Roger Ellis Mystery

Another approach to genealogy is through DNA. The “Gold Standard” of DNA testing is Y-chromosome or male line DNA. This is passed nearly unchanged from father to son. All male direct line descendants of a male (same surname) should have very similar Y chromosome markers. There is an Ellis DNA Project that tries to sort out the many Ellis families lines. Unfortunately, we do not have a male Ellis at this point that meets the criteria and who has agreed to the test. If you are an Ellis male that meets the criteria for the Y-DNA test, please consider submitting a cheek swap or saliva sample for Y-chromosome testing. Your identity can be kept private. You will get email addresses of a short list of likely matches or the administrators of the samples.

Another DNA test is the autosomal DNA test now offered by Ancestry.com and others. It compares 1000’s of DNA segments samples from all chromosomes, so it can help identify any of a persons distant cousins that have taken the test. One advantage of this test is that anyone (male or female) can take the test. With this test, you will get a very large list of possible matches, ranked by how close the match is. You will also get links to their family tree if they have a public tree on Ancestry, with notifications if a common ancestor is found between your tree and one of your DNA matches. This test is more of a shot-gun instead of the rifle shot of the DNA-Y test. From experience to date, matches of 5th or 6th cousins (with common ancestor born after around 1700) are reasonably certain. More distant than that, it becomes likely that a match could be coincidental.

When one member of our family did the test, when they included Joseph Ellis of Henrico and Elizabeth Perkins as ancestors in their family tree. This person found about 6 match that also showed a Constantine Perkins born 1682 in Henrico Co. Va. as an ancestor. At first it was believed this confirmed this part of the Ellis tree, but this is less than certain. The person’s mother’s DNA tested. She also had 4 of the matches. However, the family trees of these matches also included another one of the mother’s direct ancestors. It looks like a couple of the matches could be on the father’s side, so the test really is not giving a solid conclusion. The Perkins family is a large and well documented family, so it looks like many family trees include Constantine Perkins. It seems that if you look over a thousand family trees at random, it is likely that you will find Constantine, or any other of the early Virginia patriarchs. They could have millions of descendants. While not really conclusive, the matches seem to support the Tuckahoe connection. Even if the connection is not through Elizabeth Perkins and Joseph Ellis, there seem to be many connections between the Perkins and Ellis families. There is even a Col. Charles Ellis Perkins born 1742. His father, Nicholas Perkins opened a ferry across the Dan River in Halifax Co. (later Pittsylvania Co. Va.) around 1758. The Perkins family was one of the largest slaveholding families of the area. Elizabeth did not marry Joseph Ellis until 1745, so it why was a child born three years earlier be given the middle name Ellis? Was there another marriage between a Perkins and Ellis for which no record has been found? There was a group of Perkins in Philadelphia (including a gunsmith that became superintendent of the Harper’s Ferry Arsenal) , but we don’t know if they were related.

Another member of the family found that the line of Jeremiah Ellis of Surry, Virginia shows up in a number of DNA matches. A second line that matched is the family from Washington County, Pennsylvania that moved to Kentucky. That family seems to tie back to Ellis Ellis that lived in Pennsylvania and died in Virginia. Several records seemed to be tied to this line although not all were tied to Washington County, Pennsylvania. Ellis Ellis was tied to the Philadelphia area. More analysis is required of the DNA data to determine if it is possible to pinpoint the correct Ellis ancestry via DNA.