^{page 343 note 1} Canterbury Tales, B, 60 ff.

^{page 343 note 2} Studies in Chaucer, i, 418–19.

^{page 343 note 3} xxiv, 124 ff.

^{page 343 note 4} Ed. Michel, ii, 83–84; ii, 118.

^{page 344 note 1} Benoit says that Æneas went to Lombardy (vv. 28127–30) and Guido that he “peruenit italiam et in tusciam se recepit” (ed. 1489, sig. n. 4, 2 verso). Neither of them speaks of Lavinia. The Roman de la Rose (ed. Michel, ii, 321) mentions her along with Helen: “N'onques Helaine ne Lavine Ne furent de color si fine.”

^{page 344 note 2} Or in Guido. For the sake of simplicity, I confine the argument to Benoit. In my opinion Benoit, rather than Guido, was in Chaucer's mind in this passage of the Book of the Duchesse, as well as in other passages of that poem (to be quoted presently). But, if Guido be substituted for Benoit, the considerations which I shall urge will be equally cogent.

^{page 345 note 1} It will not help matters to refer to Benoit's initial summary, vv. 151 ff. If Chaucer had read these lines, he would probably have read vv. 1199 ff. as well. For convenience, I use Joly's numbering of the lines in Benoit.

^{page 345 note 2} The mss. have Antilegius, or the like. The emendation, which improves the metre, is adopted by Skeat in his Index of Proper Names (Oxford Chaucer, vi, 360).

^{page 346 note 1} See also Guido, ed. 1489, sig. 1 3, verso—1 4, recto.

^{page 346 note 2} Benoit uses the form Dares (as well as Daires) elsewhere. We have no critical text of this part of the Roman de Troie. So far as I know, he does not employ the full form Dares Phrygius anywhere, but the name was well-known.

^{page 346 note 3} Guido does not cite Dares here.

^{page 346 note 4} See also Guido, sig. m ff. In a passage of the Troilus (i, 197–210) which is not from the Filostrato Chaucer recurs to the treason of Antenor.

^{page 346 note 5} See also Guido, sig. m 4, 2 verso, col. 2.

^{page 347 note 1} Guido has merely “Cassandra vero quasi demens affecta sola fugit et mmerue templum intrauit vbi suorum omnium excidium grauiter lamentatur.” Virgil, Aen., ii, 403–6, would not have given Chaucer the hint. There is another passage in Benoit (vv. 10355–84 Joly) which may have been in Chaucer's mind. It mentions both “Ylion” (“Abatuz sera Ylion,” v. 10366) and “Troie” (“Ha riche Troie, a quel eissil Sereiz dessi qu'a poi livrée,” vv. 10380–1), as Chaucer does. Cf. Guido, sig. h 3, verso (when, however, Ilion is not separately mentioned). This lament, like the other, is in neither Dictys nor Dares.

^{page 348 note 1} It will hardly be contended that Chaucer read Benoit's summary account of Jason, Hercules, Medea, and Laomedon in vv. 151–63 and did not go on to vv. 989 ff.

^{page 348 note 2} The fact that Chaucer was familiar with what Jean de Meun says of Medea is of course no reason for believing that he had not also read Benoit's account of her. There is nothing inconsistent in the two narratives. Benoit simply does not tell that part of Medea's story which concerns the murder of her children, and Jean de Meun does tell it (ii, 84). There is nothing in Benoit that would discredit Jean de Meun.

^{page 348 note 3} See Dr. Karl Young's investigation, The Origin and Development of the Story of Troilus and Criseyde (Chaucer Society, 1908), where full references to previous studies may be found.

^{page 348 note 4} Troilus, i, 652–65; see Her., v, 139–46, and (for Apollo and Admetus) cf. Ars. Am., ii, 239–42.

^{page 349 note 1} See p. 351, below.

^{page 349 note 2} “For-why men rede That love is thing ay ful of bisy drede” (Troilus, iv, 1644–5) is Ovid's “Res est solliciti plena timoris amor,” as Francis Junius noted, centuries ago, in his copy of the Folio of 1598 (Bodleian Library, ms. Jun. 9). See Heroides, i, 12 (Penelope to Ulysses). The Filostrato has no such line, but it was a stock quotation. It is found, for example, in the Ovidian flosculi given by Vincent of Beauvais, Speculum Historiale, vi, 107 (Venice, 1494, fol. 67 recto).

Further evidence of Chaucer's reading in Ovid is seen in the use which he makes of the Remedium Amoris (cf. B. Duch., v. 568) in the Troilus. (1) “For who-so list have helping of his leche, To him behoveth first unwrye his wounde,” i, 857–8; cf. R. A., 125–6: “Adgrediar melius tum, cum sua vulnera tangi Iam sinet;” (2) “For thilke ground that bereth the wedes wikke, Bereth eak thise holsom herbes, as ful ofte Next the foule nettle, rough and thikke, The rose wexeth, swote and smothe and softe,” i, 946 ff.; cf. R. A., 45–46: “Terra salutares herbas eademque nocentes Nutrii, et urticae proxima saepe rosa est.” (3) In R. A., 135 ff., Ovid prescribes occupation as a remedy for love (see especially vv. 139, 144, 149–50, 205–6); next comes absence (214 ff.); somewhat later the patient is directed to seek a new love in order to recover from the old (451 ff.), and this advice is fortified by the example of Agamemnon, who got rid of his passion for Chryseis by transferring his affections to the captive (“si prima sinat syllaba, nomen idem,” 476) whom he took away from Achilles—“et posita est cura cura repulsa nova” (484). Compare Troilus, iv, 421 ff.:—

None of the passages quoted under 1, 2, and 3, is in the Filostrato. Nos. 1 and 2 are in the flosculi in Vincent of Beauvais (Speculum Historiale, vi, 144, fol. 68 recto-verso, ed. Venice, 1494), but the parallel in No. 3 is pretty significant, especially in view of the interest that Chaucer would of course take in Ovid's testimony about Chryseis. (4) “And eek, as writ Zanzis, that was ful wys, ‘The newe love out chaceth ofte the olde,‘” iv, 414–15, is Ovid's “Successore novo vincitur omnis amor” (R. A., 462, quoted by Skeat), which comes just five lines before the Chryseis passage; but here Boccaccio intervenes: “E come io udii già sovente dire, Il nuovo amor sempre caccia l'antico” (Filostr., iv, 49). The lines stand also among Vincent's flosculi (vi, 115, fol. 68 verso).

Pandar's advice to Troilus about the love-letter (ii, 1023 ff.) is partly from Ars Am., i, 457–68. Possibly, as Skeat suggests, Trail., ii, 1027, is reminiscent of Her., iii, 3 (an epistle which Chaucer certainly knew when he wrote the Troilus).

In the Parliament of Fowls, vv. 253–6, Chaucer shows his acquaintance with Fasti, i, 415–38 (Priapus).

^{page 350 note 1} I have believed (and taught) for years that the House of Fame is earlier than the Troilus. But the relative age of the two poems makes no difference to the present argument, since both are older than the Man of Law's introduction.

^{page 350 note 2} See Skeat's notes (Oxford Chaucer, iii, 251–2), where interesting marginalia from mss. may be found.

^{page 350 note 3} See Root, pp. 130–31.

^{page 350 note 4} Heroides, ii, 13, 78, 81–83, 89, 108, 111–12.

^{page 350 note 5} Her., ii, 23–24, 31–44, 53.

^{page 351 note 1} Her., iii.

^{page 351 note 2} Her., vi.

^{page 351 note 3} Her., ix.

^{page 351 note 4} Her., x.

^{page 351 note 5} P. 134, note 1.

^{page 351 note 6} Here, as elsewhere, I have of course made the freest use of Professor Skeat' s notes and other apparatus.

^{page 351 note 7} Troil., iv, 1543–5 (“Satyry and Fauny,” “halve-goddes”); Met, i, 192–3; of. Boccaccio, Gen. Deor., viii, 13 (ed. 1511, fol. 64 verso). Not in Filostr. or R. R.

^{page 351 note 8} Troil., iii, 729–30 (Herse, Aglaurus); Met., ii, 708 ff.—Troil., iii, 1703–4 (Pyrois, etc.); Met, ii, 153–4.—Troil., v, 663–5 (Phaethon); Met., ii, 1 ff. (None of them in Filostr. or R. R.)

^{page 351 note 9} Troil., iv, 1538–40 (Athamas); Met., iv, 447 ff. Not in Filostr. or R. R. For Chaucer's form Athamante see the acc. Athamanta, Met., iv, 466, 470.—Troil., v, 211–12 (Ixion); Met., iv, 460. The information about Ixion might have been derived from R. R., ii, 272; but note that Met., iv, 460, stands in the story of the madness of Athamas, which is not in R. R.

^{page 352 note 10} Trail., v, 319 (Ascalaphus); Met., v, 534 ff. Not in Filostr. or R. R. The Man of Law mentions the Metamorphoses in his introduction (B, 92–93), and speaks of the Pierides or false Muses (see Met., v, 294–678).

^{page 352 note 11} H. F., vv. 1229–32 (Marsyas); Met., vi, 382–400. Dante, Par., i, 20–21, is insufficient. A passage which Michel gives from some fifteenth-century mss. of R. R. is full enough (i, 360), but does not afford the name Marcia; it has Marse (Dante, Par., i, 20–21, Marsia; Met., vi, 400, “Marsya nomen habet”).—Troil., ii, 64–70 (Progne and Tereus); Met., vi, 412 ff. Not in Filostr. or R. R.—Troil., i, 699–700 (Niobe); Met., vi, 312. Not in Filostr. or R. R.

^{page 352 note 12} H. F., vv. 919–24 (Icarus); Met., viii, 183 ff. R. R., i, 173–4, and Dante, Inf., xvii, 109–11 (cf. Par., viii, 125–6) are insufficient.—Troil., v, 1457 ff. (Atalanta); Met., viii, 271 ff. Not in Filostr. or R. R.

^{page 352 note 13} Troil., iv, 1138–9 (Myrrha); Met., x, 298 ff. (especially 500–2). Not in Filostr.; R. R., ii, 332–3, is insufficient.—H. F., vv. 589–92 (Ganymede); Met., x, 155–161.

^{page 352 note 14} B. Duck, vv. 62 ff. (Ceyx and Alcyone); Met., xi, 410 ff.—H. F., vv. 69–76 (Morpheus); Met., xi, 592–614.—Troil., iv, 789–91 (permanent reunion of Orpheus and Eurydice); Met., xi, 61–66. (Not in Filostr. or R. R.) Chaucer's “feld of pitee” is Ovid's “arva piorum,” and his “Elysos” may have come from amarginal “campos Elysios.”—Chaucer's mention of Midas's ears in Troil., iii, 1388–9 (not from Filostr.), might of course come from R. R., i, 360, if the latter passage were not (as it seems to be) an interpolation. But his characterization of Midas as “ful of coveityse” is another matter, for R. R. does not speak of the Golden Touch. Chaucer had read Ovid's account (Met., xi, 100 ff.); for covetousness see Met., xi, 118–19, 132, 136, 141.—Troil., iv, 120–6 (Laomedon), may be compared, tentatively, with Met., xi, 199–208, and Her., v, 139; but things do not exactly fit (see also Bode, Scriptores Rerum Mythicarum, 1834, i, 43–44 (Mythogr. i, caps. 136, 137), 138 (Mythogr. ii, cap. 193), 174 (Mythogr. iii, cap. 5, § 7). Benoit (vv. 25814–19) speaks of the walls of Troy as built by Neptune and dedicated by Apollo (cf. Gower, C. A., i, 1152–5).

^{page 352 note 15} Ovid's House of Fame, Met., xii, 39 ff.

^{page 352 note 16} Troil., iv, 1548–53 (Simois, in an oath); Met., xiii, 324–7. Not in Filostr. or R. R.

^{page 352 note 17} Parl. F., v. 289 (Byblis); Met., ix, 454 ff. Not in R. R. Cf. Boccaccio, Amorosa Visione, xxv, 14 ff. (Biblide).

^{page 352 note 1} Troil., iv, 25 (Quirinus); Met, xiv, 772 ff. (especially 828 ff.). Cf. next note. See also H. F., v. 589.

^{page 352 note 2} Troil., iv, 25 (“Thou cruel Mars eek, fader to Quirine”); Met., xv, 863 (“invicti genitor Gradive Quirini”). Not in Filostr. or R. R. Cf. preceding note.

^{page 353 note 1} Vv. 3247–4222.

^{page 353 note 2} Vv. 143–336.

^{page 353 note 3} Vv. 271–2008.

^{page 354 note 1} ii, 84; see Root, p. 127.

^{page 354 note 2} vii, 396.

^{page 354 note 3} Benoit does not mention the murder. He says that the gods were angry with Jason for his faithlessness and avenged Medea “trop asprement” (vv. 2024–6). In the Heroides the death of the children is contemplated, but particulars are not given (vi, 159–60; cf., xii, 211–12), although there is a suggestion of actual bloodshed (“caede cruenta suo,” vi, 162, words which would be amply justified by the slaughter of Absyrtus, Her., vi, 129–30, even if Medea had not killed her children). Gower says that Medea “slew” Jason's two sons “before his eye” (C. A., v, 4210–6). Boccaccio (Amorosa Visione, xxi–ii) has much to say of Jason and his three wives (Hypsipyle, Medea, and Creusa), but does not mention the murder. In his De Geneaologia Deorum, however, he speaks of the children as killed with a sword—“gladio laniari” (xiii, 26, ed. 1511, fol. 98 recto).

^{page 355 note 1} Mr. Lounsbury himself said very frankly that the theory “has nothing in its favor that can strictly be called evidence” (Sludies in Chaucer, i, 419).

^{page 355 note 4} The fact that Chaucer refers to Medea's killing her children in the Hypsipyle part of the story rather than in the Medea part is of no consequence, since he makes the story of these two heroines one continuous narrative, under a single title (Legenda Ysiphile et Medee, Martyrum). The Man of Law does not say that we shall find a separate story of Medea, or anything in particular in such a story. We have a right to use anything that we can find in the double story, or, indeed, in the Legend of Good Women anywhere, in justifying his reference.

We may note, by the way, that Chaucer's account of the Apollonius story contains a hideous detail not to be found in Gower (“Whan he hir threw upon the pavement,” B, 85). Yet even Mr. Root, who found this discrepancy significant in 1906 (Poetry of Chaucer, p. 184, note 2), now believes that it affords no ground for thinking that Chaucer is not glancing at Gower in the passage (p. 138, note 1). One may be allowed to regard the slight discrepancy between the Man of Law's introduction and the Legend of Hypsipyle and Medea as equally non-significant.

Again, should we be troubled by a slight discrepancy in Medea's case, when, in the same introduction, there is a more striking discrepancy which no one regards as of any consequence? I refer, of course, to the Man of Law's declaring that Chaucer “no word he wryteth” of “thilke wikke ensample of Canacee,” despite the fact that, in the ballade (both versions), this unfortunate lady is quite particularly celebrated as one of the heroines who, though famous for her truth, must make no “boost ne soun” of it in comparison with Alcestis (or “my lady”). The line in question—“and Canace, espyed by thy chere”—is a manifest allusion to the eleventh epistle of the Heroides (27–36). When Chaucer wrote the ballade, he probably meant to include Canace among his Good Women, but, by the time he wrote the Man of Law's introduction, he seems to have changed his mind and did not hesitate to laugh at Gower for telling her story. However that may be, the discrepancy is more serious than that in the case of Medea. Yet nobody holds that the ballade was not written until 1390.

^{page 357 note 1} On the resemblances between the two works see Bech, Anglici, v, 365 ff.; Skeat, Oxford Chaucer, iii, xl–xlii; Macaulay, Gower, iii, 545–6; Kittredge, Modern Philology, i, 2, and vi, 438–9; Tatlock, Development and Chronology, pp. 128–30.

^{page 357 note 2} Vv. 2450 ff., 2805.

^{page 357 note 3} Vv. 2462–8.

^{page 357 note 4} Vv. 2531–5.

^{page 357 note 5} Vv. 2171 ff., 2203 ff.

^{page 358 note 1} Vv. 2891–7.

^{page 358 note 2} Vv. 2941 ff.

^{page 358 note 3} Vv. 2941–9.

^{page 358 note 4} A, vv. 410–11; B, vv. 422–3. The lines are the same in A and B.

^{page 358 note 5} In version B. In A, Alcestis is mentioned, making nineteen. If she is counted, the figures are twelve out of nineteen.

^{page 358 note 6} As follows, in the order of the ballade:—Penelope (vv. 2621 ff.), Isolde (v. 2501), Helen (v. 2529), Lucretia (vv. 2632 ff.), Polyxena (vv. 2590 ff.), Cleopatra (vv. 2572 ff.), Thisbe (vv. 2578 ff.), Dido (vv. 2552–3), Phyllis (vv. 2554–5), Canace (vv. 2587–9), Ariadne (vv. 2556–8). The omissions are Esther, “Marcia Catoun,” Lavinia, Hero, Laodamia, Hypsipyle, and Hypermnestra.

^{page 358 note 7} Vv. 2605–60.

^{page 358 note 8} Vv. 2640–6.

^{page 358 note 9} Vv. 2550–96.

^{page 358 note 10} Dido (Chaucer, Legend, iii), Phyllis (viii), Ariadne (vi), Medea (iv), Cleopatra (i), Thisbe (ii), Philomela (vii). I count Philomela and Progne as one, since the story of one involves the other. The only omissions are Hypsipyle (iv), Hypermnestra (ix), and Lucretia (v),—and Lucretia is omitted in order that she may be specially mentioned a few lines later (vv. 2632–9). Cf. Tatlock, pp. 128–9.

^{page 359 note 1} Canace and Polyxena. Penelope (also in Chaucer's ballade) is reserved (like Lucretia) for particular mention a few lines later (vv. 2621–31).

^{page 359 note 2} Vv. 2573–5.

^{page 359 note 3} Vv. 2605 ff. The interval is only eight lines.

^{page 359 note 4} Briefly, as already noted, but with reproduction of the peculiar manner of her death (viii, 2571–7).

^{page 359 note 5} iii, 1331 ff.

^{page 359 note 6} iv, 77 ff.

^{page 359 note 7} v, 3247 ff.

^{page 359 note 8} vii, 4754 ff.

^{page 359 note 9} v, 5231 ff.

^{page 359 note 10} v, 5551 ff.

^{page 359 note 11} iv, 731 ff.

^{page 359 note 12} vii, 1917 ff.

^{page 359 note 13} B, vv. 510 ff. (=A, vv. 498 ff.).

^{page 360 note 1} That is, all except Cleopatra and Philomela (see p. 361, note 2, below).

^{page 360 note 2} viii, 2647–56.

^{page 360 note 3} Vv. 2927 ff.

^{page 360 note 4} See Lowes, Publications of the Modern Language Association, xix, 593 ff.; xx, 780 ff.; Tatlock. pp. 121–2; Root, p. 143.

^{page 360 note 5} There is nothing surprising in Chaucer's neglecting to mention, in the the Man of Law's introduction, two of the legends which he had already written—Cleopatra and Philomela. He was in no wise bound to compile a complete list.

^{page 361 note 1} Even if Chaucer should never complete the Legend, he was well aware that he had mentioned Helen, Hero, and Laodamia in the ballade, and Dejanira and Briseis (with very brief summaries) in the Some of Fame (vv. 397–8, 402–4), and that, besides referring to Penelope in the ballade, he had commended her “trouthe” in the Troilus, v, 1778 (“Penelopeës trouthe and good Alceste”; see also Anelida, vv. 81–82). Thus, if worst came to worst, and he never finished the Legend of Good Women, he could allege that he had at least spoken of every one of the heroines whom the Man of Law names (save Hermione alone)—“if not in o book,” then at all events “in another.” This, to my mind, seems quite conscientious enough for a humorist and the greatest of English raconteurs.

^{page 361 note 2} Two of the legends that we have are ignored in the Man of Law's introduction,—Cleopatra (i) and Philomela (vii). Now, though the omission of Philomela might be explained by supposing (with Mr. Root) that it had not yet been written, no such explanation will apply to Cleopatra, which Mr. Root admits was already in existence. Let us be thankful for the omission of these two names, since it proves (for anybody who needs proof in such a case) that Chaucer was not aiming at frigid mathematical accuracy. For the same reason he does not mention the legends in the order in which they stand.

Mr. Root believes that “the Legend of Good Women was not authoritatively published during the poet's lifetime” (p. 152). None of his arguments seem to me to have any force. One of them, however, may be noted. It relates to the unfinished condition of Hypermnestra. “It is inconceivable,” he says, “that [Chaucer] should have deliberately given out the ninth legend unfinished, when a dozen lines or so would have served to conclude it” (p. 151). Now the story of Hypermestra is complete. The legend ends (as we have it) with the first line of the moral:—“This tale is seid for this conclusioun.”—Three to five lines would have sufficed to complete the application. The chances are that Chaucer did finish Hypermnestra, and that we have lost these verses by accident. This is much more likely than that he should have left the legend at loose ends when ten or fifteen minutes with his facile pen would have remedied the defect.

^{page 362 note 1} See Tatlock, pp. 128–9.

^{page 362 note 2} viii, 2550–96. Cf. p. 358, above.

^{page 362 note 3} I count Progne and Philomela as one, for obvious reasons.

^{page 362 note 4} I append the number of each story in Chaucer.

^{page 363 note 1} I make haste to add that Gower's list contains also the names of Deidamia, Dejanira, Canace, and Polyxena, whose legends Chaucer had not written, so that, if taken ad amussim, it proves too much. But, after all, Gower was not writing for strict bibliographical purposes!