^{page 465 note 1} I notice one error: Buecheler's edition of 1893 is said to be the third edition of Jahn's Juvenal of 1851 : it is the third edition of Jahn's Persius and Juvenal of 1868.

^{page 465 note 2} Mr Wilson adds indeed in a footnote ‘accipiet is an easier reading’; but knowledge is not to be thrown in loose fragments at the learner's head : it is for an editor to balance alternatives, to prefer the more probable, and to tell the learner why he prefers it. At vii 112 the note runs ‘conspuiturque sinus: a charm to avert the wrath of Nemesis, who punished boastfulness ; cf. Petron. 74. Some editors interpret is spluttered over (Lewis), gaining support in the prefix (con-)’: the student is left to choose for himself, and to frame for himself the question ‘does one, when lying and boasting, splutter more than at other times?’ The same failure to think matters out and reach assured conclusions causes other inconveniences of a different sort: at vii 42 Mr Wilson gives the usual and correct explanation of ‘sollicitas imitatur ianua portas,’ and the student, quite satisfied, is about to pass on, when his eye alights on this footnote : ‘J. Jessen, 1.1. p. 505, suggests that portas is a scribal error for porcas.’ Surely, thinks he, the editor would never have mentioned this unless there were either some defect in the explanation given or some superiority in the conjecture : which can it be? and what oan it be? And then he wastes his time in the attempt to guess.

^{page 466 note 1} His explanation of x 189 is almost the same as Mr Mayor's, and surely wrong.

^{page 467 note 1} I use this tern to exclude examples like causa terroris repentini, where the adjective adds percision.