Jefferson's Daughters

by Catherine Kerrison

  From the outset, however, the simplicity of that plan is foiled by nineteenth-century record-keeping at both the federal and District levels. The national government did not mandate that the names of all persons in a household be entered in the federal census until 1850, rendering wife, children, other relatives, servants, slaves, and any other live-ins unidentified between 1790 and 1840. Nor did the city fathers order the registration of births in the District of Columbia until 1874. (Deaths were registered beginning in 1855.) What may have been a straightforward if somewhat tedious task of reading through the census records of 1830 and 1840 and the city’s birth rolls instantly becomes more complicated. Instead children’s names must be retrieved from baptismal records contained in scattered church records.

  This strategy presents several immediate problems. Not all Washingtonians were Christians or churchgoers, so before I even begin, the net that may have gathered Harriet’s children is pocked full of holes. Even of those who were, many were not baptized as infants; some Presbyterians, for example, required an adult confession of faith for baptism. Identifying the churches presents another challenge. The earliest city directories, in 1822 and 1827, listed the churches that existed in Washington within the most likely years of Harriet’s marriage, but most have since changed their names or locations or been absorbed into other congregations. In other words, one must first construct a genealogy of the churches to know where to look for any existing records, and then hope the records have survived the various transformations and moves over almost two centuries. Even if the records can be located, they would not have been kept with any consistency. Instead they reflect the idiosyncratic preferences of the particular clergyman with respect to what information he considered important to record, over what period of time, and with what standardization of spelling and legibility of penmanship. Only two of the original churches remaining in the city today possess records dating to 1822: First Presbyterian Church and St. John’s Episcopal Church. My searches in each church’s baptismal records for children with the distinctive Hemings family first names turn up empty.

  These few existing church records do not yield any likely brides named Harriet, either. But my hopes rise when I discover that, beginning in 1811, all marriages had to be registered with the District. Extracting the name of every Harriet married between 1822 and 1830 from Marriage Licenses of Washington D.C. 1811 Through 1830, I compile a list of fifty-eight. The marriage record lists the names of the bride and groom, their race (only if they were “black” or “mulatto”), and their wedding date—nothing more; no parents’ names or witnesses, or date and place of birth. There is not a single Hemings listed in the entire book. But fifty-eight Harriets constitute a much more discrete sample to research than an entire city population of thirteen thousand. More important, I learn the names of their husbands, whose tracks as heads of household, workers, taxpayers, property owners, and devisors of property are much easier to follow than those of women, who tend to disappear from the historical records once they marry.

  The most logical place to begin this work is the Daughters of the American Revolution library, which has collected voluminous genealogical records to help researchers determine whether their ancestors fought during the Revolutionary War. There one finds a database created from the work of the DAR’s Genealogical Record Committee (GRC). This early-twentieth-century project preserved by transcription a mass of early church, town, county, and court records, including wills. They even transcribed gravestone inscriptions. Entering the name of each Harriet (duplicating the entries with various spellings, such as Harriot, Harriott, and Harriett) and each husband into the database, and searching in the District of Columbia, Maryland, and Virginia, I am able to eliminate several possibilities. For example, baptism records at Rock Creek Church reveal that, in spite of her suggestive last name, Harriet Free had been baptized there in 1800, Harriet Higdon at Christ Church in May 1807, Harriet Dyer in the Presbyterian church in Alexandria in 1809, Harriot Hughes in February 1801, and Harriet O. Graves at Rock Creek Church in 1809. With their births and parents firmly established in the Washington area records, these Harriets could not have been Hemingses. Baptized at the First Baptist Church in 1831, Harriet Sales was a woman of color and so could also be eliminated.

  Marriage records taken from both church and District records in the GRC database are also useful. Although Harriet Narden was married in December 1823, a date well after Hemings had arrived in Washington, the GRC also showed that she had married previously in 1819, far too early to have been Hemings. Similarly, Harriet Wallice, who married William Ridgeway in June 1823, was almost certainly the widow of Richard D. Wallace, whom she had married in 1816. (The different spellings are not enough to raise a question; clerks frequently spelled phonetically.) Harriet Nicoll, who married in November 1824, had also been married previously, and was bringing so much property to her second marriage that she protected it with a marriage settlement (what today we would call a prenuptial contract).

  Other records show how vital are the details of Madison’s account in the search for his sister. For example, funerals recorded in the GRC database lead me to eliminate Harriet L. Cruttenden, who had married Dr. Hezekiah Magruder in 1830. She died in 1836, long before the period 1863–1864, when Madison Hemings last heard from his sister. When George Sweeny, onetime chief clerk of the post office, prepared his will in 1849, his wife, Harriet Burgess, was already dead. Elsewhere, an obituary in 1851 for her brother, Thomas Shields (age thirty-five), eliminates Harriet Ann Shields, who married Washington bookbinder James P. McKean in 1830. Even if Beverley had taken his father’s name, Thomas Shields was eighteen years younger than Beverley; nor, in 1873, had Madison given any hint that Beverley was dead.

  My list of Harriets narrows my search through the District’s probate records to find a will that may have devised property to children bearing the distinctive names of Harriet’s brothers. But only a handful of the couples married in the 1820s filed a will with the District of Columbia, a sobering reminder of the importance of both class and mobility in nineteenth-century America. Not everyone possessed sufficient goods to leave a will, nor had the capital lost its reputation as a city of transients, with its ebb and flow of hopeful job seekers with each changing administration. Although none of the wills point in Harriet Hemings’s direction, they do eliminate two more Harriets from my list. Harriet Castle’s husband, John Cromwell, died in 1835 with no children, an unlikely candidate for someone who “raised a family of children.” Probate records at the National Archives reveal that Harriet Bohrer was the sister of physician Benjamin Bohrer.

  A variety of other sources help me to eliminate more contenders. Less than eight months after his marriage, Martin Steel posted a notice in the National Intelligencer: “My wife, Harriet Steel, having eloped from my bed & board without any just provocation, this is to forewarn all persons from crediting her on my account, as I am determined to pay no debts of her contracting.” Harriet Tolliver Steel clearly did not follow the conventions of respectable white womanhood, much less “raise a family of children” that would indicate she could have been Harriet Hemings.

  Harriet Bell married William J. Cooper in 1826. Her last name draws my interest because it echoes a family tradition that Sally Hemings had received a bell from her dying half sister Martha Wayles Jefferson. Bell’s husband was an English-born printer, a rising and respectable occupation in the early republic. When he died in 1871, he was mourned as a man who “filled with honor several positions of trust under the former government of the city of Washington.” This sounds like the man Madison Hemings described as his sister’s husband, and I begin to hope that I might be on the right track. Cooper was visible in the way that counted most for nineteenth-century male identity: He was a skilled craftsman, a good provider, and civic-minded, as his official positions attested. In these ways, he seems to have been exactly the kind of citizen upon whom Jefferson pinned his hopes for the survival of the republic and an intriguing possibility
for Harriet Hemings’s husband. Although he named no children in his will, the 1850 census showed he had six. With the exception of one-year-old Jane, however, none bear names connected to the Monticello families, white or black. Still, it is only when I visit Congressional Cemetery that I’m forced to eliminate Harriet Bell. There, next to the stone marking the graves of Harriet Bell Cooper and her husband, is one for Harriet’s father, Charles Bell.

  Another intriguing early possibility is Harriet Cottringer, who married Robert Young Brent in 1824. Brent’s father had served as the city’s mayor during Jefferson’s and Madison’s administrations. Certainly, then, Brent fit Madison’s description of a “man in good standing in Washington City.” But a letter written in 1823 by Charles Francis Adams, grandson of John Adams, casts doubt on the likelihood that Harriet Cottringer could have been Harriet Hemings. Remarking that it was high time that Cottringer got married, “for she has passed the grand climacteric and is now going down hill” (she was twenty-four at the time), the sixteen-year-old Adams reflected that although he did not doubt that she would make a good wife, he would prefer “something like pleasure for so disagreeable a step” as marriage. “Excellence is good,” the young man concluded, “but it is not much without beauty.” Since Jefferson’s overseer, Edmund Bacon, had said that Hemings was quite beautiful, Cottringer was not likely to have been Hemings. Further research reveals that Harriet Cottringer was from a Philadelphia family; born there in 1799 and baptized in St. Mary’s Catholic Church, she could not have been Hemings.

  A more exciting possibility arises with the case of Harriet Walker, who married John Newton in May 1825. The last name is certainly suggestive: Martha’s daughter Ellen had spoken of the “three young men and one girl, who walked away,” and we know that in spite of Jefferson’s notations, neither Harriet nor Beverley had “run.” In addition, the District’s marriage records list a William B. Walker who married just two months after Harriet Walker did. Could he possibly be Beverley? William B. Walker was a coach painter, not entirely unlikely for Beverley; Jefferson’s valet, Burwell Colbert, was a talented coach painter as well. And the man who voluntarily endured almost three additional years of slavery to wait for his sister to turn twenty-one could well have waited a mere two months for her to marry before establishing his own household. But there are thirty-five other Walkers (excluding two identified as black) in the marriage register; if Harriet and William B. were related to any of them, they obviously were not Hemingses.

  More sleuthing in both church and census records rewards me by establishing a clear connection between John and Harriet Newton and William B. Walker. In October 1839, William’s wife, Maria, stood as sponsor for the baptism of John and Harriet’s infant daughter. The close relationship of Harriet Newton and her sister-in-law connects Harriet Walker and William B. Walker as brother and sister, raising my hopes further. But then the 1830 census finally eliminates Harriet Walker from contention: Although William B. Walker and his wife were white, John Newton was a free man of color, and Madison had been emphatic that his sister had married a white man.

  Aside from checking baptismal records, I am only able to identify most of my fifty-eight Harriets through the historical imprints of their husbands. However, a sizable minority, twenty of the fifty-eight men, seem to have disappeared from the landscape after their marriages were recorded by the District’s clerk. This is particularly puzzling given the advent of the Internet, which, coupled with a phenomenal rise of interest in family genealogy as a hobby, has made this kind of search possible in ways that it would not have been thirty years ago. Databases such as Ancestry.com and Family Search that make available both original records as well as individually uploaded family trees allow me to cast a net nationwide, rather than assuming that these husbands remained in the capital district. Even so, the trail stops cold for John Anchors, who married Harriet Ann Hess, Joseph Askins and Harriet Wilson, John Barry and Harriet Farland, and many more. Nor does my reading of Washington’s 1850 and 1860 census records reveal their presence. Their invisibility reminds us of the haphazard nature of record-keeping and record survival. In any event, remembering Olive Rebecca Bolling’s comment in the 1940s that Harriet Hemings’s family was “prosperous and lives right here now,” the invisibility of these men in the records makes me believe it less likely that any of the Harriets who married them was Harriet Hemings. Prosperous men own real estate, run businesses, and pay taxes—all of which would be discoverable in the records.

  Two last Harriets of my original fifty-eight still offer intriguing possibilities, in cases that are at once suggestive and questionable. Harriet Simpson married Rezin Pumphrey in January 1824. Pumphrey was a laborer who, even as late as the 1860 census, when he identified himself as a farmer, still had not managed to amass any real estate at all; his personal property amounted to only one hundred dollars. He had lived in Virginia, where, if we can rely on the distinctiveness of his name, the 1820 census seems to have counted him twice, in both Brooke and Ohio counties (now West Virginia). This is not necessarily to be wondered at, when it took eighteen months to compile the census, for it shows the imperative for landless men to be on the move to find work. It also suggests that he was not skilled in a viable trade. His name appears in the Washington newspapers twice, both times in reports of violence. A meeting with a John Long broke out into a fight in which Long assaulted Pumphrey and left him for dead in 1853; and Pumphrey was injured during a shooting at the Navy Yard in 1858.

  None of what we can glean about Rezin Pumphrey, then, comports with Madison’s description of a “man in good standing.” Not even the names of their ten children, which include John, Martha, Ann, and Mary—all names that could be found at Monticello during Harriet’s childhood—persuade me to take another look. Rather, along with their other children’s names (Sophia and Lloyd), they all echo the Pumphrey family line. There are also about two dozen other Simpsons in the District’s marriage register, and Simpson is a common name in Prince George’s and Montgomery counties in Maryland as well. It is likely that Harriet was related to one of those families; indeed, the wedding announcement of a Rezin S. Simpson in Baltimore in 1821 may indicate a tie of long standing between the two families long before Rezin and Harriet’s marriage.

  Nonetheless, one wonders. Could Olive Rebecca Bolling have been referring to the Pumphreys when she talked about the prosperity of Harriet’s descendants? By the 1940s, a Pumphrey family had been in the funeral and undertaking business in the capital district for over one hundred years. In 1928, they purchased a home in Rockville, Maryland, where their undertaking business still operates; and the Bethesda location on Wisconsin Avenue continues to serve as the business’s headquarters, as it has since 1934. In other words, this visible and thriving business, managed by a family with a distinctive name, could well explain why Bolling thought Hemings’s descendants were prosperous. However, the line of Pumphrey undertakers was not Rezin’s; rather, it descended from his brother William. The Pumphreys were a large and prolific family in the Washington-Baltimore corridor; Bolling could be forgiven if she confused one branch with another. It is possible, then, that Harriet Hemings married Rezin Pumphrey, and if Madison thought his sister had done well, it could be because she had exaggerated her success in her reports to her brother, either to cover her own discomfiture or to keep him from worrying about her.

  But given what we have learned about the successful lives of Hemings family descendants from Monticello historian Lucia Stanton and the Getting Word project, it seems unlikely that Harriet Hemings was so imprudent as to choose a luckless laborer to raise her fortunes. Or perhaps, I admit to myself as I consult my list for other possibilities to explore, I just want a happier ending for Harriet Hemings than she would have known as the wife of Rezin Pumphrey. Maybe that’s why I eagerly turn my efforts to a tantalizing new trail.

  Harriet Garner (as she appears in the marriage register) or Gardner (as reported in the newspaper’s wedding announcement) married Benjamin Williamson
on July 13, 1822. If Garner was Harriet Hemings, she made her decision about her future livelihood quickly but well. Williamson was a Scottish immigrant who had arrived in the city about four years earlier. His very ethnicity may have appealed to Harriet: As a Scotsman, he had no experience of or investment in the institution of slavery, nor is there any indication for the rest of his life that he ever owned a slave. As a recent migrant, he did not have family or connections who would have scrutinized those of a prospective wife. (Interestingly, there are several examples of Scotsmen who marry women of color in this period.) He was a carpenter by trade and so could have moved in Beverley’s orbit, explaining how he would have met Harriet so soon.

  Benjamin was no ordinary carpenter, however. In today’s parlance, he was a contractor and developer. “In the ’30’s,” a Washington history recounted in 1908, “Benjamin Williamson erected a row of two-story and basement frames…and Williamson’s row was as well-known as was the grocery of Michael Sardo, erected at the southeast corner of 10th and H streets.” The savvy Scotsman knew, as Jack Eppes had advised Jefferson two decades earlier, that investing in building the capital city would yield “a handsome interest on money and would perhaps be better property than United States paper, being more permanent.”