^{page 49 note 1} Eusebius, Vit. Const, III. 13.

^{page 49 note 2} Dessau, 1237, with Seeck, P.-W. iii, 1981, and Zw. R. 1. 1187.

^{page 50 note 1} Seeck, P.-W. x. 93, s.v. ‘Julianos’ (32). His identity was established by Bidez, , Melanges Paid Thomas, pp. 57 ff.Google Scholar; cf. his edition of Julian, , Lettres, p. 71. 13Google Scholar, with critical note. So Julian may have had Semitic blood in his veins.

^{page 50 note 2} Julian, Misop. 352 ab.

^{page 50 note 3} Libanius, Or. i. 97; cf. Greg. Naz., Or. iv. 30.

^{page 50 note 4} Bidez, , Vie de ľEmpereur Julien, pp. 52 f.Google Scholar; cf. Eunap, . Vit. Soph., p. 428 (Wright, Loeb ed.)Google Scholar.

^{page 50 note 1} Libanius, , Or. xviii. 21Google Scholar.

^{page 50 note 2} Revue Beige de Philologie et ďHistoire, 1927, p. 125, n. 4.

^{page 50 note 3} Ammianus, xvi. 5. 10; cf. Cic, ad Att. v. 15. 3; Quintilian, v. 11. 21.

^{page 50 note 4} Vie, p. 52 f.

^{page 50 note 5} Untergang, iv. 458 f.

^{page 51 note 1} Sulpicius Severus, Dial. i. 27.

^{page 51 note 2} Epp. 8 and 9; cf. 13, ed. Bidez.

^{page 51 note 3} Ammianus, xvii. 11. 1.

^{page 51 note 4} Or. ii. 77 d. We do not hear that Constantius knew much Greek.

^{page 51 note 1} Ep. ad Ath. 284 d. Is his use of τριακοσιοστός (wrongly changed to Tpiamaros by Hertlein and Wright) in the same pamphlet, 276 a, an echo of the Latin use of irecenli to denote an indefinitely large number?

^{page 51 note 2} Gothofredus on Cod. Theod. xi. 39. 5; and see his note on vi. 24.1 for the use of Greek in Julian's laws. Greek occasionally appears in laws issued before his time, e.g. viii. 15. 1.

^{page 51 note 3} Bidez, Lettres, preface, p. iii; Borries, P.-W. x. 83. On the few Latin words—all technical terms—in his letters, see Wright, Loeb ed., vol. iii. 69, n. 3.