166. Some of the more remarkable parallels have been collected by Nisard (_Études sur les poètes latins de la décadence_, i. 68-91), e.g. _Med_. 163 'qui nil potest sperare, desperet nihil'. _Ep_. v. 7 'desines timere, si sperare desieris'. _Oed_. 705 'qui sceptra duro saevus imperio regit, timet timentes: metus in auctorem redit'. _Ep_. cv. 4 'qui timetur, timet: nemo potuit terribilis esse secure'. de Ira_, ii. 11 'quid quod semper in auctores redundat timor, nec quisquam metuitur ipse securus?'-_Oed_. 980 sqq.; _de Prov_. v. 6 sqq.; _Phoen_. 146, 53; _Ep_. xii. 10; _de Prov_. vi. 7; _Herc. F_. 463, 464; _Ep_. xcii. 14. 167. The arguments against the Senecan authorship are of little weight. It has been urged (a) that the MSS. assign the author a _praenomen_ Marcus. No Marcus Seneca is known, though Marcus was the _praenomen_ of both Gallio and Mela, and of Lucan. Mistakes of this kind are, however, by no means rare (cp. the 'Sextus Aurelius Propertius Nauta' of many MSS. of that poet: both 'Aurelius' and 'Nauta' are errors), (b) Sidonius Apollinaris (ix. 229) mentions three Senecas, philosopher, tragedian, and epic writer (i.e. Lucan). But Sidonius lived in the fifth century A.D., and may easily have made a mistake. Such a mistake actually occurs (S. A. xxiii. 165) where he seems to assert that Argentaria Polla, Lucan's faithful widow, subsequently married Statius. The mistake as regards Seneca is probably due to a misinterpretation of Martial i. 61 'duosque Senecas unicumque Lucanum facunda loquitur Corduba'. Not being acquainted with the works of the elder Seneca the rhetorician, Sidonius invented a new author, Seneca the tragedian. 168. See ch. on Octavia, p.78. 169. Leo, _Sen. tragoed._ i. 89-134. 170. It is not even necessary to suppose with Leo that these were the earliest of the plays and that these metrical experiments were youthful indiscretions which failed and were not repeated. Leo, i. p. 133. 171. For a detailed treatment see Leo, i. p. 48. Melzer, _de H. Oetaeo Annaeano_, Chemnitz, 1890; _Classical Review_, 1905, p. 40, Summers. 172. See p. 39 on relation of epigrams to dramas. 173. _Ann_. xiv. 52. 174. See also note on p. 42 for Leo's ingenious, but inconclusive theory for the dates of the _Agamemnon_ and _Oedipus_. 175. There is but one passage that can be held to afford the slightest evidence for a later date, _Med_. 163 'qui nil potest sperare, desperet nihil' seems to be an echo of _Ep_. v. 7 'sed ut huius quoque diei lucellum tecum communicem, apud Hecatonem nostrum inveni ... "desines", inquit, "timere, si sperare desieris".' This aphorism is quoted as newly found. The letters were written 62-5 A.D. This passage would therefore suggest a very late date for the _Medea_. But Seneca had probably been long familiar with the works of Hecato, and the epigram is not of such profundity that it might not have occurred to Seneca independently.