Hostname: page-component-586b7cd67f-r5fsc Total loading time: 0 Render date: 2024-11-30T19:59:42.408Z Has data issue: false hasContentIssue false
  • English
  • Français

The Annotations of M. Valerivs Probvs, III: some Virgilian Scholia*

Published online by Cambridge University Press:  11 February 2009

H. D. Jocelyn
H. D. Jocelyn
Affiliation:
University of Manchester
Get access Rights & Permissions [Opens in a new window]

Extract

Most of the commentaries on Greek authors which circulated in the towns of Egypt during the late Ptolemaic and early Imperial periods ignored the critical and colometrical problems which had engaged the attention of the great Alexandrian grammarians. A few, however, based themselves on texts equipped with signs, included the signs in their lemmata and offered explanations. Such commentaries must be the source of the scattered references to signs in the older marginal scholia in Byzantine manuscripts of Homer, Hesiod, Pindar and the Attic dramatists. The only Byzantine manuscript to transmit a pagan text still equipped with a large number and a variety of signs, namely cod. Venice, Bibl. Marc. gr. 454, is also the only one to transmit scholia with lemmata retaining prefixed signs. Just as texts and scholium-lemmata lost their signs in the course of transmission, so too did references to signs within scholia either disappear or become garbled. At best, a statement about the reasons for affixing a sign would turn into one about the content or style of the verse in question. The few mentions made of the great Alexandrians give no cause for thinking that we ever have a verbatim quotation of an ὑπόμνημα written by one of them in order to explain his own signs. Time and again it is demonstrable that an explanation of a sign's presence against a particular verse goes back to some writer like Aristonicus.

What survives of the ancient discussion of Latin literature is exiguous in comparison with the Greek material.

Type
Research Article
Information
The Classical Quarterly , Volume 35 , Issue 2 , December 1985 , pp. 466 - 474
Copyright
Copyright © The Classical Association 1985

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

160 See P. London, Brit. Libr. Pap. 2055 (commentary on Hom. Il. 2.751–827 from first century b.c., published by A. S. Hunt, P. Oxy. 8 (1911), no. 1086; = Erbse, , Schol. Il. 1.164–74Google Scholar), P. Cairo, Journal no. 60566 (commentary on Horn. Il. 6.236, 252–85 from second century a.d.; published by Waddell, W. G., Mélanges Maspero (Cairo, 1934–1937 [Institut français d'Archéologie Orientale du Caire, Mémoire 67]), ii. 148–51Google Scholar; = Erbse, , Schol. Il. ii. 124–8Google Scholar), P. Vienna, Oest. Nationalbibl. G. 29780 (commentary on Aristoph. Pax 410–15, 457–66 from fifth century a.d., first published by Gerstinger, H., Mitteil. aus d. Papyrussamml. der Nationalbibl. i. Wien [Vienna, 1932], i. 167Google Scholar, identified by Gronewald, M., ZPE 45 [1982], 64–9Google Scholar).

161 Very few traces of ancient symbols survive in medieval manuscripts; see above, nn. 79–80. For the signs in the lemmata to the scholia in the Venice manuscript see Dindorf, op. cit. (n. 57), pp. xx–xxiii.

162 Noteworthy in P. London, Brit. Libr. Pap. 2055 are the formulae διπλ⋯ ὅτι (col. ii. 55), τ⋯ σημεῖον πρ⋯ς…ὅτι (coll. i. 28, ii. 47, iii. 98), ⋯θετεῖ τούτους ⋯ δεῖνα ὅτι (col. ii. 63).

163 On the prehistory of the references to signs in extant scholia see the works cited in nn. 98, 115, Schrader, H., De notatione critica a ueteribus grammaticis in poetis scaenicis adhibita (Diss. Bonn, 1864)Google Scholar, Lehrs, K., Die Pindarscholien (Leipzig, 1873), pp. 104–11Google Scholar, von Wilamowitz-Moellendorff, U., ‘De Rhesi scholiis disputatiuncula’, Ind. scholl. Greifswald, 1877–8, 10–12 (= Kleine Schriften [Berlin, 1935], i. 9–13Google Scholar), Roemer, A., ‘Die Notation der alexandrinischen Philologen bei griechischen Dramatikern’, Abh. bayer. Ak. Phil.-hist. Kl. 19 (1892), 627–82 (661–5)Google Scholar, Deas, H. T., ‘The Scholia Vetera to Pindar’, HSCPh 42 (1931), 1–78 (3–4, 30, 72–6)Google Scholar, Irigoin, J., Histoire du texte de Pindare (Paris, 1952), pp. 51, 64, 104Google Scholar.

164 Gramm. Lat. vii. 534.5–6.

165 Jerome, knew of a commentary on the De rerum natura (Adu. Rufin. 1.16)Google Scholar.

166 Cf. Brink, op. cit. (n. 4), p. 40.

167 See above, n. 33.

168 Seneca's one citation of Terence, (Haut. 77 at Epist. 95.53)Google Scholar may be at second hand (see Jocelyn, H. D., Antichthon 7 [1973], 42–4Google Scholar). The same philosophical theme is handled at Persius 5.161–9 as at Hor. Serm. 2.3.259–71; significantly, Horace refers to Terence's Eunuchus and Persius to the Menandrian original.

169 Gramm. 24.4.

170 See Serv. Aen. 7.773, 8.406, Serv. Dan. Georg. 1.277, Aen. 1.44. Cf. Serv. Aen. 1.441 (secundum Probum), 10.539 (Probus dicit legendum). The same formula is applied to Donatus at Serv. Aen. 12.365, Buc. 3.38. Behind it lie injunctions like lege (Donat. Ter. Andr. 167–8, Serv. Dan. Georg. 3.187) and legendum est (Serv. Aen. 1.2 et al.).

171 See Donat. Ter. Andr. 720, Eun. 46. Cf. Serv. Dan. Aen. 10.173 (Probus subdistingui uult). For the underlying classroom injunction see Serv. Dan. Aen. 1.133 (distingue), 5.404 (hic distinguendum), 8.381 (subdistingue). The correlative iunctim legere (Serv. Dan. Aen. 9.219) suggests that the commentators had in mind oral delivery rather than punctuation as such.

172 See Donat. Ter. Andr. 875, Phorm. 372, 1005, Serv. Aen. 6.473, 782, 10.18, Serv. Dan. Aen. 3.83. Cf. Serv. Aen. 6.177 (Probus…et Donatus de hoc loco requirendum…dixerunt). For the direct putting of a query cf. Serv. Dan. Aen. 4.696.

173 See Serv. Dan. Aen. 12.174, Philarg. Georg. 1.224. For Probus ait see Serv. Aen. 3.3, Serv. Dan. Aen. 1.1, 4.359, 9.811, Philarg. Georg. 1.403, 4.134. For Probus…inquit see Donat. Ter. Phorm. 155. The same formula is applied regularly by Servius to Urbanus (Aen. 5.493) and to Donatus (Aen. 6.230)

174 See Macrob. Sat. 5.22.9. For the classroom injunction nota see Serv. Aen. 2.288; for notandum Serv. Aen. 1.147 et al.

175 See schol. Ver. Virg. Aen. 9.369, Donat. Ter. Phorm. 49. For the injunction adnotandum see Donat. Ter. Eun. 971, Serv. Dan. Virg. Aen. 11.211, Georg. 3.351, schol. Ver. Virg. Buc. 3.30, Aen. 9.402 (Asper), 10.63, 10.564, Porph. Hor. Carm. 2.11.16–17 et al.

176 Pressed literally, Probus notat and Probus adnotauit suggest the note-taking student. Probus putat (schol. Ver. Virg. Aen. 2.173, Serv. Dan. Aen. 10.303), Probus uult (Serv. Buc. 6.76, Serv. Dan. Aen. 10.173, 11.830), Probo displicet (Serv. Dan. Aen. 2.173), secundum Probum (Serv. Aen. 1.441, 10.33), sic Probus (Donat. Ter. Hec. 2), Probus tradit (Serv. Dan. Aen. 10.182) are of course completely ambiguous.

177 Cf. Jahn, op. cit. (n. 8), pp. cxl-cl (admitting, however, that the In Buccolica et Georgica commentariolum had suffered abbreviation and addition). At op. cit. (n. 4), pp. 207–8, Zetzel seems to derive Serv. Dan. Aen. 1.21–2 from a line by line commentary composed by Probus; at Medievalia et Humanistica n.s. 10 (1981), 20Google Scholar, he distinguishes a little more clearly between the work of Probus and ‘variorum’ commentaries like that of Donatus.

178 Aistermann, op. cit. (n. 4), p. i, attempted to soften the implications of Suetonius' account by conjecturing uiuus for nimis.

179 In Iuuenalis Satyras commentarii (Venice, 1486)Google Scholar. On Valla and ‘Probus’ see Anderson, W. S., Traditio 21 (1965), 383–424Google Scholar, Bartalucci, A., SIFC 45 (1973), 233–57Google Scholar. The source of Valla's knowledge can now be taken back with some plausibility as far as ninth-century Brescia (on the basis of the hexameter poem appended to a commentary on Terence on fol. 144r of cod. Munich, Bayer. Staatsbibl. lat. 14420; see Billanovich, G., IMU 17 [1974], 43–60, 22 [1979], 367–95Google Scholar). Cod. Vatican City, Bibl. Apostol. Urbin. lat. 661, fol. 62r, has ‘Probus’ applied in a contemporary (eleventh century) hand to a set of Carolingian scholia (see Sanford, E. M. ap. Kristeller, P. O., Catalogus Translationum et Commentariorum [Washington, 1960], i. 184Google Scholar).

180 Cf. Poliziano, A., Miscellaneorum centuria prima (Florence, 1489), cap. xxxiiiGoogle Scholar.

181 See above, n. 34. For continuing doubts about the Probian origin of the biography see Scarcia, R., RCCM 6 (1964), 298–302Google Scholar, Robathan, D. M. and Cranz, F. E., Catalogus Translationum et Commentariorum (Washington, 1967), iii. 204Google Scholar.

182 G. Parrasio printed at Milan in 1504 the material in foll. 76r–95v (lacking a title) of what is now cod. Naples, Bibl. Naz. Lat. 2 as the ‘Probi grammatici Instituta Artium’ and at Vicenza in 1509 the material in foil. 95v–111v (titled, fol. 95v, DE CATHOLICIS PROBI; fol. 111v, ARS PROBI…CATHOLICA) of the same codex as the ‘Probi Catholics’. On the fifth-century origin of the folia see CLA iii. 397a. Freund, W., NJbb 5 (1832), 90–5Google Scholar, showed that the first work could not be the one cited in the second as ‘Instituta artium’. Mai, A. printed in Rome in 1833 (Classicorum Auctorum e Vaticanis Codicibus Editorum Tomus V, pp. xxxiv–xli, 153–328)Google Scholar the untitled material in the sixth-century Vatican codex, Urb. lat. 1154 (CLA i. 117). von Eichenfeld, J. and Endlicher, S., Analecta Grammatica maximam partem anecdota (Vienna, 1837), pp. xi–xiiGoogle Scholar, pointed out the existence of the same material in a Vienna codex (now Naples, Bibl. Naz. lat. 1, foll. 17–49, lacking a title [CLA iii. 388, written in the seventh or eighth century at Bobbio]) and in two Paris codices (Bibl. Nat. lat. 7494, foll. 124–218, titled TRACTATVS PROBI GRAMMATICI [written in the ninth century]; 7519, titled Probi Grammatici de octo orationis partibus [written in the fifteenth century]). Lersch, L., Zeitschr. f.d. Alt. n.s. 1 (1843), 633Google Scholar, proved that Priscian knew this work as the Instituta artium of Probus. Von Eichenfeld and Endlicher printed from foll. 8–10 of the same Vienna codex a fourth lot of material titled VALERII PROBI DE NOMINE and from foll. 49–52 a fifth lot which lacks a title but which coincides in part with the DIFFERENTIAE PROBI VALERII of cod. Montpellier 306, fol. 68. The latter they called an ‘appendix ad Probi Artem Minorem’.

183 See the works cited in n. 39 and Steup, op. cit. (n. 4), pp. 138–201. The notion is enshrined in the Realencyclopädie (23.1 [1957], 59–64Google Scholar, s.v. ‘Probus. 26’).

184 Keil, H. published the four works in the fourth volume of his Grammatici Latini (Leipzig, 1864)Google Scholar under the name of Probus. For Keil's views on the process of redaction see the preface to this volume, p. xxx, and ‘De M. Valerio Probo grammatico’ in Symbola Philologorum Bonnensium in honorem Friderici Ritschelii collecta, fasc. prior (Leipzig, 1864), 91–100Google Scholar. There has not been a great deal of discussion of the problem since Keil's day; see, apart from Steup's book, Beck, J. W., De M. Valerio Probo Berytio quaestiones nouae (Groningen, 1886)Google Scholar, Fröhde, O., ‘Valeri Probi de nomine libellum Plinii Secundi doctrinam continere demonstratur’, NJbb Suppl. 19 (1893), 159–203Google Scholar, Jeep, L., Zur Geschichte der Lehre von den Redetheilen (Leipzig, 1893), pp. 73–82Google Scholar, Ihm, M., RhM 52 (1897), 633Google Scholar, Wessner, P., Bursians Jahresberichte 113 (1903), 166–9Google Scholar, Barwick, K., Hermes 54 (1919), 409–22Google Scholar, Helm, R., RE 23.1 (1957), 59–64Google Scholar, s.v. ‘Probus. 26’, Robson, C. A., MA 69 (1963), 37–54Google Scholar, Casa, A. Della, ‘La “Grammatica” di Valerio Probo’, in Argentea Aetas: in memoriam E. V. Marmorale (Genoa, 1973 [Istituto di Filol. Class. e Med. dell'Univ. di Genova]), 139–60Google Scholar.

185 On the make-up of cod. Naples, Bibl. Naz. lat. 2 see De Nonno, M., La grammatica dell' Anonymus Bobiensis (Rome, 1982), pp. xvii–xxviiiGoogle Scholar.

186 With Servius' notes on Aen. 1.194, 2.15 compare Gramm. Lat. iv. 36. 25–8, 16.35–17.2. Cledonius, , Gramm. Lat. v. 45.17–19Google Scholar (Probus in arte quam de regulis scripsit) seems to relate to Gramm. Lat. iv. 17.4–5; Priscian, , Gramm. Lat. ii. 218.22–3Google Scholar (Probus in libro qui est de catholicis nominum) to Gramm. Lat. iv. 28.23–4.

187 Cf. Riese, op. cit. (n. 42), p. 28, Kübler, op. cit. (n. 43), p. 30, Thilo, G., NJbb 149 (1894), 289–304, 421–32Google Scholar, Norden, E., RhM 61 (1906), 171–7Google Scholar (= Kleine Schriften [Berlin, 1964]), pp. 443–8Google Scholar; for the view that the commentariolum has a Probian core see Jahn, op. cit. (n. 8), p. cxli, Ribbeck, op. cit. (n. 80), pp. 163–4, Körtge, G., ‘In Suetonii de uiris illustribus libros inquisitionum capita tria’, Diss. philol. Hal. 14.3 (1900), 187–284 (235–8)Google Scholar, Wessner, P., BPhW 20 (1900), 878Google Scholar, Bursians Jahresberichte 113 (1903), 201–4Google Scholar, Marx, F., C. Lucilii Carminum Reliquiae (Leipzig, 1904), i lxxii–lxxvGoogle Scholar, Aistermann, op. cit. (n. 4), pp. 72–80. The issue has been left to one side for many years; cf., most recently, Lehnus, L., ‘Verso una nova edizione del commento Virgiliano attribuito a Probo. La Vita Vergilii’ in Scripta Philologa (Milan, 1982), iii. 179–211Google Scholar.

188 Serv. Aen. 6.177 (Probus tamen et Donatus…dixerunt), 7.543 (Probus, Asper, Donatus dicunt), 8.406 (Probus uero et Carminius…legunt) point in the same direction; likewise Serv. Aen. 6 praef. (licet…Probus et alii…reliquerint), 6.473 (quaerit et Probus et alii), 7.773 (alii…legunt, ut Probus), Serv. Dan. Aen. 1.44 (Probus et ‘tempore’ legit…qui ‘tempore’ legunt…), 11.830 (alii…legunt. Probus…). Probus' was a famous name which Servius and the compiler of the additions tended to pick out of lists of older students.

189 This is not necessarily so (despite Jahn, op. cit. [n. 8], p. cxl) in the case of Donat. Ter. Andr. 720, Eun. 46, Ad. 323.

190 I note in the Ps. Acro scholia to Horace: Serm. 1.6.95 ‘parentum’ inuenio; 1.8.19 in aliis ‘uexant’ lego; 2.3.126 in aliis ‘inpexum’; Epist. 1.7.69 in alio ‘praeuidisset’ legitur; in the commentum Donati on Terence: Andr. 236 in aliis ‘factu aut inceptu’ fuit; 599 in aliis ‘idem’ scriptum est; 978 ut in plurimis exemplaribus bonis non inferantur; Eun. 312 in quibusdam omnino non legitur; Ad. 511 hi sex uersus in quibusdam non feruntur; 601 uersus…quos multa exemplaria non habent; 706 hic uersus in quibusdam non inuenitur; Hec. 665 in ueteribus codicibus sic est; in the Berne scholia to Virgil: Georg. 4.26 in Ebrii ‘coice’;77 in Ebrii ‘nanctae’, non ‘nactae'; 87 in Corneliani ‘quiescunt’, non ‘quiescent’; 120 ‘intyba’ in Ebrii, ‘intuba’ in Corneliani; 131 in Ebrii ‘seram’; 169 in Ebrii ‘feruit’; 175 ‘forcipe’ in Ebrii et ‘forfice’ in Corneliani; 545 in Ebrii ‘Orphi’; 564 in Ebrii ‘Parthinope’; in Servius' commentary: Aen. 5.871 in non nullis antiquis codicibus…est; 6 praef. licet primos duos uersus Probus et alii in quinti reliquerint fine; 7.543 difficile in exemplaribus inuenitur; 7.568 quamquam antiqui codices habeant…antiqui codices…habent; 9.607 in aliis; 10.444…huic loco Probus alogum posuerit; in the scholia added to this commentary: Buc. 7.64 in Hebri; 7.65 in Vari et in Hebri; Georg. 1.12 antiquissimi libri ‘fudit aquam’ habuerunt…in Corn. ‘equm’, in authentico ‘aquam’, ipsius manu ‘equm’ (cf. Georg. 1.6 emendauit ipse; 1.64 induxit et reposuit; 1.66 ipsius manu adiectum; 1.67 ipsius manu adiecti sunt deletis duobus; 2.224 emendauit ipse; 4.141 ipsius autem manu duplex fuit scriptura; Aen. 3.204 hi uersus circumducti inuenti dicuntur et extra paginam in mundo; 3.226 sed sane hic uersus qui circumductus est tabs auditur); Aen. 1.21 in Probi adpuncti sunt, 2.37 antiqua tamen exemplaria ‘ue’ habere inueniuntur; 2.775 et hic uersus in plerisque dicitur non fuisse; 3.153 hic uersus in multis non inuenitur; 4.348 sane quidam in nouis et emendatis libris pro ‘detinet’ ‘demerer’ inuentum adserunt; 5.871 <in> Corneliani[s] et in Hebri inuenies (see n. 192); 9.36 in omnibus bonis…inuentum; 11.142 multa…exemplaria…habuerunt.

191 Priscian does not state his source at iii. 162.11–163. 3 — …sed melius in quibusdam codicibus sine m pharetra ablatiuus inuenitur… sed melius hic quoque in quibusdam inuenitur ‘quod scelus…aut Calydone merente’ …sed inuenitur etiam “iustitiaene prius mirer” –, but it must have been some commentary rather fuller and more specific than Servius' (contrast with Priscian Servius' notes on Aen. 1.323, 7.307, 11.126).

192 HSCPh 43 (1932), 109Google Scholar. To judge by the commentary based on his lectures (published in 1487), Pomponius Laetus had read this manuscript. For other examples of the obliteration of singularities in the scholiastic tradition see Timpanaro, S., Stud. Urb. Ser. B 31 (1957), 166–7Google Scholar (= [with additions] op. cit. [n. 135], pp. 448–50).

193 Op. cit. (n. 80), p. 152. Cf. Riese, op. cit. (n. 42), p. 4 n., Gudeman, ‘Krit. Zeich.’ (n. 52), 1926.

194 On second thoughts Riese, , NJbb 93 (1866), 467 n. 3, 472, 871Google Scholar, suggested the antisigma cum puncto. Aistermann's notion (op. cit. [n. 4], p. 65) that it was a question of the asteriscus increases the difficulty. Steup, op. cit. (n. 9), pp. 86–90, argued for a radical rewriting: in Probi his duobus uersibus obeli appuncti sunt.

195 See Part II 161 n. 156.

196 On the Greek στιγμή see above, n. 101.

197 Aen. 5.871.

198 Nothing should be deduced from Servius' use of the term codex (rather than liber or exemplar). He could have been translating something from an old commentary into terms readily comprehensible in the fifth century.

199 For the coronis in non-dramatic poetry see Part II 156. The formula reliquit aliquis is normally used in relation to the original author's autograph (cf. Hyginus ap. Gell. 1.21.1, ‘Probus’ ap. Gell. 1.15.18, Serv. Virg. Aen. 7.464, 12.120, Serv. Dan. Aen. 6.289).

200 For the belief, backed by what looked like hard manuscript evidence, that Varius and Tucca did not publish verbatim the material left by Virgil see Nisus ap. Donat. Vit. Verg. 42, Servius, , Aen. 1Google Scholarpraef., 7. 464, 12.120 (citing Caper and Hyginus), Serv. Dan. Aen. 2.566, 6.289.

201 Both Leo, , Pl. Forsch. 2 (n. 4), pp. 42–3Google Scholar, and Büchner, op. cit. (n. 4), p. 394 (cf. RE II 8.2 [1958], 1469–70Google Scholar), have Probus rejecting texts of the sort cited by Seneca at Epist. 94.28 in favour of the one issued by Varius and Tucca. Goold, G. P., HSCPh 74 (1968), 125Google Scholar, extracts from Servius' two notes the notion that the division between books 5 and 6 made by Varius was what Virgil intended and that Probus conjectured a different arrangement (cf. Zetzel, op. cit. [n. 4], pp. 52, 272 n. 19).

202 For the distinction see Varr. Ling. 9.106 Plauti aut librarii mendum si est, Ascon. p. 60. 7–8 Stangl inducor magis librariorum hoc loco esse mendam quam ut Ciceronem parum proprio uerbo usum esse credam.

203 For satis licenter cf. the notes on Aen. 5.326, 8.268, Georg. 3.443; for reports of criticism by Probus of Virgil's latinity cf. those on Aen. 2.173, 3.83, 4.359, 9.811 (referring to 5.432). Since Burman (op. cit. [n. 8], vol. iii. 534) recognised the reference to an alogus and suggested that the scholium be read either as…ut Probus huic loco corrupto alogum apposuerit or…ut Probus hoc loco corrupto alogum posuerit, some have kept the reference to corruption (cf. Bergk, op. cit. [n. 57], 110 [= Kl. phil. Schr. i. 581], Ribbeck, op. cit. [n. 80], pp. 151–2, Steup, op. cit. [n. 9] p. 85). G. Thilo (Leipzig, 1884) rightly got rid of it (ut huic loco Probus alogum adposuerit); likewise Aistermann, op. cit. (n. 4), p. xiii.

204 The first Paris list has Aristarchus and others marking with the obelus ‘bad’ verses as well as non-Homeric verses (Gramm. Lat. vii. 534.13–16).

205 See above, n. 19.

206 It is doubtless true that nunc magis est at Aen. 10.481 results from conjecture (cf. Timpanaro, op. cit. [n. 135], p. 184 n. 50), but Servius' legitur et ‘nunc magis est’ can hardly mean ‘“nunc magis est” is also conjectured’. Likewise audiam at Ter. Andr. 592 and paucis at Hec. 58 may be conjectures based on the original Greek (cf. Wessner, , Aemilius Asper [n. 3], p. 28Google Scholar), but Donatus' legitur et and quidam legunt refer merely to what has been seen in manuscripts.

207 See op. cit. (n. 69). Adnotandum cannot stand; the gerund is regularly constructed either with an accusative and infinitive (Ascon. p. 58, 27, Serv. Dan. Virg. Georg. 3.351, Porph. Hor. Carm. 2.11.16–17, 4.5.18, Serm. 1.3.1) or with a clause introduced by quod (Donat. Ter. Eun. 971, schol. Ver. Virg. Buc. 3.30, Aen. 10.564, Serv. Dan. Aen. 11.211, Porph. Hor. Epod. 2.54, Serm. 2.1.17, 2.3.187, 2.3.316; at Porph. Hor. Serm. 1.10.18 quod is easily supplied).

208 It is impossible to tell with certainty whether sed Vergilius amat aliud agens exire in laudes populi Romani is meant to belong to Probus' note or to be a criticism thereof.

209 Cf. Donat. Phorm. 49 hoc annotauit Probus, schol. Ver. Virg. Aen. 9.369 adnotant Probus et Sulpicius.

210 See above, n. 42 (especially the discussion of Steup).

211 Servius seems not to use adnotare at all.

212 Zetzel, op. cit. (n. 4), pp. 48–9, 207–8, makes the scholium wholly dependent on a commentary by Probus which prefixed the lemmata with signs. In ‘postilla’ to his original article (n. 4; see Studi di letteratura Latina imperiale, pp. 214–21) Scivoletto argues that the author of the statement in the scholium (Donatus) based it on a schedule of notae with brief explanations illustrated from the Aeneid by Probus himself.

213 Cf., however, schol. Ver. Virg. Aen. 9.373 Prob. hic posuit aptissimum hoc exemplum ex Horatio: ‘node sublustri nihil antra praeter uidit et undas’.

214 See above, pp. 470–1, on ‘hinc…parcas’ in Probi adpuncti sunt.

215 Et adnotatum could be regarded as co-ordinate with the earlier et ‘late regem’ …et ‘hinc’ rather than as linked closely with adpuncti sunt. Not, however, I think plausibly. Scivoletto, op. cit. (n. 212), p. 220 n. 74, defends the transmitted adnotandum as coming from Donatus.