Two editions, thirty-four years apart, and a shrinking canon
Abraham Bredius spent most of his career accumulating Rembrandt: buying pictures, cataloguing them, defending attributions, and eventually producing, in 1935, a corpus of 630 paintings he accepted as autograph. Published in Vienna, the Gemäldekatalog — reprinted in English translation as Rembrandt: The Complete Edition of the Paintings — was a monument of the connoisseurial tradition, organised by subject and illustrated throughout.
For the first time, a single scholar had assembled every painting he was willing to call Rembrandt's, photographed them in one place, and numbered them. Those numbers stuck. Scholarship still cites them as "Br." followed by a number, even when arguing that the work in question is not Rembrandt at all.
Bredius was born in 1855 and had been looking at Dutch pictures for more than half a century before the catalogue appeared. He had directed the Mauritshuis in The Hague, done foundational work at what is now the the Netherlands Institute for Art History
, and dug through Dutch archives with a thoroughness few matched. His eye was real, but it operated in an intellectual climate that was essentially expansionist: in the tradition running through Cornelis Hofstede de Groot before him, the working assumption was that Rembrandt's output was large, his workshop activity extensive, and that pictures deserving consideration should stay in rather than come out.

Bredius's 630 was itself a reduction from some earlier lists, but by later standards it remained very large.
Horst Gerson revised the catalogue in 1969, and the revision was not gentle. Working from the same numbered sequence — the famous Br. numbers were preserved for continuity — Gerson rejected a substantial portion of the pictures, in many cases curtly. His apparatus recorded the rejection, noted a brief reason, and moved on.
The effect, reading the 1969 edition against the 1935, is of a systematic culling: pictures that Bredius had reproduced without comment are now flagged as workshop, as school, as copies, or simply dismissed. Gerson's revised edition brought the accepted total down considerably from Bredius's 630, though the precise figure that emerges from his notes is difficult to pin down cleanly, because his rejections vary in firmness.
What Gerson was responding to, in part, was a generation of technical examination that Bredius had not had access to. X-radiography had been applied to Dutch paintings since the 1920s, and by mid-century the results were accumulating fast enough to embarrass purely visual judgement.

The post-war literature, particularly in German and Dutch journals, was already raising doubts about a significant number of accepted works. Gerson's revisions incorporated this mood without fully systematising it: he was still working as a connoisseur, but a connoisseur aware that the old assumptions of abundance were insupportable.
The 1969 edition therefore occupies an odd position in the literature. It was the standard reference for almost twenty years, universally cited, its Br. numbers appearing in museum labels and auction catalogues alike. At the same time, scholars knew it was transitional — that a more rigorous technical examination programme was needed, and that the catalogue format itself, with its single-scholar authority, was becoming epistemologically awkward.
What Gerson was responding to, in part, was a generation of technical examination that Bredius had not had access to.
The apparatus that would eventually address this was the Rembrandt Research Project, which began work in 1968 — the year before Gerson's revision appeared — and whose published Corpus would eventually replace both editions as the primary scholarly instrument, while itself proving controversial.
The practical legacy of Bredius and Gerson together is the numbering system: a Br. number identifies a picture uniquely, gives its position in the 1935 or 1969 edition, and signals immediately that the work has at least been considered in the central tradition of Rembrandt scholarship, whatever conclusion that tradition eventually reached.

Museum entries, catalogue essays at the National Gallery in London, and scholarly articles regularly lead with a Br. number before any other reference. For a catalogue first published in Vienna during the mid-1930s and revised once, in 1969, by a single scholar working without the benefit of systematic technical examination, that is a remarkable afterlife.
It was the standard reference for almost twenty years, universally cited, its Br.

