¹ For his names see below, p. 26.

² XVI, 10, 15, cf. VIII, 5, 61: Vicar, it should be noted, not the non-existent post of ‘praetorian prefect of the Spains’ as given in virtually all editions and reference works from Gronovius (1670) down to Jan, Teuffel, Schanz, P-W, OCD, Sandys and Stahl (n. 5 below, p. 6), not to mention many less authoritative works. It is sad to think of three centuries of scholars copying this error from one another without ever looking up the original text.

³ XI, 28, 6.

⁴ VI, 8, 1.

⁵ See the authorities cited by Stahl, W. H., Macrobius' Commentary on the Dream of Cicero (New York, 1952), pp. 6 f.Google Scholar, to which add, e.g., Fuhrmann, M., Philologus cvii, 1963, 308.Google Scholar I am at a loss to know from what source Stout, S. E., TAPA LXXXVI, 1955, 252Google Scholar, writes of Macrobius beginning in 390 ‘the political career which sixteen years later led to his assassination’.

⁶ Following the standard article by Georgii, H. in Philologus LXXI, 1912, 518 f.Google Scholar

⁷ REL XXXIV, 1956, 232 f.

⁸ Fuhrmann, M., Philologus CVII, 1963, 301 f.Google Scholar

⁹ Magnus Maximus is the only Emperor who appointed a non-eunuch praepositus (Zosimus IV, 37, 2)—a short-lived experiment, for we soon find the eunuch Gallicanus as his praepositus. Cf. Hopkins, M. K., Proc. Camb. Phil. Soc. CLXXXIX (n.s. 9), 1963, 68.Google Scholar

¹⁰ cf. Stahl, op. cit. (n. 5), 4–5, and the works of Mras and Wissowa there cited. Note also the manner in which he introduces Greek words: cf. especially Sat. I, 15, 16, ‘quod Graeci ἰδεῖν dicunt, nos V littera addita videre dicimus’; Comm. 1, 5, 9, etc. Contrast Ammianus, who reveals that he is a Greek by using the first person in such explanatory phrases, e.g.: ‘Machina … quam ἑλέπολιν Graeci cognominamus’ (XXIII, 4, 10). Cf. CQ n.s. XIV, 1964, 324. It has been suggested to me that this point is not valid because the characters in the Saturnalia are (for the most part at any rate) Romans and would therefore have been made to speak as Romans even if their creator had been Greek. This might certainly apply for the Saturnalia, but Macrobius in fact uses nos meaning ‘we Latin-speakers’ far more often in the Comm., where he is speaking throughout in his own person. Cf. also § II of my article cited in n. 18.

¹¹ Macrobii Opera I (1848), p. 6. It would be perfectly natural for a man writing in Rome to say that Africa was ‘sub alio caelo’. At De Bello Gild. 2 Claudian calls Africa ‘alterius convexa poli’.

¹² For the proconsul see the fanciful speculations of Leschi, L. (Études d'épigr., d'archéol. et d'hist. Africaines, 1957, 132 f.Google Scholar) on the role of ‘le dernier proconsul paien d'Afrique’ in the Donatist controversy. The vicar was fined for misusing the cursus publicus (C. Th. VIII, 5, 61), but this would have been forgotten ten years later. Cf. now Chastagnol, A., Les empereurs romains d'Espagne (1965), 277.Google Scholar

¹³ Chastagnol, A., [Les]; Fastes [de la préfecture de Rome au Bas-Empire];, 1962, 268–9.Google Scholar

¹⁴ AÈ 1912, no. 178: W.Ensslin, in P-W XIV, 2533, though cf. Mazzarino, S., Stilicone, 1942, 81, n. 6.Google Scholar Doubtless a relative of Flavius Macrobius Longinianus, whose brilliant career at court began in 399 (Chastagnol, Fastes, 255): it was normal for a man to take his relatives with him as he rose in the imperial service (cf. the notorious case of the family of Ausonius).

¹⁵ Preserved in most MSS. at the end of Bk. I of the Comm.: see Willis' ed., vol. 11, p. 94. Oddly enough Willis heads every other page throughout his edition ‘Ambrosii Theodosii Macrobii’, without explaining why he chose this particular combination.

¹⁶ Expos. Psalm., ed. Adriaen, , Corp. Christ. I, 1958, 30, 20–1; 116, 125 f.Google Scholar

¹⁷ In Isag. Porph. I., CSEL XLVIII, 1906, 31, 21 f.

¹⁸ cf. my article ‘Macrobius, Avienus and Avianus’ in CQ n.s. XVII, 1967.

¹⁹ cf. Chastagnol, Fastes 80–82.

²⁰ Indeed if any of Macrobius' names is a signum, Macrobius itself would surely be at least as likely a candidate. It has the three normal characteristics of the signum: it is a Greek word, it means something, and it ends in -ius. And it is in fact found as a signum: ‘Flavius Paranius ὁ καὶ Μακροβίος’ (P. Oxy. 1303, A.D. 336), whereas, so far as I have been able to discover, Theodosius is not. At any rate it is not included by G. Évrard in her list of names beginning Theo- which are found as signa (Mél. d'arch. et d'hist. LXXIV, 1962, 624).

²¹ C. Th. XII, 6, 33: probably the Theodosius who was primicerius notariorum in 426 (cf. C. Th. VI, 2, 25). This identification was suggested in passing by Mazzarino, S., ‘La politica religiosa di Stilicone’, Rend. Ist. Lombardo, 1938, 255 f.Google Scholar A.D. 384 was the year of the death of Praetextatus: 485 is the year of Memmius Symmachus' consulate. In the subscriptio both he and Macrobius Plotinus Eudoxius are styled only ‘v.c.’ After his consulate Symmachus would have acquired, and would certainly have used, the title ‘v.c. et inl.’ (he was careful to give Macrobius Theodosius this title in the same subscriptio).

^21a Laws were very frequently based on information supplied by the recipient: cf. Jones, A. H. M., Later Roman Empire I (1964), 351 f.Google Scholar

²² REL XXXIV, 1956, 232 f.

²³ REL XXXVI, 1958, 214 f.

²⁴ See n. 8.

²⁵ At Ep. 48, 13 he lists ‘qui Ecclesiasticorum de impari numero disputarint: Clemens, Hippolytus, Origenes, Dionysius, Eusebius, Didymus’. The subject is treated in detail by W. H. Roscher, Hebdomadenlehren der griech. Philosophien und Ärzte (1906), and cf. Robbins, F. E., ‘The Tradition of Greek Arithmology’, Class. Phil. XVI, 1921, 97 f.CrossRefGoogle Scholar

²⁶ Favonius Eulogius, for example, writing at about the same time as Jerome, though independently of both Jerome and Macrobius, also refers to the Timaeus apropos of the properties of 7.

²⁷ cf. for example, another contemporary writer, Calcidius, In Timaeum pp. 88, 16; 297, 9 Waszink, with the passages quoted by Waszink ad locc. In fact it is common to most arithmological treatises: cf. Robbins, op. cit. (n. 25), 121, and Sicherl, M., Abh. Mainz X, 1959, 704–5.Google Scholar

²⁸ Courcelle, , Lettres grecques [en Occident²] 1948, 88 f.Google Scholar

²⁹ Harv. Studies XXV, 1914, 73 f.

³⁰ Bloch, H., Conflict between Paganism and Christianity, ed. Momigliano, A., 1963, 208.Google Scholar

³¹ cf. Carcopino, J., Autour des Gracques (1928), 109, cf. 85.Google Scholar

³² As Macrobius himself explicitly comments at Comm. I, VII, 9; VIII, 2.

³³ Courcelle, Lettres grecques 10–11, following Wissowa, thinks that both Macrobius and Athenaeus were drawing on a common source; but Wissowa did not persuade Martin, J., Symposion: die Geschichte einer literarischen Form (1931), 286.Google Scholar The very fact that Macrobius does not name Athenaeus is suspicious: for it is his regular practice not to name the direct source of his information. For example he nowhere names either Gellius or Plutarch, both of whom he certainly used directly and extensively.

³⁴ Kaibel, ed. of Athenaeus, 1, 1887, XXXI f.

³⁵ And there are several other parallels: Cicero's De Oratore is set shortly before the death of Crassus, and his De Senectute shortly before the death of Cato. cf. also Plato's Phaedo and Theaetetus (which has two dramatic dates, the first as Theaetetus is carried dying from battle).

³⁶ A work with which Macrobius seems to have been familiar: cf. Courcelle, Lettres grecques 27 f.

³⁷ C.Just. I, 54, 5.

³⁸ Cavallera, F., Saint Jérôme I, 2, 1922, 23–4Google Scholar, argues that Praetextatus must have died before Pope Damasus (11th December) because the letters in which Jerome mentions Praetextatus' death (Epp. 23 and 39) do not also mention Damasus'. This is a reasonable, if not compelling inference, but even so it is unlikely that Macrobius would have been much concerned that Praetextatus had in fact died a week or two before rather than after the Saturnalia (and if he wrote as late as the 430's, he may well not have known to within a week when exactly P. did die).

³⁹ VI, 7, 1; VII, 3, 23; I, 2, 15; VII, 11, 1; I, 2, 3.

⁴⁰ Chastagnol, Fastes 257 f.

⁴¹ Philol. LXXI, 1912, 5.

⁴² GLK IV, 456.

⁴³ Chastagnol, Fastes 246–7. In view of the rather lower date for Servius here suggested, I would now modify my view (Hermes XCII, 1964, 363 f.) that he was actually responsible for the late-fourth-century revival of interest in Juvenal and other ‘Silver Age’ Latin poets. He is a manifestation rather than the inspiration of this movement. I do not think that this affects the main course of my argument about the date of the movement as a whole, or my point about the Historia Augusta.

⁴⁴ cf. my article cited n. 18 above.

⁴⁵ cf. Pease, , Cicero, Nat. Deor. I (1955), 25–6.Google Scholar

⁴⁶ cf. Plasberg's preface to his Teubner ed., 1922, 111 f.

⁴⁷ ‘Artistic propriety and convention discountenanced the introduction of the living’: Syme, R., Tacitus I (1958), 108Google Scholar (assuming a late dating for the Dialogus). cf. also Ovid, Trist. II, 467–8, ‘praestantia candor nomina vivorum dissimulare iubet’.

⁴⁸ cf. Seeck's ed., praef. pp. LXII–III.

⁴⁹ cf. Chastagnol, , Latomus XX, 1961, 749 f.Google Scholar

⁵⁰ Libanius, Epp. 1278, 1279.

⁵¹ The date of Rutilius' De Reditu: I hope to argue elsewhere for 416 rather than 417 or 415 (as recently argued by I. Lana in his Rutilio Namaziano, 1961, 1–104).

⁵² On Aen. X, 272; 388: cf. my article cited n. 18.

⁵³ cf. Gulick, Loeb ed., 1, 1927, XII.

⁵⁴ cf. Jones, R. E., AJPH LX, 1939, 307 f.Google Scholar

⁵⁵ Athenaeus 505 d-e.

⁵⁶ Hence I am dubious of Bloch, H.'s claim (Harv. Theol. Rev. XXXVIII, 1945, 207 f.Google Scholar) that the discourse assigned to Praetextatus in Sat. I, 17– 24 represents Praetextatus' religious views rather than those of Porphyry, from whose Περὶ θείων ὀνομάτων Macrobius probably derived it (Courcelle, Lettres grecques 17–20).

⁵⁷ Servian scholars seem agreed that such parallels as there are between the Servian corpus and Macrobius are to be explained by use of a common source: cf. Lloyd, R. B., Harv. Studies lxv, 1961, 306.Google Scholar I am quite unconvinced by E. Türk's recent contention that Servius Danielis drew on Macrobius (REL XLI, 1963, 327 f.).

⁵⁸ In the preface to his edition (Coll. Lat. XXVII, 1957) 7, n. 2.

⁵⁹ Courcelle (REL XXXVI, 1958, 211) argues that both are drawing on Calcidius' In Timaeum. But though Favonius certainly did (Skutsch, F., Philol. LXI, 1902, 196Google Scholar; Waszink, J. H., Studien zum Timaioskommentar I, 1964, 77Google Scholar, n. 1 and a letter to me dated 25th August, 1965), Macrobius equally certainly did not (Waszink, preface to his edition of Calcidius, 1962, XV; Mras, K., Sitzb. Berlin 1933, 255Google Scholar). And it is hard to believe that Favonius would have copied out Calcidius' already hackneyed simile in the way he did had he read Macrobius' polemic. The objections of Sicherl, M., Rh. Mus. CII, 1959, 355–8Google Scholar, are equally baseless.

⁶⁰ REL 1958, 212–3, having argued that it was used by Augustine for the last book of the Civ. Dei. M. Sicherl's calculations in Abh. Mainz X, 1959, 668, are vitiated by his belief that the Superius consularis of Byzacena to whom the work is dedicated (whom Sicherl takes to have been a pupil of Favonius) was an ex-consul, instead of a rather lowly provincial governor! Moreover, his (in any case out of date) figure for the minimum consular age is applicable only for the principate. Superius is not otherwise attested.

⁶¹ Wissowa, G., De Macrobii Sat. Fontibus, Diss. Breslau 1880, 12.Google ScholarSat. I, 24, 25, read in context, certainly need not imply that the Comm. had not yet been written (as Jan conjectured, vol. 1, p. 13): cf. § 111 of my article cited in n. 18.

⁶² This point was made by Mazzarino, op. cit. (n. 21), 256. For a convenient list of the correspondents, classified according to religion (54 pagan, 33 Christian, 47 uncertain), see McGeachy, J. A. Jr., Class Phil. XLIV, 1949, 226, n. 25.Google Scholar

⁶³ Seeck, preface to his ed., p. LX.

⁶⁴ Seeck, ib. p. XXIII.

⁶⁵ It is natural to bracket Macrobius with Servius, but it must be remembered that Macrobius is not a mere grammaticus, but a high imperial official. There are many other examples of such dignitaries writing grammatical works in this period: cf. Marrou, H.-I., St. Augustin et la fin de la culture antique,⁴ 1958, 91–3.Google Scholar

⁶⁶ As pointed out already by Hartke, W., Römische Kinderkaiser, 1951, 133, n. 1.Google Scholar

⁶⁷ Türk, E., REL 1963, 348.Google Scholar I would now modify my remarks in JRS LV, 1965, 241 on the pagan character of the Sat.

⁶⁸ For a good summary of the contents and character of the Saturnalia, cf. Glover, T. R., Life and Letters in the Fourth Century, 1901, 176 f.Google Scholar, or T. Whittaker, Macrobius, 1923, ch. I.

⁶⁹ Theodoret, HE V, 24; cf. Bloch, H., Harv. Theol. Rev. XXXVIII, 1945, 236–7, with n. 94.Google Scholar

^69a J. Martin, op. cit. (n. 33), 64.

⁷⁰ Brown, P. R. L., JRS LI, 1961, 4.Google Scholar

⁷¹ I cannot believe that Türk is right to see pagan propaganda in the representation of Vergil as ‘pontifex maximus’: as he himself admits (REL 1963, 348), this motif is common to Servius Danielis as well. It is arbitrary (and implausible: see n. 57) to claim that Servius Danielis derives from Macrobius.

⁷² By virtually all MSS.: not, however, by any of the MSS. carrying the excerpts from the De differentiis. Hence Jan conjectured, no doubt rightly (I, P. XV), that Macrobius wrote this book before his promotion. The MSS. of Aurelius Victor's Caesares similarly omit the title after his name—rightly, for although he held the illustrious post of prefect of Rome, he did not attain it till 389, while the Caesares appeared in 360 (in JRS LIV, 1964, 14, I wrongly included Victor in a list of fourth-century writers whose MSS. do record such titles of rank).

^72a See the fasti in Sundwall, J., Weströmische Studien (1915), 22.Google Scholar

⁷³ Chastagnol, Fastes 239.

⁷⁴ His edition of the first decade of Livy is attested by the famous subscriptio to be found at the end of Bks. 6, 7 and 8.

^74a Sundwall, J., Weströmische Studien 1915, 74, no. 160.Google Scholar

⁷⁵ E. Türk, Diss. Freiburg 1961 (unpublished).

⁷⁶ Holmes Dudden, F., Life and Times of St. Ambrose I, 1935, 28 f.Google Scholar, who accepts Macrobius' picture without reservations, reconciles it with Ammianus' by the simple assumption that they are writing about different sections of Roman society. But they purport to be writing about the same section, and it was precisely the absence of a literary circle such as Macrobius depicts which disillusioned Ammianus.

⁷⁷ McGeachy, J. A. Jr., Q. Aurelius Symmachus and Senatorial Aristocracy of the West, Diss. Chicago, 1942, 92.Google Scholar

⁷⁸ cf.JRS LIV, 1964, 27.

⁷⁹ Maenchen-Helfen, O. J., AJPH LXXVI, 1955, 384 f.Google Scholar

I am grateful to Professors A. Chastagnol, A. Momigliano and Sir Ronald Syme, and to P. R. L. Brown and J. F. Matthews for commenting on an earlier draft of this paper. It was read to the London Classical Society on 2nd March, 1966.