and strengthening the web of his poetry with thoughts and reflections culled wherever he could find them. His prose translation of Boethius and his study of Dante now came to help this reflective vein, and on the other hand he had made acquaintance with two of Boccaccio's masterpieces, not the Decamerone, which it is probable he never knew, but the Teseide and the Filostrato. The story of Palamon and Arcite, which, after at least one recasting, has come down to us as the Knightes Tale, represents his work on the Teseide, and Troilus and Criseyde that on the Filostrato, and these two splendid poems, full of all the colour of mediæval chivalry and love and thought, relieved ever and anon with subtle touches of humour, are the striking achievements of his middle period. In the Parlement of Foules, written in 1382, he returns, to please the Court, to the French models of his earlier days, only to show how far he had progressed since the Boke of the Duchesse of thirteen years earlier. In the Hous of Fame he is much less happy. I think there can be no doubt that Dr.

Heath is right in his conjecture (see his Introduction to the poem) that Books i. and ii. were separated from Book iii. by some interval, but the poem raises many difficulties, some of which we are not likely ever to solve. Perhaps it is not amiss to remark here that Chaucer, though one of the world's great story-tellers, is not remarkable for inventiveness. Probably all, or nearly all, of his plots are borrowed, and in the fourteenth century books from which he could borrow were not easily come at. He had brought back the Teseide and Filostrato from one of his visits to Italy, and perhaps had strained his purse to do it; but when he had used them he was thrown back on the rather jejune material he could find in the books around him. At an earlier period he had probably been driven by some such straits to compile the dreary tragedies of misfortune which we know as the Monkes Tale. In the first two Books of the Hous of Fame we find him narrating or alluding to almost all the tragedies of hapless love which he soon set himself to tell, till he wearied of them, in the Legende of Good Women. The third Book of the Fame is in quite a different style, cast in Chaucer's happy discursive vein, and only failing for lack of a climax. Legende it is the Prologue, in its two drafts, which gives him his opportunity. the nine stories of loving women which he had patience to complete, only the first three (those of Cleopatra, Thisbe, and Dido) are in any way worthy of him.

In the
Of

The Legende of Good Women was no doubt abandoned from sheer weariness with its monotonous theme, and it was perhaps Chaucer's sense that this monotony must be avoided at all costs that caused him to conceive the plan of the Canterbury Tales, of which diversity, the exchange of stories between gentle and simple, bookmen and the bookless, the religious and the irreligious, is the very essence. Once more the scheme was left unfinished, but in this case there is little to regret. If indeed Chaucer had been in the mood, he might have described the adventures of the pilgrims at Canterbury, and the final supper at the Taberd on their return to Southwark, with all the richness of humour which marks the General Prologue or that of the Wife of Bath. But there is some gain in being left with the picture of the pilgrims as still journeying along the Kentish roads, and as for the Tales, they run the whole length of the gamut, and seem to leave no note wanting. As is generally agreed, some of the tales of the gentle folk had probably been written at earlier dates, and had now only to be revised and fitted into their places, but his scheme gave Chaucer an excuse for displaying the same mastery in the broad humours of narration as he had shown in his Troilus and Knightes Tale in the fields of romance. It is too true that several of these tales must be reckoned among those which, as the Retraction phrases it, sounen into sin,' but it is as unfair to take them too seriously as it would be to expose the essential immorality of most fairy-tales, and there can be no question as to the extraordinary

skill with which the tales of the Miller, Reeve, and Summoner, no less than the gentler humours of that of the Nun's Priest, are set forth.

Along with their many masterpieces of humour and romance, the Canterbury Tales contain some poorer stories, the very feeble version of the death of Virginia, for instance, and the Manciple's tale of Phoebus and the Crow, and it is not easy to tell whether these represent earlier work foisted into the cycle, or whether we have here the fruits of Chaucer's failing powers. It needs some acquaintance with the workings of the medieval mind to imagine how, at any period of his career, he could have cared to set forth the weariful prose discourses of Dame Prudence. The Parson's sermon, long as it is, is much more endurable, and though nobody is likely, except for professional reasons, to read it through, as I have done, at least six times, the task is not so repellent as might be imagined. The prose treatise on the Astrolabe, written for little Lewis Chaucer in 1391, though only a tenth of its length, is much more formidable. But in all his prose work Chaucer is merely as any other fourteenth century writer, without a touch of the grace and humour with which his poems are filled. As a poet he needs to-day no one to praise him. He has been praised already, wisely and well, by many clever writers. All that is now needed is that the praise shall no longer be taken contentedly on trust, but that his poems, which in their freshness and restfulness must in this century have more power of pleasure-giving than ever before, should be allowed to speak for themselves to ears no longer deaf.

THE CANTERBURY TALES

(ALFRED W. POLLARD)

The Canterbury Tales are given the place of honour in this edition partly out of deference to a time-honoured precedent, which might fairly claim some weight even against the chronological arrangement which commends itself to modern scholarship, but partly also because their assignment to any other position would be misleading. In addition to two long treatises in prose they contain some 18,000 lines of verse, and it is quite certain that not all of these 18,000 lines sprang from Chaucer's brain after he had conceived the plan which was to link together this wonderful medley. That one, at least, of the tales was written at an earlier period of his career we have clear evidence. In the Prologue to the Legende of Good Women we find the Second Nun's Tale already alluded to as the Lyf of Seint Cecyle, and in its introduction the narrator is made to speak as an unworthy sone of Eve' (1. 60) instead of as a woman, and to address those that reden that I write' (1. 78) instead of the listeners to a tale told along the highway to Canterbury. Again, with our suspicions thus aroused, we note Chaucer's distinct statement that he learnt the story of Grisilde at Padua of Fraunceys Petrak,' who died in 1374, and whom the English poet may have met on his Genoa mission of 1373, when Petrarch was living at Arqua, near Padua. Chaucer was not so well off for subjects for it to be probable that if he learnt this story from Petrarch in 1373 he would have left it unused for a dozen years or more, and there is a general agreement in the belief that he wrote his English version of Petrarch's Latin shortly after his return to England. In the Monk's Tale, again, the wearisome tragedies fall into two distinct groups, one of twelve stories of old time, derived from the Bible, Boccaccio's De Casibus Virorum et Feminarum Illustrium and De Claris Mulieribus, and the Roman de la Rose; the

4

other, of five modern instances, mostly very briefly treated, and one of them recording the death of Bernabo Visconti, Lord of Milan, which occurred as late as 1385. One of the modern stories, that of Ugolino of Pisa, is partly taken from Dante, and is strikingly better than all the rest. In the early stories, though the verse is good enough, the treatment is often careless and unsympathetic, and Chaucer was clearly not interested in them. It cannot be said dogmatically that they show early work, but it seems probable that at some time towards the close of the decade 1369-1379 (to which, it must be remembered, there is strikingly little of his poetry which can be positively assigned) Chaucer began a poem on the same plan as that afterwards adopted by his follower Lydgate in his Falls of Princes, and then abandoned it till the need came to suit the Monk with an unexpected but appropriate theme, when it was revised and enlarged. The Man of Lawes Tale, once more a curiously inappropriate one, is cast in the same seven-line stanza as the Seint Cecyle and the Grisilde, and from its subject, style, and tone appears to have been written towards the close of the same period. On the other hand, the Prioress's Tale of the little chorister, though it goes back in feeling to this earlier period, is clearly written after the conception of the plan of the Canterbury Tales, as is proved by the 'quod she' with which the narration is interrupted (B 1644), while its ripe and mature beauty fully agrees with this evidence.

Whether any of the other Tales-all of which, except the Sir Thopas parody, are written in heroic couplets-should be assigned to a date earlier than the immortal General Prologue, is a point much more difficult to determine. Outside the Canterbury Tales the only extant poem in which Chaucer used the heroic couplet is the Legende of Good Women, and as this certainly preceded the Canterbury Tales as a whole, there is a general inclination to regard this as Chaucer's first essay in the couplet, rather than to give any individual Tale precedence over it. On the other hand, there is an allusion in the already oft-quoted list of Chaucer's works in the Legende to a poem enshrining

Al the love of Palamon and Arcyte

Of Thebes, thogh the story is knowen lyte.

It is difficult to believe that the reference here is to the fragment of Queen Anelida and Fals Arcyte which has come down to us, as it ought to point to a poem which kept much more closely to the loves of the two knights as narrated in the Tescide. Our natural inclination would therefore be to identify this poem with the Knightes Tale, as we now have it, but the ingenuity of Chaucer's commentators has discovered that there are ten seven-line stanzas translated from the Teseide in Anelida and Arcyle, sixteen in the Parlement of Foules, and three in Troilus and Criseyde. Hence has arisen a theory that in addition to the Anelida and the Knightes Tale Chaucer composed a more literal translation of the Teseide in seven-line stanzas, subsequently withdrew it from circulation, and used some of his old material in later poems. Ingenious as this theory is, the supposition of the writing and suppression of a poem, necessarily of considerable length, is no light matter, and if Chaucer really wrote such a poem and subsequently used fragments of it in other works it is extraordinary that he should have called attention to a tale thus cruelly treated by an entirely gratuitous reference in the Legende. As for the fragments of the Teseide found in the three seven-line poems, there is a parallel instance, of the nearly simultaneous use of the same material in two different metres, in the story of Dido and Æneas, which we find first in the octosyllabic couplets of the Hous of Fame, and again in the decasyllabic couplets of the Legende of Good Women. On the whole,

and with all deference to the great authority of the scholars who have held the opposite view, it seems best to regard the theory of a lost seven-line version of Palamon and Arcyte as a needless hypothesis. If this be so, the reference in the Legend must be almost certainly to the Knightes Tale, and this fine poem is thus brought back nearer to the period of the Troilus, with which it is so closely allied in style and temper.

If the Knightes Tale is thus brought back, other Tales, notably those of the Franklin (one of Chaucer's great successes) and the Squire, may perhaps come with it, and we need not hesitate, on the score of their metre, to relegate such poor work as the story of Appius and Virginia as told by the Doctor of Phisik, and the Manciple's tale of Apollo and the Crow, to a less happy period of Chaucer's career than that in which he was writing the Prologue and others of his finest works. Without wishing to press this point too far, it seems fair to point out that there is nothing unreasonable in supposing that when Chaucer conceived his immensely ambitious scheme of the Canterbury Tales he had a really considerable amount of material already at his disposal. It is sufficient, however, here to emphasise the fact that inclusion in the Canterbury series of itself tells us absolutely nothing as to the date at which any given poem was written, and that we must therefore place the Tales as a whole entirely outside the chronological sequence of the poet's other works.

As regards the date at which the idea took shape of a Canterbury Pilgrimage as a framework by which to connect a number of otherwise distinct stories, we have only two or three years from which to choose, and we must not attempt to pin it down too precisely to any one of them. We have various good reasons for believing that the six years which succeeded 1379 produced the Boece, Troilus, Parlement of Foules, Hous of Fame and Legende of Good Women, and it is therefore inconceivable that Chaucer should have planned the Canterbury Tales earlier than the end of 1385 or beginning of 1386. Again, no one who has read the talks by the way can doubt that the poet himself had travelled over the ground, while we know that until on 17th February 1385 he was permitted to appoint a deputy in his Comptrollership he was closely tied to his official work, a bondage of which he complains bitterly in the Hous of Fame. Chaucer's own pilgrimage, then, may have been made in 1385 or in any subsequent year, but hardly before this. On the other hand, the short poems written towards the close of his life show that the not very advanced age to which he attained pressed heavily on him, and it would be unreasonable to assign the plan of the Tales to his last decade. If, as is highly probable, the Legende was begun in 1385 and soon afterwards left unfinished in despair, everything points to the scheme of the Canterbury Tales as taking form during the next two or three years, 1386-1388. Nearer than this it is not easy to go with safety, for in drawing conclusions from the indications of date which we find in the talks by the road we must remember that Chaucer may have fitted them in either to the year in which he was writing, or back to the year in which he himself took his holiday. In the latter case the dates would be more likely to be real dates, while if we prefer to believe that they are taken from the year in which he was writing, we can hardly imagine that Chaucer was likely to trouble himself to consider too curiously whether this or that week would be a convenient one for some of his imaginary characters to make their pilgrimage. Thus, in drawing conclusions from the mention of 18th April in the talk which precedes the Man of Law's Tale (B 5), I do not think we can absolutely rule out of court the year 1386, on the ground that in that year 18th April fell in Holy Week, when the Parson and others would be much in

request for the duties which the season imposed on them,'1 or reject 1388 because 19th April then fell on a Sunday, and if Sunday travelling had been intended, something would have been said about the hearing of mass. With this caution, however, I am quite prepared to accept Professor Skeat's assurance that in 1387 'everything comes right,' since the pilgrims could assemble at the Tabard on Tuesday, 16th April, with four clear days before them, and the journey ending conveniently on a Saturday. Whether we should assign this year to that of Chaucer's own pilgrimage, or to that of his imaginary pilgrims, must remain undetermined. In any case we cannot be wrong in believing that in or about 1387 is the most probable date for the Canterbury Tales to have been begun. As to whence the idea of this particular framework for story-telling came to the poet, out of his own head' seems in every way the best answer. Certainly there is no shred of evidence to prove that he copied it from the very inferior scheme of Boccaccio's Decamerone.

The fame of Becket's shrine, the popularity of the pilgrimage to it, and the mediæval habit of turning a pilgrimage into a kind of religious holiday, are all matters of such common knowledge that they do not need illustrating here. Nor need we

stop to prove the futility of the idea once current, that the pilgrims were in so great a hurry to bring their holiday to an end as to have accomplished the then well-nigh impossible feat of travelling fifty-six miles over heavy roads in a single day. In 1358 the queen-mother Isabella, on her own pilgrimage, left London 7th June, slept that night at Dartford, slept at Rochester on the 8th, and at Ospringe on the 9th, and reached Canterbury the next day. Two years later John of France slept at Dartford 1st July, dined there next day, slept at Rochester on the 2nd, dined at Sittingbourne and slept at Ospringe on the 3rd, and reached Canterbury 4th July. The records of other fourteenth century journeys confirm the presumption that Dartford, Rochester, and Ospringe (where some traces of the old Pilgrim's House still exists) were the regular sleeping-places on the road, and there can be no doubt that Chaucer intended his pilgrims to make the journey by these stages, and to take four days over it.

As to the exact route they followed some little uncertainty prevails, owing to the line of the modern road not coinciding everywhere with that of the old pilgrim's way,' but we have references to Deptford and Greenwich in the talk before the Reeve's Tale (A 3906, 3907), to Rochester in the Host's address to the Monk (B 3116), to Sittingbourne in the quarrel between the Friar and the Summoner (I) 847), and to Boughton-under-Blee in the Canon's Yeoman's Prologue (G 556), and to the still mysterious Bobbe-up-and-doun, under the Blee,' in the Manciple's (II 2). Rochester could not possibly be reached after Sittingbourne, and guided by this fact Henry Bradshaw and Dr. Furnivall were able to correct a mistake in arrangement, found even in the best MSS., by which the five Tales of the Shipman, Prioress, Chaucer, the Monk, and the Nun's Priest (all linked together by the talks on the road) were placed immediately before that of the Second Nun, instead of between that of the Man of Law (with which the tales of the second day were begun) and that of the Wife of Bath, in which Sittingbourne is mentioned. By a less necessary alteration the position of the Tales of the Doctor and Pardoner, which in the best manuscripts come before the Shipman's group, were brought back along with it, but placed after instead of before. There are no references to place or time in these two tales, so that the alteration matters little either way, and we now have the twenty-four extant tales and fragments in a reasonable and probable order. Some of these tales (as has been mentioned in the case of the Shipman's group) are linked together by references, backwards or forwards, in the talks on the road; in other cases there is no link of any kind between 1 Skeat's Chaucer, vol. iii. p. 373.