S2. Source editor’s annotation A note in the green pocket notebook indicates that, on September 21, Joubert is at Villeneuve.
T38. The source editor’s roman report and present is remain. September 21 is separate from the main September 26 date-only marker that owns the call after septembre and before its period. No main paragraph is invented for that marker, and neither date is transferred to the following blood-and-faith entry.
T78. This complete date and raised 2 call stand alone between centered separators. No main paragraph accompanies the date-only marker. The call follows septembre before its period; neither the marker date nor the apparatus September 21 belongs to the following independent Le sang entry.
T79. This one roman annotation belongs to the date-only September 26 marker. Its distinct September 21 and present est are retained, not transferred to the following main paragraph.
T36. The ceaseless struggle and full exception remain, with little in both alternatives and the second predicate compressed. English adds a comma before and while preserving both against relations. This entry remains undated: the separate September 26 marker above it has no main paragraph of its own, and neither that date nor the annotation’s September 21 is transferred here.
T80. This paragraph begins after its own separator and is undated. Excepté retains an identifiable initial e in color. The repeated contre , compressed second predicate and complete exception with peu twice remain. All nine owners, the date-only marker and two separate annotations were read in full color, twelve original-size color crops and four native 433 margin bands. Fourteen word-loss slots remain. The explicitly scoped senior date comparison remains distinct from the source-specialist collation of the other named contours; every individual uncertainty remains qualified.
T39. The definition, first-person object and repeated good remain. The final period follows American quotation placement.
T40. The infinitive fragment and three or alternatives remain. Express keeps exprimer ’s wordplay with pressing out juice. The marked source word has italic express as its counterpart; the following English comma remains roman, and no italic extent of the source comma is certified. Plural they follows French ils without forcing a singular correction. Both source ne forms belong to the comparative construction and do not add English negation. After plus , the source preserves only qualified d in d[…] ; contextual de is not restored in French. English more juice naturally has no separate lexical word for that damaged connector, so the intact juice remains unbracketed.
T82. The complete 4, infinitive opening, three ou alternatives, plural ils and both ne forms remain. At the right edge only a qualified d survives after plus ; no complete following e is certified in color/native. In the French source, only exprimer is italic; the following comma is retained without an asserted italic extent. The pressing/expression interpretation does not reconstruct French.
S1. Source editor’s annotation A leaf dated “October 1820” contains this: “Continuation [of the] Essays, volume I (which is the ninth). Page 13, fear — mortifies the [passions]. Love alone would make them flourish...” Quotations follow.
T41. The opening question and answer, three when clauses, repeated fullness and two that-is-to-say explanations remain. [Papacy] contextually interprets papaut[…] without certifying its missing ending. The comma after that bracket is supplied English punctuation; no French stop is recovered at the damaged edge. [Possesses] is a broader contextual rendering of est e[…] possession : est and possession are intact, while the damaged relation e[…] is interpreted through en without restoring a French n . The English bracket covers the combined expression rather than claiming that its intact French words are lost. [It] contextually supplies the council as subject at the whole-word gap before est reconnu ; no particular lost French form is certified. Both c’est à dire forms remain unhyphenated in French, and English adds grammatical commas after that is to say. French retains plain Eglise . Papacy and universal/universality keep the source’s terms without a later doctrinal definition. The source annotation call belongs after octobre in the October 30 date heading, before its period.
T43. The source-editor framing, two bounded roman quotations, month-only October 1820 date and external volume/page locators remain separate from the main October 30 date that owns the call. [Of the] contextually interprets d[…] through des without certifying any missing French letters. [Passions] interprets passio[…] : neither the incomplete following n nor any later letter is certified. The English period after [passions] is supplied to divide the clauses; it is not a recovered French stop. The source uses Arabic 1 with raised er for the first volume and Roman IX with raised e for the ninth; English volume I and the ninth adapt the numeral typography without changing the ordinal values. The unusual first-volume/ninth-volume relation is not repaired or converted into an internal link. Alone and conditional would remain, and the final quoted ellipsis is printed and unfilled. No external Essays edition is authenticated.
T83. The complete 30 and raised 1 after octobre before its separate period remain. At the right edges only papaut and e after est are complete; no ending, punctuation after papaut , or following n is certified. After the final quand , background show-through and detached traces certify no whole short word, so a whole loss remains. These limits agree in original-size color and native 434. The repeated plénitude and c’est à dire , three quand clauses, lower-case c’est after the first damaged ending and plain Eglise remain. Neither source hyphens in c’est à dire nor a final doctrine is supplied. The grammatical gender and contextual referents do not certify the lost French letters.
T84. One roman paragraph retains the two separately bounded quotations, the external month/year, page 13, dash, conditional feroit , seul and printed three-dot ellipsis. The source call belongs to October 30, not to the quoted month-only leaf date. At the first line ending only d is identifiable; its right contour is abraded, with no complete following letters. At the second ending only passio is complete: the next n -like portion is interrupted and does not certify a full n , later s or terminal punctuation in either color/native witness. Both gaps remain explicit. Tome 1ᵉʳ has an angled upper flag and lower base consistent with Arabic 1, followed by raised er ; the later IXᵉ has the distinct Roman I/X and raised e . The first/ninth relation is not repaired or expanded to an authenticated external edition.
T42. The place relation and plural bodies remain. The printed date is December 7, with no visible preceding 2; it is not changed to December 27 from the later source-editor narrative. This final 1820 unit stays above the separate 1821 heading.
T85. The complete printed 7 remains; there is no visible preceding 2. The later source-editor narrative’s December 27 does not repair this main date. This is the last 1820 unit, above the rule and ANNÉE 1821 heading. All four main paragraphs and their separate annotation were read against the full color 902 image, original-size upper/right/app crops and native 434. Four main and two apparatus word-loss slots remain; no missing endings or punctuation are restored. The source-specialist facsimile collation retains all stated contour limits; it does not restore missing endings or punctuation.