^{page 191 note 1} de Herodiani fontibus et auctoritate (Berlin, 1909)Google Scholar.

^{page 191 note 2} de geschiedschrijver Herodianus en zijn bronnen (Leiden, 1913)Google Scholar.

^{page 191 note 3} Pauly-Wissowa, , Realenkyklop. iii, p. 1720Google Scholar.

^{page 191 note 4} Griechische Litteratur, p. 247.

^{page 191 note 5} Grundriss der römischen Geschichte, p. 280.

^{page 191 note 6} Wahrheit und Kunst, p. 397.

^{page 192 note 1} ‘Der historische Wert der Vita Commodi’ (Philologus, Supplementband ix, p. 113).

^{page 193 note 1} de Herodiano rerum Romanarum scriptore (Dissert. Bonn, 1881), p. 36Google Scholar, note 2.

^{page 193 note 2} Geschichtl. Litteratur der röm. Kaiserzeit, ii, p. 291.

^{page 194 note 1} cf. Boissevain in the preface to the second volume of his edition of Dio, p. xxvi.

^{page 194 note 2} xi, 19, p. 60, 5–12 (Dindorf's edition).

^{page 194 note 3} See Boissevain, Hermes, 26, p. 440–452, and in his edition of Dio, iii, p. 187.

^{page 194 note 4} Fragm. 107 Müller (= Exc. de insidiis, 44, p. 82, 12–15, De Boor; also printed in Boissevain's Dio, vol. iii, p. 760, 14–17).

^{page 194 note 5} It might be supposed that Dio and Herodian had found the story of the boy and the tablet in their source, but that one of them had introduced it in the wrong place. The resemblance between many of their expressions might then be explained in this way, that, though independent of each other, they had closely followed the same original. The objection to this is that Dio declares that the story is added by himself, as something that he has been told : ἤκουσα δὲ ἔγωγε καὶ ἐκεîνο. Nor is the boy mentioned by the other authors who tell of the murder of Domitian; cf. Suet. Domit. 17; Eutrop. vii, 23, 6; Victor, Caesares, xi, 7Google Scholar; Epitome, xi, 11. The best known instance of Dio introducing something he did not find in his originals is in lxix, 1, 3.

^{page 195 note 1} Schulz, , Das Kaiserhaus der Antonine und der letzte Historiker Roms (Leipzig, 1907), pp. 173 and 176Google Scholar. Schulz agrees with Heer.

^{page 195 note 2} Ein versetztes Fragment des Cassius Dio Hermes, 41 (1906), p. 623 ffGoogle Scholar.

^{page 196 note 1} See Schneider, , Beiträge zur Geschichte Caracallas (Dissert. Marburg, 1890), p. 18 ffGoogle Scholar.

^{page 196 note 2} See Schneider, p. 25 ff; Schulz, , Beiträge zur Kritik unserer litterarischen Ueberlieferung für die Zeit von Commodus' Sturze bis auf den Tod des M. Aurelius Antoninus (Diss. Leipzig, 1903), p. 109 fGoogle Scholar.

^{page 196 note 3} It is no. 373 in the excerpts περὶ ἀρετῆς, preserved in the Codex Peirescianus first published by Valesius (1634), and lately by me (Excerpta de virtutibus et vitiis, pars 2, Berolini, 1910), p. 394Google Scholar.

^{page 197 note 1} No better connexion is obtained if, in accordance with Baaz, p. 50, note 120, who for the rest follows Bang, the second half of our fragment is not inserted in lxxvii, 22, 2, after ἀπέκτϵινϵ, but in the same chapter at the beginning of the third paragraph.

^{page 197 note 2} cf. Wobst, Büttner, Die Anlage der historischen Encydopädie des Konstantinos Porphyrogenetos (Byzant. Zeitshr. vol. xv, pp. 88–120)Google Scholar.

^{page 198 note 1} The transpositions of some fragments from the last books of the Archaeology of Dionysius of Halicarnassus, suggested by Maji and adopted by Kiessling and Jacoby in their editions, has been discussed by me in Mnemosyne, xxxviii (1910) pp. 281 ffGoogle Scholar.

^{page 199 note 1} The description of this entry bears, as is usual with Herodian, a conventional character. A similar description is found in iv, 11, 2 ff; cf. Schulz, Beiträge, p. 102.

^{page 200 note 1} Bang justly observes (p. 627) that in Herodian's description of this massacre expressions are to be found which remind one of Dio's description of the slaughter committed in the town; compare Herodian, iv, 9, 7, with Dio, lxxvii, 23, 1.

^{page 200 note 2} The whole of Herodian's description as it stands gives one an impression that we have to deal with a fabrication highly characteristic of this author. The reference (iv, 9, 8) to the mouths in the town deserves to rank with the other passages of the Nile being coloured red with the blood shed which betray Herodian's ignorance of geography; cf. Schulz, Beiträge, p. 102. For the rest, there is nothing material in Herodian's description of the event that is not also found in Dio and in the Vita. This also applies to his statement that Caracalla pretended he would name after Alexander the phalanx to be formed out of the inhabitants. This he may have found in the source of the Vita, which is very short here (6, 2 f), or it may have been taken from Dio, who in another connexion speaks of the Ἀλεξάνδρου ϕάλαγξ (lxxvii, 7, 1.)

^{page 200 note 3} For Claudius Pompeianus see Groag in Pauly-Wissowa, Realenkyklopaedie, iii, p. 2843, no. 282.

^{page 201 note 1} For Glabrio see von Rohden in Pauly-Wissowa, i, p. 258, no. 43.

^{page 202 note 1} The view of Baaz (p. 57 f.) on the passages compared here, is entirely erroneous. He observes that with Dio, lxxiii, 3, 2–4, we should compare Herodian ii, 3, 2–3: ‘hic Dio nuntiat Pertinacem Claudium Pompeianum plurimum honorasse, quod idem de Glabrione dicit Herodianus l.l. Sed hoc de Glabrione Dio quoque scire videtur cum narret, “καὶ τοῦτο καὶ τὸν Γλαβρίωνα τὸν Ἀκίλιον ὲποίει.”’ Nevertheless Baaz thinks that no ‘consensus auctorum’ should be assumed here, for he supposes that what we read in our present text of Dio about Glabrio has been interpolated, as would appear from the words: καὶ γὰρ ἐκεῖνος τότε καὶ ἤκουεν καὶ ἔβλεπε: ‘hoc referendum est ad ea quae de Pompeiano dixerat (§2) eum Commodi temporibus propter oculorum lippitudinem run vixisse. Quae igitur de Glabrione leguntur, significant eum quoque Commodo imperante ad rem publicam non accessisse. Sed id a vero abhorret, quoniam Glabrio anno 186 p. Chr. n. una cum Commodo iterum consul fuit. Praeterea illud καὶ γὰρ ἐκεῖνος τότε καὶ ἤκουεν καὶ ἔβλεπε omnino sensu caret: nusquam enim traditur de Glabrionis aurium vel oculorum morbo.’ Baaz then supposes that the words objected to state something that is not true, and that therefore they cannot be Dio's. But after Glabrio had been consul in 186, that is to say before the worst years of Commodus, he may very well have retired into the country on account of an alleged disease of the eye or the ear, like Claudius Pompeianus. That we have no conclusive evidence for this need not surprise us, since we hear nothing whatever about him between 186 and 193. Consequently no objection can be raised to Dio's remark about him in 193, and Baaz's supposition that some reader interpolated it from Herodian (‘in summam igitur suspicionem venit ille Dionis locus nec dubitari potest, quin interpolatus sit ab homine quodam, qui Herodiani libro diligenter lecto meminerat ab eo Glabrionem hoc loco nominari’) is devoid of all probability. Equally unfounded is his final remark: ‘firmatur autem nostrum iudicium eo, quod in Xiphilini codicibus illud absurdum deest: iniuria igitur a Boissevainio in contextum receptum est.’ This passage of Dio's we possess in a twofold tradition, first in Constantine's titulus de Virtutibus et Vitiis, viz. in excerpt no. 327 (p. 379 of my edition), and secondly in the epitome of Xiphilinus (for the words of Xiphilinus without the additions from the excerpt, see Boissevain's Dio, iii, p. 679). As is often the case, the excerpt has given Dio's words in full, while Xiphilinus has abbreviated. Hence Boissevain gives the extract in his text of Dio (iii, p. 308) and rightly describes the deviations of Xiphilinus as abbreviations made by the latter. If Baaz were right, either the excerptor must have added the words in question, which is utterly impossible, or Xiphilinus must have made use of an uninterpolated manuscript of Dio, while the excerptor on the other hand employed an interpolated one. For this no analogy has ever been met with anywhere.