a collection of magical texts
Description
Internally, the collection falls in three parts:;A. a magical text, represented by 593, and duplicated by nos. 599, 594, 603, and 596, in the order named;;B. Another magical text, or series of texts, represented by nos. 600, 601, and 602, duplicated in part by nos. 597 and 598.;C. A third magical text, represented by no. 595, possibly a part of B, and resembling slightly in phraseology the Michigan papyrus no. 594a.;One's first impression is that III (593) is the original from which the "rough copies," nos. 599, 594, 603, and 596, were made: by an apprentice or pupil of the magician, or by himself for sale, or by some unskilled who wished to save the cost of professional services. But collation of the two versions of the text reveals a different relationship. There are many cases of the same error in both. There are cases in which an error of the copies is paralleled by a further corruption, or a frank omission, or a simple emendation in the codex: as though the codex were itself dependent upon the copies. In fact, the copies are right and the codex wrong in so many cases that one must regard the copies as older and better than the codex. But there are enough cases in which the codex corrects the copies to prevent assuming that the codex was made directly from the copies. They have a common ancestry, and the copies are in the better line. Perhaps the copies were made from a good manuscript when still in good condition, and the codex from the same manuscript after it had become old.
Cross-references (1)
- APIS-Text michigan.apis.3648 tier-1
About this record's data
- From the source institution — accession, description, dimensions, and dating are as catalogued by Papyri.info — APIS (Advanced Papyrological Information System) — papyri.
- AI-inferred — the image-analysis panel (deities, names, signs) is machine-generated and may be wrong.
- Approximate location — most map points are plotted at the site centroid, not the exact findspot.
- Inferred links — cross-references marked with a match method other than explicit-source-field were matched by us, not stated by the source.