----- Original Message ----- From: Charles Austin To: Harold Steiner Cc: Charles Austin Sent: 12/22/2009 9:29 PM Subject: Which William? Dear Hal: I’ve been thinking about two Austin problems which we have discussed: the possible relationships between (1) Philip and Absalom, and (2) Larkin K. and James S. The second question is easiest to deal with, because despite supporting evidence of geographic proximity (in Wayne County, Illinois, as shown by census records), the DNA evidence argues against Larkin and James being brothers. Larkin and James The 1840 federal census of Illinois showed a James Austin and a Larkin K. Austin separated by four lines on the census page , possibly making them near neighbors, in an unidentified location in Illinois (“the division allotted to Daniel Lunney [?].” ) Later census records included: 1850: James Austin with wife Delilah S. and four children (not including Francis Marion, b. 1852) in Hitesville Precinct of Coles County, but no record of an Austin named Larkin K., Larkin, L.K. or L.R. 1860: L.R. Austin with wife Louisa A. and seven children including Jasper, age 8 and Jonathan, age 7, in Township 1S, range 5E (which may have been 5W according to one source) in Wayne County. This was clearly Larkin K. despite the wrong middle initial, because Jasper and Jonathan were in the two DNA lines from Larkin. No record of James and Delilah found. (There is a James Austin, age 20, living near “L.R.”, possibly an older son of Larkin.) 1870: James Austin with wife Delilah S. and seven children, now including Francis, age 8 (b.1852), in Four Mile Township of Wayne County. No record of Larkin (or “L.R.”) found. !880: Jasper Austin, apparently age 24, with wife and two children, may be Larkin’s son Jasper despite age discrepancy (Larkin’s Jasper b. 1853), in Four Mile Township of Wayne County. 1900: Jonathan Austin, b.1855, with wife and nine children, in Four Mile township of Wayne County. This suggests that James S, and Larkin K. moved to Illinois with their families by 1840, and at least after 1860 may have both lived in Four Mile Township in Wayne County in southern Illinois. This in turn suggests at least an acquaintance and probably a family relationship, but the DNA results give a fairly small probability that they were brothers. Their donors are either a 35/37 match (kits 32203 and 118609) or a 36/37 match (kits 98075 and 118609). If Larkin and James were brothers, their father would be the MRCA of each of these pairs of donors. He would be four generations back from the donors. The probability of an MRCA within 4 generations, assuming mutation rate 0.0045, is 14.2% for a 35/37 match, and 37.8% for a 36/37 match. With an average mutation rate of 0.004 the probabilities would be slightly lower. These are “Bill Jackson” estimates, and FTDNA might give slightly higher probabilities, but they do not seem to support your hypothesis that Larkin and James were brothers. On the other hand, a low probability is not the same as impossibility. Improbable events do occur. Consider the two donor-descendents of Larkin (his great grandchildren). They are only a 34/37 match, but their documented MRCA is Larkin himself, only three generations back from them. The Bill Jackson probability of a 34/37 match having MRCA within at most 3 generations is 1.7% (using rate 0.0045). Nearly impossible! But it happened. See why they say that DNA cannot prove anything, it can only give probabilities? Absalom and Philip Each of Absalom and Philip has one line leading to a DNA donor (kit 124161 for Absalom, kit 27680 for Philip), and those donors are a 35/37 match. If Philip’s father and Absalom were brothers, their father would be the MRCA of these two donors. He is 9 generations back from 124161 and 8 generations back from 27680. The Bill Jackson probability of an MRCA within at most 9 generations, for a 35/37 match, is 47.9% with mutation rate 0.004, and 55.7% with mutation rate 0.0045. This does not preclude their being brothers, but it is not very strong evidence that they were brothers. (The chances are about 50/50, one would say.) If you have good documentation showing that Absalom was Philip’s uncle, then he probably was. The DNA result doesn’t contradict this. The number of generations needed to get the near certainty of 95% probability is 22 if the mutation rate is 0.004, and 19.6 if the rate is 0.0045, as you correctly observed. Even assuming that the average rate for the first 37 markers is 0.0045, 19.6 generations (effectively 20 generations) is still a long time. Converting it to years requires an assumption about the average length of the generations in this particular line. You have assumed 20 years between generations, but the average generational gap in the generations of the Philip line is about 30 years (1937-1758=179, and 179/6=29.83 years). For the Absalom line it is (1972-1725)/8 = 30.9, about 31 years. I would suggest using 30 years as the average generational distance in both lines of descent. If the MRCA is 20 generations back from the Absalom donor (124161), then he is 18 generations back from the Philip donor. The birthdate of this MRCA is approximately 600 years before 1972 (1372), or 540 years before 1937 (1397). Since both these figures are approximations, and since the MRCA of the two donors has 95% probability of being at most 20 generations back from each of them, we can say that the MRCA of the Absalom and Philip donors (who is also the MRCA of Absalom and Philip) was almost certainly born after 1370. That doesn’t really tell you much about the relationship between Absalom and Philip. In fact, 1370 is barely within the age of surnames. Which William? Reviewing the manuscript for your Chapter 11 (which I still have on my computer), I see that the candidates are “(KW4) William” and “(QA11) William.” It appears that (KW4) William is the son of William bc.1710 MD, and great-grandson of William bc.1657 of MD (Kent County). But (QA11) William, son of John and Rosamond, is not on the chart. Nothing is known about his DNA. The donor descendents of (KW4) William (kits 18245 and 34236) are both 34/37 matches with the donor descendents of Philip and Absalom, suggesting that Philip and Absalom are related to the line of William of Kent, but not necessarily very closely. (You can work out the Bill Jackson probabilities for the tMRCAs here.) Without knowing something about the DNA of (QA11) William, there is nothing the present DNA results can tell you about which William went to North Carolina.