^{page 186 note 1} On the controversy see now Momigliano, A., Secundo Contributo alia Storia degli Studi Classici (Rome, 1960),Google Scholar 112f.; White, P., ‘The Authorship of the Historia Augusta’, JRS lvii (1967),Google Scholar 115 ff.; Syme, R., Ammianus and the Historia Augusta (Oxford, 1968),Google Scholar 176 ff.

^{page 186 note 2} ‘Die S.H.A.’, Bayerische S.-B. 1891, Heft 4, 465 ff.

^{page 186 note 3} ‘Die S.H.A., ’, Rh. Mus. xlvii (1892),I ff., 515 ff.Google Scholar

^{page 186 note 4} By P. White, op. cit.

^{page 186 note 5} As done by Hönn, K., Quellenunter-suchungen zu den Viten des Heliogabalus und des Severus Alexander im Corpus der Scriptores Historiae Augustae (Leipzig and Berlin, 1911),Google Scholar107 ff.

^{page 186 note 6} See Löfstedt, E., Philologischcr Kommentar zur Peregrinatio Aetheriae (Uppsala, 1911),Google Scholar 256 ff.

^{page 187 note 1} Isolated fluctuations of taste are almost inevitable in any writer.

^{page 187 note 2} A proportion based on a substantial amount of Ammianus' work.

^{page 189 note 1} For eo interfecto see Oros. 5. 21. 9.

^{page 189 note 2} See Merguet, H., Lexikon zu den Reden des Cicero, iv.Google Scholar 915^b for past participles with videor in Cicero's speeches (for occisus videor see Verr. I. 72). The large majority precede the verb. Cf., e.g., Suet. Iul. 55. 2, 84. I, Vesp. 10. See also Sulp. Sev. Chron. 1. 2. 4 occisus traditur.

^{page 189 note 3} See K. Lessing, Scriptorum Historiae Augustae Lexicon, 730b for participles with videor.

^{page 189 note 4} For a table showing the most common clausulae in Spartianus, Lampridius, and Vopiscus see de Groot, A.W., Der antike Prosarhythmus, i (Groningen, 1921), III.Google Scholar These three Scriptores do not differ significantly in their preferences. Moreover, according to Zernial, H.L., Über den satzschluss in der Historia Augusta (Deutsche Akad. d. Wiss. z. Berlin, Schrift. d. Sekt. f. Altertumswiss. 2, 1956),Google Scholar 5 f. and 6 n. 1, summarizing an unpublished Louvain dissertation (A. Fiasse, ‘Les Clausules métriques with dans l'histoire Auguste’), the preferred clausulae in all six Scriptores are the same.The table of de Groot can therefore be taken as a general guide to the favoured clausulae of the whole H.A.

^{page 190 note 1} At Elag. 14. 5 both the participle and consobrinus would have had to be transferred to leave the rhythm intact.

^{page 190 note 2} For a discussion of accentual clausulae in the H.A. see Zernial, op. cit., especially chap. iii.

^{page 190 note 3} On the secondary accent which long words (e.g. ínterémptus) must have possessed, see Lindsay, W.M., The Latin Language (Oxford, 1894),Google Scholar 159 f., 161.

^{page 191 note 1} e.g. Oros, 3. 1. 17, 3. 11. 1, 3. 15. 4, Iul. Obs. 65, 66.

^{page 191 note 2} e.g. Oros. 6. 13. 5, 6. 16. 9, 6. 18. 13, 6. 19. 2, Eutrop. 4. 20, 6. 15, 7. 3, 8. 17, 8. 18, Ampelius 42. 12.

^{page 191 note 3} On Trig. Tyr. 26. 4 see below, Section 9.

^{page 192 note 1} Occidi posset would have given a double spondee.

^{page 192 note 2} Occidere posset would have given a heroic clausula, a rhythm which is extremely rare in the H.A.

^{page 192 note 3} e.g. Oros. I. 10. 6, 3. 23. 32, 5. 16. 6. 16. 9, Eutrop. 4. 20, 6. 15, 8. 18.

^{page 193 note 1} See Szantyr, J.B. Hofmann-A., Lateinische Syntax und Stilistik (Munich, 1965), 405.Google Scholar

^{page 193 note 2} See Hofmann-Szantyr, loc. cit., on Cicero's use of inversion to create a good clausula.

^{page 193 note 3} Comm. 6. 12, Sev. 11. 5, Did. ltd. 3. 7, Geta 6. 6, Gall. 14. 9, Trig. Tyr. 6. 3, Aurel. 34. 1.

^{page 193 note 4} Interfectus is not placed after its auxiliary verb by any of the Scriptores.