^{page 169 note 2} Exceptional cases occur, where Heinsius says ‘in Coloniensi legitur, etiam Modio teste’ or the like (14. 422) or even says expressly that his collation differs from Mod. here (so 8. 671).

^{page 170 note 1} l.c. p. 208.

^{page 170 note 2} Blass, strangely enough, cites this passage (note 37) to shew that Drakenborch, whilst professing to. give Modius' note from Novantiquae Lectiones, edits it and edits it wrongly. This is true, but he omits to notice that Heinsius' note contains the same mistake. This rather suggests that Drakenborch is really responsible for both; however, Barth has made the same mistake, Adversaria 1. 20.

^{page 170 note 3} l.c. note 38.

^{page 170 note 4} See his ed., preface p. 8.

^{page 170 note 5} Or ‘ Ven. vetusta,’ which means the same.

^{page 170 note 6} The Venice editions of 1483 and 1492, the Paris one of 1512. My thanks are due in this matter to Dr. J. P. Postgate and the Rev. Dr. Sinker of Trinity College. There is a Venice ed. of 1493, which I have not seen, but various circumstances make me confident its readings will not make any difference in the matter.

^{page 171 note 1} In 11. 163 the reading as given is unmetrical, but Heinsius gives a conjecture which adopts both the magnum and the positum, and no ed. before him does this.

^{page 171 note 2} The ed. Colinaei did contain conjectures in addition to the collation copied from the Gryphian: Drakenborch's Preface, p. 6.

^{page 171 note 3} Blass also believes he has given us some of Livineius' conjectures as if they were readings of C (l.c. p. 210). But I doubt the instances he gives. In 7. 497 where he assumes furorem, was really not an C, but came from Livineius, it is worth noting that the Parma ed. gave this reading long before.

^{page 172 note 1} See Blass, l.c. p. 210.