Long may you keep alive in Cambridge the traditions of Ascham, Cheke, Bentley; and when in a ripe old age you have followed to their rest the pupils whose names will always recall yours-such as Shilleto, Riddell, Clark, Cope, Munro, Fraser, Bather, Morse—may your portrait exhort generation after generation of Johnian-Salopians : SIS MEMOR ATQVE ANIMO REPETAS EXEMPLA TVORVM.

ADVERTISEMENT

HEREWITH I fulfil, so far as this volume is concerned, my promise of fourteen years ago. In the long vacation. of 1885 I had leisure to examine my collections of 32 years, and thence 'to enlarge the commentary on' these five 'satires on one uniform scale'.

I have also added a supplementary index of phrases. The commoner particles and pronouns; omnis; the verbs dico, habeo, possum, sum; are only represented by selections for all other words I hope that the present index, together with that in volume II, will be found exhaustive. When I began, I intended merely to index my additional notes, and certainly it was not a morbid craving for the familiar drudgery that led to my change of purpose, but rather an imperative sense of duty.

When, in 1850 or 1851, my friend the publisher said to me 'You ought to bring out a book', it was no special acquaintance with Juvenal that suggested the choice, but dissatisfaction with Ruperti's' edition, then holding

1 Ruperti was a diligent gleaner in the commentaries of Heyne, Jacobs, Hemsterhuis, and other scholars, chiefly belonging to the 17th and 18th centuries, but is sadly deficient in taste and scholarly instinct. Heinrich worked out some special points, in lexicography and archaeology, excellently, but his work was left incomplete and by no means supersedes Ruperti's, as Madvig and Bernays have seen. In moral sympathy with Juvenal he ranks with the best critics.

the field: 'I have a good many notes on Juvenal, and Ruperti's book is not worthy of his author'.

But if I have had a share in silencing Ruperti, I am responsible for depriving students of a most valuable help to the study of Juvenal.

The index rerum et verborum in Priestley's reprint (Lond. 1824) occupies 147 closely printed pages, large 8vo; extended by Lemaire (whose enlarged Ruperti is little known in this country) to 234 pages. The burden imposed on me by this recollection, was aggravated by another consideration. Jahn's index, of words not of phrases, which aims at, and in great measure attains to, absolute completeness', is from carelessness of execution utterly untrustworthy. As the preface was reserved for (the still lacking) volume II, destined for the commentary, we have to construct conjecturally a history of this index, which can never have passed under Jahn's eye. It was compiled, presumably by a student, in the most mechanical fashion. Thus caligas n. s. (xvi 24) cālīgantes (vi 31) and căligatus (iii 322) are classed (the last two also by Ruperti and Lemaire) as parts of one verb. Under consul we have not only viii 23 praecedant ipsas illi te consule virgas, but also xi 33 te consule, dic tibi qui sis', a not very rare equivalent to γνώθι σεαυτόν.

1 For example the word et fills two columns and 21 lines; a sheer waste of type, as it would take less time to read the satires from beginning to end than to verify several hundreds of citations.

2 This imperative with acc. is, I am told, often construed by undergraduates, under stress of examination, as abl. abs. Let them take heart of grace. 'We live by great examples in this world'. Two of the best known authorities in Latin syntax have been caught in the same trap.

Haase on Reisig's Vorlesungen (Leipz. 1839) p. 760 gives a bare reference to Iuv. xi 33, but cites xiii 3 se iudice nemo nocens absolvitur at length. From him probably, not from Jahn's index, Dräger hist. Synt. 11" 810 has borrowed the former example, which he mutilates and punctuates

ADVERTISEMENT

Under deses iv 44 desidia tardos.

ix

xv 88 nil umquam hac carne libentius edit, does duty as a present; the comparative adverb (xi 12) egregius cenat as nom. masc. adj.

Different persons bearing the same name1, or different substantives spelt alike, are huddled together (e.g. under Gallus vii 144. ix 30. xvi 1; under gallus viii 176. ix 107. xiii 233; lacerti ‘a fish' in xiv 131, with lacertis 'thews' x 11; nay, the fungi ancipites of v 146 stand in line with the functus of xi 88).

The genuinum of v 69 must be sought under geminum; olfecisse of vii 225 under officio; patriam n. s. x 142 under patrius; ignota, -um, -os 'unknown', under ignosco; vii 192 subtexit under subtego; viii (not ix) 145 tempora 'temples' under tempus 'time' (so Ruperti and Lemaire); uva of xiii 68 under unus; xiv 8 et eodem iure natantes, follows the numerous examples of ius 'right', an error avoided by Ruperti and Lemaire, but not by the Delphin editor; so plantas iii 227 of the vegetable kingdom is confused with the human planta; pullo palliolo (so Jahn in both editions reads with Buchner in iii 94) with the n. s., rēferre with referre throughout (so also Maittaire, Ruperti, Lemaire, but not the Delphin editor).

These and many other errors called aloud for a remedy;

thus seu affectas te consule ctt. I seize this opportunity of remarking that, useful as Dräger's books are, no one who values accuracy should accept any statement in them without careful verification. They contain very little original research.

I have noted a third offender, R. Bitschofsky, in a learned review (Philologische Rundschau 1 1206).

Add to the examples (from Sen. and Plin. ep.) quoted in vol. II p. 222, Sen. ep. 59 § 14 nunc ipse te consule. ben. VI 38 § 5 denique se quisque consulat et in secretum pectoris sui redeat.

1 In Jahn's index of proper names to his second edition (Berol. 1868) this blemish is removed.

b

and as my collections have for more than 30 years been used by grammarians and lexicographers, I felt bound to act. To ensure completeness and accuracy I collated my first proof with Maittaire (Tonson 1716) and Jahn (Berlin 1851); the revise with Lemaire (Paris 1825) and a London reprint of the Delphin (trade edition, ‘octava prioribus multo correctior', 1750). Whoever cares to follow me through a single page will see that the labour was severe; but it is so difficult, at least for me, hoc agere throughout a multitude of details, that I do not grudge the time, especially as Latin lexicographers, with the veteran Georges at their head, are now waking from their slumbers and taking a higher view of their responsibilities. In England at any rate, for many years, the commercial interest of such work has drowned every other. We ought not to be satisfied till we have a Latin thesaurus as extensive as Didot's Greek Stephanus, and far more complete and exact.

Jahn's index confutes the heresy that any hireling can be trusted with such a task. If genius is 'an infinite capacity of taking pains', no genius need decline an employment demanding, like sentinel's duty in the field, incessant vigilance. Of late indexes of the Delphin type1 have come again into fashion. We are grateful for them; they are certainly better than nothing; but no mere vocabulary can represent an author's mind as a concordance does, whose bulk may be kept within bounds, without sacrifice of efficiency, by a wisely compressed notation3.

1 Containing bare words, not phrases.

2 To Caesar, Holder; to Catullus, Ellis and Schwabe; to Vitruvius, Nohl; and several in Teubner's library.

3 Among other excellent remarks on index-making a great master, Pierre Bayle (avertissement sur la seconde édition of his Dictionary, vol. xvi p. 22, Par. 1820) testifies: “Je sais par ma propre expérience, et par celle de plusieurs autres, que les articles d'une table chargée d'une demipage