¹ Contrasting views on extent of rhythmic prose in Imperial Greek: E. Norden, Die antike Kunstprosa vom VI. Jahrhundert v. Chr. bis in die Zeit der Renaissance, 2 vols. (Leipzig, 1898), 2.918-22; cf. e.g. 1.423-7; U. von Wilamowitz-Moellendorff, ‘Asianismus und Atticismus’ [1900], Kleine Schriften 3 (Berlin, 1969), 223-73, at 257-9; R.L. Hunter, A Study of Daphnis & Chloe (Cambridge, 1983), 85 (though ‘rhythmical’ is not used as narrowly as ‘rhythmic’ here); M. Winterbottom, ‘On ancient prose rhythm: the story of the dichoreus’, in D. Obbink and R.B. Rutherford (edd.), Culture in Pieces: Essays on Ancient Texts in Honour of Peter Parsons (Oxford, 2011), 262–76, at 264. Hunter at 84-5 on Longus and Winterbottom at 263-5 on developments in Greek are among the most significant contributions in relatively recent times. Earlier, an especially important general treatment is the work of A.W. de Groot: A Handbook of Antique Prose-Rhythm I (Groningen, The Hague, Leipzig, 1919); Der antike Prosarhythmus I. Zugleich Fortsetzung des Handbook of Antique Prose-Rhythm (Groningen, 1921); La Prose métrique des anciens (Collection d’études latines 2) (Paris, 1926). Particularly valuable as a treatment of a single author (though he includes comments on others) is St. Heibges, De clausulis Charitoneis (Diss. Münster; Halle, 1911), supervised by W. Kroll. On the accentual system which begins in the fourth century see W. Hörandner, Der Prosarhythmus in der rhetorischen Literatur der Byzantiner (Wiener Byzantinische Studien 16) (Vienna, 1981), esp. 51-71; C. Klock, Untersuchungen zu Stil und Rhythmus bei Gregor von Nyssa. Ein Beitrag zum Rhetorikverständnis der griechischen Väter (Beiträge zur klassischen Philologie 173) (Frankfurt am Main, 1987), 219-300.

² Some typical comments on Appian's style: E.L. Bowie, ‘Appian’, in P.E. Easterling and B.M.W. Knox (edd.), The Cambridge History of Classical Literature I Greek Literature (Cambridge, 1985), 707-9, at 709: ‘stylistically undistinguished, but … far from unreadable’; similarly McGing, B.C., ‘Appian's “Mithridateios”’, ANRW 2.34.1 (1993), 496-522 Google Scholar, at 520; Weißenberger, M., ‘Das Imperium Romanum in den Proömien dreier griechischer Historiker: Polybios, Dionysios von Halikarnassos und Appian’, RhM 145 (2002), 262-81Google Scholar, at 271, ‘der für Appians Stil ungewöhnliche Aufwand an stilistischem Schmuck … (Trikolon, Parallelismus, Homoioteleuton)’ (on Proem 41; this certainly gives a misleading impression: for the superlatives cf. e.g. Hisp. 318, B Civ. 2.621, for the three features e.g. Hisp. 197, Hann. 102). More responsive to Appian's writing is an article by a scholar who works especially on Latin poetry: Henderson, J., ‘Three men in a vote: proscription and the power of the text (Appian, Bellum Civile 4.1.1-6.51)’, Histos 1 (1997), 93-112 Google Scholar. Cf. also Potz, E., ‘ Ficta, non facta dicere – und trotzdem die Wahrheit berichten. Caesar, Bellum civile 1, 43-87 und Appian, Emphylia 2, 42 f.’, Grazer Beiträge 21 (1995), 85-94 Google Scholar, at 89 and 91-2; and already H. Strebel, Wertung und Wirkung des Thukydideischen Geschichtswerkes in der griechisch-römischen Literatur. (Eine literargeschichtliche Studie nebst einem Exkurs über Appian als Nachahmer des Thukydides) (Diss. Munich; Speyer am Rhein, 1935), 92: ‘Appian … kann wenigstens nach der formalen Seite keineswegs als reiner Compilator angesehen werden, der hinter seinen Quellen verschwindet’ (73-92 for Appian and Thucydides). On shape and thought in Appian's work, see n. 23 below; an early aside in the right direction: J.P.V.D. Balsdon, review of P. Meloni, Il valore storico e le fonti del libro macedonico di Appiano (Rome, 1955), JRS 46 (1956), 199-201, at 200: ‘Appian (who is commonly assumed to have had no mind of his own)’.

³ On prose before Hegesias note recently Usher, S., ‘Eurhythmia in Isocrates’, CQ 60 (2010), 82-95 CrossRefGoogle Scholar. It seems reasonable to accept   ̄  ̆  ̄  ̆  ̄  as part of the rhythmic system, though it is less common than the other three endings: it is strongly supported by Cicero's use of atque + consonant to obtain it (cf. Hutchinson, G.O., ‘Rhythm, style, and meaning in Cicero's prose’, CQ 45 [1995], 485-99CrossRefGoogle Scholar, at 485-6). Longin. 39.4 in fact suggests its place in the system (cf. D.A. Russell, ‘Longinus’ On the Sublime: Edited with Introduction and Commentary [Oxford, 1964], 175); Quint. Inst. 9.4.107-8 quotes it from Cicero (Phil. 2.63) and finds it acceptable in delivery, despite anxieties about poetry. Even De Groot (n. 1 [1921]), 66 notes that it occurs twice in the inscription of Antiochus I of Commagene and is somewhat more common in Cicero than in unrhythmic prose; he grudgingly thinks it permitted in rhythmic prose if not actually sought. The overall case would not be affected if the ending were accounted unrhythmic.

⁴ For the changes in Latin and Proto-Italic see A.L. Sihler, New Comparative Grammar of Greek and Latin (New York and Oxford, 1995), 68-70, 75-80, 83-5.

⁵ Heibges (n. 1), 57-9 and Hunter (n. 1), 126 n. 1 adopt a different approach to mute and liquid; for our purposes it is best to be austere, and limit the possibilities for rhythmic endings. For reasons to be explained elsewhere I have not gone so far as to suppose in Appian Attic lengthening before initial rho; equality has required the same treatment in the earlier texts which are used to provide comparative data. But the issue is of no significance for the argument: thus it affects none of the sentence-endings in B Civ. 5 or Lysias 1-7 and 9-10 (used in § II below). The name Ἀπουλήιος (B Civ. 1.125, etc.) exemplifies two problems in Latin names: does ου capture the quality of the Latin vowel but not form a long syllable? does -ηι (despite editors’ accentuation and non-subscript iota) capture Latin -ēi-, including the consonantal i? A metrical inscription (IGUR III 1166.4-5) suggests that the name Πομπηιος, so common in Appian, was pronounced as three syllables. On Appian's knowledge of Latin, cf. É. Famerie, Le Latin et le grec d'Appien. Contribution à l’étude du lexique d'un historien grec de Rome (École Pratique des Hautes Études, IV^e Section, III Hautes études du monde gréco-romain 24) (Geneva, 1998), 27-32. Note that Appian is probably not a Latin name: Hahn, I., ‘Papyrologisches zum Namen Appians’, Philologus 117 (1973), 97-101 CrossRefGoogle Scholar.

⁶ Note that Klock (n. 1), 230 treats high point before direct speech as weak punctuation. To avoid suspicion of choice or change on rhythmical grounds, Mendelssohn-Viereck's text of B Civ. has been followed throughout (but not always in punctuation). Problems of text rarely affect rhythm at the end of paragraphs or sentences; but one should note e.g. 1.331, where θέαν ἀθέμιστον. had already aroused disquiet on stylistic grounds, after ἐπὶ τοῖς ἔργοις ἐς ὄψεις ἐτρέποντο ἀθεμίστους (-γοῑς ἐ̆ς ὄ̄ψεῑς |, -ο̄ντο̆ ἀ̆θε̆μί̄στοῡς |; see n. 10 below and § III). Deletion of ἀθέμιστον would give a rhythmic ending. If Appian is indeed rhythmic, textual criticism will need to bear it in mind. Thus P. Goukowsky, ‘Sur les funérailles de Pompée’, in C. Brixhe (ed.), Hellènika symmikta. Histoire, linguistique, épigraphie 2 (Paris, 1995), 55-61, at 58-9 brings forward for 2.361 the variant ἄλλως ἐπέγραψεν from Vat. Gr. 2156 (a.d. 1450); but ἄλλος gives better rhythm (and sense), on my view of mute and liquid. In considering the authorship of the new fragment Goukowsky ascribes to Appian, one would have to see if it looked rhythmic (it does) (‘Trois nouveaux extraits d'Appien’, ibid. 63-70, at 65-8; cf. Hisp. 8). (More on the contents of the MS: Amerio, M.L., ‘Ancora sui nuovi frammenti di Appiano’, Invigilata Lucernis 21 [1999], 35-42 Google Scholar.)

⁷ For a table see M.L. Samuels and J.A. Witmer, Statistics for the Life Sciences (Upper Saddle River, NJ, and London, 2003³), 686 (appendix of tables omitted from fourth edition). For the type of χ² test which compares two unknown distributions to ascertain if they are different, see W.H. Press, S.A. Teukolsky, W.T. Vetterling, B.P. Flannery, Numerical Recipes in C: The Art of Scientific Computing (Cambridge, 1992²), 620-3, esp. 622.

⁸ Out of the non-rhythmic endings in the 400 sentences of Appian, 60 were forms of   ̄  ̄  ̄  ̄ , 12 were forms of    ̄  ̆  ̄  ̄    ̄  ̆  ̄ ; of those in the twenty authors, the figures were 138 and 13.

⁹ Pherecydes of Athens was originally one of the twenty, but later banished, since it was feared so many dactylic names would produce an abnormal specimen of non-rhythmic prose. Lycurgus took his place.

¹⁰ See R.G.M. Nisbet, ‘Cola and clausulae in Cicero's speeches’ [1990], in S.J. Harrison (ed.), Collected Papers on Latin Literature (Oxford, 1995), 312-24, at 319-22 for atque and at 316-18 for the occurrence of hiatus after clausulae; Hutchinson (n. 3), 486-90 and 496 for atque. Hiatus in Appian is discussed by A. Zerdik, Quaestiones Appianeae (Diss. Kiel, 1886), 49-82: a useful collection of material, though the significance of rhythm is not glimpsed (partly a matter of date, cf. F. Novotný, État actuel des études sur le rythme de la prose latine [Eos Suppl. 5] [Lwów, 1929], 2-33 for history of study); contrast Heibges (n. 1), 54-6, though he does not distinguish between common words and others.

¹¹ Once the existence of rhythm is established, the most plausible hypothesis will be that with these words elision sometimes operates, sometimes not; the reader's rhythmic sense could have been relied on as a guide. Thus, though τοῦτο is elided 34 times, the possibility of hiatus seems guaranteed, if Appian is rhythmic, by the recurring phrase ἐς τοῦ̄το̆ ἐ̆τε̆λεύ̄τᾱ |, followed by full stop: Sic. fr. 2.5, Mac. fr. 9.4; by high point: B Civ. 5.155; by comma: Hisp. 322, B Civ. 5.352. (Rhythmic closes, potential or actual, are marked by | for their end; |̮ is used for the first close when two overlap [p. 14]. When | appears at the beginning of a quotation, it indicates that the words begin directly after a rhythmic close.) Note that P. Dur. 2 fr. A does not support, as T.F. Brunner thought, the manuscript στρατιὰν δὲ ἀγείρας at B Civ. 2.27, where rhythm would favour elision; see Welles, C.B., ‘Fragments of Herodotus and Appian from Dura’, TAPhA 70 (1939), 203-312 Google Scholar; Brunner, T.F., ‘Two papyri of Appian from Dura-Europus’, GRBS (1984), 25, 171-5Google Scholar; Babcock, R.G. and Johnson, W.A., ‘The Appian papyrus from Dura-Europus (P.Dura 2)’, Bulletin of the American Society of Papyrologists 31 (1994), 85-8Google Scholar.

¹² After Appian's rhythmic status is accepted, it will become legitimate to look again at hiatus with common words; it will be clear that even here a great many instances form part of rhythmic phrases.

¹³ At Syr. 201 πολεμούμενον δὲ καὶ̄ πλέ̆ο̆σῐ χρῆ̄σθαῑ· | one can see how the rhythmic phrase has been added to Polyb. 21.43.13; cf. K. Brodersen, Appians Antiochike (Syriake 1,1-44,232). Text und Kommentar nebst einem Anhang: Plethons Syriake-Exzerpt (Münchener Arbeiten zur Alten Geschichte 3) (Munich, 1991), 203; note the alterations of Polybius in a rhythmic direction, e.g. at Syr. 202 μηδὲ φυγάδας ἐξ αὐτῆ̄ς ὑ̆πο̆δέ̆χε̄σθαῑ |; Polyb. 21.43.15 μηδ᾿ ὑποδέχεσθαι τοὺς φεύγοντας.  B Civ. 5.20 τοῦ ὑμε̆τέ̆ροῡ εὐ̄ε̄ργέ̆τοῡ, | would have been an exception to the principle on the short first element if we had not excluded possessive adjectives; but cf. e.g. Philo, Leg. 316 τοῦ ἡμετέρου ἱεροῦ, Decal. 1 τοῦ ἡμετέρου ἔθνους (τοῦ̄ ἡ̄με̆τέ̆ροῡ ἔ̆θνοῡς |), Joseph. AJ 14.323 τοῦ ἡμετέρου ἔθνους. At B Civ. 5.34 Ἀντωνίωι μὲν ἡ περὶ ἅπαντα τέως ἐπιμέλεια ἀθρόα ἠμβλύνετο,  ἀθρόα will be neuter plural with ἅπαντα (ἀ̆θρό̆ᾰ ἠ̄μβλύ̄νε̆το̄, |); cf. Posid. F 253.88-9 Edelstein-Kidd, Philo, Abr. 199, Luc. Tyr. 18. To make sure that no instances of hiatus in 5.1-200 were missed, -α α-, -α ε-, etc., were searched for by computer. In the Civil Wars generally, there are exceptions to the picture given above, but they are infrequent. A particularly intractable example of hiatus as part of a non-rhythmic close is 4.485 | ἐς πέτρας ἢ ἀκτὰς ἐξεφέροντο ἐρήμους. (cf. Syr. 318 | ἤρξατο ἄχθεσθαι [comma follows in Viereck-Roos, Brodersen and Goukowsky], Mith. 68 ἐπέκειντο ὄπισθεν.). Medio-passive endings are occasionally elided in MSS of Imperial authors (so B Civ. 4.349 ἀφί̄κε̆τ᾿ αὐ̄τῶ̄ι, |, Plut. Brut. 37.7 ὤιχοντ᾿ ἀποπτάμενοι. [ὤιχο̄ντ᾿ ἀ̆πο̄πτά̆με̆νοῑ. |]); such a solution would not imply that we should elide all the great number of Appian's medio-passive endings in hiatus (it would become an inexplicable coincidence that they occur mostly in rhythmic phrases). For diphthongs (not simple long vowels) as part of rhythmic closes cf. B Civ. 3.107 ἀγορᾶ̄ι ἀ̆νε̆τί̆θεῑ. |, 265 ἐπὶ τῆι δε̆δο̆μέ̆νη̄ι ἀ̄ρχῆ̄ι |.

¹⁴ E. Fraenkel, ‘Kolon und Satz, II’ [1933], Kleine Beiträge zur klassischen Philologie, 2 vols. (Rome, 1964), 2.93-139, at 117 n. 1 and 135 views as a ‘Kurzkolon’ even an opening particle when followed by another word followed by ἄν; in ‘Nachträge zu “Kolon und Satz, II”’, ibid. 131-9, at 138 n. 2 he inclines to the promising designation ‘Auftakt’. At the other end of the sentence or part of it, apparent hiatus within an unrhythmic phrase can be actually at the end of a rhythmic phrase, followed by a short unrhythmic close; cf. e.g. outside the Civil Wars, Mac. fr. 9.4 ἐς τὸν τῶν Ἰσθμίω̄ν ἀ̆γῶ̄νᾱ | ἐπελθών,; Mith. 535 | ὃ περὶ τῶι ξίφει φάρμακον ἀεὶ̄ πε̆ρῐέ̆κεῑτο̄  | ἐκίρνη. (comma after περιέκειτο in Viereck-Roos and Goukowsky).

¹⁵ καί probably does not include διὰ τόδε; cf. 197. ἐστρατηγηκώς at 202, unlike στρατηγῶν at 210, refers to a praetorship; cf. Them. Or. 34.34, 2.219.2-3 Downey-Norman.

¹⁶ Cf. Syr. 40 | τετρακόσια ἀνέστη̄σᾰ ἄ̄στη̄ | καὶ περὶ τῆι πόλει τὸν ἀγῶνα πολλάκις ὑμῖ̄ν ἐ̆πέ̄στη̄σᾱ, |; Cic. Arch. 12 neque ad communem adfērrĕ frūctūm | neque in aspectum lucēmquĕ prōfērrē , |. The parallel from Mith. 86 is not in E. Gabba, Appiani Bellorum Civilium liber primus. Introduzione, testo critico e commento con traduzione e indici (Florence, 1967²), 254–6, or in P. Goukowsky and F. Hinard, Appien, Histoire Romaine, VIII, Livre XIII, Guerres Civiles, Livre I (Paris, 2008), 190. Mith. 85, which mentions not only Sulla but also the Bella Ciuilia, suggests here an ominous parallel between the two men's actions; the link perhaps even hints at Mithridates as the source for Sulla's idea on how to deal with enemies (worth considering historically, if one separates the idea of the lists?).

¹⁷ For doubts on the text of the last part, cf. Zerdik (n. 10), 66, and Viereck's apparatus. Hiatus after αἰεί/ἀεί is found at Pun. 623, B Civ. 2.577 and quite frequently in Plutarch; the diphthong in a rhythmic ending at B Civ. 2.577 suggests that the word is treated as a common one for the purposes of hiatus. On Appian's shaping of the history here in 3.194 cf. P. Goukowsky and P. Torrens, Appien, Histoire Romaine X, Livre XV, Guerres Civiles, Livre III (Paris, 2010), 140-1 (one could go further).

¹⁸ For the divine element at the end cf. e.g. Hahn, I., ‘«Appianus tacticus»’, Acta Antiqua Academiae Scientiarum Hungaricae 18 (1970), 293-306 Google Scholar, at 293-5 (thinks divine intervention very rare without a human decision as intermediary); B. Goldmann, Einheitlichkeit und Eigenständigkeit der Historia Romana des Appian (Beiträge zur Altertumswissenschaft 6) (Hildesheim, Zurich, New York, 1988), 32. Giving by a god: e.g. Hann. 224. For the antithesis of ships cf. also L. Casson, Ships and Seamanship in the Ancient World (Princeton, 1971), 169.

¹⁹ Note also Hann. 205. In B Civ. 1.413 | ὕδατός τε ὄντος ἐξ οὐ̄ρᾰνοῦ̄ πο̄λλοῦ̄ |̮ is marked as a self-sufficient phrase, which is then expanded, by the intertextuality with Thuc. 2.77.6 ὕδωρ ἐξ οὐρανοῦ πολύ (Appian confirms that ἐξ οὐρανοῦ should not be omitted there, as by the OCT and Alberti). For overlap in Cicero see Hutchinson (n. 3), 494-6; id., ‘Pope's spider and Cicero's writing’, in T. Reinhardt, M. Lapidge, J.N. Adams (edd.), Aspects of the Language of Latin Prose (PBA 129) (Oxford, 2005), 179-93, at 181 n. 5, 184, 186, 188, 190.

²⁰ Note the different twist after 2.492 (death of Caesar) | ἐπὶ τὴν σφαγὴν τὸ ξίφος ἤρεῑσε̆ πρῶ̄το̄ς,  |. At Mith. 384 the next close comes before it is expected, to highlight the one-liner, marked out by rhythmic density (cf. 352): ῾εἰ μὲν πρέσβεις εἰ̄σὶ̆ν οἵ̄δε̆, |̮ πο̄λλοί̄, | εἰ̄ δὲ̆ πο̆λέ̆μῐοῑ, | πά̄μπᾰν ὀ̆λί̆γοῑ.᾿ | Here and at B Civ. 2.360, if these examples are accepted, what would have been a final syllable lengthened by ‘metrical’ pause remains short in the overlapping rhythm that surprisingly takes over. (Similarly at 5.92 above | τοῖ̄ς ἐ̆ροῦ̄σῐν |̮ ὡ̄ς ἐ̄ν δί̆κη̄ι. |, though there -σῑν  would also be rhythmically possible:   ̄  ̄  ̄    ̄  ̆  ̄.)

²¹ Cf. Hutchinson (n. 3), 496-9 on density in Cicero; note too Hunter (n. 1), 84.

²² Cf. A. Gowing, The Triumviral Narratives of Appian and Cassius Dio (Ann Arbor, 1992), 66-70 on this meeting. For the impact of Octavian cf. also Ill. 58, with κατέπληξε; B Civ. 3.368, with ἐξεπέπληκτο; 371, with καταπεπλῆχθαι;  378 (mentioned above), with τεθη̄πό̆τω̄ν πά̄vτω̄ν |. On Appian's manipulation of events at Syr. 206 cf. Brodersen (n. 13), 211-12; and note e.g. Pun. 327.

²³ For the general possibility note Espelosín, F.J. Gómez, ‘Estrategias narrativas en la Historia de Apiano: algunos ejemplos’, Annali della Scuola Normale Superiore di Pisa Ser. 4, 1.1 (1996), 103-17Google Scholar, at 117: ‘Su obra obedecía a un diseño artístico particular … , para cuyo completo desarrollo se sirvió de determinadas estrategias narrativas y unos procedimientos estilísticos.’ Architecture on the scale of a book (or most of one): Brodersen (n. 13), 232-6; cf. id., ‘Appian und sein Werk’, ANRW 2.34.1 (Berlin and New York, 1993), 339-63Google Scholar, at 359 (even saying [358], ‘Tatsächlich gliedert Appian das Material geradezu rhythmisch, etwa wenn … ’); C. Leidl, Appians Darstellung des 2. Punischen Krieges in Spanien (Iberike c. 1-38 § 1-158a). Text und Kommentar (Münchener Arbeiten zur Alten Geschichte 11) (Munich, 1996), 300-4. On B Civ.: Magnino, D., ‘Le “Guerre Civili” di Appiano’, ANRW 2.34.1 (Berlin and New York, 1993), 523-54Google Scholar (building on, but more positive than, E. Gabba, Appiano e la storia delle guerre civili [Florence, 1956]). Of the whole work (thematically): Goldmann (n. 18); Bucher, G.S., ‘The origin, program, and composition of Appian's Roman History ’, TAPhA 130 (2000), 411-58Google Scholar (with attention to the process of composition, which is perfectly compatible with perception of an evolving design; for monarchy cf. esp. 431-6, and Brodersen [(this note), 1993], 356). Unity on level of military analysis and character of generals: Hahn (n. 18). Studies of moral and military shaping in individual episodes: e.g. E. Potz (n. 2), and id., ‘Appians Klio dichtet: Die Curio-Episode bei Appian (E 2,44, 175-45, 185) und Caesar (b.c. II 23-44)’, Philologus 142 (1998), 293-9Google Scholar; Bucher, G.S., ‘Fictive elements in Appian's Pharsalus narrative’, Phoenix 59 (2005), 50-76 Google Scholar.

²⁴ For the mean and the sample standard deviation, see M. Samuels, J. Witmer, A. Schaffner, Statistics for the Life Sciences (Harlow, 2014⁴), 63-4, 82-4, 87-8; J.E. Freund and B.M. Perles, Modern Elementary Statistics (Upper Saddle River, NJ, 2007¹²), 74-93. For the normal distribution, see Samuels, Witmer and Schaffner, 155-81.

²⁵ Cf. n. 1 (Norden, Wilamowitz). See Heibges (n. 1) for Chariton, Josephus (105-8) and Polemon (83-91); for Philo, De Groot ([n. 1], 1921), 63-4, ([n. 1], 1926), 34-5 and Winterbottom (n. 1), 264; for Longus, De Groot ([n. 1], 1926), 35, Hunter (n. 1), 84-5; for Longinus, Wilamowitz (n. 1), 259 n. 1 (‘Ein Bösenwilliger könnte manche Klauseln asianisch nennen wollen.’), Russell (n. 3), 196-7. On Roman historiography, cf. G.O. Hutchinson, Greek to Latin: Frameworks and Contexts for Intertextuality (Oxford, 2013), 238-40. As for rhetoric and its traditions, one exercise (P.Oxy. 3235 [third cent. a.d.]) looks rhythmic in what we have (the proper names in fr. 2 col. ii 5-6 notwithstanding); but the sample is much too small for us to be sure. For the Hellenistic material see recently Papanikolaou, D., ‘The Aretalogy of Isis from Maroneia and the question of Hellenistic “Asianism”’, ZPE 168 (2009), 59-70 Google Scholar.

²⁶ For Plutarch's Lives, cf. De Groot ([n. 1], 1919), 1-6, 12-13, 42-54; ([n. 1], 1926), 35-6; on the Moralia, Sandbach, F.H., ‘Rhythm and authenticity in Plutarch's Moralia ’, CQ 33 (1939), 194-203 CrossRefGoogle Scholar. Plutarch is an obvious point of comparison with Appian, and was probably so for Appian himself (his fame is implied e.g. by his appearance in the first chapter of Gellius’ Noctes Atticae; add to the papyri P. Gen. inv. 477 and 504 [second or third cent. a.d.], M. Bagnoud, S. Gindrat, V. Monteventi, S. Nasel, Th. Schmidt, ‘Nouveaux fragments d'un papyrus de la Vie de César de Plutarque (P. Gen. inv. 477 et 504)’, MH 70 [2013], 10-15). Vatic. Pal. 2 puts the comparison of Caesar and Alexander which ends Bella Ciuilia 2 after Plutarch's pair Alexander and Caesar: Manfredini, M., ‘Un nuovo testimone di Appiano in un codice di Plutarco’, Annali della Facoltà di Lettere e Filosofia dell'Università di Napoli 20 (1977-8), 105-8Google Scholar. Appian's vocabulary cannot be discussed here; but it should once more be seen in relation to his structures. Scholarship on Racine suggestively illustrates how the restriction of vocabulary can contribute to impact and the charging of words: cf. J.-G. Cahen, Le Vocabulaire de Racine (Paris, 1946), Ch. Bernet, Le Vocabulaire des tragédies de Jean Racine : analyse statistique (Travaux de linguistique quantitative 12) (Geneva and Paris, 1983); note 104-6 on Racine and Corneille; M.G. Pittaluga, Aspects du vocabulaire de Jean Racine (Biblioteca della ricerca: Cultura straniera 40) (Fasano and Paris, 1991); Louvat, B., ‘Le Vocabulaire à l’épreuve de la langue : l'exemple d’Andromaque ’, La Licorne 50 (2009)Google Scholar (http://licorne.edel.univ-poitiers.fr/document4394.php).