More from Pfeiffer, on the learned Angelo Ambrogini (i.e., Poliziano), who was tutor to the children of Lorenzo the Magnificent and lecture on Latin and Greek literature:
In [Politian’s] circle the text of Callimachus’ Hymns was most eagerly copied, and he himself translated the Bath of Pallas. In his notes to this hymn he confessed ‘that he was not afraid to correct small corruptions of the original’; but in one passage (line 136), where he was thought to have divined the true Callimachean text, we now know that he was completely astray. Only the concluding word was preserved in Politian’s Greek manuscript, and in his Latin translation he supplied the pentameter according to the sense he expected. This seemed to be confirmed by the text F. Robortello used in his edition of Callimachus’ Hymns (1555), where the Greek line agreed completely with Politian’s Latin one. But unfortunately, it is the other way round. Robortello’s manuscript is one of the interpolated manuscripts of the sixteenth century in which all the gaps of the archetype are filled by modern supplements, and the one in question is nothing but a poor translation of Politian’s Latin into Greek. That both Politian’s guess at the contents of the line and all the Greek words are wrong is now proved by another group of manuscripts, unknown to him, in which four syllables of the beginning of the pentameter are preserved.
(History of Classical Scholarship, vol. 2, pp. 45-6)