Buckle up! Because this is a bit of a long one – but an interesting one! It is a case study of sorts in reviewing bad trees online, following the evidence without presumption, and allowing for thinking a bit outside of the box. First, let me mention that there is no shame at all to the member who brought this to our attention or any previous researcher who has worked on these lines. These are very easy and common mistakes to make and are all part of the learning process while we navigate through these fascinating new waters of our individual genealogies.
Yesterday I was asked by a member of Sheldon Genealogy to add his great-grandfather to our database, and was provided the information of a Charles W. Sheldon, born 1860 in Cross Keys, New Jersey – son of Charles W. Sheldon, born 8 Oct 1828 and died 15 Aug 1886. Sure enough, we did not yet have these Sheldon’s in our database, but finding information on them was quite easy as well as connecting them as descendants of Godfrey Sheldon of Maine. However, I immediately saw a problem.
Charles was listed with his parents in census records for 1860 (age 2 months), 1865, 1870, and 1880 – all in New Jersey. We then have a marriage record for Charles W. Sheldon Jr to Ella Emma Jennings, both of Cross Keys, on 13 Aug 1884. Ella sadly passed away in 1886 at only 25 years old after a 5 month illness of tuberculosis – no children are found. Charles then remarried to Mattie E. Ewan on 30 May 1887 in Gloucester County, New Jersey. This union would last a bit longer and civil records show a son, Ralph D. Sheldon, born in 1889. However, Mattie also died young in 1896 and Charles remarried for a third time to Emily Louise Tussey on 23 Nov 1898. In the 1900 census we then find Charles, Emily, and Ralph living together. Sadly Ralph also died young, at just 16, while working as a stock boy in Barrette, New Jersey. Ralph never had the opportunity to marry or have children. From this point on Charles is found alone with wife Emily in the census records for 1910, 1915, 1920, and 1930. He then died in Camden, New Jersey on 9 Apr 1933. But there was a clear problem, if our member was the great-grandson of this Charles W. Sheldon, where was the grand father that connected them? The only record of any child was Ralph, who died young.
So, I went back and asked for the name of the grand-father, – could there have been a hidden child somewhere? Did Ralph have a son when he was a teenager? The information I was now provided with was Harry Ray Sheldon, born 20 May 1900 in Harveys Lake, Luzerne, Pennsylvania, and died in April 1969 in Ambler, Pennsylvania. I could quickly see that Harry was, in fact, the son of Charles W. Sheldon. But, not the same Charles W. Sheldon. It was a simple and honest mistake of identity by a later researcher. This Charles W. Sheldon in Pennsylvania had no connection to the Charles W. Sheldon in New Jersey.
After some searching I found that this new Charles W. Sheldon was born 8 Aug 1870 (a decade younger than the one in NJ) in Lehman, Luzerne, Pennsylvania. He married Anna Elizabeth King on 25 May 1892 in Lehman. They had a long married life with ten children, only seven of which we can currently name: Mary (1892), Eurah (1895), George Dewey (1898), Harry Ray (1900), Ralph (1902), Blanche (1905), and Jennie (1907). The unnamed children apparently died young and, sadly, Charles died relatively young as well in 1907, at age 37, of lobar pneumonia. Anna had a hard time at first but found a boarder to help with an income into the household. The boarder, living with the family in the 1910 census, was Bert Smith. He and Anna later married. First mystery solved.
The next step was to find the origins of Charles W. Sheldon. In his death certificate his father was noted as a James Sheldon, born in Wisconsin, and Sarah King, born in Pennsylvania. I could not seem to find Charles in an 1880 or 1870 census, but I noted that Charles wasn’t born until Aug of 1870 and the census was taken in June. His marriage record said he was from Lehman, in Luzerne County, so it was in that area that we should find his family living in the 1870 census. That is exactly what I found.
In the 1870 census, at a little place called Dallas, Luzerne, Pennsylvania (neighboring Lehman), was a James M. Sheldon, 59, a stone mason born in New York, with Sarah M. Sheldon, age 29, born in Pennsylvania, and Ura Sheldon, age 3. Beside the names of the parents, James and Sarah, the inclusion of a daughter Ura is additional confirmation as Charles would later name one of his own daughter Eurah. When I first saw this record, something caught my eye. This was a new family with only a 3 year old daughter and a son quickly on the way. Sarah was on the older side for a newish bride, but she was still 30 years younger than her husband. In fact, the algorithm on Ancestry.com assumed that Sarah must have been a daughter. This big age gap would be important later.
So now we had the line back another generation, but who was this James M. Sheldon, and where was he born? His son would later say in census records, and as recorded on the previously mentioned death certificate, that James was born in Wisconsin. However, James himself gave New York as his place of birth in the 1870 census (see census above). And, to further complicate it a bit, when searching for the daughter Ura, who did later spell her name Eurah like her niece, I found that she would repeatedly say that her father was born in Pennsylvania. The ONLY other record that I could find for this James in Pennsylvania was a map of Lehman Township from 1873, showing his home at the far east corner of the township right on the boarder of Dallas. I could not find him later in the 1880 census, and I could not find him earlier in the 1850 or 1860 – when he should have been.
I began doing some broad searches to see if I could find any leads. I did find several online trees that said this James M. Sheldon was the same as a James Sheldon who was born 29 Sep 1818 in Wisconsin and died 20 Feb 1873 in Wisconsin. At first glance this individual may have seemed promising, since he was born in Wisconsin – but as usual, you should never trust online trees. There were two immediate red flags – the difference in age from our James, and the fact that this James appeared to live his entire life in Wisconsin. So how would he have a family in Pennsylvania? There was solid evidence that this James actually died in Wisconsin, including a grave with a headstone at Waterloo, Wisconsin for himself with his wife, “Sarah Babcock Sheldon”. So it’s easy to see that people confused James Sheldon and Sarah King of Pennsylvania with James Sheldon and Sarah Babcock of Wisconsin. To further confuse people the James and Sarah in Wisconsin had a daughter Euretta, which is very similar to Ura – but Euretta was born in 1851 while Ura was born in 1868. To further confirm that this was the wrong James, his will filed in Jefferson County, Wisconsin in 1873 listed all of his children with no Charles included.
I continued my search, using some more creative search criteria. An individual kept coming back around, another James in Wisconsin but this one was married to a Julia Sweet and appears to have a complete family. So I also wrote this one off until I noticed that someone had posted him on a tree on Geneanet with surprising death information. They named him as James Madison Sheldon, intriguing against our James M. Sheldon, and that he had died in 1873 – like the other James so immediately suspicious. His place of birth was listed as New York but all records as an adult were in Wisconsin, which would make sense of our New York / Wisconsin discrepancy. Also, this time they gave the place of death as Little Lake, Pennsylvania. Frustratingly there was no documentation of where this information came from, but intriguingly this tree hadn’t otherwise confused this James with our James which suggested that they obtained this information independently. All of the other information for him showed his life in Wisconsin, his marriage to Julia, their children, everything was the same except for Little Lake. So I looked up this place and it is located in Kidder Township, Carbon County, Pennsylvania. At first that may not sound interesting, but Carbon County, and Kidder Township specifically, border Luzerne County. Little Lake was less than 40 miles from Lehman. That could not be a coincidence.
Now I had a new problem. While this was intriguing, there was no documentation to support it. Further, a newspaper article of Julia’s probate noted that she died intestate on 11 Nov 1872 in Wisconsin, and I didn’t see a divorce record for her and James. Meanwhile, daughter Eurah in Pennsylvania was born in 1868, so James must have been with Sarah as early as that date – an overlap of at least four yeas from 1868-1872. This lead me to an idea that needed to be tested, and it would have to be proven in order to move forward:
Hypothesis: That one James Madison Sheldon first married Julia Sweet in Wisconsin and had several children in Columbia County. That in the mid 1860’s he left his family in Wisconsin, possibly due to the war, and resettled in Luzerne County, Pennsylvania. That in Pennsylvania he remarried to Sarah King (possibly bigamous) and had two additional children.
Test 1: Daughter Eurah is consistently recorded as born in Pennsylvania in 1868 or 1869, and James is listed with Sarah and Eurah in the 1870 census. As Wisconsin is substantially distant from Pennsylvania, it’s unlikely that he would be traveling regularly between both location. Therefore, James should not be listed with Julia in Wisconsin in the 1870 census.
Answer 1: Julia was found in Dekorra, Wisconsin on 29 Jul 1870. The only person living with her is her youngest child, son Ira S. Sheldon, age 20. This confirms that Julia and James had apparently separated before 1870.
Test 2: As James and Julia were separated by 1870, can we confirm that this James died in or near Luzerne County, Pennsylvania in or about 1873? Probate records may confirm property in multiple locations.
Answer 2: Unfortunately the probate records of Luzerne County, Pennsylvania did not show a record and those for Columbia County, Wisconsin are not available online. So I needed to resort to an alternative. In many cases, when lucky, probate cases were published in the local paper. Sure enough, on 28 Feb 1874 in the Wisconsin State Register is published a petition by Ira S. Sheldon asking for the administration of the estate of his father, James M. Sheldon, who “died intestate on or about the 8th day of March, 1873, at Wilkeabarre, Pa.”
This was the final needed proof. Wilkes-Barre (correct spelling), was the county seat of Luzerne County, Pennsylvania. This absolutely confirms that the James Madison Sheldon, husband of Julia Sweet (parents of Ira) in Columbia County, Wisconsin was the same James M. Sheldon in Luzerne County, Pennsylvania where he died in 1873. It’s also interesting to note that James died 8 March 1873, but his son did not petition for the administration of the estate until February of 1874, almost a year later. This could suggest that James was completely estranged from his family in Wisconsin and that they were not immediately aware of his death.
So just how much of a hidden scandal was this? Was James a bigamist? It’s hard to say for certain. There is no actual marriage record for he and Sarah in Pennsylvania (so far) – but then that would mean they had two children together out of wedlock. If they were married, there does not appear to be a divorce record for James from Julia in Wisconsin. Obtaining a divorce was an expensive, and sometimes nearly impossible, proposition – this combined with the fact that the communities in Pennsylvania would have no reason to question them, and no way to learn, or confirm, an earlier marriage. So bigamy is a very real possibility and there is no great scenario.
More work needs to be done to further document and understand the various families mentioned here, but I believe I can confidently say that this James had two families. Let this also be an extra warning to be reticent of online trees, even those that appear to be well sourced. It’s important to always review and ask yourself, upon each point of information – does this make sense? Does it work?