¹ Cf. Grial's notes in the editions of Isidorus by J. Breul (Paris, 1601), and F. Arevalo (Rome, 1797–1803). The latter is printed in Migne, , PL LXXXII. Cf. especially col. 99Google Scholar, note g.

² l.c.

³ Not even during his six years' residence at Alexandria (probably from A.D. 372 to 377).

⁴ l.c. Text by W. M. Lindsay, Oxford, 1910. For inperitiam there are the variant readings inpuritiam and inpueritiam. From the latter pueritiam has been conjectured; it was read, for instance, by Breul.

⁵ Etym. I, III, 7–9.

⁶ v. infra on the mysticae litterae.

⁷ Attributed to Ennius, ann. 625, by Vahlen, on the authority of a twelfth-century MS., Cod. Admont. 472 f. 13b (cited by Huemer, J., Wiener Studien II, 305Google Scholar). A marginal gloss in this MS. reads: ‘Albus (sic) est tabula, ubi scribebantur nomina illorum, qui ad militiam recipiebantur, et si contigisset ut aliquis eorum fuisset interemptus, apponebatur super nomen illius theta littera, quae mortem significat. Habet enim haec quoddam iugulum. Vnde Ennius uersificator optimus: O multum ante alias infelix littera theta.’ If an Ennius did write this line, perhaps we should consider Suetonius' ‘Ennius posterior’ (cf. Suet., gramm. 1), whose two books de litteris syllabisque, item de metris may well have contained a metrical ABC. For the structure of the line, cf. Verg., catal. 9, 23Google Scholar; eclog. 3, 78; Aen. III, 321; also Sil. Ital. 14, 52.

⁸ Ibid. I, III, 9. This interpretation of T had already been given by Tertullian, (adv. Marc. III, 22Google Scholar) and Origen (comm. in Ezech. in PG 13, 799), and arose naturally from the fact that the last letter of the Hebrew alphabet—tau—meant a mark or cross. Cf. Dornseiff, F., Das Alphabet in Mystik und Magie, Leipzig, 1922, 23Google Scholar. For this information I am indebted to Mr. Hugh Last.

⁹ cf., for instance, Scaliger's note on Ausonius, epigr. 128 (in Souchay's Variorum edition, London, 1823): ‘“Tuumque nomen θ sectilis signet.” Erat enim litera condemnatoria, quae praecipiebat θανατοṽν. Martialis (VII, 37): ‘Nosti mortiferum quaestoris, Castrice, signum? Est operae pretium discere theta novum.” Quin et in castris militem ex cohorte morte dispunctum haec nota significabat, ut videmus in veteribus inscriptionibus.’ To these authorities may be added Sidonius Ap., Carmina IX, 335: ‘Isti qui ualet exarationi, districtum bonus applicare theta.’

¹⁰ cf. Wilmanns, , Exempla Inscriptionum Latinarum, Berlin, 1873, 158 n.Google Scholar: ‘Mommsenus tamen origine credit esse O cum hasta transversa et significare obiit.’ This view is developed further by Mowat, in Bulletin épigraphique IV (1884), 133 ff.Google Scholar, and by Hübner, Exempla Scripturae Epigraphicae Latinae, 1885, no. 4.

¹¹ Wessely, (Karl), Schrifttafeln zur älteren lateinischen Palaeographie, Leipzig, 1898, no. 8Google Scholar. Republished by Wessely, in Die ältesten lateinischen und griechischen Papyri Wiens, Leipzig, 1914Google Scholar (= Stud. Pal. XIV, pl. VIII).

¹² Text and photograph by Mallon, Marichal, and Perrat, in L'Ecriture latine de la capitale romaine à la minuscule, Paris, 1939, no. 17Google Scholar, pl. XII.

¹³ Most recently edited by Marichal, R., in L'Occupation romaine de la Basse Égypte, Paris, 1945. See especially p. 45Google Scholar, n. 71.

¹⁴ ‘A Register of a Cohort in Moesia,’ Raccolta Lumbroso, Milan, 1925, pp. 265 ff.Google Scholar; cf. Aegyptus IX (1928), 63 ffGoogle Scholar.

¹⁵ ‘Latin Papyri in the University of Michigan Collection’ (Michigan Papyri, vol. VII), edited by Henry A. Sanders, with contributions by James E. Dunlap, Ann Arbor, 1947, p. 35.

¹⁶ o.c, p. 28. See also pl. v.

¹⁷ Note also the deletion of the name.

¹⁸ O.C, p. LXXIII.

¹⁹ Introduction to the Study of Latin Inscriptions, New York, 1896, 231Google Scholar.

²⁰ For θ cf. the following selection from CIL in:—Pannonia Superior: 3779, 3788, 3789, 3790, 3791, 3793, 3794, 3797, 3798, 3804, 3805, 3809, 3815, 3817, 3818, 3820, 3821, 3822, 3823, 3825, 3826, 3829, 3853, 3855, 3861, 3862, 3866, 3871, 3874, 3877, 3878, 3892, 4089. Noricum: 4857, 4929, 4950, 5127, 5142, 5337, 5467, 5527, 5546, 5547, 5567, 5591, 5601, 5616.

²¹ Latin Epigraphy, 2nd edition, revised by Campbell, S. G., Cambridge, 1927, 62Google Scholar.

²² L'Occupation romaine, p. 46, n. 72. Fink (o.c, 8, n. 8) had already arrived at a similar conclusion. He also suggested the possible influence of Hellenistic military procedure.

²³ On the other hand, this meaning is not far removed from that given to the word by the jurists, ‘ignorantia facti vel iuris,’ and cf. the following instances in the Digest:—2, 1, 7, 4; 17, 1, 57; 25, 4, 2, 1; 26, 7, 40; 38, 2, 14, 4; 50, 17, 132. The nearest parallel, however, is found in Felix, Minucius, Octavius, XXXV, 5Google Scholar: ‘quamquam inperitia dei sufficiat ad poenam, ita ut notitia prosit ad veniam.’

²⁴ Cod. Theod. VI, 1, 14; cf. also VI, 24, 2; VII, 1, 11.

²⁵ The Apologia adv. Hieronymum was written in A.D. 401. The Thesaurus quotes also St. Augustine, retract. 2, 39; coll. c. Don. I, praef.; and Alcinus Avitus, epist. 87 (Peiper, , MG VI ₂, p. 96, 23Google Scholar). Cf. also Fink, o.c, p. 7 and n. 6, who points out that the word appears to be confined to patristic writers.

²⁶ There are four other examples of brevis as ‘list’ in the Augustan History; in these, however, the word does not have a purely military significance. Cf. SHA, Alex., 28, 3Google Scholar; Aurel., 35, 5; quadr. tyr., 15, 7; ibid., 15, 8. It is common in the Codes of Theodosius and Justinian. Cf. also Seeck in P-W s.v.

²⁷ Вρέβιον and the other spellings, βρέουιον βρέουιον and βρέουιον are quite frequent in the fourth century, but are not clearly attested earlier. The term persists in the papyri of the fifth and sixth centuries, and sporadic instances occur until very much later. Cf. the citations in Du Cange, and, for the papyri, Wessely, , ‘Die lateinischen Elemente in der Gräzität der ägyptischen Papyrusurkunden,’ Wiener Studien XXIV–XXV (1902–1903)Google Scholar; Meinersmann, , Die lateinischen Wörter und Namen in den griechischen Papyri, Leipzig, 1927Google Scholar, and Preisigke, Wörterbuch. L-S-J ⁹ quote also IG 12 (9) 907, 15 (Chalcis, A.D. 359). Since it is safe to assume that the use of the term ‘brevis’ in Latin must have preceded its adoption into Greek, we can be certain that the word was in general currency by the beginning of the fourth century, and probably much earlier.

²⁸ Thompson, E. A., A Roman Reformer and Inventor, Oxford, 1952, p. 97Google Scholar (De reb. bell. v, 5). I owe this reference to Professor Thompson. It is not listed in the Thesaurus. It is true that in CIL VI, 1585a (most recently edited by Arangio-Ruiz, V. in Fontes Iuris Romani Anteiustiniani III, Negotia, Florence, 1943, no. 110Google Scholar), an inscription of A.D. 193, matri[culam] has been read in 1. 8, et in matri[culam referri, quod sine in]iuria cuiusqua[m fiat], but since no other example is found before the fourth century, it would seem preferabe to restore in 1. 8 in matri[cem or in matri[ces. Cf. CIL VIII, 6948, ‘secundum matricem publicam,’ and Tertullian, de fuga in persec., 13: ‘in matricibus beneficiariorum.’ It was the form μάτριξ that was adopted in Greek, though surviving examples are late. Cf. Ensslin in P-W s.v. ‘matricula.’

²⁹ It is not within the province of this discussion to enter into the controversy about the date and composition of the Historia Augusta. To take extreme instances, neither a Constantinian nor a fifth-century date would seriously affect our argument: the former would only confirm what we already suspect, that ‘brevis’ was a recognized technical term well before the middle of the fourth century; whereas the latter supposition would not disprove this, for the general use of the Greek equivalent about that date implies a considerably earlier use in Latin.

³⁰ De re mil. I, 8, 13 ff. Cf. Schenk, Dankfrid, Die Quellen der Epitoma Rei Militaris (Klio, Beiheft XXII, 1930Google Scholar).

³¹ De mag. I, 47: μάρτυρες Κέλσος τε καὶ Πάτερνος καὶ Κατιλίνας, οὐχ ὁ συνωμότης ἀλλ᾿ ἕτερος, Κάτων 〈τε〉 πρὸ αὐτῶν ὁ πρῶτος καὶ Φροντῖνος, μεθ᾿ οὓς καὶ Ῥενᾶτος, Ῥωμαῖοι πάντες. The order in which Lydus cites these authorities would seem to assign Catiline to a date sometime between Paternus (executed A.D. 183) and Vegetius, i.e. either to the third or to the fourth century.

³² So in the Digest. Tarrutenius in Dio Cassius (LXXI, 12, 3, and LXXII, 5, 1). (SHA Comm. 4, 1Google Scholar, has the genitive Tarruteni.) The latter spelling is generally, favoured by historians. See Berger in P-W, s.v., and PIR ¹, T 24. For the background of these jurists see also Kunkel, W., Herkunft und soziale Stellung der römischen Juristen (Weimar, 1952), 219 ffGoogle Scholar.

³³ For Menander, cf. PIR ² A, 1100Google Scholar; for Macer A, 379) 381 (M. Aemilius Macer Faustinianus, whom Krüger identifies with the jurist).

³⁴ Cf. Schenk, o.c, p. 7. But the fact that neither Vegetius nor Lydus mentions either Menander or Macer robs this argument of much of its force.

³⁵ That Macer made extensive use of the two earlier writers is proved by his citing Paternus at least once, and Menander four times in the few brief extracts from his two books de re militari that are preserved. He cites Paternus in Dig. 49 16, 12, and Menander in 38, 12, 1; 48, 19, 14; 49, 16, 13 bis. The other titles attributed to Macer are 29, 1, 26; 35, 2, 92; 49, 17, 11. These concern soldiers' wills and castrense peculium. Dirksen suggests that in 38, 12, 1, where Paulus and Menander are cited, Paternus and Menander are meant. But Paulus is cited in 49, 16, 13, also, and frequently in the excerpts from Macer's other works. Cf. Berger in P-W, s.v. ‘Tarruntenus Paternus,’ and Dirksen, H. E., Hinterlassene Schriften II, Leipzig, 1871, 422Google Scholar.

³⁶ By an odd coincidence, three of the four papyri that contain theta nigrum or an equivalent term are of the Trajanic-Hadrianic period: Schriftt. 8 seems to be late Trajanic or early Hadrianic. P. Mich. VII, 435Google Scholar, is dated by Sanders (p. 28 f.) between A.D. 109 and 119, with slight indications pointing to 4th July, A.D. 116. Racc. Lumbr. 265 f. is usually given a date not long after A.D. 102. If, however, a suggestion of Stein's is adopted, the date may well be A.D. 116 also. (Stein, A., Die Legaten von Moesien, Budapest, 1940, 112–13Google Scholar. He concludes that—stano—in l. 30 is what remains of a consular dating, ‘L. Vipstano Messalla M. Vergiliano Pedone cos’ (= A.D. 115); and that the procurator, Latinianus, is the Cornelius Latinianus of Digest 48, 5, 28, 6). P. Fay. 105, alone is of a later date. Marichal (o.c, p. 49) ascribes it to about A.D. 175.