Repertories of conjectures can be of high value to editors of classical texts, and even more so to editors of Greek tragedy, who very often deal with a corrupt text and need to consider many conjectures. Such collections also ascribe conjectures to their original authors (as far as possible) and, more importantly, they draw the editors' attention to conjectures which have been neglected by all other editors. Even conjectures deemed very unlikely can still be helpful, because they could lead an editor to a better proposal of his own. Last, but not least, a repertory allows scholars to consider everything that has been suggested on any single passage and avoid offering (independently) conjectures already made in the past.

The repertory of conjectures on Prometheus Bound is the first volume of A New Repertory of Conjectures on Aeschylus. T. has reviewed all editions of Prometheus, as well as similar repertories compiled by N. Wecklein (1885 and 1893), R.D. Dawe (1965) and M.L. West (1990). He provides precise references to the editions or secondary works, where the conjectures were published. Identifying the place of a conjecture's publication was especially demanding for Wecklein's lists, which only included the names of the scholars who proposed the conjectures. T. has been able to add unknown conjectures to his repertory and attribute others to earlier scholars than to those to whom they have been assigned so far. More significantly, T. also reports a number of conjectures which are in fact found in Byzantine manuscripts, which he studied within the framework of a project on a new critical edition of this play.

T.'s work is a thorough contribution of permanent value. I noticed one omission from T.'s bibliography: P.J. Finglass, ‘Unpublished Conjectures at Leiden on the Greek Dramatists’, GRBS 49 (2009), 187–221, at 196. According to Finglass, ἐγεγήθει at PV 157 and ἔννοιαν at PV 446 should be ascribed to Valckenaer (1715–1785), instead of Elmsley (1810) and Wakefield (1793) respectively. It is very possible that more corrections like these will become necessary and, indeed, a loose sheet in my copy of T.'s book lists sixteen ‘Addenda and Corrigenda’. They mostly consist of conjectures which T. has now found to have been anticipated by medieval manuscripts. The editors of the New Repertory of Conjectures on Aeschylus may wish to consider publishing these databases in electronic format, which would facilitate regular updates, even by external contributors.