^{page 282 note 2} The credit goes in the first place to M. van der Valk, who saw the connection between Aethiopica 9 and Julian or. 1 and 3 (Mnem. ser. 3 ix [1941], 97–100Google Scholar); his argumentation was unfortunately not careful enough and people found ways round it.

Proof that Heliodorus was writing after 350 is not in itself proof of his dependence on Julian, but Professor Keydell has recently shown in an important article (Polychronion, Festschrift Fr. Dölger, 345–50Google Scholar) that certain details in Heliodorus's narrative have been inappropriately transferred from Julian's; the lower limit for Aethiopica 9 is therefore c. 358.

A caution must be entered against one of Colonna's arguments. The MS. V has the subscription Colonna has discovered that the source of this subscription was not Georgius Cedrenus but Theodorus Melitenus, an eleventh-century chronographer, and in deference to this gentleman he turns Theodosius into Theodosius I (379–95) and makes Heliodorus bishop of Tricca between 379 and 395. Since the man has identified two different Heliodoruses, however, it will not be a rash surmise that he has transferred to the first the Theodosius he found associated with the second (cf. Rattenbury's, exemplary discussion, vol. i, vii–xiii; the other Theodosius was Theodosius III, 716–17). If Heliodorus ever was bishop of Tricca, Colonna's date may well be right, but it will be right by accident.Google Scholar

^{page 282 note 3} Complete silence, for instance, in Weinreich, , Der griechische Liebesroman (Zurich, 1962), 34–40Google Scholar, and Lesky, , Geschichte der griechischen Literatur (Munich, 1963), 922–4.Google Scholar

^{page 282 note 4} If Colonna's apparatus is any guide, there are one or two things Rattenbury missed (6. 8. 4. 6 Colonna; Rattenbury, without remark).

^{page 282 note 5} It may be worth while setting out a list of the places where the text has been handled in detail outside the editions, since several scholars who have written on Heliodorus have overlooked previous literature (lest the same charge be levelled at the bibliography presented here, it should be said that some articles have appeared in periodicals that are not generally accessible, Italian periodicals in particular). The contributions are not of equal value; the most useful are those of Naber, Jackson, and Wifstrand (Rattenbury's review of Wifstrand’s suggestions in C.R. Ix (1946), 111 does not do them justice).Google Scholar

Struve, C. L., Opuscula selecta (Leipzig, 1854), vol. i. 245–52.Google Scholar

Naber, S. A., Mnem. N.S. i (1873), 145–69, 313–53.Google Scholar

Haupt, M., Hermes vii (1873), 372.Google Scholar

Rohde, E., Der griechische Roman (Leipzig, 1876), 458 n. 5.Google Scholar

Ph. Neimke, , Quaesliones Heliodoreae (Halle, 1889).Google Scholar

Headlam, W., Proc. Catnb. Phil. Soc. xxxvi (1893), 14Google Scholar; J. Phil, xxiii (1895), 263–5.Google Scholar

Earle, M. L., C.R. x (1896), 3.Google Scholar

Prager, P., Philologisch-historische Beiträge Curt Wachsmuth zum 60 ten Geburtstag überreichi (Leipzig, 1897), 89–91.Google Scholar

Richards, H., C.R. xx (1906), 109–13.Google Scholar

Rattenbury, R. M., C.Q. xix (1925), 178Google Scholar; C.R. xli (1927), 53–5, lii (1938), 114–15, lx (1946), 111Google Scholar; Proc. Camb. Phil. Soc. clxix (1938), 2–3, clxxviii (1941–1945), 1–2.Google Scholar

Jackson, J., C.Q. xx (1926), 33, xxix (1935), 52–7, 96–112.Google Scholar

Castiglioni, L., Mélanges Bidez (Toulouse, 1935). 80–1Google Scholar; Gnomon xiv (1938), 305–11, xvii (1940), 406–13.Google Scholar

Morel, W., Mnem. ser. 3 ix (1941), 281–2Google Scholar; Gymnasium lxx (1963), 545–8.Google Scholar

Wifstrand, A., EIKOTA 5, Bull, de la Soc. Royale de Lettres de Lund, 1944–1945, 2. 26–41.Google Scholar

Trypanis, C. A., C.R. lx (1946), 109–10.Google Scholar

Merkelbach, R., Rh. Mus. c (1957), 99–100Google Scholar; Studien zur Textgeschichte und Textkritik (Cologne, 1959), 182–4.Google Scholar

Loenertz, R.-J., Byzantion xxix–xxx (1959–1960), 1–6.Google Scholar

van Krevelen, D. A., Philologus cv (1961), 157–60.Google Scholar

Whittle, E. W., Cl. Phil. Ivi (1961), 178–9.CrossRefGoogle Scholar

Arnott, W. G., Hermes xciii (1965), 253–5Google Scholar; Philologus cix (1965), 308–9.Google Scholar

Most of the articles in this list contain emendations, and as a catalogue of emendations it purports to be complete. A few articles have been included that contain explanations of particular words or phrases or point out significant borrowings; if this. part of the list had been going to be complete, it would have been difficult to justify stopping short of recording everything that has ever been written about Heliodorus.

^{page 283 note 1} His note appears on a copy of Hirschig's text that now belongs to Mr. P. W. Martin of New College; it is due to Mr. Martin's interest and helpfulness that it can be made public. The copy seems to have been Jackson's working text of all the novelists except Heliodorus: his notes on Heliodorus stop at 1. 12. 2. 5 (at 1. 4. 2. 7 occurred to him).

^{page 284 note 1} Professor Keydell points out an instance in a similar phrase at Agathias, , Hist. 5. 3Google Scholar fin. where one MS. has

^{page 284 note 2} This clause bears a suggestive but doubtless fortuitous resemblance to Xen, . Hell. 7. 1. 15Google Scholar

^{page 285 note 1} The same defence is suggested by Professor Keydell.

^{page 287 note 1} The occurrences of hiatus in Heliodorus will be collected and examined in a later article.

^{page 287 note 2} Following Wifstrand, , EIKOTA 5. 30.Google Scholar

^{page 287 note 3} It is of no consequence whether the reading of Z is the original reading preserved undamaged or the corrupt reading of the other MSS. corrupted back to the original reading.