Let's Just Assume . . .

Setting aside all those comments which can be made about the risks of assuming, let's recall for a moment the scientific necessity of the willingness to formulate hypotheses. If we can't test an educated guess, how will we ever know whether it's true—or false?

In order to figure out who these other Rileys were, whose names were swirling around the one-page document drawn up at the end of our William F. Riley's life in 1906, means we need to formulate a reasonable hypothesis of who they might be, then devise a way to test it. It seems possible that one might have been a grandchild—possibly even a married grandchild, thus doubly hiding the Riley surname in her roots. Thus, our task is to figure out just how that Mary Riley and Bettie Taylor, mentioned in the Soldiers Home records, might be related to William F. Riley—if at all.

To test that hypothesis means we need to build a tree, starting from the Riley child we know, and moving forward in time to her children, thus William Riley's grandchildren. If any scenarios match—or, conversely, do not match—the names we have listed in William's Tennessee Soldiers' Home records, then we have our answer.

To start, let's revisit what we know about William's daughter Mary. She it was for whom a newspaper transcription was attached to her Ancestry.com record by another subscriber, claiming she was involved in a paternity suit. Whether that was, indeed, our Mary Riley, we may have to ferret out. But for now, let's just stick to what we can find about this Mary.

Finding records on the right Mary Riley can be a challenge, considering such a common name, so I'll stick close to the record trail, especially from the start, when connections with the parents we already know are more apparent. Thus, we start with her father William's first census record after his 1852 marriage to Eliza Jane Thompson in Washington County, Tennessee. In 1860, that was the Riley household in nearby Greene County, where daughter "Mary E." was now six years of age.

By the time of the 1870 census, Mary was predictably ten years older, but her family had moved from Tennessee to Putnam County, Indiana.

It was in March of 1871 that the news article appeared in a Greencastle, Indiana, newspaper, alleging Mary was the claimant in a paternity suit, but until I can verify that report by accessing it myself, my only recourse is to fast forward to the next census record and see if there was a child in Mary's household who was born around that time. However, when we access the 1880 census, we find Mary in a home with two daughters—yet neither of them was born in 1871.

Mary, by this time, has married local shoemaker John Shellenbarger—on June 1, 1872, the same day as her sister Rachel married William Taylor Froggett—and continued to live in Indiana at least through her appearance in the 1930 census. Even if she did return to Tennessee to be with her mother at the passing of her father William, she would not have been referred to as Mary Riley; her name by that point would have been Mary Shellenbarger.

If both of William's daughters—Mary and Rachel—were out of the running for being the true identity of either the Mary Riley or Bettie Taylor named in the Tennessee document, who could those two Tennessee entities be? Neither of Mary Shellenbarger's daughters would fit; their names were wrong. And Mary's sister Rachel didn't seem to have any daughters to fill the parameters for the identity of Bettie Taylor, either.

There was one tantalizing other bit of evidence, however: that only Bettie Taylor in the 1900 census, back in Washington County, Tennessee, just happened to be born in 1871, herself.