subject: incandescent-lamp
status: passed_with_open_items
reviewed_by: independent review agent (fourth reviewer)
reviewed_on: 2026-10-02
checks:
- check: 'conformance: no errors for this subject'
  result: pass
  detail: 'whole repository run: 15 subjects and 107 shared nodes, 0 errors, 92 warnings; 2 warnings are for this subject (lamp-swan-in-use and lamp-edison-oct-1879-test are established but anchor_checked is secondary)'
- check: headline present, 70-95 characters
  result: pass
  detail: '95 characters (at the upper limit)'
- check: headline claim is established/refuted at high confidence with a primary check, or status is draft
  result: pass
  detail: 'lamp-edison-sole-inventor: refuted, high, anchor_checked primary; headline_status draft'
- check: headline_status is draft or review before the gate passes
  result: pass
  detail: draft
- check: assessment present with question, answer, headline, text, basis, would_settle
  result: pass
  detail: 'answer: partly'
- check: assessment has exactly three key_points
  result: pass
  detail: '3'
- check: assessment states no percentage
  result: pass
  detail: ''
- check: taxonomy entry (area and question types)
  result: pass
  detail: ''
- check: every claim has statement_kind
  result: pass
  detail: ''
- check: every searched gap names its next step
  result: pass
  detail: ''
- check: sources manifest exists
  result: pass
  detail: ''
- check: every source has an authenticity block
  result: pass
  detail: ''
- check: log has entries
  result: pass
  detail: ''
- check: timeline present (dates exist, so a timeline is required)
  result: pass
  detail: ''
quotes:
  found_in_stored_sources: 0
  not_found_in_stored_sources: 39
  no_stored_source_to_check_against: 0
  note: 'Nothing is stored in the subject folder, so the tool cannot match any quote. The reviewer opened the sources and matched the quotes by hand (see notes); the quotes that rest on the New York newspaper pages could not be matched because the Library of Congress host was not reachable.'
rounds:
- round: 1
  reviewed_on: 2026-10-02
  note: 'Two earlier review passes recorded no review.yaml; their findings are entered here as round 1 with the author responses taken from build/subjects/incandescent-lamp/log.yaml (entries L-10 to L-18; the requester cited them as L-13, L-16, L-17 and L-18, which in the file are L-13 to L-18 plus L-11 and L-12 for the blocking items). The author commits are 8495c38 (the revision content was swept into that commit by a parallel commit on the shared index) and b2dfecd (the empty commit that describes it). Every response was read against the sources by this reviewer, who did not write the dig.'
  requests:
  - id: R1-1
    target: 'claims.yaml headline, search_summary, assessment'
    severity: blocking
    finding: 'The headline settled Edison versus Swan ("first practical to make and sell") although the dig''s own claims leave it open (Swan lamps at Cragside 1880 and the Savoy on 28 Dec 1881 precede Pearl Street on 4 Sep 1882); the headline was 115 characters.'
    author_response: {action: fixed, note: 'L-11 and L-15: headline replaced by a 95-character line that says what the documents show and what stays open (Edison''s priority over Swan); headline_status stays draft; "first practical to make and sell" removed from lamp-edison-sole-inventor and made a contested claim, lamp-first-practical-priority; assessment block added (answer partly, three key points).', commit: 8495c38}
    reviewer_check: 'Counted the headline (95 characters). Read lamp-first-practical-priority against 47 F. 454, the Edison Papers essay, The Times transcription and the IET blog: contested, low confidence, positions and a named disputant, and its statement matches what the sources say (Wallace J. quotes verified in 47 F. 454; "the first lamp that answers the purpose we have wished" verified in the Edison Papers essay; Savoy and Cragside dates verified in the transcription and the blog). The Pearl Street date rests on the New-York Tribune page, which could not be opened (see open_items). The wording "patents date to 1841" in the new headline is dealt with under R2-1.'
    resolved: true
  - id: R1-2
    target: 'search_summary, headline, claims on the "six who built lamps"'
    severity: blocking
    finding: '"At least six people built lamps before Edison" rested on patents (designs, not builds) and on the four 1890s court cases, which are the Goebel cases the dig itself refutes; Swan''s 1878 lamp was a rod that blackened the globe (Chemical News 39).'
    author_response: {action: fixed, note: 'L-11: the "six built" and "four court cases" wording removed; patents are described as designs (S3); Goebel counted as two cases against one; Swan''s December 1878 lamp stated as a rod that blackened its globe; lamp-edison-sole-inventor no longer says Lodygin or Woodward "built" lamps.', commit: 8495c38}
    reviewer_check: 'Searched claims, threads, tests and timeline for "six", "built" and "four courts": none left in surface text (the remaining hits are comments in basis fields that describe the correction). Opened Chemical News 39 (28 Mar and 18 Apr 1879): the chairman Swan "described an experiment ... a slender rod of carbon enclosed in an exhausted globe", the rod "glow with great splendour", "the glass became coated with a sooty deposit on its inside" and the rod "became curved to the form of a bow, as if from softening"; Proctor''s note says Swan "exhibited an electric lamp ... which had broken down". Matches lamp-swan-1878.'
    resolved: true
  - id: R1-3
    target: lamp-10000-ways-quote
    severity: blocking
    finding: '"Not documented in Edison''s lifetime" cannot rest on Dyer and Martin 1910 (should read "earliest version found", with the absence caveat in the confidence reasons).'
    author_response: {action: fixed, note: 'L-12: reworded to "earliest version found"; confidence held at provisional (absence_anchor); confidence_reasons say why; weight lowered from 4 to 3.', commit: 8495c38}
    reviewer_check: 'Opened Dyer and Martin vol. II (archive.org, edison.-his-life-and-inventions-volume-ii, 1910), pp. 615-616: Mallory says Edison had "made over nine thousand experiments" on the storage battery and replied "Results! Why, man, I have gotten a lot of results! I know several thousand things that won''t work."; "his experiments had run up to over ten thousand". Matches the claim. Also opened the Quote Investigator page: it places the same 1910 text as its earliest evidence and lists 1946 (999) and 1982 (10,000); it also lists a 1921 American Magazine interview and a 1882 Christian Union article by Edison that the dig does not record (R2-9, non-blocking).'
    resolved: true
  - id: R1-4
    target: 'claims: lamp-sawyer-man-priority (a), lamp-edison-bamboo (b), lamp-de-moleyns-patent (c), lamp-de-changy-belgium (d), lamp-lodygin-1872-74 and lamp-lodygin-first-claim (e), lamp-swan-1878, lamp-swan-early-1879, lamp-swan-1850s (f)'
    severity: blocking
    finding: 'Factual errors: (a) the Supreme Court holding in 159 U.S. 465 (1895) was misstated; (b) bamboo and Japan wording; (c) the de Moleyns 1841 abridgment (p. 62) says only a means of generating the electric light by pulverized charcoal in an exhausted globe, and "glows" and "incandescent" are not its words; (d) de Changy''s nationality; (e) Kommersant doc 5338615 versus 6822124 misattribution; (f) Swan date wording (Chemical News 39 "immediately preceding Christmas"; Chemical News 42 "almost two years ago" and "about twenty years ago").'
    author_response: {action: fixed, note: 'L-13 (correction of L-04: the Court said the claims except the third are too indefinite, priority becomes unnecessary, decree affirmed) and L-14 (b) to (f): bamboo and Japan restated from the opinion, de Moleyns restated from the abridgment, de Changy''s nationality left unsettled, the 1947-48 campaign statement moved to doc 5338615, Swan wording restated.', commit: 8495c38}
    reviewer_check: '(a) fixed: read 159 U.S. 465 in the Library of Congress scan (OCR): "we are all agreed that the claims of this patent, with the exception of the third, are too indefinite to be the subject of a valid monopoly", priority "becomes unnecessary", "the decree of the Circuit Court is Affirmed", third claim admitted not infringed; heading misprint "1815" confirmed. (b) fixed: "thirty or forty different woods", "despatched a man to Japan", "to different parts of Japan and China sent him about forty different kinds of bamboo" verified; the OCR reads "palmieaf". (c) fixed in lamp-de-moleyns-patent and its node (abridgment p. 62-63 verified: "A means of generating the electric light, by constantly supplying the electrodes at their junction, with pulverized charcoal"; "The whole arrangement is enclosed in an exhausted glass globe"; no "glow" or "incandescent"; the Supreme Court says "finely pulverized boxwood charcoal or plumbago was used for an incandescent electric lamp"), BUT NOT fixed downstream: see R2-1. (d) fixed: Fontaine "M. de Changy, in Belgium" and "M. Jobart, of Brussels" and Howell and Schroeder "a French civil engineer, obtained a Belgian patent" in 1856 verified. (e) the misattribution is fixed (doc 5338615, page dated 6 May 2022, says the campaign began with the article of 9 May 1947 and that within the following year Soviet newspapers reported that Lodygin made the lamp before Edison), BUT the author''s statement that doc 6822124 does not contain the privilege number 1619 and the 2 Oct 1872 priority date is false: see R2-3. (f) fixed: "immediately preceding Christmas", "almost two years ago", "this little lamp, which I showed here two years ago in action", "about twenty years ago" all verified.'
    author_response_round2: {action: fixed, note: 'The two items the reviewer sent back are corrected: (c) the downstream surface text is restated in L-20 (see R2-1), and (e) the Kommersant statement is corrected in L-19 (see R2-3). Re-opened the abridgment pp. 62-63 and 97-98 and Kommersant doc 6822124 in this round.'}
    reviewer_note: 'Parts (a), (b), (d), (e), (f) verified again (Supreme Court OCR; Fontaine; Kommersant 6822124 and 5338615; Chemical News 39 and 42). Part (c) is not closed because the same de Moleyns wording survives in lamp-earliest-patent: see R2-1 and R3-1.'
    reviewer_note_round4: 'Round 4 (third reviewer): part (c) is now closed. Abridgment pp. 62-63 and 97-98 re-read on archive.org (patentsforinvent1766grea): de Moleyns 9053 ''A means of generating the electric light, by constantly supplying the electrodes at their junction, with pulverized charcoal ... The whole arrangement is enclosed in an exhausted glass globe'', no ''glow'', no ''heated''; King 10,919 ''intensely heated'', carbon ''enclosed in a ... vacuum'' (OCR ''Tooricelieu''). 159 U.S. 465 (tile.loc.gov, pdftotext) read for (a) (''with the exception of the third, are too indefinite'', ''becomes unnecessary'', ''Affirmed'') and (b) (''thirty or forty different woods'', ''despatched a man to Japan'', ''to different parts of Japan and China sent him about forty different kinds''; De Moleyns ''finely pulverized boxwood charcoal or plumbago ... incandescent electric lamp''). (d) Fontaine pp. 186-187 ''M. Jobart, of Brussels''; (e) Kommersant 6822124 and 5338615 re-read; (f) Chemical News 39 and 42 re-read (''immediately preceding Christmas'', ''almost two years ago'', ''showed here two years ago in action'', ''about twenty years ago''). The downstream wording is fixed in lamp-earliest-patent (R3-1).'
    resolved: true
  - id: R1-5
    target: 'lamp-goebel-claim, lamp-de-changy-belgium, lamp-lodygin-1872-74, the dig description, thread-national-priority, the assessment'
    severity: blocking
    finding: 'Uneven standard: Goebel was "refuted" on absence while Swan''s 1850s work, de la Rue, Davy and Rohde stayed searched gaps; de Changy and Lodygin were "established" on Fontaine alone; the meaning of "invent" differed between claimants; the thread text "every claimant is dug to the same standard" was unproven.'
    author_response: {action: fixed, note: 'L-15: one standard (S1 to S4) stated in the description and applied to everyone; Goebel moved from refuted to contested at low confidence with positions and named disputants; de Changy, Lodygin and Woodward performance moved from established to proposed (single channel); the first-practical claim split out as contested; the thread sentence removed; lamp-lodygin-first-claim and lamp-canada-invented-claim scoped to "first to describe or patent".', commit: 8495c38}
    reviewer_check: 'Verified the Goebel record in 54 F. 678, 56 F. 496, 57 F. 616 and 61 F. 834 (quotes in lamp-goebel-claim all match; two cases against one is right because 61 F. 834 is the appeal in the Wisconsin case). The standard is stated and the claim states are now consistent with S1 and S2 for Goebel, de Changy, Lodygin and Woodward. NOT fully fixed: the assessment text still says "One standard was used for every claimant" and still says the national stories "each ... stretch [one real record] into first", which pre-judges Goebel (contested) and de Changy (the Belgian first claim is not dug at all); the divergence_note calls Goebel''s record "the weakest" although it is the one with about forty affidavits. See R2-2.'
    author_response_round2: {action: fixed, note: 'The assessment text, divergence note and thread no longer say one standard was used for every claimant or that every national story stretches a record into ''first''; they say one stated standard (S1 to S4) is applied, depth differs, Goebel is contested, de Changy''s first-claim is not dug (L-20; see R2-2).'}
    reviewer_note: 'Not closed: see R2-2 and R3-2 (assessment.text still lists Latimer as a national story and says one standard is applied to every claimant while lamp-swan-in-use is established on a blog and a transcription). Goebel (contested), de Changy and Lodygin (proposed), the thread and the divergence note are consistent with S1 to S4.'
    reviewer_note_round4: 'Round 4 (third reviewer): the assessment text no longer lists Latimer as a national story, names the Cragside exception and sends Cragside to its own proposed claim (lamp-swan-cragside, low); Goebel is contested, de Changy and Lodygin proposed, the thread and divergence_note say what the records are. Goebel statements re-matched in 54 F. 678 (''extremely improbable'', ''some forty affidavits'', ''experimental toys''). Residual wording (the ''one exception'' sentence, while the Savoy transcription and the 22 Oct 1879 transcription are also established below primary) is carried as non-blocking R4-4.'
    resolved: true
  - id: R1-6
    target: lamp-goebel-claim
    severity: non_blocking
    finding: 'Goebel courts counted as two cases against one: Colt 54 F. 678 "extremely improbable"; Seaman 57 F. 616 plus the Seventh Circuit appeal 61 F. 834; Hallett 56 F. 496 "fair preponderance".'
    author_response: {action: fixed, note: 'L-16 A1: two against one, with precise quotations and the different burdens of proof.', commit: 8495c38}
    reviewer_check: 'Verified in the four opinions: "extremely improbable that Henry Goebel constructed a practical incandescent lamp in 1854"; "The weight of evidence on this motion is in the direction that he made these lamps, or lamps similar in general appearance"; "assuming they were made, they do not anticipate the invention of Edison. At most they were experimental toys"; "simply impossible, under the circumstances, to believe that a lamp so constructed could have been made by Goebel before 1872"; "a fair preponderance of testimony in support of the Goebel claim"; $20,000 bond; "without designing to suggest what result should be reached upon final hearing"; "such an atmosphere of improbability"; "a defense which puts the case in doubt".'
    resolved: true
  - id: R1-7
    target: 'lamp-goebel-claim, dig description'
    severity: non_blocking
    finding: '"No dated document before 1882", not "no record".'
    author_response: {action: fixed, note: 'L-16 A2.', commit: 8495c38}
    reviewer_check: 'Wording present in the claim, tests.yaml and the thread. The earliest dated traces (Curtis "in the early part of 1882", Eaton "in May, 1882", Dreyer "in 1882", the offer of 28 Nov 1882) verified in 54 F. 678; Goebel''s own application for US 266,358 was filed on 23 January 1882 (OCR of the USPTO scan), a dated document that the dig mentions only in the node.'
    resolved: true
  - id: R1-8
    target: 'lamp-goebel-claim (US 266,358 and US 47,632)'
    severity: non_blocking
    finding: 'The rejection of the second claim of US 266,358 comes from 54 F. 678, not from the patent; US 47,632 needed an anchor.'
    author_response: {action: fixed, note: 'L-16 A3: moved to the node and the basis; US 47,632 added as an anchor and a node.', commit: 8495c38}
    reviewer_check: 'Opened both patents at the USPTO image server and OCR''d them: US 266,358 (filed 23 Jan 1882, patented 24 Oct 1882; two claims, both for flattened, spirally coiled wire sockets); US 47,632 (hemmers for sewing machines, 9 May 1865). 54 F. 678 says "The second claim, as drawn in the original application, was rejected by reference to the Edison patent now in suit."'
    resolved: true
  - id: R1-9
    target: 'lamp-goebel-claim, transmission'
    severity: non_blocking
    finding: 'Curtis "early part of 1882".'
    author_response: {action: fixed, note: 'L-16 A4, new events tx-g-10 to tx-g-12, node lamp-1882-early-curtis-hears-goebel-story.', commit: 8495c38}
    reviewer_check: '54 F. 678: Curtis "first heard of Henry Goebel in the early part of 1882" and was told "as early as about 1850". Matches.'
    resolved: true
  - id: R1-10
    target: 'lamp-latimer-patent'
    severity: non_blocking
    finding: 'The Maxim and United States Electric Lighting rivalry is not in US 252,386.'
    author_response: {action: fixed, note: 'L-16 A5: anchored to the Queens Library essay hosted by the Edison Papers.', commit: 8495c38}
    reviewer_check: 'US 252,386 names only "the United States Electric Lighting Company ... of same place". The essay (opened) says Latimer worked for Maxim, "then chief engineer at the U.S. Electric Lighting Company", that it was "only one of Edison''s competitors", Bridgeport from March 1880, the move to Brooklyn in 1880.'
    resolved: true
  - id: R1-11
    target: 'lamp-latimer-patent, lamp-latimer-invented-claim'
    severity: non_blocking
    finding: 'Add US 247,097 (Nichols and Latimer); anchor King 1845 in the refutation.'
    author_response: {action: fixed, note: 'L-16 A6.', commit: 8495c38}
    reviewer_check: 'US 247,097 opened (filed 18 Apr 1881, dated 13 Sep 1881; "the method of mounting the carbons or connecting them with the wires"). King 1845 (abridgment p. 97) is an anchor of lamp-latimer-invented-claim.'
    resolved: true
  - id: R1-12
    target: 'lamp-edison-high-resistance, lamp-edison-sole-inventor'
    severity: non_blocking
    finding: 'Edison sole-inventor wording should come from 47 F. 454 and 54 F. 678.'
    author_response: {action: fixed, note: 'L-16 A8: the courts are quoted.', commit: 8495c38}
    reviewer_check: 'Wallace J. (47 F. 454): "the first to make a carbon of materials and by a process which was especially designed to impart high specific resistance to it ... By doing these things he made a lamp which was practically operative and successful"; Colt J. (54 F. 678): the second claim "covered a broad and fundamental invention, namely, an incandescent lamp, composed of a carbon filament, hermetically sealed in an all glass chamber exhausted to a practically perfect vacuum, and having leading-in wires of platinum". Lane-Fox (two patents October 1878, one March 1879, and a letter to the Times of December 1878) and Edison''s French patent 130,910 of 28 May 1879 (52 F. 300) verified.'
    resolved: true
  - id: R1-13
    target: 'lamp-edison-sole-inventor (Lodygin and Woodward "built"; "central station by 1882")'
    severity: non_blocking
    finding: 'The facts were not among that claim''s anchors.'
    author_response: {action: fixed, note: 'L-16 A9: removed or moved to the claims that carry them.', commit: 8495c38}
    reviewer_check: 'No "built" statement about Lodygin or Woodward remains in the surface text; lamp-woodward-evans-performance (Electrical World, 14 Apr 1900) verified: "The first incandescent lamp was constructed at Morrison''s brass foundry in Toronto, and was a very crude affair", nitrogen, investors "declined", "left for Europe".'
    resolved: true
  - id: R1-14
    target: lamp-latimer-edison-role
    severity: non_blocking
    finding: 'Anchor mismatch (the Edison Papers carbon-filament essay does not mention Latimer).'
    author_response: {action: fixed, note: 'L-16: anchor is the Queens Library essay, class synthetic, anchor_checked secondary.', commit: 8495c38}
    reviewer_check: 'Essay verified: Engineering Department of the Edison Electric Light Company from about 1885, Legal Department about 1889, testimony, Edison Pioneers 1918.'
    resolved: true
  - id: R1-15
    target: 'transmission.yaml tx-g-01, tx-g-07, tx-g-08, tx-g-09, tx-l-04; threads.yaml overlay_notes'
    severity: non_blocking
    finding: 'tx-g-01 called the "earliest trace" (Eaton May 1882, Curtis early 1882 and Dreyer 1882 are earlier or equal); read: no needed for tx-g-07, tx-g-08, tx-l-04; thread notes without source ids.'
    author_response: {action: fixed, note: 'L-16 B3.', commit: 8495c38}
    reviewer_check: 'tx-g-01 now says it is not the earliest; tx-g-07, tx-g-08 and tx-g-09 are read: no with notes; tx-l-04 cites the Kommersant page that was read and says no Soviet text was; overlay_notes carry source ids.'
    resolved: true
  - id: R1-16
    target: 'sources/MANIFEST.yaml (Davy edition, springe)'
    severity: non_blocking
    finding: 'Davy edition is the American one (Bradford and Inskeep); the springe "debatable" wording is unverified.'
    author_response: {action: fixed, note: 'L-16: manifest corrected; springe marked UNVERIFIED; Canadian Encyclopedia and the notebook viewer recorded as unreachable.', commit: 8495c38}
    reviewer_check: 'archive.org metadata for elementschemica00davygoog: publisher Bradford and Inskeep, 1812, from the New York Public Library; the OCR text of pp. 85-86 matches the claim ("more than half the volume of the charcoal became ignited to whiteness", "a most brilliant ascending arch of light", battery of 2000 double plates). The Canadian Encyclopedia returned HTTP 403 to this reviewer as well.'
    resolved: true
- round: 2
  reviewed_on: 2026-10-02
  requests:
  - id: R2-1
    target: 'claims.yaml search_summary, headline, assessment.text, assessment.key_points[0], lamp-edison-sole-inventor, lamp-lodygin-first-claim, tests.yaml dt-which-first (continuation of R1-4 c)'
    severity: blocking
    finding: 'The de Moleyns claim itself is now correct (the abridgment says only "a means of generating the electric light" with pulverized charcoal between platinum coils "enclosed in an exhausted glass globe", and the dig''s own basis says it "reads like an arc fed with charcoal"), but the surface text still says the 1841 patent describes "electric lamps with a heated conductor in an exhausted globe" (search_summary and key point 1), "glowing conductors in evacuated glass" (lamp-lodygin-first-claim), "a lamp that glows in an exhausted globe ... de Moleyns (1841)" (assessment text), and the headline says "patents date to 1841". Only the Supreme Court (1895) calls the 1841 patent incandescent; the abridgment (pp. 62-63, opened) does not, and the full specification was not read. The 1845 King abridgment (pp. 97-98, opened) is the sound one for a heated carbon conductor ("intensely heated"; carbon "enclosed in a Torricellian vacuum"; the platinum form is under "a glass shade", not stated as exhausted). The dig''s own S3 (a patent shows a design, not a lamp) applies as well to "a lamp that glows".'
    evidence: 'Abridgments of Specifications relating to Electricity and Magnetism (1859), archive.org patentsforinvent1766grea, pp. 62-63 and 97-98; 159 U.S. 465 (Library of Congress scan); the dig''s lamp-de-moleyns-patent basis.'
    asked: 'Restate the 1841 patent in the surface text as an electric light fed with pulverized charcoal in an exhausted globe (the Supreme Court calls it incandescent), carry the heated-conductor wording only for 1845 (and say "carbon" there), and make the headline and key point 1 say what the 1845 patent shows or "patents go back to 1841 and 1845"; add lamp-de-moleyns-patent, lamp-starr-king-patent and lamp-earliest-patent to the assessment basis (key point 1 rests on them but they are not listed). Fix by a new log entry, not an edit of L-14.'
    author_response: {action: fixed, note: 'Verified in the abridgment (pp. 62-63: ''a means of generating the electric light ... pulverized charcoal ... enclosed in an exhausted glass globe''; pp. 97-98: King ''intensely heated'', carbon ''enclosed in a Torricellian vacuum''). Restated in search_summary, assessment text, key point 1, headline (95 characters, draft), lamp-edison-sole-inventor, lamp-lodygin-first-claim (and their would_change_if) and dt-which-first; 1841 is now ''an electric light fed with powdered charcoal in an exhausted globe (the Supreme Court calls it incandescent)''; ''heated carbon conductor in a vacuum'' is carried for 1845 only. Assessment basis now includes lamp-de-moleyns-patent, lamp-starr-king-patent and lamp-earliest-patent. Log L-20.'}
    reviewer_note: 'Partly fixed, verified: abridgment pp. 62-63 and 97-98 re-read (de Moleyns: ''a means of generating the electric light ... pulverized charcoal ... enclosed in an exhausted glass globe'', no ''glow'', no ''heated''; King: ''intensely heated'', carbon ''enclosed in a Torricellian vacuum'', platinum under ''a glass shade''). The search_summary, headline, assessment text, key point 1, lamp-edison-sole-inventor, lamp-lodygin-first-claim and dt-which-first are now correct and the assessment basis lists the three claims. 159 U.S. 465 read: ''pulverized boxwood charcoal or plumbago'' was used ''for an incandescent electric lamp'' in De Moleyns''s 1841 patent. One residue: lamp-earliest-patent (an established claim and a basis of key point 1) still says ''No patent earlier than de Moleyns''s of 1841 for an electric light with a heated solid in an enclosed globe'', which describes the 1841 patent as a heated solid (see R3-1). Headline ''1845'' claim is supported by the King abridgment, 40 F. 21 and 54 F. 678 (a carbon conductor in a Torricellian vacuum), read; the full British specifications were not read (open item).'
    reviewer_note_round4: 'Round 4 (third reviewer): re-checked every surface line that names 1841: search_summary, assessment text, key point 1, headline, lamp-edison-sole-inventor, lamp-lodygin-first-claim, dt-which-first, timeline people and a2 now say an electric light fed with pulverized charcoal in an exhausted globe (the Supreme Court, not the abridgment, calls it incandescent), and carry the heated carbon conductor in a vacuum for 1845 only; the assessment basis lists the three claims; lamp-earliest-patent is restated (R3-1). Small residues are non-blocking (R4-3, R4-5).'
    resolved: true
  - id: R2-2
    target: 'assessment.text, divergence_note, thread-national-priority description (continuation of R1-5)'
    severity: blocking
    finding: 'The assessment text says "One standard was used for every claimant" and that the national stories "(Goebel, Lodygin, de Changy, Woodward and Evans, Latimer) each hold one real record and stretch it into first"; the thread says "Each story takes one real record and stretches it into first. The pattern is similar in every country"; the divergence_note says Goebel and the Canadian sale "have the weakest records". These say more than the claims: Goebel is contested at low confidence (preliminary rulings split two to one; about forty affidavits per Colt J.), de Changy is proposed and his Belgian "first" claim is not dug at all (the dig says so in the thread), Latimer is not a national claim, and the depth and anchors differ (for example lamp-swan-in-use is established at moderate on a blog and a transcription, lamp-10000-ways-quote is refuted on one anchor and absence, de Moleyns and King rest on abridgments). The dig removed the thread sentence "every claimant is dug to the same standard" but the same assertion remains in the assessment.'
    evidence: 'claims.yaml assessment.text, divergence_note; threads.yaml thread-national-priority; 54 F. 678 ("some forty affidavits"), 56 F. 496, 57 F. 616, 61 F. 834.'
    asked: 'Say that one stated standard (S1 to S4) is applied, that depth differs, and that the Goebel and de Changy claims are not shown to be stretched (Goebel: contested; de Changy: proposed, first-claim not dug). Drop "weakest records" or replace it with what the record is (testimony of about forty affidavits, no dated document before 1882).'
    author_response: {action: fixed, note: 'Agreed. Assessment text, divergence_note and thread-national-priority reworded: one stated standard (S1 to S4), depth differs, Lodygin and Woodward-and-Evans are the stretched ones on the documents read, Goebel contested (about forty affidavits, no dated document before 1882), de Changy''s first-claim not dug, Latimer not a national claim; ''weakest records'' replaced. Log L-20.'}
    reviewer_note: 'Mostly fixed, verified in the text: the thread and divergence_note now say what the records are (about forty affidavits per Colt J., no dated document before 1882) and no longer say ''weakest'' or that every story stretches a record. Two residues remain in assessment.text: ''Several national stories (Lodygin, Woodward and Evans, Latimer)'' still lists Latimer as a national story, which this request said he is not; and ''One stated standard (S1 to S4) is applied to every claimant'' is contradicted by lamp-swan-in-use, which is established at moderate although its Cragside half rests on a blog about a letter not read (S1 makes that single-channel); see R3-2.'
    reviewer_note_round4: 'Round 4 (third reviewer): assessment.text now names two national stories (Lodygin; Woodward and Evans) and a Latimer misattribution, says Goebel is contested and de Changy''s first-claim is not dug, and names the Cragside exception; the thread and the divergence_note match. Verified in claims.yaml and threads.yaml.'
    resolved: true
  - id: R2-3
    target: 'lamp-lodygin-1872-74 basis, node lamp-1874-07-11-lodygin-russian-privilege, timeline.yaml a6 note, sources/MANIFEST.yaml src-kommersant-lodygin, log L-14 (e) and L-17 (12) (continuation of R1-4 e)'
    severity: blocking
    finding: 'The author states that Kommersant doc 6822124 "gives NO privilege number" and that the number 1619 and the priority date 2 October 1872 "are not in this page". They are. The page (dated 11 July 2024, opened by the reviewer) says: "11 июля 1874 года родное патентное ведомство выдало Александру Лодыгину «привилегию» № 1619. Интересно, что «приоритет» был указан задним числом, от 2 октября 1872 года." ("On 11 July 1874 the home patent office issued Alexander Lodygin privilege No. 1619. Interestingly the priority was dated retroactively, 2 October 1872"). The node text and the timeline note ("no privilege number in the page read") put a false statement about a source on the page. The Lomonosov prize and General Electric are indeed not in it (correct), and the "1947-48 campaign" sentence is in doc 5338615 (correct).'
    evidence: 'https://www.kommersant.ru/doc/6822124 (retrieved by the reviewer 2026-10-02); https://www.kommersant.ru/doc/5338615.'
    asked: 'Correct the statement in a new log entry, a new revision of the node, the timeline note and the manifest; restore the number and the backdated priority date with their single newspaper source (a later newspaper history, no scan of the privilege), or say that they are not used because they are not independently documented.'
    author_response: {action: fixed, note: 'Disputed my own earlier statement: the reviewer is right. Re-opened Kommersant doc 6822124: ''11 июля 1874 года ... «привилегию» № 1619 ... «приоритет» был указан задним числом, от 2 октября 1872 года''. Number and priority date restored with their single source in lamp-lodygin-1872-74, node lamp-1874-07-11-lodygin-russian-privilege (rev 2), the timeline note (pin rev 2) and the manifest; L-14 (e) and L-17 (12) corrected by L-19. The prize and General Electric are confirmed absent from that page.'}
    reviewer_note: 'Verified. Kommersant doc 6822124 (dated 11.07.2024) read by this reviewer: ''11 июля 1874 года родное патентное ведомство выдало Александру Лодыгину «привилегию» № 1619. Интересно, что «приоритет» был указан задним числом, от 2 октября 1872 года.'' Number and date are restored in lamp-lodygin-1872-74, node lamp-1874-07-11-lodygin-russian-privilege (rev 2, history entry), timeline a6 (rev 2), the manifest and log L-19, with the single newspaper source and ''no scan of the privilege''. The page has no Lomonosov prize or General Electric (checked). The 1947-48 sentence is in doc 5338615 (read).'
    resolved: true
  - id: R2-4
    target: 'lamp-swan-before-edison statement'
    severity: blocking
    finding: 'The sentence runs "Edison''s laboratory kept a thread lamp lit 13 1/2 hours ... and applied for its patent on 4 November 1879; the lamp he showed in October 1880 had a carbon ''extremely thin - a mere hair'', and he said blackening of the globe stopped once the vacuum and contacts were good." By the grammar "he" is Edison, but both quoted statements are Swan''s, from his 20 October 1880 lecture (Chemical News 42, pp. 227-230: "The carbon is extremely thin - a mere hair"). A reader of this contested headline claim would credit Edison with a lamp shown in October 1880 and with the blackening explanation.'
    evidence: 'Chemical News 42 (5 Nov 1880), archive.org s713id13691440.'
    asked: 'Name Swan in both places.'
    author_response: {action: fixed, note: Verified in Chemical News 42 (the 'mere hair' and blackening statements are Swan's). Swan is now named in both places in lamp-swan-before-edison. Log L-20.}
    reviewer_note: 'Verified. Chemical News 42 pp. 229-230: ''The carbon is extremely thin - a mere hair'' and ''there was no blackening of the globes'' are Swan''s own lecture. lamp-swan-before-edison now names Swan in both places (''in his October 1880 lecture Swan described his own lamp''s carbon ...''; ''said (of his own experiments)'').'
    resolved: true
  - id: R2-5
    target: 'sources and transcription of L-17 (2) and (3), lamp-edison-patent basis, lamp-edison-oct-1879-test basis'
    severity: non_blocking
    finding: 'The log and two basis notes say the words "application filed November 4, 1879" and "not gained until October, 1879" were not found in 52 F. 300 and were dropped. They are in the opinion (static.case.law f/52/cases/0300-01): "Lacombe, Circuit Judge. On January 27, 1880, under an application filed November 4, 1879, letters patent No. 223,898 were issued to Thomas A. Edison" and "The knowledge that practical stability could be given to a carbon filament was not gained until October, 1879." The correction itself was a false statement about a source, and the second sentence is a court''s statement relevant to lamp-first-practical-priority.'
    evidence: 'https://static.case.law/f/52/cases/0300-01.json'
    asked: 'A new log entry correcting L-17; consider citing the second sentence as a court statement in lamp-first-practical-priority (it is a court''s view, from the patent suit, not a dated record).'
    author_response: {action: fixed, note: 'Verified: 52 F. 300 has ''under an application filed November 4, 1879'' and ''was not gained until October, 1879''. My earlier statement was wrong; corrected by log L-19, 52 F. 300 re-cited in lamp-edison-patent, the basis notes corrected, and the second sentence cited as a court''s view in lamp-first-practical-priority.'}
    reviewer_note: 'Verified. static.case.law 52 F. 300: ''under an application filed November 4, 1879, letters patent No. 223,898 were issued'' and ''The knowledge that practical stability could be given to a carbon filament was not gained until October, 1879'' are present. L-19 corrects L-17; lamp-edison-patent cites 52 F. 300 again; the second sentence is cited as a court''s view in lamp-first-practical-priority.'
    resolved: true
  - id: R2-6
    target: 'lamp-swan-in-use, assessment.key_points[2]'
    severity: non_blocking
    finding: 'Under the dig''s own S1 a claim resting on a later secondary account of a document not read is single-channel and is not "established". The Cragside half of lamp-swan-in-use rests only on the 2023 IET archives blog (the blog says the arc lamps "were replaced by Armstrong in 1880 with 45 incandescent lamps" and that Armstrong wrote on 17 January 1881; the letter was not read); the Savoy half rests on a transcription of The Times of 29 December 1881 (verified: "when the stage was for the first time lit up by the electric light" and "The Swan incandescent lamps were used"; the original page not compared). Key point 3 states "Swan lamps lit Cragside (1880) and the Savoy stage (1881)" without the limit. The Savoy date alone precedes Pearl Street, so the conclusion survives.'
    evidence: 'https://ietarchivesblog.org/2023/03/09/cragside-harnessing-the-power-of-water/ ; https://gsarchive.net/carte/savoy/electric.html'
    asked: 'Split Cragside into its own proposed claim, or word key point 3 as "per an archive blog" and "per a transcription of The Times".'
    author_response: {action: fixed, note: 'Key point 3 and the assessment text now say ''per a blog'' and ''per a transcript''; Cragside and the Savoy remain one claim at moderate and secondary, shown as a limit by the conformance warning. Not split into a separate claim (the Savoy date alone precedes Pearl Street). Log L-20.'}
    reviewer_note: 'Verified as worded: key point 3 and the assessment text now say ''per a blog'' and ''per a transcript''; the Savoy half matches the Times transcription (gsarchive.net: ''for the first time lit up by the electric light'' and ''The Swan incandescent lamps were used''). Not split into its own claim; the substantive S1 issue left is carried in R3-2 and open_items.'
    resolved: true
  - id: R2-7
    target: 'nodes lamp-1878-12-swan-rod-lamp-newcastle-chemical-society, lamp-1879-02-03-swan-shows-lamp-lit-and-phil (N14)'
    severity: non_blocking
    finding: 'The node `time` values carry days that no source read gives: 18 Dec 1878 (from summaries the dig did not keep, the node says so) and 3 Feb 1879 (from search summaries, access cited_only). date_basis is "derived". Chemical News prints no day (windows 1-24 Dec 1878 and 20 Oct 1878 to 20 Oct 1879 are fine).'
    evidence: 'Chemical News 39 pp. 133 and 168; Chemical News 42 pp. 227-230.'
    asked: 'Use the middle of the window as `time` with date_basis inferred, or say in the node that the day is the best single date from secondary pages.'
    author_response: {action: fixed, note: 'Both Swan nodes revised to rev 2: `time` at the middle of the window, date_basis inferred, text says the days came from secondary pages; timeline pins updated. Log L-20.'}
    reviewer_note: 'Verified. Both Swan nodes are rev 2 with time at the middle of the window, date_basis inferred, history entries naming R2-7 and text that says the 3 Feb and 17 Jan days come from secondary summaries.'
    resolved: true
  - id: R2-8
    target: 'lamp-swan-before-edison, lamp-swan-early-1879 (Swan''s own dating)'
    severity: non_blocking
    finding: 'In the same lecture Swan dates the start of his carbon-in-globe work: "in October, 1877, I sent to Mr. Stearn a number of carbons, made from carbonised cardboard, with the request that he would get them mounted for me in glass globes by a glassblower". The dig quotes the "two years ago in action" and "twenty years ago" passages but not this one, which is the self-account that bears most on the Swan side of the priority dispute (S4: a self-account, primary for what Swan claimed).'
    evidence: 'Chemical News 42 (5 Nov 1880), pp. 227-230.'
    asked: 'Record it in lamp-swan-before-edison (positions or basis) as Swan''s claim, with the same S4 caveat.'
    author_response: {action: fixed, note: 'Verified in Chemical News 42: ''in October, 1877, I sent to Mr. Stearn a number of carbons, made from carbonised cardboard ... mounted for me in glass globes by a glassblower''. Added to lamp-swan-before-edison as Swan''s claim with the S4 caveat.'}
    reviewer_note: 'Verified. Chemical News 42 p. 229: ''in October, 1877, I sent to Mr. Stearn a number of carbons, made from carbonised cardboard, with the request that he would get them mounted for me in glass globes by a glassblower, and then exhaust the air''. In lamp-swan-before-edison positions/basis with the S4 caveat.'
    resolved: true
  - id: R2-9
    target: 'transmission.yaml tx-10000-ways, lamp-10000-ways-quote'
    severity: non_blocking
    finding: 'The Quote Investigator page the author read also reports a 1921 American Magazine interview in which Edison himself recounts the incident without a number, and an 1882 article by Edison in The Christian Union ("the enormous number of things that won''t work"). Neither is in the chain, and the claim''s would_change_if mentions "Edison''s own record". The refutation is therefore not about an absence of any Edison remark of this kind; the claim states only the earliest version found.'
    evidence: 'https://quoteinvestigator.com/2012/07/31/edison-lot-results/'
    asked: 'Add both as chain events (read: no, via Quote Investigator) and keep the claim at provisional.'
    author_response: {action: fixed, note: 'Verified in Quote Investigator (1921 American Magazine interview by B. C. Forbes; 1882-06-15 Christian Union article by Edison). Both added to tx-10000-ways as read: no; the claim stays provisional and says only ''earliest version found''; basis note added.'}
    reviewer_note: 'Verified through Quote Investigator (read through a page-summary fetch, not the raw page, which a bot check blocks): 1882 Christian Union (''the enormous number of things that won''t work''), 1910 Dyer and Martin, 1921 American Magazine (B. C. Forbes), 1946 and 1982. Both earlier items are in tx-10000-ways as read: no; the claim stays provisional and says ''earliest version found''. Dyer and Martin vol. II pp. 615-616 re-read: matches.'
    resolved: true
  - id: R2-10
    target: 'refuted claims lamp-10000-ways-quote, lamp-canada-invented-claim, lamp-lodygin-first-claim (S2, owner decision)'
    severity: non_blocking
    finding: 'Under S2 the absence of an earlier document never refutes; the 10,000-ways claim is refuted at provisional on one anchor plus absence (the claim says so), the second half of the Canadian target ("Edison''s lamp was built on their patent") is refuted by differing designs while the sale is a gap, and the Lodygin target was narrowed to "first to describe or patent" so that the refutation holds. Each is stated openly, and the reviewer does not object to the states, but the owner should know the dig leans on narrowing the target in two places.'
    evidence: 'US 181,613 against US 223,898 (both OCR''d); Dyer and Martin pp. 615-616.'
    asked: 'None required; owner to confirm that "refuted at provisional on absence" and narrowed targets are acceptable under N-rules before publication.'
    author_response: {action: deferred, note: 'No author change required; noted for the owner: the 10,000-ways refutation is at provisional on one anchor plus absence, and the Lodygin and Canada targets are narrowed (stated in the claims and in L-15). Deferred to the owner.'}
    reviewer_note: 'Non-blocking and deferred to the owner by the author; accepted as an open item (refuted at provisional on one anchor plus absence; narrowed Lodygin and Canada targets). The owner decision is listed in open_items.'
    resolved: true
  - id: R2-11
    target: 'timeline.yaml people'
    severity: non_blocking
    finding: '"Mathew Evans" in the people list against "Matthew" in claims.yaml (Electrical World, 1900: "Matthew Evans, a hotelkeeper"; the Canadian patent list prints the initial M.).'
    evidence: 'Electrical World and Engineer 35:15, 14 Apr 1900.'
    asked: 'Use one spelling.'
    author_response: {action: fixed, note: "Verified 'Matthew Evans' in Electrical World 1900; the timeline people list now uses Matthew."}
    reviewer_note: 'Verified: ''Matthew Evans, a hotelkeeper'' (Electrical World 14 Apr 1900); the timeline people list reads Matthew.'
    resolved: true
- round: 3
  reviewed_on: 2026-10-02
  reviewer: 'independent review agent (second reviewer); did not write the dig and did not take part in rounds 1 and 2'
  requests:
  - id: R3-1
    target: 'claims.yaml lamp-earliest-patent (statement, would_change_if); timeline.yaml people entry for de Moleyns (check wording)'
    severity: blocking
    finding: 'lamp-earliest-patent (established, provisional, anchor primary; a basis claim of assessment key point 1) says "No patent earlier than de Moleyns''s of 1841 for an electric light with a heated solid in an enclosed globe has been found" and its would_change_if speaks of "a lamp lit by an electrically heated solid". Read as written, the 1841 patent is that kind of patent. The abridgment (pp. 62-63, read again by this reviewer) says only an electric light fed with pulverized charcoal between platinum coils in an exhausted glass globe; "heated" is in the King abridgment (1845, pp. 97-98). This is the R2-1 error in the one claim R2-1 did not name. Fontaine (p. 169), the claim''s own anchor, calls the 1845 patent "The first document on the question" of lighting by incandescence.'
    evidence: 'Abridgments (1859) pp. 62-63 and 97-98 (archive.org patentsforinvent1766grea djvu text); Fontaine 1878 p. 169; 159 U.S. 465 (tile.loc.gov scan).'
    asked: 'Restate lamp-earliest-patent so that it separates the two things: no patent earlier than 1841 was found for an electric light in an enclosed globe (de Moleyns), and none earlier than King''s 1845 for a heated conductor (Fontaine''s "first document"); keep it provisional (absence). New log entry; do not edit L-20.'
    author_response: {action: fixed, note: 'Verified in the abridgments (pp. 62-63, 97-98) and Fontaine p. 169. lamp-earliest-patent restated as two separate absences (1841 light in an enclosed globe; 1845 heated conductor in a vacuum), would_change_if, timeline note and people entry, and dt-which-first reworded; provisional kept. L-21.'}
    reviewer_note: 'Verified by the third reviewer. Abridgment pp. 62-63 and 97-98 and Fontaine p. 169 (''The first document on the question that we have found, is an English patent of the 4th November, 1845'') read; Fontaine does not mention de Moleyns (no hit for ''Moley''). lamp-earliest-patent now states two separate absences (1841 electric light in an enclosed globe, in the 1859 abridgments; 1845 heated conductor in a vacuum), provisional, with would_change_if, the timeline people entry and note a2, and dt-which-first matching. The claim does not mention that 159 U.S. 465 describes the 1841 patent as an incandescent lamp, which would put a heated solid in an exhausted globe before 1845 if the Court''s reading were right; this is non-blocking (R4-3).'
    resolved: true
  - id: R3-2
    target: 'claims.yaml assessment.text (sentences on "national stories" and "One stated standard ... applied to every claimant"); description (S1 and "applied to Edison''s side as well"); lamp-swan-in-use'
    severity: blocking
    finding: 'The assessment says "Several national stories (Lodygin, Woodward and Evans, Latimer) each hold a real record that is stretched" (Latimer is a refuted misattribution, not a national story; R2-2 said so) and "One stated standard (S1 to S4, in the description) is applied to every claimant". S1 says a claim resting on a later account of a document not read is single-channel and is never established, yet lamp-swan-in-use is established (moderate, secondary) and its Cragside half rests on an IET blog about Armstrong''s letter of 17 January 1881, which was not read. The blog (read by this reviewer) says arc lamps "were replaced by Armstrong in 1880 with 45 incandescent lamps" and that the letter of 17 January 1881 was written "shortly after he utilised Joseph Swan''s newly invented incandescent light bulbs"; it does not itself say Swan lamps were lit in 1880. The sentence therefore says more than the claims show, on the contested Swan side, and the page would tell readers the standard is even when the dig''s own conformance warning shows it is not.'
    evidence: 'build/subjects/incandescent-lamp/claims.yaml; https://ietarchivesblog.org/2023/03/09/cragside-harnessing-the-power-of-water/ (read); https://gsarchive.net/carte/savoy/electric.html (read); conformance warning for lamp-swan-in-use.'
    asked: 'Either move the Cragside half to its own proposed claim (the Savoy half can stay), or word the assessment so that it names this exception ("applied, except that the Cragside date rests on a blog"); say "Lodygin and Woodward and Evans (and a Latimer misattribution)" rather than listing Latimer as a national story; and say in the Cragside node and claim that "1880" and "Swan" come from two passages of the blog (45 incandescent lamps in 1880; Swan bulbs in the letter of January 1881), not one.'
    author_response: {action: fixed, note: 'Verified against the IET blog. Latimer removed from the national-story list; Cragside moved to a new proposed claim lamp-swan-cragside (low), with the blog''s two passages stated; lamp-swan-in-use is now the Savoy only; assessment text and description name the exception; Cragside node rev 2. L-21.'}
    reviewer_note: 'Verified by the third reviewer. IET blog read: ''the arc lamps were replaced by Armstrong in 1880 with 45 incandescent lamps'' and, in another paragraph, the letter of 17 January 1881 ''shortly after he utilised Joseph Swan''s newly invented incandescent light bulbs''; lamp-swan-cragside (proposed, low, secondary) states the two passages and that the letter was not read; its node (rev 2) says the same; lamp-swan-in-use is the Savoy only (Times transcription at gsarchive.net read: ''for the first time lit up by the electric light'', ''The Swan incandescent lamps were used'', auditorium lit since the opening). The assessment text lists Lodygin and Woodward and Evans as the national stories and a Latimer misattribution, and names the Cragside exception. Wording residue is non-blocking (R4-4).'
    resolved: true
  - id: R3-3
    target: 'claims.yaml lamp-lodygin-1872-74 (statement), node lamp-1873-lodygin-carbon-rod-lamps'
    severity: non_blocking
    finding: 'The statement says Kommersant gives 1872 for Lodygin''s first demonstration "in St Petersburg". The page says "В 1872 году Лодыгин в первый раз демонстрирует свою лампу" with no place; St Petersburg appears elsewhere only for his return to the city and his experiments (and the Academy report in Fontaine is St Petersburg). The node place note is worded more carefully than the claim.'
    evidence: 'https://www.kommersant.ru/doc/6822124 (read).'
    asked: 'Drop "in St Petersburg" from the claim statement or say the place is the Academy''s.'
    author_response: {action: fixed, note: 'Verified: Kommersant gives no place for the 1872 demonstration. "in St Petersburg" removed from the claim; the Lodygin node (rev 2) says the place is the Academy''s city. L-21.'}
    reviewer_note: 'Third reviewer: ''in St Petersburg'' is gone from the claim and the node, as asked. But the claim now says ''no place is given for it'', and the Kommersant page (doc 6822124, read) says later that Lodygin''s lamps reached the US ''five years after their successful demonstration in Petersburg'' (''после их успешной демонстрации в Петербурге''). The 1872 sentence itself gives no place. Non-blocking; see R4-6.'
    reviewer_note_round5: 'Fourth reviewer: Kommersant doc 6822124 re-read. The sentence "В 1872 году Лодыгин в первый раз демонстрирует свою лампу" names no place; elsewhere the page says the lamps reached the United States "спустя пять лет после их успешной демонстрации в Петербурге". The claim now says exactly that ("the sentence that gives 1872 names no place; elsewhere the page speaks of a demonstration in Petersburg") and the node (rev 3) says the same. Resolved.'
    resolved: true
  - id: R3-4
    target: 'threads.yaml overlay_notes (Britain)'
    severity: non_blocking
    finding: 'The Britain note says the drift is to "Edison copied Swan", which no record read supports. No claim, source or transmission chain in the dig records that story as made by anyone; it is the dig''s own construction of a drift and carries no source id for it.'
    evidence: 'build/subjects/incandescent-lamp/threads.yaml, claims.yaml, transmission.yaml (no matching entry).'
    asked: 'Remove the drift sentence or anchor it to a source that shows the story.'
    author_response: {action: fixed, note: 'Agreed: no source or chain shows an "Edison copied Swan" story; the drift sentence is removed from the Britain note. L-21.'}
    reviewer_note: 'Verified by the third reviewer: threads.yaml overlay_notes (Britain) no longer contains an ''Edison copied Swan'' drift; it says the dig records no popular story that has drifted from the Swan records and files none.'
    resolved: true
  - id: R3-5
    target: 'claims.yaml lamp-edison-patent (statement, optional)'
    severity: non_blocking
    finding: 'Edison''s US 223,898 has four claims (claim 3 the coiled filament; claim 4 the method of securing the contact wires and carbonising), read in the USPTO scan. The statement names claims 1 and 2 only and is correct; an earlier line of this review (the fresh-pass paragraph of the first review) says "two claims" and is wrong.'
    evidence: 'US 223,898 (USPTO image server, OCR by this reviewer).'
    asked: 'None for the dig. Optional: say "its first two of four claims". The review text is corrected in the notes below.'
    author_response: {action: fixed, note: 'No change needed; as an optional wording, the claim now says "first two claims (of four)". L-21.'}
    reviewer_note: 'Verified by the third reviewer: the claim reads ''first two claims (of four)''; US 223,898 (USPTO PDF, OCR by this reviewer) has claims 1 to 4 (1 electric lamp for giving light by incandescence of a carbon conductor of high resistance; 2 carbon filaments with an all-glass receiver; 3 the coiled filament; 4 the method of securing the leading wires), patented Jan. 27, 1880.'
    resolved: true
- round: 4
  reviewed_on: 2026-10-02
  reviewer: 'independent review agent (third reviewer); did not write the dig and took no part in rounds 1 to 3'
  requests:
  - id: R4-1
    target: 'claims.yaml search_summary, assessment.key_points[1], lamp-edison-sole-inventor (statement), lamp-edison-high-resistance (statement)'
    severity: blocking
    finding: 'Three surface lines say US courts "upheld Edison''s 1880 patent for a high-resistance carbon filament in an all-glass bulb" (search_summary; key point 2; lamp-edison-sole-inventor). What the courts held valid and infringed is the second claim, a carbon filament in an all-glass exhausted receiver (lamp-edison-patent-upheld says so correctly). The first claim, the high-resistance filament made as described, was held limited to leading wires secured by carbonised cement and not infringed, and the complainant did not appeal that (52 F. 300). On appeal Lacombe J. refused to read high resistance into the second claim: "the further limitations, which defendant seeks to read into this claim, viz., that the filament must be one of high specific resistance, or of at least 100 ohms resistance, cannot be accepted"; the draughtsman "omit[ted] the phrases ''of high resistance'' and ''made as described''". High resistance is in Wallace J.''s account of what Edison accomplished (47 F. 454: "to unite the characteristics of high resistance, small radiating surface, and durability in a carbon conductor") but the appeal decision that stands did not make it a limit of the upheld claim. A reader of key point 2 is told the courts upheld a high-resistance filament patent; the courts upheld claim 2, and the higher court said high resistance is not needed for it. lamp-edison-high-resistance (established, high) says high resistance is what "set Edison''s lamp apart, as his patent and the courts described it" without this limit.'
    evidence: '47 F. 454 and 52 F. 300 (static.case.law JSON, read by this reviewer); US 223,898 (USPTO PDF, OCR by this reviewer: claim 1 "a carbon conductor of high resistance", claim 2 carbon filaments and an all-glass receiver, claims 3 and 4).'
    asked: 'Say in the three surface lines that the courts upheld the patent''s second claim (a carbon filament in an all-glass exhausted bulb), or name Wallace J. for the high-resistance account; add to lamp-edison-high-resistance that claim 1 (high resistance) was held not infringed and that 52 F. 300 refused to import a high-resistance limit into claim 2. New log entry; do not edit L-21.'
    author_response: {action: fixed, note: 'Verified in 52 F. 300 and 47 F. 454 (claim 1 limited and not infringed, not appealed; Lacombe J. refused a high-resistance limit on claim 2; Wallace J. credits high resistance). Search summary, assessment text, key point 2, lamp-edison-sole-inventor and lamp-edison-high-resistance reworded; would_change_if widened. L-22.'}
    reviewer_note: 'Fourth reviewer, verified personally on static.case.law. 52 F. 300 (2d Cir., 4 Oct 1892, Lacombe J.): the four claims are quoted (claim 1 "a filament of carbon of high resistance, made as described"; claim 2 "carbon filaments with a receiver made entirely of glass ... from which receiver the air is exhausted"; claims 3 and 4); "The circuit court held that the first claim was by its phraseology limited to lamps in which ... the leading-in wires are secured ... by cement carbonized in situ ... This construction of the first claim has been acquiesced in by the complainant, which has not appealed"; "the further limitations, which defendant seeks to read into this claim, viz., that the filament must be one of high specific resistance, or of at least 100 ohms resistance, cannot be accepted"; the draughtsman omitted "of high resistance" and "made as described"; "The decree of the circuit court is therefore affirmed". 47 F. 454 (Wallace J.): "What he actually accomplished was to unite the characteristics of high resistance, small radiating surface, and durability in a carbon conductor"; "he made a lamp which was practically operative and successful"; claim 1 not infringed because the leading wires are not secured by cement carbonized in situ. 54 F. 678: claim 2 "covered a broad and fundamental invention, namely, an incandescent lamp, composed of a carbon filament, hermetically sealed in an all glass chamber exhausted to a practically perfect vacuum". Read in the revised dig: search_summary, assessment text, key point 2, lamp-edison-sole-inventor and lamp-edison-high-resistance now say the courts upheld claim 2 (a carbon filament in an all-glass exhausted bulb), name Wallace J. for the high-resistance account, and record that claim 1 was held not infringed and that 52 F. 300 refused a high-resistance limit; would_change_if is widened; lamp-edison-patent-upheld already said claim 1 was not infringed. I did not OCR the USPTO scan of US 223,898 myself; the claim wording rests on the Court''s verbatim quotation of the four claims in 52 F. 300. Resolved.'
    resolved: true
  - id: R4-2
    target: 'claims.yaml headline'
    severity: non_blocking
    finding: '"No one person invented the bulb" is a universal that the refuted claim (lamp-edison-sole-inventor: Edison as first or only inventor) does not carry; it also denies Swan sole credit although lamp-swan-before-edison and lamp-first-practical-priority are contested and the assessment answer is "partly Edison". The second clause ("Edison v Swan is unsettled") is accurate.'
    asked: 'Consider "Edison did not invent the bulb alone: carbon lamps were patented by 1845; Edison v Swan is unsettled" (95 characters or fewer) or leave the headline draft with an owner decision.'
    author_response: {action: fixed, note: 'Agreed; headline now "Edison did not invent the bulb alone: carbon lamps were patented by 1845; Edison v Swan is open" (95 characters, draft). L-22.'}
    reviewer_note: 'Fourth reviewer: headline counted by script, 95 characters ("Edison did not invent the bulb alone: carbon lamps were patented by 1845; Edison v Swan is open"); headline_status draft; headline_claim lamp-edison-sole-inventor is refuted at high with a primary check. "Not alone" matches the refuted target (first or only inventor); "carbon lamps were patented by 1845" is carried by the King abridgment (re-read, archive.org patentsforinvent1766grea pp. 97-98: "continuous metallic and carbon conductors, intensely heated by the passage of a current of electricity, to the purposes of illumination"; carbon "enclosed in a Torricellian vacuum") and by Colt J. in 54 F. 678 ("Starr suggested in the King patent a lamp composed of a thin pencil of carbon, inclosed in a Torricellian vacuum"), as a design (S3). "Edison v Swan is open" matches lamp-first-practical-priority (contested, low). Resolved.'
    resolved: true
  - id: R4-3
    target: 'claims.yaml lamp-earliest-patent (statement 2)'
    severity: non_blocking
    finding: 'The second absence says nothing earlier than King''s 1845 patent for a heated conductor in a vacuum has been found, but 159 U.S. 465 (read) calls the 1841 de Moleyns patent one "for an incandescent electric lamp" with pulverized charcoal, which, if the Court''s description were right, would be a heated solid in an exhausted globe in 1841. The dig keeps de Moleyns separate (the abridgment does not say heated), which is right, but the claim should say that the Court''s reading is the one place the second absence could fail.'
    asked: 'Add one clause to lamp-earliest-patent statement or would_change_if: the full specification of 9053 (not read) could show a heated charcoal conductor in an exhausted globe, which would move absence 2 to 1841.'
    author_response: {action: fixed, note: Verified (159 U.S. 465; full specification unread). lamp-earliest-patent and its would_change_if say absence (2) could move to 1841. L-22.}
    reviewer_note: 'Fourth reviewer: lamp-earliest-patent statement and would_change_if now say absence (2) could fail at 1841 because 159 U.S. 465 (tile.loc.gov scan, pdftotext, re-read by me) says "in the English patent of 1841 to De Moleyns, ''finely pulverized boxwood charcoal or plumbago'' was used for an incandescent electric lamp" and the full specification of No. 9053 was not read. Abridgment pp. 62-63 (re-read): "A means of generating the electric light, by constantly supplying the electrodes at their junction, with pulverized charcoal ... The whole arrangement is enclosed in an exhausted glass globe"; no "heated" or "incandescent". I also searched the 1859 abridgments'' text for lighting entries up to 1845: none before de Moleyns for an electric light in a globe; the 1845 King entry is the next. Resolved.'
    resolved: true
  - id: R4-4
    target: 'claims.yaml assessment.text ("the one exception is named"); description (last sentence)'
    severity: non_blocking
    finding: 'The text says one stated standard is applied to every claimant and "the one exception is named: the Cragside date". Two other established claims (lamp-swan-in-use, the Savoy, on a transcription whose original page was not compared; lamp-edison-oct-1879-test, on the Edison Papers transcription without the notebook image) are also below the "read directly" bar of S1 and shown only as conformance warnings and in the description. Whether a transcription of a dated document counts as reading it is a judgment the dig makes in favour of these two and against Fontaine-quoted material (de Changy, Lodygin), which is a different case (a later book quoting a report) but is not explained.'
    asked: 'Replace "the one exception" with "the exceptions are named" and add one sentence to the description saying why a transcription of a dated document (Times, Batchelor''s notebook) is treated differently from a later book quoting a report (Fontaine).'
    author_response: {action: fixed, note: 'Agreed: "the exceptions are named", and the description explains the transcription versus Fontaine distinction as a stated judgment. L-22.'}
    reviewer_note: 'Fourth reviewer: assessment.text now says "the exceptions are named" and lists the Cragside date (its own proposed claim lamp-swan-cragside), the Savoy report and the 22 Oct 1879 notebook entry; the description gives the stated judgment (a transcription of a dated document versus Fontaine''s quotation of an unseen report) and says both transcription-based claims stay at secondary and moderate. The Times transcription (gsarchive.net) re-read: "when the stage was for the first time lit up by the electric light, which has been used in the auditorium ever since the opening"; "The Swan incandescent lamps were used"; it does not say the auditorium lamps were Swan''s, and the claim does not either. Resolved.'
    resolved: true
  - id: R4-5
    target: 'claims.yaml assessment.headline; assessment.key_points[0]; tests.yaml dt-which-first; threads.yaml thread-edison-legend description'
    severity: non_blocking
    finding: '(a) assessment.headline says "patents for electric lamps go back to 1841"; the abridgment says "electric light"; "lamp" and "incandescent" are the Supreme Court''s words (R2-1 in spirit). Key point 1 ends "Edison was not first to patent one" without saying which of the two. (b) dt-which-first says "five different firsts" while the assessment says "at least four"; its row 1 gives Davy''s arc for "heating a solid", which lamp-davy-arc says does not bear on the incandescent lamp; row 2 answers "a patent on an incandescent lamp" with a patent whose abridgment does not say incandescent; the finding sentence reads "row (3) is the earliest found ..." before row (3) is given; and row 5 gives Edison without any search for an earlier wired supply (nothing in the dig searched for one). (c) thread-edison-legend says "the rest are settled in claims.yaml" although the 10,000-ways claim is provisional and the 40-hour lamp is only "not found in any source opened".'
    asked: 'Reword (a) to "an electric light" and name both patents in key point 1; align the count in dt-which-first and say row 5 is the only central supply examined; replace "settled" with "filed".'
    author_response: {action: fixed, note: '(a) fixed: "an electric light", both patents named; (b) fixed: dt-which-first count, Davy, row 2, row order and row 4 caveat; (c) fixed: "filed". L-22.'}
    reviewer_note: 'Fourth reviewer: (a) and (c) verified: assessment.headline says "an electric light"; key point 1 names de Moleyns (1841, electric light in an exhausted globe) and King (1845, heated carbon conductor); thread-edison-legend says "filed" and describes the 10,000-ways claim as provisional and the 40-hour figure as "not found in any source opened". (b) is only partly done: Davy is out of row 1, row 2 carries the abridgment caveat, the rows are ordered and row 4 (central supply) says nothing earlier was searched for; but the predicts block still says "on all five" although four rows are listed, finding (3) ends "so row 4 is contested" (the contested row is 3, as the resolves block says), and the note says the test does not decide Swan versus Edison "on 4". Left unresolved as non-blocking: see R5-2.'
    resolved: false
  - id: R4-6
    target: 'claims.yaml lamp-lodygin-1872-74 (statement and basis); node lamp-1873-lodygin-carbon-rod-lamps note (R3-3 follow-up)'
    severity: non_blocking
    finding: 'The statement now says Kommersant gives no place for the 1872 first demonstration. The sentence that gives 1872 has none, but the same page says Lodygin''s lamps reached the United States "five years after their successful demonstration in Petersburg" ("спустя пять лет после их успешной демонстрации в Петербурге"). The place is therefore stated by the page, for a demonstration it does not date.'
    asked: 'Say "the sentence that gives 1872 names no place; elsewhere the page speaks of a demonstration in Petersburg" or drop the parenthesis.'
    author_response: {action: fixed, note: Verified the Kommersant wording; claim and node (rev 3) now say where Petersburg is and is not named. L-22.}
    reviewer_note: 'Fourth reviewer: Kommersant doc 6822124 re-read (see R3-3). The claim statement, its basis and node lamp-1873-lodygin-carbon-rod-lamps (rev 3) say the sentence that gives 1872 names no place and that the page elsewhere speaks of a demonstration "in Petersburg" which it does not date. The page also says the first samples lasted "30-40 minutes" ("работали первые образцы всего 30-40 минут"), as the node says. Resolved.'
    resolved: true
  - id: R4-7
    target: 'claims.yaml lamp-first-practical-priority (positions, ground), assessment key point 3'
    severity: non_blocking
    finding: 'The Swan side is dated against Pearl Street (a central station, 4 Sep 1882), so "both Swan dates come before Pearl Street" and the position "Swan''s lamps in use first". The dig''s own lamp-menlo-park-dec-1879 records Edison lamps lit and seen by visitors at Menlo Park from 30 Dec 1879 ("sixty lights per the Herald"), which is before Cragside (1880 to Jan 1881) and the Savoy; it is a laboratory, not a customer, and the claim is about practical use, but the position ground and key point 3 compare Swan''s installations with Edison''s system only. A reader of key point 3 is not told that Edison''s lamps were in public view two years before Pearl Street.'
    asked: 'Add Menlo Park (30 Dec 1879, laboratory, contested as an exhibition) to the Edison position ground, or say in key point 3 that the comparison is installations for users against Edison''s first district service.'
    author_response: {action: fixed, note: 'Agreed; key point 3 and the Edison position mention Menlo Park (30 Dec 1879, laboratory, contested). L-22.'}
    reviewer_note: 'Fourth reviewer: key point 3 and the Edison position of lamp-first-practical-priority now mention Menlo Park (30 Dec 1879, a laboratory, contested as an exhibition) beside Pearl Street, as asked. The New York newspaper pages (loc.gov) returned a Cloudflare 403 to me, so I did not re-read the Herald and Tribune; the Menlo Park wording rests on the third reviewer''s reading and the node text. One wording point remains: see R5-1. Resolved.'
    resolved: true
- round: 5
  reviewed_on: 2026-10-02
  reviewer: 'independent review agent (fourth reviewer); did not write the dig and took no part in rounds 1 to 4'
  requests:
  - id: R5-1
    target: 'claims.yaml assessment.key_points[2]; lamp-first-practical-priority (Edison position ground)'
    severity: non_blocking
    finding: 'Key point 3 says "Edison lamps were lit at Menlo Park from 30 Dec 1879" and the position ground says "lit and seen by visitors from 30 Dec 1879". The Herald of 31 Dec 1879 is dated Menlo Park, 30 Dec, and says the place was "thronged with visitors"; its title in the node is "The Horseshoes Still Shining", and the dig itself dates the 22 Oct 1879 test and the 16 Nov 1879 lamp earlier. "From 30 Dec" reads as the day lighting began. The newspapers show visitors and lights on 30 and 31 Dec, not a first lighting.'
    evidence: 'build/nodes/lamp-1879-12-31-menlo-park-visitors-no-formal-exhibition.yaml; lamp-edison-oct-1879-test; Edison Papers Doc. 1831. (The loc.gov pages were not reachable to this reviewer.)'
    asked: 'Say "seen by visitors at Menlo Park on 30-31 Dec 1879" (or "by 30 Dec").'
    author_response_round5: {action: fixed, note: 'Verified (Herald 31 Dec 1879, dateline Menlo Park 30 Dec, "thronged with visitors"; no first lighting reported). Key point 3 and the position ground say "seen by visitors at Menlo Park on 30-31 Dec 1879". L-23.'}
  - id: R5-2
    target: 'tests.yaml dt-which-first (predicts, result.finding, note) (continuation of R4-5 b)'
    severity: non_blocking
    finding: 'The question lists four rows and says "at least four ... and a system", but predicts still says "the same name should come first on all five"; finding (3) ends "so row 4 is contested" although row 3 (a practical lamp) is the contested one; the note says the test "does not decide the Swan versus Edison question on 4". Row 4 (central supply) is Edison, the only supply examined.'
    evidence: 'build/subjects/incandescent-lamp/tests.yaml (read in full).'
    asked: 'Use "all four" and "row 3" consistently, or say five if a fifth row is meant.'
    author_response_round5: {action: fixed, note: Verified in tests.yaml; "all four" and "row 3" now consistent. L-23.}
  - id: R5-3
    target: 'claims.yaml assessment.text ("On the fourth, Edison''s company is the one documented"); assessment.basis'
    severity: non_blocking
    finding: 'The assessment says Edison''s company is "the one documented" for a lamp run from a central supply; dt-which-first says nothing in the dig searched for an earlier supply. A reader of the assessment alone is not told that. Also, key point 3 rests on lamp-swan-in-use, lamp-swan-cragside and lamp-menlo-park-dec-1879, none of which is in assessment.basis (lamp-first-practical-priority, which cites them, is).'
    evidence: 'claims.yaml; tests.yaml dt-which-first row 4.'
    asked: 'Add "no earlier central supply was searched for" to the assessment text and list the three claims in the basis.'
    author_response_round5: {action: fixed, note: 'Agreed; assessment text says nothing earlier was searched for, and the basis lists lamp-swan-in-use, lamp-swan-cragside and lamp-menlo-park-dec-1879. L-23.'}
  - id: R5-4
    target: 'timeline.yaml people (Lewis Latimer); threads.yaml thread-national-priority overlay_notes (Russia)'
    severity: non_blocking
    finding: 'The Latimer entry says "Carbon-making patent 252,386 (1882) and, with J. V. Nichols, a lamp-mounting patent (1881), the first assigned to the United States Electric Lighting Company". The 1881 patent (US 247,097, OCR of the USPTO scan) has no assignment line (Nichols of Brooklyn and Latimer of New York); only 252,386 says "assignor to the United States Electric Lighting Company". "The first" is ambiguous and can be read as the 1881 patent. The Russia overlay note says "a patent dated 11 July 1874 per Kommersant"; the page says "privilege" (привилегия) No. 1619 with a backdated priority, as the claim says.'
    evidence: 'build/subjects/incandescent-lamp/timeline.yaml; US 247,097 and 252,386 (OCR in the first reviewers'' notes); Kommersant doc 6822124.'
    asked: 'Say "the 1882 patent was assigned to ..."; say "privilege" in the Russia note.'
    author_response_round5: {action: fixed, note: 'Verified in the OCR of US 247,097 (no assignment line) and US 252,386; timeline Latimer entry and Russia note ("privilege") fixed. The Greener and Staite (1846) and Roberts (1852) lead is logged in L-23, no claim change.'}
open_items:
- 'FOURTH REVIEWER, FINAL STATUS (passed_with_open_items): every blocking request in rounds 1 to 4 (R1-1 to R1-5, R2-1 to R2-4, R3-1, R3-2, R4-1) is resolved by a reviewer''s own reading, and I re-verified R4-1 myself in 52 F. 300, 47 F. 454 and 54 F. 678. The earlier open_items below that say the status stays revisions_requested were true when written (round 4) and are superseded by this one. Non-blocking leftovers: R4-5 (b) residue and R5-1 to R5-4 above; they should be shown to the author as follow-ups and do not change what a reader is told about who invented the lamp. The subject is NOT added to build/site.yaml by this review.'
- 'Fourth reviewer, read personally: 52 F. 300, 47 F. 454, 54 F. 678 (static.case.law JSON); 159 U.S. 465 (tile.loc.gov, pdftotext); abridgments pp. 62-63 and 97-98 and a text search of the 1859 abridgments for lighting entries to 1845 (archive.org); Kommersant 6822124 (re-read, 1619, 2 Oct 1872 backdated priority, Petersburg wording, 30-40 minutes); the IET Cragside blog (45 incandescent lamps in 1880; Swan bulbs and 17 Jan 1881 in a separate passage); The Times transcription at gsarchive.net (needs a browser user agent); Chemical News 42 pp. 227-230 (Stearn Oct 1877; "a mere hair"; "no blackening"; "two years ago in action"; "as much better than carbonised cardboard"; Swan reports Upton said Edison''s use of carbon in such a vacuum was "entirely new"); the Google Patents text of US 575,002 (citizen of Russia, Pittsburg; filed 4 Jan 1893; dated 12 Jan 1897; molybdenum and tungsten).'
- 'Fourth reviewer, NOT reached or not re-read (listed so nobody assumes they were checked again): the Library of Congress newspaper pages (NY Tribune 5 Sep 1882, NY Herald 21 and 31 Dec 1879, NY Tribune 1 Jan 1880) returned a Cloudflare 403 to me, so lamp-pearl-street and lamp-menlo-park-dec-1879 rest on the second and third reviewers'' reading of the OCR; the USPTO scan of US 223,898 (claim wording taken from the Court''s quotation in 52 F. 300); Fontaine pp. 169-173 and 186-187 (not re-opened by me); Chemical News 39; the Edison Papers essay and Doc. 1831 (not re-opened by me); Davy 1812; Dyer and Martin pp. 615-616; Quote Investigator; US 181,613, 252,386, 247,097, 266,358, 47,632 (OCR files of earlier reviewers only); Electrical World 14 Apr 1900; Scientific American 12 Sep 1874; Howell and Schroeder; Schroeder; 40 F. 21, 56 F. 496, 57 F. 616, 61 F. 834 (not re-read by me); the Canadian Encyclopedia, Library and Archives Canada, Espacenet, the Edison Papers notebook image. The full specifications of British patents 9053 and 10,919, the Comptes rendus of 1858, the St Petersburg Academy report of 1874, Armstrong''s letter of 17 January 1881 and the original Times page were not read by anyone.'
- 'Lead for a later round (not a request): 159 U.S. 465 also says that an English patent of 1846 to Greener and Staite was "for an incandescent light" and used "charcoal, reduced to a state of powder"; 54 F. 678 names Roberts (1852). Neither is in the dig. They fall after 1845 and do not change the 1841 and 1845 findings, but a later dig on prior art should read them.'
- 'Owner decisions (carried, all three still open): (1) publish with headline_status draft (the headline names an open question, and its second half rests on contested claims, not on the headline claim); (2) the 10,000-ways claim stays refuted at provisional on one anchor plus absence, and the narrowed Lodygin and Canada refutation targets are acceptable (R2-10); (3) the Lodygin privilege number 1619 and the backdated priority date 2 Oct 1872 are shown although they rest on one newspaper history with no scan (Kommersant says both; its own text also says the Russian patent came "only a year" after the 1872 petition, which does not fit a grant in July 1874; the dig shows the number and date as that newspaper''s statement only).'
- 'Round 4 (third reviewer): R1-4, R1-5, R2-1, R2-2, R3-1, R3-2, R3-4 and R3-5 are resolved by this reviewer''s own reading; one new blocking request is open (R4-1, what the courts upheld) and six non-blocking (R4-2 to R4-7, plus R3-3). Status stays revisions_requested. The subject must not be added to build/site.yaml.'
- 'Third reviewer, not reached: the Library of Congress newspaper pages (NY Tribune 5 Sep 1882, NY Herald 21 and 31 Dec 1879, NY Tribune 1 Jan 1880) returned HTTP 403 to this reviewer; lamp-pearl-street and lamp-menlo-park-dec-1879 rest on the second reviewer''s reading of the OCR text. Not reached or not re-read by this reviewer: Google Patents (bot check), Quote Investigator, the Canadian Encyclopedia, the Edison Papers notebook image, Scientific American 12 Sep 1874, Electrical World 14 Apr 1900, US 181,613, 252,386, 247,097, 266,358 and 575,002 (rest on earlier reviewers), Howell and Schroeder, Schroeder, Dyer and Martin (re-read in earlier rounds only), Davy 1812.'
- 'Third reviewer, read personally: abridgments pp. 62-63 and 97-98 (archive.org); Fontaine pp. 169-173 and 186-187; 159 U.S. 465 (tile.loc.gov); 40 F. 21, 47 F. 454, 52 F. 300, 54 F. 678 (static.case.law); US 223,898 (USPTO PDF, OCR); Chemical News 39 and 42 (archive.org); Edison Papers essay and Doc. 1831; IET blog; The Times transcription at gsarchive.net (needs a browser user agent); Kommersant 6822124 and 5338615.'
- 'Reached by this reviewer, correcting the first reviewer: the Library of Congress newspaper pages are reachable through the loc.gov JSON endpoint (fo=json, then the fulltext_service text). Read: NY Tribune 5 Sep 1882 p. 8 ("The company at 3 p.m. turned on the electric light in part of the first of the districts ... The central station ... is at No. 237 Pearl-st., just below Fulton-st."; "yesterday" dates it to 4 Sep; "There are no lamps lighted in Pearl-st."), NY Herald 21 Dec 1879 p. 5 ("first public exhibition ... announced to take place on New Year''s Eve at Menlo Park"; "mercury pump"), NY Herald 31 Dec 1879 p. 3 (dateline Menlo Park, 30 Dec; "has been thronged with visitors"; "fixed no day for a general inspection"; "the public exhibition may be said to be actually going on now"; "the total number of lights being sixty"), NY Tribune 1 Jan 1880 p. 8 (headline "NO PUBLIC EXHIBITION PROMISED OR GIVEN", "The unauthorized announcement", "disclaimed any intention of giving any formal exhibition"). lamp-pearl-street and lamp-menlo-park-dec-1879 are verified in OCR text (not in the page images).'
- 'Not reached by this reviewer: The Canadian Encyclopedia, Library and Archives Canada, the Edison Papers notebook image (N-79-07-31:105; the transcription Doc. 1831 was read), Espacenet, Google Patents (blocked); Quote Investigator was read only through a page-summary fetch (the raw page is behind a bot check); the OCR of the text pages of US 266,358 and US 575,002 did not complete (their headers, titles and grant dates, Patented Oct. 24, 1882 and Jan. 12, 1897, were read; the filing dates, the "citizen of Russia, residing in the city of Pittsburg" wording and the molybdenum and tungsten claims rest on the first reviewer).'
- 'Not read by anyone (author or reviewers): the full specifications of British patents 9053 (1841) and 10,919 (1845) (abridgments, 40 F. 21 and 54 F. 678 only); the Comptes rendus of 1858 (Jobard) and the St Petersburg Academy report of 1874 (Fontaine is the only channel); Armstrong''s letter of 17 January 1881; the original Times page of 29 December 1881 (transcription only); the Canadian specification of patent 3,738; Swan''s patents and the Lit and Phil minutes of 1879; Davy''s 1802 papers; Rohde (2007); the springe.de wording; Hammer (1900); Sprengel''s 1865 paper; the Edison patent application file and the Patent Office interference record; the Linda Hall and Newcastle University pages on 3 Feb 1879.'
- 'Anchors below primary behind established claims (conformance warnings): lamp-swan-in-use (a blog and a transcription; see R3-2) and lamp-edison-oct-1879-test (the Edison Papers transcription, Doc. 1831, read by this reviewer: "On from 1.30 AM till 3 pm 13 1/2 hour and was then raised to 3 gas jets for 1 hour then cracked glass & busted"; the notebook image not opened). Both must be shown on the page as limits.'
- 'Abridgment-anchored claims (de Moleyns, King, and the refutations of Lodygin and of Edison as sole inventor) have anchor_checked primary although the full specifications were not read; King is also quoted by 40 F. 21 and 54 F. 678 (read). A provisional-confidence absence claim (lamp-earliest-patent) carries "established" as its state; accepted, as before.'
- 'Owner decisions: (1) whether to publish with headline_status draft (the headline names an open question; its second half rests on contested claims); (2) whether the 10,000-ways claim may stay refuted at provisional on absence, and whether the narrowed Lodygin and Canada refutation targets are acceptable (R2-10); (3) whether the Lodygin privilege number 1619 and the backdated priority date, resting on one newspaper history with no scan, are shown (restored by the author; verified in the Kommersant page). The review does not add the subject to build/site.yaml.'
needs_manual_check:
- 'Matched by hand against the sources the reviewer opened: lamp-de-moleyns-patent (abridgment pp. 62-63, 159 U.S. 465); lamp-earliest-patent (Fontaine p. 169); lamp-starr-king-patent (abridgment p. 97-98, Colt J.); lamp-lodygin-refractory-metals ("a citizen of Russia, residing in the city of Pittsburg", filed 4 Jan 1893, granted 12 Jan 1897); lamp-woodward-evans-patent ("that will not unite chemically with the carbon when hot"; Scientific American 12 Sep 1874 list, OCR "3,:38 ... H. Weodward and M. Evans, Toronto, Ont. ... Woodwara and Evans'' Electric Light. Aug. 3, 1874"); lamp-swan-1878; lamp-swan-early-1879; lamp-first-practical-priority (Wallace J.; Edison Papers essay); lamp-edison-oct-1879-test ("on from 1.30 AM till 3 pm 131/2 hour and was then raised to 3 gas jets for 1 hour then cracked glass & busted", Doc. 1831, N-79-07-31:105).'
- 'Not matched, source not reached: the New York newspaper quotes in lamp-pearl-street and lamp-menlo-park-dec-1879.'
- 'Refuted claims name their target and rest on primary text (Rule 9): lamp-lodygin-first-claim, lamp-canada-invented-claim, lamp-latimer-invented-claim, lamp-10000-ways-quote, lamp-edison-sole-inventor. The 10,000-ways claim also rests on absence (provisional, absence_anchor).'
notes: |
  Independent review, 2026-10-02, of the revised dig (author commits 8495c38 and b2dfecd). The reviewer did not write the dig or review it before and did not edit any dig file.
  Sources opened by the reviewer (host in brackets): US patents 223,898, 181,613, 252,386, 247,097, 266,358 and 47,632 (USPTO image server, OCR by the reviewer) and 575,002 (Google Patents text); Federal Reporter 40, 47, 52, 54, 56, 57 and 61 F. (static.case.law); 159 U.S. 465 (tile.loc.gov scan); the 1859 British abridgments pp. 62-63 and 97-98 (archive.org); Chemical News 39 and 42 (archive.org); Fontaine 1878 (archive.org); Davy 1812 pp. 85-86 and the archive.org record (publisher Bradford and Inskeep); Dyer and Martin 1910 vol. II pp. 615-616 (archive.org); Howell and Schroeder 1927 and Schroeder 1923 (archive.org, passages on de la Rue 1820, de Changy 1856, the 50,000 rouble prize, Grove 1840); Scientific American 12 Sep 1874 and Electrical World and Engineer 14 Apr 1900 (archive.org OCR); Edison Papers essay, Doc. 1831 PDF, Latimer essay and Edison and Swan United page (edison.rutgers.edu); the IET archives blog; The Times transcription (gsarchive.net); Kommersant docs 6822124 and 5338615; CBC 2017, Canada's History and Quote Investigator.
  Round 1 (the earlier passes): R1-1, R1-2, R1-3 and the non-blocking items R1-6 to R1-16 are resolved; R1-4 and R1-5 are not, for the reasons in their reviewer_check lines (R2-1 to R2-3). Because two blocking round 1 requests stay open and four blocking round 2 requests are new, the status is revisions_requested.
  Fresh pass: more than twenty claims were checked against the sources (lamp-davy-arc, lamp-de-la-rue-1840, lamp-de-moleyns-patent, lamp-earliest-patent, lamp-starr-king-patent, lamp-de-changy-belgium, lamp-lodygin-1872-74, lamp-lodygin-first-claim, lamp-lodygin-refractory-metals, lamp-woodward-evans-patent, lamp-woodward-evans-performance, lamp-canada-invented-claim, lamp-swan-1850s, lamp-swan-1878, lamp-swan-early-1879, lamp-swan-before-edison, lamp-swan-in-use, lamp-first-practical-priority, lamp-edison-oct-1879-test, lamp-edison-patent, lamp-edison-high-resistance, lamp-edison-bamboo, lamp-sawyer-man-priority, lamp-edison-patent-upheld, lamp-goebel-claim, lamp-latimer-patent, lamp-latimer-invented-claim, lamp-10000-ways-quote, lamp-edison-sole-inventor). Every established or refuted claim was checked except lamp-pearl-street (the newspaper page was not reachable). Patent numbers and dates agree with the scans (Edison 223,898 signed 1 Nov 1879, patented 27 Jan 1880, two claims; Woodward 181,613 filed 4 Jan 1875, dated 29 Aug 1876, carbon in "a rarefied gas that will not unite chemically with the carbon when hot"; Latimer 252,386 filed 19 Feb 1881, dated 17 Jan 1882, assigned to the United States Electric Lighting Company; 247,097 filed 18 Apr 1881, dated 13 Sep 1881; Goebel 266,358 filed 23 Jan 1882, dated 24 Oct 1882; Lodygin 575,002 filed 4 Jan 1893, dated 12 Jan 1897). The court dates (5 Oct 1889, 4 Oct 1892, 18 Feb 1893, 21 Apr 1893, 20 Jul 1893, 1 May 1894, 11 Nov 1895) match the reporter data; the July 1891 judgment date is from 57 F. 616 and the decree date of 23 July 1891 from 54 F. 678.
  Nodes: all 41 lamp nodes were listed with time, precision, date_basis, window and place. Dates agree with their sources; places are given only where a source names them and carry Nominatim orientation notes; state-level places for the court rulings are the districts the opinions name. The two exceptions are in R2-3 (a node text that misstates its source) and R2-7 (days that no source read gives). The Davy node and the Cragside node are honest about year-level precision.
  Standard across claimants: Goebel (contested, low), de Changy and Lodygin (proposed, single channel through Fontaine), Woodward performance (proposed) and Swan 1850s (gap) are held to S1 to S4 consistently; the Edison side is held to it too (lamp-first-practical-priority is contested at low confidence; the 22 Oct 1879 test stays secondary). The residual unevenness is the assessment wording (R2-2) and the Cragside anchor (R2-6).
  Assessment: the answer partly and the three key points are in line with the claims except for the 1841 wording (R2-1) and the Cragside limit (R2-6). headline_status draft is justified: the headline names an open question and its first half is carried by more than the headline claim.
  Repository text: no personal names beyond the historical figures and the cited authors were found in the dig and its nodes; no model identifiers or contact details; no copyrighted images; quotations are short and attributed.
  Mapping of the log entries the requester cited: L-11 (blocking 1 and 2), L-12 (blocking 3), L-13 and L-14 (blocking 4), L-15 (blocking 5), L-16 (non-blocking), L-17 (dates and places), L-18 (still unchecked). Two statements in L-14 and L-17 are false about their sources (R2-3, R2-5); the log is append-only, so the correction must be a new entry.

    author_response: {action: fixed, note: Verified 'Matthew Evans' in Electrical World 1900; timeline people list now uses Matthew.}
  Second review (round 3), 2026-10-02, by an independent review agent (second reviewer) that did not write the dig and took no part in rounds 1 and 2. No dig file other than this one was edited. Sources opened by this reviewer: Kommersant 6822124 and 5338615; the 1859 abridgments (pp. 62-63, 97-98); Fontaine 1878 (pp. 169-173, 184-187); Chemical News 39 and 42; Federal Reporter 40, 47, 52, 54, 56, 57 and 61 (static.case.law); 159 U.S. 465 (tile.loc.gov, pdftotext); US 223,898, 181,613, 252,386 and 247,097 (USPTO scans, OCR); Dyer and Martin vol. II pp. 615-616; Electrical World 14 Apr 1900; Scientific American 12 Sep 1874 (OCR "Woodwara and Evans' Electric Light. Aug. 3, 1874"); the Edison Papers essay and Doc. 1831; the IET blog; the Times transcription; the loc.gov text of the four newspaper pages; Quote Investigator (summary only).
  Round 1 and 2 requests: R1-1 to R1-3, R1-6 to R1-16 and R2-3 to R2-11 are resolved by this reviewer's own reading. R1-4, R1-5, R2-1 and R2-2 are not, because of R3-1 and R3-2. Because two blocking requests are new and four earlier blocking ones stay open, the status remains revisions_requested.
  Fresh pass (all matched against the sources unless marked): lamp-davy-arc (not re-opened; the first reviewer's reading kept), lamp-de-moleyns-patent, lamp-earliest-patent (R3-1), lamp-starr-king-patent, lamp-de-changy-belgium (Fontaine pp. 186-187: 27 Feb 1858, "hermetically sealed glass tubes", "incandescence of platinum", "lit in the hollow of my hand", "in 1857, M. de Changy, in Belgium, substituted platinum for the carbon"), lamp-lodygin-1872-74 (Fontaine: 1873, "a high prize", "Lomonossow prize", "the first who had the idea"; Kommersant; R3-3), lamp-lodygin-first-claim, lamp-woodward-evans-patent (US 181,613: filed 4 Jan 1875, patented 29 Aug 1876, "rarefied gas that will not unite chemically with the carbon when hot", "becomes luminous"), lamp-woodward-evans-performance, lamp-canada-invented-claim, lamp-swan-1850s, lamp-swan-1878, lamp-swan-early-1879, lamp-swan-before-edison, lamp-swan-in-use (R3-2), lamp-first-practical-priority, lamp-edison-oct-1879-test, lamp-edison-patent (signed 1 Nov 1879, patented 27 Jan 1880, four claims), lamp-edison-high-resistance (47 F. 454 and 52 F. 300 quotes; French patent 130,910 of 28 May 1879; Canadian 10,654 of 17 Nov 1879), lamp-edison-bamboo, lamp-menlo-park-dec-1879, lamp-pearl-street, lamp-sawyer-man-priority (317,076 issued 12 May 1885 per the Supreme Court scan; "too indefinite"; "becomes unnecessary"; "Affirmed"), lamp-edison-patent-upheld, lamp-goebel-claim (all quotations; "some forty affidavits"; Seaman J. agrees with Hallett J. only that the St. Louis evidence lacks "the measure of proof demanded" and rejects his rule, which matches the node label), lamp-latimer-patent (252,386 filed 19 Feb 1881, dated 17 Jan 1882, assigned to the United States Electric Lighting Company; 247,097 filed 18 Apr 1881, dated 13 Sep 1881), lamp-latimer-invented-claim, lamp-10000-ways-quote, lamp-edison-sole-inventor. lamp-lodygin-refractory-metals and the Goebel patents: headers only (see open_items). Every established claim was checked except lamp-davy-arc (not re-opened) and the filing details of lamp-lodygin-refractory-metals.
  Nodes: all 41 lamp nodes were listed with time, precision, date_basis, window and place. Dates, windows and places are consistent with the sources read. Node ids that carry a day the sources do not give (lamp-1879-02-03-swan-shows-lamp-lit-and-phil, and "boston" and "st-louis" in two court ids) cannot be renamed under the append-only rule; their labels and text are honest (the Boston and St. Louis names come from 57 F. 616). The Pearl Street, Menlo Park and Tribune nodes now match the newspaper text.
  Headline: 95 characters, headline_status draft, the headline claim lamp-edison-sole-inventor refuted at high confidence with a primary check; "carbon lamps were patented by 1845" is carried by the King abridgment, 40 F. 21 and 54 F. 678 (a design, per S3). Assessment: answer partly, three key points, no percentages; it says no more than the claims except the two points in R3-2 and the residue in R3-1. Repository text: no model identifiers and no personal names beyond historical figures and cited authors were found in the files read.
  Correction to the first review: it counted Edison's US 223,898 as having two claims; the patent has four. A duplicated author_response line that had been pasted at the end of the earlier notes block was removed (it duplicated R2-11).
  Third review (round 4), 2026-10-02, by an independent review agent (third reviewer) that did not write the dig and took no part in rounds 1 to 3. No dig file other than this one was edited. Round 1 to 3 blocking requests R1-4, R1-5, R2-1, R2-2, R3-1 and R3-2 are now resolved by my own reading (see each reviewer_note_round4 or reviewer_note); R3-4 and R3-5 are resolved; R3-3 stays open as non-blocking (R4-6). Conformance: 0 errors, 2 warnings for this subject (lamp-swan-in-use and lamp-edison-oct-1879-test below primary); review_dig: all mechanical checks pass; the quote check cannot match stored sources (none stored) and lists 17 quotes for manual check, of which I matched by hand the abridgment, Fontaine, Supreme Court, case-law, Chemical News, Edison Papers, IET and Kommersant quotes listed above.
  Fresh pass: every established or refuted claim whose source I could reach was re-read against its anchor (de Moleyns, earliest patent, King/Starr, Lodygin first claim and 1872-74, Swan 1878, early 1879, before-Edison, Savoy, Cragside, first practical, 22 Oct 1879, Edison patent (four claims), high resistance (new finding R4-1), Pearl Street not reached, Sawyer and Man, patent upheld, bamboo, Goebel quotes, 10,000-ways via the earlier readers). Result: one new blocking finding (R4-1), the rest non-blocking (R4-2 to R4-7). Places and dates in the 41 lamp nodes were listed again: consistent with the sources, with the disclosed conventions (patent-grant node placed at the inventor's address; Jobard node placed in Belgium for de Changy's work; Lodygin node placed in the Academy's city; Swan nodes at window mid-points with date_basis inferred). No model identifiers, contact details or private names in the subject files or nodes. Because R4-1 is blocking and has no author response yet, the status stays revisions_requested; owner decisions listed in open_items stand.
  Fourth review (round 5), 2026-10-02, by an independent review agent (fourth reviewer) that did not write the dig and took no part in rounds 1 to 4. Only this file was edited. Conformance: 0 errors (15 subjects, 107 shared nodes, 92 warnings repository-wide; 2 for this subject: lamp-swan-in-use and lamp-edison-oct-1879-test, anchored below primary and shown as limits). review_dig: all mechanical checks pass; the quote check lists 39 quotes it cannot match (nothing is stored in the subject folder).
  Open requests: R4-1 (blocking) is resolved by my own reading of 52 F. 300, 47 F. 454 and 54 F. 678; R3-3 and R4-2, R4-3, R4-4, R4-6 and R4-7 are resolved; R4-5 is left unresolved (non-blocking residue, R5-2). All blocking requests of rounds 1 to 4 now carry resolved: true from a reviewer who verified them. Status: passed_with_open_items.
  Fresh pass: re-read against their anchors the claims lamp-de-moleyns-patent, lamp-earliest-patent, lamp-starr-king-patent, lamp-lodygin-1872-74, lamp-lodygin-first-claim, lamp-lodygin-refractory-metals (Google Patents text), lamp-swan-before-edison and lamp-first-practical-priority (Chemical News 42), lamp-swan-in-use (Times transcription), lamp-swan-cragside (IET blog), lamp-edison-patent, lamp-edison-high-resistance, lamp-edison-patent-upheld, lamp-edison-sole-inventor, lamp-edison-bamboo and lamp-sawyer-man-priority (159 U.S. 465: "too indefinite", "becomes unnecessary", "Affirmed", "thirty or forty different woods", "despatched a man to Japan", "about forty different kinds"), lamp-goebel-claim (54 F. 678 outcome "Injunction granted"). Surface lines checked for "first", "invented", "practical", "built", "sold", "upheld" and "settled": headline, search_summary, assessment (headline, text, key points), claim statements, would_change_if lines, tests, threads and timeline notes. Nothing says more than the claims show except the five non-blocking points in R4-5 (b) and R5-1 to R5-4. Counts: 39 claims (17 established, 5 refuted, 4 contested, 4 proposed, 9 searched gaps), 41 lamp nodes, six national stories, four firsts in the assessment, headline 95 characters. Node dates and places listed again for all 41: consistent with the sources and the disclosed conventions; Swan nodes at window mid-points with date_basis inferred. No model identifiers, contact details or private names in the subject files or nodes (grep for model names, e-mail patterns and session ids found none outside this file).
  Not a finding but worth the owner's eye: the dig treats a patent abridgment as primary for what a patent claims (anchor_checked primary) although neither full specification was read; the dig says so in the claims and in open items, and King is also quoted by two court opinions that I read.