Catalogue concordance
| Catalogue | Number | Confidence | How it was checked |
|---|---|---|---|
| Bartsch | 20 | Confirmed | Verified against the catalogue entry for no. 170, which the project now holds in photographic form. |
| New Hollstein | 170 | Confirmed | Verified against the catalogue entry for no. 170, which the project now holds in photographic form. |
| Wikidata | Q65024742 | Confirmed | Wikidata Special:EntityData confirmed P31 print, P170 Rembrandt (Q5598), P571 1638, P2079 etching, P136 self-portrait, P217 1943.3.7201. Identifies the work and the plate, not this impression. |
In the older numbering the plate is simply no. 20 — Bartsch 20, Hollstein 20, White & Boon 20 — and only the New Hollstein renumbers it, as 170. That is why a citation reading B. 20 and one reading NH 170 can describe the same plate without either being wrong, and why a bare number with no catalogue attached is not a citation at all. Catalogue raisonné: A systematic catalogue of all known works by an artist in a given medium, numbering each and describing its states. For Rembrandt's prints the current standard is the New Hollstein volumes; Bartsch's catalogue of 1797 and White & Boon's of 1969 remain in wide use, and they number and divide the works differently.
| Catalogue | States | Evidence | Source |
|---|---|---|---|
| Bartsch, 1797 | 1 | The Rijksmuseum renders the reference as “Bartsch 20-1(1)” — state 1 of 1. | Rijksmuseum |
| Hollstein, original series | 1 | The Rijksmuseum renders the reference as “Hollstein Dutch 20-1(1)”. | Rijksmuseum |
| White & Boon, 1969 | 1 | The National Gallery of Art's catalogue raisonné field reads “White/Boon 1969, no. 20, State only.” New Hollstein confirms the correspondence directly, printing “White & Boon only state” against its own state II. | National Gallery of Art, corroborated by New Hollstein 170, p. 31 |
| New Hollstein, as cited by the National Gallery of Art | 3 | The NGA record's bibliography cites “The New Hollstein … no. 170, state ii/iii”, a three-state reading. The catalogue as published divides the plate into four, and lists the NGA's impression (1943.3.7201, retouched with brush and grey ink) under state II. | National Gallery of Art |
| New Hollstein, Rembrandt (Hinterding & Rutgers) | 4 | Read directly from the catalogue entry for no. 170, pp. 31–32, which sets out four states with their diagnostic features and lists the impressions known in each. | New Hollstein 170, pp. 31–32 |
CaveatThe work in this state is explicitly not Rembrandt's, so an impression in state III is a posthumous rework of his plate. That makes the II-versus-III determination the most consequential open question about NP-RMB-001 — and it turns on the line-work described above, not on the signature.
Counts in the middle column are impressions this project has located and transcribed, not impressions that exist: 15 of the 27 census rows state a state at all, and the catalogue’s own tallies — given inside each state description in the atlas — are larger. The watermark column is summarised at /watermarks, including the point that the catalogue lists none at all for states III and IV.
1679 – 1786
New Hollstein records the plate as possibly in the 1679 inventory of Clement de Jonghe, as “13 conterf.[eytsel] van Rembr.[andt]”; then Pieter de Haan's sale of 1767, lot 2, to Fouquet; then Claude-Henri Watelet's sale of 1786, lot 365. This is the plate's history, not this sheet's — but this sheet came off that plate, so it bounds where it came from.
Source: New Hollstein Dutch & Flemish, Rembrandt, no. 170, pp. 31–32 and plates 84–87
c. 1786 – 1846
Rembrandt's plates passed en bloc to Basan for 500 livres. Pierre-François Basan died in 1797; the plate descended to Henry-Louis Basan, who sold it around 1809 to Auguste Jean, and on Jean's death in 1820 to his widow, whose collection was auctioned in 1846 as lot 540. It is in this period that the plate was rebitten and reworked by a hand that was not Rembrandt's, producing the third state, and later had its signature clumsily redrawn in the Basan workshop, producing the fourth. If this sheet is state III, it was printed inside this window.
Source: New Hollstein Dutch & Flemish, Rembrandt, no. 170, pp. 31–32 and plates 84–87
1846 – 1993
Bought in 1846 by Auguste Bernard, by descent to his son Michel Bernard, who sold it in 1906 to Alvin-Beaumont, who sold it in 1937 to Robert Lee Humber; on Humber's death in 1970 to his heirs, and sold in 1993 through Artemis London with R.M. Light. The catalogue records the plate as last in a private collection in the United States.
Source: New Hollstein Dutch & Flemish, Rembrandt, no. 170, pp. 31–32 and plates 84–87
The custody chain of the single sheet this project holds is a separate timeline, with its own gaps, at /provenance. This project does not appraise, authenticate or value objects, and publishes no figures of that kind.
It identifies a plate. Bartsch 20 and New Hollstein 170 name one piece of worked copper and the composition printed from it. Every impression of that plate carries the number, whether it was pulled in 1638 or in the nineteenth century, whether it is brilliant or exhausted, whether it is in Amsterdam or in a private collection.
It does not identify a sheet. Nothing about the number tells you which impression you are holding, what state it is in, what paper it is on, or when it was printed. Those are properties of the object in front of you, and each one has to be established separately. The census of 27 located impressions exists precisely because the number is shared and the sheets are not.
It does not identify an artist on its own. Bartsch numbers each artist’s prints separately, so “Bartsch 20” names a different print for every artist in the catalogue. “Rembrandt, Bartsch 20” is the citation; “Bartsch 20” alone is an ambiguity.
It does not settle a state. Three of the four catalogues that number this plate recognise no states at all, so for those three there is nothing a state field could say. Where a state is asserted, it is only meaningful with the catalogue named beside it.