Parallela Graeca et Romana 20A: Sources and Narrative Structure

This paper focuses on the Greek tale of Parallela Graeca et Romana’s chapter 20. The differences in the Greek story between the direct and the indirect tradition perhaps depend on the intent of Parallela’s author to emphasize its parallelism with the Roman tale. As for the source, the direct tradition indicates Euripides’ Erechtheus, while the indirect tradition (Stobaeus and Clement of Alexandria) indicates Demaratus’ Tragodoumena. The Greek tale is probably founded on the account of Erechtheus’ plot in Lycurgus’ Adversus Leocratem 98-100, perhaps integrated with information from other sources. The divergence between the direct and the indirect tradition regarding the source can be explained (according to Jacoby’s theory) by the presence in the originary text of both the names of Demaratus and Euripides (the latter introduced by μέμνηται, as in the pairs of sources in De fluviis). It is possible that in the originary text Demaratus was introduced to justify the presence in the story of information missing in Euripides. Key-Words: Plutarch, Parallela Graeca et Romana, Euripides, Erectheus, Demaratus, Indirect tradition. Riassunto Il contributo si occupa della narrazione greca del capitolo 20 dei Parallela Graeca et Romana. Le differenze nella narrazione greca tra la tradizione diretta e la tradizione indiretta dipendono forse dall’intento dell’autore dei Parallela di enfatizzare il parallelismo con la narrazione romana. Per quanto riguarda la fonte, la tradizione diretta indica l’Eretteo di Euripide, la tradizione indiretta (Stobeo e Clemente Alessandrino), invece, i Tragodoumena di Demarato. La narrazione greca si basa probabilmente sull’esposizione della trama dell’Eretteo in Licurgo, Contro Leocrate 98-100, forse integrata con informazioni provenienti da altre fonti. Le divergenze tra tradizione diretta e tradizione indiretta nell’indicazione della fonte possono essere spiegate (concordemente con la teoria di Jacoby) con la presenza nel testo originario sia del nome di Demarato che di quello di Euripide (quest’ultimo introdotto da μέμνηται, come nelle coppie di fonti citate nel De fluviis). È possibile che nel testo originario il nome di Demarato fosse stato introdotto per giustificare la presenza nel racconto di informazioni assenti in Euripide. Parole-chiave: Plutarco, Parallela Graeca et Romana, Euripides, Eretteo, Demarato, Tradizione indiretta. Ploutarchos, n.s., 15 (2018) 43-58 ISSN 0258-655X Received: 25th October 2018 Accepted: 30th Decembeer 2018


ISSN 0258-655X
Ploutarchos, n.s., 15 (2018) [43][44][45][46][47][48][49][50][51][52][53][54][55][56][57][58] T he pseudo-Plu tar chan work Pa ralle la Grae ca et Ro mana 1 consists, as is known, of a brief in troduction and 41 pairs of tales, one regarding the Greek world and the other the Roman world 2 . It came down to us both through the direct tradition 3 , a part of which (family Σ) is in the form of an epitome, and through the indirect tradition, primarily through Stobaeus, John Lydus and (in some cases) Clement of Alexandria. With respect to the indirect tradition, the Parallela offer a generally more concise presentation, although there are cases in which they contain information not present in the indirect tradition or where there are qualitative differences 4 . These relationships between the direct tradition and the indirect tradition have been explained by Jacoby under the hypothesis that there exists a single recension, on which depend both Stobaeus and John Lydus as well as (through an epitome) the text of the Parallela in its more complete form and in the further epitomated form of the Σ family 5 . 1 The work is generally considered not authentic on the basis of external (cf. above all Á. Ibañez Chacón, 2012) and internal criteria (cf. D. Wyttenbach, 1821, pp. 78-79; K. Ziegler, 1965, pp. 276-277;E. Valgiglio, 1992, p. 4024;M. Cannatà Fera, 2000, p. 388;B. Scardigli, 2004;contra cf. G. D'Ippolito, 1999, p. 340, who thinks that the work can be considered authentic because it belongs to the genre of paradoxograpy). 2 On the nature and the subject of the work, cf. A. De Lazzer, 2000, pp. 21-27. 3 The text of the work is included in the corpus Plutarcheum already in Par. gr. 197 (F), X-XI sec., used thereafter by Planudes for his Plutarch's edition in Ambr. 126 C inf. (α), 1294-1295cf. Á. Ibañez Chacón, 2016, p. 27;Á. Ibañez Chacón, 2017, pp. 341-343 (both with previous bibliography). The inclusion of the Parallela minora in the corpus Plutarcheum has been explained on the basis of its topics, which probably seemed analogous to that of Vitae parallelae, but perhaps also to those of Aetia Graeca and Aetia Romana (cf. C. Müller, 1861, p. LII;F. Jacoby, 1940, pp. 79-81;F. Carlà-Uhnik, 2017, p. 2610). The work is generally identified with διηγήσεις παράλληλοι, Ἑλληνικαὶ καὶ Ῥωμαικαί, n. 28 in Lamprias' catalogue, but, if the identification is correct, it can prove only that in the III-IV sec. d.C. the work was attributed to Plutarch (cf. A. De Lazzer, 2000, pp. 18-19; and, for the problematic nature of the catalogue, Á. Ibañez Chacón, 2012, p. 160). 4 For a status quaestionis on the relationship between the direct and indirect tradition and a description of the manuscripts cf. A. , pp. 82-139. 5 F. Jacoby, 1940. The reconstruction of the relationship between the direct and indirect tradition proposed by Jacoby has recently been accepted by A. De Lazzer, 2000, pp. 86-89; however, he refutes the hypothesis (made by Jacoby) that underlying the tradition there is a συναγωγή, "quale testo dotto, infarcito di informazioni e citazioni accumulate senza un criterio compositivo" and also expresses doubts on the hypothesis of a single recension. Also R.M. Piccione, 1998, pp. 170, 174-177, 180 thinks that the text of the Parallela and of Stobaeus depend autonomously on a common source.

Ploutarchos, n.s., 15 (2018) 43-58 ISSN 0258-655X
Of the numerous sources cited in the text of the Parallela and in the indirect tradition, many are probably invented authors (Schwindelautoren in Jacoby's definition 6 ) and only a few can definitely be held to be authentic 7 ; among these is Euripides, who is cited in 20Aa 8 , 310D and in 26A, 312A (with an indication of the title of the works, the Erechtheus and the Meleager, respectively) and in 24A, 311D (with no indication of the title, but with a clear reference to the plot of the Hecuba) 9 . As for what concerns the sources, the text poses two main problems: 1) to establish which of the otherwise unknown authors are fictitious and which are actually existed; 2) for the already known authors, to establish if the tales are really taken from their works (eventually in combination with other sources) or are (partially or to tally) invented (this problem concerns also the unknown authors recognized as real ones) 10 .
The aim of this paper is to examine the case of 20A, 310D, which is of par ticular interest as we possess the in direct tradition both for the Greek tale (Stob.,III 730 = 20Ab and Lyd.,Mens. 4.147 = 20Ac) and for the Roman tale (Lyd.,Mens. 4.147 = 20Bb;Clem.Al.,Protr. 3.42,7 = 20Bc). The Greek tale summarises the central episode of Euripides' Erechtheus, where the cha racter of the same name sacrifices a daughter in obedience to an oracle that prophesied his victory in the war against the Thracians led by Eumolpus 11 if he did so, while the parallel Roman tale reports that Marius at the time of the war with the Cimbri acted in the same way following a warning received in a dream.
Demaratus in book III of Tra go doumena; when Eumolpus, king of Thracia, was at war with Erechtheus, lord of Attica, he learned from an oracle that he would defeat his enemies if he sacrificed his eldest daughter to Persephone. On reaching Athens, he told his wife Praxithea of the prophecy of the Pythia and led his daughter to the altars where he sacrificed her; he was then victorious in battle.
Erechtheus, lord of Attica, performed this deed, out of trust not in a dream but an oracle, and defea ted his enemies 14 .
When Marius was fighting the Cimbri and was being worsted, he saw in a dream that he would conquer if he sacrificed his daughter before the battle; for he had a daughter Calpurnia. Since he placed his fellow-citizens before the ties of nature, he did the deed and won the victory. And even to this day there are two altars in Germany which at that time of year send forth the sound of trumpets. So Dorotheüs in the fourth book of his Italian History 16 . 13 Translation of this and other passages from the Parallela in F.C. Babbitt, Plutarch.
Moralia, IV, Cambridge (Mass.) -London, 1936. 14 The translations of the passages from Stobaeus and John Lydus are mine. 15 Μάριος is transmitted only by families Φ and Σ, whereas the other manuscripts have Μάνιος. Μάριος is also the reading of the indirect tradition. 16 The text of the epitome is here omitted as it is outside the scope of the present paper.
When Marius the great was fighting the Cimbri and the Teutons, he saw in a dream that he would defeat his enemies, if he sacrificed his daughter to the Averters of evil. By placing the in terests of his fellow-citizens before those of his family, he did the deed and defeated his enemies.
Erechtheus the Athenian and Ma rius the Roman sacrificed their own daughters: the former to Persephone, as Demaratus relates in the first book of his Subjects of Tragedy, the latter, Marius, to the "Averters of evil", as Dorotheus relates in the fourth book of his Italian History 17 .
As far as the Greek tale is concerned, the text of the Parallela is (as usual) more concise than that of Stobaeus. However, the many differences or omissions do not seem to be random but, rather, tend to emphasis the parallelism of the Roman and the Greek tales, which is consistent with the aim of the work to select similar events (ὅμοια 305B): 1. the use of the generic ἔμαθε instead of the reference to the oracle (χρησμὸν ἔλα βεν) eliminates the difference with respect to the Roman tale, where Marius receives the information in a dream (ὄναρ εἶδεν; cf. also John Lydus κατ' ὄναρ ... οὐκ ὀνείρῳ ἀλλὰ χρησμῷ) 18 ; 2. the indication τὴν θυγατέρα, with the omission of Stobaeus' specification that it was the eldest daughter who is sacrificed (τὴν πρεσβυτάτην τῶν θυγατέρων), corresponds to the identical expression present in the Roman tale of the Parallela (similarly in the indirect tradition of the Roman tale τὴν ἑαυτοῦ θυγατέρα in John Lydus, τὰς αὑτῶν ... θυγατέρας in Clement of Alexandria, with reference to the daughters of both characters); 3. the use of προθύω 19 in both the Greek and the Roman tales (compared to the simple θύω in the indirect tradition and, only in the Greek tale, in the epitome):

ISSN 0258-655X
Ploutarchos, n.s., 15 (2018) [43][44][45][46][47][48][49][50][51][52][53][54][55][56][57][58] in this case, it seems uncertain whether this is a conscious choice on the part of the author of the Parallela or whether the compound verb was already present in the originary text and the preverb was omitted in the indirect tradition (and, for the Greek tale, in the epitome); 4. finally, we cannot exclude the possibility that in the direct tradition the omission of the different deities to whom the two girls had been sacrificed (Persephone for the daughter of Erechtheus in Stobaeus and Clement of Alexandria and the Averters of evil [ἀποτροπαίοις] for the daughter of Marius in John Lydus and Clement of Alexandria) is also attributable to the desire to make both tales as similar as possible.
Just how complex the situation is and why we cannot talk about coherent epitomation 20 for the text of the Pa ralle la is indicated by two cases (in the Roman tale) in which, compared to the indirect tradition, we find further information 21 that cannot be explained as interventions or 'inventions' of the au thor of the Parallela: 1. the indication of the name of Marius' daughter, Calpurnia (while the daughter of Erechtheus remains ano ny mous); 2. the reference (totally extrinsic to the core of the parallel tales) to the altars which emit the sounds of the σάλπιγξ.
It is likely, then, that the Roman tale of the Parallela provides us with information present in the originary text but omitted in the indirect tradition, which usually offers a fuller and more detailed account.
A particularly major problem regards the source of the Greek tale, which the Parallela identify as the Erechtheus of Euripides. An initial hypothesis might indicate an indirect knowledge of the tragedy through the oration Against Leo crates of Lycurgus (98-101, p. 67 f. Conomis) 22 , which also contains the citation of an ample rhesis made by Erechtheus' wife Praxithea in favour of the sacrifice of their daughter (fr. 12 Sonnino = 14 Jouan = 360 Kannicht) 23 . The following arguments support this hypothesis: 1. a substantial coincidence between the account of the plot in Lycurg., 98-99 (which mentions the war against Eumolpus, Erechtheus' wife -Praxithea -, the oracle, the sacrifice of their daughter, the fact that the enemies were subsequently driven from the region χρήσαντος δ' αὐτῷ τοῦ θεοῦ, τὴν θυγα τέρα εἰ θύσειε πρὸ τοῦ συμβαλεῖν τὼ στρατοπέδω, κρατήσειν τῶν πολεμίων, ὁ δὲ τῷ θεῷ πιθόμενος τοῦτ'ἔπραξε, καὶ τοὺς ἐπιστρατευομένους ἐκ τῆς χώρας ἐξέβαλε "The god's answer to him was that if he sacrificed his daughter before the two sides engaged he would defeat the enemy and, submitting to the god, he did this and drove the invaders from the country" 24 ) and the tale of the Parallela and of the indirect tradition; 2. in the Roman tale (which is not otherwise recounted and which, as in other cases 25 , seems to have been in vented along the lines of the Greek story 26 ) the observation that Marius placed the interests of his fellow-citizens before those of his family (προκρίνας ... τῆς φύσεως τοὺς πολίτας) presents a clear analogy with the text of Lycurgus, in which Euripides is prai sed for choosing to represent a myth that provides the citizens with an incentive to love their country (τὸ τὴν πατρίδα φιλεῖν 100, τοὺς γ᾽ ἄνδρας ἀνυπέρβλητόν τινα δεῖ εὔνοιαν ὑπὲρ τῆς πατρίδος ἔχειν 101 "men should show towards their country a devotion which cannot be surpassed") in the same way as Praxithea, who loved her country more than her daughters (τὴν πατρίδα μᾶλλον τῶν παίδων φιλοῦσαν 101) 27 . The expression προκρίνας τῆς φύ σεως τοὺς πολίτας seems to echo the vocabulary of the rhesis of Praxithea cited by Lycurgus, which contains several references to the πόλις (ll. 5, 8, 11, 16, 21, 23, 27, 40, 42, 47, 52) and to the πολῖται (ll. 13, 50) and where in l. 38 the woman states that her daughter is not her πλὴν <εἰ> φύσει "except in nature's way" 28 (cf. also φύσει γὰρ οὐσῶν φιλοτέκνων πασῶν τὼν γυναικῶν 101 "All women are by nature fond of children").
If the hypothesis that the Greek tale is derived from the text of Lycurgus is correct, the expression εἰ θύσειε πρὸ τοῦ συμβαλεῖν τὼ στρατοπέδω (Lycurg., 99) might have led to the use of the com pound προθύω in the Parallela and (perhaps) in the originary text as well.
On the contrary, other elements seem to cast doubt on the use of Lycurgus' text: 1. the fact that Erechteus refers the content of the oracle to his wife (present both in the Parallela and in Stobaeus), although implicit in the rhe sis of Praxithea (from which it may theoretically have been deduced), is absent in Lycurgus' account of the plot;  Connelly, 1996, pp. 79-80. 28 On the value of the expression, the text of which is corrupted, cf. M. Sonnino, 2010, pp. 280-281, ad loc., which considers as being noteworthy the integration εἰ of F. Ferrari, 1978, p. 234.
We thus cannot exclude the possibility that underlying the Greek tale (instead of or as well as the text of Lycurgus) is a knowledge of the story of the Erechtheus 29 , either first-hand or (more probably) through the mediation of another source that has since been lost. That the tragedy (or at least its plot) was known in the first few centuries AD (probable time in which the Parallela were written) 30 is shown first of all by the citation (partially integrated, but virtually certain) of the work in a commentary on Thucydides, found in a papyrus scroll (POxy VI, 853, col. X 2 ad 2.15.1 [... ἐν Ἐρε]χθεῖ Εὐριπί[δης ...) dating from the 2nd century AD and written probably in the 1st century AD, and secondly by a concise allusion to the plot in Aristid., Panathen. 87: λέγεται γὰρ Ἐρεχθεὺς μὲν ἐν τῷ πρὸς Εὔμολπον τούτῳ πο λέμῳ τὴν θυγατέρα ὑπὲρ τῆς πόλεως ἐπιδοῦναι, τοῦ θεοῦ χρήσαντος, προσαγαγεῖν δ' αὐτὴν κοσ μήσασα ἡ μήτηρ ὥσπερ εἰς θεωρίαν πέμπουσα· Erechtheus is said in this war against Eumolpus to have given his daughter on behalf of the city because of the god's oracle; and her mother is said to have led her forth after adorning her as if for a festival 31 .
That Aelius Aristides may have been the source of the Greek tale can probably be excluded, firstly because the passage makes no mention of Euripides or the tragedy and, secondly, because they lack such elements as the name of Erechtheus' wife and his victory over his enemies, while the active role played by the girl's mother does not seem to find correspondence in the Parallela or in Stobaeus 32 . The mention of the myth in Aelius Aristides as an example of the ability to set aside personal interest in favour of Athens (Panathen. 86) 33 and the citation of fr. 13 Sonnino = 15 Jouan = 360a Kannicht (φιλῶ τέκν', ἀλλὰ πατρίδ' ἐμὴν μᾶλλον φιλῶ "I love my children, The most problematic aspect of the source of the Greek tale is that Stobaeus and Clement of Alexandria attribute it to Demaratus' Tragodoumena. It is important to establish first if this au thor can be considered authentic or invented, like many others unknown au thors cited as sources in the Parallela and in the indirect tradition (cf. supra, p. 45). Other works by Demaratus are cited as sources in the Parallela (16A, 309D-E = 42 F5 Jacoby) and in the De fluviis (9.5; 9.3 = 42 F6-7 Jacoby), a pseudo-Plutarchan work characterised by considerable similarities to the Parallela (collection of isolated mythological tales, citation of otherwise unknown sources -some of which common to the Parallela -, pre sence of rare expressions and of parti cular syntactic turns) 37 and which, L. Di Gregorio, 1980, p. 55 n. 76;M. Sonnino, 2010, p. 288. 35 The verse has been transcribed several times in the margins of the text of the Praecepta gerendae rei publicae in the Par. gr. 1957;cf. J.-C. Carrière, 1984, p. 104. 36 The citations of this and other passages from the Erechtheus in Plutarch (Nic. 9 = fr. 10, are not necessarily proof that the two authors had direct knowledge of this tragedy, as they could have found them in anthological collections; see P. Carrara, 1977, p. 12;L. Di Gregorio, 1980, p. 55 (who notes that the citation of fr. 13 Sonnino = 15 Jouan = 360a Kannicht in Plutarch shows that it must have been well known in the early 2 nd century AD); M. , p. 249. Whereas A. Tuilier, 1968  some claim, is attributable to the same author 38 . Demaratus is also attributed, in the scholia on Apollonius Rhodius and in the Bibliotheca of Pseudo-Apollodorus, with some fragments regarding the saga of the Argonauts (42 F1-3 Jacoby) 39 . Ini tially, Jacoby had hypothesised that these fragments might belong to the Tragodoumena, thus implicitly admitting the authenticity of Demaratus and this work 40 ; but he later accepted Wendel's proposal 41 to attribute them to the my thographer Demaretes 42 , based on the fact that the manuscript tradition of F 1-2 Jacoby presents the form Δημαρέτης (-ετος), while Δημάρατος is conjectured 43 . If the fragments 42 F1-3 Jacoby of Demaratus are attributed to Demaretes, falls the main ar gument for the authenticity of Demaratus himself, cited only in the Parallela and in the De fluviis 44 ; he is therefore likely to be an invented author 45 .
This passage is hence one of the five cases in the Parallela in which the direct and the in di rect tradition cite completely diffe rent sources 46 , which in this case are authentic (Euripides) and probably in vented (Demaratus). These cases we re explained by Jacoby 47  40B, 315F ... ὡς Ἀριστείδης Μιλήσιος καὶ Ἀλέξανδρος ὁ Πολυΐστωρ ἐν τρίτῳ Ἰταλικῶν. "So Aristeides the Milesian, and also Alexander Polyhistor in the third book of his Italian History"), where the epitomation activity may have been less significant. In the passage in question, Jacoby's hypothesis seems to be supported by the introduction of the reference to Euripides through the verb form μέμνηται, which is an unicum in the Parallela, where the reference to the source is generally introduced by ὡς, καθά o καθάπερ, at times accom pa nied by ἱστορεῖ or φησιν 48 . In the grammatical and erudite tradition, μέμνηται is often used to introduce the reference to a source after giving some information and, in particular, in those cases where an author mentions facts that are also (or primarily) known through another, previously cited author; particularly significant in relation to this passage is the sch. E., Tr. 228, which reports the information acquired by Parmeniscus (ὡς Παρμενίσκος φησιν) regarding the existence of two rivers named Crathis and then points out that Euripides (in the passage to which the comment refers) mentions the second of the two (οὗ νῦν μέμνηται ὁ Εὐριπίδης). As Jacoby observed 49 , in the De fluviis it is a highly common practice to juxtapose two authors as sources, introducing the second through the indication μέμνηται δὲ τούτων 50 . Similarly, in this case we can hypothesis that the mention in the originary text of Demaratus as the source was followed by the indication μέμνηται Εὐριπίδης ἐν Ἐρεχθεῖ: the indirect tradition conserved only the first reference and the Parallela only the second 51 .
The attribution of the tale to a fictitious source (Demaratus) and on ly secondarily to the authentic one (Eu ripides) in the text on which both the Pa rallela and the indirect tradition depend could explain the presence in the latter of some elements that (inasmuch as can be deduced from the fragments and the testimony in our possession) seem to be absent in the Erechtheus and of which there is in any case no trace in the text of Lycurgus. Specifically: 1. the indication of Persephone as the goddess to whom the daughter of Erechtheus is sacrificed 52 . According to Diggle and Collard, this information might 48 Á. Ibañez Chacón, 2012, p. 163 and n. 20 seems to underestimate the difference. 49 F . Jacoby, 1940, pp. 133-134. 50 De fluv. 1.3;9.3;9.4;9.5;11.4;14.3;17.4;18.3;cf. also A. De Lazzer, 2003, pp. 64-66. 51 Á . Ibañez Chacón, 2012 hypothesises that the indication of the Erechtheus of Euripides as the source in the Parallela might have been inserted by the epitomator who, in summarising a well-known story, may have preferred to make reference to an important and famous author instead of an almost unknown writer like Demaratus. 52 On the presence of this element in the indirect tradition cf. Á. Ibañez Chacón, 2012, p.
2. the specification that the eldest daughter was sacrificed: fr. 12 Sonnino = 14 Jouan = 360 Kannicht cited by Ly curgus mentions two other daughters (δύο ... ὁμοσπόρω) in l. 36 but does not specify which was destined for the sa crifice. Moreover, the information con trasts with the testimony of Apollod., III 15.4, who indicates that the youngest of the daughters (τὴν νεωτάτην) was sacri ficed 55 .
It is therefore possible that the compiler of the originary text started from the version of the myth present in Euripides (which he probably knew through the mediation of Lycurgus and/or other authors) and expanded it with the inclusion of these elements (invented or gleaned from other sources unknown to us). The divergence between the direct and the indirect tradition regarding the source can be explained (according to Jacoby's theory) by the presence in the originary text of both De maratus Tragodoumena (probably in vented to justify the presence in the story of infor mation missing in Eu ripides 56 ) and Euripides Erechteus (in tro duced, as in the pairs of sources in De fluviis, by 162. According to A. Martínez Díez, 1975, p. 237, ἐὰν τὴν πρεσβυτάτην τῶν θυγατέρων Περσεφόνῃ θύσῃ (Stobaeus) can be interpreted in the sense that «la hija al morir pasaría a ser considerara potencia ctónica, o equivalente del semple hecho del morir». 53 Cf. Ch. Collard and J. Diggle apud Ch. Collard -M.J. Cropp -K.H. Lee, 1995, p. 188 ad loc. (Cropp does not put the conjecture in the text); J. Diggle, 1997, p. 106. The proposal was taken into consideration by M. Sonnino, 2010, p. 361 ad loc., who highlights the relationship that could be established with the testimony of 'Demaratus' (cf. supra, n. 45). 54 On the reasons for the sacrifice to Kore in the Heraclidae and on the possible analogies with the Erechtheus cf. J. Wilkins, 1995, pp. 104-105 ad loc. 55 On the various versions of the myth with regard to the overall number of Erechtheus' daughters and to the ones involved in the sacrifice cf. A. Martínez Díez, 1975, pp. 232-234;J.B. Connelly, 1996, pp. 60-61 and n. 52;F. Jouan -H. Van Looy, 2002, p. 97;M. Sonnino, 2010, pp. 100-102. 56 Cf. A. De Lazzer, 2000, pp. 46-47, in the exposition of F. Jacoby, 1940, p. 135. A similar case is that of 21A, 310E, where the Parallela present major differences compared to the text of Parthenius (indicated as the source), while in Stobaeus the tale is attributed to a μέμνηται, which does not imply an exact match between the tale pre sented and the Euripides tragedy).
At this point we can envisage two different hypotheses for the text on which the Parallela and the indirect tra dition depend: 1. a single and more extensive tale, similarly to what we read in Sto baeus (containing therefore also the indi cations on Erechtheus' eldest daughter and Persephone). This would have been followed by the indication of the Tragodoumena of Demaratus as the source and the reference to Euripides introduced by μέμνηται. The author of the Parallela (or the epitome from which the text of the work is derived) may then have summarised the tale, omitted some details (including those not present -as far as we know -in Euripides' text) and indicated Euripides alone as the source. This reconstruction has the advantage of a direct juxtaposition of the names of the two sources in the originary text (as in Parallela 1B and 40B and in various passages from the De fluviis 57 ); moreover, the presumed epitomation would find ample confirmation throughout the Parallela.
2. A larger tale, attributed to De maratus, may have been followed by a more concise tale, attributed to Euripides, following the Erechtheus more closely. Sto baeus would then have reported the first tale (summarised in Clement of Alexandria and John Lydus), while the Parallela would have followed the second. The fact that the Parallela lack that information which (as far as we know) did not even appear in Euripides' text would then not be the result of epitomation, but would depend on using Euripides as (direct or indirect) source. This reconstruction is in line with the hypothesis of Jacoby 58 , according to which the cases of double citations might be explained as the result of the desire to juxtapose an invented story alongside the known tale of a known author or (as in our case) to tell it with the addition of variations.
Although the first reconstruction undoubtedly appears more plausible, the problem (which is likely to remain as such) nevertheless shows the comcertain Sostratus, a probably fictitious author (or invented starting from the name of a real author): see the analysis of A. Ibáñez Chacón, 2010, which on p. 63 hypothesises the existence of an intermediate source between Parthenius and the Parallela (cf. also F. Jacoby, 1940, pp. 130, 140-143). 57 Cf. supra p. 53 and n. 50. 58 F. Jacoby, 1940, p. 135. Jacoby's theory is questioned by A. De Lazzer, 2000, pp. 46-49, above all with reference to the possibility of accurately identifying the origin and meaning of the double citations.