§1. nobis ipsis, codd.: e nobis ipsis Gertz.
utilitatis etiam. Ioan. gives etiam utilitatis, which Spalding quotes also from Goth.
§2. alte refossa. This (the reading of N) I have found also in Ioan. and Prat.: alter effossa BH: altius effossa Harl. 4995 M Harl. 4950, 4829 Burn. 244 Bodl. Dorv.: alte effossa Harl. 2662, 11671.
fecundior fit. Fit appears as a correction in T and Vall.: it does not occur in B M Prat. H T Ioan. S Harl. 4995 or 2662. Perhaps fecundior is the true reading, and est is to be supplied in thought: Introd. p. lv.
effundit B Prat. Ioan. N and most codd.: effunditur b H. et fundit Vall.2 M, Harl. 4995, Halm and Meister.
parentis: parentium Ioan.: parentum Dorv. Harl. 4950 Burn. 244 C: parentibus bH Bodl.
§4. iam hinc. Obrecht iam hunc: see note ad loc. Harl. 2662 and 11671 agree in iam hic.
§6. scriptorum. This reading, attributed to Badius by Halm and Meister, is found in Ioan. Harl. 4995 Burn. 243 Harl. 2662 (the last corr. from -em). It is also in the editio princeps (Campanus), and the ed. Andr. Becher reports it as a correction in Vall.
§9. sequetur Bn and Bg N Sal. Dorv. Harl. 2662, 4950, 4829, 11671: persequetur b Harl. 4995 Burn. 243: prosequetur HM Bodl. and Prat. Prosequetur (Spald. and Bonnell) may be right: there is a graphic touch about the compound.
§10. ut provideamus obelized by Halm (after Bursian): but see note. Becher proposed provideamus ut resistamus et ... coerceamus: Krüger suggests rather resistamus et provideamus ut ... coerceamus: Jeep, ut provide eamus, also, for efferentes se, efferventes. The passage is discussed by Kiderlin (Blätter f.d. bayer Gymn. 1888, p. 85), who recommends the excision of et before efferentes, as it is found in no MS. He translates: ‘Aber gerade dann, wenn wir uns jene Fähigkeit (schnell zu schreiben) angeeignet haben (bei solchen, welche noch nicht schnell schreiben können, fehlt es an Ruhepausen obnehin nicht), wollen wir innehalten, um vorwärts zu blicken, die durchgehenden Rosse wollen wir gleichsam mit den Zügeln zurückhalten.’ He considers ut provideamus a necessary addition, in order to make the meaning of resistamus clear. ‘Was jeder Besonnene beim Schreiben thut, dass er manchmal innehält, um vorwärts zu blicken, d.h. um sich zu besinnen, welche Gedanken nun am besten folgen und wie sie am besten ausgedrückt werden, rät hier Quint. seinen Lesern.’ The best MSS. read resist. ut provid. efferentes equos frenis: Hb Bodl. Burn. 243 give ut for et: Harl. 4995 has resist. ut prohibeamus ferentes equos fr. quib. coerc.: 4950 and Burn. 244 resist. ut prohibeamus efferentes equos quos fr. quib. coerc. The reading et efferentes se is due to Burmann. Something might be said for et ferentes se: ‘ferre se’ is often used by Vergil of ‘moving with conscious pride,’ e.g. Aen. i. 503: v. 372: viii. 198: ix. 597: xi. 779.
§12. patruo. Harl. 2662 and 11671 both give patrono: which, with other coincidences, establishes their relationship to the Guelferbytanus (Spald.).
§14. quod omni, see note ad loc.: edd. vett ex quo.
§15. plura et celerius Prat. N: and so now Becher reports from B and Ambrosianus ii. Et had escaped Halm’s notice, and Meister follows, plura celerius.
sed quid: sed is supplied by the old edd., but does not appear in any MS. Halm (ii. p. 369) conjectures at, which may easily have slipped out after obveniat.
§17. quae fuit: (manent) quae fudit Harl. 4995 (as also Goth. Voss. 2 and Vall.)
§19. urget. Kiderlin supports (in Blätter f. d. bayer. Gymn. 1888, p. 86) his proposal to read urgetur, which would however give a different antithesis. ‘When we write ourselves, our thoughts outstrip our pen, but when we dictate we forget that the scribe is writing under similar conditions, and give him too much to do.’
§20. in intellegendo. This conj., which is due to H. J. Müller and Iwan Müller, has been adopted by Becher and Meister: legendo BM Ioan, and most codd. (Halm). See note ad loc. The true reading may be si tardior in scribendo aut incertior, et in intellegendo velut offensator fuit. This is supported by et diligendo (bH Burn. 243 Bodl.), for which Spalding conjectured et delendo, Gertz in tenendo (‘significatur notarium imperitum et oscitantem verba quae dictantur non statim intellegere aut fideliter tenere, ut saepius eadem dictanda sint’). A number of codd. (Ioan. Vall. Harl. 4995, 4950, 4829, Burn. 243 and 244, Dorv.) have inertior for incertior: but this gives no antithesis to tardior: it appears, however, in ed. Colon. 1527. The same codd. (and also M) have fuerit, for fuit, which may be right.
concepta Regius: conceptae codd. Becher points out that concipere and excutere are ‘termini technici’: cp. Scrib. ep. ad C. Jul. Callist. p. 3 R ne praegnanti medicamentum quo conceptum excutitur detur: and Ovid, excute virgineo conceptas pectore flammas.
§21. altiorem. This reading, ascribed by Halm and Meister to ed. Colon. (1536) I have found in Harl. 2662 (A.D. 1434) and 11671 (A.D. 1467). B N Ioan, and other codd. aptiorem: Prat. apertiorem, and so a later hand in Vall.
frontem et latus interim obiurgare. B, Prat. M, Ioan., Harl. 2662, 4950, 4829, 11671, Burn. 244 and Dorv. all give simul et interim: Harl. 4995 (again in agreement with the 2nd hand in Vall.) and Burn. 243 have simul vertere latus et interim (the reading of many old edd.): so Bodl. except that it omits et. It is to b that we must apply for what must be at least a trace of the true reading; and b gives sintieletus, which H shows as sintielatus. Considering how liable s (ſ) and f are to be confused, I venture to think that ſinti may conceal fronte.
Bursian’s femur et latus (Halm and Meister) is not so near the MSS.: it is based on ii. 12. 10 and xi. 3. 123 (quoted ad loc.), but the latter passage would warrant frontem quite as much as femur, and frontem ferire seems to have been considered by Quintilian a more extravagant action than femur ferire, of which he says ‘et usitatum est et indignantes decet et excitat auditorem.’ In any case the man who is in the agony of composition is as likely, if alone, to ‘rap his forehead’ and ‘smite his chest,’ as to ‘slap his thigh.’
Frotscher and Bonnell’s sinum et latus cannot be supported by any parallel for such an expression as sinum caedere, ferire, obiurgare. Becher approves Gertz’s conjecture semet interim obiurgare, which is adopted also by Krüger (3rd ed.) as = increpare: ‘obiurgat semet ipse scribens et convicium sibi facit ut stulto, si quando tardior in inveniendo est.’
Another interesting conjecture is put forward by Kiderlin (Blätter f. d. bayer. Gymn. 1888, p. 87). He proposes to read (on the lines of b) singultire, latus int. ob. This would need to be taken of those more or less inarticulate sounds which the solitary writer addresses πρὸς ὃν θυμόν, when there is no one there to listen. Kiderlin refers to singultantium in 7 §20, of broken utterance: but we cannot take the reference here of ‘sobs’ or ‘gasps’: the writer is not practising with a view to theatrical effect, he is supposed to be indulging in little peculiarities that become ridiculous in another’s presence. As an alternative Kiderlin suggests singultu latus interim obiurgare, comparing for the ablative §15 cogitationem murmure agitantes. Singultus is common enough: and Kiderlin thinks that as singultire is nearer the MSS. than singultare, it may possibly have been used here by Quintilian.
§22. secretum in dictando. So bH Harl. 4995, 4950, Burn. 243, Bodl., M, Dorv.: quod dictando BN Prat. Ioan., Harl. 2662, 4829, 11671, Burn. 244 (corr. to in). With the reading quod dictando perit, atque liberum ... nemo dubitaverit (Halm and Meister) it is senseless to quote 2 §20 (Bonn., Meister, and Dosson) as parallel. Krüger (3rd ed.) reads secretum dictando perit. Atque liberum arbitris, &c.
§23. mihi certe iucundus. After these words H has videmoni (and so the cod. Alm.): Flor. vindemoni. This word greatly puzzled Spalding, and has been allowed to disappear from the critical editions of Halm and Meister. Jeep transformed it into mihi certe vitae inani iucundus, &c. An ingenious suggestion is made by Mr. L. C. Purser (in the Classical Review, ii, p. 222 b). He thinks that it may be “the gloss of a monk, on a somewhat ornate passage about poetry, who recollected how (as Bacon says in his ‘Essay on Truth’) one of the Fathers had in great severity called Poesie vinum daemonum.” Cp. Advancement of Learning ii. 22. 13, where Mr. Wright tells us that Augustine calls poetry vinum erroris ab ebriis doctoribus propinatum, Confess. i. 16; and that Jerome, in one of his letters to Damasus, says Daemonum cibus est carmina poetarum, while both these quotations are combined in one passage by Cornelius Agrippa, de Incert. &c. c. 4. Hence the phrase vinum daemonum may have been compounded.—If the gloss is to be credited to the copyist of H (as seems probable), it perhaps arose from something that caught his eye in the Bambergensis four lines further down, where tendere ani(mum) is shown in a form that could easily be mistaken by a sleepy scribe.
§24. ramis, referred by Halm and Meister to ed. Camp., appears in Harl. 4995: it is reported by Becher also from the Vallensis. All other codd. rami.
voluptas ista videatur most codd.: videatur ista voluptas N.
§25. oculi. Kiderlin thinks it allowable to infer from the words ex quo nulla exaudiri vox that aures aut has fallen out before oculi. Cp. §28 nihil eorum quae oculis vel auribus incursant.
velut tectos: velut rectos all codd. There is the same confusion at ix. 1. 20 where M has recteque for tecteque (i.e. tectaeque). For Becher’s explanation of the vulgate tectos (first in ed. Leid.) see ad loc. Kiderlin (Blätter f. d. bayer. Gymn. 1888, p. 88) is not satisfied, and objects that for tectos teneat we should have expected tegat. The figure also seems to him out of place, as the context speaks not of the attack of an enemy, but of the distractions which draw the mind of the student away from his task: §23 avocent, respexit: §24 ad se trahunt: §25 aliud agere. He proposes, therefore, velut recto itinere, comparing iv. 2. 104 ut vi quadam videamur adfectus velut recto itinere depulsi, and ii. 3. 9 et recto itinere lassi plerumque devertunt. Itinere may first have fallen out, and then recto may have been changed to rectos.—Halm conjectured velut secretos, or coercitos; Wrobel, velut relictos.
§26. haud deerit: aut deerit BN Ioan, and all codd. except a later hand in Vall. Kiderlin (Blätter l.c.) comments on the infrequent use of haud in Quintilian, though haud dubie 1 §85 (where however GH have aut) must have escaped him (cp. i. 1. 4); and founding on the consensus of the MSS. for aut he proposes to read aut non deerit or aut certe non deerit. But haud goes closely with deerit, and does not (like non, ac non) introduce an antithesis to supererit. Aut deerit might be made to mean that the sleepless man is to work: but this would be too cruel!
§29. et itinere deerremus: et ita ne BN Ioan. Harl. 2662, 4829, 11671, Dorv. and Ball.: ita erremus HMb Bodl. (erramus). The reading in the text is given by Halm and Meister as from the old editions: it occurs in Vall. and Harl. 4995.
§31. crebra relatione appears in Harl. 4995 (and Vall.) corrected from crebro relationi which is the reading of B Ioan. and all codd. Jeep suggested crebra dilatione, Kiderlin crebriore elatione. Other proposals are crebra relictionis, q. i. c., repetitione, Gottfried Hermann (in Frotscher), crebra relictione, q. i. c., et repetitione, Zumpt (in Spald. v, p. 423). Becher thinks crebro may be right, adverbs being often used in Latin where we should use adjectives: crebro would then go closely with morantur and frangunt.
§32. adiciendo ‘for making additions’: so Bursian, Halm, and Becher. BN Prat. Ioan, and most codd. have adicienda: b adiciendi sint: Harl. adjiciendi sit. Meister adopts adicienti from ed. Col. 1555: so Spalding: cp. iv. 5. 6 quo cognoscenti iudicium conamur auferre (where B has cognoscendi).
ultra modum esse ceras velim: Ioan, omits esse, and is thus in agreement with N.
§3. habet: habeat, Halm quoting from ed. Camp. Habeat occurs in Burn. 243: most codd. have habet, but some (H and Bodl.) give habent.
§1. ἕξιν parantibus: for the ex imparantibus of Bn N and Ioan. Bursian added non est huius. So Halm. Harl. 4995 gives nec exuberantis id quidem est operis ut explicemus.
factum est iam, Halm and Meister: est etiam all codd. except Ioan, which has factum etiam.
iam robustorum: so all codd. except bHFT which omit iam: and Harl. 4995, Burn. 244 which give iam robustiorum.
§2. id Messallae: B Ioan. M and most codd. Ball. and Dorv. however give M. id Messalae: and Harl. 4995 Marco id Messalae. The spelling Messallae is adopted in the text as more correct.
§4. eadem: so most edd. and Spalding, followed by Mayor and Krüger (3rd ed.): eandem all codd., with the single exception of M, and so Halm and Meister, though without giving any indication of the meaning. The only way to explain eandem seems to be to continue the sentence in thought sc. quae non proprie, or quae apud poetas: cp. eandem i. 9. 1. The sense will then be: ‘the poet’s inspiration has an elevating influence, while his licences of style do not carry with them in advance, or involve, the corresponding ability to use the language of ordinary prose: something is left for the reproducer.’ This suggests that there may be something in the reading of B (also Vall. and Harl. 4995), which have no non with praesumunt, at least if we may read eadem: ‘poetical licence implies that the orator can say the same things propriis verbis.’ Bursian suggested nec (for et) verba ... praesumunt.
§5. post quod. Harl. 4995 again agrees with Goth. and Voss. 2, praeter quod: so Vall.
§13. reus sit. Krüger (3rd ed.) revives Halm’s conj. rectene reus sit, to correspond with rectene occiderit and honestene tradiderit in what follows: along with Gertz’s quaeramus, an to correspond with veniat in iudicium an, Becher, however (Philol. xiv, p. 724), has pointed out that if the object of such a change is to secure complete symmetry, we should need to read, ‘Cornelius rectene codicem legerit’ quaeramus, an ‘liceatne magistratui ... recitare’: otherwise, in the other two cases the text ought to run, ‘Milo quod Clodium occidit’ veniat in iudicium, an..., and ‘Cato quod Marciam tradidit Hortensio’ an. Qnintilian has avoided this excess of parallelism without coming into conflict with logic.
Just as at iii. 5. 10 we have Milo Clodium occidit, iure occidit insidiatorem: nonne hoc quaeritur, an sit ius insidiatorem occidendi?, so here the finita or specialis causa shows the form of a positive statement (Cornelius reus est), as frequently in Seneca. Reus sit and legerit are motived only by the disjunctive interrog.: it might have run ‘utrum dicamus, Cornelius reus est,’ or only ‘Corn. quod legit ... reus est.’ The infinita quaestio, on the other hand, appears as in the above example in the form of a question, and this form the writer adheres to in the two following finitae and infinitae quaestiones. The finita quaestio rests on the generalis quaestio: acquittal of the charge (here laesa maiestas) depends on the answer to violeturne, &c. In a word, it is as if Quintilian had written (as at iii. 5. 10) Cornelius quod codicem legit, reus est: nonne hoc quaeritur: violeturne, &c.
§14. dum adulescit profectus, B Harl. 2662, 4995, 4829, Burn. 244, Ball.: inventus Hb Bodl. Burn. 243: Bonnell’s conj. invenis appears in Dorv. Bursian and Jeep conj. dum adul. profectui sunt util.
quia inventionem, Halm: quae inventionem all codd. Qy. quod?
§16. materia fuerit. Meister suggests erit: perhaps rather fuerit—necesse erit.
§17. assuescere Zumpt: assuefieri Philander. All MSS. have assuefacere. Frotscher wrote inanibus se simulacris ... assuefacere, and was followed by Halm. Most MSS. also (B Ioan. Ball. Harl. 2662, 4995, 4829, 11671) give difficilis digressus: but in view of the consensus for assuefacere the alternation difficilius digressos (H Bodl. Dorv. Harl. 4950 Burn. 243) is worth considering: inanibus simulacris would then go (though awkwardly) with detineri (for the rhythm cp. x. 2. 1), and the rest of the sentence makes excellent sense.
§18. transferrentur N Dorv. Ball. Harl. 2662.
§20. decretoriis Harl. 4995, probably from a correction in Vall.: Voss. 2 and Goth. (Spald.) derectoriis BJ Ball. Dorv. Burn. 244: detectoris b: delectoris H: delectoriis Bodl.: de rhetoriis Harl. 2662, 4829, 11671: vel rhetoricis M.
satis so most codd. But Bodl. Dorv. Burn. 243 litis: Hb sitis.
§21. idoneus bHM: si idoneus Bn Bg Sal.: sudoneus N: is idoneus Halm.
§22. sustinere Halm and Meister: sustineri Bn Bg HN Sal.
recidet occurs in Dorv., and is reported by Becher as a correction in Vall.: all other codd. recidere.
§23. diligenter effecta all codd. Regius proposed una diligenter effecta, Badius una enim diligenter effecta, and so many edd. Una would come in well before quam; but Becher rightly holds that it is unnecessary, the opposition being not quantitative alone, but qualitative as well. He reports una enim as a correction in the Vallensis.
quidque. Fleckeisen proposed quicquid; see Madvig on de Fin. v. §24.
§1. vacui nec otium patitur. The reading in the text, which is quite satisfactory, occurs in Harl. 4995, 4950, and Dorv. Bn and Bg give vacuum otium pat., and are followed by N Ioan. Harl. 2662 and 11671. For otium patitur b (followed by HFT) gives the remarkable reading experientium (experientiam Burn. 243, Bodl.), which reminds one of the confusion at the opening of ch. v: may the true reading perhaps be nec ἕξιν parantibus otium patitur? Jeep suggested expetit otium: nec perire otium patitur has also been suggested.
§2. desit. After this word there is a considerable space left blank in Bn and Bg, as well as in some later MSS., e.g. Harl. 2662 and 11671. In Harl. 4995 there is no blank, but in the margin the words ‘hic deficit antiquus codex.’
inhaeret ... quod laxatur: a later hand in Vall., Meister, and Krüger. BMN give inhaeret ... quae laxatur, which appears in ed. Camp. (and Halm) as inhaerent ... quae laxantur.
§4. tandem Madvig, Emend. Liv. p. 61, tamen libri.
§5. redire. I find this reading in Bg Ioan. C Harl. 2662, 4995, 4829, and restore it to the text, in place of regredi (Halm and Meister), which seems to have arisen out of redi HF, and occurs in Harl. 4950, Burn. 243, 244, and Dorv.
§6. domo Harl. 4995: domū B Ioan. MN Sal.
§7. utrimque Bonnell and Meister. The codd. give utrumque. Gesner (followed by Halm: cp. i. §131) proposed utcumque: Spalding utique: Jeep si tutius utcumque quaerendum est (cp. iv. 1. 21), founding on the reading of b strict * * * (margine adcisa), which reappears in HFT (strictius—strutius).
§1. praemium quoddam Harl. 4995, probably following a correction in the Vallensis: primus quid amplius Bn Bg Ioan. Sal. HFTM Harl. 2662, 4950. Amplissimum Stoer.
intrare portum Bn Bg H Ioan. N Sal. and most MSS. Halm adopts Meiser’s conj. instar portus. On this reading the advocate who has nothing but (solam) the scribendi facultas, and who therefore is found wanting at a crisis, is compared to a harbour which seems to promise a refuge to every ship at sea, but which really (owing to rocks and sand-banks) can afford protection only when the sea is calm, and so not praesentissimis quibusque periculis. Neither of the two justifies the expectations formed. But it must be admitted that the comparison of a man to a harbour is awkward. Other suggestions are monstrare portum: instaurare p.: and in terra portum (?) Jeep.
§2. statimque. I follow Krüger (3rd ed.) in the punctuation: see ad loc. The editors print statimque, si non succ.
§3. quae vero patitur, &c. In the text possit (for sit of MSS.) is due to Frotscher, omittere (for mittere) to Bonnell. Ratio (for oratio Bn Bg H Ioan. M) occurs in Harl. 4995. Krüger (3rd ed.), following Gertz, reads quae vero patitur hoc ratio ut quisquam sit orator aliquando? mitto casus: quid, &c. Aliquando he takes as = ‘only sometimes,’ ‘not always’ (i.e. tum demum cum se praeparare potuerit). For mitto casus (‘praeteritio’) he compares v. 10. 92: xi. 2. 25.
§5. quid secundum ac deinceps: so Harl. 4995. The MSS. clearly point to this reading, though Halm and Meister print ac sec. et deinc. Bn and Bg (as also N Ioan. and Sal.) have ac sec. ac dein.: but in Bg above the first ac the letter d appears (evidently for quid, not ad as H), and over the second ac, et is written, and is adopted by HFTM. In place of the first ac Harl. 2662 gives atque, and so Spalding reports Guelf. (with which 2662 is frequently in agreement). The Carcassonensis also has quid secundum.
§6. via dicet ducetur, bHFM Harl. 4950 Burn. 244: ducet ducetur Bn Bg Ioan. Sal. Dorv. Harl. 4995 shows the variant viam discet (as Goth. Voss. 2 Vall.) Meister, following Eussner, inverts the words, reading ducetur, dicet to avoid a ‘tautology’: cp. iii. 7. 15: ix. 4. 120. Bonnet changed ducetur into utetur. Kiderlin cannot believe that Quintilian wrote ducetur ... velut duce, and suggests that certa may have fallen out after serie (Rhein. Mus. 46, p. 24). This gives, he thinks, additional point to the clause introduced by propter quod: men who have had but little practice do not always speak methodically (via), but in telling stories they have no difficulty in keeping to the thread of their discourse, because the sequence of events is ‘a trusty guide.’
§8. paulum, BM Harl. 2662, 4829, 11671, Burn. 244, Dorv.: paululum bHN Ioan. Harl. 4995, 4950, Burn. 243, Bodl.
sed ipsum os coit atque concurrit, Halm, by adding os to the reading of B (Harl. 2662, 4995). sed ipsum os quoque concurrit, Spalding after Gesner. In Ioan. I find sed id ipsum coit atque conc., which may show that we ought to read os ipsum.
elocutioni, b: om. B (also N Ioan. Harl. 2662 Sal.) ‘haud scio an recte,’ Halm.
§9. observatione una, Harl. 4995 M Dorv. and Meister: observationen (-nū Bg) in luna Bn Bg Ioan. N Sal. Harl. 2662, 4829, 11671: observatione (-um H) in una bH: observatione simul Halm.
§13. superfluere video, cum eo quod, Harl. 4995, Voss. 2 Goth. Spald. and most edd.: superfluere video: quodsi Halm, and a later hand in Vall. (Becher): videmus superfluere: cum eo quodsi Meister, followed by Hild and Krüger (3rd ed.). The commonest MS. reading is superfluere cum eo quod (BHFTN Sal. Ioan. Harl. 2662, 4829, 11671, Burn. 243, Bodl., Dorv.), from which video seems to have disappeared: the later hand in Bg gives videantur.
Meister seems to be right in retaining cum eo quod, though his adoption of videmus for video is unnecessary, considering mirabor in the same sentence. Cum eo quod (see ad loc.) is defended by Günther (de Conj. Caus. apud Quint. usu: Halle, 1881, p. 24): he holds that it is more probable that video dropped out of the text than that it ‘in illo corrupto cumeo latet’ (Halm). Becher (Phil. Runds. I, n. 51: 1638) denied that ‘cum eo quod’ could mean ‘mit der Einschränkung dass,’ either in Cic. ad Att. vi. 1. 7 or anywhere in Quintilian. He found the necessary limitation in quodsi (‘wenn dagegen’: Cic. ad Fam. xii. 20) and supported Halm’s reading (which is also that of Par. 2. sec. m.), explaining the whole passage as follows: ‘Ich bin kein Freund des extemporierten Vortrages: wenn aber Geist und Wärme belebend wirkt, trifft es sich oft, dass der grösste Fleiss nicht den Erfolg eines extemporierten Vortrages erreichen kann.’ But in his latest paper (Programm des Gymnasiums zu Aurich) he advocates the reading and explanation adopted in the text.
§14. ut Cicero dictitabant. The reading is far from certain, but it seems best to adhere (with Halm) to the oldest MS., Bn, which is in agreement with N Sal. Ioan., Harl. 2662, 11671, and Dorv. The best alternative is ut Cicero dicit aiebant (C, Par. 1, also in margin of Harl. 4950: Bonnell-Meister): b H Bodl. and Burn. 243 give dicit agebant, which shows that the older codex from which b is derived probably had this reading, if indeed it is not a mistake for dictitabant. Bg gives dictabant: Harl. 4995 Goth. Voss. 2, Par. 2, sec. m. aiebant: Regius conjectured ut Cicero ait dictitabant: so ed. Camp, and Meister, cp. xii. 3. 11. For the inclusion of Cicero among the veteres cp. ix. 3. 1 ‘ut omnes veteres et Cicero praecipue.’
§16. tum intendendus. Krüger (3rd ed.) brackets tum (which is omitted in bHM) on the ground that this sentence does not contain, like the next (addit ad dicendum ...) a new thought, but rather (after the parentheses pectus est enim ... mentis, and ideoque imperitis ... non desunt) forms only a further development of what went before (omniaque de quibus dicturi erimus, personae ... recipienda): hence also the repetition of participles, habenda ... recipienda ... intendendus. H. 2662 gives tamen (and is here again in agreement with Guelf.).
addit ad dicendum, B: addiscendum (om. addit) bHFT. The loss of addit seems to have given rise to interpolation: M shows addit ad discendum stimulos habet et dicendorum expectata laus. Bonnell prints Ad dic. etiam pudor stim. habet et dic. exp. aus: so Vall. For the gerund used as subst. cp. pudenda xi. 1. 84: i. 8. 21: praefanda viii. 3. 45: desuescendis iii. 8. 70 and xii. 9. 17 num ex tempore dicendis inseri possit.
§17. pretium, all codd.: praemium Halm, following Regius.
§18. praecepimus, edd. vett, occurs in Harl. 4995 and Vall.2: other codd. praecipimus.
§19. cum ... sint consecuti bHM: cum ... sunt consecuti Bn Bg N. I cannot follow Becher in adopting the indicative here, as at 2 §6 (tradiderunt), where see note. Here cum is more or less causal: there it is antithetical. In point of form the two sentences are no doubt very much alike. Here the meaning seems to be ‘he who wishes to acquire extemporalis facilitas must consider it his duty to arrive at the point where..., seeing that many,’ &c.
Gertz put a full stop at tutior, and for cum read quin, holding that, on the traditional reading (i.e. with extemporalis facilitas as subject), potest would be expected instead of debet. This suggestion is adopted in Krüger’s third edition. H. J. Müller suggested Nam ... sunt consecuti.
§20. tanta esse umquam debet. This conj. of Herzog I find in the cod. Dorv., and receive it into the text; Halm and Krüger adopt Jeep’s tanta sit umquam. Bn Bg N Ioan. Harl. 2662 give tanta esse umquam fiducia: M has tantam esse umquam fiduciam: Vall. esse unquam tantam fid.: Harl. 4995 esse tantam unquam. Regius made the addition of velim after facilitatis: Becher thinks it may have dropped out before ut non. Meister follows: perhaps rather tantam velim (tm) esse unquam.
§22. consequi, Spald.: non sequi bH: sequi MC Harl. 4995, 4950: om. Bn, Bg, N Sal. Ioan. Harl. 4829. Becher would omit it, explaining utrumque non dabitur as ‘vim omnem et rebus et verbis intendere.’
§23. satis Krüger (3rd ed.) brackets, considering it to be the result of a dittography, and comparing what follows deinde ... aptabimus vela et disponemus rudentes. It seems however quite genuine.
§24. non labitur. Perhaps the most that can be said for this reading (which is that of Spalding, following earlier edd.) is that it is undoubtedly better than non capitur, which occurs in Bn Bg H Ioan. M and most codd., and is adopted by Halm and Meister. Capitur is explained in the Bonnell-Meister ed. by reference to such phrases as ‘altero oculo capi’ and ‘mens capta’ alongside of ‘mente captus’ in Livy: it is not ‘lamed’ or ‘weakened.’ This can hardly stand. Another reading is rapitur, which Halm thought might be right: but the notion of ‘snatching away’ seems too violent for the context, though appropriate enough in the passages quoted in support, vi. pr. §4 a certissimis rapta fatis, and Hor. Car. iv. 7. 8 quae rapit hora diem. Hild suggests animo (or mente) non labitur: Jeep non carpitur (cp. Sen. Nat. Quaest. 2. 13 totum potest excidere quod potest carpi): Becher non abit (cp. ix. 4. 14 abierit omnis vis, iucunditas, decor). The passage invites emendation: non cadit might stand alongside of Becher’s non abit, or such a future as servabitur or retinebitur could take the place of the negation, though we should then look for deperdet instead of deperdit.
non omnino B and codd.: omnino non Gesner, followed by Halm.
§25. est alia exercitatio, Harl. 2662 (Guelf.), 4995, 4950, 4829, 11671, Burn. 244, M, C, and so Krüger (3rd ed.): est illa BH Bodl. Burn. 243 Dorv.: est et illa Spalding Halm and Meister (cp. ix. 3. 35 est et illud repetendi genus, quod...).
utilior (Halm and Meister, following Spalding and ‘edd. vett.’) Vall.2, Harl. 4995: all other codd. utilitatis (Halm: ‘ex utilis magis?). In support of his proposal to read maioris utilitatis, Kiderlin (Blätter f. d. bayer. Gymn. 24, p. 90) compares ii. 4. 20 quod non simplicis utilitatis opus est: and xi. 1. 60 quod est sane summae difficultatis.
§26. quam illa: so all codd. Gertz quam in illa (sc. exercitatione), and so Meister. This is opposed by Becher (Bursian’s Jahresb. 1887, p. 49), ‘Zu componitur ist Subjekt exercitatio cogitandi totasque m. vel silentio (dum tamen ... ipsum) persequendi, d.h. dem Sinne nach tacita oratio, wie dum t. q. dicat i. s. i. zeigt, zu illa ist Subjekt vera oratio; componitur oratio aber ist nicht auffälliger als explicatur exercitatio.’
§27. ut Cicero ... tradit. Krüger (3rd ed.) follows Gertz in transferring this parenthesis to the end of the previous sentence, after ubique. Becher rejects it as a gloss.
aut legendum b M: om. BN Sal.: vel ad legendum Vall. Becher would omit it, on the ground that the whole chapter is concerned only with writing and speech, and even with writing only so far as it promotes the ‘facultas ex tempore dicendi.’
§28. innatans Stoer: unatrans BN Ioan. Sal.: inatrans bH: iura trans Harl. 2662: intrans FM Vall.2.
§29. an si, Meister (following ed. Camp.): ac si bHFT Burn. 243: an Bn Bg M.
debent, all codd.: debemus Krüger (3rd ed.) after Gertz. Either seems quite appropriate to the conditional use of the participle: ‘when men are debarred from both, they ought all the same,’ &c.
sic dicere. The grounds on which I base this emendation are stated in the note ad loc. Bn Bg HN and most codd. have inicere, which looks as if some copyist had stumbled over the repetition of the letters -ic in what I take to be the original text, whereupon the preceding tamen (or tam̅) would assist the transition to inicere. Cp. the omission of sic in most codd. in ut sic dixerim 2 §15. Halm (after Bursian) wrote id efficere, and so Meister. Other attempted emendations are vincere M, Harl. 4950, Burn. 244 Vall.2: tantum iniicere Harl. 4995: inniti or adniti edd.: id agere Badius: evincere Törnebladh.
§32. et in his: in his Halm and Meister: ne in his BN Ioan. HMC Dorv. Bodl.: ne in iis Harl. 2662: vel in iis Spald.: vel in his Bonnell and Krüger (3rd ed.). I venture on et, which seems to help the antithesis with in hoc genere above: v. ad loc.
velut summas ... conferre. So Bonnell (Lex. p. 139) Halm, Meister, Krüger (3rd ed.). The MSS. vary greatly: vel in summas in (sine bH: sive Harl. 4995) commentarium Bn Bg Dorv. Bodl. Harl 2662: velin summas et (suprascr. in) commentarium N: vel insinuamus sine commendarios M: commentarioram et capita Harl. 4950. Other conjectural emendations are velut in summas commentarium Spald.: mihi quae scr. velut in commentarium summas et c. conf. Zumpt: nec in his quae scrips. velim summas in commentarium et capita conferri Frotscher; vel in his quae scrips. rerum summas (cp. Liv. xl. 29. 11 lectis rerum summis) in commentarios conferre Jeep: ex iis quae scrips. res summas in commentarium et capita conferre, Zambaldi,—(on the ground that with conferre, ex his gives a better sense than in his). To these may perhaps be added et in his quae scrips. velut summas in commentariorum capita conferre.
In the Blätter f. d. bayer. Gymn. (1888) 24, pp. 90-91 Kiderlin discusses the whole passage. Keeping to the reading of the oldest MSS. (ne in his) he proposes ne in his quae scripserimus erremus: ‘damit wir nich bei dem Vortrage dessen, was wir geschrieben haben, den Faden verlieren’: cp. the use of errare xi. 2. 20 and 36. He rejects the various conjectures suggested above for vel in summas on the ground that it is impossible to explain ‘summas in commentarium et capita conferre.’ What is the meaning of ‘entering the chief points in a note-book and heads’ (‘den Hauptinhalt in ein Gedenkbuch und einzelne Hauptabschnitte einzutragen’—Bonnell-Meister)? Can the note-book and the ‘heads’ be conjoined in this way? You can make an entry in your notes, but not in ‘capita’: ‘in ein Gedenkbuch kann man eintragen, in Hauptabschnitte aber nicht.’ Baur’s version is excluded by the order of words: ‘den Hauptinhalt und die einzelnen Punkte in ein Gedenkbuch eintragen.’ Lindner’s is even less satisfactory: ‘welcher zufolge man auch von dem, was man geschrieben hat, den Hauptinhalt nach gewissen Hauptabschnitten eintragen soll.’
Kiderlin thinks the context shows that the essence of Laenas’s advice was to enter the chief points in a memorandum. This demands the elimination of the unmeaning et which wrongly conjoins commentarium and capita. Again as summa and caput are synonyms for ‘Hauptpunkt’ (cp. iii. 11. 27 and vi. 1. 2) one of the two may very well be a gloss: and the vel in vel in summas seems to show that these words were originally a marginal gloss to explain (in) capita. Kiderlin therefore proposes to transform the text as follows: ne in his quae scripserimus erremus [vel in summas] in commentarium capita conferre.
quod non simus, Regius, Frotscher, Becher, Meister, Krüger (3rd ed.): quod simus Bn Bg Ioan. M Dorv.: and so Halm: non simus bHT Bodl. In explanation of quod simus Spalding says ‘ubi satis fidere possumus memoriae ne scribendum quidem esse censeo’; and so Prof. Mayor (Analysis, p. 56), ‘We are even hampered by writing out at all what we intend to commit to memory: bound down to the written words, we are closed against sudden inspirations.’
hic quoque, Bn Bg and most codd.: hoc quoque Harl. 4995: id quoque bHM.