Mr. Carkeet's analysis used a variety of sources and collections in an attempt to establish a supporting paper trail or find a supporting nugget to augment the book. I think a singular mention or a 'brief item" in a Detroit newspaper or an indisputable reference in a letter or diary would have altered his conclusion and allowed for further investigation. 

Here in Washington it has long been mentioned a nephew of Twain's purchased the Halycon House in Georgetown --> https://en.m.wikipedia.org/wiki/Halcyon_House

Tour guides, real estate agents and newspaper articles have long reported it.

Much less attention toward his real connection to Thomas Nelson Page or other Washingtonians.

Sent from my iPhone

> On Jan 9, 2017, at 10:46 AM, Shoshana Bailey <[log in to unmask]> wrote:
> 
> I have never answered Carkeet's conclusions and I'm not now but=20
> I will say one thing. He based a lot of his conclusion on the assumption tha=
> t I had put up a family tree and that is how I came up with the Clemens/Lang=
> donlinks. Now Barbara is addressing that same issue.=20
> I had NOT attached a tree to my DNA for the first several years. The DNA mat=
> ches I got with Clemens and Langdon were sent to me organically by Ancestry a=
> nd FTDNA (who, at the time, didn't have trees.)
> It was only after getting over 100 matches that I attached a tree.=20
> Regards,
> Susan Bailey
> 
> Sent from my iPhone
> 
>> On Jan 9, 2017, at 10:26 AM, Barbara Schmidt <[log in to unmask]> wrote:=
> 
>> =20
>> Hal --
>> =20
>> David Carkeet's research on the flawed methodology used to establish DNA
>> connections to Clemens via ancestry.com family trees was the game changer.=
> 
>> His essay on his months long research and first hand experiences in workin=
> g
>> with the DNA reports is online at:
>> =20
>> http://www.twainquotes.com/Carkeet/AncestryReport.html
>> =20
>> In a nutshell, anyone can plug in a false or inaccurate family tree at
>> ancestry.com after submitting a DNA test, and the ancestry.com database
>> will generate a list of people who are also related to people in the tree
>> one believes is their own tree.  Just about anyone can be distantly relate=
> d
>> to someone else who ties in to the Clemens family tree in some fashion if
>> the family tree branches are traced back far enough.  =46rom Carkeet's rep=
> ort:
>> =20
>> "... 'What you are seeing is the result of endogamy, intermarriage within a=
> 
>> population group. In the year 1700, the population of the United States wa=
> s
>> approximately 250,000 people.' In other words, it is a small world. Or,
>> more to the point, it was a small world, with so few people in the pocket
>> of immigrant history that I share with my matches that if I select a
>> specific name from the past (in this case, with my false genealogy, an
>> ancestor of Samuel Clemens), some match of mine from among my more than
>> 6,000 matches will descend from that ancestor."
>> =20
>> Barb