Thomas, John, and Francis Baldwin of Chester County: A Rare DNA Marker

THOMAS, JOHN, AND FRANCIS BALDWIN OF CHESTER COUNTY: A RARE DNA MARKER by Donn Devine, CG, FNGS T homas, John, and Francis Baldwin came to America in 1682. But unlike legendary brothers of so many undocumented accounts who went off in separate directions, these three all settled in Chester County. Through much of the next century, many of their descendants continued to reside there, in Delaware County set off from it in 1789, or in adjoining New Castle County, now across the circular boundary line in Delaware. Where they went their separate ways was in their choice of churches and, subsequently, the extent of the records available about their descendants. Most of John’s and Francis’s progeny continued in the Quaker tradition, and Quaker meeting records have provided support for genealogies that are quite complete, even for lines that left Pennsylvania for the Carolinas or western Virginia. In contrast, many of Thomas’s descendants adhered to Anglican churches, such as St. Paul in Chester or St. Thomas in Marcus Hook, or the Swedish Lutheran Holy Trinity (“Old Swedes”) Church in Wilmington, none of which maintained as complete records as the Quakers. Other descendants joined Baptist congregations, which maintained even fewer records. Lines of descent from John1 and Francis1 through the 18th and well into the 19th century are well documented (see Baldwin Documentary Evidence, below), but nothing is known of some of the third brother Thomas’s mid-18th century descendants of the fourth and later generations. Many Baldwins who have been unable to document ancestors earlier than the late 18th or early 19th centuries have hypothesized descent from Thomas1 but have not been able to overcome the absence of extant records from the mid-18th century relating to some of his descendant lines. This study uses DNA from the male Y chromosome, which is passed on in the male line from father to son essentially unchanged except for infrequent random changes called mutations (or polymorphisms). Some preliminary but useful findings are reported here, but further testing of appropriate descendants is still needed to resolve remaining open ques Donn Devine has been a member of the board of directors of the Genealogical Society of Pennsylvania since 2008. He is the archivist of the Catholic Diocese of Wilmington and a semi-retired Delaware attorney. He has also served on the boards of the National Genealogical Society and the Board for Certification of Genealogists. 102 Pennsylvania Genealogical Magazine Vol. 49 (2014) © Genealogical Society of Pennsylvania 2014 Thomas, John, and Francis Baldwin of Chester County 103 tions. This study builds on two ongoing independent Baldwin searches based on conventional documentary research. One seeks ancestors of Pratt J. Baldwin, born in 1862 and reared as an orphan in Delaware County (see below).1 The other seeks origins for four Baldwins, born in the last two decades of the 18th century of unknown parents, whose lives in western Pennsylvania, Ohio, and Indiana followed parallel migratory routes.2 The latter four, as will be seen, have been determined from records to be brothers, although their parents are still unknown. The principal finding of this DNA study to date is that descendants of a number of the unconnected lines share a distinctive and rarely occurring marker with a known descendant of the immigrant brother Thomas1 through his son Anthony2 and showing clearly that they share a Baldwin ancestor who is not shared by descendants of the immigrant Francis1. Whether their common ancestor was Thomas himself, his son Anthony, or a later descendant in that line remains an open question, although the significant number and wide distribution of descendants with that marker by the late 1700s suggests an earlier, rather than later, first occurrence. As yet, no known descendants of John1, the third brother, have come forward to be tested.3 What makes this interim discovery significant is 1 This paper, below. 2 Marti Baldwin Swanson, editor, The Baldwin Family Bulletin (Ringwood, Ill.: the editor) 1-11 (2001–2011), three issues per year, hereafter cited as Swanson. 3 DNA consists of two long strands, comprising complementary pairs of bases designated A, C, T, and G from the initial letters of their chemical names. The base pairs are arranged in a precise sequence that seldom changes from generation to generation. The Y chromosome, found only in males, is about 50 million base-pairs in length. Groups of adjoining base-pairs that determine hereditary characteristics are called genes. Other groups or single positions that have no known function, but that are also inherited substantially unchanged, are called markers. Because they vary among individuals, they are useful for genealogical studies. Over a number of generations, a one-marker change may occur in a single individual, which is passed on to that person’s descendants, and distinguishes them from those in other lines of descent. Two types of marker changes are used in genealogy. The first is called a singular nucleotide polymorphism (SNP, pronounced “snip”). It is a change or mutation (polymorphism) in one of the four nucleotide bases—A, C, T, or G—at a single location along the DNA chain. These changes occur infrequently, and generally identify larger population groups, rather than families. They are used in genealogy for studies of maternal-line or mitochondrial DNA (mtDNA), and more recently, for studies of the autosomal or numbered chromosomes, which include segments inherited from multiple ancestral lines, and in anthropology for studies of deeper ancestry and population origins and migrations. The second type of DNA mutation, and the one used for male-line DNA testing as in the present study, uses markers called microsatellites or short tandem repeats (STRs). These markers consist of short repeated sequences—most often three, four, or five of the letter-designated bases A, C, T, or G—that occur the same number of times in each succeeding generation, unless there is a chance mutation. It is the number of times the sequence is repeated that is reported as the marker value, for comparison purposes. See Megan Smolenyak Smolenyak and Ann Turner, Trace Your Roots with DNA: Using Genetic Tests to Explore Your Family Tree (Emmaus, Pa.: Rodale Books, 2004); and Thomas H. Shawker, M.D., Unlocking Your Genetic History: A 104 Pennsylvania Genealogical Magazine Vol. 49 that the same rare marker value found in the Thomas descendant occurs in a number of other Baldwins tested, whose record lineages do not go back far enough to link them to one of the three immigrant brothers, but their DNA clearly matches the line descended from Thomas. Representative male-line descendants from several unconnected Baldwin lines were tested, along with two known descendants of Thomas1 and descendants of Francis1 whose record lineages are well documented. The DNA test results report numerical values at a number of points along the Y-chromosome called markers. Mutations at these locations are passed along to descendants and identify their lines, distinguishing them from collateral lines where no change has occurred. DNA EVIDENCE: PROBABILITY AND INHERENT WEAKNESS DNA evidence is easy to understand when it is used negatively to disprove a proposed relationship or identity. However, when exactly or closely matching DNA samples are used positively as evidence of a relationship, the findings aren’t quite so straightforward. The likelihood of relationship shown by two matching samples can be expressed only in terms of probabilities, from the highly certain to the very unlikely. Genealogists are frequently uncomfortable with relationships that can, at best, be called only probable. They look for more certainty, such as that offered by the Genealogical Proof Standard (GPS), while recognizing that even conclusions based on that high standard are subject to revision if new evidence is discovered. To take one example of DNA findings, there is a high probability that a grandson’s Y-DNA will exactly match his paternal grandfather’s when 37 markers are tested. However, there is a small remaining probability— perhaps 2 percent—that one or even two of the 37 Y-DNA markers will show a one-step difference over the two-generation interval between them. While improbable, it does happen, and it is the continued presence of these infrequent marker changes in descendants that provides evidence of close or more distant relationships through DNA.4 The numerical probabilities associated with DNA comparisons are merely an effort to describe the degree of uncertainty—the weight of the evidence. There are two inherent weaknesses in the use of DNA evidence Step-By-Step Guide to Discovering Your Family's Medical and Genetic Heritage (Nashville: the author, 2004). 4 An average rate of .004 mutations per marker per generation was used, based upon a number of studies, although rates for individual markers vary over a considerable range. See, for example, a range of .00036 to .00964 in Concetta Burgarella and Miguel Navascués, “Mutation rate estimates from 110 Y-chromosome STRs combining population and father-son pair data,” European Journal of Human Genetics 19(1) (2011): 70-75 and a range of .00378 to .0744 in Kaye N. Ballantyne et al., “Mutability of Y-Chromosomal Microsatellites: Rates, Characteristics, Molecular Bases, and Forensic Implications,” American Journal of Human Genetics 87(3) (10 September 2010): 341353. 2014 Thomas, John, and Francis Baldwin of Chester County 105 to support genealogical conclusions, both of which appear in this study. The first is found in comparing results between two individuals. While a non-match is conclusive evidence that the two are not biologically related within the period of recorded history, a close or even an exact match shows only that they share a common ancestor somewhere in their respective lineages. The closeness of the match at best gives only a statistical estimate for how many generations back their common ancestor may have lived. By comparing closely matching results from at least three different descendants of a known common ancestor from two different lines of descent—a process sometimes called triangulation—it is often possible to determine the precise generation in which a single marker mutation first occurred in an otherwise matching DNA pattern, thus providing a distinctive marker that identifies further descendants of that particular line. The second inherent weakness in the use of DNA is that identifying the ancestor who first displayed a mutation from the ancestral pattern through the triangulation method depends on the credibility of the reported lineages of the sample donors, based upon traditional documentary research. THE BALDWIN DNA EVIDENCE Recent Baldwin DNA testing has uncovered what is sometimes called a private marker. Here it distinguishes one documented descendant of the immigrant Thomas1 through his son Anthony2 from a number of other Baldwin men, all known to descend from Thomas’s brother Francis1. All of the Baldwin men tested here belong to a population subgroup designated I2b1, also called I-M223. It is part of the major Haplogroup I, which is found extensively in Scandinavia, but also in smaller numbers in Great Britain, Ireland, and along the North Sea coast of western Europe. The group is suggestive of a Viking or Norman origin for the all-male paternal line.5 The marker that distinguishes the two groups of related Baldwins is a particular location designated YCA-IIa along the Y chromosome DNA strand. It is 28th in the 37-marker panel of results as reported by Family Tree DNA, where the tests were done. The known descendants of the immigrant Francis1 all have the value 19 at that location, the value that occurs most frequently in Haplogroup I and is called the mode, or modal value (see Table 1). 5 Wikipedia, “Haplogroup I,” http://en.wikipedia.org. Viewed 28 Feb 2012. 106 Vol. 49 Pennsylvania Genealogical Magazine Modal Values at 67 STR Markers for Closely Matching Baldwin Haplotypes PANEL 1 (1-12) Panel 1 (1 – 12) Locus 1 2 3 4 5 6 7 8 9 10 11 12 DYS# 393 390 19 391 385 a & b 426 388 439 389-1 392 389-2 Allele 15 23 12 17 15 10 14 17 11 13 11 14 14 15 459 a & b 16 455 17 454 18 447 19 437 20 448 21 449 22 23 24 25 ---- 464 a – b – c – d ---- 14 8 10 11 11 26 14 20 27 11 13 14 Pane 3 (26 – 37) Locus 26 27 28 29 30 31 32 33 34 35 36 37 Panel 2 (13 – 25) Locus 13 DYS# 458 Allele DYS# 460 GATA H4 YCA-II a & b 456 607 576 570 CDY a & b 442 438 Allele 11 10 19 14 14 16 19 35 12 10 Panel 4a (38 – 47) Locus 38 21 38 16 39 40 41 42 43 44 45 46 47 DYS# 531 578 395 S1a & b 590 537 641 472 406S1 511 Allele 12 8 16 16 8 11 10 8 10 10 Panel 4b (48 – 60) Locus 48 DYS# 425 49 413a 50 413b 51 557 52 594 53 436 54 490 55 534 56 450 57 444 58 481 59 520 60 446 12 21 21 15 11 12 12 14 9 15 26 20 13 Panel 4c (61 – 67) Locus 61 62 63 64 65 66 67 Allele DYS# 617 568 487 572 640 492 565 Allele 13 12 12 11 12 12 11 Table 1. Modal Values for the Chester County, Pa., Baldwin Project Modal values are a calculated measure of a central tendency in a group of statistics and provide a convenient reference point for discussing differences among the members of the group. They do not necessarily represent values displayed by any specific member, nor among related group members do they necessarily represent the values carried by their common ancestor, although in many cases they may, as can sometimes be demonstrated. The marker values (alleles) shown here are the one that occur most frequently at each of the tested marker positions, among the group of 20 men considered in this study. They all have patterns, called haplotypes, that closely match known descendants of three English brothers from Oxfordshire who came to Chester County, Pa., in 1682. Haplotype is the term that describes a pattern of markers in a DNA segment inherited intact from the same parent. Almost all of the Y chromosome is inherited intact by a male from his father, so its pattern or motif is usually called a haplotype. The various numerical values that may occur at a particular marker are called alleles. Modal values like the alleles shown here may, but do not always, represent the haplotype of the most recent known common ancestor of the group, in this case William Baldwin, father of the three immigrant brothers, who lived at Hook Norton, Oxfordshire, in the mid-1600s. The rare allele 17 at marker YCA II in a number of descendants is more likely a mutation or derived state of that marker, and the more common allele 19 is the original or ancestral state carried by William, but tests of more known descendants are needed for confirmation. The small differences among the closely matching haplotypes in the study are shown in Table 2. However, two known descendants of his brother Thomas1 through his son Anthony2 have a distinctive and very rare 17 at this marker (see Figure 1). They share this unusual characteristic with several other Baldwins whose earliest known Baldwin ancestors were born in Pennsylvania in the late 1700s but have not been connected through record evidence to any of the three immigrant brothers (see Figure 2). The value 17, in combination with the value 21 at the associated marker 2014 Thomas, John, and Francis Baldwin of Chester County 107 Shown here are five lines connected by documented lineages to two of the three immigrant Chester County Baldwin brothers. Boxed names mark the earliest individual in each line for whom the marker values that distinguish the lines of Thomas and Francis have been determined, based on tests done to date, and also for the living donors of samples tested. Lineages have not been independently verified; they are based on compiled genealogies discussed in the text and recent family records of the sample donors. William, born 1716, is known from the will of the first-generation Francis to be his grandson, a relationship consistent with two other descendants’ DNA, but the identity of William’s father is at best probable. YCA-IIb found in all of the Baldwins considered here, occurs at a frequency of only one in 10,000 among members of Haplogroup I.6 6 Leo Little, “Y-STR Allele Frequencies” RootsWeb Freepages. Viewed 6 Jan 2009 at http://freepages.genealogy.rootsweb.ancestry.com/~geneticgenealogy/yfreq.htm/. Frequency tables for the two markers designated YCA-IIa and YCA-IIb show that less than 0.5 percent of 5,528 samples from Haplogroup I had a YCA-IIa value of 17, making it quite rare in relation to the modal value 19, which occurred in 85% of Haplogroup I samples. When combined with the overall 87% frequency of the YCA-IIb modal value 21, the combined values 17 and 21 could be expected at a frequency of no more than .005  .87 = 0.45%, less than once in every 225 samples, while the Pennsylvania Baldwin cluster shows nine of these rare occurrences among 20 samples. 108 Pennsylvania Genealogical Magazine Vol. 49 Records do not identify earlier ancestors for the lineages shown here, but each carries the rarely occurring value 17 at marker YCA-IIa, a very strong indication that all three descend from the first-generation Thomas Baldwin of Chester County, born 1657. A test from a descendant of his immigrant brother John is needed to be sure the latter didn’t also carry the 17. Further tests from descendants of Thomas through sons other than Anthony are also needed to show whether the 17 mutation first appeared in Thomas himself, rather than in his son Anthony, whose two known descendants tested—BAL 45 and BAL 77 (Figure 1)—have both inherited that distinctive value. The modal value 19 was also found in two other persons who have record descent attributed to Thomas1 through his son Joseph2. As will be seen, this attribution was characterized by its compiler as only the most probable (their descent from Thomas’s1 brother Francis1 cannot be ruled out7), so these lineages could not be reliably used for triangulation to determine the generation in which the distinctive allele 17 first appeared (see Table 2). 7 Judson C. Ammons, Descendants of William Baldwin (Hendersonville, N.C.: the author, 2003). Available from the author, 320 Colony Lane, Hendersonville, NC 287918612, jtaz1957@bellsouth.net. 2014 Thomas, John, and Francis Baldwin of Chester County 109 Table 2. Differences Among Closely Matching Baldwin Haplotypes This table shows how closely the 20 samples matched when compared at the 37 markers for which all were tested. Five of the 20 displayed the modal values at all markers. Differences appeared only in the 12 marker columns shown here, including the significant YCA-IIa marker, where nine samples showed the distinctive allele 17 identified with descent from the immigrant brother Thomas Baldwin. None varied by more than four steps from the modal values, which is consistent with descent from a common ancestor nine to 10 generations in the past. Of nine samples that were tested at an additional 30 markers, only two varied from the additional modal values—one at a single marker, the other at two markers. It cannot be overemphasized that this study is a work in progress. The findings here are based on incomplete testing and are subject to change as further testing is completed. For example, if a test from the lines of Thomas’s1 sons Thomas2 or William2 shows the same distinctive 17 as does their brother Anthony’s2 line, it would confirm (1) that that value first appeared in their father Thomas1, and (2) that Thomas’s son Joseph2, having inherited the same 17, could not have been the father of John3 attributed to him on the basis of records, but must have been instead 110 Pennsylvania Genealogical Magazine Vol. 49 from a line carrying the more usual 19. Alternatively, if a descendant of either Thomas2 or William2 has the more usual 19, it would provide clear evidence that the mutation first appeared in their brother Anthony 2, not in their father Thomas1, and that all those now showing a 17 at that marker are descended from Anthony2. THE BALDWIN DOCUMENTARY EVIDENCE Baldwin researchers are the beneficiaries of two massive conventional genealogies of the past two centuries. While source citations of the modern type are largely absent, sources are often described or can be determined from the text itself. In both cases, the sources from which the authors worked have been preserved and are available to current researchers. Judge Charles Candee Baldwin compiled his two-volume The Baldwin Genealogy from 1500 to 18818 and The Baldwin Genealogy Supplement9 based largely on correspondence with family members and genealogists in the area, including the respected Gilbert Cope for Pennsylvania lines. Baldwin’s extensive correspondence and copies of the records sent to him are preserved in the library of the Western Reserve Historical Society, Cleveland, Ohio. Over a century later, Aubrey H. Baldwin, Jr., produced his Revised Baldwin Genealogy in five volumes, published in a limited edition of bound photocopies of his typescript and deposited in several major libraries, but also widely available on microfiche through the FamilySearch branch library system.10 His revision and update of the earlier work was based largely on typescript extracts of church, land, tax, probate, and census records from areas where Baldwin families settled. Images of these extracts, arranged by state and county, are available on FamilySearch microfilm.11 Other recent Baldwin genealogies are more specific in scope. Frank C. Baldwin’s The Baldwins from Virginia Westward12 extended lines that were in Virginia before 1800, based on a review of Charles Candee Baldwin’s correspondence, as well as original records that have since become available, and on Aubrey Baldwin’s research. Judson C. Ammons, 8 Charles Candee Baldwin, The Baldwin Genealogy from 1500 to 1881 (Cleveland, Ohio, 1881), also online at http://www.ebooksread.com/authors-eng/c-c-charlescandee-baldwin/the-baldwin-genealogy-from-1500-to-1881-dla/1-the-baldwingenealogy-from-1500-to-1881-dla.shtml/. 9 Charles Candee Baldwin, The Baldwin Genealogy Supplement (Cleveland, Ohio, 1889). 10 Aubrey H. Baldwin, Jr., Revised Baldwin Genealogy (Philadelphia: 1990). Available on FamilySearch microfiche: vols. 1-4: 6103878, 6103879, 6089174, 6089173; vol. 5, covering Pennsylvania Baldwin families, 6093213. 11 FamilySearch microfilms 1421707, 1421708, principally Baldwin extracts from deeds, wills, and other public records. 12 Frank C. Baldwin, The Baldwins from Virginia Westward (Evanston, Ill.: the author, 1985, reprinted 1999). 2014 Thomas, John, and Francis Baldwin of Chester County 111 in Descendants of William Baldwin,13 built on Aubrey Baldwin’s work by extending North and South Carolina lines beyond the fourth and fifth generation down to more recent ones. Marti Baldwin Swanson, editor, The Baldwin Family Bulletin,14 focuses on the descendants of four brothers, their parents not yet identified, who were born in the late 1700s, probably in western Pennsylvania, and who moved through Ohio into Indiana. In addition to the inherent problems associated with DNA evidence, two further problems frequently appear when attempts are made to integrate it with documented lineages to determine the generation in which a mutation first appeared. The first of these problems is the inaptly named “non-paternity event” (NPE), where genetic evidence from a DNA comparison shows that a well-documented and never-questioned lineage contains a father-son relationship somewhere in the surname line that is genetically impossible—a father-son pair who are unrelated biologically, whatever other evidence points to their relationship. However, the paternity, which happened, is not the issue. Rather, it is the patrilineal succession to the biological father’s surname that failed. “Non-patrilineal transmission” (NPT), as used in more recent genetic literature, describes such occurrences more accurately.15 The most frequent cause of NPTs in the period before vital records was an unrecorded first marriage by a woman who brought to a second marriage a child or children whom the second husband raised as his own, or an orphaned child taken in by relatives or neighbors without leaving a record that the child was not biologically their own. The second problem found in integrating DNA with recorded lineages is a father-son kinship link that does not meet GPS requirements, and so can be described at best only as “probable.” Most often such deficiencies involve a kinship link that depends on evidence from a single source without independent confirmation, is from a chain of sources that may not have transmitted the information reliably, or is deduced from indirect evidence but with an unresolved inconsistency or subject to alternate interpretations. Other published lineages for donors of the DNA samples used here have at least one father-son connection qualified by the compiler as only probable, but they have been accepted where there were no conflicts with other published lineages or with the DNA evidence itself. In each case, the DNA evidence was consistent with the probable conclusion from records and adds independent support for the earlier conclusion, 13 Ammons, Descendants of William Baldwin (note 7). 14 Swanson, Baldwin Bulletin (note 2) 15 See, for example, Turi E. King and Mark A. Jobling, “Founders, Drift and Infidelity: The Relationship Between Y-Chromosome Diversity and Patrilineal Surnames” Molecular and Biological Evolution 26 (May 2009): 1093. 112 Pennsylvania Genealogical Magazine Vol. 49 while not necessarily confirming it. DNA tests can also show unexpected biological breaks in lines that had good support from records. One early participant in our study has a solid record lineage back to the immigrant brother Francis1 but his DNA test results place him in a different haplogroup, a major population group that split off thousands of years ago from the one that our three Chester County brothers belonged to. A major mismatch in a surname line provides conclusive evidence for an NPT, although no place for it has been identified in the line of the disappointed tester.16 The distinctive 17 at marker YCA-IIa is also found in three Baldwin families that closely match known descendants of the three immigrant brothers, but who have not yet found records of their own Baldwin ancestors earlier than the late 1700s. One of these families is first known in present Delaware County, Pennsylvania, a second in adjoining New Castle County, Delaware, and the third in western Pennsylvania, from the unidentified Baldwin parent of the four brothers who are the subjects of The Baldwin Family Bulletin (see Figure 2). Although parents of the four brothers are unidentified, their fraternal relationship has been established through direct or indirect documentary evidence and confirmed by their closely matching DNA results. All share the distinctive 17 marker and have been the subject of intensive study for the past decade, as reported in The Baldwin Family Bulletin. They are represented by DNA samples BAL 6, 11, 12, 22, 25, and 63. The Delaware County line that shares the rare distinctive 17 is attributed to Eli Baldwin of Lima, Delaware County, born about 1790, as its most probable progenitor, and is represented by sample BAL 1. The New Castle County line with the 17 marker descends from William Baldwin, born in 1770, who married Susannah Crozer in 1793 in Delaware17 and died there in 1814. His line is represented by BAL 85. PROBLEM LINKS IN DOCUMENTED LINES, WITH POSSIBLE CAUSES AND CORRECTIONS Most problems in establishing relationship links arise when evidence is attributed to the wrong individual, often a person of the same name and approximate age. In the following discussion, the lineages are identified by the DNA sample number of the present-day descendant, a number preceded by BAL. Most of the problem links addressed can be found among Baldwin descendants listed in Aubrey Baldwin’s Revised Baldwin Genealogy. 16Thomas P. Doherty, PhD, Wilmington, DE, unpublished report on Descendants of Levi4 Baldwin (ca. 1760–1825), described here but without reporting lineage details or DNA results with permission of his client, whose anonymity was protected by designating his sample BAL 2. 17 See Baldwin-Crozer marriage bond, 20 July 1793, New Castle, Delaware; image online in Delaware Marriage Records, 1744–1912, at Ancestry.com. 2014 Thomas, John, and Francis Baldwin of Chester County 113 Descendants of the three immigrant Baldwin brothers in that work are numbered in separate lines from each of the brothers, Thomas, John, and Francis. Those numbers, preceded by T, J, or F to indicate the line in which they appear, will be used in the discussion following to clarify references among many recurrences of the same names. The problematic links were usually identified as probable or suggested. THE LINE OF ELI, BAL 1 This DNA sample is from a grandson of Pratt Baldwin, who died 17 June 1945. By his own account, Pratt was born 4 March 1862 in Delaware County, Pennsylvania, and both his parents, whose names he did not know, died while he was an infant; he was “put out” with a family named Frame and was still with them at age 18 in 1880, 18 but left them before his term expired at age 21.19 The probable identity of his parents has been inferred from the manner he was entered in the Delaware County pauper book,20 a chronological register of individuals who came under the care of the county directors of the poor. His name is interlined above and bracketed with a 27 November 1864 entry for Penrose Baldwin, white, 1 year 1 month old, from Middletown, born Philadelphia, bound on 7 November to Eli Baldwin, his grandfather. The bracketed entry for Pratt has only his name with a ditto mark in the date column and the date “5th 9th month” [5 September] of his 1870 indenture to P. Miles Frame as an apprentice farmer for a term of 12 years 6 months, consistent with coming of age 21 years from birth on 4 March 1862. A Philadelphia birth registration exists for Penrose Baldwin, born 22 December 1862 to Alfred and Mary Baldwin, 1 Alroy Street, registered 31 December 1862,21 but no record of birth has been found for Pratt. Indentures of both Penrose and Pratt are found in the Poor Directors’ Indenture Book. Penrose’s term of service—19 years, 1 month and 15 days after 5 November 1864, ending 20 December 1883—is consistent with his registered birth date 22 December 1862. However, his age “1 yr 1 mo.” shown in the 27 November 1864 pauper book entry should have read “1 yr 11 mo.” to agree with his recorded birth date and reaching his majority the day following expiration of his indenture. 18 U.S. Census 1880, Pennsylvania Population Schedules, Delaware County, Birmingham Twp., stamped page 550, household 130, NARA T9 Roll 1125, FHL 1255126 (head of household P. Miles Frame). 19 Recollections of Pratt Baldwin’s granddaughter, Elizabeth Baldwin Devine, wife of the author, abstracted 1977. 20 Delaware County Directors of the Poor, 1845–1867 pauper book (Delaware County Archives, Lima, Pa.), 10-11. 21 Philadelphia Department of Records, Vital Statistics, birth certificate issued 2 Dec 1980 Penrose B born 22 Dec 1862, parents Alfred Baldwin, Mary Baldwin Baldwin, 1 Alroy St, registered 31 Dec 1862. 114 Pennsylvania Genealogical Magazine Vol. 49 Penrose’s father, Alfred Lee Baldwin, is found in the 1850 Census, age 10, in the household of Eli Baldwin, 58, with Esther 47, and Marshall, 16.22 Alfred has not been positively identified in the 1860 Census but may be the Afred [Alfred] Baldwin, 18, apprentice gilder, in the household of Geo. P. McLean, 48, looking glass and picture frames maker, in Philadelphia’s 13th Ward.23 Alfred’s death on 9 December 1863, age 23, appears in a newspaper notice and was recorded on a gravestone in Lima M.E. Church cemetery, where Eli and Esther Baldwin are also buried.24 Some evidence raises doubts about the fraternal relationship of Pratt and Penrose: (1) the possible but unlikely less than 10-month interval between their births; (2) grandfather Eli’s failure to show the same concern for both boys; and (3) inconsistency between Pratt’s received account that his parents died in a common calamity,25 and Penrose’s 1918 obituary listing his mother, then Mrs. Phillip Hallworth, as a survivor living in Haddonfield, New Jersey.26 Pratt’s belief that his parents died together in a common calamity, as his granddaughters recall,27 is inconsistent with the documented survival of Penrose’s mother after his 19 August 1918 death28 in Asheville, North Carolina, where he was a retail pharmacist.29 Notwithstanding those conflicts, the Delaware County Poor Book entry, where Pratt’s name is interlined and bracketed with the earlier entry of Penrose’s name, remains strong evidence of a relationship of Pratt to Penrose known to the Directors of the Poor, and thus relating Pratt through the Baldwin male line to the deceased Alfred and his father Eli. Such a relationship is supported by the DNA haplotype of Pratt’s grandson, showing descent from the line of the immigrant Thomas Baldwin through his son Anthony, and probably through Anthony’s poorly documented 18th-century progeny who are Eli’s most likely ancestors. If Pratt and Penrose were only half-siblings with different 22 Eli Baldwin household, 1850 U.S. Census, Middletown Township, Delaware County, Pennsylvania, penned page 370, #135/136 (NARA M432, roll 776); image online at Ancestry.com. 23 Geo. B. McLean household, 1860 U.S. Census, Philadelphia Ward 13, Philadelphia County, Pennsylvania, stamped page 872, #393/450 (NARA M653, roll 1163); image online at Ancestry.com. 24 Delaware County American, 30 Dec 1863; Works Progress Administration, Delaware County Graveyard Survey Vol, B, Lima M.E. Church collections of Delaware County Historical Society. 25 Author’s abstracts, Pratt Baldwin recollections by his granddaughters Elizabeth Baldwin Devine and the late Jane Baldwin DeShan, the author’s wife and sister-in-law respectively. 26 Philadelphia Inquirer, Penrose Baldwin death notice, vol. 179 (13 Sep 1918) #75. 27 Pratt Baldwin recollections, above. 28 North Carolina Board of Health, 1918, Penrose Baldwin death certificate #12441. 29 Penrose Baldwin death notice, above. 2014 Thomas, John, and Francis Baldwin of Chester County 115 mothers, the circumstance could explain a family reluctance to recognize Pratt’s relationship. In Baldwin, Revised, Eli is attributed as an eighth child to Levi5 of New Castle County, Delaware, from the line of Francis, citing Levi’s will, in spite of its very definite reference to “my seven children,” “any of them,” and “each of them,” and the lack of any reference to an Eli.30 The will of Levi’s mother, in which she made a bequest to a grandson “Eli senr,” suggests that there may have been a second grandson of that name, in addition to Eli, the son of her deceased son, Eli.31 However, Levi’s account as his mother’s executor shows clearly that the bequest was to his nephew Eli, not to a son. Whether or not Levi had a son Eli, the identification of him in Baldwin, Revised, with the Pennsylvania-born Eli Baldwin of Lima, Delaware County and his wife, Esther, and his placement in the lineage of Francis as #F18032 have no evidentiary basis and are not consistent with Eli having a YCA-II DNA marker 17 as found in his probable grandson Pratt’s line. The Pennsylvania-born Eli of Lima is mentioned in two works by the prominent Chester and Delaware counties genealogist Gilbert Cope. He describes Eli and his first wife Elizabeth Walton of Lima as parents of Abigail Baldwin, who married Jacob Farra about 1840,33 and in another work that lists Eli’s second marriage to Esther Clark, widow of Abel Green, noting that they had five or six children, but naming only Marshall, of Media, after listing children of her first marriage.34 Although it is clear from the DNA evidence that Eli was descended from the immigrant Thomas, no descendant of Thomas’s line has yet been identified as a probable father of an Eli born in the 1790s. THE LINE OF WILLIAM, BAL 85, BORN 1770, LATER OF WILMINGTON, DELAWARE William Baldwin of Wilmington, Delaware, born 1770 and progenitor of a line with the distinctive 17 marker, cannot be the William Jr. #F69 listed in Baldwin, Revised in the line of Francis1 as the son of #F22 William3, through #F5 John2. Accordingly, identifying that Francis1 descendant with the William of Christiana Hundred who married Susannah Crozer and lived in Delaware must be a mistake. Baldwin, Revised lists two 30 New Castle County, Delaware, Will Record S-1, p. 80. 31 New Castle County, Delaware, Will Record W-1, p. 853. 32 Aubrey Baldwin assigned a separate series of descendant numbers to each lineage in his work. Here, descendant numbers from that work will be preceded by the letters T, J, or F to distinguish the lines of Thomas, John, and Francis respectively. 33 Gilbert Cope, Genealogy of the Sharpless Family (Philadelphia: the author, 1887), 1216. 34 Gilbert Cope, Genealogy of the Smedley Family (Lancaster, Pa.: the author, 1901), 370. 116 Pennsylvania Genealogical Magazine Vol. 49 possible alternatives from Thomas’s line who are more likely fathers to William, progenitor of the BAL 85 line: #T165 William, who resided in Delaware State, and #T57 William, a son of Anthony3, if his later description in Baldwin, Revised as the tailor of that name who relocated to Ohio is also mistaken. Not knowing that the William who married Susannah Crozer had a son named Harlan, it is quite possible that Aubrey Baldwin misidentified the two Williams, both sons of an Anthony Baldwin and born the same year, mistakenly linking them to the wrong fathers. The most reasonable hypothesis consistent with the DNA evidence is that the William who married Susannah Crozer was the son of William born 1770 to #T21 Anthony3 and his wife Mary Harlan. The undated #T165 William, fifth child of #T67 Anthony4 who was born 10 May 175135 could not be BAL 87 progenitor, although he lived in Delaware, because his father was too young to have had five children by 1770. THE LINES OF BAL 67 AND BAL 69 John Baldwin Jr. of Chatham, North Carolina, and his wife Frances are the documented progenitors of these two lines, both displaying the modal 19 at YCA-IIa. Both have documented lines to the same early Baldwin forebears—a John Baldwin, born about 1732 in Chester County, Pennsylvania, who moved to Orange County, North Carolina, by 1755, and with wife Frances had sons John (BAL 67 a descendant) and Eli (BAL 69 a descendant). Ammons identifies this John Jr. with the #T44 John Jr. in Baldwin, Revised, without a birthdate, but noted as a son who received a land grant in Rowan County, North Carolina, at the same time as his father #T16 John3, born 1714, and his first wife, Phylis. The older John was the first grandson of the immigrant brother Thomas through his son Joseph2, born 1685. However, Ammons cautioned that the parents of John, born about 1732, were determined from indirect and incomplete evidence, and proposed them as most probable, while noting the possibility that the John who married Frances could be the John, son of William, from a related Baldwin line. Baldwin, Revised lists a John, numbered 15 1/2, as a grandson of Francis1 through his son #F2 William2, who had relocated to North Carolina by 1724, but gives no further information about him. The DNA from the two descendants of John Baldwin, born about 1732, who married Frances both show 19 at marker YCA-IIa, making their descent more probable from the immigrant Francis, rather than from Thomas, whose known descendants show the distinctive marker 17. As noted 35 Historical Society of Delaware, The Records of Holy Trinity (Old Swedes) Church, Wilmington, Del., From 1697 to 1773, With an Abstract of the English Records From 1773 to 1810 and Catalogue and Errata of the Records of Holy Trinity (Old Swedes) Church (Wilmington: the author, 1890), 561. 2014 Thomas, John, and Francis Baldwin of Chester County 117 earlier, final resolution depends on comparison with DNA samples from known descendants of Thomas’s other sons. THE LINES OF BROTHERS JOSEPH, ELIAS, WILLIAM, AND THOMAS BALDWIN: BAL 6, BAL 22, BAL 11 & 25, AND BAL 12 & 63 These four Baldwins have been shown to be brothers from records, although nothing pointing to their parents’ identities has been found. They are variously described in later census entries for their children as born in Pennsylvania or Virginia, so probably in that part of western Pennsylvania that was administered as part of Virginia until after the Revolution. The fraternal relationship of Elias, born 1772, and William, born 1781, and their marriages to Crooks sisters Mary and Amy respectively was recorded in both family lines.36 William’s relationship to Thomas, born 1783 is evidenced by an 1835 affidavit of William’s son William Crooks Baldwin, identifying Indiana Baldwin, daughter of Thomas, as his cousin and of age to marry.37 The relationship of Joseph, born about 1779, to Thomas was confirmed through the marriages of half-siblings William and Clarissa Rebecca Cook to Baldwins known to be cousins—William to Mary Ann Baldwin, daughter of Joseph, in 182438 and Clarissa to William A., son of Thomas, in 1843.39 ALTERNATIVE LINKS AND CURRENT HYPOTHESIS By 1770, a carrier of the 17 marker, likely born before 1750 and married, had a son Joseph in western Pennsylvania/Virginia, followed over the next 18 years by sons Elias, William, and Thomas. At the same time, another carrier of the 17 marker was still in eastern Pennsylvania, married and had a son William. Some 20 years later, probably in what is now Delaware County, a carrier of the 17 (perhaps William’s father, but more likely a younger Baldwin) had a son Eli. Anthony2 Baldwin, the earliest certain carrier of the 17 marker, had two sons by his first marriage, neither of whom had children, and five by his second marriage, born between 1729 and 1742, who had among them 18 sons, but the sons’ births between 1763 and 1786 place them too late to have fathered the unconnected lineages, with heads born between 1770 and 1788.40 36 Swanson, Bulletin 1#1:3, citing Merle L. Shanklin, The Shanklin Family of Montgomery County, Indiana (1969), p. 17. 37 Swanson, Bulletin 2#1:9, notes 7 and 9, citing Indiana State Marriage Records, Montgomery County, 15 November 1835. 38 Swanson, Bulletin 8#3:5, citing Ohio Marriage Records, Butler County, v. 1 p. 194, 17 Feb. 1824. 39 Swanson, Bulletin 8#3:5, citing Indiana Marriage Record Book 2, p. 88, 6 Nov. 1843, noting that the online database incorrectly states 11 Nov. 1843. 40 Baldwin, Revised, Thomas Section, (note 9), 12, 18-20. 118 Pennsylvania Genealogical Magazine Vol. 49 Without a known descendant of Anthony in a position to have fathered the unconnected lines, and considering the distinctive marker’s wide geographic dispersion by 1770, it may well have first appeared in Thomas1 himself, rather than in his son Anthony2, born in 1690, and have been spread by his other sons, as well as by Anthony. However, only further DNA testing of well-documented descendants of Thomas’s1 second-generation sons Thomas, William, or Joseph will put the matter to rest. To conclude this interim account, it is reasonably certain that all Baldwins of suspected Pennsylvania origin whose Y-DNA shows the distinctive YCA-IIa marker 17 are descended from Thomas Baldwin of Oxfordshire, who came to Chester County, Pennsylvania, with his brothers John and Francis in 1682. Whether they all descended through his son Anthony remains to be determined by further DNA testing of known descendants of Thomas through his other sons. _________ Note: The Baldwin Surname DNA Study website is online at http://tinyurl.com/baldwin-ft and http://tinyurl.com/BaldwinStudy. The project has a small financial aid fund that can assist prospective Baldwin test subjects, if cost stands in the way of their testing. Contact the author for further information at donndevine@aol.com.