Abramelin Manuscripts: Wolfenbüttel and the Wellcome Copies

The word “Abramelin” names a work transmitted through different manuscripts. A date in its story, a date in a title, and the date of a surviving copy answer different questions.

Wolfenbüttel: Cod. Guelf. 47.13 Aug. 4°, seventeenth century
Wellcome MS.820: German copy dated 1726
Wellcome MS.821: German copy dated about 1775

The Wolfenbüttel manuscript

The Herzog August Bibliothek describes Cod. Guelf. 47.13 Aug. 4° as a paper volume of 428 leaves from the seventeenth century. Its Abramelin material occupies folios 1–31v; the following leaves contain Jewish prayers and other texts. The catalog also notes two different hands. The Abramelin portion is therefore only one component of a much larger bound manuscript. A date or characteristic assigned to the entire volume need not describe every component in the same way.

Kollatsch identifies this Wolfenbüttel copy as the oldest complete witness of high textual quality. That is a scholarly assessment of the surviving textual tradition, not a claim that the physical manuscript was written in the fourteenth century. His dating argument uses linguistic and textual clues as well as manuscript context. The catalog provides a comparatively modest object date, “seventeenth century,” and is the right authority for what the library holds.

Two later German witnesses at Wellcome

Wellcome MS.820 is dated 1726 and described as a German version with three books. Its title claims older origins and a 1619 translation; those statements belong to the manuscript’s wording, while 1726 is the catalogue’s date for the surviving object. The record also preserves a note about an apparent transcription error in the title’s historical date. The quoted historical number is a title claim, not the date of the surviving copy.

Wellcome MS.821 is catalogued as a German translation written around 1775 and containing four books. The record supplies a shelfmark, physical description, ownership history, and the language. Its four-book form makes it useful when discussing material omitted from familiar three-book English editions, but its later copying date does not make it an earlier witness than the Wolfenbüttel manuscript.

What differs among the witnesses

The Wolfenbüttel volume dates to the seventeenth century, Wellcome MS.820 to 1726, and Wellcome MS.821 to about 1775. Their catalogues also describe different book counts and languages. The later German copies record a period when the work was still being translated, copied, and reorganized. Their historical value does not depend on being older than Wolfenbüttel: the changes in arrangement show how the text continued to circulate. A passage or square may occupy a different numbered book even when it belongs to the same broader tradition.

The surviving objects also differ from the dates stated inside their narratives and titles. MS.820 makes a claim about an earlier translation, while its paper copy is dated 1726. Those two dates describe different events. The institutional records identify the held volumes and their contents; they do not turn later copying into evidence that the narrated events happened when the text says they did.

Worked comparison: one square in two book maps

Follow one purpose-labelled letter square through the competing structures. In Mathers’s French-derived edition, the preparatory method occupies Book Two and the squares occupy Book Three. In the four-part German structure described by Kollatsch, the life story is followed by a separate recipe collection, then the extended initiation, and finally the square appendix as Book Four. The same kind of diagram therefore sits after a different amount and kind of preceding material. A reference to “Book Three” can point to the operation in one structure and the square collection in another.

The changed map also changes the argument made by the whole work. In both forms, the squares come after a period in which prayer, discipline, and the angelic encounter establish the operator’s authority; they are not presented as self-starting charms. The four-book form inserts an additional collection of practical material and preserves a longer program, while the familiar English form compresses that route. Comparing the tables of contents thus reveals a substantive difference in ritual logic, not simply a binder’s choice about where to place a chapter.

Source records

Continue in this archive

← Abramelin archive overview · Short occult-library overview