In spite of the labours of Sedlmayer,1 Ehwald2 and Palmer,3 it cannot be said that there exists a completely satisfactory edition of Ovid's Heroides. One or all of these editors sometimes leave a corrupted text, sometimes adhere too closely to a manuscript reading, and sometimes introduce untenable emendations. A new edition is called for, with revised collati ons of the known manuscripts, and an augmented apparatus criticus, exhibiting the large class of what I may term the ‘Vulgate’ manuscripts, which represents a tradition different from, and probably later than that of our chief authority, the excellent but unfortunately incomplete eleventh-century Parisinus (P), which, like all primary manuscripts of this type, contains many readings or corruptions which should not on account of its mere authority be accepted slavishly. Light is often thrown on the text from less important sources, and the truth may be recovered from manuscripts of the ‘Vulgate’ family, where our better authority fails. In this connexion the study of my own manuscript (O) has led me to some conclusions in certain passages which I venture to set forth as suggestions towards an improved constitution of the text of the Heroides. Though it is clear from the preceding investigation into its nature that O is no primary manuscript, but belongs rather to that large class which has passed through various stages of reproduction involving alterations (as is natural in the case of so popular an author as Ovid) due either to carelessness of copyists or, more rarely, to deliberate alteration, it presents at the same time a phenomenon not unusual with such manuscripts, inasmuch as it often supports the best tradition, and in some cases preserves a reading which yields the truth, or from which the truth can be elucidated. I have used the following symbols:

P = Parisinus 8242 s. xi.

E = Etonensis s. xi.

G = Gueferbytanus s. xii.

V = Schedae Vindobonenses s xii.

O = my manuscript s. xiv.

D = Dresdensis s. xiii.

ω = all or the majonly of the ‘Vulgate’ manuscripts.

ζ = some of these manuscripts.