ancestor within the last: 10 - 14 generations is 21.89% 18 generations is 44.37% 22 generations is 63.14% 26 generations is 76.93% 30 generations is 86.19% 34 generations is 92.03% Kelly's results are based upon a 12-marker test and he is also descended from Joseph, putting an additional two generations distant from the time of John, the immigrant. Assuming that Peter K. Cracroft and Kelly D. Cracraft could not have had a common ancestor in the last 10 generations, their 12 marker comparison, with one variation in results, shows that the probability that they shared a common ancestor within the last: 10 - 14 generations is 17.61% 18 generations is 33.96% 22 generations is 48.11% 26 generations is 59.85% 30 generations is 69.3% 34 generations is 76.75% The results for (Private) Craycroft are more difficult to interpret because the earliest verified ancestor in this line was born in 1779, more than 100 years after the arrival of John, the immigrant. While he is only six generations removed from the known ancestor, Benjamin Joseph, there are at least another five generations farther back to any possible common ancestor to Peter. Assuming that Peter K. Cracroft and (Private) Craycroft could not have had a common ancestor in the last 12 generations, their 12 marker comparison, with one variation results, shows that the probability that they shared a common ancestor within the last: 12 - 16 generations is 18.83% 20 generations is 35.63% 24 generations is 49.83% 28 generations is 61.43% 32 generations is 70.66% 36 generations is 77.87% Matthew's results are also problematic. He is only six generations removed from his earliest verified ancestor, Charles, but Charles was born about 150 years after the arrival of John, the immigrant. This would indicate that Matthew is at least 13 generations away from any potential common ancestor. Assuming that Peter K. Cracroft and Matthew Craycraft could not have had a common ancestor in the last 13 generations, their 12 marker comparison, with one variation in results, shows that the probability that they shared a common ancestor within the last: 13 - 17 generations is 19.36% 21 generations is 36.35% 25 generations is 50.58% 29 generations is 62.11% 33 generations is 71.24% 37 generations is 78.36% It is possible that a 25-marker test performed on these last samples could well change these percentages, but a lack of funds at this time prevents further testing. HYPOTHOSES For the lines tested, the DNA testing has proven that the family surnames Cracraft, Craycraft, Craycroft all share a common ancestor. DNA testing alone cannot tell us who this common ancestor was or exactly how far back in time we connect with him. Since Peter Cracroft is a known descendant of Ragemar de Welles we have used his test results as a baseline for the purposes of determining relationships. Based on interpretations by Eileen Krause of Family Tree DNA we can make the following assumptions: 1. Peter, John Shaver Craycroft and Rev. Leo Craycroft may belong to the same branch because they share a 16 at DSY464c 2. Rev. Leo and John Shaver may be from a sub-branch of Peter's branch because of the variation at DYS439. (DYS439 is what is called a "fast mover", i.e. this marker changes at a faster, but unknown, rate than other markers.) 3. Dan Craycraft and Timothy Craycraft may be from the same sub-branch because they share a 17 at DYS464c. 4. Matthew Craycraft, Kelly Cracraft and (Private) Craycroft may descend from the same branch as Dan Craycraft or Rev. Leo Craycroft but this is difficult t determine as their tests are only for 12 markers. 5. Robert L. Craycroft is known to descend from a "non-paternity event", that is one of his male ancestors is not a Craycroft. From the results of (Private) sample we know that this occurred within the last five generations. Further DNA matching has determined that Robert is genetically related to a completely different family of Lithuanian heritage. 6. Justin David Cracroft's sample has presented us with another mystery. His line of descent is relatively well proven, but his DNA results are so far varied from Peter Cracroft's and everyone else's that there is no possible genetic relationship between Justin and Peter Cracroft. Patrick Cracroft-Brennan's research has given him some doubt as to the actual parentage of Charles Cracroft, the 19th person in Justin's lineage. Perhaps this result is a possible confirmation of these doubts. Patrick Cracroft-Brennan, who has always been skeptical of John (son of Robert and Martha Cracroft) being identified with the Maryland immigrant, is not surprised at these findings. If, for example, John the immigrant was the same person as John the apprenticed fishmonger and 2nd son of John Cracroft of Burgh-le-Marsh and Fulnetby, then the common ancestor with Peter Cracroft would be back a further six generations. The problem is that in the 16th and 17th centuries there were a lot of Cracroft "branches" which have yet to be grafted to the main tree and John the immigrant could descend from any of these. For all practical purposes, this does not really help in pinpointing who this common ancestor might be. We have no plans to do any further testing at this time. We have proven what we set out to prove, that the American families who share variations on the England Cracroft surname all descend from a common English ancestor. It proves that we are all one family regardless of how we spell our ancestral surname. We did discover one disturbing piece of information in our testing which I have briefly touched upon. Robert Lynn Craycroft, who is a member of our Internet group, has not been able to tie his family line (Benjamin Joseph Craycroft) into our Joseph Cracraft line, the John and Ann Craycroft line or the English Ragemar Cracroft line. The results of his DNA testing indicate that he does not share our DNA. His results were off by more than one number in four markers. This result is far more than genetic mutation might account for. To resolve the issue, Robert recruited another member of his family line, (Private) Craycroft. This man branches off from Robert's line at Edward Scott Craycroft born 1809. As you can see from the DNA results, his results matched mine at 12 markers. This result confirms that Robert Lynn's line had a break in the male DNA Y-chromosome chain, most probably due to an adopted son or an illegitimate child. We encountered a similar problem when we tested Glen Woodrow. His DNA totally did not match any of the other tested volunteers. This has been explained by descendants of this line that claim that family "history" believes that a female Craycraft in that line had a son who assumed the Craycraft name. As noted in Item 6 above, Justin David Cracroft's DNA has given results we were not expecting. Justin's Haplotype indicates that he is of Nordic extraction, rather than Western European as the Cracroft line is. Regarding Justin's results another, very important, possibility was recently (September, 2007) raised by Patrick. One fact that we have overlooked during the DNA analysis is that Ragemar, the earliest known ancestor was in fact of Danish origin. This is included in the Nordic haplotype. It is possible we have gotten so caught up in our English roots that we have expected to see results representing the Western European Haplotype that we have lost sight of Ragemar's origins. Perhaps we should have been expecting a Nordic or Danish haplotype, which is represented by Justin's results. It is possible that Peter's DNA results actually represent a genetic break in the family line and Justin's DNA results are indicative of a direct genetic descent from Ragemar. What we know for certain is that Justin and Peter share a common ancestor in Richard Cracroft, known to be living on 9 Apr 1490 when he was named in his father's will, proved on that date. Richard had only two known sons, William (married to Elizabeth Howson) and John (married to Jane Claymond). Justin descends from William and Peter from John. It is curious to note that while 17 generations separate Justin and Richard, only 13 generations separate Peter and Richard. This could be an anomaly or an indicator. Only further DNA testing of this line will answer the question. At this time there are no other candidates coming forward to volunteer a DNA sample however. (NOTE: Larry Franz Cracraft was tested after Robert Lynn Craycroft and Stephen Patricka Cracroft-Brennan edited this report in 2007. Larry was tested by FTDNA in July/August, 2012 and I have added his results to this report. Dannie Lee Craycraft) APPENDIX A So that those not familiar with genealogical DNA testing may better understand the meaning of the results of our DNA project here is an explanation of how this testing is conducted and its significance. This testing is referred to as Y-Chromosome STR testing. You may or may not know that females carry two X-chromosomes and males carry 1 X- and 1 Y-chromosome. It is the Y- chromosome that Surname DNA testing is performed on and this is why only males can participate in the testing. DNA testing for female lineage is called mtDNA (mitochondrial DNA) and tracks genes from the mother to all her children. As such it is only useful when tracking the maternal lines. DNA is often called the 'molecule of life'. Within it are the code words that help build proteins and thus help our bodies to function. However, most of it doesn't encode for anything at all, these regions are called "junk DNA" and they contain no medical information whatsoever. Family Tree DNA, the service we have used for testing, currently offers testing for 12, 25, 37 and 67 markers found in this junk DNA. At these markers, the sequences of bases, represented by T (Thymine), G (Guanine), A (Adenine) or C (Cytosine) repeat themselves many times. These are also called STR's or Short Tandem Repeats. An example would be at marker DYS391 for Peter Cracroft the sequence TCTA is repeated 11 times before a different sequence is seen. This returns a result of DYS391 = 11. For DYS391 the number of repeats can range from 7 to 14. Since the Y-chromosome only comes from the father his sons will have this same number of repeats at each marker. Very occasionally these repeats will increase or decrease, usually only one at a time. This is called a mutation and happens when the DNA is copied slightly incorrectly within the body. This is a natural phenomenon. These mutations are very useful because otherwise every male would have exactly the same Y-chromosome. The fact that related males that share the same surname will have identical Y-chromosomes (or at least a very close match), is why Y-chromosomes are so useful to genealogists. It is also very easy to tell if two males aren't related at all. Mutations can be expected to appear over long periods of time, i.e. over the span of 10-25 generations. But a DNA sample with more 2 or 3 mutations from the baseline sample is highly unlikely to be related to that person. Mutations are generally thought to occur once in every 500 "transmission events" or male births, in other words 0.2% per generation. So given the number of markers Family Tree DNA tests here is the expected rate of mutation: 12 markers = ~ 41 generations 25 markers = 20 generations 37 markers = ~ 13 generations 67 markers = ~7 generations What this means is that in a 12-marker test we can reasonably expect to see .984 mutations. In a 25-marker test we may see 1 mutation. In a 37-marker test we may see 1 mutation. In a 67-marker test we may see .938 mutations. Effectively we may see only 1 mutation in any test, although 2 or 3 mutations is possible. Anything more than that would eliminate that sample as being related. (This explanation is drawn from DNA Heritage, another DNA testing organization.)