Checked against the record. · Checked 2026-10-04
Corrected 2026-10-04. A version of this story that circulated said Marzio took over the Houston museum from Philippe de Montebello. He did not: the Handbook of Texas records de Montebello as director from 1969 to 1974, then 'He was succeeded by William Agee', who left in 1982, the year 'Peter Marzio became director'. The likely source of the error is a biography site that says de Montebello's move to the Metropolitan Museum opened the job in 1982.
The football scholarship is Marzio's own phrase, in the Houston Chronicle of 19 March 2006. A reader who coached at a Division III school pointed out that Division III schools may not award athletic scholarships. Division III was created in 1973, after Marzio's time, and no record from Juniata itself saying what his aid was called has been found, so the page gives the scholarship in his words rather than as the college's record.
The Goya story is told from his own 2006 account; the Boston Globe obituary independently confirms the projected image in a lecture, the Frick visit, and his poor academic record in high school. A collection figure was left out because two sources disagree on the starting number (14,000 against 13,000 works).
Checked against the record. · Checked 2026-09-27
Checked against the record. · Checked 2026-09-26
Checked and confirmed. · Checked 2026-09-25
Published without a source list. Every figure in it has now been read against Welch's 1909 history and all of them hold exactly: the seven thousand gallons a day imported from Europe by way of Philadelphia, the two thousand ties treated daily, the five hundred thousand on hand, Carl W. Tiffany as superintendent and about fifty men. Welch, writing in 1909, calls the plant only a few months old and the process "somewhat of an experiment". Checked 1 September 2026.
Checked and confirmed. · Checked 2026-09-24
Published without a source list. Every claim has now been read against Welch's 1909 history and all of it holds, including the date, the two dynamos in the engine room of the brick works, the day and night current, and the phrase about a night long to be remembered, which is the book's own. Checked 1 September 2026.
Checked and confirmed. · Checked 2026-09-23
Published without a source list. Every claim has now been read against the 1991 county industrial survey and holds, including the architectural description and the count of seven built and three surviving. The superlative needed the most work, because saying a thing suffered the worst losses means counting all its rivals: the survey gives all six house types, and the Double Mountain House lost four of seven where the next worst lost two of twenty-three. One thing the original leaves out is worse still - Kistler also had twelve four-family buildings, and not one of them survives. Checked 1 September 2026.
Checked and confirmed. · Checked 2026-09-21
Published without a source list. Every claim is in Africa's 1883 county history, which devotes a chapter to the killing and prints the deposition sworn at Paxton on 19 April 1744. Worth knowing: Welch's 1909 town history tells it differently, saying Armstrong's wife and children were murdered in 1752 and that he lived to avenge them. That is the "Captain Jack" legend, and Lytle's 1876 county history had already headed it "a popular error" and set out its unreliability. The post follows the documented account. Checked 1 September 2026.
Checked against the record. · Checked 2026-09-20
Checked and confirmed. · Checked 2026-09-18
Published without a source list. Every claim in it has now been read against the books it names and all of them hold, in the wording the original gives them: Welch 1909 for Pollock and for the "good authority" footnote naming John Dougherty, Lytle 1876 for the 1850 laying-out and for the four mountains, and the 1937 sesquicentennial book for both the 1840 date and the postmaster's $6.59. One thing the check added: the 1937 book itself says Pollock was appointed postmaster on 1 April 1833, seven years before the same book says the town was laid out, so even that single source does not settle it. Checked 1 September 2026.
Checked against the record. · Checked 2026-09-16
Checked against the record. · Checked 2026-09-15
Checked against the record. · Checked 2026-09-14
Checked against the record. · Checked 2026-09-13
Checked against the record. · Checked 2026-09-12
Checked against the record. · Checked 2026-09-11
Checked against the record. · Checked 2026-09-10
A claim corrected by the town's own nomination. This page opened by calling 1908 the year “when the Pennsylvania Railroad's main line was Mount Union's front door”. By 1908 it was not. The National Register nomination for the historic district records that a high line viaduct was built over the old Pennsylvania canal bed in 1906 and 1907 specifically so the main line could bypass the town with its express trains, and that the original right of way beside the station had become a yard siding for transfers to and from the East Broad Top. The station was a junction, not a main line stop, and that is the more interesting thing anyway.
The same correction was made on the other page about this depot on the same day. A fix that reaches one of two pages about the same building is not a fix.
Unchanged and still documented: the ganister before the rails, the 1899 Haws plant as the first works in the country built purely for refractory brick, and the October 1915 report of the railroad's spending, the new tower and the double siding to Mapleton.
Checked against the record. · Checked 2026-09-10
A frequency claim stays out, and now there is a reason rather than a shrug. An earlier version of this page said dozens of trains a day passed here. That was removed for want of a source, which is a weak reason on its own. The district nomination has now been read for it: across 186,000 characters it mentions railroads fifty two times and stations twenty times and gives no figure for train frequency anywhere. It does give something better, and something that cuts against the claim. The high line viaduct of 1906 and 1907 carried the main line and its express trains around the town, and the right of way beside this building had by then become a transfer siding for the East Broad Top. Whatever traffic this platform saw, the through expresses were not part of it.
ADDED, for the same reason: the high line, the transfer siding, and the position of the upgraded main line beyond the district's northern boundary.
Unchanged, because both are documented: the ganister before the rails, from the two plant nominations, and Welch's account of Dibert and Haws coming to look at the mountain.
Checked against the record. · Checked 2026-09-10
Every number on this page now has a source, and it is one document. The page carried six employment figures and a chain of ownership with nothing behind any of it. All of it is in Nancy S. Shedd's inventory of the county's historic engineering and industrial sites, prepared for the National Park Service in 1991, on pages 161 and 162: sixty-two women in 1922, the Greif purchase that same year, 314 in 1938, 591 by 1948, 725 by 1959 and “about 100 more than the combined work force of the two refractories”, Genesco in 1971 and the closure of both plants in 1974. The page's figures were accurate. They simply could not be checked by a reader, and now they can.
ADDED from the same entry: 174 employees by 1928, which the page skipped, and the dimensions, which make the growth concrete. The plant started on a lot 120 feet by 40 and the surveyed building is 450 by 120.
ADDED, and it is the better half of the story: the survey's own explanation of why the workforce was almost entirely female. Three refractory plants and the East Broad Top yards employed a large workforce of men, which “gave the company a large pool of females from which to draw its workers.” The page had the fact and not the cause, which left it looking like a coincidence.
ADDED: what happened to the building. Dallco Industries of York has made women's garments there since 1978 and was expanding when the survey was written. A page that ends in 1974 reads as though the place were derelict.
One discrepancy inside the source, noted and not hidden. The survey's header gives a construction date of 1919 while its own history says the factory opened in 1920. Those are not necessarily in conflict, and the page says opened in 1920, which is what the narrative states.
Checked against the record. · Checked 2026-09-10
Dates put back, and they were this site's own doing rather than the scanner's. The passage below promises to print the record exactly as it appears, and it was doing so apart from one thing: sentences kept losing their year. “In an imposing brick church” for “In 1877 an imposing brick church”, “organized in by Rev. Richard Allen” for “organized in 1816 by”.
The cause was a rule meant to strip page numbers where a running head had run into the text. It removed any two to four digit number standing between two lowercase words, which is what a page number looks like there, and also what a year looks like in a sentence.
The rule now refuses to touch a number between 1600 and 1999, and it still removes page numbers, which was checked both ways. The dates on this page were restored one at a time from the full text of the book itself.
Checked against the record. · Checked 2026-09-10
Dates put back, and they were this site's own doing rather than the scanner's. The passage below promises to print the record exactly as it appears, and it was doing so apart from one thing: sentences kept losing their year. “In an imposing brick church” for “In 1877 an imposing brick church”, “organized in by Rev. Richard Allen” for “organized in 1816 by”.
The cause was a rule meant to strip page numbers where a running head had run into the text. It removed any two to four digit number standing between two lowercase words, which is what a page number looks like there, and also what a year looks like in a sentence.
The rule now refuses to touch a number between 1600 and 1999, and it still removes page numbers, which was checked both ways. The dates on this page were restored one at a time from the full text of the book itself.
Checked against the record. · Checked 2026-09-10
Dates put back, and they were this site's own doing rather than the scanner's. The passage below promises to print the record exactly as it appears, and it was doing so apart from one thing: sentences kept losing their year. “In an imposing brick church” for “In 1877 an imposing brick church”, “organized in by Rev. Richard Allen” for “organized in 1816 by”.
The cause was a rule meant to strip page numbers where a running head had run into the text. It removed any two to four digit number standing between two lowercase words, which is what a page number looks like there, and also what a year looks like in a sentence.
The rule now refuses to touch a number between 1600 and 1999, and it still removes page numbers, which was checked both ways. The dates on this page were restored one at a time from the full text of the book itself.
Checked against the record. · Checked 2026-09-10
Dates put back, and they were this site's own doing rather than the scanner's. The passage below promises to print the record exactly as it appears, and it was doing so apart from one thing: sentences kept losing their year. “In an imposing brick church” for “In 1877 an imposing brick church”, “organized in by Rev. Richard Allen” for “organized in 1816 by”.
The cause was a rule meant to strip page numbers where a running head had run into the text. It removed any two to four digit number standing between two lowercase words, which is what a page number looks like there, and also what a year looks like in a sentence.
The rule now refuses to touch a number between 1600 and 1999, and it still removes page numbers, which was checked both ways. The dates on this page were restored one at a time from the full text of the book itself.
Checked against the record. · Checked 2026-09-10
Services replaced with the ones the library actually names. This page listed things the library was assumed to offer rather than things it says it offers. Describing a service that does not exist creates a promise nobody has agreed to keep, and it is the same error as inventing a fact.
What the library's own site names, and what this page now carries: Summer Programs, Library by Mail, Ask A Librarian, makerspace studio appointments, interlibrary loan requests, Power Library, tax resources, voter registration and a Friends of the Library. The address is given as 330 Penn St., Huntingdon, PA 16652.
WHAT CAME OUT, and why it is not the same as saying the library lacks these things. Story hours, public computers and air conditioning are not named anywhere on the pages read. A library homepage does not list everything a library has, so this is not evidence the library lacks them; it is simply not something this page can assert. If the library confirms them, they go back in.
Hours are deliberately not printed. A third party directory listing gives one set of opening hours and the library's own page gives a different set. The library's page wins, and hours change quietly, so this page links to them rather than copying them.
Checked against the record. · Checked 2026-09-10
Dates put back, and they were this site's own doing rather than the scanner's. The passage below promises to print the record exactly as it appears, and it was doing so apart from one thing: sentences kept losing their year. “In an imposing brick church” for “In 1877 an imposing brick church”, “organized in by Rev. Richard Allen” for “organized in 1816 by”.
The cause was a rule meant to strip page numbers where a running head had run into the text. It removed any two to four digit number standing between two lowercase words, which is what a page number looks like there, and also what a year looks like in a sentence.
The rule now refuses to touch a number between 1600 and 1999, and it still removes page numbers, which was checked both ways. The dates on this page were restored one at a time from the full text of the book itself.
Checked against the record. · Checked 2026-09-10
The page had the wrong cause, and so did this site's own list of refuted claims. The page said the stone house fell “just as the railroad age overtook the ferry age”, which puts the ferry's end in the late 1850s. Three county histories put it a generation earlier and blame the canal. Welch in 1909 says the ferry flourished before the canal was built. The Commemorative Biographical Encyclopedia of 1897 says it plainly: the ferry “was operated for many years, and not until the building of the canal, about the year 1830, was it abandoned.” The same book, writing up James Drake separately, says he ran it “until about 1830, when the construction of the canal through the Narrows took place, of which he was the first foreman.”
This site's own list of refuted claims had “the canal replaced the ferries” listed as a refuted claim, on the reasoning that a ferry crosses a river and a canal runs along a valley, so one cannot replace the other. That reasoning is tidy and it is not evidence. Nobody had looked in the county histories, and all three of them say the canal is what ended this ferry. The entry is gone and the correction is on the page.
A genuine disagreement between sources, left as one. The page said the pioneer stone house by the ferry was “destroyed by a train wreck around 1858 or 1859”. That is in Welch, and Welch hedges it twice: “probably 1858 or '59”, and “according to information received from D. E. Shaver of town”, a man's recollection half a century later. Africa's 1883 history says the house “was burned several years ago”, and Jordan's 1913 history says it was “subsequently destroyed by fire”. Two books say fire, one says a train ran into it. The page no longer states either as fact.
Confirmed and dated exactly. The charter really was 1848, and the day can be given: meetings were held along the route early in 1848, one at Scottsville on 28 February, and on 25 March the legislature incorporated the Drake's Ferry and East Broad Top Railroad Company. Lytle adds that no steps were ever taken to build it, and that Mount Union was laid out in 1850 as the transfer point between that railroad and the canal.
ADDED: the 1787 road from the Burnt Cabins, twenty four miles and forty nine perches, and the route the ferry served, named town by town from the encyclopedia rather than summarised.
Sources now openable. This page cited the Juniata Valley encyclopedia with no link. Both it and Jordan's history run to several volumes, so the right volume was established by searching the full text of each rather than by picking one: the ferry passage is in volume 1 of the encyclopedia, and the railroad charter passage is in volume 1 of Jordan.
Checked against the record. · Checked 2026-09-10
Corrected on geography, on direction, and on who is making the claim.
A canal is not a ferry, but it did end one, and this note overstated the correction. The original page said the canal “had replaced the earlier ferries, offering faster passage than the river crossings that came before”, and ended “travelers depended on the canal boats to cross.” The second half is still wrong: a canal boat does not carry you across a river, and nobody crossed the Juniata on one. But this note went on to say that a canal cannot replace a ferry at all, and that is not right either. Three county histories say the building of the canal, about 1830, is exactly what ended Drake's Ferry a few miles from here, and the ferryman became the canal's first foreman. The mechanism was wrong in the original sentence; the connection was not. See the page about Drake's Ferry, where it is set out with the sources. The original said the canal "had replaced the earlier ferries, offering faster passage than the river crossings that came before", and ended "travelers depended on the canal boats to cross." A ferry carries you across a river; a canal carries you along a valley. They are different jobs and one does not replace the other. Both sentences are cut.
Direction. The original had Dickens going "down the Juniata River." He was going to Pittsburgh, which from Harrisburg is west and UP the Juniata, and he was on the canal beside the river, not on the river. His route, from his own chapter headings, was Baltimore to York to Harrisburg by coach, then the canal boat over the Alleghenies to Pittsburgh.
Whose claim is it. This matters most. Dickens does not name the Juniata anywhere in American Notes, not once. The local connection rests entirely on Charles H. Welch asserting in 1909 that Dickens "came this way", and Welch is geographically right, because the Main Line Canal from Harrisburg to Hollidaysburg ran up the Juniata past Mount Union and there was no other way for that boat to go. But the site was presenting a local historian's inference as though Dickens had recorded it himself. It is now attributed to Welch, and the quotation is attributed to Dickens.
Quotation restored from the book. The original quoted a single trimmed sentence by way of Welch. Welch's page is set in badly worn type and reads "Despite oddities" where Dickens wrote "And yet despite these oddities", and prints "1 heartily" for "I heartily". So the passage is quoted from Dickens's own text and Welch is cited for the local claim rather than for the words. The fuller passage is also simply better than the fragment that was there.
CUT for want of a source: "The stone bridge visible here was built in 1906." No source was found for the date. It is not contradicted either, and it stays on the list of things to look for.
The four-miles-an-hour packet speed and walking the towpath are Dickens's own: he describes the brisk walk on the towing-path in the quoted passage.
Checked against the record. · Checked 2026-09-10
Services replaced with the ones the library actually names. This page listed things the library was assumed to offer rather than things it says it offers. Describing a service that does not exist creates a promise nobody has agreed to keep, and it is the same error as inventing a fact.
What the library's own site names, and what this page now carries: Summer Programs, Library by Mail, Ask A Librarian, makerspace studio appointments, interlibrary loan requests, Power Library, tax resources, voter registration and a Friends of the Library. The address is given as 330 Penn St., Huntingdon, PA 16652.
WHAT CAME OUT, and why it is not the same as saying the library lacks these things. Story hours, public computers and air conditioning are not named anywhere on the pages read. A library homepage does not list everything a library has, so this is not evidence the library lacks them; it is simply not something this page can assert. If the library confirms them, they go back in.
Hours are deliberately not printed. A third party directory listing gives one set of opening hours and the library's own page gives a different set. The library's page wins, and hours change quietly, so this page links to them rather than copying them.
Checked against the record. · Checked 2026-09-10
The count is back, because it was right all along, and taking it out was this site's mistake rather than the original post's. That is worth saying plainly.
An earlier check found that the post's closing sentence contradicted its own arithmetic: it gave 14 officers and 127 enlisted men lost to combat, 68 to disease, and then said disease had killed more. It caught that correctly. But it then removed the whole casualty count and left a note saying the number would stay out until somebody put a source under it. The source is William F. Fox's Regimental Losses in the American Civil War, published in 1889 and still the standard reference for exactly this question. Its table for the Fifth Pennsylvania Reserves gives 14 officers and 127 men killed or died of wounds, 68 dead of disease, accident or imprisonment, and a total enrollment of 1,046. Fox's own line beneath the table reads “141 killed = 13.5 per cent”. 141 and 68 make 209. Every figure in the original post was correct and only the conclusion was wrong.
Cutting a true fact for want of a source is the error that feels like care while you are making it. It is the third time this project has caught itself doing it.
A different sentence on this page was actually wrong. It said four companies were recruited partly out of Huntingdon County. Africa's 1883 county history says two, and names them: the Huntingdon Infantry under Captain Andrew S. Harrison, which became Company G, and the Scott Infantry under Captain George Dare, which became Company I.
A distinction Fox makes and the page did not. The page listed Gettysburg and the North Anna among the regiment's battles. Fox separates the actions in which the regiment lost men from those at which it was merely present, and Gettysburg, Malvern Hill, Mine Run, the North Anna and Totopotomoy are in the second list. The page now keeps the two apart.
ADDED, and it is the reason to run this post at all: both Huntingdon County officers named in the 1883 history were killed. Fox records Lieutenant Colonel George Dare, in command of the regiment, killed at the Wilderness, and Major Frank Zentmyer among the killed at Fredericksburg. Dare had raised the Scott Infantry; Zentmyer had taken over that company from him.
A spelling note, left as found: the 1883 county history spells him Zentmyer and Fox spells him Zentmeyer. The county's own book is followed here.
Checked against the record. · Checked 2026-09-10
Checked against the record. · Checked 2026-09-09
Partly confirmed, and the page now claims only the part that is.
Confirmed at the Historic American Buildings Survey report on the Town of Mount Union, HABS No. PA-5974: a footnote reads “Ivan Phillips, interview by author, Mt. Union, Pa., June 11, 1991.” That establishes both the interviewing and the 1991 date, in the survey's own citation.
Not found here: the other eight names the page listed — Fred S. Beatty, Josh Booker, Marge Cassatt, Mike Dimoff, Sylvia Hoenstine, Carl Peduzzi, John Curry and Robert Wagner. Neither the Mount Union data pages nor the Kistler data pages name them. They very likely come from the fuller documentation or from Kim E. Wallace's Brickyard Towns of 1993, and they are not asserted on the page while the source for them is unopened. Not found is not the same as not real.
A name trap avoided. The Kistler report does contain a Hoenstine — William Ray Hoenstine, a resident who with Superintendent Edgar Kurtz bought the first two company houses in 1944. That is not Sylvia Hoenstine, the interviewee. Same surname in the same small town, thirty years apart, and treating the one as evidence for the other would have been exactly the kind of match that reads as confirmation and is not.
Also in those footnotes, and it corroborates another page here: “Nancy Shedd to author, April 9, 1992”. Shedd is the county historian whose 1988 Historic American Engineering Record report underlies the National Register nomination for the Mount Union Refractories site.
Checked against the record. · Checked 2026-09-09
Source upgraded from Wikipedia to two county histories already in the corpus, and the page gained the thing it was missing: evidence.
Confirmed, Africa 1883, on the trade itself. At a conference held by Weiser with Shikellamy and other Indians at Chambers' Mill in June 1747, complaint was made against the traders James Dunning and John Powel; Dunning was accused of stealing forty-seven deer-skins and three horses “upon the heads of Joniady (Juniata) River”. The page had been asserting that traders came first and dealt in skins; it can now name two of them, and date them to 1747.
Confirmed, Lytle 1876, on the road: the traders' route ran “through the Tuscarora Valley, Shade Gap, Black Log, Aughwick, Woodcock Valley, Hartslog Valley, Water Street, Frankstown, Hollidaysburg, and crossing the Allegheny mountains at or near Kittanning Point”, and “it was this trail that gave Huntingdon county its early importance. It was the great highway between the east and the west.” That replaces the vaguer claim that the paths are under some of our roads with the route itself, named valley by valley.
The Great Cove correction is kept but its numbers are cut. A correction earns more scrutiny than the thing it corrects, not less. The geography is sound and is the point: Great Cove is McConnellsburg, in what is now Fulton County, and the corpus places the Great Cove settlements in disputed land outside this county. But the casualty figures the page carried - around fifty killed or captured, and Sheriff Potter reporting forty-seven - are not in any book in this corpus and were not sourced anywhere opened. A correction that itself carries unsourced numbers is not finished, so the numbers are out and the geography stays.
A coincidence worth naming, because it could have produced a false citation. Africa's 1747 passage contains the figure forty-seven, for deer-skins stolen from an Indian trader. The page's Great Cove paragraph also carried a forty-seven, for people. They have nothing to do with each other, and a keyword search for the number would have joined them.
Checked against the record. · Checked 2026-09-09
Sourced for the plant, not for the photograph's decade. The page had no source at all. Welch's 1909 account of the tie-treating plant, page 226, establishes what the works were and at what scale, and that is now on the page with his figures.
What is still not sourced: that this particular photograph is from the 1950s. No source opened here dates the image. The claim that ties were still being treated here four decades after Welch wrote is a reasonable reading of a yard photograph, and it is not proven by anything cited. If the image turns out to be another decade, the Welch material stands and only the caption moves.
Checked against the record. · Checked 2026-09-09
Two dates cut, both unsourced, and the same year in both.
The original said "The Shapiro family opened their first movie house in 1915, the same year the powder plant opened." Two searches of the Pennsylvania newspaper archive and a search of the nine books in the local corpus produced no source for either. Neither is contradicted. Both are simply not shown, and a year that carries the whole sentence has to be sourced or dropped.
On the powder plant the record does set a bound, and it does not help 1915: the earliest evidence found for the Aetna Explosives works at Mount Union is March 1916, and it is already operating then with hundreds of men. That is consistent with a 1915 opening and equally consistent with 1914 or 1916. It cannot settle the year, so the page no longer names one.
Replaced with what the record does carry, which is a longer story than the one that was there. George Shapiro has his own profile among the businessmen in Welch's 1909 history, so the family were established merchants in Mount Union at least six years before the date the page was hanging everything on. The Mapleton centennial history records Shapiro and Sherr opening a Mapleton branch of their Mount Union store in 1924, and a Joe Shapiro appears on the 1937 county sesquicentennial committee.
The powder works detail is from the Harrisburg Telegraph of 25 March 1916, reporting the town's William Penn Highway boosters: they wanted a bridge over the Juniata so the men could walk to the Aetna works, "a little more than a mile below the town", from their homes. The 1,500 figure is from the same paper of 21 November 1916, reporting an order that would put that many men to work, with R. O. Carter as superintendent. Both are attributed to the paper rather than stated flatly, because a reported employment forecast is a forecast.
Still open: when did the Shapiro movie house actually open, and when did the Aetna works start? A Mount Union Times run would probably settle both, and this county's own papers are not in either archive searched here.
Checked against the record. · Checked 2026-09-09
The specific claim is not sourced, and the page now says so rather than asserting it.
The page said the company painted interiors in bright deep greens, blues, browns and greys, that residents complained, and that the company switched to beiges and pastels. Two federal documents were opened and searched: the Historic American Buildings Survey for the Town of Mount Union (PA-5974) and for the Town of Kistler (PA-5976). Neither mentions colours, beiges, pastels, or a complaint about paint.
Where to look. The Kistler survey names the publication the whole study fed into: Kim E. Wallace, “Brickyard Towns: A History of Refractories Industry Communities in South-Central Pennsylvania”, published in 1993 by America's Industrial Heritage Project and HABS/HAER. If the colour dispute is documented anywhere, that book is the first place to check. It was not available here.
What is sourced, and it is a bigger story than the paint. The Mount Union survey records that the houses were “solid frame structures, nicely painted and slate roofed” with interiors “painted and papered”, porcelain sinks and city water, and plans to have them “electrically lighted and scientifically sewered”. It quotes the aim: “the employees and their families should be comfortably and decently housed; and . . . they should be encouraged to keep their houses as nicely as possible.” Housing was part of what the companies called a programme of social and cultural transformation.
And the part the page was leaving out. The same programme “included a policy of curtailing alcohol consumption”, and “employees who flagrantly ignored admonishments to abstain were fired.” A page calling this “one of the gentlest labor stories” while omitting that the company could sack a man for drinking was telling half of it.
A misquotation of this site's own, corrected. An earlier version of this page rendered the workforce as “Poles, Greeks and Serbians”. The survey quotes the local newspaper of 1916 as “Slavish, Roumanians, Bulgarians, Italians, Greeks and Serbians” as well as “native Americans, and southern Negroes.” Poles are not in it; Roumanians, Bulgarians and Italians were dropped. The period terms are kept as the paper printed them because paraphrasing them hides what was actually said.
Checked against the record. · Checked 2026-09-09
Source upgraded from Wikipedia to two primary records: the Nobel Foundation's own API for the prize, and Phillips's autobiography published by the Nobel Foundation for everything about Juniata.
Confirmed at the Nobel Foundation: the 1997 Nobel Prize in Physics, awarded to Steven Chu, Claude Cohen-Tannoudji and William D. Phillips, one third each, with the citation "for development of methods to cool and trap atoms with laser light" - which is the wording the page now uses rather than a paraphrase.
Confirmed in his own words: he began at Juniata College in the autumn of 1966 and graduated in 1970. He describes Juniata as "a small school in Huntingdon, Pennsylvania, founded and strongly influenced by the pacifist Church of the Brethren."
ADDED, and it is much the better story: his parents were each the first in their families to go to college, both went to Juniata, and graduated in 1930 and 1936 without ever meeting - until a Juniata professor who knew them both suggested his father telephone a young Juniata alumna. His aunt, his sister and his younger brother all went there too. A page about a Nobel Prize was carrying only the graduate, and the county's actual claim is a family.
A name trap worth recording, because it nearly went the wrong way. "Juniata" appears twice in this autobiography meaning two different places. Phillips's father was born in 1907 in "Juniata, a community on the edge of Altoona" - a neighbourhood in BLAIR County, not the college and not this county. The connection was checked against the sentences that name Juniata College and Huntingdon, not against the word on its own.
Also from the record and not previously on the page: Phillips was born on 5 November 1948 in Wilkes-Barre, Pennsylvania, so he is not a native of this county. The page no longer implies otherwise; the Juniata connection is the real one and it does not need him to have been born here.
Checked against the record. · Checked 2026-09-09
Sourced, and the source is unusually direct: this newspaper printed the book the rest of this site is built on, and its editors wrote it.
Confirmed from the book's own title page: “Printed at The Mount Union Times Office. A. S. & C. H. Welch, Editors. Mount Union, Pa. 1909-1910.”
Confirmed from Welch's closing pages: “We started this historical sketch of Mount Union last summer. The first installment of it was printed in the Mount Union Times September 10, 1909. At the beginning it was expected that only a brief history would result and no thought of making a book was entertained.” The county history that this site cites on page after page began as a newspaper serial and grew because the material kept coming.
ADDED: the town had a second paper. Welch records the County Bulletin, run by Orthello Z. Shaver, which John S. Kirk entered in February 1900 and which had “since passed out of existence” by 1909. Kirk moved to the Times in October 1900 and was foreman there while the history was being set. A defunct rival is the kind of thing a town forgets, and it is on the record.
Not claimed: when the Times itself was founded. Welch never says, and nothing else opened here does. The page describes what the paper was doing in 1909 and 1910 rather than how long it had been doing it.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
ADDED, and it is the best documented thing on the page: the borough of Kistler was incorporated to keep a school out. The survey sets out the sequence and quotes the company president directly. By 1924 Wayne Township was renting three rooms above the company store for the school. The board wanted a portable, ready-made building of its own. R. P. M. Davis, then president, objected that it was neither adequate nor appropriate for a town of this reputation, and wrote that a “squatty, ready-made, ugly school building right in the center of Kistler would be a great misfortune.” To block it he had Kistler incorporated as a borough, separating its school administration from the township, and classes continued in the company's building. The survey cites the Mt. Union Times of 4 September 1924, under the headline “Kistler Will Become A Borough”.
A company name handled carefully. The survey calls him U. S. Refractories' president at this point, not Mount Union Refractories'. That is not a contradiction: the same company took the new name in 1922 after opening a plant in Clearfield County, two years before the school fight. The page says “the company president” rather than picking one of the two names and getting the year wrong.
Confirmed and unchanged: the store and school were one building, sited along the state road across Kistler Green from the community building, and it was demolished in 1990. The survey adds that a ranch house was built on the site.
ADDED: the 1936 brick school at the end of Cedar Street, built with Works Progress Administration help, and still the Kistler elementary school when the surveyors were there.
NOT USED, because the source will not carry it. The survey reports that the community building held a theatre, a poolroom and a dance hall and burned in 1933 without being replaced, and that one resident remembered visiting it near the end only to register to vote. That is a fine detail but it belongs to the community building, which is a different building from the one in this photograph.
Checked against the record. · Checked 2026-09-09
Sourced. This page previously cited an internal catalogue reference and a photograph credit, neither of which a reader can open, and it carried a population claim with nothing behind it: that Mount Union went “from about a thousand to over three thousand” between 1900 and 1910.
Confirmed, and better than the estimate. Welch's history, page 390: “When the plant was installed, in 1900, Mount Union had a population of 1086. It now has, including suburbs, upward of 3200.” The rough figures were right and the real ones are more interesting.
Two qualifiers kept, because dropping them is how a good number goes wrong. Welch's 3,200 is “including suburbs”, so it is not the borough count on its own. And his book is 1909, not 1910, so the span is nine years rather than the decade the page implied. The page now says both.
ADDED from the same passage, which is the reason the town grew: by 1909 there were two large fire brick plants, a large steam tannery, extract works, creosoting works and two railroads, employing more than a thousand men between them, where in 1900 work had only just started on the Ganister Works.
Checked against the record. · Checked 2026-09-09
ADDED, from the survey's own text and footnotes: the African American families in Kistler were recruited from the area around Saluda, South Carolina. That is not a general statement about migration. It names the place they were brought from, and it is the kind of detail a family in this town might still be able to match to a grandparent.
The hedge is the source's own and it is reproduced exactly. The survey says “although no strict segregation was reported in Kistler, most of the African-American families lived on Pine Street.” It would be easy to write that up as either a segregated town or an unsegregated one. The document supports neither. It reports no formal rule and an actual concentration on one street, and it rests on interviews with residents, Mary Alexander and Ressie Costlow of Kistler, recorded in 1989 and 1991. The page now carries the same hedge the surveyors used.
Confirmed, unchanged: the six house names come off Nolen's plan key, and the survey lists three double house designs and three one and a half storey single family designs. The Vermont Farm House is one of the single family three, which is consistent with the twenty seven figure on this page.
Spelling corrected across the Kistler pages: Mann and MacNeille.
Checked against the record. · Checked 2026-09-09
The same three errors as the sister page, which had been corrected while this one was left alone. A correction that reaches one page and not the other is not a correction, and every page on this site is now searched for each refuted claim.
WRONG, now fixed: “a coal freight”. Every contemporary account calls it a fast preference freight. The coal belongs to a victim rather than to the train — Chester A. Minds was a coal operator in the Clearfield region.
Number changed to the one a source actually gives: the page said “Seven of them were members of a single family”. The Evening Public Ledger says Minds “and six members of his family” died. The Harrisburg Telegraph's list of identified dead in that party runs to nine. The sources do not agree on a count, so the page now quotes the one that names its own figure and attributes it.
CUT for want of a source: “photographers sold real-photo postcards of the wreck across the state.” Real-photo wreck postcards were an ordinary trade in 1917, but nothing opened here says it happened for this wreck, and “across the state” is a scope claim needing its own source. No source was found, which is not the same as finding it untrue.
Checked against the record. · Checked 2026-09-09
Corrected, and one claim in the original was simply wrong.
WRONG, now fixed: the original called it a 'coal freight'. Every contemporary account calls it a fast preference freight. The coal in this story belongs to a victim, not to the train: Chester A. Minds was a coal operator in the Clearfield region. That is the likeliest route by which 'coal' migrated onto the freight.
CUT for want of a source: the original said 'photographs of the wreck were sold as postcards across the state.' Real-photo wreck postcards were an ordinary trade in 1917, but no source opened here says it happened for this wreck or that they sold statewide, and 'across the state' is a scope claim that needs its own source. No source was found for it, which is not the same as finding it untrue. It stays on the list of things to look for.
CUT: 'the deadliest day in Mount Union's history.' A superlative needs every rival counted and none has been. What the record does support is the Evening Public Ledger's own description of the crash as the most disastrous the Pennsylvania Railroad had suffered in years.
Tightened: the original said relief crews came 'from the Altoona shops.' The sources say a relief train set out from Altoona carrying Superintendent N. W. Smith of the Central Division with doctors and nurses. Altoona is sourced; the shops as the specific institution are not.
Numbers the sources disagree on, therefore left out: the infant son is 'Chester A. Minds, Jr., two weeks old' in the Harrisburg Telegraph and 'Owen Minds, aged three weeks' in the Democratic Watchman; Maud Minds is 28 in one and 35 in the other. The death toll is 'at least twenty' (Harrisburg Telegraph, AP), 'twenty persons, nineteen passengers and a porter' (Evening Public Ledger) and 'nineteen people instantly killed' (Democratic Watchman). The Ledger's precise formulation is the one used above because it accounts for the difference.
Confirmed: the date and the 12.08 a.m. time; Mercantile Express as train No. 6, out of Altoona at 10.35 p.m.; the express standing at Mount Union station with brake trouble while the crew tested airbrakes; dense fog; the rear sleeper split; all the dead in the car Bellwood; Chester A. Minds of Ramey, Clearfield County coal operator and Penn football star, bound for his wife's father's funeral; the powder-plant accommodation train pulling out just before; the Altoona relief train at one o'clock with Superintendent N. W. Smith, doctors and nurses, more physicians picked up at Huntingdon; and the carload of coffins strewn over the wreckage. Blame was placed at the time on freight engineer A. T. Cook of Harrisburg, who said the block signal at 'MO' tower showed white; the Interstate Commerce Commission and the Public Service Commission both investigated.
Checked against the record. · Checked 2026-09-09
Sourced. The page previously cited an internal catalogue reference and a photograph credit, neither of which a reader can open, and described the building only in general terms about the brick boom.
Confirmed from Welch: the Crum building stands on the corner of Jefferson Street and Railroad Avenue, on the site of the former Broad Top House; James Nevell's barber room was in it, opposite the Beers Hotel; and McKelvy and Fields, real estate and insurance agents, had their office there.
The building date is left as the late 1850s for the hotel it replaced, because Welch gives 1858 on one page and 1859 on another. Nothing opened here dates the Crum building itself, and the page no longer implies a decade for it.
Checked against the record. · Checked 2026-09-09
Sourced. The page previously cited an internal catalogue reference and a photograph credit, neither of which a reader can open, and it described the operation only in general terms.
Confirmed and greatly expanded from Welch, page 226, written while the plant was new: erected by the Pennsylvania Railroad Co. on their grounds northeast of town along the Juniata River; seven thousand gallons of creosote used daily, imported from Europe and stored in large tanks at Philadelphia; two thousand feet of railroad siding; a daily capacity of about two thousand ties; roughly five hundred thousand ties on hand; Carl W. Tiffany superintendent; approximately fifty men employed.
A hedged claim, carried as his hedge. Welch writes that “so far as we can learn, this process of tie treatment to creosote is somewhat of an experiment, this being the first industry of this kind built by the Pennsylvania Railroad Co.” That is a first-of-its-kind claim and he does not assert it flatly. Neither does this page: it is given as what he could learn, in his words. Upgrading a source's hedge into a fact is the specific way this kind of sentence goes wrong.
Dating. Welch's book is 1909 and he says the plant was erected “during the last year” and opened “only a few months ago”, which places construction in 1908 and opening in 1908 or early 1909. The page says what he says rather than fixing a year he does not give.
Checked against the record. · Checked 2026-09-09
Checked against the record. · Checked 2026-09-09
Citation corrected. An earlier pass on this page cited the nomination at npgallery.nps.gov/GetAsset/90000398_text. That address answers HTTP 200 with a 1,623 byte PNG placeholder for ANY reference number, including a deliberately invented one, so a reader following it would have got a broken image and no document. There is no digitized text asset at NPGallery for this reference, and the reason turns out to matter: the nomination is filed under the National Archives' withdrawn series. The 54-page bundle actually read is at NARA RG 79.
Confirmed from that bundle: organized 1911, plant completed 1912, third such concern in Mount Union in under fifteen years; quarries pushed out to Three Springs because the company was last on the scene; the East Broad Top serving both the quarries and the coal supply; and 'By 1919, Mount Union Refractories had expanded from eight to 29 kilns and was employing 885 men to satisfy the demand created by World War I.' The original's 'eight kilns to start, twenty-one more added' comes to the same twenty-nine and was right.
ADDED, and it is why the withdrawn filing matters: the same bundle holds the PHMC letter of 17 June 1996 and the reconnaissance survey behind it. After North American Refractories ended production in 1990 the property was sold or leased to a salvage company that dismantled the buildings for scrap. The survey of 3 April 1996 found only three buildings standing, due to come down that summer, and concluded the property 'has ceased to meet the criteria for listing in the National Register of Historic Places because the qualities which caused it to be listed have been lost.' The Bureau began the process of removal. So this plant was listed and then delisted, and the site must not imply it is still on the Register. The beehive kilns, buildings 'H' and 'I' on the survey map, are among the demolished.
Restored. This note previously recorded that the sentence 'Two former Harbison-Walker managers struck out on their own in 1911' had been cut as unsourced. That was an error on this project's part, not on the page's. The Historic American Buildings Survey of Mount Union states on page 4 that Mt. Union Refractories was organised by two former managers for Harbison-Walker, R. P. M. Davis and Clinton V. Hackman, and footnotes it. The claim is sound and it is set out in full on the page about Hackman. The contemporary record names R. P. M. Davis of Harrisburg as president and general manager, F. D. Halstead as secretary and treasurer, and C. V. Hackman as superintendent of the local plant. The nomination's own framing cuts against it: the company was 'last on the scene' and took the leftover quarries, which is not the story of insiders leaving with what they knew. No source was found for it, which is not the same as finding it untrue.
Hedge kept: 'silica brick capital of the world' is the town's own claim, which is exactly how both nominations phrase it. Kistler is across the river in Mifflin County and belongs under across-the-line.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
Refuted, and refuted by this site's own sourced page before any outside source was opened.
The original called 1936 "the most destructive flood ever recorded in the Juniata Valley." The page on the same flood at flood-of-36-devastated-hc already carried the Geological Survey's gage records and says the opposite: on two of the four Huntingdon County gages the June 1889 flood was higher. Aughwick Creek at Orbisonia, 18,400 second feet in 1889 against 16,400 in 1936. Standing Stone Creek, maximum known stage 9.38 feet on 1 June 1889. Two pages on one site were making opposite claims about the same flood, and only one of them had a source. That is the cheapest fact check there is and it should have been run first.
Causation softened. The original said the flood "hit so hard, across so many towns, that Congress passed the Flood Control Act that same year." The Act is real and was signed on 22 June 1936. The Corps of Engineers' own published history of the law credits the floods of 1935 AND 1936 together rather than 1936 alone, but every route to that document could not be opened here, so it is not cited: a source that was not opened does not go under a sentence. What IS opened and cited is the contemporaneous report of the $320,000,000 omnibus bill naming Pennsylvania among the states with emergency projects that could not wait. So the page now says the flood reached far enough to move Washington, which the record supports, rather than that it single-handedly caused the Act, which was not shown.
DATE. The flood keeps St. Patrick's Day in its name, and the naming is right, but readers should know the crest here came in the small hours of 18 March. The month is used rather than a bare 17 March so this page and the gage page cannot be read as contradicting one another.
Checked against the record. · Checked 2026-09-09
Sourced. The page had no source and said nothing about who the Floods were.
Confirmed from Welch, page 431: the Messrs. Flood were born near Nossville in the lower end of Huntingdon county and spent their early days there; J. D. Flood, the senior member of the firm, spent time in Virginia and in Tyrone and was later in the furniture business in Lewistown; the Floods came to Mount Union in the spring of 1906; J. D. Flood married Marie Boggs of Lewistown and was the father of two daughters.
That the firm arrived in 1906 is worth having beside a 1910 photograph: the wagon in the picture belonged to a business four years old, run by men who had grown up at the other end of this county and come back to it.
Not claimed: the date of the photograph itself. The page says 1910 and nothing opened here establishes that. Welch's account fixes when the firm arrived, not when the picture was taken.
Checked against the record. · Checked 2026-09-09
Source upgraded from Wikipedia to the FBI's own Wall of Honor record for Special Agent Terry R. Anderson, which is a federal account of this case and names the place, the abductor, the seven days and the outcome.
Confirmed there: 17 May 1966 as the date Anderson was killed; near Shade Gap, Pennsylvania; William Hollenbaugh as the abductor; a 17-year-old girl held seven days; the victim rescued and Hollenbaugh killed; and Anderson's own details, born Washington, Iowa in January 1924, Drake University and Notre Dame.
ADDED: Hollenbaugh's nickname, “Mountain Man”, which is the FBI's own word for him and is how the region remembered him.
CUT: “the largest FBI manhunt in U.S. history at that time.” Two faults in one phrase. It is a superlative, and no source opened here supports it. And it was not an FBI manhunt: the search drew in state police, local officers, the National Guard and civilian volunteers, so describing it as the Bureau's operation takes the work off a great many local people. The page now says a manhunt, and lets the federal agent's death carry the scale.
Still unsourced, and left out rather than asserted: the abduction date of 11 May, the rescue on 18 May, and State Trooper Grant H. Mixell as the man who fired the fatal shot. These are almost certainly right and they are in every retelling, but the FBI page does not carry them and this county's own newspapers are not digitised for 1966 in either archive available here. The Pennsylvania State Police or the county historical society would settle all three, and that is the route rather than another search.
A note on naming. The FBI does not name the victim, and that is deliberate practice. Her name is kept here because she has spoken publicly about it for decades and is inseparable from the story in this county; the page carries nothing about her beyond what she has made public herself.
Checked against the record. · Checked 2026-09-09
Source upgraded. This story previously rested on Wikipedia alone. It now rests on the National Historic Landmark nomination itself (NRHP ref. 93001614), plus I. C. White's 1885 report for the Second Geological Survey, which is in the project corpus.
Credit corrected. The original gave the discovery to John Frazer alone. The nomination gives it to three men: the rocks 'profoundly represent an understanding achieved by Henry D. Rogers, the director of the first survey and his two assistants, James C. Booth and John F. Frazer.' Frazer was an assistant, not the man on his own.
Number corrected. The original said 'several layers of sandstone.' The nomination is specific: the 1836 survey found 'there were in fact three ridge forming sandstones, not just one as previously thought.'
Measurement corrected. The original said the pillars sit 'about ten feet apart.' White's 1885 report, quoted in the nomination, says 'isolated piles of rock 50' high, with passage-ways from one to ten feet wide between them.' Ten feet is the top of the range, not the typical gap.
Sharpened, all from the nomination: the watercolour is by George Lehman, it was the frontispiece to Rogers's 1858 final report rather than merely 'in it', and Lehman is 'commonly believed' to be the first professional artist hired by a state or federal survey to sketch sites for government reports. That hedge is the nomination's own and is carried through rather than upgraded. Juniata College owns the MAJOR section, donated by the Africa family, with the remainder held by the John V. Eisenberg heirs and used by a local game and hunt club. NHL criteria 1 and 2, theme XIII Science, Earth Science, Geology.
Location tightened: 'between Huntingdon and Alexandria' becomes the nomination's own placement, on Warrior Ridge about a mile and a half northwest of Huntingdon along the old Huntingdon-Hollidaysburg Turnpike. White's 1885 description does place it 'near the road to Alexandria', so the original was not wrong, only vague.
Checked and held: the designation really is a National Historic Landmark, not a National Natural Landmark, which is what a geological feature designated in 1993 looks like it ought to be. The nomination is headed 'NATIONAL HISTORIC LANDMARK NOMINATION' and the designation date is 4 November 1993. The suspicion was checked and did not survive, which is the correct outcome to record.
Checked against the record. · Checked 2026-09-09
Sourced. The page had no source at all and described the counter without ever saying who kept it.
Confirmed from Welch, page 406, written while Peduzzi was in business: born in Lombardy in northern Italy on 16 March 1867; fourteen years in Italy; five years through the cantons of Switzerland, a stone mason by trade; France at nineteen for four years; to America in 1890, settling in the West in the region of Montana and Washington; travelled through thirty-one states.
A NAME NOTE. The chapter heading reads PEDUZZI and the first line of the text reads “Perduzzi”. The index, the tile in the doorway and every other mention say Peduzzi, so that is what the page uses; the variant is a slip in the book rather than a second spelling worth preserving.
Checked against the record. · Checked 2026-09-09
A generalisation the record contradicts for the named family. The page said “Italian families like the Peduzzis and Melonis came for the brick work”. For the Peduzzis that is not what happened. Welch has T. C. Peduzzi reaching America in 1890 and going to Montana and Washington. The W. H. Haws plant, the first brick works in Mount Union, was not built until 1899, nine years later, and he was a stone mason by trade rather than a brick worker.
The broader statement may well hold for other Italian families in this town, and this page is no longer the place it is asserted from: a claim about why a whole community came needs its own source, not one family's doorway.
The Meloni half is untouched here because nothing opened settles it either way. It is neither confirmed nor refuted, and it is no longer stated as though the Peduzzi record supports it.
Checked against the record. · Checked 2026-09-09
Source upgraded from Wikipedia to the Smithsonian's own published study of Saxton, by Arthur H. Frazier, 1975. The page was already careful and its central claim survived; three things changed.
Age corrected, or rather un-asserted. The page said he "went to work at twelve as an apprentice to a watchmaker." Frazier says: "When he was eighteen years old, Saxton was serving an apprenticeship under David Newingham, a local silversmith, watchmaker, and engraver, in his native village of Huntingdon, Pennsylvania." That places him in the apprenticeship at eighteen and gives no starting age at all, so twelve is not shown - and it is not refuted either, because an apprenticeship recorded at eighteen could have begun years earlier. The page now says what the source says and names the master, which it could not do before.
Date settled. The page hedged between "around September 25 or October 16". Frazier gives 16 October 1839 flatly. The hedge is no longer needed.
Small embellishment removed. The page had him at "the window of his office". Frazier says a second-storey window of the Mint, and says nothing about whose office it was.
Hedge made more precise, in both directions. Frazier's sentence is carefully built and the page now mirrors it: the plate is "perhaps the earliest daguerreotype ever taken in the United States, and certainly represents the oldest American daguerreotype extant." Perhaps for the taking, certainly for the surviving. The page had been weaker than the source allowed on one half and is now exact on both.
ADDED, and it is the reason to run this story at all: the cigar box and the ordinary reading glass, and the letter of recommendation from John McCahan of the Huntingdon Gazette. A county newspaperman vouching for a local boy is how the oldest surviving American photograph starts.
Checked against the record. · Checked 2026-09-09
Partly sourced, and the page says which parts.
Confirmed at NASA's own Apollo 11 mission page: the crew of Neil Armstrong, Michael Collins and Edwin “Buzz” Aldrin; launch on 16 July 1969; splashdown on 24 July 1969; the landing on 20 July 1969; and “Houston, Tranquility Base here, the Eagle has landed.” NASA's page also dates the mission to fifty-seven years ago, which is what the page says.
Not sourced here, and listed rather than quietly kept: that Armstrong stepped off about six hours after the landing; that Aldrin followed nineteen minutes later; that the pole would only go about two inches into the ground; the West Point salute; and the estimate of six hundred million watching, or one in five people alive. Two NASA pages were opened for these and neither carries them - the mission overview gives no timings and the anniversary feature is a photographic essay. They are almost certainly right and they are the sort of detail NASA's mission reports do carry; the report itself is the route, not another search of the same two pages.
A NOTE ON WHAT THIS PAGE IS. This is a national anniversary, not Huntingdon County history. That is a fine thing for a county page to post and the facts in it still have to be right, but it is worth being clear that nothing on it connects to this county, and the fact-checking effort this site can spend is better aimed at the county's own past.
Checked against the record. · Checked 2026-09-09
A number replaced, and the site's own pages are what caught it. The page said the town was finished in 1917 with ninety four dwellings, seventeen of them two family houses and seventy seven single family houses. That count could not be traced to any document. The survey gives the Sanborn Map Company's count of 1919: 111 residences, ten of them four family buildings, twenty one double houses and eighty single family houses.
The eighty is confirmed by arithmetic on this site's other Kistler pages, which were each taken from a separate survey report. Twenty seven Vermont Farm Houses and twenty three Norman Cottages are two of the three single family designs. The third, the Georgian Cottage, makes up the balance to eighty. The seventy seven could not be made to work with any of them.
Where the missing ten went. The four family buildings were never on Nolen's plan at all. They were put up at the southern end of the village, set apart from the rest, and torn down in the 1930s. Ranch and split level houses stand on Beaver Road where they were. A count taken after that demolition would not include them, which is the likeliest history of the lower figure.
What the page was leaving out, which is the substance of this check. The page praised the village and stopped there. Nolen's own chapter says what the good houses were for, and the survey quotes him: the population being largely foreign, there was a distinct necessity for a lead in the direction of Americanism, better given by good living conditions than by exhortation. A page about this town that prints the compliment and not the purpose is telling half of it. Both halves are now on the page, in his words rather than a paraphrase of them.
The dining room detail is the surveyors' own, not an inference: the scale of many of the houses was “sometimes so small” that the space designated as the dining room was not large enough for a dining table.
Spelling: the architects were Mann and MacNeille. The survey spells it that way on its title page and in the measured drawings. The page had MacNeill.
Checked against the record. · Checked 2026-09-09
A false claim removed. The original said this 1879 engraving showed the valley "before factories and railroads filled" it. By 1879 both railroads had long since arrived. Lytle's 1876 history records that the Pennsylvania Railroad was completed in 1850, twenty-nine years earlier, and that Mount Union was laid out that same year. Africa's 1883 history records the East Broad Top opened to Three Springs on 1 July 1874, five years earlier. The Pennsylvania main line runs through the Narrows itself, so an 1879 view of this gorge is a view of a railroad corridor, not of a valley before railroads.
A production note removed from the published text. The opening paragraph ended with a note naming an internal file and a date, which was never meant for a reader. Eleven other pages carried the same line and were fixed at the same time. Where the image comes from is now recorded in the source list instead, in plain words.
ADDED, and it is the better story: the name. Africa's history reports the murder of the trader John Armstrong with James Smith and Woodworth Arnold at the Narrows about the middle of February 1744, and says that event named the gorge. Lytle gives a whole chapter to the competing Captain Jack legend and titles it "A popular error," which is the hedge carried through here rather than an assertion that the legend is simply false.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
Source upgraded from Wikipedia to two federal pages and one nineteenth-century biographical record, and the page gained more than it lost.
Confirmed, from the Commemorative Biographical Encyclopedia of the Juniata Valley of 1897, which is in the project's own corpus: "David McMurtrie Gregg, born April 10, 1833, at Huntingdon; graduated at West Point, 1855 ... colonel Eighth Pennsylvania Cavalry in January, 1862; brigadier general of volunteers, November 29, 1862", and service through the Gettysburg campaign. The page had his birthplace and his rank right and can now date both.
Confirmed at the National Park Service: on the afternoon of 3 July 1863, "United States cavalry under General David Gregg stopped Jeb Stuart's Confederates from turning the Federal flank and threatening Meade's rear." That is the site's claim that Stuart never got around behind the Union line, in the Park Service's own words.
ADDED, all from the Park Service: Gregg was outnumbered; McIntosh's and Custer's 3,200 men attacked from three sides; the Union lost 254 men and the Confederates 450. And the Gregg Cavalry Shaft, erected in 1884, which the Park Service calls the only monument commemorating the services of both Union and Confederate cavalry at Gettysburg. A superlative, so it is carried as theirs and attributed rather than asserted here.
Two things dropped for want of a source. "About three miles east of town" - no distance was found on any page opened. And "one of the most important cavalry actions of the war", which is a ranking claim and needs its own support; the casualty figures and the Park Service's own account of what Stuart failed to do carry the weight better than an adjective would.
The hand-to-hand detail is left out for the same reason. It is very likely true and it is attributed in secondary accounts to a Confederate officer of Lee's Brigade, but that account was not opened here.
Checked against the record. · Checked 2026-09-09
Sourced, from a page image of the book rather than its OCR, which had the chapter only as an index entry.
Confirmed, Welch page 334: “The Mount Union Fire Department was organized July 28, 1903. At the present time there are about 50 members, who have enlisted their services voluntarily for the extinguishing of public fires.” The page had said fire protection “before 1903” with nothing behind it. The year was right and can now be given to the day.
ADDED: the first slate of officers, all named on that page — Samuel S. Mosser president, R. O. Johnson vice president, Stewart A. Kyper secretary, Arthur Johnson assistant secretary, Frank B. Davis financial secretary, G. B. M. Kepler treasurer, Charles Wenzel chief, Frank Moore assistant chief. And Welch's own complaint that the department badly needed equipment, which is a more honest picture than a salute on its own.
A cross-check that held. Charles Wenzel appears as fire chief here, and independently on this site's page about the Harbison-Walker works at full steam in 1909, which describes him as chief of the Mount Union fire department and the borough's street commissioner. Two pages built from different sources agree about the same man in the same year.
Checked against the record. · Checked 2026-09-09
Source upgraded from Wikipedia to the official recipient record, and everything the page already claimed survives.
Confirmed: born 15 April 1837 at Huntingdon, Huntingdon County, Pennsylvania; Medal of Honor for action at Chickamauga, Georgia on 20 September 1863; the action itself, rallying broken troops under fire. The record gives the citation verbatim, so the page now quotes it instead of paraphrasing, and the paraphrase was leaving out the point of the action: the ground was held long enough to get the wagon trains and the batteries away.
ADDED from the record. His rank at Chickamauga was captain, in the Ordnance Department; brevet brigadier general was his highest rank later, and the page had been carrying only the eventual rank. The medal was presented on 8 July 1902, thirty-nine years after the action. He died 29 May 1921 in New York and is buried at West Branch Cemetery, New Jersey. The medal itself is in Special Collections at the Georgetown University Libraries.
One nuance the record carries and the page does not need to hide: his service was credited to Harrisburg, Dauphin County. He was born in Huntingdon; the credit is an administrative fact about where he was accounted for, not about where he came from.
The ambassadorship to France, 1897 to 1905, was already checked in an earlier pass and is unchanged.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
This page carried a correction that was itself wrong, and it is now reversed. The honest version of what happened matters more than a tidy note, so here it is.
The page originally said Hackman “was one of two former Harbison-Walker managers who organized Mount Union Refractories in 1911.” An earlier check cut that sentence for want of a source, cut it from a second page as well, and recorded that nothing supported it. That was wrong. The Historic American Buildings Survey's written history of Mount Union states it flatly on page 4: “The third company, Mt. Union Refractories, was organized by two former managers for Harbison-Walker, R. P. M. Davis and Clinton V. Hackman.” The survey footnotes it to Trisha Colyer's article in The Valley Log of 9 May 1990, and adds a second, independent thread: R. P. M. Davis appears in the Harbison-Walker correspondence of the 1903 strike as superintendent at the Farrandsville plant.
Why the first check went wrong, because the reasoning is the part worth keeping. It found the Harrisburg Telegraph naming Davis as president and general manager and Hackman as superintendent, and treated that as evidence against two men having founded the company. It is not evidence against it. Two men who start a firm hold two different posts in it, which is exactly what the newspaper shows. Absence of the claim from the National Register nomination was then read as the claim being false, when a nomination is not the document that would carry it.
The lesson, and and this site now checks every page for it: a negative verdict needs a source, the same as a positive one. Cutting a true fact for want of a source feels like discipline while you are doing it and reads as a hole in the work afterwards.
Still good, and unchanged: Welch's 1909 portrait captions Hackman as superintendent of the Harbison-Walker brick plant, and the Harrisburg Telegraph of 21 November 1916 names him at Mount Union Refractories. The undated claim that he ran the Harbison-Walker plant “from 1902 until 1905” is still out; it had no source and it contradicts Welch, who was printing in 1909.
ADDED from the same survey page: eight kilns at the start, twenty one more during the First World War, the Clearfield County plant of 1922 and the change of name to U. S. Refractories, and the 1930 merger into North American Refractories.
Kistler is in Mifflin County, and the page says so.
Checked against the record. · Checked 2026-09-09
Source upgraded from Wikipedia to two books already in this project's corpus, which is where this should have been checked first.
Confirmed, Commemorative Biographical Encyclopedia of the Juniata Valley, 1897 - a contemporary account written while he was still a young man: "Martin Grove Brumbaugh, Huntingdon, Pa., was born in Penn township, Huntingdon county, April 14, 1862, son of George B. and Martha P. Brumbaugh, who now reside in Marklesburg." It also gives the whole early career: common schools and Juniata College, graduating 1881; country schools taught in 1877 and 1878; Millersville in 1882; teaching at Juniata until 1884; elected county superintendent of schools and re-elected in 1887, serving six years; Howard University as a post-graduate student from 1890, then the University of Pennsylvania.
Confirmed, Souvenir Historical Book of the county Sesqui-Centennial, 1937, which lists him among Huntingdon County men who reached prominence: "Martin G. Brumbaugh, Governor of Pennsylvania, 1915-1919."
Sharpened. The page said he was "born overlooking the Raystown branch" and "raised in Woodcock Valley". Both are defensible - Marklesburg sits in that country, and Africa's history places a frontier fort "in Woodcock Valley, near Marklesburg" - but the record says Penn township and names the parents at Marklesburg, so the page now says that. The Woodcock Valley tie is real and is given where it belongs: his great-grandfather Jacob bought land there.
CUT: "Pennsylvania's 26th governor". An ordinal is a checkable number and it could not be checked. The Historical and Museum Commission's own governors page answers with the agency's landing page instead - a redirect that returns a success code and none of the content, which is the failure that looks most like an answer. The years 1915 to 1919 are sourced and carry the point without it.
The line about him teaching at fifteen is arithmetic from two sourced facts in the same paragraph, born April 1862 and teaching country schools in 1877, not a claim from elsewhere.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Wikimedia Commons record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
Sourced from the Historic American Buildings Survey report on the Town of Mount Union, HABS No. PA-5974. The page previously had no source.
Confirmed and made exact: the page said “the hundred company houses on Ganister Hill”. The survey says Harbison-Walker owned 104 houses in 1911, all two-storey frame, twenty-seven of them duplexes, all but three grouped on the hillside south and west of the plant to form Ganister Hill. A round hundred was close; 104 is what the record says.
The nationalities corrected to the ones actually recorded. The page listed “Macedonia, Serbia, Croatia, and Russia”. The survey quotes the town's own newspaper of 28 January 1916, the Twice-A-Week Republican, listing the company's employees as “Slavish, Roumanians, Bulgarians, Italians, Greeks and Serbians” as well as “native Americans, and southern Negroes.” Serbians appear in both lists. Croatians and Russians do not appear in the source, and Bulgarians, Roumanians and Italians were missing from the page — which matters here, because this page is built around a Bulgarian-Macedonian newspaper. The period terms are kept as the paper printed them; paraphrasing them into modern words hides what was actually said about these people.
Not sourced, and removed: that families boarded “five to ten single countrymen at a time”. Boarding was common and the figure is plausible; nothing opened here gives it.
ADDED, because it is the honest half of an “uplift” story: the Red Cross ran an Americanization Program teaching sewing, fancy work and cooking, and the survey itself reads the whole effort as an attempt to make newcomers “less challenging and more controllable”. A page about new Americans that records only the welcome is not describing what happened.
Checked against the record. · Checked 2026-09-09
A correction that did not propagate, which is worse than not making it. An earlier pass refuted the claim that the 1936 flood was the most destructive ever recorded in the Juniata Valley, using the Geological Survey's gage records: on Aughwick Creek at Orbisonia the June 1889 discharge was about 18,400 second feet against 16,400 in 1936, and on Standing Stone Creek the highest stage in the whole record is 9.38 feet on 1 June 1889. That correction was applied to one page. This page still said the Juniata “broke all records”, so the site was contradicting itself again, and this time the contradiction was introduced by fixing only half of it.
All four flood pages on the site were checked together. Two state the 1889 finding correctly. One, the Corbin Bridge page, does it the right way and is left alone: it reports that a centennial souvenir book “called the flood of March 17, 1936, probably the most destructive the Juniata Valley had ever experienced,” attributed to the book, carrying the book's own “probably”. That is how a strong claim from a weak source should appear.
Sourced, the building itself: Welch records that Mount Union had two planing mills, one run by steam and the other owned and operated by A. S. Welch with electric power in connection with his lumber business. The page had never said which mill this was, and still does not (nothing opened here identifies it), but a reader now knows there were two and whose they were.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
Sourced at last. Every detail this page carried — James Nevell, November 1896, the Beers Hotel, the Broad Top House — came from Welch's history and none of it was cited. The page now names the book and the pages.
Corrected: the page called the Broad Top House an early hotel “in canal days”. It was not. Welch places it “along the railroad opposite the station” and dates its building to the late 1850s, by which time the Pennsylvania Railroad had been through for years. This is a railroad-era hotel, and the site has confused the canal era with the railroad era before.
A date the source will not settle. Welch gives two. On page 178, in Doyle's own biography, the Broad Top House is erected “in 1859”. On page 650 it “was built by James G. Doyle in 1858”. The same book, two different years. There is no way to choose between them from the book alone, so the page says the late 1850s and gives the one date the book is consistent about: Doyle died in the unfinished building on 28 April 1859.
ADDED, and it is the best thing on the page: the hotel was torn down and its material re-erected on the North Side, in use by Mr. Crum when Welch wrote. The building that replaced it on the corner carries his name, and so did the timber of the one it replaced.
Checked against the record. · Checked 2026-09-09
Corrected on arithmetic and on a missed fact that is better than anything in the original. Arithmetic: the original says '23 years straight' and '1937-1960'. May 11 1937 to January 7 1960 is 22 years and about 8 months. Rounding up to 23 is an overstatement of a checkable number, so the rewrite says twenty two. Name corrected: it was the National Republican Congressional Committee, which he chaired from 1953. Sharpened: 'a key player on the Ways and Means Committee' is vague. The source says that at his death he was its ranking member. Say the actual position. The missed fact: he died in office, on January 7 1960 at Bethesda Naval Medical Center, several weeks after brain surgery. That is why the service ends in 1960, and the original presented the end date as if it were a retirement. A man who served twenty two years and died in the job is the story; the original had the dates and missed what they meant.
Official record unreachable, recorded so nobody repeats the attempt. The Biographical Directory of the United States Congress is the right source for this page and all three routes to it none of the three sites that carry it could be opened from here. congress.gov answers 403, congress.gov answers 403, and history.house.gov answers HTTP 200 with a Cloudflare interstitial rather than the record - a wall that reports success. The arithmetic correction made in an earlier pass, from twenty-three years to twenty-two, stands on its own: May 11 1937 to January 7 1960 is twenty-two years and about eight months, and that is a calculation rather than a citation. The dates themselves still want the Directory behind them.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress item record, so a reader can open it.
The prose makes only general statements about company towns, which the federal survey documentation supports and other pages of this site cite in detail. It carries no number or date of its own to check.
Checked against the record. · Checked 2026-09-09
Photograph citation corrected from an internal credit line to the Library of Congress record, so a reader can open it. The text on this page is descriptive and still carries no source of its own; that is the next pass, not this one.
Checked against the record. · Checked 2026-09-09
The same number, corrected on the same day on every page that carries it. This page said ninety four company dwellings at Kistler, finished in 1917. So did the page about the design of the village. Neither figure could be traced to a document, and the survey gives the Sanborn Map Company's count of 1919: 111 residences, eighty single family, twenty one double, and ten four family buildings that were not on Nolen's plan and were pulled down in the 1930s.
A correction that reaches the page you happen to be reading is not a correction. This project has already had one figure fixed on one page and left standing on another, so both Kistler counts were changed together and every story on the site is now searched for the old number.
Unchanged, because it is documented: the General Refractories figures in the first paragraph, which come from the Mount Union survey and are given place by place rather than as a total.
Checked against the record. · Checked 2026-09-09
Sourced at the Library of Congress record for the map itself.
Confirmed: “Mount Union, Huntingdon Co., Pa. 1906”, published at Morrisville, Pennsylvania by Fowler & Kelly, c1906; drawn by T. M. Fowler, Thaddeus Mortimer Fowler, 1842-1922; a sheet 34 by 59 centimetres; number 811 in the Library's catalogue of panoramic maps of cities in the United States and Canada. The page had his initials right and can now give his name, his dates and his publisher.
Corrected, gently. The page said the view was “drawn from the air before airplanes”, which is a good line and slightly misleading: it was not drawn from the air at all. The Library's own note reads “Perspective map not drawn to scale.” Fowler worked from the ground and constructed the viewpoint. The page now says that, which is the more impressive fact.
The growth claim replaced with sourced figures. The page said the town “tripled in size” in the decade after the map. Welch, writing in 1909, gives 1,086 in 1900 and upward of 3,200 including suburbs by the time he wrote — which is close to a trebling but over the nine years before 1909, not the decade after 1906. The page now carries his figures and his dates instead of an unsourced span.
Checked against the record. · Checked 2026-09-08
Checked against the record. · Checked 2026-09-06
Checked against the record. · Checked 2026-09-04
Checked against the record. · Checked 2026-09-03
Checked against the record. · Checked 2026-09-02
Checked against the record. · Checked 2026-09-01
This post asked readers to settle a disagreement between the county's 1937 history (laid out 1840 by William Pollock for John Sharrer) and the federal survey (laid out and lots offered for sale in 1849). The contemporary newspaper answers most of it. The Huntingdon Journal for 4 and 11 September 1849 prints the lot sale advertisement over the names of George W. Speer and John Dougherty, dated 1 September 1849 - confirming the survey's 1849, and showing that Lytle's 1876 history named the right two men but dated it a year late. The same paper for 5 July 1843 shows John Sharer already keeping a warehouse at Mount Union, so the name is older than the lot sale and the 1937 account is not simply wrong. Added 1 September 2026.
Checked against the record. · Checked 2026-08-31
Checked against the record. · Checked 2026-08-30
Checked against the record. · Checked 2026-08-29
Checked against the record. · Checked 2026-08-29
Checked against the record. · Checked 2026-08-28
Checked against the record. · Checked 2026-08-27
Checked against the record. · Checked 2026-08-27
Checked against the record. · Checked 2026-08-26
Checked against the record. · Checked 2026-08-25
Checked against the record. · Checked 2026-08-25
Checked against the record. · Checked 2026-08-24
Checked against the record. · Checked 2026-08-23
Checked against the record. · Checked 2026-08-23
Checked against the record. · Checked 2026-08-22
Checked against the record. · Checked 2026-08-22
Checked against the record. · Checked 2026-08-21
Checked against the record. · Checked 2026-08-21
Checked against the record. · Checked 2026-08-20
Checked against the record. · Checked 2026-08-19
Checked against the record. · Checked 2026-08-16
Checked against the record. · Checked 2026-08-15
Checked against the record. · Checked 2026-08-14
Checked against the record. · Checked 2026-08-13
Checked against the record. · Checked 2026-08-10
Checked against the record. · Checked 2026-08-09
Checked against the record. · Checked 2026-08-07
Checked against the record. · Checked 2026-08-06
Checked against the record. · Checked 2026-08-05
Checked against the record. · Checked 2026-08-01
The original asserted a 140-foot tower, a Statue of Justice on top of it, and the French Renaissance style. Its last word on the subject is the brickwork reaching the second story. The 1909 Historic Huntingdon returned nothing on a new courthouse. SAH Archipedia could not be opened. So: Unsourced, not false. Also cut: "has dominated the town skyline for nearly 145 years." 1883 to 2026 is 143 years, and "nearly 145" is a strange way to say it. "Dominated the skyline" is also an unsourced scope claim. What replaced it is entirely from Africa 1883 and is better than what it replaced: a grand jury declaring the old courthouse inadequate, six bids spread over twenty two thousand dollars, the county taking the lowest one, the offices decamping to the Morrison House, and a very precise construction snapshot dated November 24 with the brick at the second story. All of it checkable, none of it invented. One claim kept from the original: the 1883 date, which is consistent with a building whose brickwork was at the second story in late November 1882.
Checked against the record. · Checked 2026-08-01
Four claims cut. This post took a source that explicitly says it does not know, and made it certain. 1. The hedge was upgraded. Africa 1883 says flatly "The origin of the name of the township is not clearly settled." That converts an explicit non-answer into a confident narrative. 2. "Iroquois" is not in the source. Africa says "Indian warriors" in Maguire's 1845 statement and nothing about which nation. 3. "the Great Indian Warpath" is not in the source either. What Africa actually says is that the maps show "the Indian path leading from Frankstown to the Bald Eagle's Nest." Different path, different name. Two separate inventions of specificity in one sentence. 4. "The village of Warriors Mark was founded in 1768" is unsourced. Africa gives the earliest land warrants in the township as July 28, 1766, to John Baynton and Samuel Wharton, nine surveys; further warrants in 1767 for Samuel Wallace and others; and farms improved in 1777 by Nathan and Thomas Rickets. No 1768, and no village founding date at all. Removed. 5. "today it is the northernmost township in the county" is an unsourced geographic superlative. Removed. Correct in the original: the township was erected in 1798. Africa nails it down further than the original did, to the January 1798 term of the Court of Quarter Sessions, and reproduces the petition and the court's order. What the original missed, and it is much better than what it invented: Krider's four forked oak trees standing at the corners of a square with stones grown into the forks. That is a physical thing a person showed the historian, in a specific place, and it is far more arresting than "ceremonially marked trees."
Checked against the record. · Checked 2026-08-01
Supersedes an earlier check made the same day, with a third source that turns a thin post into a good one. What could not be sourced earlier, and still cannot be: the 140-foot tower, the Statue of Justice, and "French Renaissance." All three remain cut. Four sources checked, none of them carries it. Still unsourced, still not false. What the new book did give: the 1937 Souvenir Historical Book, published for the county's sesqui-centennial, describes the 1842 courthouse that the present one replaced and openly mourns it. "It was a beautiful colonial structure, with tall white columns across the front and a beautiful white colonial belfry. Unfortunately that beautiful building, which should have been preserved was torn down and the present one erected in 1882." That is a named, dated, published local voice grieving a demolition fifty five years after it happened, and it is a far better post than a tower height would have been. The 1937 book says "erected in 1882." Africa 1883 shows the brickwork only reaching the second story on November 24, which means it was not finished in 1882. The sources do not agree on a single completion year, so the original asserts none. It says the 1842 building came down and the present one went up starting in 1882, which both sources support. Also still cut: "has dominated the town skyline for nearly 145 years" is both an unsourced scope claim and bad arithmetic, since 1882 to 2026 is 144 years and 1883 to 2026 is 143. Ordinal corrected: the present courthouse is the county's THIRD, not its second. Africa and the 1937 book agree on the sequence: the first stood in the middle of Third Street between Penn and Allegheny, the second went up in 1842 on the present site, and the third replaced it.
Checked against the record. · Checked 2026-08-01
Checked against the record. · Checked 2026-08-01
1. "Pennsylvania's largest lake" is wrong as written. The Army Corps of Engineers' own Raystown page says "the largest lake located entirely in Pennsylvania." Dropping "entirely" turns a true, careful claim into a false one, and it is the kind of thing a reader from Erie corrects in public. Same family as "only one left" versus "only one ever." The post now quotes the Corps' hedge instead of upgrading it, and makes the distinction the hook. 2. "since its creation in 1974" is corrected to 1973. The National Inventory of Dams, the Corps' own federal dam register, gives Raystown Dam a Year Completed of 1973. 3. "has spawned lake monster folklore since its creation in 1974" is CUT. That is a duration claim, and duration claims need their own source exactly like scope words do. No published source found for when the story began or where the name came from. Rather than hedge it or drop the legend entirely, the original says plainly that the check could not source it and asks the readers who would actually know. 4. "Folklore or fact? Nobody knows." is removed. "Nobody knows" is itself a claim about the state of knowledge. Sourcing note: the Corps' Raystown page could not be opened directly and was read through the Internet Archive snapshot of 2020-08-14. The National Inventory of Dams was read from the Corps' own public nationwide CSV export, which opened directly.
Checked against the record. · Checked 2026-08-01
The two copies differ only in their final clause. This one calls the relocation "one of the era's largest rail" projects; the other calls it "a significant railroad rerout." Both descriptions are attached to an event the check could not source at all, so the difference is moot and both are cut. The sourced facts stay identical across copies because they came from something real. The invented flourish drifts because it came from nowhere. All corrections identical to 8913b1cfd1de0fc4: the 1857 purchase date restored, the Huntingdon to Hollidaysburg section correctly shown as out of service by 1876 and therefore already dead before the 1889 flood, the unsourced 1892 main line relocation cut, the Johnstown conflation removed, and no canal end date stated because the sources disagree on it.
Checked against the record. · Checked 2026-08-01
Birthplace held up, but not from where the original got it. Neither Africa 1883 nor the 1897 Juniata Valley encyclopedia nor the 1909 Historic Huntingdon contains the name Purdue at all. Wikipedia gives Huntingdon County with NO citation on that sentence. The claim is carried by Purdue University's own Archives and Special Collections: "John Purdue was born in Germany Valley (near Shirleysburg), Pennsylvania on October 31, 1802." Africa 1883 independently places Germany Valley in the south part of Huntingdon County, east of the Aughwick, which is what lets Germany Valley and Huntingdon County sit in the same sentence. Two claims cut, both for source conflict. 1. "gave the money AND LAND." Three sources, no agreement on the land. Purdue's official history page says the 100 acres came "from local residents" and only the $150,000 came from Purdue. Purdue's Archives says he gave "one hundred acres of land that he had purchased on the west bank of the Wabash." Wikipedia says the donation was raised to $150,000 and 100 acres. All three agree on the $150,000 and only that. The land is dropped, not averaged, not picked between. 2. "on ONE condition: it had to carry his name." Every source that lists his stipulations lists more than one, and they do not agree on what the other one was. Purdue's Archives: "the college bear his name and that he be awarded a lifelong seat on the board of directors." Wikipedia, citing Kriebel's The Midas of the Wabash: "for the college to be located in Battle Ground, Indiana and for his surname to be associated with it." The name condition is in every account, so it stays. The count "one condition" is wrong and is replaced with "among his conditions." The second condition is dropped because the sources contradict each other on what it was. 3. "made his fortune in dry goods" is narrowed. The Archives biography has him brokering crops and hogs on commission first, then storekeeping, then banking, newspapers, railroads and an agricultural works. Dry goods is one chapter, not the fortune. What the original missed, and it is the whole reason to run this story: the extreme poverty, the iron foundry father, school at eight and out at twelve. Also the one graduate at the first commencement, with Purdue himself giving the address. Straight from the university's own archives.
Checked against the record. · Checked 2026-08-01
Corrected. Confirmed from Africa 1883: J. C. Blair began as a small retail stationery trade on Penn Street between Fourth and Fifth in 1869, moved into manufacturing what he sold, employed one hundred skilled workers by 1883, and in 1881 bought the Presbyterian Church building at Penn and Sixth and converted it into a factory he named the Keystone Building, fifty by seventy feet and four stories. Africa's telephone list separately records him as 'Blair, J. C., manufacturing stationer.' None of this specific, checkable, vivid material was in the original. A newspaper search for the firm's reach was attempted and did not produce usable evidence . The page does not need it, because the sourced detail is stronger. Dropped by implication, the legal pad: the original is titled the legal pad and opens 'That pad of paper on your desk has Huntingdon roots', which reads as a claim that the legal pad originated here. The rewrite claims tablets, which is sourced, and claims no invention. If the legal-pad origin is to be claimed later it needs its own source, and the title should change with it. 'By the trainload' is deliberate colour about volume, not a figure.
Checked against the record. · Checked 2026-08-01
Supersedes an earlier check of this story, with 8 quoted sources against 4. The earlier check was right about what it caught and missed something bigger. Six American Indian intellectuals founded it at a meeting McKenzie convened, and he deliberately kept himself out of the credit. He was forced out of Fisk. He closed the student newspaper, abolished the student council, banned most extracurricular activities, and students who broke those rules were expelled and labelled Bolshevists. W. E. B. Du Bois wrote in The Crisis, 'Men and women of Black America: Let no decent Negro send his child to Fisk until Fayette McKenzie goes.' Students protested in 1924 and 1925 and he resigned in 1926. Two independent articles agree, which matters because this is a serious claim about a real person. The Fisk University article, which is separate from the McKenzie article and notably more balanced, says he was 'eventually forced to resign when his strict policies on dress code, extracurricular activities, and other aspects of student life led to student protests in 1924 and 1925.' The same article's bibliography carries a Journal of Southern History paper devoted to the strike (Lamon, 49#2, pp. 225-44), so this is documented scholarship, not an internet grievance. And the other side is also sourced, so the original carries it: the Fisk article credits him with developing Fisk into the premier all-Black college in the United States, winning its academic recognition from the Carnegie Foundation, Columbia and Chicago, and raising a million dollar endowment. He was also genuinely the first American sociologist to specialize in Indian affairs and he taught at Wind River. He is a complicated man, not a villain and not a hero, and the rewrite says so in as many words. Nothing here is a false-statement finding. This is an omission finding, and for an elected official omission is the expensive kind: a reader who knows Fisk history closes that gap in the comments, and the correction is more visible than the original. The page already says 'helped found', 'helped bring together', 'run by Native people themselves', and even 'He was not from Huntingdon County. He came to it.' Its Juniata chronology is correct: he returned in 1927 and retired in 1941. Its only problem is the Fisk omission.
Checked against the record. · Checked 2026-08-01
Three of four numbers held. One was dropped, and it was dropped because the source fights with itself. The American Canal Society's Juniata Division index sheet says in its narrative: "The Juniata Division required 88 locks to overcome a change in elevation of 584 feet in the 127 miles between Duncan's Island and Hollidaysburg." That is where the original post's three numbers come from and 584 feet and 127 miles are uncontested anywhere in the document. BUT the same document's own lock by lock table carries this note: "This totals 86 locks, not the 88 in McCullough & Leuba." The compilers tabulated the locks themselves, got 86, and flagged that it disagrees with their own cited authority. That is a conflicting number inside a single source. The lock count is dropped. Not averaged, not footnoted, not shipped with an "about." Two superlatives cut, both unsourced. 1. "it carried something unprecedented." No source calls the Juniata Division unprecedented and it plainly was not, since it was one division of a system already under construction. 2. "the crown jewel of the Main Line of Public Works." What the original missed: the packet boat ran between Huntingdon and Hollidaysburg, so the inaugural run started here. That is the local hook and the original buried it. And the opening was not the finish. The same paragraph records that faulty workmanship was found afterward and locks and aqueducts had to be rebuilt before the canal could fully operate the following year. Seventy seven of its locks were wooden. That is a better and truer story than a crown jewel. Spelling note: the source document spells it "Huntington" throughout, an old variant. Corrected silently to Huntingdon in the original, flagged here.
Checked against the record. · Checked 2026-08-01
Corrected. Two of its supporting claims did not survive that scrutiny. Confirmed: Smith had the town laid out in 1767 and named it for Selina, Countess of Huntingdon, who had donated to the College of Philadelphia where Smith was provost. The substance of the correction is right. A specific attribution to a named relative is exactly the sort of detail a reader will ask about, and it is the riskiest kind of claim to carry unsourced in a post whose entire subject is getting the record right. Corrected, 'the county histories of 1876 and 1883 both agree': the check verified 1883 (Africa) and 1909 (Historic Huntingdon). The check did not open any 1876 history and cannot confirm one says this. The rewrite names the two books actually read. Hedge restored, and this matters: Africa 1883 does not state the naming as fact. It says the name 'is believed to have been given' by Dr. Smith in her honor. The rewrite shows both, the 1883 hedge and the 1909 flat statement, which is both more honest and more persuasive. Quote the source's own hedge.
Checked against the record. · Checked 2026-08-01
Five numbers and one role claim removed. This post was almost entirely numbers, and almost none of them held up. 1. "co-founded" is contested and is removed. Wikipedia says "co-founded with Chip Davis in 1974," citing an Omaha World-Herald piece the check could not open. Berkey's OWN official site does not claim it. It calls him "featured keyboardist with Mannheim Steamroller." When a man's own site gives him a smaller title than a third party does, the smaller title is the one to print. Used his wording. 2. "Over 50 years with Mannheim Steamroller" is removed. His own bio page says 38 years and the page is undated, so it cannot be turned into a current figure. Two sources, no agreement, no date. 3. "10 original Christmas albums" versus "11 original Christmas albums." That is the cleanest possible case for the drop rule. Neither number ships. 4. "more than 30 albums" is removed. No source found. 5. "over 300 classical pieces" is removed and replaced with his own site's figure, over 400 published titles, in his own site's words. His number, not ours, and it is bigger anyway. 6. "multiplatinum-selling" is removed. 7. "across every major idiom" is removed. That is a scope word and it needs its own source. What survived: birth date and birthplace (May 24, 1942, Huntingdon, Pennsylvania), Juilliard, and the Huntingdon connection, which is the only part of this story the page actually cares about. Sourcing note: the birth date and birthplace come from Wikipedia citing Pitcher, John, "New CD shows Berkey's range," Omaha World-Herald, 20 December 2009. Everything else traces to berkey.com, his own site.
Checked against the record. · Checked 2026-08-01
Confirmed: the date (May 1, 1947, before dawn); two 16-foot sheets of steel lying lengthwise on an open gondola of a westbound freight, which shifted so their ends stuck out over the tracks on both sides; the train struck was the Pennsylvania Railroad's New York to St. Louis sleeper-coach flyer, running westbound on the parallel track; the steel ripped open its side; and the third-train claim, which is real. An eastbound freight hit the steel protruding on the other side moments later, its engine and tender were thrown from the rails, and all four main tracks of the PRR were tied up. Corrected, train name: the original calls it the 'American Limited'. Both papers call it 'The American'. There is no 'Limited' in the name. Corrected, death toll: the original says four dead. Four was the first-day wire figure (Nome Nugget). By the next day the AP reported FIVE dead and named all five: Joseph Selnick jr., 18, soldier, Pleasant Gap, Pa.; Pvt. Robert E. Deavers, 37, Zanesville, Ohio; Pfc. Herbert B. Huetsch, 21, Columbia, Ill.; Michael Burdock jr., Monessen, Pa.; and Mrs. Jean D. Sims, 26, Clifton, Va. Wikipedia's citation to a Pittsburgh Press headline of May 2 1947 ('Fifth Victim Dies After PRR Crash') corroborates the fifth death, but the check did not open that paper. The rewrite says 'five people were dead by the next day', which is exactly what the opened source supports and also explains why other accounts say four. Dropped, injured count: the numbers conflict badly. 40 (first-day AP), 46 (Pittsburgh Press headline per Wikipedia, not opened), 51 in the Evening Star headline against 45 injured plus 6 critical in its own body. No injury figure appears in the rewrite. Dropped, distance: the first-day AP says five miles west of Huntingdon, the May 2 AP follow-up says 4 miles west. Two AP dispatches disagree, so the distance is dropped and the rewrite says only 'west of Huntingdon'. Dropped, 'the fourth passenger car': no source supports it. The only traceable statement, in the Wikipedia entry citing an AP story in The State (Columbia S.C.), says the FIRST coach. The check did not open The State, so first coach is not asserted either. The rewrite says the steel ripped open the side of the train and names no car. Dropped, '1:47 a.m.': traces only to the same Wikipedia entry and its unopened AP citation. The two papers opened say 'pre-dawn hours' and 'early yesterday'. The rewrite says 'just before dawn'. Name conflict flagged: the Evening Star prints the mother's surname 'Simms' in the AP Wirephoto caption and 'Sims' twice in the article body, in the same issue. The rewrite uses Sims, the majority spelling in that issue. The story the original missed: the two-week-old baby who came out of the wreckage wrapped in a towel with a broken leg and lived, while her mother did not, and the honeymoon couple in the same hospital. Both are in the rewrite. Tone: the original ended 'Tag somebody who grew up along these tracks, bet they've never heard this one' on a wreck that killed five people including a young mother. Replaced with a genuine question about J. C. Blair Memorial Hospital.
Checked against the record. · Checked 2026-08-01
The post says "built in 1805." Africa 1883 says "About the year 1796 a company was formed ... to erect a furnace called 'Huntingdon,' on the Warrior's Mark Run, in Franklin township." Nine years off, and note the source's own hedge, "about the year," which the rewrite keeps. The post says it was "built in 1805 by Alsatian immigrant George Anshutz." It was built by a company of three, and Africa is unusually specific about who did what: Mordecai Massey owned the land, Judge John Gloninger of Lebanon contributed the money, and Anshutz contributed the skill. Giving one man all of it is a factual claim and it is wrong here. This is the same defect corrected in this site's page on General Gregg. Cut, no source: "Alsatian." Africa says only that Anshutz's skill was "acquired in the manufacture of iron in Europe." Cut, no source: the list of "seven original structures" (stone blast furnace stack, brick company store, ironmaster's mansion, gristmill with equipment, mill house, worker's house, office building). The check did not open a survey that enumerates them. Cut, no source: "By 1810, Huntingdon County's iron industry roared with eleven furnaces, eight forges, three rolling mills, and twelve naileries." Four precise counts in one sentence. The word "nailery" does not appear anywhere in the county histories held here. Cut entirely rather than partially. Complication the original glosses: the original calls it "a complete iron plantation ... with seven original structures," implying continuity from 1805. Africa says the FIRST location "was not favorable, and after some time was abandoned, and another furnace was built about a mile farther down the run." So whatever stands today is the second works, not the original seat. Any post about surviving structures has to say which furnace it means. What the original missed: the three-way division of land, money and skill; the one horse and pair of oxen, carried with the source's own hedge "it has been said"; and the charcoal run lasting until 1870.
Checked against the record. · Checked 2026-08-01
The original's central premise is wrong. It calls Jacob Higgins "a Huntingdon County officer." Africa 1883 says, twice and unambiguously, that he was from Blair County: "authority from Governor Curtin to Lieut.-Col. Jacob Higgins, of Blair County, to recruit a nine months' regiment." This matters more than a normal date error because the entire reason the original exists is to claim him as ours. He is not ours. What IS ours, and what the original never mentions, is that four of the regiment's ten companies were raised in Huntingdon County, with named captains. The 125th mustered out May 18, 1863. The Twenty second Cavalry organized in February 1864, with Higgins as colonel. That is roughly nine months later, not immediate, so the original's claim of an immediate move is removed. The rewrite says only that this regiment's men went home when their nine months were up, which is what the muster-out record shows. Africa follows the Twenty second through 1864 engagements but the check did not read a statement that Higgins personally served to the end of the war, so the original stops where the source stops. Africa gives Higgins no middle initial anywhere the check found. The initial may well be right in other records, but since the check did not open one, the rewrite uses the name the source uses. Correct in the original: the nine month enlistment, the May 1863 end date (May 18, 1863 exactly), and the move to the 22nd Pennsylvania Cavalry. Gregg of Company H here is not the same man as the Gregg in the Huntingdon Hero page corrected earlier, who was the cavalry officer at East Cavalry Field. Do not merge them.
Checked against the record. · Checked 2026-08-01
Corrected. The post was largely right and one detail was wrong in a way a visitor would notice from the parking lot. Confirmed from PA DCNR: iron production topped 3,000 tons annually by the early 1880s, and DCNR adds the detail the original omitted, that this made it one of the largest charcoal furnace operations in the state. Freedom Iron began producing locomotive tires, railroad car wheels and axles in 1856 'utilizing iron produced at the Greenwood Works', which does establish the link the original asserted. John Armstrong Wright 'helped found the Pennsylvania Railroad and the city of Altoona'. Added, and it is the best line available: 'The iron was hauled by wagons over Stone Mountain to Freedom Iron Works to be turned into wrought iron.' Concrete, physical, sourced, and it was not in the original at all. Corrected, 'The furnace stacks still stand at Greenwood Furnace State Park today': not supported. What DCNR says is that Furnace Stack #2 was restored as a monument in 1936, and that six original buildings and the cemetery remain. Plural stacks 'still standing' implies surviving original structures and overstates what is there. The rewrite says exactly what DCNR says. Corrected, 'John Armstrong Wright led the company': DCNR does not give him a title at Greenwood or at Freedom Iron. What it says is that John A. Wright and Company purchased the works in 1847. The rewrite says that instead. Timeline untangled: the original ran the early 1880s tonnage straight into the Freedom Iron products, which began in 1856. Both are true and they are decades apart. The rewrite keeps them in order rather than implying one figure describes the other date.
Checked against the record. · Checked 2026-08-01
1. "sweeping away bridges and homes around Huntingdon". No source. Removed. 2. "Entire families evacuated from Portstown as floodwaters crept across the countryside." No source. Portstown is a real Huntingdon neighborhood and the claim is entirely plausible, but the check could not source it, so it is out. Worth chasing in a local paper later; it is exactly the kind of detail a reader will confirm in a comment. Source found: USGS Water-Supply Paper 799, The Floods of March 1936, Part 2, Hudson River to Susquehanna River Region, published 1937. A primary federal measurement document with gage-by-gage records for four Huntingdon County streams. Downloaded and read directly from pubs.usgs.gov, 406 pages. The finding, and it reverses the emotional shape of the original: the original said 1936 "remains one of our county's most significant natural disasters," which is hedged enough to survive, but the Survey's own tables show that on TWO of the four Huntingdon County gages the June 1889 flood was higher than 1936. Aughwick Creek at Orbisonia: 18,400 second-feet in June 1889 against 16,400 in 1936. Standing Stone Creek near Huntingdon: maximum known stage 9.38 feet on June 1, 1889. Said precisely, because this is exactly where a post like this goes wrong: 1936 did NOT under-perform everywhere. On Great Trough Creek it was more than four times the previous recorded peak, and on the Raystown Branch at Saxton it was 80,500 second-feet against a previous record of 29,000. But those two gages only have records reaching back to 1930 and 1911. The two gages whose record reaches back to 1889 both put 1889 on top. The post says only what the tables say, gage by gage, and does not generalize to "the county." Also handled: "torrential rains combined with snowmelt" is not asserted in the rewrite. The Survey measured both, publishing daily March precipitation tables and separate tables of snow depth still on the ground, but the check did not open a narrative sentence stating the cause, so the original does not state one. Not available, and worth knowing for next time: there was no USGS gage on the Juniata River main stem at Huntingdon in 1936. The nearest main-stem gages were Frankstown Branch at Williamsburg and the Juniata at Newport. No post should ever quote a 1936 river level "at Huntingdon."
Checked against the record. · Checked 2026-08-01
Corrected. The core claim holds and traces to an official state source, but the original overstated it in three separate ways and one of them is a superlative in capital letters. Confirmed: PA DCNR's own Trough Creek park history says, without hedging, 'During 1835, for the first time in the United States, good iron was produced using coke as a fuel at Mary Ann Furnace.' DCNR also confirms Savage leased the furnace in 1832, renamed it, and built Savage Forge. Africa 1883 independently records Mary Ann Furnace and Forge on Big Trough Creek, owned by John Savage, conducted by George Thompson. The claim is contested and the original should show it. Wikipedia's coke article says the first use of coke in an American iron furnace was around 1817 at Isaac Meason's Plumsock works in Fayette County. But a flat 'FIRST in the entire U.S.' in capitals invites a reader with a search engine to produce Meason in ten seconds. The rewrite attributes the claim to the state park's own history and keeps DCNR's qualifier 'good iron', which is the word that makes the claim defensible. Dropped, 'John Savage's innovation': DCNR says Savage leased and renamed the furnace. No source opened credits him with the technique. Sole credit for an innovation is a claim of its own and this is the same trap that hit this site's page on General Gregg. Dropped, 'changed American metallurgy forever' and 'We were decades ahead of national iron technology': unsourced significance and unsourced scope. Dropped, 'birthplace of American coke-iron smelting': stronger than what DCNR actually says, and directly exposed to the Meason objection.
Checked against the record. · Checked 2026-08-01
The sourced part is almost word for word right, which is why the one unsourced sentence stands out so badly. Confirmed: Africa 1883 says Cromwell "was erected from Shirley and Springfield townships in January, 1836, and named 'in honour of Col. Thomas Cromwell, deceased, who was an early settler and a distinguished and hospitable citizen.'" January 1836, the two parent townships, and the honorific language all check out. One claim cut: "a practice rarely documented in Pennsylvania's township records." No source. It is a scope claim about statewide practice, resting on nothing, and it is doing all the work in the original. Removed. One claim tightened: "an early settler who arrived around 1785." Africa does not say Cromwell arrived in 1785. It says that in the fall of 1785 he and George Ashman "began to take warrants for vacant lands in the valleys in the vicinity of Orbisonia." Taking up land warrants is not the same as arriving. The post now says what the source says. What the original missed entirely, and it is the whole reason the man got a township: Bedford Furnace. Africa calls it "the first iron-smelting establishment west of the Susquehanna" and the 1897 encyclopedia independently calls it "the first furnace west of the Susquehanna river." Two independent published sources for the superlative, which is the bar. The September 10, 1798 Pittsburgh Gazette advertisement is Africa's own citation of a contemporary newspaper. Not used: Africa also records a claim that "the first bar-iron made in this country that found its way to the market of that city was produced at this forge," but he writes it as "it has been claimed that," and a hedge that heavy on a superlative that large does not belong in a post. Left out rather than repeated with the hedge. Not used: no death date for Thomas Cromwell was found in any corpus book, so the original says only that he was already gone, which is what the 1836 naming language establishes.
Checked against the record. · Checked 2026-08-01
1. "William Lovell ran the only building, a public house." Wrong twice over. Africa 1883: "When the village was laid out three buildings stood on its site." And Lovell's tavern was "the first in the village, previous to 1815," which is eighteen or more years AFTER the village was laid out. He did not run the only building and he was not there at the founding. 2. "By 1849 postal confusion forced the town to pick ONE name." No source. Africa says only "Measures for the incorporation of the borough were inaugurated in 1849." Not one word about the post office or name confusion. Invented causation, removed. 3. "The citizens voted for Cassville." Wrong, and it is the opposite of what happened. Africa: the name "was selected by a committee consisting of Robert Speer, D. Stever, and D. Clarkson." A three man committee, not a town vote. The original post even asks readers "What name would you have voted for?" off a vote that never happened. Correct in the original: the two former names, and the 1853 incorporation (March 3, 1853, by act of the Legislature). Spelling: Africa's borough section spells it Chilcotestown and the family Chilcote. Used the town spelling the source uses. Not asserted: why Cassville. Also fixed: the original ended "Tag someone from Cassville's Name," which is a broken template string, not a call to action.
Checked against the record. · Checked 2026-08-01
The open question was the 1959 birth year, which Wikipedia does not carry. The library of congress name authority file does, and it is a better source than anything else in this story's chain: record n90644672 gives 'Raschka, Christopher, $d1959-' and a birthDate of 1959-03-06, with LC's own note reading 'Chris Raschka (born March 6, 1959 in Huntingdon, Pennsylvania)'. The post's year was right and the rewrite can now give the exact day. Confirmed: two Caldecott Medals, The Hello, Goodbye Window in 2006 and A Ball for Daisy in 2012. Added: Yo! Yes? was a Caldecott Honor Book in 1994, so there are three Caldecott recognitions, not two. The post undersold him. Corrected, the from-here trap again: 'From a small rural Pennsylvania county to the top of American children's illustration' implies he was raised here. He was not. He was born here and grew up in suburban Chicago, with part of his childhood in Austria, his mother's homeland. The rewrite claims the birth, which is genuinely ours and genuinely documented, and says plainly that the raising was elsewhere. Claiming only what is true is also what makes it safe to say loudly.
Checked against the record. · Checked 2026-08-01
Confirmed: about 3,000 bituminous coal miners in the Broad Top field of District No. 2 idle on a strike order issued by the district officers and approved by the international president of the United Mine Workers. Corrected, 'more than 3,000': every paper says 'About 3,000' or 'About three thousand'. The post upgrades the source's own hedge into a floor. The rewrite says 'about three thousand'. Corrected, the July 10 date: no source supports it. The AP dispatch is datelined Johnstown, Pa., July 11, 1920 and reports the men 'are idle', so the walkout began on or before July 11 but the day itself is not established. This is the publication-date-versus-event-date trap in a new form. The rewrite says 'By July 11, 1920'. The Nogales reprint redates the same wire copy 'July 14', which is a reprint artifact and not a second event. Dropped, 'near Dudley': no source opened mentions Dudley, or any specific mine or town. The dispatch describes the Broad Top field as a whole, filed out of Johnstown. Dropped, 'Most locals today don't realize our backyard was a flashpoint in America's coal labor wars': two invented scope claims in one sentence, neither with a source. Corrected, 'for higher wages': this is the real find. It was not a demand for a raise. The miners alleged the operators refused to pay the full mining rate already due under the government agreement that ended the previous winter's general strike, plus the district wage scale signed after it. The operators held that the new scale wiped out a Broad Top sub-district rate of ten cents a ton above the district base rate that had been in effect since early in 1917. Men walking out over ten cents a ton they said was already theirs is a sharper and truer story than 'higher wages'. Ocr note: the Ward County Independent prints 'District No. 3'. Three other papers print District No. 2, and Johnstown is UMW District 2, so No. 2 is correct and the Minot copy is a typo. Name not asserted: all four papers say only 'International President Lewis of the United Mine Workers'. None gives a first name, so the rewrite says 'the international president of the United Mine Workers' and names no one. Almost certainly John L. Lewis, but that needs its own source before it goes in print. The OCR is too garbled to quote: 'Broad Ton field hnn thm hnrt..t haul to Philadelphia markets from bituminous mines and its outnut I ilmii.u- Important now because of a coal famine'. Worth confirming from a cleaner copy.
Checked against the record. · Checked 2026-08-01
Corrected, and this one carried the single riskiest thing a post can carry: a direct quote in quotation marks that does not match its own cited source. The quote is wrong as cited. The post has Braddock replying "the English Shoud Inhabit & Inherit the Land." The post's own claimed source, the National Park Service page on Shingas, gives the reply as "No Savage Should Inherit the Land." Both phrases circulate in accounts of this exchange, which descends from Charles Stuart's captivity narrative, but the original is quoting one wording and citing a page that prints another. The rewrite uses the NPS wording and says so in the text, so a reader can check it in one click. Causation invented, and this is the bigger fault. The post says "Enraged, Shingas switched sides to the French." No source says that. What the sources say is that the Lenape stayed neutral in the early stages, declined to assist Washington at Fort Necessity in 1754 and Braddock in 1755, had no desire to be French subjects, and aligned with France only reluctantly after Braddock's defeat left France dominant in the region. Turning a considered and reluctant political decision by a Native leader into an emotional outburst is a real distortion, and it is the kind that reads badly. Dropped, 'killing and capturing hundreds': no source opened supports a number. The sources say numerous raids, and that both Pennsylvania and Virginia put a bounty on him. Confirmed and kept: 'Shingas the Terrible' is genuine. Added, because leaving it out makes a one-sided portrait of a real person: 'Although he was an implacable foe in battle, he was never known to treat his prisoners of war with cruelty.' This is the same defect class as this site's page on Fayette McKenzie, a real human being flattened into one trait. Softened: 'sparked a frontier war that changed Pennsylvania forever' overstates single-cause history. The rewrite says the answer helped set the frontier on fire, which the sequence supports.
Checked against the record. · Checked 2026-08-01
It was still the wrong outcome, exactly the failure a cautious check warns about: cutting a true fact for lack of a source reads as discipline in the moment and as a hole in the work later. It says: "When the Pennsylvania Railroad was being projected through the county more than a century ago the company established a station here and named it Mapleton because of the many huge maple trees standing in the area." The centennial says the railroad company named the station, for the trees. Naming by a corporation is a sharper fact than naming by drift. The centennial says "A town was laid out and building encouraged, but by 1858 only three houses stood here." So a town WAS laid out, it simply did not take. The post now says both: it was laid out, and it did not take. Both sources agree 1858 is the year the place was still almost nothing, three houses, not the year it was laid out. Independent corroboration gained: the 1966 centennial separately confirms Africa 1883 on the 1851 owners (M. F. Campbell and John Donaldson) and on only three houses standing by 1858. Two sources eighty three years apart agreeing on both details is much firmer ground than Africa alone. New material added: Campbell gave the borough land for all of its churches and the cemetery, and the Presbyterian church still in use was built by Campbell, Clayton and Mrs. John Donaldson from brick made in Campbell's own factory. Ocr note: the centennial's scan renders the capital C in Campbell, Clayton and Cemetery as G throughout ("Gampbell", "Gayton", "Gemetery").
Checked against the record. · Checked 2026-08-01
Confirmed in a primary source, same day, page one, and then made much stronger. The Harrisburg Telegraph of September 20, 1916, page 1, city edition, carries the story datelined "Mount Union, Pa., Sept. 20" and reports the explosion as having happened "at 6 o'clock this morning." That kills the newspaper-date-versus-event-date trap outright: the paper and the event are the same day, and the paper says so. Every core claim in the original holds: September 20, 1916, six men killed, cotton dry house, completely destroyed, Aetna Explosives Company, Mount Union. Even the word "terrific" in the original is the newspaper's own word. Four claims were cut on 1 August 2026, and the first cut was wrong. 1. "over 1,000 workers at peak production." No source had been found for it then. The Historic American Buildings Survey report on the town (HABS No. PA-5974) says the powder plant "employed at least 1,000 men and women until closing at the end of 1918," so the figure is back in the story, in the survey's words: at least 1,000, not over 1,000 at peak production. 2. "The plant churned out smokeless powder for World War I." Cut as written. The papers call it a powder plant and an explosives plant, and the Evening Public Ledger of October 25, 1916 has Aetna and du Pont planning to convert auxiliary factories to dyestuffs once "the demand for explosives and munitions becomes normal," which confirms munitions but not smokeless powder specifically. Worth noting the United States was not in the war in September 1916, so "for World War I" needs care it was not given. 3. "killing six workers in an instant" loses "in an instant." Embellishment on top of a death toll. 4. "the cost of that industrial boom was measured in lives lost to catastrophic accidents" was a plural claim with a single incident under it. The Harrisburg Telegraph's weekly Mount Union column reported fires in that exact building on August 5, August 19 and September 2, 1916, in the seven weeks before it blew up. Three separate notices, each treated as ordinary town news. The post no longer has to assert a pattern. It can just print the paper. Not used: the Evening Public Ledger of January 8, 1916 reports an earlier Aetna fire "supposed to have been caused by spontaneous combustion due to overheating the building" with a loss figure the OCR renders as "J75.C0O." The figure is almost certainly $75,000 but the OCR is not clean enough to print a dollar amount from, so it is left out. The three krishocker.com and spicerweb.org URLs originally attached to this story are now unnecessary. The post stands on the newspaper.
Checked against the record. · Checked 2026-08-01
One claim cut, and it was the whole hinge of the original post: "Then the Pennsylvania Railroad built its great hubs at Tyrone and Altoona instead." Africa 1883 does not say that. What it says is general and hedged: "the proximity to other points more advantageously located has since the completion of the railway built them up while Birmingham has steadily declined." It names no hub, no Altoona, and it puts the railway decline LATE, after the village's best years. Naming Tyrone and Altoona as the cause is invented specificity, a known trap. Cut rather than hedged. The original also had the sequence backwards. Africa is explicit that Birmingham failed as a speculation FIRST, in the 1790s, for reasons that had nothing to do with railroads, and then had its most prosperous stretch from about 1836 until the railroad came. The real reason it failed is far better than the invented one and it is documented at length: Cadwallader sold the town by lottery to absentee buyers, and the ground did not match the paper. Africa reproduces the entire deed of dedication, which is where the ten dollar ticket, the five dollar deposit, the ballot and the five sworn freeholders all come from. Date handled carefully: the deed was subscribed by John Cadwallader and his wife Catherine on Dec. 13, 1797, and the plan was not recorded until Feb. 26, 1799. The post uses the 1797 laying-out date, which is what Africa's own summary sentence uses.
Checked against the record. · Checked 2026-08-01
Confirmed by the college itself, which is the institution that holds the records: Nancy (Slaughter) Lee was the first African American student to graduate from Juniata College, in 1932. She was a Huntingdon native and a graduate of Huntingdon High School, the daughter of Levi and Anna Bertha Slaughter, enrolled after finishing high school in 1928, and was part of the Day Students' Club all four years. Her senior portrait in the 1932 Alfarata is the first image of a Black student to appear in any Juniata yearbook. Dropped, 'a time when many American colleges still turned Black students away': true in substance and unsourced as written. 'Many' is a scope word and needs a source of its own, and the original does not need it. The specific facts carry more weight than the generalisation. Dropped, 'walked across the stage': nobody has sourced a description of the ceremony. It is a small invented image and this is not a post to invent images in. Dropped, 'made history quietly, one commencement at a time': she graduated once. The phrase means nothing. The yearbook detail is the strongest thing available and it was not in the original at all. Two leads came out of the check: (a) 'Her family traces back to Charles Slaughter, who escaped slavery in Virginia and settled in Huntingdon County in 1845.' If that holds it is one of the best local stories on this site and deserves its own post, not a clause in someone else's. (b) 'After graduating from Juniata, Nancy (Slaughter) Lee '32 later moved to Philadelphia, where she passed away in 1957.' Hedge worth keeping in mind: the television coverage of her story says 'little is known about her time at the college.' Both are leads still being followed.
Checked against the record. · Checked 2026-08-01
Error: "The road was completed by 1819." It was not. The 1991 county survey of historic engineering and industrial sites says "By 1821 the road was complete from Huntingdon to Blairsville, a distance of seventy-seven miles," and gives the construction date range as 1815 to 1821. Two years early. Correct in the original: the 1810 authorization (the company was incorporated March 20, 1810), the seventy seven miles, and Huntingdon to Blairsville. Those all check out exactly. Cut, no source: "one of over 200 turnpike companies building more than 3,000 miles of toll roads across Pennsylvania." It is also doing nothing for a county page. Gone. Cut, no source: "replacing colonial Indian trails and dirt roads with the commercial superhighways of the early republic." Rhetoric asserting a causal replacement nobody sourced. What the original missed, and it is far better than what it had: the road is still there and still unpaved. The survey describes a three and a half mile stretch over Warrior Ridge between Huntingdon and Alexandria as extant and relatively unaltered, and lands the best line in the whole entry, that it is still marked by the same potholes, ruts, washouts and mud that travelers complained about at the time. A history post that ends at a place a reader can drive to today is worth five that end in 1821. Also recovered: the project stalled for four years after incorporation and was only rescued by a citizens' meeting in the county courthouse in January 1814, and the company ran out of money and paid contractors in scrip. Africa 1883 separately indexes a "Huntingdon, Cambria and Indiana turnpike shinplaster," which is that same scrip. Not used: Africa's line that "a considerable length of [it] passed through a rugged and sparsely settled region, unable to contribute any substantial aid" is quoted inside the 1991 survey rather than read in Africa directly at that page, so it is left out of the original.
Checked against the record. · Checked 2026-07-31
Checked against the record. · Checked 2026-07-30
Checked against the record. · Checked 2026-07-29
The two standard county histories confirm nearly all of it. (1) 'Middle of Third Street' is RIGHT: Africa 1883 says 'Third, then called Smith Street,' which reconciles the 1794 document's 'middle of Smith St.' (2) 1797 is RIGHT: the 1909 history says 'the probability, however, is that the building was first used in 1797.' (3) The William Smith bell donation is RIGHT and documented by a surviving 1798 receipt. (4) The 1842 second courthouse is RIGHT: Africa 1883 says it was 'completed and occupied in August, 1842.' Only two problems remain. First, the weight: the 1798 receipt says 350 pounds, but the 1909 history says the bell itself 'weighed 254 pounds.' The rewrite also surfaces the genuinely moving fact the original missed: the bell outlived the courthouse and went on to ring for schoolchildren until it cracked.
Checked against the record. · Checked 2026-07-29
Partly confirmed, one clear error, several unsourced. Confirmed by the 1897 Juniata Valley encyclopedia: Thomas White was an immigrant from Ireland, is remembered as having taken part in the Boston tea party, was a soldier in the Continental army, born May 1739, died September 13, 1820. Important: the source hedges ('is remembered as having taken part'), and the original post correctly preserved that hedge with 'county histories remember him among the men'. Good instinct, kept in the rewrite. The error: the original calls him 'a young Irish tailor'. The source says 'farmer, nail-maker, and soldier'. There is no tailor anywhere in it. And born in 1739 he would have been about 34 at the 1773 tea party, so 'young' is a stretch too. Still unsourced and therefore removed: raising 21 children, membership in the American Philosophical Society, three sons serving in 1812, and son Ezekiel being captured at Lundy's Lane in 1814 and dying in a British prison camp. That last one is a vivid, checkable claim and it needs its own research before it is ever stated here.
Checked against the record. · Checked 2026-07-29
Africa 1883 gives the whole arc: 'In 1872, W. A. Orbison and C. W. Welch, under the firm-name of Orbison & Welch, established... the Huntingdon Car-Works... at the end of which time operations were suspended.' Then: 'In July of 1880, J. G. and M. C. Blain, under the firm-name of Blain Brothers, purchased the establishment, which they enlarged to four times its previous capacity, and added an extensive foundry for casting car-wheels... hence the change in the name of the works.' So Blain Brothers Car and Car Wheel Works in the 1880s is correct. One claim removed: 'Proof of it sits in Washington today, in the Smithsonian's collection of American business history.' The check could not source that at all, and it is the kind of specific institutional claim a reader can check in thirty seconds. Dropped rather than hedged. The rewrite uses what the original missed entirely: the works had already failed and been shut down before the Blains bought it and made it the town's biggest employer.
Checked against the record. · Checked 2026-07-25
Checked against the record. · Checked 2026-07-24
Checked against the record. · Checked 2026-07-23
Checked against the record. · Checked 2026-07-22
Checked against the record. · Checked 2026-07-21
Checked against the record. · Checked 2026-07-20
Checked against the record. · Checked 2026-07-15
Checked against the record. · Checked 2026-07-14