Who Lit the O'Leary Barn

1. What this report rests on

The Board of Police and Fire Commissioners sat from 23 November to 4 December 1871 and examined fifty-one witnesses. A shorthand writer took down every word, and the transcript, about 1,100 handwritten pages, survives at the Chicago History Museum. The museum's online transcription was unreachable while this report was being written (its server returned an error on every attempt), so the transcript was not read directly. Three substitutes stand in for it, and the report says at each point which one a statement comes from.

The first is the daily report of every sitting printed by the Chicago Tribune, one of the city's morning papers, from 24 November to 5 December 1871, read in full from the Library of Congress's online newspaper collection. Two of those pages are missing from that collection (page 4 of the issue of 30 November and page 2 of 5 December, which report the sittings of Wednesday 29 November and Monday 4 December); they were recovered from another set of scans. These reports are reported speech with some direct quotation, written the same evening. The second is the set of affidavits, sworn statements taken on 15 October 1871 by Michael McDermott, a notary public and the city surveyor, from Catherine and Patrick O'Leary and from Daniel Sullivan, printed in the Tribune of 20 October and reprinted by Andreas. The third is Richard Bales's own transcription of the witnesses to the fire's origin, printed as Appendix D of his 2002 book and recovered here in fragments through an online search service. Where a witness's actual words matter, the fragment is quoted and marked as such.

The other primary texts are the Board's report of 12 December 1871 as printed in the Chicago Evening Post, another daily; the Chicago and out-of-town newspapers from 9 October onward (about 120 issues read in machine-transcribed text); Edwards' city directories for 1869 to 1871, the annual lists of residents with their trades and addresses; the interviews of neighbours and firemen that Andreas took in 1884 and 1885 and printed in volume 2 of his History of Chicago (pages 701 to 719, with his plan of the premises on page 717); the weather observations of the Army Signal Service, then the national weather service, as transcribed by H. A. Musham, whose 1940 study is the longest older account of the fire's first hours; the Greeley-Carlson atlas of 1891, a property atlas of the city; the Sanborn fire-insurance map of 1906; the county's present-day lot boundaries for the block; and a stereoscopic photograph of the O'Leary cottage taken after the fire by J. H. Abbott.

The books and articles engaged are Musham (1940), who affirms the cow story; Bales (1997 article, 2002 book, 2005 paperback), who accuses Daniel Sullivan; Carl Smith (2011 essay, 2020 book), who holds the cause unknowable; Fred Fedler (1985), a journalism scholar who traced how the cow story got into print; Herman Kogan and Robert Cromie (1971), two Chicago journalists; and the press coverage of the fire's 150th anniversary in 2021. Bales's book could be read only through two search services that return fragments of a book rather than its pages, because its digital copies are lending copies that cannot be read or searched in full: 128 Google Books queries returned 534 snippets, each tagged with its page number, from 187 different pages; and 435 search-inside queries to the Internet Archive returned 1,439 fragments of about 100 characters each, most of them repeats, which were joined into 76 passages. The fragmentary reading of Bales is the largest gap in the report's sources; section 11 lists the others.

2. The premises, to the foot

The O'Leary lot was the east half of Lot 12 in Block 38 of the School Section Addition, the 1835 subdivision that laid out this part of the West Side: 25 feet on DeKoven Street by 100.75 feet deep, running north to a 15-foot alley opened in 1855. Its west line is 133.4 feet east of the 1871 east line of Jefferson Street. These figures come from the lot dimensions on the 1891 atlas and from the present-day parcel of the city's fire training academy, which now occupies the site and whose corners still sit on the 1835 lot lines. Everything below is placed in that frame: x is feet east of the Jefferson line, y is feet north of the DeKoven building line, the line along which the house fronts stood.

Two disagreements in the record could not be settled and are carried through the model as alternatives: the barn's orientation (Andreas's plan against Musham's) and the alley width (15 ft by city ordinance, 16 on Andreas's plan, "about twelve" in Sullivan's testimony). The model uses 15.

3. The night, in the witnesses' words

The household went to bed early. Catherine O'Leary had milked at half past four, fed the horse beside the fence at about seven and put him in the barn. Patrick O'Leary and his wife swore on 15 October, and again at the inquiry, that neither had carried a lamp to the yard or barn that evening and that both were asleep when roused.

She milked her cows at 4½ p. m. and 5 a. m., as Mrs. O'Leary peddled the milk. Mrs. O'Leary fed the horse beside the fence at about 7 o'clock p. m., and then put him in the barn. She had no lamp in the yard or barn that night or evening. Patrick O'Leary testifies that he was not in the barn during that day or night; left the feeding of the cows and horse to his wife and daughter; that both were in bed when awakened by Denis Ryan, of No. 118 DeKoven street.

Affidavit before Michael McDermott, 15 October 1871; Chicago Tribune, 20 October 1871, p. 2; Andreas II:708. "Denis Ryan" is the neighbour called Dennis Regan below.

Patrick told the Board that his wife went to bed about eight and he about half past, that both barn doors stood open, the alley door and the south door nailed back so it could not be shut, and that "on the right hand side of the barn, going north, there were some shavings and wood, and in a shed outside were some coal and more wood". Daniel Sullivan, the drayman with a wooden leg who lived across the street at No. 134, had sat with the O'Learys that evening and left when they turned in. His account is the only description of the fire's first minutes.

He left the house, and, while on the opposite side of the street, at twenty or twenty-five minutes past 9 o'clock, he saw fire in Leary's barn. He ran across the street as fast as he could, he has a wooden leg, and cried out "Fire" as loud as he could, which was very loud, as he has strong lungs. He entered the barn, intending to cut the ropes with which the cows were tied. He cut two of them loose, but as they did not seem inclined to leave, and the fire was increasing, he thought he had better depart. … The fire appeared to be on the right side of the barn [where, Leary said, the shavings were kept. Rep.]. … He had never noticed shavings on the right side of the barn. Wood and shavings were kept in the shed adjoining the barn. There was no fire in the shed when he got there. … There was a vacant shed, fronting on the alley, opposite Leary's barn, in which the boys of the neighborhood were in the habit of congregating; in front of it, facing Taylor Street, was a vacant house, which, he understood, was often occupied by vagrants and loafers as a lodging-place.

Daniel Sullivan before the Board, 25 November 1871, as reported in the Chicago Tribune of 26 November 1871, p. 1 (machine transcription corrected against the scan). His affidavit of 15 October has "he found the hay in the loft on fire".

Sullivan's own words on where he sat survive in Bales's transcription: "I sat down on the sidewalk, a big high sidewalk that was there. There was a great big high fence not as high as that door. I sat [against it]", by a fence "at the head of White's house or head of his lot" (Bales, Appendix D, transcript p. 256 and following; chapter 4, p. 113). In his affidavit six weeks earlier he had said only that "he went a few lots east of O'Leary's on the opposite side of DeKoven street". The Board's report placed him "on the sidewalk on the south side of DeKoven street, and nearly opposite" the premises, and Andreas in 1885 had him "sitting in front of his own house, nearly opposite the O'Leary place". The four placements are different spots along the south side of the street; sections 5 and 6 test each of them.

Dennis Regan of No. 112, a neighbour down the street, had also called at the O'Learys', about half past eight, found them in bed, gone home to bed himself and been woken by a neighbour's cry. The McLaughlins in the front cottage had company that evening for a cousin newly arrived from Ireland. Mrs McLaughlin's testimony gives the party as five young men and two young women; there were two tunes on the fiddle and two dances; half a gallon of beer was fetched from outside once or twice; nothing was cooked and no stove was lit; and, she swore, nobody was sent to the barn for milk. William Lee, in the middle cottage two doors east, saw the flames and at first "thought the flames were in James Dalton's shed", ran to Goll's drug store on the corner, which held the key of the street alarm box, and was refused the key. James Dalton, next door east, remembered in 1884 that "at about fifteen minutes to nine o'clock" he "heard a woman's scream of terror" from the direction of the O'Leary premises through his west windows, and that about ten minutes later his mother-in-law saw reflected firelight on Forbes's cottage. The scream has only one published explanation: Musham attributed it to Mrs O'Leary and the kicked lamp. Two young women were dancing in the McLaughlins' front room, forty feet from Dalton's windows, that evening, which is at least as likely a source for a scream as Mrs O'Leary in the barn; there is no evidence to choose between them.

The light of the barn burning was seen from the second storey of Shults's saloon at the corner, from a window on Twelfth Street, from a porch on Lytle Street and from the courthouse cupola downtown. Richard Riley, who saw it from Twelfth and Canal, reached DeKoven Street "while the fire was still confined to the barn" and found that "Dennis Sullivan had just returned to his own house, across the street from O'Leary's, from the vain attempt to save something from the barn. Everything was destroyed except a calf, which was placed in the lot next to Sullivan's."

4. The clock

Forty-six statements of the time were tabulated from the affidavits, the inquiry reports, the Andreas interviews and the fire department's records (the table is published with this report as timeline.csv). They fall into three clusters.

The department's cluster does not support the third. If the glow was visible from Dearborn Street, downtown, at 9.10 by a fireman's watch and the first engine was already pumping on five burning buildings at 9.17½ by its own clock, ignition on department time was no later than about 8.55 and probably 8.45 to 8.50. Andreas drew that conclusion in 1885 ("the fire must have originated by nine o'clock or before; the flames had gained sufficient brilliancy to illuminate the heavens half an hour before an alarm was sounded"), and Bales built on it. What the tabulation adds is the size of the offset: the neighbourhood's clocks and the department's clocks disagree by ten to fifteen minutes, where the two headline times, 8.45 and 9.30, suggest forty-five. The 9.30 rests on no timepiece: it derives from Sullivan's estimate, which the Board adopted and the Chicago Evening Journal, an afternoon paper, printed, and no witness who cited a clock or watch gave it.

Three consequences follow. First, the O'Learys had been in bed for less time than the later hour suggests. Patrick told the Board that his wife went to bed about eight and that he followed her half an hour later; set against ignition at 8.30 to 8.50 by the neighbourhood's clocks, that puts him in bed about ten minutes (none to twenty) and his wife about forty (thirty to fifty) when the barn caught, on times for which the O'Learys cited no clock. Musham's remark that they were "very good sleepers" assumes them "in bed and asleep at 8:30 p.m. and for the next half hour", and Bales puts them asleep at 8.30 with the fire no later than 8.45; the shorter interval for Patrick weakens the remark as far as it rests on the time asleep, and whether they could sleep through the McLaughlins' party is a different question, which the timing does not settle. Second, Sullivan's sequence is internally consistent once his absolute times are discarded: an hour at the O'Learys' from about a quarter past seven, twenty to twenty-five minutes on the sidewalk, the fire seen at about 8.40. Third, the fire had been burning fifteen to thirty minutes before any engine arrived, which explains why it was beyond control when they did; and the Board, whose inquiry was chiefly into the department's conduct, had an interest in the later time, which shortens that interval.

5. Bales's case against Sullivan, tested

Richard Bales used the Chicago Title tract books, the company's ledgers recording every sale of every lot, together with the full transcript and the maps made after the fire, to argue in 1997 and 2002 that Sullivan could not have seen the fire from where he said he sat, that his story therefore covers his own presence in the barn, and that he, perhaps with Regan, started the fire with a pipe, a match or a lamp while visiting his mother's cow. In interviews he is "100 percent convinced"; in the book he writes "No one knows what really happened in the barn on October 8, 1871" and presents his reconstruction as a sequence of steps each qualified by "perhaps". His argument has six parts. Each is stated in his words as far as they could be recovered, then tested.

5.1 Where Sullivan sat

Bales: Sullivan "was sitting on the sidewalk, by a fence 'at the head of White's house or head of his lot.' William White did own land east of Sullivan's property; it was across the street, on the south side of DeKoven Street, and southeast of the O'Leary and McLaughlin homes. But in 1871 William White owned two contiguous parcels of land on DeKoven Street, a 'west parcel' and an 'east parcel,' each twenty-five feet wide. In front of which one did Sullivan sit?" He answers by measuring "conservatively" from the midpoint of the west parcel (pp. 113, 195 to 199).

Test. Bales's reading of the tract books (pp. 137 to 138, 195 to 199) has Catherine Sullivan, Daniel's mother, buying all of Lot 18 in 1864 and the west half of Lot 19 in 1865, Daniel living on that west half of Lot 19, and White owning "the adjoining land to the east, the east half of lot 19", deeded to him in 1864, with "the adjoining east parcel" added in 1870. Bales puts Sullivan in front of the west parcel because White lived at No. 130 before he owned the east parcel, and adds that "even if Sullivan sat in front of a fence on White's east parcel, his view of the barn still would have been obstructed by Dalton's house". On that reading White's west parcel is x 233.5 to 258.5 (box P2 on the plan) and Sullivan's own house is the parcel west of it, x 208.5 to 233.5 (box P1). The 1891 atlas and the 1906 Sanborn map number the half-lots 136, 134, 132, 130 from west to east, which would put No. 134 one parcel further west (P3); the deeds are the better evidence and are followed here, but the model reports every box. One small new fact bears on why Sullivan sat at White's fence at all: Edwards' directory for 1871 lists "White William, teamster, Chase, Hanford & Co., r. 130 DeKoven" and "Sullivan Daniel, teamster, Chase, Hanford & Co., r. 135 DeKoven" (the 135 is presumably a misprint for 134). The two men drove for the same firm, an oil merchant on South Water Street. If White was Sullivan's workmate, his sitting at White's fence requires no explanation beyond acquaintance, and Bales's question "Why did Sullivan walk across the street, walk past his own home, and sit in front of White's house?" can be answered: White drove for the same firm.

5.2 The blocked view

Bales: "Sullivan could not have seen the fire from his position in front of William White's house, as at least one neighboring home blocked his view." His figure on p. 198 is captioned "Sullivan's alleged view of the barn (blocked by Dalton's house and Forbes's house)", and a diagram by the graphic artist Douglas A. Swanson adds that "even if the Forbes house was set back as far as the Dalton house, a corner of Forbes's home would have nonetheless obstructed Sullivan's line of sight". He anticipates the reply: "one might argue that perhaps Sullivan saw the fire above Dalton's house. However, Sullivan testified that when he entered the barn, the fire had not yet broken through the roof. Sullivan stated in a sworn affidavit that when he went into the O'Leary barn, 'he found the hay in the loft on fire.'" And: "it seems almost certain that Sullivan did not see the fire above Dalton's house. Sullivan never claimed in either the affidavit or his testimony that when he first saw the fire, it had broken through the roof of the barn" (pp. 115 to 118). Sullivan's own words about the barn's interior, in the transcript: "Q. What seemed to be on fire? A. I couldn't tell. I got excited. Q. It was filled with fire? A. It was all smoke and fire and flames."

Test. Section 6 puts the block into a three-dimensional model and asks, for every candidate seat, whether any part of the barn's walls was in view, whether its ridge was, and how far above the ridge a flame had to reach to be seen. The result is that from where Bales seats him, the midpoint of White's west parcel (x 246), Bales is right, and the model gives him a larger margin than he claimed. The barn's walls and ridge are never visible in the model, and nothing below about sixteen feet above the ridge is: the median threshold is 15.6 feet with the eye at the height of a man seated on the raised sidewalk, 11 to 20 feet across the central 80 percent of the model's 400 random draws of the uncertain dimensions, and 14 to 16 feet across the three eye heights tried. From that seat Sullivan could have noticed the fire only as flame well above the roof or as light on smoke and neighbouring roofs, and Bales is correct that Sullivan never described either. The argument considers only one of the four placements of Sullivan in the record. His affidavit says "a few lots east of O'Leary's on the opposite side"; the Board's report says "nearly opposite"; Andreas, who interviewed the neighbourhood in 1884, says "in front of his own house". By the deeds Bales himself recovered, his own house was the parcel next west of White's (x 208.5 to 233.5). From there the barn's walls show in 17 percent of the model's draws, the ridge in 21 percent, and a flame within five feet above the ridge in 61 percent; from the middle of that frontage the threshold is three feet above the ridge in the median draw and between minus one and plus seven feet in the central half. A hay-loft fire in a plank barn can be expected to break through the roof and light the sky within minutes; three of the neighbours first noticed the fire as reflected light on a wall or a window, and Sullivan's phrase "saw fire in Leary's barn" is consistent with his having first seen a glow that he then identified as the barn. From the two half-lots further west, the barn's walls were in plain view through the gap between the O'Learys' two cottages and Dalton's set-back house in 96 to 100 percent of draws, and from directly opposite the O'Leary lot in about 40 percent. The dimensions that decide the case, Dalton's width and setback, the cottage heights and the barn's height, are known to plus or minus three feet at best, and for the parcel west of White's the answer changes within that tolerance. The blocked-view argument shows that one of Sullivan's four recorded placements, the one he gave the Board six weeks after the fire, is hard to reconcile with a direct view of the barn. It does not show that he lied about the fire, and it is not sufficient evidence on which to accuse him of starting it.

5.3 The distance he ran

Bales's Appendix B adds half of White's west parcel (12½ feet), the whole Sullivan frontage (75 feet) and 2½ feet to the middle of the passage, 90 feet in all; takes the diagonal across the 60-foot street, 108.17 feet; and adds the passage from the street to the barn, 100.5 less the barn's 16 feet, 84.5 feet: "about 193 feet, which is more than one-half the length of a football field", a distance it does not "seem possible" for a man "hampered by his wooden leg" to hobble. The sum is right for his placement and his assumptions (a straight diagonal, the Andreas orientation of the barn). It is also not a long distance. A man who had driven a dray for fifteen years on that leg covers 193 feet in thirty to fifty seconds at a hobbling four to six feet a second, and Sullivan himself said he ran "as fast as he could". The distance is not evidence for or against Sullivan's account.

5.4 The time

Bales shows from Andreas's witnesses that the fire began by 8.45 and that Sullivan's "twenty to twenty-five minutes past nine" cannot stand. Section 4 agrees and adds that the same is true of the Board's and the department's headline times; Sullivan's error is the same as that of every witness who cited no timepiece, and the Board's report, which Bales reads as a cover-up, makes it too. Since the official record as a whole makes the same error, Sullivan's error about the time tells nothing about him in particular.

5.5 The motive to be in the barn

Sullivan told the Board that "his mother kept a cow, and he frequently went to Leary's barn to get feed; and thus knew how the cows were tied". Bales reads this as the reason Sullivan was in the barn that night. The same statement is also the reason he could find the cows in the dark and cut two loose, which is what he said he did; and Bales's own tract books show that his mother had owned a house and barn of her own on Lot 21 since September 1870, which weakens the reason for him to be in O'Leary's barn at all. The peg-leg story, that the leg was found stuck in the ruins next day, is told by Musham as fact and by Bales from a fireman's reminiscence; if true, it shows only that Sullivan went into the burning barn, which every account, Bales's included, accepts.

5.6 Regan

Bales asks whether Regan was "an accomplice" because the affidavit says Regan woke the O'Learys and the inquiry testimony says Sullivan did, because Regan heard the fiddle while running past when Mrs McLaughlin said the music had stopped, and because Regan did not mention Sullivan. Catherine Sullivan's testimony ("her son and Dennis Regan woke the O'Learys") reconciles the first point; the other two are the kind of disagreement to be expected from people describing a frightening ten minutes six weeks later. None of the three points is inconsistent with Regan and Sullivan having each run to the barn independently.

6. The sightline model

The block was modelled as a height map, in which every building is represented as a box with a gabled roof, every fence as a thin wall, and the street and lots as ground level. For each candidate seat, and for each of three eye heights (seated on the ground, seated on the raised plank sidewalk, standing), the model asks whether a straight line from the eye to a point on the barn clears everything in between. Ten points on the barn's south and east walls test whether any of the barn's body was visible; above each of the barn's four corners and its centre the model searches step by step for the lowest height that can be seen, and reports the lowest of the five as feet above the ridge. Every uncertain dimension is drawn at random from the range in Table 2; 400 draws were made for each of the six seats and three eye heights, 7,200 in all. The code, the input file with a source line for every range, and the raw results are published with this report.

The percentages are the share of the 400 draws in which the feature was visible; the last column is the middle value of the draws, with the values below which 10 and 90 percent of them fall. A negative "height above ridge" means the barn's wall itself is visible below the eaves. Seated on the ground and standing give the same picture within a few points and move the threshold by about a foot either way (for Bales's seat, medians of 16.4, 15.6 and 14.3 feet); the full table for all three eye heights is summary.csv.

The model does not contain trees, the shed the O'Learys kept coal and wood in (placed in the passage in 40 percent of draws), any building on the vacant lots, smoke, or the light of the fire itself. It treats fences as opaque and buildings as solid to the ground, though Dalton's stood on four-foot posts; a line of sight under the house was considered by Bales (p. 115) and rejected here for the same reason, the space would have been cluttered. The model's inputs are known only to within a few feet, and for the parcel west of White's the result changes within that range, so the model cannot determine whether Sullivan could see the barn from there. What it establishes is how sensitive Bales's argument is to those dimensions.

7. Each theory against the record

8. Ranking and sensitivity

The ranking in the verdict panel is the central judgment. Every figure in it depends on two choices: that the scenarios in section 7 are the complete set, and that the evidence in sections 3 to 7 carries the weights given there. The figures are judgments, and they are printed so that a reader can substitute other weights. The three readings that move the ranking most are shown here, each applied alone with the others held at the central values.

Sullivan's column includes Regan as accomplice; "Other" is arson, spontaneous heating and unclassified accident. Reading A corresponds to Bales's position, reading B to Musham's, and reading C to the Cohn family's. None of the three is supported by evidence that can be verified, and the central row is therefore retained.

Two further observations constrain the ranking. The yard and barn were in complete darkness: the sun set at 5.30, twilight ended before 6, and the moon had set at three in the afternoon and would not rise until after midnight. Whoever was in the barn carried a light or struck one; this is why an open flame heads the ranking and why smoking is counted within that scenario rather than separately. The fire began on the shavings side and reached the loft within minutes; any of the flame scenarios produces that sequence, and the chimney and hay scenarios produce it less readily.

9. What is new here and what was already known

Already known. The Board of Police found in 1871 that the cause could not be determined from the evidence; Andreas restated that in 1885 and Carl Smith in 2011 and 2020, and this report agrees. Bales showed that the case against Mrs O'Leary was never more than hearsay, a conclusion the Chicago City Council adopted by resolution on 28 October 1997, and Fedler showed that the legend was in print before any reporter could have invented it as later claimed; both stand. DeBartolo found in 1997 and 1998 that Cohn's recorded will contains no confession; that finding is his, and it is repeated here only because the contrary is often printed.

Where this report differs. It differs on one point, Bales's identification of Sullivan. The sightline argument and the tract-book research are Bales's own, and any model of the block depends on the latter; but at the precision the sources allow the sightline result does not establish whether Sullivan could see the barn, and the other parts of the case (the distance, the time, the motive and Regan) are not sufficient on their own.

New evidence. Five documents not previously used in the published discussion (as far as could be checked against a book read only in fragments): the 1871 directory entries showing Sullivan and White as teamsters for the same firm, Chase, Hanford & Co.; the Tribune's reports of the sittings of 29 November and 4 December, on pages missing from the Library of Congress set and recovered from another scan; the silence of the Illinois Staats-Zeitung, the city's German-language daily, on the inquiry and on the Board's report; the dating of the Evening Journal's interview with Mrs O'Leary to 21 October; and a second Chicago extra of 9 October carrying the cow story.

New analysis of old evidence. The sightline model adds no document to the record. It takes its dimensions from the 1891 atlas, the 1906 Sanborn map, the present-day parcel, one photograph and Bales's deeds, and computes what each placement of Sullivan already in the record would have let him see, carrying the uncertainty of every dimension through. The three-cluster reading of the 46 time statements, and the ten-to-fifteen-minute offset between the neighbourhood's clocks and the department's, is a re-reading of the Tribune's and Andreas's own reports. The observation that the 5 a.m. milking was in darkness, so that a lamp was kept in the barn as a matter of routine and its remains explain nothing, is an inference from computed sunrise and the O'Learys' stated routine. The finding that Bales's 193 feet depends on the seat he chose is a re-working of his own Appendix B arithmetic. Whether Bales makes the milking-in-darkness point could not be checked, since his chapter 2 was read only in snippets.

10. What would overturn this verdict

Five kinds of evidence would overturn it. The full transcript, read in the original, might hold a statement by Sullivan or Regan about what they saw from the street that fits no placement at all, or an admission by anyone that they were in the barn. A record of Dalton's house (a builder's contract, an insurance survey, a photograph) would fix its width and height: with the house two feet lower or five feet narrower, the ridge is visible from White's parcel in most draws; with it two feet higher, nothing under ten feet of flame is. The 1870 or 1872 directory, or a deed, would settle whether No. 134 was the east half of Lot 18 or the west half of Lot 19. A text from October to December 1871 in which Mrs O'Leary, or anyone in the house, admits being in the barn would change the ranking; none was found in about 120 issues of 1871 papers, but the Times of 18 October and the Journal extra of 9 October were read only through later quotation. Documentation of Cohn's story from before 1942 would do the same for the alley scenario.

11. What this report could not do

The inquiry transcript was not read. The museum's transcription site was down throughout, its image server is behind a bot wall, and Bales's Appendix D could be recovered only in fragments of about 100 characters. Bales's book itself was read through the two fragment services described in section 1 (128 Google Books queries, 534 page-numbered snippets from 187 pages; 435 Internet Archive search-inside queries, 1,439 fragments joined into 76 passages); it was not read in full. His Figures 54, 64 and 78 and his 1997 article were not seen; the Appendix B arithmetic and the deeds were recovered as text. Every quotation from him above is a verbatim fragment and carries its page number where the fragment did.

Carl Smith's 2020 book was read through its excerpt in American Heritage magazine and through interviews. Sawislak 1995, Pauly 1984, Miller 1990, Cromie 1958 and Kogan and Cromie 1971 were not obtained, and Ahern's 1911, 1915 and 1921 pieces and Cohn's 1942 to 1944 documents are known only through later quotation. The barn's orientation and the width of the passage remain unsettled between Andreas's plan and Musham's; the model carries both. The heights of Dalton's house, of the O'Learys' two cottages and of the cottage at No. 133 rest on one photograph and on typical construction. The 1870 and 1872 directories, the HathiTrust digital library, the Chicago Collections portal and the museum's image server were unreachable; no paywall or bot wall was bypassed.

12. Sources and reproducibility

Primary. Chicago Tribune, 11 October to 16 December 1871 (Library of Congress, Chronicling America, lccn sn82014064; 30 November p. 4 and 5 December p. 2 from fultonhistory.com); the affidavits of 15 October 1871 (Tribune 20 October, p. 2); the Board's report (Chicago Evening Post, 12 December 1871, transcribed at chicagology.com; Andreas II:709 to 711); Report of the Board of Police for the year ending 31 March 1872, pp. 12 to 13 (archive.org); about 100 out-of-town issues, 9 October to 1 December 1871 (Chronicling America); Edwards' Chicago directories for 1869 to 1870 and 1871 (archive.org); A. T. Andreas, History of Chicago, vol. 2 (1885), pp. 701 to 719 (archive.org, page images and OCR); the Greeley-Carlson Atlas of Chicago, vol. 1 (1891 to 1892), plates 27 to 28 (Northwestern University); Sanborn 1906, vol. 1, sheet 110 (Library of Congress); Cook County parcel 17-16-324-029; J. H. Abbott's stereograph of the O'Leary cottage (Chicago History Museum, ICHi-02741, via greatchicagofire.org); Signal Service observations as transcribed by Musham; astronomical times computed for Chicago local mean time.

Secondary. H. A. Musham, "The Great Chicago Fire," Papers in Illinois History 1940 (archive.org, full text); Richard F. Bales, The Great Chicago Fire and the Myth of Mrs. O'Leary's Cow (McFarland, 2002; Google Books rppt7mgfiKMC snippets; Open Library search-inside of archive.org item greatchicagofire0000bale); Carl Smith, "The O'Leary Legend" (greatchicagofire.org, 2011) and Chicago's Great Fire (2020, via American Heritage, November 2020); Fred Fedler, "Mrs. O'Leary's Cow and Other Newspaper Tales" (AEJMC 1985, ERIC ED257108, full text); R. S. Critchell, Recollections of a Fire Insurance Man (1909, archive.org); the Chicago landmark designation report (1971); NWS Chicago, "The Great Chicago Fire of 1871" (2021); Anthony DeBartolo, Chicago Tribune, 8 October 1997 and 3 March 1998 (Cohn; the Newberry Library's mock trial of March 1998); Steve Mills, "Mrs. O'Leary, Cow Cleared by City Council Committee," Chicago Tribune, 6 October 1997; Knight-Ridder, 12 October 1997 (Baltimore Sun); the Chicago City Council's resolution of 28 October 1997, its text as reproduced by the Glessner House Museum, "Chicago Fire Stories Part III: Catherine O'Leary" (19 October 2021), the primary record being the Journal of the Proceedings of the City Council of the City of Chicago for that date, which was not reached online; Associated Press, 7 October 2021 (the John Marshall Law School mock trial).

Reproducibility. analysis/plan_inputs.json holds every dimension with its source; analysis/sightline_mc.py runs the model (python3 sightline_mc.py 400, about five minutes) and writes summary.csv and one results file per seat; analysis/draw_figures.py draws the two figures; analysis/timeline.csv is the time table; maps/plan/plan_coords.py derives the coordinate frame from the atlas and the parcel. Notes on the 1871 press (press/REPORT.md), the block plan (maps/REPORT.md), the later literature (literature/REPORT.md) and the recovery of Bales's text (bales/REPORT.md, with the fragment log) record the URL and fetch date of every file used.