Chaucer's Works (ed. Skeat) Vol. II/Troilus Intro

The probable date is about 1380-2, and can hardly have been earlier than 1379 or later than 1383. No doubt it was in hand for a considerable time. It certainly followed close upon the translation of Boethius; see p. vii above.

The chief authority followed by Chaucer is Boccaccio's poem named Il Filostrato, in 9 Parts or Books of very variable length, and composed in ottava rima, or stanzas containing eight lines each. I have used the copy in the Opere Volgari di G. Boccaccio; Firenze, 1832.

Owing to the patient labours of Mr. W. M. Rossetti, who has collated the Filostrato with the Troilus line by line, and published the results of his work for the Chaucer Society in 1875, we are able to tell the precise extent to which Chaucer is indebted to Boccaccio for this story. The Filostrato contains 5704 lines; and the Troilus 8239 lines, if we do not reckon in the 12 Latin lines printed below, at p. 404. Hence we obtain the following result.

In other words, Chaucer's debt to Boccaccio amounts to less than one-third of the whole poem; and there remains more than two-thirds of it to be accounted for from other sources. But even after all deductions have been made for passages borrowed from other authors, very nearly two-thirds remain for which Chaucer is solely responsible. As in the case of the Knightes Tale, close investigation shews that Chaucer is, after all, less indebted to Boccaccio than might seem, upon a hasty comparison, to be the case.

As it was found impracticable to give Mr. Rossetti's results in full, I have drawn up lists of parallel passages in a somewhat rough way, which are given in the Notes, at the beginning of every Book; see pp. 461, 467, 474, 484, 494. These lists are sufficiently accurate to enable the reader, in general, to discover the passages which are in no way due to the Filostrato.

§ 3. I have taken occasion, at the same time, to note other passages for which Chaucer is indebted to some other authors. Of these we may particularly note the following. In Book I, lines 400-420 are translated from Petrarch's 88th Sonnet, which is quoted at length at p. 464. In Book III, lines 813-833, 1625-9, and 1744-1768 are all from the second Book of Boethius (Prose 4, 86-120 and 4-10, and Metre 8). In Book IV, lines 974-1078 are from Boethius, Book V. In Book V, lines 1-14 and 1807-27 are from various parts of Boccaccio's Teseide; and a part of the last stanza is from Dante. On account of such borrowings, we may subtract about 220 lines more from Chaucer's 'balance'; which still leaves due to him nearly 5436 lines.

§ 4. Of course it will be readily understood that, in the case of these 5436 lines, numerous short quotations and allusions occur, most of which are pointed out in the notes. Thus, in Book II, lines 402-3 are from Ovid, Art. Amat. ii. 118; lines 716-8 are from Le Roman de la Rose ; and so on. No particular notice need be taken of this, as similar hints are utilised in other poems by Chaucer; and, indeed, by all other poets. But there is one particular case of borrowing, of considerable importance, which will be considered below, in § 9 (p. liii).

§ 5. It is, however, necessary to observe here that, in taking his story from Boccaccio, Chaucer has so altered and adapted it as to make it peculiarly his own; precisely as he has done in the case of the Knightes Tale. Sometimes he translates very closely and even neatly, and sometimes he takes a mere hint from a long passage. He expands or condenses his material at pleasure; and even, in some cases, transposes the order of it. It is quite clear that he gave himself a free hand.

The most important point is that he did not accept the characters of the three chief actors, Troilus, Criseyde, and Pandarus, as pourtrayed by Boccaccio; he did not even accept all the incidents which gave occasion for their behaviour. Pandarus is no longer the cousin of Criseyde, a young and dashing gallant, but her middle-aged uncle, with blunted perceptions of what is moral and noble. In fact, Chaucer's Pandarus is a thorough and perfect study of character, drawn with a dramatic skill not inferior to that of Shakespeare, and worthy of the author of the immortal Prologue to the Canterbury Tales. I must leave the fuller consideration of these points to others; it is hardly necessary to repeat, at full length, the Prefatory Remarks by Mr. Rossetti, whilst at the same time, if I begin to quote from them, I shall hardly know where to stop. See also Ten Brink's English Literature, and Morley's English Writers, vol. v.

§ 6. It has been observed that, whilst Chaucer carefully read and made very good use of two of Boccaccio's works, viz. Il Filostrato and Il Teseide, he nowhere mentions Boccaccio by name; and this has occasioned some surprise. But we must not apply modern ideas to explain medieval facts, as is so frequently done. When we consider how often MSS. of works by known authors have no author's name attached to them, it becomes likely that Chaucer obtained manuscript copies of these works unmarked by the author's name; and though he must doubtless have been aware of it, there was no cogent reason why he should declare himself indebted to one in whom Englishmen were, as yet, quite uninterested. Even when he refers to Petrarch in the Clerk's Prologue (E 27-35), he has to explain who he was, and to inform readers of his recent death. In those days, there was much laxity in the mode of citing authors.

§ 7. It will help us to understand matters more clearly, if we further observe the haphazard manner in which quotations were often made. We know, for example, that no book was more accessible than the Vulgate version of the Bible; yet it is quite common to find the most curious mistakes made in reference to it. The author of Piers Plowman (B. text, iii. 93-95) attributes to Solomon a passage which he quotes from Job, and (B. vii. 123) to St. Luke, a passage from St. Matthew; and again (B. vi. 240) to St. Matthew, a passage from St. Luke. Chaucer makes many mistakes of a like nature; I will only cite here his reference to Solomon (Cant. Tales, A 4330), as the author of a passage in Ecclesiasticus. Even in modern dictionaries we find passages cited from 'Dryden' or 'Bacon' at large, without further remark; as if the verification of a reference were of slight consequence. This may help to explain to us the curious allusion to Zanzis as being the author of a passage which Chaucer must have known was from his favourite Ovid (see note to Troil. iv. 414), whilst he was, at the same time, well aware that Zanzis was not a poet, but a painter (Cant. Tales, C 16); however, in this case we have probably to do with a piece of our author's delicious banter, since he adds that Pandarus was speaking 'for the nonce.'

There is another point about medieval quotations which must by no means be missed. They were frequently made, not from the authors themselves, but from manuscript note-books which contained hundreds of choice passages, from all sorts of authors, collected by diligent compilers. Thus it was, I strongly suspect, that Albertano of Brescia was enabled to pour out such quantities of quotations as those which Chaucer copied from him in his Tale of Melibeus. Thus it was that borrowers of such note-books often trusted to their strong memories for the words of a quotation, yet forgot or mistook the author's name; as was readily done when a dozen such names occurred on every page. A MS. of this character is before me now. It contains many subjects in alphabetical order. Under Fortitudo are given 17 quotations which more or less relate to it, from Ambrose, Gregory, Chrysostom, and the rest, all in less than a single page. And thus it was, without doubt, that Chaucer made acquaintance with the three scraps of Horace which I shall presently consider. It is obvious that Chaucer never saw Horace's works in the complete state; if he had done so, he would have found a writer after his own heart, and he would have quoted him even more freely than he has quoted Ovid. 'Chaucer on Horace' would have been delightful indeed; but this treat was denied, both to him and to us.

§ 8. The first and second scraps from Horace are hackneyed quotations. 'Multa renascentur' occurs in Troil. ii. 22 (see note, p. 468); and 'Humano capiti' in Troil. ii. 1041 (note, p. 472). In the third case (p. 464), there is no reason why we should hesitate to accept the theory, suggested by Dr. G. Latham (Athenæum, Oct. 3, 1868) and by Professor Ten Brink independently, that the well-known line (Epist. I, 2. 1)—

'Troiani belli scriptorem, maxime Lolli,'

was misunderstood by Chaucer (or by some one else who misled him) as implying that Lollius was the name of a writer on the Trojan war. Those who are best acquainted with the ways of medieval literature will least hesitate to adopt this view. It is notorious that first lines of a poem are frequently quoted apart from their context, and repeated as if they were complete; and, however amazing such a blunder may seem to us now, there is really nothing very extraordinary about it.

We should also notice that Lollius was to Chaucer a mere name, which he used, in his usual manner, as a sort of convenient embellishment; for he is inconsistent in his use of it. In Book i. 394, 'myn autour called Lollius' really means Petrarch; whereas in Book v. 1653, though the reference is to the Filostrato, Bk. viii. st. 8, Chaucer probably meant no more than that Lollius was an author whom the Italian poet might have followed. Cf. my note to the House of Fame, 1468, where the name occurs for the third time. We may also notice that, in Book iii. 1325, Chaucer bears testimony to the 'excellence' of his 'auctor.' The statement, in Book ii. 14, that he took the story 'out of Latin' is less helpful than it appears to be; for 'Latin' may mean either Latin or Italian.

§ 9. I have spoken (§ 4) of 'a particular case of borrowing,' which I now propose to consider more particularly. The discovery that Chaucer mainly drew his materials from Boccaccio seems to have satisfied most enquirers; and hence it has come to pass that one of Chaucer's sources has been little regarded, though it is really of some importance. I refer to the Historia Troiana of Guido delle Colonne, or, as Chaucer rightly calls him, Guido de Columpnis, i.e. Columnis (House of Fame, 1469). Chaucer's obligations to this author have been insufficiently explored.

When, in 1889, in printing the Legend of Good Women with an accuracy never before attempted, I restored the MS. reading Guido for the Ouyde of all previous editions in l. 1396, a clue was thus obtained to a new source for some of Chaucer's work. It was thus made clear that the Legend of Hypsipyle and Medea was primarily derived from this source; and further, that it was from Guido that Chaucer derived his use of Ilioun to mean the citadel of Troy (Leg. of Good Women, 936, and note). In the Nonne Prestes Tale, B 4331, as was pointed out by Tyrwhitt long ago, the dream of Andromache is taken from Guido. And I find in Lounsbury's Studies in Chaucer, ii. 315, the significant but insufficient remark, that 'it was in Guido da (sic) Colonna's work that Chaucer found the martial deeds of Troilus recounted in full, the slaughter he wrought, and the terror he inspired.' Hence we naturally come to the question, what incidents in Troilus are expressly due to Guido?

§ 10. Before answering this question, it will be best to consider the famous crux, as to the meaning of the word Trophee.

When Lydgate is speaking of his master's Troilus, viz. in his Prologue to the Falls of Princes, st. 3, he says that Chaucer

'made a translacion Of a boke which called is Trophe In Lumbarde tong,' &c.

No book or author is now known by that name; and, as Chaucer was in this case much indebted to Boccaccio, critics have jumped to the conclusion that Trophee means either Boccaccio or the Filostrato; and this conclusion has been supported by arguments so hopeless as to need no repetition. But it is most likely that Lydgate, who does not seem to have known any Italian, spoke somewhat casually; and, as Chaucer was to some extent indebted to Guido, he may possibly have meant Guido.

So far, I have merely stated a supposition which is, in itself, possible; but I shall now adduce what I believe to be reasonable and solid proof of it.

We have yet another mention of Trophee, viz. in Chaucer himself! In the Monkes Tale, B 3307, he says of Hercules—

'At bothe the worldes endes, seith Trophee, In stede of boundes, he a piler sette.'

Whence, we may ask, is this taken? My answer is, from Guido.

§ 11. If we examine the sources of the story of Hercules in the Monkes Tale, we see that all the supposed facts except the one mentioned in the two lines above quoted are taken from Boethius and Ovid (see the Notes). Now the next most obvious source of information was Guido's work, since the very first Book has a good deal about Hercules, and the Legend of Hypsipyle clearly shews us that Chaucer was aware of this. And, although neither Ovid (in Met. ix.) nor Boethius has any allusion to the Pillars of Hercules, they are expressly mentioned by Guido. In the English translation called the Gest Historiale of the Destruction of Troy, ed. Panton and Donaldson (which I call, for brevity, the alliterative Troy-book), l. 308, we read:—

'But the wonders that he wroght in this world here In yche cuntré ben knowen under Criste evyn. Tow pyllers he pight in a place lowe Vppon Gades groundes, that he gotton had.'

And again, further on, the Latin text has:—'Locus ille, in quo predicte Herculis columpne sunt affixe, dicitur Saracenica lingua Saphy.' To which is added, that Alexander afterwards came to the same spot.

When Lydgate, in translating Guido, comes to this passage, he says:—

'And of the pyllers that at Gades he set, Which Alexsaundre, of Macedone the kyng, That was so worthy here in his lyuynge, Rood in his conquest, as Guydo list to write, With all his hoost proudely to visyte ... And these boundes named be of all Of Hercules, for he hymselfe theim set As for his markes, all other for to lette Ferther to passe, as Guydo maketh mynde'; &c. Siege of Troye, ed. 1555, fol. B6.

We can now easily see that, when Lydgate speaks of the book 'which called is Trophe in Lumbarde tong,' he is simply copying the name of the book from Chaucer, though he seems also to have heard some rumour of its being so called in Italy.

§ 12. Why this particular book was so called, we have no means of knowing ; but this does not invalidate the fact here pointed out. Of course the Latin side-note in some of the MSS. of the Monkes Tale, which explains 'Trophee' as referring to 'ille vates Chaldeorum Tropheus,' must be due to some mistake, even if it emanated (as is possible) from Chaucer himself. It is probable that, when the former part of the Monkes Tale was written, Chaucer did not know much about Guido's work; for the account of Hercules occurs in the very first chapter. Perhaps he confused the name of Tropheus with that of Trogus, i.e. Pompeius Trogus the historian, whose work is one of the authorities for the history of the Assyrian monarchy.

§ 13. It remains for me to point out some of the passages in Troilus which are clearly due to Guido, and are not found in Boccaccio at all.

Book I. 145-7:—

'But the Troyane gestes, as they felle, In Omer, or in Dares, or in Dyte, Who-so that can, may rede hem as they wryte.'

The reference here is simply to Guido's history, whence, and not at first hand, both Chaucer and his readers could easily get the required information. Guido constantly refers to these authors; and, although he speaks disrespectfully of Homer, he professes to put great faith in Dares and Dytes, whose names he frequently cites as being those of his best authorities.

With the description of Troilus in ll. 1072-1085, it is interesting to compare the words of Guido, in Book VIII. 'Troilus vero, licet multum fuit corpore magnus, magis fuit tamen corde magnanimus; animosus multum, set multam habuit in sua animositate temperiem; dilectus plurimum a puellis cum ipse aliqualem seruando modestiam delectaretur in illis. In viribus et strenuitate bellandi uel fuit alius Hector uel secundus ab ipso. In toto eciam regno Troie iuuenis nullus fuit tantis viribus nec tanta audacia gloriosus .' The latter part of this description should be compared with Book II. 157-161, where the very phrase 'Ector the secounde' is used; see also ll. 181-189.

§ 14. Book II. 618. 'The yate ... Of Dardanus.' The six gates of Troy are named in Guido, Book IV, 'Quarum vna Dardanides, secunda Tymbrea, tercia Helyas, quarta Chetas, quinta Troiana, vltima Anthenorides vocabantur.'

'The furst and the fairest fourmet was Dardan.' Allit. Troy-book, l. 1557.

Lydgate keeps the form 'Dardanydes'; cap. xi. fol. F 5.

§ 15. Book IV. 204. 'For he was after traytour to the toun.' The treason of Antenor is told by Guido at great length; see 'Boke xxviii' of the allit. Troy-book, p. 364; Lydgate, Siege of Troye, Y 6, back. Cf. Dictys Cretensis, lib. iv. c. 22.

Book IV. 1397, &c. 'For al Apollo and his clerkish lawes,' &c. Guido gives rather a long account of the manner in which Criseyde upbraided her father Chalcas at their meeting. Chaucer says nothing about this matter in Book V. 193, but he here introduces an account of the same speech, telling us that Creseyde intended to make it! I quote from Book XIX. 'Sane deceperunt te Apollinis friuola responsa, a quo dicis te suscepisse mandatum vt tu paternas Lares desereres, et tuos in tanta acerbitate Penates sic tuis specialiter hostibus adhereres. Sane non fuit ille deus Apollo, set, puto, fuit comitiua infernalium Furiarum a quibus responsa talia recepisti.' Cf. allit. Troy-book, 8103-40; and observe that Lydgate, in his Siege of Troye, R 3, back, omits the speech of Criseyde to her father, on the ground that it is given in Chaucer. Yet such is not the case, unless we allow the present passage to stand for it. In Book V. 194, Chaucer (following Boccaccio) expressly says that she was mute!

Book IV. 1695-1701. This last stanza is not in Boccaccio; but the general sense of it is in Guido, Book XIX, where the interview ends thus:—'Set diei Aurora quasi superueniente uicina, Troilus a Brisaida in multis anxietatibus et doloribus discessit; et ea relicta ad sui palacii menia properauit.' Lydgate, at this point, refers us to Chaucer; Siege of Troye, fol. R 2, back. The allit. Troy-book actually does the same; l. 8054.

§ 16. Book V. 92-189. These fourteen stanzas are not in Boccaccio. The corresponding passage in Guido (Book XIX) is as follows:—

'Troilus et Troiani redeunt, Grecis eam recipientibus in suo commeatu. Inter quos dum esset Diomedes, et illam Diomedes inspexit, statim in ardore veneris exarsit et eam vehementi desiderio concupiuit, qui collateralis associando Brisaidam cum insimul equitarent, sui ardoris flammam continere non valens Brisaide reuelat sui estuantis cordis amorem; quam in multis affectuosis verbis et blandiciis necnon et promissionibus reuera magnificis allicere satis humiliter est rogatus. Set Brisaida in primis monitis, vt mulierum moris est, suum prestare recusauit assensum; nec tamen passa est quin post multa Diomedis verba, ipsum nolens a spe sua deicere verbis similibus dixit ei: "Amoris tui oblaciones ad presens nec repudio nec admitto, cum cor meum non sit ad presens ita dispositum quod tibi possim aliter respondere."'

Book V. 799-805. The description of Diomede in Boccaccio (Fil. VI. 33) is merely as follows:—

'Egli era grande e bel della persona, Giovane fresco e piacevole assai, E forte e fier siccome si ragiona, E parlante quant'altro Greco mai, E ad amor la natura aveva prona.'

The account in Guido (Book VIII) is as follows:—'Diomedes vero multa fuit proceritate, distensus amplo pectore, robustis scapulis, aspectu ferox; in promissis fallax; in armis strenuus; victorie cupidus; timendus a multis, cum multum esset iniuriosus; sermonibus sibi nimis impaciens, cum molestus seruientibus nimis esset; libidinosus quidem multum, et qui multas traxit angustias ob feruorem amoris.' Cf. allit. Troy-book, ll. 3794-3803; Lydgate, Siege of Troye, fol. K 1, back.

Book V. 810. To gon y-tressed, &c. Perhaps suggested by the remark in Guido (Book XIX) that Cressid's hair was unbound in her hour of deepest sorrow:—'aureos crines suos a lege ligaminis absolutos a lactea sui capitis cute diuellit.' Cf. IV. 736.

Book V. 827-840. Troilus is not described by Boccaccio. Guido's description of him has already been quoted above; see remarks on Book I. 1072; pp. lvi, lvii.

Book V. 1002-4. The parallel passage in Guido has already been quoted, viz.: 'Amoris tui oblaciones ad presens nec repudio nec admitto.' See remarks on l. 92; p. lviii.

Book V. 1013. Obviously from Guido; the passage follows soon after that last quoted. 'Associauit [Diomedes] eam vsquequo Brisaida recipere in sui patris tentoria se debebat. Et ea perueniente ibidem, ipse eam ab equo descendentem promptus adiuit, et vnam de cirothecis, quam Brisaida gerebat in manu, ab ea nullo percipiente furtiue subtraxit. Set cum ipsa sola presensit, placitum furtum dissimulauit amantis.'

For this incident of the glove, cf. allit. Troy-book, l. 8092.

Book V. 1023-1099. This passage is not in Boccaccio. Several hints for it seem to have been taken from Guido, Book XIX, whence I quote the following.

'Nondum dies illa ad horas declinauerat vespertinas, cum iam suas Brisaida recentes mutauerat voluntates,' &c.. 'Et iam nobilis Troili amor ceperat in sua mente tepescere, et sic repente subito facta volubilis se in omnibus variauit. Quid est ergo quod dicitur de constancia mulierum,' &c.

'Tunc ilico Diomedes superuenit. . qui repente in Troilum irruit, ipsum ab equo prosternit, ab eo auferens equum suum, quem per suum nuncium specialem ad Brisaidam in exennium destinauit, mandans nuncio suo predicto vt Brisaide nunciet equum ipsum eius fuisse dilecti. . . . Brisaida vero equum Troili recepit hilariter, et ipsi nuncio refert hec verba: "Dic secure domino tuo quod ilium odio habere non possum, qui me tanta puritate cordis affectat . . . . [Diomedes] Brisaidam accedit, et eam suplex hortatur vt sibi consenciat in multitudine lacrimarum. Set illa, que multum vigebat sagacitatis astucia, Diomedem sagacibus machinacionibus differre procurat, ut ipsum afflictum amoris incendio magis affligat, et eius amoris vehemenciam in maioris augmentum ardoris extollat. Vnde Diomedi suum amorem non negat, etiam nec promittit."'

In l. 1039, read he, i. e. Diomede; see my note on the line, at p. 499.

In l. 1037, the story means the Historia Troiana; and in l. 1044, in the stories elles-where means 'elsewhere in the same History.' The passage (in Book XXV) is as follows:—

'Troilus autem tunc amorem Brisaide Diomedi obprobriosis verbis improperat; set Greci Diomedem ... abstraxerunt' ...

'Interim Brisaida contra patris sui voluntatem videre Diomedem in lecto suo iacentem ex vulnere sibi facto frequenter accedit, et licet sciuisset illum a Troilo dudum dilecto suo sic vulneratum, multa tamen in mente sua reuoluit; et dum diligenter attendit de se iungenda cum Troilo nullam sibi superesse fiduciam, totum suum animum, tanquam varia et mutabilis, sicut est proprium mulierum, in Diomedis declinat amorem.'

Cf. Troy-book, ll. 9942-59; Lydgate, Siege of Troye, fol. U 4.

Book V. 1558-60. The treacherous slaughter of Hector by Achilles is in Guido, near the end of Book XXV. See my note to l. 1558, at p. 503.

Book V. 1771. 'Read Dares.' This merely means that Guido cites Dares as his authority for the mighty deeds of Troilus. In Book XXV, I find:—'Scripsit enim Dares, quod illo die mille milites interfecit [Troilus] ex Grecis'; cf. l. 1802 below. So in the allit. Troy-book, ll. 9877-9:—

'As Dares of his dedis duly me tellus, A thowsaund thro knightes throng he to dethe, That day with his dynttes, of the derffe Grekes.'

So Lydgate, Siege of Troye, fol. U 3, back:—

'And, as Dares wryteth specyally, A thousand knightes this Troyan champyowne That day hath slayne, rydyng vp and downe, As myne auctour Guydo lyst endyte; Saue after hym, I can no ferther wryte.'

I. e. he only knew of Dares through the medium of Guido. In fact, Dares (capp. 29, 31, 32) has 'multos,' not 'mille.'

Book V. 1849-1855. The introduction of this stanza is quite irrelevant, unless we remember that, in Guido, the story of Troy is completely mixed up with invectives against idolatry. In Book X, there is a detailed account of the heathen gods, the worship of which is attributed to the instigation of fiends. See the long account in the allit. Troy-book, ll. 4257-4531, concluding with the revelation by Apollo to Calchas of the coming fall of Troy. Cf. Lydgate, Siege of Troye, fol. K 6. Of course, this notion of the interference of the gods in the affairs of the Greeks and Trojans is ultimately due to Homer.

§ 17. With regard to the statement in Guido, that Achilles slew Hector treacherously, we must remember how much turns upon this assertion. His object was to glorify the Trojans, the supposed ancestors of the Roman race, and to depreciate the Greeks. The following passage from Guido, Book XXV, is too characteristic to be omitted. 'Set o Homere, qui in libris tuis Achillem tot laudibus, tot preconiis extulisti, que probabilis racio te induxit, vt Achillem tantis probitatis meritis vel titulis exultasses?' Such was the general opinion about Homer in the thirteenth and fourteenth centuries.

§ 18. This is not the place for a full consideration of the further question, as to the sources of information whence Boccaccio and Guido respectively drew their stories. Nor is it profitable to search the supposed works of Dares and Dictys for the passages to which Chaucer appears to refer; since he merely knew those authors by name, owing to Guido's frequent appeals to them. Nevertheless, it is interesting to find that Guido was quite as innocent as were Chaucer and Lydgate of any knowledge of Dares and Dictys at first hand. He acquired his great reputation in the simplest possible way, by stealing the whole of his 'History' bodily, from a French romance by Benoît de Sainte-More, entitled Le Roman de Troie, which has been well edited and discussed by Mons. A. Joly. Mons. Joly has shewn that the Roman de Troie first appeared between the years 1175 and 1185; and that Guido's Historia Troiana is little more than an adaptation of it, which was completed in the year 1287, without any acknowledgment as to its true source.

Benoît frequently cites Dares (or Daires), and at the end of his poem, ll. 30095-6, says:—

'Ce que dist Daires et Dithis I avons si retreit et mis.'

In his Hist. of Eng. Literature (E. version, ii. 113), Ten Brink remarks that, whilst Chaucer prefers to follow Guido rather than Benoît in his Legend of Good Women, he 'does the exact opposite to what he did in Troilus.' For this assertion I can find but little proof. It is hard to find anything in Benoît's lengthy Romance which he may not have taken, much more easily, from Guido. There are, however, just a few such points in Book V. 1037-1078. Thus, in l. 1038, Criseyde gives Diomede Troilus' horse; cf. Benoît, l. 15046—'lo cheval Vos presterai.' L. 1043 is from the same, ll. 15102-4:—

'La destre manche de son braz Bone et fresche de ciclaton Li done en leu de gonfanon.'

Ll. 1051-7 answer to the same, beginning at l. 20233; and l. 1074 is from the same, l. 20308:—'Dex donge bien à Troylus!' I doubt if there is much more.

For some further account of the works ascribed to Dares Phrygius and Dictys Cretensis, both duly edited among the 'Delphin Classics,' I must refer the reader to Smith's Classical Dictionary.

§ 19. The whole question of the various early romances that relate to Troy is well considered in a work entitled 'Testi Inediti di Storia Trojana, preceduti da uno studio sulla Leggenda Trojana in Italia, per Egidio Gorra; Torino, 1887'; where various authorities are cited, and specimens of several texts are given. At p. 136 are given the very lines of Benoît's Roman (ll. 795-6) where Guido found a reference to the columns of Hercules:—

'Et les bonnes ilec ficha Ou Alixandre les trova.'

This hint he has somewhat elaborated, probably because he took a personal interest in 'columns,' on account of their reference to his own name—'delle Colonne.' I believe that the notion of Alexander finding Hercules' Pillars is due to a rather large blunder in geography. Hercules set up his pillars 'at the end of the world,' viz. at the straits of Gibraltar, whereas Alexander set up his at another 'end of the world,' viz. at the furthest point of India which he succeeded in reaching. So says his Romance; see Alexander and Dindimus, ed. Skeat, l. 1137; Wars of Alexander, l. 5063. The setting up of pillars as boundary-marks seems to have been common; cf. Vergil, Æn. xi. 262. Among the points noticed by Gorra, I may mention the following:—

1. Some account (p. 7) of the Ephemeris Belli Troiani by Dictys Cretensis, who, it was pretended, accompanied Idomeneus to the Trojan war. Achilles is depicted in dark colours; he is treacherous towards Agamemnon; falls in love with the Trojan princess, Polyxena; and slays Hector by a stratagem. It appears to have been a work of invention, resting upon no Greek original.

2. Some account (p. 17) of the Historia de Excidio Troiae of Dares Phrygius, a work which (as was pretended) was discovered by Cornelius Nepos. This also, in the opinion of most critics, was an original work. At p. 115, there is a comparison of the lists of Greek leaders and the number of their ships (cf. Homer, Il. ii.) as given by Dares, Benoît, and Guido.

3. At p. 123, there is an enumeration of points in which Guido varies from Benoît.

4. At p. 152, is an account of some Italian prose versions of the story of Troy. Such are: La Istorietta Trojana, with extracts from it at p. 371; a romance by Binduccio dello Scelto, with extracts relating to 'Troilo e Briseida' at p. 404; a version of Guido by Mazzeo Bellebuoni, with extracts relating to 'Paride ed Elena' at p. 443; an anonymous version, with extracts relating to 'Giasone e Medea' at p. 458; a version in the Venetian dialect, with extracts relating to 'Ettore ed Ercole' at p. 481; another anonymous version, with extracts at p. 493; and La 'Fiorita' of Armannino, Giudice da Bologna, with extracts at p. 532.

5. At p. 265, is an account of Italian poetical versions, viz. Enfances Hector, Poema d'Achille, Il Trojano di Domenico da Montechiello, Il Trojano a stampa (i.e. a printed edition of Il Trojano), and L'Intelligenza. At p. 336, Boccaccio's Filostrato is discussed; followed by a brief notice of an anonymous poem, also in ottava rima, called Il cantare di Insidoria. It appears that Boccaccio followed some recension of the French text of Benoît, but much of the work is his own invention. In particular, he created the character of Pandaro, who resembles a Neapolitan courtier of his own period.

The most interesting of the extracts given by Gorra are those from Binduccio dello Scelto; at p. 411, we have the incident of Diomede possessing himself of Briseida's glove, followed by the interview between Briseida and her father Calcas. At p. 413, Diomede overthrows Troilus, takes his horse from him and sends it to Briseida, who receives it graciously; and at p. 417, Briseida gives Diomede her sleeve as a love-token, after which a 'jousting' takes place between Diomede and Troilus, in which the former is badly wounded.

For further remarks, we are referred, in particular, to H. Dunger's Dictys-Septimius: über die ursprüngliche Abfassung und die Quellen der Ephemeris belli Troiani; Dresden, 1878 (Programm des Vitzthumschen Gymnasiums); to another essay by the same author on Die Sage vom trojanischen Kriege, Leipzig, 1869; to Koerting's Dictys und Dares, &c., Halle, 1874; to A. Joly's Benoît de Sainte-More et le Roman de Troie, Paris, 1871; and to an article by C. Wagener on Dares Phrygius, in Philologus, vol. xxxviii. The student may also consult E. Meybrinck, Die Auffassung der Antike bei Jacques Millet, Guido de Columna, und Benoît de Ste-More, printed in Ausgaben und Abhandlungen aus dem Gebiete für Romanischen Philologie, Marburg, 1886; where the author concludes that Millet was the originator of the story in France. Also W. Greif, Die mittelalterlichen Bearbeitungen der Trojanersage; Marburg, 1886.

§ 20. A few words may be said as to the names of the characters. Troilus is only once mentioned in Homer, where he is said to be one of the sons of Priam, who were slain in battle, Iliad, xxiv. 257; so that his story is of medieval invention, except as to the circumstance of his slayer being Achilles, as stated by Vergil, Æn. i. 474, 475; cf. Horace, Carm. ii. 9. 16. Pandarus occurs as the name of two distinct personages; (1) a Lycian archer, who wounded Menelaus; see Homer, Il. iv. 88, Vergil, Æn. 5. 496; and (2) a companion of Æneas, slain by Turnus; see Vergil, Æn. ix. 672, xi. 396. Diomede is a well-known hero in the Iliad, but his love-story is of late invention. The heroine of Benoît's poem is Briseida, of whom Dares (c. 13) has merely the following brief account: 'Briseidam formosam, alta statura, candidam, capillo flauo et molli, superciliis junctis, oculis venustis, corpore aequali, blandam, affabilem, uerecundam, animo simplici, piam'; but he records nothing more about her. The name is simply copied from Homer's, Il. i. 184, the accusative being taken (as often) as a new nominative case; this Briseis was the captive assigned to Achilles. But Boccaccio substitutes for this the form Griseida, taken from the accusative of Homer's Chryseis, mentioned just two lines above, Il. i. 182. For this Italian form Chaucer substituted Criseyde, a trisyllabic form, with the ey pronounced as the ey in prey. He probably was led to this correction by observing the form Chryseida in his favourite author, Ovid; see Remed. Amoris, 469. Calchas, in Homer, Il. i. 69, is a Grecian priest; but in the later story he becomes a Trojan soothsayer, who, foreseeing the destruction of Troy, secedes to the Greek side, and is looked upon as a traitor. Cf. Vergil, Æn. ii. 176; Ovid, Art. Amat. ii. 737.

§ 21. In Anglia, xiv. 241, there is a useful comparison, by Dr. E. Köppel, of the parallel passages in Troilus and the French Roman de la Rose, ed. Méon, Paris, 1814, which I shall denote by 'R.' These are mostly pointed out in the Notes. Köppel's list is as follows:—

Troilus. I. 635 (cf. III. 328).—Rom. Rose, 8041. 637.—R. 21819. 747.—R. 7595. 810.—R. 21145. 969—R. 12964.

II. 167.—R. 5684. 193.—R. 8757. 716.—R. 5765. 754.—R. 6676. 784 (cf. III. 1035).—R. 12844. 1564.—R. 18498.

III. 294.—R. 7085. 328; see I. 635. 1035; see II. 784. 1634.—R. 8301.

IV. 7.—R. 8076. 519.—R. 6406. 1398.—R. 6941.

V. 365.—R. 18709.

Some of the resemblances are but slight; but others are obvious. The numbers refer to the beginning of a passage; sometimes the really coincident lines are found a little further on.

The parallel passages common to Troilus and Boethius are noted above, pp. xxviii-xxx.

An excellent and exhaustive treatise on the Language of Chaucer's Troilus, by Prof. Kitteredge, is now (1893) being printed for the Chaucer Society. A Ryme-Index to the same, compiled by myself, has been published for the same society, dated 1891.

§ 22. I have frequently alluded above to the alliterative 'Troy-book,' or 'Gest Historiale,' edited for the Early English Text Society, in 1869-74, by Panton and Donaldson. This is useful for reference, as being a tolerably close translation of Guido, although a little imperfect, owing to the loss of some leaves and some slight omissions (probably) on the part of the scribe. It is divided into 36 Books, which agree, very nearly, with the Books into which the original text is divided. The most important passages for comparison with Troilus are lines 3922-34 (description of Troilus); 3794-3803 (Diomede); 7268-89 (fight between Troilus and Diomede); 7886-7905 (Briseida and her dismissal from Troy); 8026-8181 (sorrow of Troilus and Briseida, her departure, and the interviews between Briseida and Diomede, and between her and Calchas her father); 8296-8317 (Diomede captures Troilus' horse, and presents it to Briseida); 8643-60 (death of Hector); 9671-7, 9864-82, 9926-9 (deeds of Troilus); 9942-59 (Briseida visits the wounded Diomede); 10055-85, 10252-10311 (deeds of Troilus, and his death); 10312-62 (reproof of Homer for his false statements).

At l. 8053, we have this remarkable allusion; speaking of Briseida and Troilus, the translator says:—

'Who-so wilnes to wit of thaire wo fir [futher], Turne hym to, and talke there ynoughe!'

I.e. whoever wishes to know more about their wo, let him turn to, and there find enough. This is a clear allusion to Chaucer's work by its name, and helps to date the translation as being later than 1380 or 1382. And, as the translator makes no allusion to Lydgate's translation of Guido, the date of which is 1412-20, we see that he probably wrote between 1382 and 1420 ; so that the date 'about 1400,' adopted in the New Eng. Dictionary (s. v. Bercelet, &c.) cannot be far wrong.

§ 23. Another useful book, frequently mentioned above, is Lydgate's Siege of Troye, of which I possess a copy printed in 1555. This contains several allusions to Chaucer's Troilus, and more than one passage in praise of Chaucer's poetical powers, two of which are quoted in Mr. Rossetti's remarks on MS. Harl. 3943 (Chaucer Soc. 1875), pp. x, xi. These passages are not very helpful, though it is curious to observe that he speaks of Chaucer not only as 'my maister Chaucer,' but as 'noble Galfride, chefe Poete of Brytaine,' and 'my maister Galfride.' The most notable passages occur in cap. xv, fol. K 2; cap. xxv, fol. R 2, back; and near the end, fol. Ee 2. Lydgate's translation is much more free than the preceding one, and he frequently interpolates long passages, besides borrowing a large number of poetical expressions from his 'maister.'

§ 24. Finally, I must not omit to mention the remarkable poem by Robert Henrysoun, called the Testament and Complaint of Criseyde, which forms a sequel to Chaucer's story. Thynne actually printed this, in his edition of 1532, as one of Chaucer's poems, immediately after Troilus; and all the black-letter editions follow suit. Yet the 9th and 10th stanzas contain these words, according to the edition of 1532:—

'Of his distresse me nedeth nat reherse; For worthy Chaucer, in that same boke, In goodly termes, and in ioly verse, Compyled hath his cares, who wyl loke. To breke my slepe, another queare I toke, In whiche I founde the fatal desteny Of fayre Creseyde, whiche ended wretchedly.

Who wot if al that Chaucer wrate was trewe? Nor I wotte nat if this narration Be authorysed, or forged of the newe Of some poete by his inuention, Made to reporte the lamentation And woful ende of this lusty Creseyde, And what distresse she was in or she deyde.'

1. MS. Cl.—The Campsall MS., on vellum, written before 1413; prepared for Henry, Prince of Wales, afterwards Henry V, as shewn by his arms on leaf 2. The poem occupies leaves 2-120; each page usually contains five stanzas. Two pages have been reproduced by the autotype process for the Chaucer Society; viz. leaf 1, recto, containing stanzas 1-5, and leaf 42, verso, containing stanzas 249-251 of Book II, and stanza 1 of Book III. This is a beautifully written MS., and one of the best; but it is disappointing to find that it might easily have been much better. The scribe had a still better copy before him, which he has frequently treated with supreme carelessness; but it is some consolation to find that his mistakes are so obvious that they can easily be corrected. Thus, in Book I, l. 27, he writes dorst for dorste, though it ruins the grammar and the metre; in l. 31, he actually has hym for hem, to the destruction of the sense; in l. 69, he has high (!) for highte; and so on. It therefore requires careful control. In particular, the scribe gives many examples of the fault of 'anticipation,' i.e. the fault whereby the mind, swifter than the pen, has induced him to write down letters that belong to a later syllable or word, or to omit one or more letters. Thus in Book I. l. 80, he omits u in pryuely, writing pryely; in l. 126, he omits and before hoom; in l. 198, he omits lewede; in l. 275, he omits gan; &c. But the faults of 'anticipation' appear most clearly in such startling forms as addermost for aldermost, I. 248, where the former d is due to the one that is coming; assent for absent, IV. 1642, for a like reason; estal for estat, because the next word is royal, I. 432; þyn for þyng, because the next word is myn, I. 683; nat for nas, because the next word is not, I. 738; seynt for seyn, because the next word is that, V. 369; shad for shal, because the next word is drede, V. 385; liten for litel, because weten follows, IV. 198; make for may, because the line ends with wake, III. 341; fleld for feld, II. 195. Sometimes, however, the scribe's mind reverts to something already written, so that we find Delphebus for Delphicus, because Phebus precedes, I. 70; bothen for bothe, because deden precedes, I. 82; falles for fallen, after unhappes, II. 456; daunder for daunger, III. 1321; tolle for tolde, III 802; &c. Downright blunders are not uncommon; as incocent for innocent (where again the former c is due to the latter), II. 1723; agarst for agast, III. 737; right for rit, V. 60. We even find startling variations in the reading, as in III. 1408:—

'Reson wil not that I speke of shep, For it accordeth nough[t] to my matere.'

Certainly, shep (sheep) is irrelevant enough; however, Chaucer refers to sleep. And again, the line in II. 1554, which should run—

As for to bidde a wood man for to renne

appears in the startling form—

As for to bydde a womman for to renne.

As all the variations of 'Cl.' from the correct text are given in the foot-notes, it is not necessary to say more about these peculiarities. I must add, however, that, as in Boethius, I have silently corrected yn to in in such words as thing; besides altering ee and oo to e and o in open syllables, writing v for u, and the like. See above.

The Campsall MS., now in the possession of Mr. Bacon Frank, has been printed in full, as written, for the Chaucer Society; and I have relied upon the accuracy of this well-edited print.

2. MS. Cp.—MS. No. 61 in Corpus Christi College, Cambridge, described in Nasmith's Catalogue, p. 40, as 'a parchment book in folio neatly written, and ornamented with a frontispiece richly illuminated, containing Chaucer's Troilus, in four [error for five] books.' It is a fine folio MS., 12 inches by 8½. This MS., noticed by Warton, has not as yet been printed, though the Chaucer Society have undertaken to print it, upon my recommendation. It contains many pages that are left wholly or partially blank, obviously meant to be supplied with illuminations; which shews that it was written for some wealthy person. On the left margin, near the 83rd stanza of Book IV, is a note of ownership, in a hand of the fifteenth century—'neuer foryeteth: Anne neuyll.' This probably refers to Anne Neville, wife of Humphrey, duke of Buckingham (who was killed at Northampton in 1460), and daughter of Ralph Neville, earl of Westmoreland, and of Joan Beaufort, daughter of John of Gaunt. That is, she was John of Gaunt's granddaughter; and it seems reasonable to infer that the MS. was actually written for one of John of Gaunt's family. This probability is a very interesting one, when we consider how much Chaucer owed to John of Gaunt's favour and protection.

The MS. is slightly deficient, owing to the omission of a few stanzas; but not much is missing. It is of a type closely resembling the preceding, and gives excellent readings. I have therefore taken the opportunity of founding the text upon a close collation of Cl. and Cp., taking Cl. as the foundation, but correcting it by Cp. throughout, without specifying more than the rejected reading of Cl. in passages where these MSS. differ. In this way the numerous absurdities of Cl. (as noted above) have been easily corrected, and the resulting text is a great improvement upon all that have hitherto appeared. In a few places, as shewn by the foot-notes, the readings of other MSS. have been preferred.

3. MS. H.—MS. Harl. 2280, in the British Museum. An excellent MS., very closely related to both the preceding. Printed in full for the Chaucer Society, and collated throughout in the present edition. It was taken as the basis of the text in Morris's Aldine edition, which in many passages closely resembles the present text. It is certainly the third best MS. One leaf is missing (Bk. V. 1345-1428; twelve stanzas).

4. MS. Cm.—MS. Gg. 4. 27, in the Cambridge University Library; the same MS. as that denoted by 'Cm.' in the foot-notes to the Canterbury Tales, and by 'C.' in the foot-notes to the Legend of Good Women. A remarkable MS., printed in full for the Chaucer Society. It exhibits a different type of text from that found in Cl., Cp., and H. The most noteworthy differences are as follows. In Bk. ii. 734, 5, this MS. has quite a different couplet, viz.:

Men louyn women þour al þis toun aboute; Be þey þe wers? whi, nay, with-outyn doute.

Bk. ii. 792 runs thus:—

How ofte tyme may men rede and se.

Bk. iv. 309-15 (stanza 45) runs thus:—

What shulde ye don but, for myn disconfort, Stondyn for nought, and wepyn out youre ye? Syn sche is queynt that wont was yow disport , In vayn from this forth have I seyn twye; For medycyn youre vertu is a-weye; O crewel eyen, sythyn that youre dispyt Was al to sen Crisseydes eyen bryght.

Bk. iv. 638 runs thus:—

Pandare answerde, of that be as be may.

After Bk. iv. 735, MS. Cm. introduces the following stanza, which, in the present text, appears a little later (ll. 750-6) in a slightly altered form.

The salte teris from hyre eyȝyn tweyn Out ran, as schour of aprille, ful swythe; Hyre white brest sche bet, and for the peyne, Aftyr the deth cryede a thousent sithe, Syn he that wonyt was hir wo for to lythe, Sche mot forgon; for which disauenture Sche held hire-selue a for-lost creature.

Bk. iv. 806-33 (four stanzas) are omitted; so also are the 18 stanzas referring to Free-Will, viz. Bk. iv. 953-1078. Bk. v. 230-1 runs thus:—

To whom for eueremor myn herte is holde: And thus he pleynyd, and ferthere-more he tolde.

We cannot believe that Bk. iv. 309-15, as here given, can be genuine ; but it seems possible that some of the other readings may be so. The stanza, Bk. iv. 750-6, as here given, seems to represent the first draft of these lines, which were afterwards altered to the form in which they appear in the text, whilst at the same time the stanza was shifted down. However, this is mere speculation; and it must be confessed that, in many places, this MS. is strangely corrupted. Several stanzas have only six lines instead of seven, and readings occur which set all ideas of rime at defiance. Thus, in I. 1260, paste (riming with caste) appears as passede; in I. 1253, ryde (riming with aspyde) appears as rydende; in III. 351, hayes (riming with May is) appears as halis; &c.

Yet the MS. is worth collating, as it gives, occasionally, some excellent readings. For example, in Bk. i. 143, it preserves the word here, which other MSS. wrongly omit; and, in the very next line, rightly has to longe dwelle, not to longe to dwelle.

The MS. has been, at some time, shamefully maltreated by some one who has cut out several leaves, no doubt for the sake of their illuminated initials. Hence the following passages do not appear: I. 1-70; I. 1037—II. 84; III. 1-56; III. 1807—IV. 112; IV. 1667—V. 35; V. 1702—end (together with a piece at the beginning of the Canterbury Tales).

5. MS. H2.—Harleian MS. 3943, in the British Museum. Printed in full for the Chaucer Society in 1875, together with a most valuable line by line collation with Boccaccio's Filostrato, by Wm. Michael Rossetti. Referred to in Prof. Lounsbury's Studies in Chaucer, i. 398, as 'much the worst that has been printed,' where his object is to depreciate its authority. Yet it is well worth a careful study, and it must be particularly borne in mind that it consists of two parts, written at different dates, and of different value. In Bell's Chaucer, we read of it:—'Unfortunately it is imperfect. The first few leaves, and the whole of the latter part of the poem, appear to have been destroyed, and the deficiency supplied by a later copyist.' The late hand occurs in I. 1-70, 498-567, III. 1429-1638, IV. 197—end, and Book V.; and thus occupies a large portion of the MS. Moreover, two leaves are lost after leaf 59, comprising III. 1289-1428; these are supplied in Dr. Furnivall's edition from Harl. 1239, which accounts for the extraordinary disorder in which these stanzas are arranged. The MS. also omits III. 1744-1771, and some other stanzas occasionally.

This is one of those curious MSS. which, although presenting innumerable corrupt readings (the worst being Commodious for Commeveden in III. 17), nevertheless have some points of contact with an excellent source. All editors must have observed a few such cases. Thus, in II. 615, it happily restores the right reading latis, where the ordinary reading gates is ludicrously wrong. In III. 49, it supplies the missing word gladnes. In V. 8, it has 'The Auricomus tressed Phebus hie on lofte,' instead of 'The golden tressed'; and this reading, though false, lets us into the secret of the origin of this epithet, viz. that it translates the Latin auricomus; see note to the line. In the very next line, V. 9, it preserves the correct reading bemes shene, riming with grene, quene, where other MSS. have bemes clere, a reminiscence of the opening line of Book III. Hence I have carefully collated this MS., and all readings of value are given in the Notes. See, e. g. III. 28, 49, 136, 551, 1268, 1703, &c.

6. MS. Harl. 1239 (B. M.). 'It is an oblong folio, written from the beginning in a small, clear character, which ceases at an earlier place [III. 231] than the change occurs in MS. 3943 [IV. 197], leaving the remainder comparatively useless as an authority.'—Bell. Dr. Furnivall has printed the passages in III. 1289-1428, and III. 1744-1771, from this MS. to supply the gaps in H 2 (see above); we thus see that it transposes several of the stanzas, and is but a poor authority.

7. MS. Harl. 2392 (B. M.). A late MS. on paper, not very correct; once the property of Sir H. Spelman. As an example of a strange reading, observe 'O mortal Gower,' in V. 1856. Still, it has the correct reading sheene in V. 9; and in III. 49, supplies the rare reading gladnesse, which is necessary to the sense.

This MS. has a large number of notes and glosses. Some are of small interest, but others are of value, and doubtless proceeded from the author himself, as they furnish useful references and explanations. I here notice the best of them.

II. 8. 'Cleo: domina eloquencie.' This view of Clio explains the context.

II. 784. Side-note: 'nota mendacium.' A remarkable comment.

II. 1238-9. 'Leuis impressio, leuis recessio.' Clearly, a proverb.

III. 933. 'Dulcarnon: i. fuga miserorum.' This proves that Chaucer confused the 47th proposition of Euclid with the 5th; see note.

III. 1177. 'Beati misericordes'; from Matt. v. 7.

III. 1183. 'Petite et accipi[e]tis'; a remarkable comment.

III. 1415. 'Gallus vulgaris astrologus; Alanus, de Planctu Nature'; see note.

III. 1417. 'Lucifera: Stella matutina.'

III. 1466. 'Aurora: amica solis'; shewing the confusion of Tithonus with Titan.

IV. 22. 'Herine (sic), furie infernales; unde Lucanus, me pronuba duxit Herinis.' This proves that Chaucer really took the name from Lucan, Phars. viii. 90, q. v.

IV. 32. 'Sol in Leone'; i. e. the sun was in Leo; see note.

IV. 600. 'Audaces fortuna iuuat'; error for 'Audentes'; see note.

IV. 790. 'Vmbra subit terras,' &c.; Ovid, Met. xi. 61.

IV. 836. 'Extrema gaudii luctus'; see note.

IV. 1138. 'Flet tamen, et tepide,' &c.; Ovid, Met. x. 500.

IV. 1504. 'Non est bonum perdere substantiam propter accidens.'

IV. 1540. 'Styx, puteus infernalis.' Chaucer's mistake.

V. 8. 'The gold-tressed Phebus,' glossed 'Auricomus Sol'; which is from Valerius Flaccus; see note.

V. 319. Reference to Ovid's Metamorphoses; see note.

V. 655. 'Latona, i. luna'; shewing that 'Latona' is mis-written for 'Lucina.' Cf. IV. 1591.

V. 664. Reference to Ovid, Metam. ii. See note.

V. 1039. For 'she,' MS. has 'he,' correctly (see note); side-note, 'Nota, de donis c. d.', i. e. of Criseyde to Diomede.

V. 1107. 'Laurigerus'; see note.

V. 1110. 'Nisus,' glossed 'rex'; 'douhter,' glossed 'alauda'; see note.

V. 1548. 'Parodye: duracio'; see note.

V. 1550. 'Vnbodye: decorporare.'

There are many more such glosses, of lesser interest.

8. MS. Harl. 4912 (B. M.). On vellum; rather large pages, with wide margins; five stanzas on the page. Imperfect; ends at IV. 686. A poor copy. In III. 49, it retains the rare reading 'gladnes,' but miswritten as 'glanes.'

9. MS. Addit. 12044 (B. M.). On vellum; five stanzas to the page. Last leaf gone; ends at V. 1820. Not a good copy. In III. 17, it has 'Comeued hem,' an obvious error for 'Comeueden,' which is the true reading. In V. 8, it has 'golden dressed,' error for 'golden tressed.' Note this correct form 'golden'; for it is miswritten as 'gold' or 'golde' in nearly all other copies.

The next four are in the Bodleian Library, Oxford.

10. Arch. Seld. B. 24[**bold?] is the Scottish MS., dated 1472, described in the Introduction to the Minor Poems, where it is denoted by 'Ar.,' and fully collated throughout the Legend of Good Women, where it appears in the foot-notes as 'A.' It seems to be the best of the Oxford MSS., and has some good readings. In III. 17, it has 'Commeued tham' for Commeueden,' which is near enough for a MS. that so freely drops inflexions; and the line ends with 'and amoreux tham made.' In III. 49, it correctly preserves 'gladness.'

11. MS. Rawlinson, Poet. 163. Not a very good copy. It omits the Prologue to Book III. At the end is the colophon:—

I take 'Tregentyll' to be the scribe's name. Besides the 'Troilus,' the MS. contains, on a fly-leaf, the unique copy of the Balade to Rosemounde, beneath which is written (as in the former case) 'tregentil' to the left of the page, and 'chaucer' to the right; connected by a thin stroke. See my 'Twelve Facsimiles of Old English MSS.'; Plate XII.

12. MS. Arch. Seld. supra 56. Small quarto, 8 inches by 5½, on paper; vellum binding; writing clear. A poor copy. The grammar shews a Northern dialect.

13. MS. Digby 181. Incomplete; nearly half being lost. It ends at III. 532—'A certayn houre in which she come sholde.' A poor copy, closely allied to the preceding. Thus, in III. 17, both have moreux for amoreux; in III. 2, both have Adornes; in III. 6, both absurdly have Off (Of) for O; and so on.

14. MS. L. 1, in St. John's College, Cambridge. A fair MS., perhaps earlier than 1450. Subjoined to the Troilus is a sixteenth century copy of the Testament of Creseide. Quarto; on vellum; 10 inches by 6½; in 10 sheets of 12 leaves each. Leaf g 12 is cut out, and g 11 is blank, but nothing seems to be lost. It frequently agrees with Cp., as in I. 5, fro ye; 21, be this; 36, desespeyred; 45, fair ladys so; 70, Delphicus; 308, kan thus. In I. 272, it correctly has: percede; in 337, nouncerteyne. In II. 734, it agrees with H.; 735 runs—'And whan hem list no lenger, lat hem leue'; a good line. In II. 894, it has 'mosten axe,' the very reading which I give; and in II. 968, stalkes.

15. MS. Phillipps 8252; the same MS. as that described in my preface to the C. text of Piers the Plowman, p. xix, where it is numbered XXVIII.

16. A MS. in the Library of Durham Cathedral, marked V. ii. 13. A single stanza of Troilus, viz. I. 631-7, occurs in MS. R. 3. 20, in Trinity College Library, Cambridge; and three stanzas, viz. III. 302-322, in MS. Ff. 1. 6, leaf 150, in the Cambridge University Library; all printed in Odd Texts of Chaucer's Minor Poems, ed. F. J. Furnivall, Chaucer Society, 1880, pp. x-xii. In 1887, Dr. Stephens found two vellum strips in the cover of a book, containing fragments of a MS. of Troilus (Book V. 1443-1498); see Appendix to the Report of the Cambridge Antiquarian Society, May 24, 1887; pp. 331-5.

The MSS. fall, as far as I can tell, into two main families. The larger family is that which resembles Cl., Cp., and H. Of the smaller, Cm. may be taken as the type. The description of Cm. shews some of the chief variations. Observe that many MSS. omit I. 890-6; in the John's MS., it is inserted in a much later hand. The stanza is obviously genuine.

§ 26. 'Troilus' was first printed by Caxton, about 1484; but without printer's name, place, or date. See the description in Blades' Life of Caxton, p. 297. There is no title-page. Each page contains five stanzas. Two copies are in the British Museum; one at St. John's College, Oxford; and one (till lately) was at Althorp. The second edition is by Wynkyn de Worde, in 1517. The third, by Pynson, in 1526. These three editions present Troilus as a separate work. After this, it was included in Thynne's edition of 1532, and in all the subsequent editions of Chaucer's Works.

Of these, the only editions accessible to me have been Thynne's (1532), of which there is a copy in the Cambridge University Library; also the editions of 1550 (or thereabouts) and 1561, of both of which I possess copies.

Thynne's edition was printed from so good a MS. as to render it an excellent authority. In a few places, I fear he has altered the text for the worse, and his errors have been carefully followed and preserved by succeeding editors. Thus he is responsible for altering io (= jo) into go, III. 33; for creating the remarkable 'ghost-word' gofysshe, III. 584; and a few similar curiosities. But I found it worth while to collate it throughout; and readings from it are marked 'Ed.' The later black-letter copies are mere reproductions of it.

§ 27. The present edition has the great advantage of being founded upon Cl. and Cp., neither of which have been previously made use of, though they are the two best. Bell's text is founded upon the Harleian MSS. numbered 1239, 2280, and 3943, in separate fragments; hence the text is neither uniform nor very good. Morris's text is much better, being founded upon H. (closely related to Cl. and Cp.), with a few corrections from other unnamed sources.

Thanks to the prints provided by the Chaucer Society, I have been able to produce a text which, I trust, leaves but little to be desired. I point out some of the passages which now appear in a correct form for the first time, as may be seen by comparison with the editions by Morris and Bell, which I denote by M. and B.

I. 136; derre, dearer; M. B. dere (no rime). 285. meninge, i. e. intention; and so in l. 289; M. B. mevynge. 388. M. B. insert a semicolon after arten. 465. fownes (see note); M. B. fantasye (line too long). 470 felle, fell, pl. adj.; M. B. fille, i. e. fell (verb). 590. no comfort; M. comfort; B. eny comfort. 786. Ticius (see note); M. Syciphus; B. Siciphus. 896. Thee oughte; M. To oght (no sense); B. The oght (will not scan). 1026. See note; put as a question in M. B.; B. even inserts not before to done. 1050. me asterte; M. may sterte; B. me stert (better).

II. 41. seyde, i. e. if that they seyde; M. B. seyinge (will not scan). 138. were (would there be); M. B. is. 180. wight; M. B. knyght (but see l. 177). 808. looth; M. B. leve. 834. Ye; M. B. The. 1596. For for; M. B. For.

III. 17. Comeveden (see note); M. Comeneden; B. Commodious. him; M. B. hem. 33. io (= jo); M. B. go. 49. M. B. omit gladnes. 572. Yow thurfte; M. Thow thruste; B. Yow durst. 584. goosish; M. goofish; B. gofisshe. 674. M. Thei voide [present], dronke [past], and traveres drawe [present] anon; B. They voyded, and drunk, and travars drew anone. Really, dronke and drawe are both past participles; see note. 725. Cipris; M. Cyphes; B. Ciphis. 1231. Bitrent and wryth, i. e. winds about and wreathes itself; M. Bytrent and writhe is; B. Bitrent and writhen is. Wryth is short for writheth; not a pp. 1453. bore, i. e. hole; M. boure; B. bowre. 1764. to-hepe, i. e. together; M. B. to kepe.

IV. 538. kyth; M. B. right (no sense). 696. thing is; M. B. thynges is. 818. martyre; M. B. matere (neither sense nor rime).

V. 49. helpen; M. B. holpen. 469. howve; M. B. howen. 583. in my; M. B. omit my. 927. wight; M. B. with. 1208. trustinge; M. B. trusten (against grammar). 1266. bet; M. B. beste. 1335, 6. wyte The teres, i. e. blame the tears; M. B. wite With teres. 1386. Commeve; M. Com in to; B. Can meven. 1467. She; M. B. So. 1791. pace; M. B. space (see note).

It is curious to find that such remarkable words as commeveden, io, voidee, goosish, to-hepe, appear in no Chaucerian glossary; they are only found in the MSS., being ignored in the editions.

A large number of lines are now, for the first time, spelt with forms that comply with grammar and enable the lines to be scanned. For example, M. and B. actually give wente and wonte in V. 546, instead of went and wont; knotles for knotteles in V. 769, &c.

I have also, for the first time, numbered the lines and stanzas correctly. In M., Books III. and IV. are both misnumbered, causing much trouble in reference. Dr. Furnivall's print of the Campsall MS. omits I. 890-6; and his print of MS. Harl. 3943 counts in the Latin lines here printed at p. 404.

§ 28. It is worth notice that Troilus contains about fifty lines in which the first foot consists of a single syllable. Examples in Book I are:—

That | the hot-e fyr of lov' him brende: 490. Lov' | ayeins the which who-so defendeth: 603. Twen | ty winter that his lady wiste: 811. Wer' | it for my suster, al thy sorwe: 860. Next | the foule netle, rough and thikke: 948. Now | Pandar', I can no mor-e seye: 1051. Al | derfirst his purpos for to winne: 1069.

So also II. 369, 677, 934, 1034, 1623 (and probably 1687); III. 412, 526, 662, 855 (perhaps 1552), 1570; IV. 176, 601, 716, 842, 1328, 1676; V. 67 (perhaps 311), 334, 402, 802, 823, 825, 831, 880, 887, 949, 950, 1083, 1094, 1151, 1379, 1446, 1454, 1468, 1524.

It thus appears that deficient lines of this character are by no means confined to the poems in 'heroic verse,' but occur in stanzas as well. Compare the Parlement of Foules, 445, 569.

Troilus contains a considerable number of proverbs and proverbial phrases or similes. See, e. g., I. 257, 300, 631, 638, 694, 708, 731, 740, 946-952, 960, 964, 1002, 1024; II. 343, 398, 403, 585, 784, 804, 807, 861, 867, 1022, 1030, 1041, 1238, 1245, 1332, 1335, 1380, 1387, 1553, 1745; III. 35, 198, 294, 308, 329, 405, 526, 711, 764, 775, 859, 861, 931, 1625, 1633; IV. 184, 415, 421, 460, 588, 595, 622, 728, 836, 1098, 1105, 1374, 1456, 1584; V. 484, 505, 784, 899, 971, 1174, 1265, 1433.

§ 30. A translation of the first two books of Troilus into Latin verse, by Sir Francis Kinaston, was printed at Oxford in 1635. The volume also contains a few notes, but I do not find in them anything of value. The author tries to reproduce the English stanza, as thus:—

'Dolorem Troili duplicem narrare, Qui Priami Regis Trojae fuit gnatus, Vt primùm illi contigit amare, Vt miser, felix, et infortunatus Erat, decessum ante sum conatus. Tisiphone, fer opem recensere Hos versus, qui, dum scribo, visi flere.'

For myself, I prefer the English.

§ 31. Hazlitt's Handbook to Popular Literature records the following title:—'A Paraphrase vpon the 3 first bookes of Chaucer's Troilus and Cressida. Translated into modern English ... by J[onathan] S[idnam]. About 1630. Folio; 70 leaves; in 7-line stanzas.'

P. 8, Book I, met. 4, l. 8. For thonder-light a better reading is thonder-leit; see p. xliii, and the note (p. 422).

P. 10; foot-notes, l. 10. Read: C. vnplitable; A. inplitable.

P. 26, Book II, met. 1, l. 11. For proeueth read proeveth.

P. 29, Book II, pr. 3, l. 3. Delete the comma after wherwith.

P. 48, Book II, pr. 7, l. 86. For thas read that.

P. 50, Book II, pr. 8, l. 17. For windinge read windy. See pp. xlii, 434.

P. 58, Book III, pr. 3, l. 68. For all read al.

P. 62, l. 4. Counted as l. 10; it is really l. 9.

P. 63, Book III, pr. 5, l. 41. For of read of (in italics).

P. 74, Book III, pr. 10, l. 6. For has read hast.

P. 111. The side-number 215 is one line too high.

P. 122, Book IV, met. 6, l. 24. Delete the square brackets; see pp. xlii, xliii.

P. 124, Book IV, pr. 7, l. 61. MS. C. has confirme; and MS. A. has conferme. But the right reading must be conforme; for the Latin text has conformandae.

P. 159, Book I, 204. For cast read caste.

P. 160, Book I, 217. The alternative reading is better; see note, p. 463.

P. 160, Book I, 239. For yet read yit (for the rhyme).

P. 162, Book I, 284. For neuer read never.

P. 163, Book, I, 309. For Troylus read Troilus.

P. 163, Book I, 310. For thyng read thing.

P. 165, Book I, 401. Alter ! to ?

P. 166, Book I, 406. For thurst read thurste.

P. 166, Book I, 420. For deye read dye (for the rhyme).

P. 171, Book I, 570. For euery read every.

P. 172, Book I, 621. For Troylus read Troilus (as elsewhere).

P. 173, Book I, 626. Delete the comma after 'fare.'

P. 174, Book I, 656. For y read I.

P. 174, Book I, 657. Insert ' at the beginning.

P. 181, Book I, 879. For the read thee.

P. 192, Book II, 113. Delete ' at the end.

P. 194, Book II, 170. Insert ' at the beginning.

P. 205, Book II, 529. For penaunc read penaunce.

P. 208, Book II, 628. For swych read swich.

P. 229, Book II, 1294. Insert ' at the beginning.

P. 234, Book II, 1461. For streyt read streght, as in MS. H.

P. 260, Book III, 522. Delete the comma after laft.

P. 260, Book III, 535. For made read mad or maad.

P. 261, Book III, 558. For lengere read lenger.

P. 264, Book III, 662. For thondre read thonder.

P. 271, Book III, 885. For ringe read ring.

P. 282, Book III, 1219. For sweet read swete.

P. 312, Book IV, 318. For to the peyne read to my peyne.

P. 390, Book V, 1039. For she read he. Cf. note, p. 499; and p. lx, l. 3.

P. 431, note to Prose 5, 35; l. 3. Delete for which I find no authority. (In fact, postremo is the reading given by Peiper, from one MS. only; most MSS. have postremae, the reading given by Obbarius, who does not recognise the reading postremo).

P. 463. Note to I, 217. Add—So too in Barbour's Bruce, i. 582: 'Bot oft failyeis the fulis thocht.'

P. 479, last line; and p. 480, first line. For represents the Pers. and Arab. dū`lkarnayn, lit. two-horned; from Pers. dū, two, and karn, horn—read represents the Arab, zū`lkarnayn, lit. two-horned; from Arab. zū, lord of, hence, possessing, and the dual form of karn, horn.

Notes to I. 948, 951; II. 36, 1335; III. 1219. Dr. Köppel has shewn (in Archiv für das Studium der neueren Sprachen, xc. 150, that Chaucer here quotes from Alanus de Insulis, Liber Parabolarum (as printed in Migne, Cursus Patrologicus, vol. ccx). The passages are:—

Fragrantes uicina rosas urtica perurit (col. 582).

Post noctem sperare diem, post nubila solem; Post lacrimas risus laetitiamque potes (583).

Mille uiae ducunt homines per saecula Romam (591).

De nuce fit corylus, de glande fit ardua quercus (583).

Dulcius haerescunt humano mella palato, Si malus hoc ipsum mordeat ante sapor (592).

P. 498, Note to V, 806. Add—L. 813 is due to Dares; see p. lxiv, note.

P. 499, Note to V, 1039, l. 6. For the rest is Chaucer's addition read the statement that she gave it to Diomede is due to Benoît; see p. lxii. Again, just below, read The incidents of the 'broche' and 'pensel' are also due to the same; see p. lxii.