About the Evidence

What follows is how to audit this book.

The Archive and the floor

Every factual claim in this Guide traces to the OunceDesk Research Archive — principally the Mint's own reports and abstracts (1803–1971), the coinage statutes, the Eckfeldt–DuBois manual of 1842, the American Journal of Numismatics and The Numismatist runs in which the 1804-dollar literature was actually written, and the great sale catalogues: Woodward's Mickley (1867), Haseltine's Type Table (1881), Chapman's Stickney (1907). All are public-domain publications held in full; quotations from them are verbatim, with original spelling and grading language preserved. The research pack behind this book was owner-reviewed before drafting began, and nothing here was written from memory.

Deliveries are not dates

The single doctrine underlying every number in this book: the Mint's registry records coins delivered per calendar year, not coins bearing a date — the Mint's own printed NOTE says so, and Chapter 5 quotes it in full. Wherever a per-date or per-variety figure appears, it is an estimate or an allocation, and it is labelled with the party who made it. The 85,634 figure the trade carries for 1803 is published here as what the arithmetic shows it to be — the Mint's 66,064 plus the 19,570 delivered in 1804 — an interpretation, not a record.

Modern data, and what it is

Census positions, pedigree chains, survival estimates, certification numbers, and post-1930 auction results are drawn from PCGS CoinFacts, observed 2026-08-23, and are attributed to PCGS wherever they appear — not to the authors PCGS may itself be paraphrasing. Census figures carry their class and their as-of date; a count without both labels is treated as unusable. Where a pedigree link rests on a single source, the text says so or leaves the discrepancy standing — the Mickley lot number, the Sanford sale year, and the bank-counter date are all carried unreconciled, with both readings shown.

The reference works not held

Four standard works remain in copyright and outside the Archive: Bolender (1950), Newman & Bressett (1962), Bowers & Borckardt (1993), and Ferguson (2015). They are named where their systems or conclusions are load-bearing — the B-numbers, the BB-numbers, the Class I/II/III taxonomy, the 1834 diplomatic-set origin — and they are never quoted. Where a claim rests on their authority alone, this book says so in the sentence: the 1834 origin of the Class I dollars is carried as PROBABLE, with the missing primary documents listed as open orders.

Uncertainty is content

The contradictions in this series' printed record are presented, not harmonized: 72,920 against 79,920 for 1796; 423,515 against 429,515 for 1799; May 9, 1843 against Frossard's 1846 for Stickney's acquisition; the moving census of the 1804 dollar itself. The traditions the Archive cannot document — the half-dollar press of 1794, the China shipwreck — are framed as tradition and legend respectively, with their earliest located tellings cited. Where the record is silent, the silence is stated: no hoard, no melt order, no primary document for the diplomatic sets, no ledger line splitting 1794 from 1795. The gaps are listed in the open orders because a reader may be the person who closes one.