Professor R. J. Tarrant notes the unexpected shift in the thought of the poem: earlier it was Ovid who was guilty of delaying in sending Tuticanus any sign of his friendship. Ovid might be postponing the real point of the letter for reasons of tact: Tuticanus has acted as though his long association with Ovid meant nothing to him, but Ovid does not want to complain of this openly, and so stresses his own failure to send Tuticanus a letter.
33-36. The set of adynata is remarkable for the way Ovid makes each of them relate to his own hardships; even Boreas and Notus have a specific connection, since Ovid complains so often of the climate of Tomis.
35. TEPIDVS BOREAS ... SIT. A comparable inversion of nature described at Ibis 34 'et tepidus gelido flabit ab axe Notus' (before Ovid will forgive his enemy).
35. PRAEFRIGIDVS appears here for the first time in Latin; it occurs later in Celsus and the elder Pliny. Praegelidus, however, is found at Livy XXI 54 7.
36. ET POSSIT FATVM MOLLIVS ESSE MEVM. The personal reference in the last element of the series of adynata is a clear break with the conventions of the topic. The last (and therefore greatest) curse in the Ibis has a similar personal reference: 'denique Sarmaticas inter Geticasque sagittas / his precor ut uiuas et moriare locis'.
37. LAPSO FHILT LASSO BCM. Lapso ... sodali seems to me the preferable reading, since it contrasts Ovid's former life in Rome with his disgrace and exile; but lasso is well attested and can be construed easily enough. Unfortunately, parallels from the poems of exile are of little use, since in most of them the one word could easily be read for the other: 'tu quoque magnorum laudes admitte uirorum, / ut facis, et lapso [uar lasso] quam potes adfer opem' (EP II iii 47-48), 'fac modo permaneas lasso [uar lapso], Graecine, fidelis, / duret et in longas impetus iste moras' (EP II vi 35-36), 'regia, crede mihi, res est succurrere lapsis [uar lassis], / conuenit et tanto, quantus es ipse, uiro' (EP II ix 11-12), 'digne uir hac serie, lapso [uar lasso] succurrere amico / conueniens istis moribus esse puta' (EP III ii 109). Professor R. J. Tarrant cites similar variants in the text of Seneca at HF 646 & 803 and Thy 616 & 658.
A clear decision can be made, however, for the phrase res lassae; it is certified as the correct term by the parallel phrase res fessae, for which see Aen III 145 'quam fessis finem rebus ferat' and Aen XI 335 'consulite in medium et rebus succurrite fessis', cited by Luck at Tr I v 35. For res lassae in Ovid, compare Tr I v 35 'quo magis, o pauci, rebus succurrite lassis', Tr V ii 41 'unde petam lassis solacia rebus?', EP II ii 47 'nunc tua pro lassis nitatur gratia rebus', and EP II iii 93 'respicis antiquum lassis in rebus amicum'; in each of these passages lapsis is found as a variant for lassis. Similarly, the sixth-century codex Romanus reads lapsis at Virgil G IV 449 'uenimus hinc lassis quaesitum oracula rebus'.
38. HIC CVMVLVS NOSTRIS ABSIT ABESTQVE MALIS. Festus defines cumulus as a heap added to an already full measure (s.u. auctarium, 14 Muller, 14 Lindsay). The transferred sense is common in Cicero (Prou Cons 26, S Rosc 8, Att XVI iii 3), and is found elsewhere in Ovid at EP II v 35-36 'hoc tibi facturo uel si non ipse rogarem / accedat cumulus gratia nostra leuis' and Met XI 205-6 'stabat opus: pretium rex infitiatur et addit, / perfidiae cumulum, falsis periuria uerbis'.
38. ABSIT ABESTQVE. The more natural abest absitque cannot be placed in a pentameter.
39. PER SVPEROS, QVORVM CERTISSIMVS ILLE EST. Similar line-endings at Ibis 23-24 'di melius! quorum longe mihi maximus ille est, / qui nostras inopes noluit esse uias' and EP I ii 97-98 'di faciant igitur, quorum iustissimus ipse est, / alma nihil maius Caesare terra ferat'.