There are some times when we think we have identified details about the right person in our family tree—but not enough to be sure beyond a doubt that we have figured out the right connection. So it was with George Edmund McClellan's frustrating widow Celestia, the one who nearly took the money and ran, after his passing in Wellborn, Florida, in 1866.
You'd think a name like Celestia R. Holman would be rare enough to not suffer the genealogist's plague of two possible candidates for the same identity. Yet, trying to discover just where George's widow originated, I ran into not one, but two Celestia Holmans of about the same age. One showed up in the 1850 census in Chautauqua County, New York. The other one was living in the household of someone named Rhodolphus Richards in Macomb County, Michigan.
If you think a name as unusual as Rhodolphus Richards would merit a fast pass to my answer, think again. Rhodolphus was about as difficult to find in other records as was Celestia.
There was one other detail that piqued my curiosity. Rhodolphus, it turned out, had a wife named Aurilla. And, as we had already observed from Celestia's 1904 death certificate, Celestia's mother happened to be named Orrilla.
What were the chances? Could Aurilla Richards be Orrilla Grover? Looking at that 1850 census, Aurilla Richards was forty six years of age, a quite reasonable twenty nine years older than Celestia.
There was another reason I thought this scenario might be possible—and no, it wasn't mere wishful thinking. In researching Celestia's stated parents, I discovered her father, Levi Sawyer Holman, had died young, in 1834. Certainly, his equally young widow would need to marry again.
Researching further, I discovered the origin of Celestia's middle name—Relief—which she no doubt received, thanks to her paternal grandmother, Relief Sawyer.
Finding an online source for marriage records of the time—and then hoping they contained the appropriate information on the bride's former name—was a nice thought, but an impractical task, considering not only the many possible spelling permutations involved in an Aurilla Holman/Rhodolphus Richards marriage, but also the lack of knowledge of which state in which to locate the marriage records. After all, if Aurilla Richards was really Orrilla Grover, she could have married her husband in any of the states leading to Michigan from their origin in New England.
Where there isn't a direct route to our answer, there are side roads leading to more clues. One, for instance, involved one of the Richards children also listed in that 1850 census, the Richards' six year old daughter, also named Aurilla. It turns out, the younger Aurilla married well, earning both her husband as well as herself a brief biography in one of those local history books. The biography chose to use the spelling of Orrilla, and listed her parents as Rodolphus and Orrilla, encasing the name Holman in parentheses as if to indicate it was the elder Orrilla's maiden name.
But if that left enough doubt in the air, thankfully another discovery helped quell concerns. When Orrilla Richards died in 1894 at the age of ninety, her headstone helpfully included this information:
Orrilla Grover Richards
Oct. 4, 1803—Mar. 5, 1894
Underneath her entry was that of her second husband, Rodolphus S. Richards, confirming for us the connection between Celestia's mother and this Orrilla Richards.
While I am not sure how a fairly successful farmer in Michigan would have the means to send his step-daughter to the first college in the country to accept women as students—let alone have the connections to marry off said step-daughter to a well-to-do widower in Florida—at least now I know how Celestia connected to the place she called home as a girl. Whether we ever uncover the story of how she met George McClellan in Florida, I don't know, but at least we now have a clue what drew her back to the midwest from the disaster following her married years in Florida.