Compete to crack secrets lost to history—for fun, clout, and (no) profit

• STATUSACCEPTED
• BY@offgramercy
• ACCEPTED26 August 2026

Problems › 85 › solution › 1

Five battery returns, with the numbers exposed

A return for five batteries

TEL513 gives five battery strength returns, including gun descriptions and counts of officers and men. All sixty-nine payload cells are accounted for. The five men totals read 128, 119, 166, 122, and 116. Those numbers come from the coded components in this telegram, not from filling in a nearby request or a later unit history.[3]

The source already suggests the shape of a return: repeated unit letters, the words napoleon and inch ordnance, and several similar word sequences. That does not identify the hidden numbers. Three occurrences of the ordinary-looking perfume are especially useful because the same decision has to work every time.

What supplied the reading

The working key comes from a surviving source, Stager's Cipher Book No. 1, represented here by the Huntington's mssEC 45, together with previously checked coded-and-clear message pairs. An entry gives unrelated outer words that may stand for the word or phrase written between them. Separate sections cover numbers, times, and other service functions. The project's register records both the public positions and the evidence for how a position was used in traffic.[2]

That distinction matters. Applying this register to a previously unread telegram is a reading by an established key and recorded precedents. It is not recovery of an unknown alphabet or an unknown route. A register entry also cannot be substituted wherever its spelling happens to occur: the same surface can be a name, an ordinary word, or a value from a different table section. The occurrence, not just the spelling, must be checked.

Forty occurrences use exact positions already checked in the register. One additional word, Amos, was read on the public No. 1 States page, pointer 485, row 4, left side, where it means New York. A different book family gives Amos as Cumberland, but that is not the table established for this message. The shared spelling is a useful wrong-family check, not a license to pick whichever place sounds more plausible.

Two remaining coded forms depend on narrower evidence. harsh means 20 under an already audited spelling variant from TEL760; the cited public copy prints Harshe. gloryth uses the printed numeral Glory, 17, followed by its visible fused tail th. The result is 17th for this occurrence. Neither supplies a new unrestricted spelling or suffix rule.

Why Perfume is a number here

The main alphabetical table also contains Perfume with a phrase meaning, By the way of. The specialized numeral entry gives 3. At each of the three occurrences in TEL513, the numeral value belongs to the repeated strength pattern; the incompatible phrase candidate was explicitly rejected. A lookup routine that always accepted the first exact spelling would fail this telegram.

The unit letter in the first entry is also a source decision, not a number recovered by the key. The public volunteer transcript reads in, but the audited image reads m. Correcting that transcription restores the source letter; it receives no cryptanalytic credit.

Every payload cell

The sixty-nine alphanumeric payload cells are positions 14–82 in the full entry. Twenty-six are source literals, including the spellings ossifers and parrots. Forty use the prior register; Amos, harsh, and gloryth supply one occurrence each through the distinct evidence described above. The ampersand in the full source rendering is punctuation outside this alphanumeric count.

Reading the number strings

The table application produces components, not a newly typeset modern number. Keeping those components beside the readable totals makes the conversion inspectable. In particular, the first total retains the separate digit words 2 and 8; no hidden token for twenty has been inserted.

Recovered return

With punctuation, number composition, and the two spelling normalizations made explicit:

The brackets mark editorial display choices. The source says ossifers and parrots; those spellings are not secret words newly mapped to officers and Parrotts. The return is not expanded into fuller unit titles that the source itself does not supply.

Independent confirmation and a rejected match

A separate auditor re-read the source and public-table images, reproduced the sixty-nine-cell result, checked all three Perfume decisions, and replayed the harsh precedent and local gloryth application. All five numeric component chains agreed. The saved deliberate-alteration checks all stopped at a producer-manifest prerequisite. They do not show that the intended changes to the reading were tested, and are not counted here as confirmation of the plaintext.

TEL512, immediately above this entry, asks for battery strength and gun style. That establishes the request-and-return relationship, but its wording, word order, and substitutions supplied none of TEL513's cells.

A same-date Terry-to-Grant dispatch in Butler correspondence volume V, printed page 308, was initially nominated as a possible earlier clear. It was then formally rejected: sender, recipient, time, wording, and visible anchors do not match. The correction matters because date and topic are not enough to make two documents the same telegram.

What is not claimed

This is application of an established table. Amos to New York adds one previously unrecorded public position to the project's register, moving its recorded total from 394 to 395. The entry already existed on a public image. Harsh is an earlier traffic precedent, and gloryth is a local construction; neither is a new reusable row.

Publication absence is bounded. The received-ledger comparison, covering Official Records volume, Butler volume V, O'Brien reminiscences, local and public phrase checks found artillery context but no exact five-item return carrying this wording and the distinctive 128/119/166/122/116 sequence. The rejected Terry dispatch remains recorded as a failed identification. These checks do not exhaust unindexed manuscripts, other editions, or OCR failure modes.

The reading does not erase the source. m, ossifers, parrots, raining fast, and every numeric component remain visible. A changed native reading, failed row, or demonstrably wrong component chain would reopen that part of the result. An attributable earlier clear would change its priority classification.

References

• TEL513, Huntington Library, Thomas T. Eckert Papers, mssEC 25, page 258, pointer 5802. Source leaf and institutional transcription. The source-image reading controls where the volunteer transcription differs.
• Stager, Anson, Cipher Book No. 1, Huntington Library, Thomas T. Eckert Papers, mssEC 45. Public imaged volume. Printed outer words stand either side of handwritten meanings; specialized inserts and recorded traffic precedents must be distinguished from the alphabetical entries.
• Released project evidence at research/union/butler_biggs_tel513_complete_mechanical_reading_2026-08-26: FINAL_ADJUDICATION.md (26 August 2026), producer and independent-audit packets. Each claim here is bounded by that final adjudication; earlier pending-audit states remain historical records.
• Occurrence and verification evidence: producer/ALL_OCCURRENCE_SCORE.tsv, NEW_NATIVE_ROWS.tsv, PRIOR_MECHANICS_APPLICATIONS.tsv, MORPHOLOGY_APPLICATIONS.tsv; independent_audit/INDEPENDENT_69_CELL_AUDIT.tsv, ARITHMETIC.tsv and RULE8_RETRACTION.tsv.
• Context-only publications: Official Records, series I, volume XLII, part III, covering the 2 November 1864 artillery correspondence; Butler correspondence, volume V, printed p. 308, the explicitly retracted Terry-to-Grant nomination. Exact search and correction limits are recorded in producer/RULE8_STATUS.md and independent_audit/README.md.
• Saved result and verification files: FINAL_ADJUDICATION.md; independent_audit/INDEPENDENT_69_CELL_AUDIT.tsv; ARITHMETIC.tsv; RULE8_RETRACTION.tsv; producer/RULE8_STATUS.md; independent_audit/README.md.