2. First published in Cory, W., Ionica (London 1858) 7Google Scholar, and frequently anthologised since.

3. E.g. Webster, T. B. L., Hellenistic Poetry and Art (London 1964) 108Google Scholar, talks of the ‘simplicity and seriousness’ of the poem.

4. E.g. Edmonds, J. M., The Greek Bucolic Poets (London and Cambridge, Mass. 1928), p. xGoogle Scholar: ‘The “nightingales” of Callimachus's famous little poem on Heracleitus are best explained as the name of his old friend's collected poems.’ Gow (n. 1 above) Vol. II, 192, also regards it as ‘somewhat tempting’ to suppose that aēdones was the title of a book of poems by Heraclitus.

5. Mckay, K. J., The Poet at Play (Leiden 1962) 91–2Google Scholar.

6. Fraser, P. M., Ptolemaic Alexandria (Oxford 1972) Vol. I, 579Google Scholar.

7. Snell, Bruno, in Glotta 37 (1958) 1–4Google Scholar, provides an excellent analysis of Callimachus's use of sound and metre in this epigram. This is a valuable method of approach, and should certainly not be ignored in any overall assessment of the poem's merits. Snell too, incidentally, comments (p.1) on the simple language and syntax employed.

A recent book on Callimachus (Meillier, Claude, Callimaque et son Temps, Lille 1979Google Scholar, now gives a more detailed analysis of the poem (pp. 221-225 and notes). Since the author's conclusions differ from my own, I have postponed a discussion of them until the end of the article, rather than trying to deal with each point as it arises.

8. See LSJ sub voc.

9. Callimachus gives the ‘prosaic’ phrase eis dakrua agein a new dimension by his use of the singular and by the deliberate ambiguity caused by the insertion of me between the preposition and its noun. It is both ‘he brought me to a tear’ and ‘he brought a tear to me’. The sudden shock of the result is heightened by the double motion.

10. Brooks, Cleanth and Warren, Robert Penn, Understanding Poetry (New York 1960, rep. 1976) 196Google Scholar: ‘The image in poetry is never present merely as description, as report, as documentation; it has, at least, some aura of significance, and it may have.… rather complex meanings.

11. Herodotus VI, 17.

12. Homer, , Iliad I, 475 etc.Google Scholar; Iliad I, 592Google Scholar.

13. Homer, , Iliad III, 241Google Scholar; Iliad XVIII, 134Google Scholar.

14. Homer, , Odyssey X, 174–5Google Scholar.

15. Leschē, as pointed out by Gow (n. 1 above) Vol. II, 191, can mean either (i) a place where people converse (e.g. Hesiod, , WD 493Google Scholar) or (ii) conversation (e.g. Callimachus Fr.178,16). For Gow it is ‘evident’ that Callimachus is more likely to have said (ii); but the en is difficult to defend in this sense. En leschēi einai or echesthai might be acceptable, but the prepositional phrase is difficult to defend with kataduein. Hence some have been led to accept meaning (i), and to point out that such leschai existed in Alexandria and in many other towns. Aelian (NA 6, 58) seems to echo the phrase when he says that Egyptian priests, in trying to calculate when the Phoenix will re-appear, are forced to admit hoti ton men hēlion en tais leschais kataduein agousi scholēn (that they devote their time to ‘making the sun sink in their leschai’) — i.e. that they are wasting their time in mere empty talk. However this scarcely helps, as it is by no means clear whether he means ‘in their conversations’ or ‘in their talking-places’. Virgil's phrase cantando condere soles (Ecl.9,51), on the other hand, seems clearly to imply that he took lesche to mean ‘conversation’. Faced with this problem, Gow accepts the reading ēelion, which goes back to Diogenes Laertius, and with Bentley excludes en altogether. He may well be right. If he is, the instrumental dative ‘with conversation’ seems to fit my interpretation even better than the prepositional phrase does.

16. Tetrapalai seems to have been coined by Callimachus himself. The colloquial tripalai (‘three times long ago’) is reduced to total absurdity by Aristophanes (Knights 1153ff.) with his compounds mounting to trismuriopalai (‘thirty thousand times long ago’); but there is certainly no hint of jesting in the Callimachean form.

17. Xenos/xeinos is more often used of the guest than of the host (LSJ sub voc). This might be used as an argument for placing the meetings at Alexandria, but the rule is by no means invariable.

18. The possibility that Callimachus ever travelled anywhere beyond Cyrene, his birth-place, and Alexandria, his place of work, has long been the subject of speculation without any clear conclusions having been reached. (Cf. Gow, n. 1 above, Vol. II, 152): ‘Whether he ever travelled beyond Egypt is uncertain.’) There is certainly no evidence of any sort to suggest that he was ever in Halicarnassus; but equally there is not enough evidence to prove that he was not. Fragment 178, 32-3, has been used in an attempt to show that the poet had never travelled by sea. But the lines can scarcely bear this interpretation. They are the reply of a sea-faring merchant to a ‘land-lubber’ who asks a question about his homeland: ‘Lucky you, if you lead a life that has no connection with the sea!’ This surely refers to the present life of a land-based scholar and poet. It need in no way imply that he has never in his life been anywhere by boat.

Another tantalising suggestion has been made to me in discussion by Mr. J. H. Betts. Although no very convincing reason can be put forward for Callimachus to have visited Halicarnassus, it may be pointed out that only a few miles offshore from Halicarnassus lies the island of Cos, where the summer palace of the Ptolemies is situated — and for poets such as Callimachus and Heraclitus there was another very strong reason to visit Cos, for it was the home of Philitas, a poet and literary theorist who is generally regarded as the founding father of ‘Alexandrian’ poetry. The debt of Callimachus to Philitas cannot be exactly assessed, since we know too little of the works of the older poet. But the preface to the Aetia (fragment 1,9-10) apparently makes a highly complimentary reference to Philitas, and much later Properties (III, 1,1) links the two as the poets who inspire him. There is, I think, only one possible source of evidence for a visit to Cos by Callimachus, and that is the seventh Idyll of Theocruitus. The scene of this poem is Cos, and many attempts have been made to identify characters in it with contemporary poets, among them Callimachus (= Lycidas?? =Aristis??). However, pleasant as it would be to find Callimachus (and even Heraclitus) on Cos, it has to be admitted that none of the arguments in favour of this are of any value. A personal acquaintance between Callimachus and Philitas remains no more an attractive possibility.

19. MacLeish, Archibald, Poetry and Experience (Cambridge, Mass. 1961), especially 50ffGoogle Scholar.

20. Page, D. L., Poetae Melici Graeci (Oxford 1962Google Scholar), Fragmenta adespota no. 976. Attributed by some, possibly correctly, to Sappho.

21. Hesiod, , WD 203ff.Google Scholar; Bacchylides 3,98; Euripides Fragment 588 etc. For references to nightingales and their habits, including that of singing after dark, see Thompson, D. W., A Glossary of Greek Birds (London and Oxford 1936) 16–22Google Scholar.

22. Callimachus may here be echoing Aeschylus, Choephori 394-6: ‘And when may mighty Zeus bring down his hand upon them (epi cheira baloi), alas, alas, and split their heads open?’ In this passage too we have a god ‘laying hands’ on mortals in a context linked with the inevitability of what is to come. But there is all the difference in the world between the question of Aeschylus, with its hesitant optative (baloi), and the bold affirmation of Callimachus with its triumphant future indicative (balei).

There is one other possible point of connection with the phrase ouk epi cheira balei. In a poem such as this, in which the sense and imagery are so closely united, has Callimachus any purpose in using this phrase other than to raise an echo of Aeschylus? Certainly the image is vivid and effective; the picture of Death ‘laying hands on’ his prey fits well into the overall context. But is there anything more than this? It is perhaps worth while to recall that there occurs in Demosthenes (312,2) and Herodotus (IV, 115) the phrase to epibalon (‘what is cast upon one’, ‘what falls to one's lot’), a phrase which is connected with epiballō and which is in many ways reminiscent of teon moron at the beginning of the poem. It has to be admitted that neither Demosthenes nor Herodotus uses it to refer to death. But despite this I feel the echo may be there. Again I must stress the power of the negative in the final line. Is there in the poet's last words a conscious rejection of the concept of death which was expressed in the first line? Death is not moros, not to epibalon, and in this poem the poet has proved that this is so. I cannot pretend that I am convinced by this, but I put it forward as a tentative suggestion.

23. A very different interpretation of this epigram has recently been offered. Its starting-point is the view expressed by Thomson, J. A. K. (CR 55, 1941, 28Google Scholar) that since the death of Heraclitus took place so long ago, the casual mention of it cannot have been the first time that Callimachus had heard of it. ‘The old man — for that Callimachus is now old follows from tetrapalai — may have thought that time had healed the wound. A word reopened it.’ The stumbling-block in accepting this is, it seems to me, the pou of line 4. Whether it is taken in a local sense (‘somewhere or other’) or in the vaguer sense of ‘perhaps’, ‘I suppose’, it clearly stresses the poet's lack of knowledge of his friend's fate, and so makes it much more reasonable to assume that this was in fact the first time that he had heard of it. Thomson's view, however, is the basis for further conclusions drawn by Claude Meillier (n. 7 above). His argument runs somewhat as follows. Since Callimachus had for many years known about his friend's death, pou cannot mean ‘perhaps’, and is to be accounted for by seeing it as part of a spatial opposition, complementary to the temporal opposition of present and long distant past. It must therefore mean ‘somewhere vague far away’, and the other side of the opposition is provided by the specific en leschēi, which should therefore indicate somewhere definite, and so must be taken to mean ‘in the leschē’, ‘in the Conversation-hall’. And since several such structures are known to have existed in Alexandria, the probability is that Callimachus and Heraclitus met in Alexandria rather than elsewhere. Against this interpretation it may be pointed out that when Diogenes Laertius (AP 7, 130.3) uses the identical phrase alla su men pou in a way which echoes Callimachus he clearly does not take pou to mean ‘somewhere’. And in any case Meillier's position, I feel, is no stronger than the extremely shaky foundation on which it is built.