A copy of an edition of Lucretius printed by Misinta in Brescia in 1496 was sold at auction in
London in 1900. Neither this nor any other copy of this edition can be located, and it has been
suggested that it may have been a ‘ghost’ generated by a cataloguing error. The authors present
compelling new evidence that there was no error. They discuss the ownership of the unique copy
and consider the probable nature of the text. The very fact that there was a fifth incunabular edition
of Lucretius, printed, like the editio princeps, in Brescia, is of considerable interest, and it is to be
hoped that a copy (or ‘the copy’) will come to light.