The baby in the park: a genetic mystery solved?
Some names have been changed below; some have not.
As previously mentioned, I've been getting a lot of useful distraction from the Ancestry.com website of late. Back in December I reported that I'd had a call with a lady whose mother was found abandoned in a Philadelphia park as a three-week-old baby, in 1917. DNA suggests that she and her siblings are my fourth cousins or thereabouts. As it happens, my American grandmother was born in Philadelphia in 1899, but the genetic distance is too great for the baby to have been in her immediate family, or even first cousins; and she had loads of second cousins, never mind more distant possibilities.
To give a fuller picture, here are my top seven hits from Ancestry.com, with all names but one blanked out. These are Ancestry.com users who have submitted their DNA and who match with mine. For six of the seven, I had enough genealogical information to confidently identify them. None is closer than second cousin once removed, ie (in both cases) my great-great-grandparents are their great-grandparents, my great-grandparent is their grandparent's sibling, my grandparent is their parent's first cousin, and my parent is their second cousin. Those coded yellow are descended from at least one of the parents of Rebecca Hibbard née Wickersham, my American grandmother's mother, who died in childbirth in 1905. Those coded orange are descended from the parents of Jean Stewart née McElroy, my Irish grandmother's mother (who I remember well; she lived until 1985). Like most people, I have eight pairs of great-great-grandparents, but only the Wickershams and McElroys seem to have direct descendants on Ancestry.com.
So "Bella" stood out. It became clear after some back and forth that her brother "Derek" and sister "Patricia" are also linked to me on Ancestry.com. Patricia and I share about half the DNA that I share with Bella, and Derek and I half that again; our relationship could be anything from third cousin to much more distant. It was their mother who was found abandoned in a Phildelphia park as a baby. That story is told in a newspaper article from 25 August 1917:
Bella, Patricia and Derek, all now in their 70s, knew nothing more about their mother's origins, and were somewhat frustrated by the DNA results that they got and also by not always getting hugely helpful information from others who they had contacted on the site. I corresponded back and forth quite a lot with Patricia, and with her friend Susan who was doing some of the research on the ground (if hampered by the pandemic situation).
I spent some time thinking about it, and eventually sent over a list of forty Ancestry.com users who I know I am related to through my American grandmother. Patricia and Susan, god bless them, cranked through my list and found nothing at all. Not a single one of my grandmother's forty identified DNA relatives had also a DNA link with Patricia.
This was disappointing but really not so surprising. I get the sense that some bits of DNA are more "sticky" than others; you may get quite a large lump from a distant ancestor, you may get nothing at all from a closer ancestor. I have identified genealogical links with tenth and eleventh cousins with whom I share scraps of genetic material from mutual ancestors born in the 16th century. On the other hand, I also have a known third cousin with whom I apparently share no DNA at all - apparently the chance of this is around 10%, and we are both genetically linked to other known relatives, so it's not like there has been any messing with the records.
Patricia then gave me access to her own records and requested that I just try anything that might seem to work. The Ancestry.com search interface is somewhat frustrating, and also must respect privacy; after a couple of false starts, I tried inputting the surnames of my grandmother's great-grandparents and seeing how many hits that got from Patricia's DNA matches. If there were a significant number of hits, I tried to link them via genealogy to my own ancestors with those surnames.
I was pleasantly surprised to find that this actually did produce results. My American 5xgreat-grandparents' surnames (my grandmother's great-grandparents) were Hibbard, Charlton, Smith, Locke, Wickersham, Shallcross, Belt and Bordley. Smith is obviously useless, with far too many hits to tell us anything; of the others, only Hibbard and Belt pulled up anything resembling a decent number of credible hits, which meant I could provisionally eliminate the other five. Encouraged, I tried the surnames of the mothers of my Hibbard, Smith and Belt 5xgreat-grandparents - respectively Talcott, Whitehouse and Dulany. Only Talcott produced results of the same strength as I had from Hibbard. When I tried the next generation back, the maternal grandmothers of my Hibbard 5xgreat-grandfather, I again got very good results for their surnames, Leavens and Lyman.
So I felt pretty sure that Bella, Patricia and Derek were all descended from my Hibbard/Talcott ancestors, who were born in Connecticut in the 1750s and died in Vermont (her in 1831, him in 1845). The Hibbard family actually have an official genealogy published in 1901, which made things a little easier.
Er, yeah. That's 13 children, and even knocking off the one who died young, and the one I'm descended from (Lyman, the third son), that still leaves eleven. (Research revealed that Electa did in fact marry and have children, contra this record.)
I tried the same trick as before, inputting the surname of each child's spouse into Patricia's records to see which produced the best hits. Again, this weeded out quite a lot of them, leaving only two that seemed particularly promising. I tried again for those children's spouses, and unhelpfully both of them had results that looked equally plausible, neither more than the other. But then I looked at the next generation, and things became clear. One of the lineages I was following had moved to California, and never came back. The other family had settled in a New England town which I will call Hilltown, about 300 miles from Philadelphia and about 150 miles from Concord, VT. The only descendant of the right age to have fathered a baby in 1917 was a travelling salesman, born around 1870, who I will call Bill. It's not impossible of course that someone based in California could have fathered a child in Philadelphia, but a travelling salesman living only a couple of states away seems a much better bet.
Bill, the travelling salesman from Hilltown, New England, would have been my grandmother's third cousin. They probably did not know of each other's existence. (How many of your third cousins do you know about?) When the baby in the park was conceived in 1916, Bill had a wife and three young sons back in Hilltown, the kids all fourth cousins of my father's. It's entirely possible that he died in 1942 unaware that he had a daughter. His sons all married in due course, and some of their children may still be living; if I am right, Bill's grandchildren, all born with his surname, are half first cousins to Patricia and her siblings, and all of them are fifth cousins to me.
Edited to add: I was seriously off track here. Later research revealed a much better answer.
Having got this far, I then had a look at Patricia's other DNA hits to see if anything else could be learned. She has a lot more close relatives on Ancestry.com than I do - starting of course with her siblings Bella and Derek, and then another five who are all genetically her second cousin or closer. The top two of those five, I quickly realised, were both descended from a couple who I will call Hugh and Peggy, both born in the 1890s, who married in 1919 in Philadelphia. The other three were all related to Peggy but not to Hugh. It seemed pretty clear to me. The baby in the park's mother was certainly Peggy. The baby's father was definitely not Hugh.
Peggy's family lived less than a mile from Fairmount Park in Phildelphia, where the baby was found. She is recorded as being a professional musician in the 1910 and 1920 censuses. She and Hugh appear to have had a baby together in 1916, but did not get married until he returned from the war in 1919. Their marriage did not last, and the 1930 census records that Peggy and their child were living in Philadelphia while Hugh was living with a new wife on the West Coast. Hugh died in the 1930s, and Peggy successfully applied for a pension as his widow, with dependent child, from the Veterans' Administration, suggesting that their divorce, and Hugh's other relationship, were never formalised. As noted above, several of Patricia's DNA connections are descended from their child born in 1916; they had no other children together. (Actually I have no genetic proof that the 1916 baby's father was Hugh, except that he seeems to have acknowledged his own paternity.)
Reading between the lines, I speculate that Peggy and Hugh had split up around the time that their child was born in the first half of 1916, and somehow she and Bill got together - perhaps only once, perhaps more - towards the end of the year, with the August 1917 baby in the park as a result. But by the time the baby was born, Peggy and Hugh had reconciled. Hugh had just been drafted for the war, and the new baby was surplus to the requirements of the rekindled relationship. So Peggy took a sad walk to the park that warm August evening. (Or possibly her mother did, if the reports of the older woman in the area are correct.) I find this really heart-breaking: she gave up her baby to a completely uncertain future, for the sake of a relationship which had already failed once, and was destined to fail again.
The note left with the baby said that "The mother died at childbirth at the age of 22. The father, a professional singer, travels, but has now gone to the war." If I am right, the only true fact here is that the father travelled for a living. The mother had not died, was 27 rather than 22, and it was she who worked as an professional entertainer. The baby's father was too old to be drafted; it was the mother's fiancée who was just about to go to the war. To quote G.K. Chesterton's Father Brown:“Suppose someone sent you to find a house with a green door and a blue blind, with a front garden but no back garden, with a dog but no cat, and where they drank coffee but not tea. You would say if you found no such house that it was all made up. But I say no. I say if you found a house where the door was blue and the blind green, where there was a back garden and no front garden, where cats were common and dogs instantly shot, where tea was drunk in quarts and coffee forbidden - then you would know you had found the house.”
The story ends with a twist. The baby in the park was adopted, grew up, married her childhood sweetheart, whose parents were of Jewish and Scottish ancestry (which is rather helpful in distinguishing his genetic relatives from hers), and had three children, Bella, Patricia and Derek. Eventually she and her husband retired, after a life mainly spent in Illinois where he was a lecturer. And here's a truly extraordinary coincidence: the place they chose for their retirement, and where both eventually died, was her biological father's home - Hilltown, New England (whose population is only 12,000, on a good day). Bill had been dead for years at this point, but one of his sons was still living elsewhere in the state. I wonder if he or any of his brothers ever returned to Hilltown and unknowingly bumped into their half-sister? Edited to add: As previously mentioned, I was way off here, and the Hilltown relatives were second cousins not siblings to the baby in the park. Still, they may well have unwittingly bumped into each other.
One final reflection: the available genetic information can be somewhat hit-and-miss. As noted above, of my top seven DNA matches in the system, three are descended from one of my eight pairs of great-great-grandparents and three from another (one of those three has a different great-great-grandmother to me and the other two). I have not identified any Ancestry.com user descended from any of my other twelve great-great-grandparents. (I've had more luck with 23andMe in this regard.) Of Patricia's hits, the relatives of the mother of the baby in the park are very much more strongly represented than the family of the baby's father, or Patricia's own Scottish-Jewish father. So I think the warning for anyone else hoping to resolve their family past through Ancestry.com is that you may not be as lucky as Patricia. But I'd be happy to try and help.
As previously mentioned, I've been getting a lot of useful distraction from the Ancestry.com website of late. Back in December I reported that I'd had a call with a lady whose mother was found abandoned in a Philadelphia park as a three-week-old baby, in 1917. DNA suggests that she and her siblings are my fourth cousins or thereabouts. As it happens, my American grandmother was born in Philadelphia in 1899, but the genetic distance is too great for the baby to have been in her immediate family, or even first cousins; and she had loads of second cousins, never mind more distant possibilities.
To give a fuller picture, here are my top seven hits from Ancestry.com, with all names but one blanked out. These are Ancestry.com users who have submitted their DNA and who match with mine. For six of the seven, I had enough genealogical information to confidently identify them. None is closer than second cousin once removed, ie (in both cases) my great-great-grandparents are their great-grandparents, my great-grandparent is their grandparent's sibling, my grandparent is their parent's first cousin, and my parent is their second cousin. Those coded yellow are descended from at least one of the parents of Rebecca Hibbard née Wickersham, my American grandmother's mother, who died in childbirth in 1905. Those coded orange are descended from the parents of Jean Stewart née McElroy, my Irish grandmother's mother (who I remember well; she lived until 1985). Like most people, I have eight pairs of great-great-grandparents, but only the Wickershams and McElroys seem to have direct descendants on Ancestry.com.
So "Bella" stood out. It became clear after some back and forth that her brother "Derek" and sister "Patricia" are also linked to me on Ancestry.com. Patricia and I share about half the DNA that I share with Bella, and Derek and I half that again; our relationship could be anything from third cousin to much more distant. It was their mother who was found abandoned in a Phildelphia park as a baby. That story is told in a newspaper article from 25 August 1917:
Bella, Patricia and Derek, all now in their 70s, knew nothing more about their mother's origins, and were somewhat frustrated by the DNA results that they got and also by not always getting hugely helpful information from others who they had contacted on the site. I corresponded back and forth quite a lot with Patricia, and with her friend Susan who was doing some of the research on the ground (if hampered by the pandemic situation).
I spent some time thinking about it, and eventually sent over a list of forty Ancestry.com users who I know I am related to through my American grandmother. Patricia and Susan, god bless them, cranked through my list and found nothing at all. Not a single one of my grandmother's forty identified DNA relatives had also a DNA link with Patricia.
This was disappointing but really not so surprising. I get the sense that some bits of DNA are more "sticky" than others; you may get quite a large lump from a distant ancestor, you may get nothing at all from a closer ancestor. I have identified genealogical links with tenth and eleventh cousins with whom I share scraps of genetic material from mutual ancestors born in the 16th century. On the other hand, I also have a known third cousin with whom I apparently share no DNA at all - apparently the chance of this is around 10%, and we are both genetically linked to other known relatives, so it's not like there has been any messing with the records.
Patricia then gave me access to her own records and requested that I just try anything that might seem to work. The Ancestry.com search interface is somewhat frustrating, and also must respect privacy; after a couple of false starts, I tried inputting the surnames of my grandmother's great-grandparents and seeing how many hits that got from Patricia's DNA matches. If there were a significant number of hits, I tried to link them via genealogy to my own ancestors with those surnames.
I was pleasantly surprised to find that this actually did produce results. My American 5xgreat-grandparents' surnames (my grandmother's great-grandparents) were Hibbard, Charlton, Smith, Locke, Wickersham, Shallcross, Belt and Bordley. Smith is obviously useless, with far too many hits to tell us anything; of the others, only Hibbard and Belt pulled up anything resembling a decent number of credible hits, which meant I could provisionally eliminate the other five. Encouraged, I tried the surnames of the mothers of my Hibbard, Smith and Belt 5xgreat-grandparents - respectively Talcott, Whitehouse and Dulany. Only Talcott produced results of the same strength as I had from Hibbard. When I tried the next generation back, the maternal grandmothers of my Hibbard 5xgreat-grandfather, I again got very good results for their surnames, Leavens and Lyman.
So I felt pretty sure that Bella, Patricia and Derek were all descended from my Hibbard/Talcott ancestors, who were born in Connecticut in the 1750s and died in Vermont (her in 1831, him in 1845). The Hibbard family actually have an official genealogy published in 1901, which made things a little easier.
Er, yeah. That's 13 children, and even knocking off the one who died young, and the one I'm descended from (Lyman, the third son), that still leaves eleven. (Research revealed that Electa did in fact marry and have children, contra this record.)
I tried the same trick as before, inputting the surname of each child's spouse into Patricia's records to see which produced the best hits. Again, this weeded out quite a lot of them, leaving only two that seemed particularly promising. I tried again for those children's spouses, and unhelpfully both of them had results that looked equally plausible, neither more than the other. But then I looked at the next generation, and things became clear. One of the lineages I was following had moved to California, and never came back. The other family had settled in a New England town which I will call Hilltown, about 300 miles from Philadelphia and about 150 miles from Concord, VT. The only descendant of the right age to have fathered a baby in 1917 was a travelling salesman, born around 1870, who I will call Bill. It's not impossible of course that someone based in California could have fathered a child in Philadelphia, but a travelling salesman living only a couple of states away seems a much better bet.
Bill, the travelling salesman from Hilltown, New England, would have been my grandmother's third cousin. They probably did not know of each other's existence. (How many of your third cousins do you know about?) When the baby in the park was conceived in 1916, Bill had a wife and three young sons back in Hilltown, the kids all fourth cousins of my father's. It's entirely possible that he died in 1942 unaware that he had a daughter. His sons all married in due course, and some of their children may still be living; if I am right, Bill's grandchildren, all born with his surname, are half first cousins to Patricia and her siblings, and all of them are fifth cousins to me.
Edited to add: I was seriously off track here. Later research revealed a much better answer.
Having got this far, I then had a look at Patricia's other DNA hits to see if anything else could be learned. She has a lot more close relatives on Ancestry.com than I do - starting of course with her siblings Bella and Derek, and then another five who are all genetically her second cousin or closer. The top two of those five, I quickly realised, were both descended from a couple who I will call Hugh and Peggy, both born in the 1890s, who married in 1919 in Philadelphia. The other three were all related to Peggy but not to Hugh. It seemed pretty clear to me. The baby in the park's mother was certainly Peggy. The baby's father was definitely not Hugh.
Peggy's family lived less than a mile from Fairmount Park in Phildelphia, where the baby was found. She is recorded as being a professional musician in the 1910 and 1920 censuses. She and Hugh appear to have had a baby together in 1916, but did not get married until he returned from the war in 1919. Their marriage did not last, and the 1930 census records that Peggy and their child were living in Philadelphia while Hugh was living with a new wife on the West Coast. Hugh died in the 1930s, and Peggy successfully applied for a pension as his widow, with dependent child, from the Veterans' Administration, suggesting that their divorce, and Hugh's other relationship, were never formalised. As noted above, several of Patricia's DNA connections are descended from their child born in 1916; they had no other children together. (Actually I have no genetic proof that the 1916 baby's father was Hugh, except that he seeems to have acknowledged his own paternity.)
Reading between the lines, I speculate that Peggy and Hugh had split up around the time that their child was born in the first half of 1916, and somehow she and Bill got together - perhaps only once, perhaps more - towards the end of the year, with the August 1917 baby in the park as a result. But by the time the baby was born, Peggy and Hugh had reconciled. Hugh had just been drafted for the war, and the new baby was surplus to the requirements of the rekindled relationship. So Peggy took a sad walk to the park that warm August evening. (Or possibly her mother did, if the reports of the older woman in the area are correct.) I find this really heart-breaking: she gave up her baby to a completely uncertain future, for the sake of a relationship which had already failed once, and was destined to fail again.
The note left with the baby said that "The mother died at childbirth at the age of 22. The father, a professional singer, travels, but has now gone to the war." If I am right, the only true fact here is that the father travelled for a living. The mother had not died, was 27 rather than 22, and it was she who worked as an professional entertainer. The baby's father was too old to be drafted; it was the mother's fiancée who was just about to go to the war. To quote G.K. Chesterton's Father Brown:“Suppose someone sent you to find a house with a green door and a blue blind, with a front garden but no back garden, with a dog but no cat, and where they drank coffee but not tea. You would say if you found no such house that it was all made up. But I say no. I say if you found a house where the door was blue and the blind green, where there was a back garden and no front garden, where cats were common and dogs instantly shot, where tea was drunk in quarts and coffee forbidden - then you would know you had found the house.”
The story ends with a twist. The baby in the park was adopted, grew up, married her childhood sweetheart, whose parents were of Jewish and Scottish ancestry (which is rather helpful in distinguishing his genetic relatives from hers), and had three children, Bella, Patricia and Derek. Eventually she and her husband retired, after a life mainly spent in Illinois where he was a lecturer. And here's a truly extraordinary coincidence: the place they chose for their retirement, and where both eventually died, was her biological father's home - Hilltown, New England (whose population is only 12,000, on a good day). Bill had been dead for years at this point, but one of his sons was still living elsewhere in the state. I wonder if he or any of his brothers ever returned to Hilltown and unknowingly bumped into their half-sister? Edited to add: As previously mentioned, I was way off here, and the Hilltown relatives were second cousins not siblings to the baby in the park. Still, they may well have unwittingly bumped into each other.
One final reflection: the available genetic information can be somewhat hit-and-miss. As noted above, of my top seven DNA matches in the system, three are descended from one of my eight pairs of great-great-grandparents and three from another (one of those three has a different great-great-grandmother to me and the other two). I have not identified any Ancestry.com user descended from any of my other twelve great-great-grandparents. (I've had more luck with 23andMe in this regard.) Of Patricia's hits, the relatives of the mother of the baby in the park are very much more strongly represented than the family of the baby's father, or Patricia's own Scottish-Jewish father. So I think the warning for anyone else hoping to resolve their family past through Ancestry.com is that you may not be as lucky as Patricia. But I'd be happy to try and help.