Fan chart of direct ancestors.
Rings are generations, wedges are people, solid is Confirmed and pale is Probable. The record is uneven and the fan shows it: a line that stops short is a line not yet traced, not a family that ended.
Find a personThe linesLifetimes on a timelineWhere they were born and died
Play the fly-through · every generation as a corridor, Earl and Kristin in front
The family as a corridor of generations: Earl and Kristin in front, each row behind them one generation further back.
Each row is one generation; each light is a person with a page of their own. Solid is Confirmed, hollow is Probable. People with no birth year on record are placed by generation.
A sourced family record
The Proeger Family
From a German emigrant glassworker buried in Pittsburgh and an Azorean immigrant buried in Alameda County to a family in Austin, Texas — and, on Kristin's side, a German-Texan line out of Austin County, a Norwegian one out of Clear Lake, South Dakota, and a Caffey line that runs back through Ireland to a man born in 1690. Every statement below carries the tag the record research gave it — confirmed, probable, or not found — and a citation to the document it rests on. One section rests on something else and says so: the DNA.
The two charts below show five generations each, which is as far as a side-by-side chart can usefully go. Several lines run deeper than that, and they run to very different depths — that is the subject of the deep lines, where each surname gets its own chain and the uneven depth is the finding rather than a defect.
Record research from 13 September 2026 · every claim cited · nothing inferred without saying so
All sixteen slots of the outermost five generations are filled; the chains run further on the line pages.
How to line up this chart with the fan above
The fan's rings are labelled G1 (centre) through G5 (outer edge); the columns below carry the same labels. Each person occupies the same position in both — the fan shows them as a coloured wedge, this chart shows them as a card with a name and a status chip.
Confirmed on a card = solid wedge on the fan. Probable confirm or correct → on a card = pale wedge on the fan. The colour itself is the line — the same hue in both views.
Earl's ancestors — the five-generation chart, as a table
G1 · Subject
G3 · Grandparents
G4 · Great-grandparents
G5 · 2×-great-grandparents
Kristin's ancestors — the five-generation chart, as a table
G1 · Subject
G3 · Grandparents
G4 · Great-grandparents
G5 · 2×-great-grandparents
Kristin's maiden name is Caffey, established from the Travis County marriage record S2. Her father's line is now three generations deep and on records, not on testimony: he was Thomas Alan Caffey (1946–2016) of Austin, and his parents and all four of their parents are named below. The identification needed no inference at all — his own obituary names "his daughter Kristin Proeger and her husband Earl of Austin" S31.
Her mother's half is no longer empty. On 13 September 2026 the four boxes behind Lavonne Walker filled in one pass, and the generation above them filled with them: her parents were Clarence Melvin "Melvin" Walker (1920–1991) and Dorothy La Vonne Peters (1926–2022), and Dorothy's parents were Benjamin LaRoy "Pete" Peters Sr. and Gertie Selma Selland — a Norwegian-American couple from Clear Lake, South Dakota S36S38S40. The outermost Peters box is still soft and is marked as such; on the Selland side the Røldal parish registers now carry the line back another two generations on records S74S76. Lavonne herself is living, confirmed by the family S24, so she appears by name and relationship and nothing else.
⚠️ Walker was Dorothy's married name, not her maiden name. Hers was Peters, and her mother's was Selland — both established from records rather than from the surname on the grave S39S40. No name has been guessed and no placeholder is a real person.
How to read this record
Every claim is tagged
Most family trees flatten everything into the same confident typeface. This one does not. The three research tags below are the whole point of the exercise: they say what a document proves, what the weight of evidence suggests, and what nobody has found yet. A fourth marks what the family stated directly, where no document has been retrieved at all — it is kept separate on purpose and is never quietly promoted to confirmed. A fifth, added on 13 September 2026, marks biological evidence: a DNA result is not a document, and it is not testimony either. Two further labels mark where this page withholds information rather than lacking it.
Attested by a document that was actually retrieved and read — a census sheet, a government index, a grave marker, a family-written obituary.
Strongly supported but not nailed down: a user-contributed tree, a masked index field, or a chain of consistent circumstances. Believable, not proven.
Searched for and not located. This is a statement about the search, not about the person — the record may well exist somewhere unsearched.
Supplied by the family and recorded on their word — no document has been retrieved for it. Note the diamond rather than the dot: this is provenance, not proof, and it stands apart from the three record tags above S24.
Biological evidence consistent with a claim. It corroborates origins; it does not by itself establish a person or a link. Note the bar rather than a dot or a diamond: no DNA result on this page is ever counted as confirmed, because no document was read for it S46.
The person is living. By the privacy rule for this page, name and relationship are shown, and at most a year the family has already published. Full dates of birth, ages, places of residence and record identifiers are held back.
A branch that is actively being worked on. The node exists in the structure so the shape of the family is honest; the detail is simply not in yet.
The findings
Both families
Unfinished business
What is still unknown — and the record that would close it
Ordered roughly by value divided by effort. Each one names the specific document that would settle it, so nobody has to rediscover where to look. Reconciled against the research on 26 September 2026; questions that have closed move to the resolved list at the foot.
Which parish on Faial
The island is settled. Frank Silva Sr's November 1946 obituary calls him "a native of Fayal, the Azores" Confirmed S67, and the 1896 lodge registers of UPEC Council No. 10 and IDES Council No. 1 had already written Fayal against a matching Francisco da Silva S64S65. What is not settled is the parish he was born in — and without it there is no Azorean baptism to find, because the registers are organised parish by parish. ⚠️ This card used to name the 1888 Horta passport register as the document that would close it, behind a gate needing a person's click. Both parts are out of date. The page images were never gated, and on 19 September 2026 the whole of 1888 (passports 1–401) and the tail of 1887 were read line by line: no Francisco da Silva, and no Francisco of any surname aged 23–26 Not found. He is on the 1900 census S47 as married thirteen years with a wife who came in 1895 — a young married man who emigrated alone.
What would close it — his 1946 California death certificate, which names his parents and sometimes the town. His naturalization, about 1904, came before the 1906 federal reform, and Alameda County papers of that date usually give the country only, not the town; they are still worth pulling for the court and date. More passport volumes are not the route. Also possible: the UPEC beneficiary file for certificate 2223 or the IDES file for certificate 16362. ⚠️ Frank F. Silva, born at Fayal on 12 April 1863, whose 1901 passport application was read on 26 September 2026, is a different man: he came in 1877 and was naturalized at Sacramento in 1887, while ours came in 1888 and was still an alien in 1900 Not found. ⚠️ The near-match in the Azorean parish indexes — a Francisco Silveira Marques born June 1863 and christened at Horta — stays deliberately not attached: one row out of 1,926 for that name and date range is a coincidence, not an identification.
The Coffman generations above David Carl
This is the least secure stretch of Earl's side. (It used to be called the deepest as well; since 25 September 2026 the Pröger line reaches back to about 1673, so it is no longer that.) David Carl Coffman (1812–1883) is Confirmed — the Page County birth register names him as Dee's father S82 and both his and Emily's gravestones were read S83. But Samuel and Susanna Coffman above him, and Martin "Hawkbill" Coffman above them, rest on one compiled genealogy of 1940 Probable S66 — and the one index entry found since points away from it. The book's 1814 death date for Samuel is doubtful (a Samuel Coffman still heads a household at New Market in 1820), and Barbary Coffman is eliminated as his widow.
🆕 Page 98 of the typed Shenandoah County bond abstracts has now been read at the image (26 September 2026): “Coffman Saml. / Coffman Susanna / Feb 12, 1799 / dr of David, son of Geo.” S664. On the same page the abstracter writes “dr of” for the bride and “son of” for the groom, so Samuel’s father was George, and Susanna was a David Coffman’s daughter Probable: a single typed abstract, not the bond itself, and it disagrees with the 1940 book and with Earl’s grandfather’s chart, which say Martin S103.
What would close it. (1) The original 1799 bond or consent papers in the Shenandoah County clerk’s office or at the Library of Virginia, and George Coffman’s own will or estate, Shenandoah County. (2) The Page County marriage bond of David C. Coffman and Emily Thornhill, c. 1842–43, at the Library of Virginia — the likeliest document to name David C.'s own father. (3) Martin Coffman's will, Shenandoah County. The book cites Will Bk. H, p. 338 for a will of 1809; the county's will is dated 1805, in Will Book E/F, and names seven sons without a Samuel. With the bond abstract naming George, Martin as Samuel's father is now Contradicted. ⚠️ The Page County vital registers index this family as KAUFMAN, the censuses as COFFMAN; one spelling alone returns nothing and looks like proof of absence.
Who joins Charles Caffey to the Caffeys above him
The lower join is now proven. Charles Caffey’s Montgomery County, Alabama estate file was read on 26 September 2026: on 11 March 1868 his administrator swore that “H. V. Caffey” was one of his nine heirs at law Confirmed S663. With Hooper, 17, in Charles’s 1850 household S484, Hooper Van Caffey was Charles’s son. What is still open is the join above Charles. He is joined to John Caffey Sr (1751–1826) only because John’s 1821 will names a son Charles Probable; the estate file does not name Charles’s own father. Above John, Michael Caffey (d. 1754, Dorchester County, Maryland) is real, but whether he is John's father or grandfather is unresolved; the "John Caffey born 1690 in Ulster" on the online family tree has no record at all behind it (checked 26 September 2026).
What would close it — the 1821 will of John Caffey itself, read word for word from the image, and John Sr’s estate sale, to see whether Charles bought at it; then the 1785 Guilford County estate for Michael and John.
The Walker generations between Daniel Jonah and 1737
Daniel Jonah Walker (b. 1890) and his mother Nancy Jane Higginbotham are Confirmed. His father, William B. Walker (1857–1914) S787, is Probable confirm or correct →: Daniel is in his parents' house at Palacios in 1910 S406. Above William, Daniel Green Walker × Mary Jane Bross are now his probable parents Probable confirm or correct →: William's own 1914 death certificate names them S787, his mother lived with him in 1900 as Mary J. Daniels S808, and his brother Edgar's certificate gives Daniel's middle name S809. The online tree's joins back through South Carolina to Thomas Walker (1737–1802) of Lunenburg, Virginia have no source. So Thomas Walker hangs above an unproven gap, and everything above him is already quarantined as contradicted.
What would close it, in order — (1) the 1900 census household of William B. Walker is found, in Jackson County, not McLennan or Bosque: it has a son Jackson J., born March 1890, Daniel Jonah's month and year, but not his name S808, so a Texas birth or school record for Daniel Jonah is still wanted; (2) the 1860 or 1870 census of Daniel G. and Mary J. Walker at Columbia, California, with a son William B. His certificate gives his birth there on 11 April 1857; the birthplace line reads “Columbia, Calif.” (re-read from the image on 27 September 2026) S787, and the online tree puts Daniel Green's 1873 death at Columbia, Tuolumne County. An earlier version of this plan looked for the family at Columbia, Brazoria County, Texas; no record here supports that. Each is a single index search and one image; the 1870 search for Daniel Green's household found nothing (checked 26 September 2026); by 1880 he was dead and his widow was Mary Dannels in Angelina County, Texas S809, and in 1900 Mary J. Daniels in William's household S808. A Texas death index entry for Willie Bernard Walker has no birthplace column, so it cannot settle this.
How Mons Selland got from Røldal to Dakota
The parish is settled. Four entries in the Røldal church books carry the line Confirmed. Mons was born 9 May 1868 and baptised 21 June 1868, son of the farmer Thore Thoresen Selland and Ingeborg Ivarsdatter S74; the out-migration register enters him in 1889 for "Dakota, Amerika" S75. His father's 1820 baptism S76 and both parents' burials in 1912 S77 close the two generations above him. Selland is the farm, not the family.
What is still open — the crossing itself. A church out-migration entry names no ship, port or sailing date; that needs an emigrant-ship index or a US arrival record, then his naturalisation and the move to Clear Lake, Deuel County. One lead worth an hour: four Lone people from Røldal leave in the same 1889 cohort, one a woman born 1870 — and Mons married Julia Lone, born 1870. The register writes her given name as Gyri, so the identification is Probable at best; the Deuel County marriage record would settle it.
Anna Silva's maiden name
Anna "Annie" L. Silva, Frank Silva Sr's wife, is on four censuses, a death index and a headstone, and not one of them gives her maiden name Not found.
What closes it — her 17 December 1938 Alameda County death certificate, which carries the maiden name and her parents. The printed California Death Index gives its state file number, 71446 S136 — the number to order it against from California Vital Records. One concrete step from closing. Second route: the 1984 Stanislaus County death certificate of her daughter Mamie C. Silva, whose Modesto Bee obituary names no parents.
John H. Timberman's father, Samuel
John H. Timberman (c. 1826) is Confirmed — his son's 1919 death certificate S93 and the 1858 Upper Alloways Creek birth return S374 both name him. His 1857 Swedesboro marriage return calls him "son of Samuel Timberman", so Samuel is Probable on one record. John's mother is not named anywhere, and the 1850 New Jersey index has no matching Timberman household.
What closes it — a Samuel Timberman household in the 1830, 1840 or 1850 census of Gloucester or Salem County, New Jersey (or across the river in Pennsylvania), with a boy of John's age; or Samuel's will or estate in Gloucester County.
Peter Hines Martin's parents
Peter Hines Martin (1793–1849), a shoe-store keeper at Gallatin, Tennessee, is on records, and his 1849 will S172 puts his son Joseph D. Martin in the children's division. His own parents, John and Sarah Martin, are named on his grave S105 and nowhere else.
What closes it — a Virginia record of about 1793 for a John and Sarah Martin with a son Peter: a will, an estate division, or a tax list that follows the family to Davidson or Sumner County, Tennessee, where Peter married Jane Bell in 1818 S106.
Whether James Brock Senr is James Brock's father
James Brock Senr is on the same 1830 Harlan County, Kentucky census page as James Brock (b. about 1804), aged 50–60 S462. That is the whole case Probable, and a thin one: neighbours sharing a name are not proof of father and son. James and Betsy (Howard?) Brock themselves rest on their daughter Elizabeth Hoskins's 1911 death certificate S454.
What closes it — Harlan County, Kentucky tax lists of the 1820s–30s, which list "Sr" and "Jr" side by side, or a deed or will of James Brock Senr naming his children.
Henry S. Roberts's parents
Henry S. Roberts (b. Nov 1854, West Virginia), a Wheeling glass blower, is confirmed as Anna Blanche Roberts's father. His parents were both born in Ohio, per the 1900 census, and are not named anywhere yet.
What closes it — his own death certificate (Ohio or West Virginia), which names parents; failing that, the 1860 census, which may show the family on either side of the river, since West Virginia was part of Virginia until 1863.
Antone Rose's island of origin
The 1900 and 1910 censuses say only "Portugal" for Antone Rose (b. Feb 1840); the 1920 census says Azores S314 but names no island. He died between April 1910 and January 1920. If he was also from Faial, like Frank Silva Sr, the Silva–Rose case for one island community strengthens; if not, it weakens.
What closes it — his naturalization record, or his Alameda County death record of 1910–20.
Mildred Rose's second marriage
Her death is found — 7 October 1960, Alameda County, as Mildred A Springer S316 — so she remarried after Frank Silva Jr died in 1924. Her granddaughter Patricia (Proeger) Collins names him William Springer (1 October 2026), and the 1950 census finds Mildred and William Springer, both born in California, in Oakland. The Springer marriage itself has not been seen.
What closes it — the Alameda County marriage record. The 1927 "Mildred Silva b. abt 1898" and 1930 "Mildred A. Rose" index candidates need checking against the images.
The 1981 divorce
Bill Proeger and Nancy Coffman's 1964 marriage is now on a record; the 1981 divorce still rests on the family obituary alone. It is not in the Texas Divorce Index 1968–2011.
What closes it — the decree at the Travis County District Clerk, if it was filed there.
Margaretha Bullinger's baptism
Margaretha Bullinger of Eichenau married Johann Caspar Pröger at Triensbach on 12 March 1737 S599; the register does not name her parents. One candidate baptism exists — an Anna Margaretha Bullinger, 12 September 1712 at Lendsiedel — but that entry calls her family alhier (of Lendsiedel), not of Eichenau, so it is held, not attached. The Bullingers at Triensbach turned out to be a different family, from Oberspeltach. Her age at death in 1762, which would have tested the candidate, is illegible on the burial entry.
What closes it — a page-by-page read of the Lendsiedel baptisms c. 1708–18 for an Eichenau Bullinger, and the Lendsiedel banns entry for March 1737, which may name her father.
Who Hanß Pröger of Seibotenberg was
Hanß Pröger, farmer at Seibotenberg in Michelbach an der Heide parish, is the earliest person on Earl's side — and he is named once, as the groom's father in the 1704 marriage S607 Probable. His dates, his wife and his parents are unknown. Three other Pröger households farm at Seibotenberg between 1709 and 1726 S610, probably his other sons.
What closes it — the Michelbach an der Heide burials after 1704 for a Hanß Pröger with his age, and the parish's marriages of the 1660s, which would name his wife and father.
What became of Georg Melchior Julius's brothers
Georg Melchior Julius Prüger had a full brother, Johann Michael (b. 15 April 1797), and a half-brother, Georg Friedrich, born 27 February 1807, ten weeks after their father's death. Both baptisms are read from the register images. Nothing is known of either man after childhood.
What closes it — the Triensbach and Kirchberg marriages and burials of about 1815–45, and the Württemberg emigration index under both names. Low value to the main line; worth doing only for cousins.
The Peters and Hermann generations in Mecklenburg
The 1910 census of Vernon, Kay County, Oklahoma, read from the image on 26 September 2026, puts Benjamin LaRoy Peters, aged 4, in the household of Benjamin and Edna B. Peters S659, and the 1903 marriage return names Benjamin's parents as John Peters and Sophia Hermann. What is still open is Mecklenburg: John C. J. Peters's birth in 1834 at Müritz, and his parents Gottlieb Peters and Catharina Schumacher at Kirch Grubenhagen, rest on the online tree alone.
What closes it — the Kirch Grubenhagen and Moltzow church books for a John (Johann) C. J. Peters born 26 October 1834; the Iowa state census of 1885 or 1895 for John and Sophia with Benjamin at home would add a second record naming them together.
Thomas Alan Caffey's death certificate
A theoretical gap rather than a real one. No single government document names both his parents: Texas death records after 2000 are in no free index. The 1950 census and two obituaries name them instead, three times over.
What closes it — the Travis County certificate from Texas DSHS, about $20. Low priority. Better value: ask Kristin, who will have photographs and can probably name the Wimberley house in a sentence.
Louise's grave
Her 2008 obituary states interment at Forest Park Lawndale Cemetery, Houston, where her husband lies in Section 27 Lakeview Gardens. No memorial exists for her there. All sixteen Coffman memorials at that cemetery were checked, under Coffman and under Timberman.
What closes it — cemetery office records, or creating the memorial. ⚠️ A Thelma "Sissy" Bradford Coffman (1906–1995) is buried in the same section; her relationship is unknown, and she should not be assumed to be a second wife — Louise's obituary says she and Robert were married fifty years, to his death.
What is left on the Portal to Texas History
The Portal sweep of 13 September 2026 is finished and its pages are written up in
the Documents section. Portal search is behind an
ALTCHA human-attestation check-box again; ocr.txt and
metadata.untl.xml still answer. Ticking the box asserts that a person is present,
so it is not automated around: it needs one click by a person.
Earl's side. Whether Robert W. Coffman got the Pasadena acting-superintendent job: in August 1961 he was one of two deputy superintendents named as candidates S78; a Pasadena or Houston paper from the first two weeks of September 1961 settles it.
Kristin's side. Narrower Kretschmar queries — Willie Frank Kretschmar, Hattie "Liessi" Ladusch (her gravestone reads Elizabeth), Lucille Lorine Kretschmar. The bare surname returns 842 results, and the 62 pages already captured are an unrelated New Ulm, Sealy and Bellville family.
Recently resolved · 10 questions closed since 17 September 2026
- 26 Sep — When Iona Belle Brock married William Bowles: 24 December 1880, the date in the family typescript S72, stays Confirmed. The June 1880 census, where she is still single at home S166, fits a December wedding. Her card.
- 26 Sep — Eva Barbara Aufläder's baptism: found, 28 April 1737 at Gaggstatt, daughter of the smith Johann Georg Aufläder of Mistlau. Her card.
- 25 Sep — Minnie Kochendörfer's 1867 crossing: closed as unfindable. Hamburg, New York and Baltimore lists were searched and are clean negatives; she most likely sailed from Bremen, whose pre-1920 lists are largely destroyed. Three of her brothers' crossings were found instead. Her card.
- 24 Sep — The 1964 marriage: 20 August 1964, Harris County, Texas, licence 0302006, from the county register image S199.
- 24 Sep — John H. Timberman's father: named as Samuel Timberman on the 1857 marriage return (Probable). The question moves up to Samuel, above.
- 23 Sep — Mildred Rose's death: 7 October 1960, Alameda County, as Mildred A Springer S316.
- 19 Sep — Frank Silva Sr in the 1888 Horta passport register: read in full; he is not in it. The parish question stays open by another route, above.
- 17–18 Sep — The Timberman and Martin lines above Louise: John H. Timberman × Rebecca Anna Harris S93; Nannie Chenault's parents S96; Peter Hines Martin × Jane Bell S105S106. The line.
- 17 Sep — Which Julius Proeger holds which patent: all six are the father's, Julius Karl Proeger (1844–1924), a molder; the son was a clerk. The patents.
- 17 Sep — Anna Blanche Roberts's parents: Henry S. Roberts and Rose A. Williams S90S91S92. Her card. ⚠️ Hers is a different Roberts family from that of Earl Theodore Roberts (b. ~1918, Massachusetts; son of Leon H Roberts and Bessie H Henderson), Aileane Marie Silva's third husband and Earl's step-grandfather. They share a surname and nothing else; do not merge them.
Loose ends that belong to nobody yet
Sharon Lynn Proeger (born McLennan County, Texas) and Dale Luther Proeger (born Washington County, Texas) are the only other Proeger births in Texas history. Both are invisible online and therefore probably living. Their relationship to this family is Not found, and no relationship should be assumed S3.
A large, well-documented and apparently unrelated Proeger population exists in Ohio (Wooster, Killbuck, Orrville), Kansas (Emporia, Natoma), Iowa (Belmond) and Illinois (Sterling). None has been connected to this family, and none should be without a corroborating record.
Both families
Add to this record
Know something this record doesn't?
Corrections, missing names, dates, stories, and above all documents — a photograph with writing on the back, a family Bible page, an obituary clipping — are all welcome. Tell us how you know it and it goes into the record as family testimony, kept separate from what a document proves until one is found. Earl reads every submission personally and nothing appears on this page until he has reviewed it.
One request before you write: this page includes living people, who appear in full, and minors, who appear by name only — no dates, no schools, no places. Please don't send birth dates, addresses, schools or contact details for any child.
JavaScript is switched off, which is fine — the button above still works. You'll land on a plain confirmation page instead of seeing the reply here, and you can use your browser's back button to return to the record.