^{page 124 note 1} L. G. W., 1572-4. Chaucer's lines are based on Ovid's Epistle of Hypsipyle, Heroides, 6, 159-160:

^{page 125 note 1} The disparity is briefly noticed by Bech, Anglia, 5, 374-5.

^{page 125 note 2} Studies in Chaucer, 1, 418.

^{page 126 note 1} London, Chaucer Society, 1907, p. 175. Since this article was written, the hint has been taken up by Miss E. P. Hammond in her Chaucer, A Bibliographical Manual, New York, 1908, p. 252. She advances no evidence in its favor, but seems ready to accept it on Lounsbury's authority.

^{page 126 note 2} This discrepancy Lounsbury failed to notice. Bech was aware of it, apparently, though he makes nothing of it: ‘Die auch sonst ungenaue angabe gemacht wird: Thy litel children hanging by the hals.‘ Anglia, 5, 375.

^{page 126 note 3} Medea, 1278.

^{page 126 note 4} Medea, 1006.

^{page 126 note 5} Metamorphoses, 7, 396.

^{page 126 note 6} ', Bibliotheca Historica, 4, 54, 7.

^{page 127 note 1} Baumeister, Denkmäler des klassischen Altertums, 1887, 2, 874-5, 902-8.

^{page 127 note 2} Bibliotheca, 1, 9, 28.

^{page 127 note 3} Fabulae, 25.

^{page 127 note 4} Roman de la Rose, ed. Michel, 14198-14203.

^{page 128 note 1} It would be interesting to know whence Jean de Meun's estrangla is derived. This I have not been able to determine.

^{page 129 note 1} Oxford Chaucer, 3, 272.

^{page 130 note 1} Beginning at line 14115.

^{page 130 note 2} It is in this passage that we find the account of Medea strangling her children, already quoted.

^{page 131 note 1} Skeat has pointed out Chaucer's indebtedness for this phrase, Oxford Chaucer, 3, 252.

^{page 131 note 2} Heroides, 2, 2.

^{page 132 note 1} One may notice here also the similarity between ‘for Jason’ and ‘Por ce que de Jason les ot.‘ Cf. above, p. 127.

^{page 133 note 1} Roman de Troie, ed. Joly, 2014-2022.

^{page 133 note 2} Historia Trojana, ed. 1486, cap. 6.

^{page 134 note 1} There is nothing to show that he was in any way acquainted with Ovid's version in Metamorphoses, 7, or that he even knew of its existence.

^{page 134 note 2} Anglia, 5, 324-332.

^{page 134 note 3} Oxford Chaucer, 3, xxxviii.

^{page 135 note 1} Though Ovid alludes to the episode in the letter of Hypsipyle, as noticed above, there is no suggestion of it in Medea's own letter. Medea vaguely hints at possible revenge; but only a few lines before the end of her epistle she reproaches Jason for giving to their children a cruel stepmother: ‘Si tibi sum vilis, communis respice natos: Saeviet in partus dira noverca meos.‘ Eeroides, 12, 187-8.

^{page 135 note 2} One may even notice the similarity of phrase between ‘For thy Jason’ of the Man of Law and ‘for Jason’ in this passage.

^{page 136 note 1} If one declines to accept this explanation of the discrepancy between the legend and the reference to it, one must assume either that Chaucer had forgotten his own work, or that he was deliberately misrepresenting it. The first of these alternatives is possible, but hardly probable. One may forget a good deal in the course of years; but surely it is unlikely that Chaucer should forget not only a detail, but the whole tendency of his poem. The second alternative is hardly worth considering unless on the hypothesis that he was planning a revision of the story along new lines. For such an hypothesis there is no color of probability.

^{page 136 note 2} Cf. Lydgate's well-known reference to the unfinished state of L. G. W. in his Fall of Princes (Morris's Aldine Chaucer, 1, 80).

^{page 138 note 1} The date 1390 or later for the Man of Law's Prologue rests on the assumption that the Man of Law's reference to the ‘cursed stories’ of ‘Canacee’ and ‘Tyro Apollonius’ is aimed at Gower's Confessio Amantis, the first edition of which appeared in 1390. For an excellent statement of the case see Tatlock, Development and Chronology, pp. 172-5. In a foot-note on p. 184 of my own Poetry of Chaucer I cast doubts on this generally accepted view. Further study has shown me that my note was ill-advised. Despite the inaccuracy of the reference to ‘Tyro Apollonius,’ which I there noticed, the allusion must be to something recent, or the whole passage becomes pointless. I must accept Dr. Tatlock's just strictures on my note, and hereby revoke it in ‘my retracciouns.’

^{page 138 note 2} Ll. 94-114.

^{page 138 note 3} B. 66-77.

^{page 138 note 4} P. 23.

^{page 138 note 5} Testament of Love, 1, 2, 87-93 (Chaucerian and Other Pieces, edited by Skeat).

^{page 139 note 1} Avcite ‘falsed fair Anelida the quene,’ Anelida, 147. Criseyde ‘falsed Troilus,’ Troilus, 5, 1053. See also Booh of the Duchess, 1234; Troilus, 3, 784; 5, 1056, 1845; Cant. Tales, A 3175; F 627.

^{page 139 note 2} Conf. Amantis, 2, 2150; 5, 5182. One may consult further Bradley-Stratmann's Middle English Dictionary s. v. falsen, and for its use in Old French, Godefroy s. v. fausser. Lydgate uses the word in a reference to Jason and Medea in Temple of Glass, 63, where the context suggests Chaucer's influence.

^{page 140 note 1} The question is briefly noticed without definite conclusions by Koch, The Chronology of Chaucer's Writings, Chaucer Society, 1890, p. 45, and by Ten Brink in Englische Studien, 17, 19-20 (1892). Bech discusses the matter at some length, Anglia, 5, 379 (1882), but concludes that ‘wir müssen … auf ein bestimmtes datum [for the conclusion of Chaucer's work on the Legend] verzichten.‘

^{page 140 note 2} Koch, however, suggests in passing that ‘at least [the legend] of Cleopatra must have been finished when Chaucer was writing the Prologue (see 1. 566).‘ Chronology, p. 45.

^{page 140 note 3} Publications of the Modern Language Association, 20, 802-818.

^{page 141 note 1} Op. cit., pp. 809-810.

^{page 141 note 2} Development and Chronology, pp. 122-8. To Tatloek's argument that the large proportion of lines beginning with and is ‘simply Chaucer's rapid narrative style,‘ one may add that in Tennyson's Marriage of Geraint 176 out of 849 lines, or more than 20%, similarly begin with and. In his earlier Morte d'Arthur the proportion of such lines is less than 14%, while in Ulysses there are but 4 such lines out of 70, or less than 6%.

^{page 141 note 3} Chaucer's Legend of Good Women, London, 1902.

^{page 141 note 4} Anglia, 5, 379.

^{page 141 note 5} Op. cit., p. 89. The italics are Mr. Bilderbeck's. For the elaborate and ingenious argument by which Mr. Bilderbeck seeks to support his hypothesis the reader must be referred to his own pages. It involves for one thing the date 1390 for the so-called B version of the Prologue, a position which is no longer tenable. Dr. Lowes has pointed out (Publications of the Modern Language Association, 20, 801-3) in a foot-note that Bilderbeek's calculations are in one important particular self-contradictory.

^{page 143 note 1} L. G. W., B 481-4. The A version reads lyve for tyme, but is otherwise identical. On this passage cf. Bech, Anglia, 5, 379.

^{page 143 note 2} Op. cit., p. 130.

^{page 143 note 3} See the articles by Lowes in Publications of the Modern Language Association, 19, 593-683; 20, 749-864, and Tatlock, op. cit., pp. 86-131. In my Poetry of Chaucer (1906), pp. 142-3, in a brief discussion of the two forms of the Prologue, I said that the question of priority had never been satisfactorily settled, but that the probabilities seemed to me to favor the priority of A. I could not agree with Dr. Lowes in denying the identification of Alcestis with Queen Anne in the B version; and granting the identification, I could see no adequate reason why Chaucer should have effaced it in A. Since my book appeared, Dr. Tatlock's restatement and amplification of Dr. Lowes' argument has put the matter in a new light. He accepts the identification of Alcestis with the Queen, and gives an adequate reason for its effacement in the later version. My own studies lead me to put this reason in a somewhat different form, as will appear presently in this paper; but essentially my position is the same as his. For convenience in referring to the printed text, I continue to use the designations A and B as employed by Skeat.

^{page 144 note 1} Op. cit., p. 128. Cf. Bilderbeck, op. cit., p. 74.

^{page 145 note 1} Skeat, Oxford Chaucer, 3, xxxvii; Bech, Anglia, 5, 318; Tatlock, op. cit., p. 129.

^{page 145 note 2} L. G. W., 917-921. It may be worth while to notice that in the Merchant's Tale, which Tatlock (p. 217) sees reason for dating ‘not later than 1394,’ there is an allusion to the story of Pyramus and Thisbe, which in one line seems to echo the phraseology of the Legend. Cf.: ‘Thogh they were kept ful longe streite overal’ (E 2129), and ‘Maidens been ykept, for jelosye, / Ful streite, lest they diden som folye,’ L. Q. W., 722-3. Ovid merely says: ‘Sed vetuere patres’ (Met., 4, 61).

^{page 146 note 1} It is possible that each may have consulted also the account in Livy, Book i, chapters 57-60. Cf. Beeh, Anglia, 5, 333-6.

^{page 147 note 1} Livy is briefer and more explicit: ‘Paucis interiectis diebus Sex. Tarquinius inscio Collatino cum comité uno Collatiam venit.’ 1. 58. 1. Similarly Boccaccio in De Claris Mulieribus: ‘Nee multis interpositis diebus urgente insania, clam castris relictis nocte uenit Collatium (sic).’ Cap. xlvi, ed. 1539, fol. xxxii (b). One wonders whether Boccaccio's clam or Livy's inscio Collatino may not be responsible for Gower's priveliche and Chaucer's privy halke. The idea is not expressed by Ovid, though it is, perhaps, sufficiently implied in the context.

^{page 147 note 2} The error is repeated earlier in Gower's version of the story, where he is describing the visit of Tarquin and Collatine to the latter's house:

Cf. Macaulay's note in his ed. of Gower, vol. 3, pp. 534-5.

^{page 148 note 1} I have given this piece of evidence for what it is worth, though fully recognizing its doubtful character. Though Ovid explicity mentions Collatia in the passage just quoted, earlier in his story (line 741) he is very vague. Gower and Chaucer may have been independently misled by the earlier passage. Still there are the verbal parallels noticed above. Chaucer and Gower further agree in making Lucretia swoon at the moment of Tarquin's outrage, a detail which I have found in no other version. Chaucer says:

Gower reads:

Ovid merely has:

One may also compare Chaucer's lines 1740-44 with Gower's 4838-43 and Ovid's 759-60.

^{page 149 note 1} Op. cit., p. 128. Cf. also pp. 112-113, and Lounsbury, Studies in Chaucer, 3, 335-9.

^{page 150 note 1} Dr. Tatlock suggests (pp. 110-111) that ‘Chaucer had in mind to read [the poem] aloud in a circle of his friends, presumably at court.‘ To the passages which he cites in support of this conjecture may be added Phyllis, 2397-8 and 2401-2. If it was Chaucer's practice to read aloud at court or elsewhere individual legends from time to time as they were composed, the consciousness that their first audience was to be auricular would explain the intrusion of such passages as Dr. Tatlock cites.

^{page 150 note 2} We must assume some such circulation, or at least a reading aloud, to explain Usk's allusion to the Prologue and Gower's to Cleopatra noticed above p. 138.

^{page 150 note 3} Op. cit., p. 131.