The lost catalog, and the palimpsest that gave it back 

Parchment was expensive, so it was reused. A scribe would scrape or wash the old ink from a sheet, cut and refold it, and write something new on the recovered surface. The result is a palimpsest, and the erased layer is often still faintly present in the fibres. Modern multispectral and ultraviolet imaging, sometimes with computational processing, can lift that undertext back into legibility. The technique gave back the Archimedes Palimpsest and its treatise on method, and it is now standard practice in manuscript studies.

The manuscript at the centre of this story is the Codex Climaci Rescriptus, a collection of leaves long associated with Saint Catherine's Monastery in Sinai and now largely held by the Museum of the Bible in Washington. Its visible text is Christian material in Syriac. Beneath it lie earlier Greek pages, and scholars had already identified some of the undertext as a late antique astronomical work, the Codex Climaci having been assembled from more than one older book.

In 2022 a team including Victor Gysembergh, Emmanuel Zingg and Peter Williams published a study of one such folio in the Journal for the History of Astronomy. Working from multispectral images, they read a passage giving coordinates for stars in the constellation Corona Borealis, stated as the positions of the stars at the extremities of the figure. The passage was not written in Ptolemy's coordinate system and its numbers did not match the Almagest.

The dating argument runs through precession. Because stellar longitudes shift steadily, a set of coordinates carries a timestamp: you can ask which epoch makes the numbers self-consistent. The researchers reported that these values fit an epoch in the second century BC, close to the working life of Hipparchus, with an accuracy in the region of one degree. On that basis they argued the passage most likely derives from his lost catalog, transmitted through the later astronomical text.

It is worth stating the limits plainly. This is a small fragment, a handful of stars rather than a catalog. The reading of individual numerals in an erased undertext involves judgement. The attribution is an inference from epoch and coordinate system, not a signature, and other specialists have discussed alternative routes of transmission. The finding is best described as strong evidence rather than proof, which is roughly how the authors themselves framed it.

Even held carefully, it is a remarkable thing. It suggests that Hipparchus worked to about a degree, that he used a coordinate frame Ptolemy later replaced, and that more of him may be sitting unread inside reused parchment in library collections. Manuscript imaging projects continue, and the same folios are being re-examined.

Two thousand years on, the question he was really asking has not changed at all, only the size of the instrument answering it.