The only conclusion that I came to is that not all Farnsworth's are related based on DNA test results of seven members of the Farnsworth DNA Surname Project on familytreedna.com.Three members match exactly 25 out of 25 markers which indicates a 99.12% probability that they share the same ancestor within the last 20 generations.Those same three persons only match 12 or 13 out of 25 markers with the other members which means that they are not related at all.If you have 6 or more mismatches out of 25 markers than you are totally unrelated. This is fact and not a conclusion on my part as you have described it. There are reasons that the DNA Analysis reflects that not all Farnsworth's are related and one of them does happen to be adoptions.Another reason is that children were born out of wedlock and carried the surname of the mother.I am beginning to believe that this possibility exists within my own family since I have hit a stone wall with my 3g grandfather. I believe the main reason for the DNA mismatches is how the surname Farnworth or Farnsworth was originated in the 1300 and 1400's.The surnames were taken from basically from four categories: 1. First Names( William, the son of John became William Johnson.) 2. Occupational(John the Cook became John Cook) 3.Location or topography(Richard Hill or Woods or Farnworth) 4. Nicknames.(William Long or Short or Curley). Note:Farnworth is the name of a couple of communities in England where some think that is where the Farnsworth surname may have originated from. I think it would be interesting to have the DNA results from an ancestor from the 1500.s or older to compare with today's results.However I don't think that it would result in 900 distinct lines out of 1000. David, why don't join the DNA Surname Project? We have one member in the group who has traced his family to Matthias of Groton,Ma and another who has traced his family to Thomas of Bordentown, NJ.Wouldn't it be interesting to see which line you fall into if you don't already know?