Plus in amore ualet Mimnermi uersus Homero (Prop. 1.9.11). As Fedeli’s note illustrates, there are parallels for comparing carmina breuia with Homeric epic, and no doubt in casting this line in such strong terms Propertius was deliberately being provocative. The reference to Mimnermus, nonetheless, reflects an important fact of literary history, namely the powerful influence exerted by this early elegist on the later literary tradition. He was regarded as a founder of the genre of love elegy by Hellenistic poets, and figures prominently in the prologue to Callimachus’Aetia (fr. 1.9-12 Pf., a passage usefully discussed in Appendix B to the book under review). In this light, Mimnermus stands with Hipponax and Antimachus as an early poet whose exiguous literary remains are overshadowed by an exalted reputation. Those remains, however, are of absorbing interest, and suggest that Mimnermus was an interesting poet in his own right. Accordingly, a new edition of Mimnermus with full commentary—the first devoted solely to this poet—needs no justification.
Allen has profited greatly from the recent labours of M. L. West, B. Gentili and C. Prato, whose critical editions have set the study of the poet’s text on a securer footing. 1 A critical edition, however, even one with an apparatus as detailed as that of the Teubner, is no substitute for a commentary in which issues of text and interpretation can be discussed discursively, and it is the commentary that will make A.’s work essential reading for students of early Greek poetry. In fact, A. shows himself to be the model of the conscientious commentator; all the relevant problems are subjected to a patient, critical examination with judicious citation of secondary literature. That our understanding of Mimnermus does not emerge radically altered should not be regarded as a failing; A.’s achievement lies in the collection and consolidation of the pertinent material, thus providing a firm basis for future scholarship.
The introduction is preceded by a useful collection of testimonia. This selection of illustrative material makes no pretence to being exhaustive, as does the uncritical compilation of S. Szádeczky-Kardoss, Testimonia de Mimnermi uita et carminibus (Szegedini 1959), nor has A. annotated these texts critically, as did Gentili-Prato their similar collection; rather, the testimonia serve conveniently to underpin the discussion of the issues treated in the introduction. Here A. sets out what is known concerning the poet’s date, life, name, and writings. He argues clearly for a floruit in the middle of the seventh century, setting his birth ca 670 B.C., and surveys the evidence rooted in fr. 9 for Mimnermus’ connections with Smyrna and Colophon, arguing with West against the Alexandrian tradition that the latter was the poet’s city.
So far as Mimnermus’ writings are concerned, A. contends plausibly that the Alexandrian edition filled no more than a single papyrus roll and that Nanno was a general title for the collection of Mimnermus’ poems which comprised not only erotic and sympotic elegy, but also the Smyrneis. This historical elegy has recently received important illumination from the appearance of the new Simonides papyrus that is contained in volume LIX of the Oxyrhynchus Papyri (1992), which appeared too late to be taken account of by A. (these texts have been re-edited by West in the second edition of IEG). As reconstructed by West ( ZPE 98  1-14), the remains of the poem on the battle of Plataea (frs 10-17 West 2) suggest a similar work on the kind of ample scale—perhaps several hundred lines long—imagined for Mimnermus’Smyrneis. Particularly striking are the possible implications of frs 10 and 11 that Simonides’ elegy was prefaced with a formal hymn (to Achilles, on West’s view); this is reminiscent of Pausanias’ statement (9.29.4 = Mimn. fr. 14 = 13 W 2) 2 that Mimnermus’ poem on the battle of the Smyrnaeans against Gyges and the Lydians had a προοίμιον in which the poet mentioned two generations of Muses. In addition, that the new Simonides was found among the remains of a roll that also contained sympotic poems supports A.’s contention that Mimnermus’ poem was not long enough to fill an entire papyrus roll, and so may well have formed part of the Nanno. It is, of course, possible that testimony concerning Mimnermus has influenced reconstruction of the Simonides papyrus in certain ways, but the important possibility remains that we now have evidence for a poem that was closely parallel to the Smyrneis.
The title Nanno remains mysterious. Tradition tells us that Nanno was Mimnermus’ αὐλητρίς and lover, but authority for this does not appear to go back any earlier than Hellenistic poets such as Hermesianax (fr. 7 Powell) and Posidippus ( AP 12.168 = 3086-3092 Gow-Page HE). These sources do not inspire confidence; after all, in the same poem Hermesianax detected in the familiar ἠ’ οἵη of the Hesiodic Catalogue the name of the poet’s lover, Eoia (fr. 7.24). But A. is probably right in saying that “It is hardly likely … that she is wholly fictional, a late classical or Alexandrian invention” (p. 18). It seems unlikely that an early book would bear a title like Nanno, and so it is probable that the collection was given this designation in the Alexandrian period. That Posidippus mentions Nanno in close proximity to Antimachus’ Lyde is suggestive, perhaps reflecting a link between the two poets. West sees that connection in terms of textual history, raising the possibility that the Nanno came to Alexandria in an Antimachean edition ( Studies in Greek Elegy and Iambus [Berlin and New York 1974] 75), but I find this unlikely; Antimachus’ activity as an editor has been much overrated. 3 More likely is the position that we are dealing here with the detritus of a Hellenistic construct of literary history, according to which Mimnermus and Antimachus were claimed as predecessors for Alexandrian poetic practice. Nanno was elevated to the position of the poet’s mistress from some minor reference or references, perhaps in a poem placed near the beginning of the collection (possibly an invocation or prelude in which Nanno is named as αὐλητρίς). 4
A. prints a text of each fragment with context and an apparatus criticus. In the wake of recent editions, there is little room for radical improvement; A.’s text is sober and generally well considered. This does not mean, however, that A. tamely reproduces the standard editions; he has clearly thought through each textual problem in a sensible and independent manner. At fr. 1.6, for example, he follows Gentili-Prato (and the paradosis) against West in printing καλόν, but he improves on the Teubner text by accenting ὅμως paroxytone instead of perispomenon (γῆρας, O(/ T’ AI)SXRON ὅμως KAI KALON TIQEI=). This is clearly a superior reading, presenting a more pointed line (“… old age, which makes even the handsome man ugly”), while avoiding the dubious understanding of ὁμῶς endorsed by Gentili-Prato. The text of both IEG and the Teubner is rejected at fr. 2.1, where A. prints Bergk’s οἷά τε φύλλα φύει πολυανθέος ὥρῃ (the usual version reads πολυάνθεμος ὥρη with the majority of the MSS 5). A. argues that this text avoids the unmarked change of subject in line 2 (φύλλα is regularly understood as object of φύει, but subject of αὔξεται) and that the intransitive use of φύει finds a precise parallel at Il. 6.149, the passage that is widely regarded as Mimnermus’ model. Bergk’s text is plausible, but the alteration strikes me as unnecessary. φύει in the Iliadic passage is a dubious parallel: it is the only instance of the present stem used intransitively in Homer, and, as Kirk notes ( ad loc.), it stands in awkward proximity to familiar transitive φύει in the preceding line; accordingly, Kirk finds Brandreth’s φύετ’ attractive. In the elegiac fragment, it is more natural to construe φύει as transitive. The “uncomfortable change of subject” in the O(/TE-clause that troubles A. is not intolerable; Homer provides a number of examples. An instructive parallel is Od. 13.20-23:
καὶ τὰ μὲν εὖ κατέθηχ’ ἱερὸν μένος ἀλκινόοιο,
αὐτὸς ἰὼν διὰ νηός, ὑπὸ ζυγά, μή τιν’ ἑταίρων
βλάπτοι ἐλαυνόντων, ὁπότε σπερχοίατ’ ἐρετμοῖς.
These lines exhibit the same shift in syntax as the Mimnerman passage: τά (like φύλλα, neuter plural) is the object of the verb in the main clause, but without any signal becomes the subject of βλάπτοι in the MH/-clause (the subject shifts again with σπερχοίατ’). These lines seem to have confused some scribes, who corrupted τιν’ to TIS in a few MSS; but there is no reason to adjust the text.
At fr. 2.10 A. replaces the transmitted βέλτιον with βέλτερον. This is an attractive conjecture, although it must be noted that A. has been anticipated in proposing it by H. Friis Johansen (see Friis Johansen-Whittle on Aesch. Suppl. 1069). This emendation assumes a corruption that finds an exact parallel (unmentioned by A. or Friis Johansen) at Od. 17.18, where βέλτιον has displaced βέλτερον in a couple of later MSS; but there is little cause to judge βέλτιον suspect in the present passage, in which the paradosis is unanimous. Citing Chantraine’s discussion of Homeric comparatives in –ίων ( Grammaire homérique [Paris 1958] 1.257), A. ( ad loc.) points to the “antiquity and priority of βέλτερος,” but it is hard to see how this point is decisive in a post-Homeric poet like Mimnermus. The controversy over the scansion of comparatives in –ίων, to which A. alludes, concerns the admission of forms with short iota to Attic tragedy, but such forms are regular in Ionic, Doric, and Attic comedy (in addition to the authorities cited by A., cf. Headlam-Knox on Herodas 2.91; Collard on Eur. Suppl. 1101b-1102a); there is nothing odd about the scansion of βέλτιον in Mimnermus. More relevant is Friis Johansen’s observation that BELTίων is not otherwise securely attested until late in the fifth century.
Special difficulties are raised by fr. 5, in which the excerpt in Stobaeus is complemented and possibly supplemented by Theognidea 1017-1022. Here A. follows the Teubner editors in printing only the lines preserved by Stobaeus, rejecting the possibility supported by West and others that the Theognidea transmits a fuller version of the text. On this point A. seems to be much influenced by D. Young’s dogmatically unitarian view of ‘Theognis,’ according to which the Megarian adapted pieces of poems by other poets to make constituent elements of his own poems (“Even if one cannot readily embrace Young’s belief in the full unity of the sylloge, it is hard not to conclude … that somebody—’Theognis’—tampered with Mimnermus’ lines on old age and made them part of his own poem” [p. 60]). The influence of Young is also apparent on p. 70, where A. cites Thgn. 793-796 as Theognis’ “adaptation” of Mimermus fr. 7, citing with approval Young’s judgment that the ‘substitution’ of τοί for τίς in 796 was made by Theognis for reasons of euphony (it should be noted that in IEG West prints the Theognidean lines as two separate couplets). Young’s views on the Theognidea have always struck me as a rather desperate—at times, even perverse—bid to claim all that is transmitted under the name of Theognis for that poet. A. (following Young) believes that “there can be no doubt that some poets occasionally used and revised the work of other poets, dead or alive” (p. 60), but it is unlikely that this practice was as purposeful or common as A. supposes. The “revision” of Solon fr. 6.3 by ‘Theognis’ noted by Clement (and adduced by A., p. 60) is more likely a variation on a common theme of moralizing poetry than a conscious reworking of an earlier text. Solon’s (fr. 20) reply to Mimnermus, on the other hand, seems to belong to a different category, since in that case Solon explicitly names the earlier poet (cf. also Sim. fr. 19 West 2). Moreover, the subtle adjustments that ‘Theognis’ is alleged to have made to these ‘borrowed’ texts (see A. p. 60 for Mimn. fr. 5) seem to presuppose a preoccupation with the letter of the text that is most unlikely in the archaic and classical periods, when performance was almost certainly the principal means of publication, and such variants are better regarded as indicative of the vagaries of transmission—either oral or literate.
All this ultimately obscures a rather important point about early elegy, and that is that we owe a large amount of what survives to the work of anthologists, and this probably means that we have a highly selective picture of the genre. Our single most important source for Mimnermus is the anthology compiled by Stobaeus, which preserves most of the longer passages (frs 1-5, 8, 14, [24-25]), and it must be remembered that Stobaeus excerpted texts according to discernible criteria, ostensibly to serve as an aide-memoire for his son, but he also clearly assembled his anthology with a view to moral edification; he selected portions of texts that seemed to him best to illustrate some general point, and so we have a large number of gnomic passages. 6 I do not believe that we have many complete poems from this source, however whole some passages may seem. A., for example, believes that frs 1 and 2 are complete (p. 32), but is forced to argue for inceptive δέ in both cases. The new Simonides papyrus can teach a cautionary lesson here: what Stobaeus gave to us as one continuous text beginning with the pentameter ἓν δὲ τὸ κάλλιστον Χῖος ἔειπεν ἀνήρ we know now to be made up of two excerpts (frs 19-20 West 2). West suggests ( ZPE 98  10) that these two passages were originally separated by a heading, τοῦ αὐτοῦ, but it is also possible that we possess the results of the telescoping of a longer elegy, an unnerving prospect. It is also likely that certain subjects are less well attested than would be the case if we possessed a broader selection of early elegy. Athenaeus and Strabo preserve tantalizing glimpses of what appears to have been rich mythic narrative (Mimn. frs 11 and 12). Stobaeus, however, seems to have had little interest in myth; from Theognidea 697-718 he excerpts two purely gnomic passages, 699-702 and 717-718, omitting the most interesting section, the striking mythic paradigms that fill the intervening lines. What would almost certainly find no place in the anthologies is material concerned exclusively with the particular social setting or occasion of the poems, and this may in part account for the absence of Nanno from the extant fragments. For example, we know from a variety of sources that Archilochus composed an elegy or elegies on the victims of shipwreck (frs 8-13 West 2), but fr. 13, preserved by Stobaeus, contains general reflection only. In addition, there are occasions when Stobaeus altered texts to create a more general reference (e.g., Herodas 1.67 and 6.37, where vocatives are generalized): see Hense, RE 9.2584.
In the commentary the reader will find a wealth of information on a wide range of interpretative issues; the poetic habits of Mimnermus are also discussed more synoptically in Appendix A. Like most commentators on early elegy, A. is much concerned to examine Mimnermus’ affinities with the language of Homeric epic, and, accordingly, A. carefully sets out the Homeric pedigree of forms and phrases. Occasionally A. pursues these concerns to the exclusion of issues of meaning. An interesting example is fr. 1.9, where in discussing the phrase ἀτίμαστος δὲ γυναιξίν α. notes that ἀτίμαστος occurs only here and that ἄτιμος is the Homeric word, but he offers no discussion of how ἀτίμαστος is suited to the erotic context; Solon fr. 24.5 (παιδός τ’ ἠδὲ γυναικός) and Tyrtaeus fr. 10.29 (ἐρατὸς δὲ γυναιξί), both adduced by A., tell us nothing of the rôle of τιμή in women’s attitudes towards their lovers. Perhaps the best parallel is fr. 5, in which ἥβη is characterized as τιμήESSA and γῆρας described as ἐχθρὸν ὁμῶς καὶ ἄτιμον, but the context is generalized. Anacreon’s girl from Lesbos may dishonour a prospective lover in fr. 358 PMG by rejecting him, but no form of τιμή used. In fact, it is striking that τιμή does not seem to be part of the vocabulary of heterosexual love in early Greek poetry. Mimnermus’ ἄτιμαστος γυναιξίν is very likely sound, but I crave some discussion of the meaning of the phrase. In this light, West’s γυναικί (“hateful to his children and dishonoured by his wife”) might deserve more serious consideration than the quick dismissal it receives from A. Although it is unclear just how the passage as a whole would cohere, the shift from the erotic pursuits of youth to the breakdown of relations within the family would have interesting affinities with fr. 3, in which early beauty gives way to discord between father and sons. The connection between hexameter and pentameter in this latter fragment is also puzzling.
In general A.’s notes are convincing and helpful, but no commentator can persuade the reader on every point, and I mention a couple of passages where I believe that a different view is possible. Fr. 2.4-5: τοῖς [sc. φύλλα] ἴκελοι πήχυιον ἐπὶ χρόνον ἄνθεσιν ἥβης , πρὸς θεῶν [[scanned as one syllable]] εἰδότες οὔτε κακὸν . This last clause has attracted considerable attention from scholars. A. rejects the view that Mimnermus is here referring to ignorance of evil (or ignorance of the difference between good and evil), **preferring to see in the passage a reference to “ignorance of imminent, fateful evil” (p. 44). In this light, οὔτε κακὸν is understood as a polar expression with κακόν the only meaningful component; A. judges ἀγαθόν to be “simply rhetorical.” It always strikes me as dangerous to treat any word or phrase as effectively void of meaning. In the present passage I would prefer to take EI)DO/TES OU)/TE KAKON / OU)/ T’ ἀγαθόν closely with πρὸς θεῶν: “knowing neither bad nor good that comes from the gods.” As is often the case, such language refers broadly to our dispensation from the gods which includes both bad and good (cf. Il. 24.527 ff.); we enjoy our brief time of youth because we cannot foresee the vicissitudes of fortune that will undercut our happiness. The reference then will be to what Herodotus calls the κύκλος τῶν ἀνθρωπηίων πρηγμάτων (1.207), according to which good fortune must yield to bad. This reading is consonant with the rest of the fragment, in which there is a return to the idea of dispensation at lines 15-16, οὐδέ τίς ἐστιν , ᾧ ζεὺς μὴ κακὰ πολλὰ διδοῖ. Here the emphasis is on the inevitable κακά. Gentili-Prato aptly quote Od. 4.236 f., A)TAR QEOS A)/LLOTE A)/LLW| / ZEUS ἀγαθόν TE κακόν TE DIDOI=, which also provides a close verbal parallel for OU)/TE KAKON OU)/T’ ἀγαθόν.
Fr. 12.5 ff. τὸν [sc. ἠέλιον] μὲν γὰρ διὰ κῦμα φέρει πολυήρατος εὐνή … (8) εὕδονθ’ ἁρπαλέως χώρου ἀφ’ ἑσπερίδων … . In his note on εὕδονθ’ ἁρπαλέως α., like most recent commentators, takes participle and adverb together (rather than the adverb with φέρει), translating ‘sleeping pleasurably.’ This understanding of ἁρπαλέως has been advocated by M. S. Silk, CQ 33 (1983) 326-328, who states that “The truth is that in all its classical occurrences ἁρπαλέος means ‘pleasing’ and ἁρπαλέως means ‘gladly’ (i.e., the subject of the verb is ‘pleased to …’).” This understanding, however, strikes me as too restrictive and, in some cases, misleading. In the LSJ entry, which Silk largely rejects, the Mimnerman passage was the only instance in which the adverb was rendered ‘gladly, pleasantly’; other early instances were understood in senses approximating more closely to that of ἁρπάζω, from which ἁρπαλέος was believed to derive. That etymology has now been rejected as secondary in favour of derivation from the root of ἄλπνιστος and ἔπαλπνος, but the word was early connected with the ἁρπ root (cf. Frisk and Chantraine s.v.), and, whatever the primary meaning of ἁρπαλέος, this popular etymology seems to have affected usage (cf. also Garvie on Od. 6.250). In this light, it is probable that the word connoted something stronger than ‘pleasant.’ When applied to a noun such as κέρδεα ( Od. 8.164), it suggests that gain is positively attractive or desirable (Garvie [ ad loc. ] aptly translates, “profit which they greedily seize”). The phrase ἥβης ἄθεα … ἁρπαλέα (Mimn. fr. 1.4) is comparable: the flowers of youth exert a strong attraction. When the adverb is used, it seems to suggest eagerness or intensity. The Homeric examples illustrate the point well. At Od. 6.249 f. Odysseus is given his first meal since washing up on the shore of Phaeacia, and he falls to it with gusto (ἦ τοι ὁ πῖνε καὶ ἦσθε … / ἁρπαλέως), and the reason for his eagerness is given in line 250 (δηρὸν γὰρ ἐδητύος ἦεν ἄπαστος). Silk’s understanding of ἁρπαλέος is simply inadequate to the context. Similarly at 14.109 f., Eumaeus gives Odysseus his first meal after returning to Ithaca: O( D’ ἐνδυκέως KRE/A T’ H)/SQIE PI=NE/ TE OI)=NON / ἁρπαλέως A(KE/WN, κακὰ δὲ μνηστῆρσι φύτευεν. Odysseus has just learned in detail of the activities of the suitors. Once again the hero eats with intensity (both ἐνδυκέως and ἁρπαλέως), but there is also a dangerous silence, as he plans destruction for the suitors. There is little pleasure here. In the case of Mimnermus’ description of Helius as EU(/DONQ’ ἁρπαλέως, I suggest that the adverb connotes intensity and means something like ‘soundly’, suggesting a deep sleep following his daily πόνος. We might compare Odysseus at the end of Od. 5, τῷ δ’ ἄρ’ ἀθήνη , ἵνα μιν παύσειε τάχιστα , φίλα βλέφαρ’ ἀμφικαλύψας. What is remarkable in the Mimnerman passage is the fact that Helius, whom Pindar describes as ἁγνὸς θεός ( Ol. 7.60), endures ever-lasting πόνος as his lot (cf. line 1)—a condition that is the antithesis to the usual state of divinity in early Greek thought (cf. Aesch. Suppl. 100, πᾶν ἄπονον DAIMONI/WN)—and the description of the god as sleeping soundly after his toil reinforces this point.
A few marginalia. P. 33 (on fr. 1.1): A.’s note wrongly implies that Aphrodite alone is called ‘golden’: this may be true of epic, but not later texts. Bacchylides, for example, calls Artemis χρυσέα δέσποινα λαῶν (11.117). P. 34 (on fr. 1.3): κρυπταδίη φιλότης. Cf. Pind. Pyth. 9.39, κρυπταὶ κλαίδες ἐντὶ σοφᾶς the same ode also provides a parallel for μείλιχος in an erotic context, μείλιχος O)RGA/ (43): see the discussion by A. Köhnken in A. Hurst (ed.), Pindare (Entretiens sur l’Antiquité classique 31: Vandoeuvres-Geneva 1984) 71-111, at 89 f. P. 37 (on fr. 1.6): it is incorrect to speak of ὅμως as “Verdenius’ emendation,” when, as A. notes, it was proposed by Doederlein early in the last century. P. 52: A.’s introductory note is somewhat odd; he recognizes in κάλλιστος an implication of homosexual ἔρως, but it is hard to see how this coheres with the pentameter, in which the issue concerns a father’s relation to his sons (see above). P. 60 (fr. 5): “One might object, too, that there is a grotesqueness in the description of erotic sweat [at Thgn. 1017] which is out of place in a reflective elegy of the seventh century.” What about Sappho fr. 31.13 Voigt, a lyric on a similarly high stylistic level? P. 62 (on fr. 5.3): for the description of Tantalus’ punishment in Pind. Ol. 1.56 ff., see R. D. Griffith, “The Mind is its Own Place: Pindar, Olympian 1.57 f.,”GRBS 27 (1986) 5-13. P. 67 (on fr. 6.2): if κίχοι indicates personification, should not μοῖρα be capitalized, as editors of Aeschylus do at Cho. 911? P. 73 (on fr. 8.1): for the early meaning of ἀληθεία, see also C. H. Kahn, The Verb ‘Be’ in Ancient Greek ( Foundations of Language Suppl. 16, Dodrecht 1973) 363 ff. P. 74 (on fr. 9.4): on the striking phrase ὕβριος ἡγεμόνες, see now N. R. E. Fisher, Hybris: a Study in the Values of Honour and Shame in Ancient Greece (Warminster 1992) 214 ff. P. 91 (on fr. 11.3): for Mimnermus’ ὑβριστῇ Πελίῃ α. rightly notes the relevance of Pindar’s description of Pelias as ἄθεμις ( Pyth. 4.109). It is, however, pertinent to point out that in the same passage Pindar also emphasizes the king’s ὕβρις: (of Jason’s parents) ὑπερφιάλου (i.e., Pelias) δείσαντες ὕβριν (111 f.). P. 92 (on fr. 11.5 and fr. 15.11): ὠκέος ἠελίοιο. Citing fr. 2.7-8, A. notes that the epithet here likely “refers to the swiftness of the Sun’s course across the sky.” This may well be correct, but in both Mimnerman passages the adjective occurs in close proximity to mention of the Sun’s ἀκτῖνες, and it is possible that there is also some suggestion of heat, as in the case of the similar use of rapidus in Latin (a parallel noted by A.): see D. R. Shackleton Bailey, Propertiana (Cambridge 1956) 317; Mynors on Verg. Georg. 4.425. P. 93 (fr. 11.7): θεῖος Ἰήσων. A. describes the adj. as “ornamental”, but such descriptions are hazardous, especially in discussing a poet like Mimnermus. Here θεῖος may underline the point made earlier (possibly in the lost protasis of the opening section) that Jason could not have made his journey successfully without divine aid. It is uncertain what version of the Argonautic myth Mimnermus was following. The Naupactia, reflecting an influential tradition, seems to have portrayed Aphrodite as Jason’s special helper (frs 6 Bernabé = 7A Davies; cf. A. R. 2.424, ἐν γὰρ τῇ κλυτὰ πείρατα κεῖται ἀέθλων): see V. J. Matthews, Phoenix 31 (1977) 199; for other traditions, see Campbell on A. R. 3.8. P. 94 (fr. 12.2-3): Aeschylus seems to have echoed these lines in fr. 192.3-6 Radt, ἵν’ ὁ παντόπτης ἥλιος αἰεὶ [Griffith’s καμάτων ἵππους is attractive] / θερμαῖς ὕδατος , with χρῶτ’ … ἵππων picking up ἵπποισίν τε καὶ αὐτῷ and ἀναπαύει recalling ἄμπαυσις. P. 96: Alex. Eph. should be cited as fr. 38 SH. P. 101 (on fr. 12.3): the material preserved in Hyginus Fab. 183 (p. 153 Marshall) seems to possess no independent value from the evidence of the Titanomachia, since recent editors believe that Hyginus is referring to that poem (fr. 7 Bernabé = 4B Davies). P. 101 ( ibid.): on ῤοδοδάκτυλος, see now E. Irwin, “Roses and the Bodies of Beautiful Women in Early Greek Poetry,”EMC n.s. 13 (1994) 1-13. P. 107 (on fr. 12.9): on the Ethiopians, see now B. MacLachlan, “Feasting with the Ethiopians: Life on the Fringe,”QUCC 40 (1992) 15-33.
The book is clearly produced. There is a relatively small number of typographical errors and slips; few will cause any confusion. I mention some of the more noteworthy: οἶ ( app. crit. to fr. 1.4, p. 31); τρυχοῦται (text of fr. 2.12, p. 40); χρυσοστεφάνοιο (p. 46); τοῦ ( H.Aph. 237, quoted p. 54); αἴ γάρ ( lemma to fr. 6.1, p. 67); Theogn. 795 (p. 70); ἵπποισίν (text of fr. 12.3); A. W. James (not ‘Jones,’ p. 102); Od. 5.984 (p. 107). The author-date method of citing secondary sources inevitably produces omissions from the bibliography: Gentili 1965 (p. 73) is Maia 17, 366 ff.; Boedeker 1974 (p. 102) should be Aphrodite’s Entry into Greek Epic (Leiden).
It is the function of the commentator both to illuminate the text at hand and to lay a solid foundation for further scholarship. A. has accomplished these tasks with conspicuous success, producing an elegant libellus that will long remain indispensable to students of early elegy and of Greek poetry in general.
 M. L. West, Iambi et elegi graeci 2 (Oxford 1972; second ed. 1992); B. Gentili and C. Prato, Poetae elegiaci: testimonia et fragmenta 1 (Leipzig 1979; repr. 1988). The second edition of West appeared too late to be consulted by A.
 A.’s numeration of the fragments is essentially that adopted by West; in what follows I note any discrepancies.
 I propose to discuss Antimachus’ edition of Homer elsewhere.
 Cf. the brief invocations placed at the beginning of the Theognidean syllogy (1-18).
 Bergk’s text is supported by one ms. , which gives πολυάνθεος, but the false accent may suggest that this reading is itself a corruption of πολυάνθεμος.
 See D. A. Campbell, “Stobaeus and Early Greek Lyric Poetry,” in D. E. Gerber (ed.), Early Greek Poetry and Philosophy: Studies in Honour of Leonard Woodbury (Chico 1984) 51-57.