'The Cokes Tale of Gamelyn.' That was done because it happened to be found next after the Cook's Tale, which, instead of being about Gamelyn, is about Perkin the reveller, an idle apprentice.

The fitness of things ought to shew at once that this Tale of Gamelyn, a tale of the woods, in the true Robin-Hood style, could only have been placed in the mouth of him 'who bare a mighty bow,' and who knew all the usage of woodcraft; in one word, of the Yeoman. (Gandelyn is the name of an archer in Ritson's Ancient Songs, i. 82). And we get hence the additional hint, that the Yeoman's Tale was to have followed the Cook's Tale, a tale of fresh country-life succeeding one of the close back-streets of the city. No better place can be found for it.

There is yet one more Tale, found only in some of the earlier printed editions, but in none of the MSS., viz. the Ploughman's Tale. This is admittedly spurious, in the sense that it is not Chaucer's; but it is a remarkable poem in its way. The author never intended it for an imitation of Chaucer, nor pretended any disguise about it; on the contrary, he says plainly that he was the author of the well-known poem in alliterative verse commonly known as Pierce the Ploughman's Crede. It can only have been inserted by inadvertence, but we need not blame Thynne for doing this, since otherwise the poem would not have been preserved at all, no MS. of it being now in existence.

The next question that presents itself is this-Have we any means of telling which of the Tales are of early, and which of late workmanship? In reply to this, we may note, in the first place, the following facts and probabilities.

The Knight's Tale was almost certainly re-written from beginning to end. In the first instance Chaucer took a good deal of it from Boccaccio's Teseide, and gave it in the name of Palamon and Arcite; see Prologue to Legende of Good Women, 1. 420. This he would naturally do just after or just before writing his Troilus', in which he follows the same author, and

1 Several lines are common to Troilus and to the Knight's Tale, shewing that the former and Palamon and Arcite' were probably in hand together.

he would naturally employ the seven-line stanza. But this is not all, for it is obvious upon comparison (though I do not remember to have seen it pointed out before) that Chaucer also pressed into his service, when writing the Knight's Tale, a poem also in the seven-line stanza, which has been preserved under the title ‘Of Queen Annelida and False Arcite.' In this poem, after three introductory stanzas, he quotes three lines from Statius, beginning—‘Iamque domos patrias,' &c.; and it is not a little remarkable that the very same three lines reappear as a heading at the beginning of the Knight's Tale in many of the MSS. It is interesting to note the traces of resemblance between this poem and the Knight's Tale, but it must be admitted that they are very few, such as these :

'With Emely her yonge suster schene'—

which reappears in the Knight's Tale, l. 114; with a few similar phrases. For example, the first three lines of the prologue run thus:

O thou fiers God of armes, Mars the rede,
That in thy frosty country called Thrace,
Within thy grisly temples full of drede'—

which may be compared with the Knight's Tale, 1111-1115. The general story is, however, widely different, and Chaucer used up the latter part of it, not in the Knight's Tale, but in the Squire's Tale. I draw attention to this poem chiefly in support of my opinion, to which I shall have occasion to recur, that the early draught of Palamon and Arcite was in seven-line stanzas.

It must next be noted that Mr. Furnivall, who has drawn up, tentatively, a list of Chaucer's works in their supposed order, puts down amongst the works of the 'Second Period,' i. e. prior to the Canterbury Tales, that Tale which is now known as the Second Nun's, though formerly called by Chaucer himself the Life of Saint Cecile. Of this result there has never been a doubt; Tyrwhitt says expressly, 'The Tale of the Nonne is almost literally translated from the Life of St. Cecilia in the Legenda Aurea of Jacobus Januensis. It is mentioned by Chaucer as a separate work in his Legende of Good Women,

1. 426, under the title of the Life of Seint Cecile, and it still retains evident marks that it was not originally composed in the form of a Tale to be spoken by the Nonne1.' It is, then, little more than a translation, and it is in seven-line stanzas.

Mr. Furnivall assigns to this Second Nun's Tale the conjectural date of 1373; now this is the very year when Chaucer met Petrarch at Padua (see note to E 27), and learnt from him the tale of Griseldis, now known as the Clerk's Tale. This tale is, for the most part, a translation, and it is in seven-line stanzas.

The Prioress's Tale has a Proem much better suited for a formal poem than for a Tale to be told, being much in the same strain as one of the author's other poems, known as Chaucer's A. B. C. Moreover, it is (by an oversight) still called a song; see B 1677. This poem is also in seven-line stanzas.

The Monk's Tale is in a very peculiar metre, which appears nowhere else in Chaucer, except in the above-mentioned poem called the A. B. C. (perhaps written before A.D. 1369), and in some of Chaucer's latest but very short poems, such as the Envoy to Bukton, and the Ballad of the Visage2 without Painting; so that, considered with reference to metre, this Tale may be of any date. The main part of it shews no great originality, and seems to me rather early than late.

Having premised these considerations, I wish now boldly to state that we have, in fact, one test of earliness or lateness on which we may rely, I believe, with great confidence. It is a test so obvious that it is a wonder to me that no one, as far as I know, has pointed it out before; I mean the test of rhythm. The canon I propose is simply this. Nearly all of Chaucer's tales that are in stanzas are early, and nearly all that are in the usual couplets are late. We have seen that this is known

6

1 In the Proem, the Nun calls herself an unworthy son of Eve.'

2 Oddly spelt Vilage in the MSS.; but the poem is imitated from Boethius, and has special reference to the passage-This ilke Fortune hath departyd and vncoueryd to the bothe the certeyn visages, and eke the dowtos visages of thy felawes'; Chaucer's Boethius, ed. Morris, p. 62.

to be true in the case of the Second Nun's Tale, that it is highly probable in the case of the Clerk's Tale (of which more hereafter), and there is nothing against it in the case of the Monk's Tale, written in the same metre as a poem which is said to have been his very first, or nearly so, if there be any truth in the statement that it was written for the use of the Duchess Blanche, who died in 1369. At the same time, I suppose 'Palamon and Arcite' to have been written in stanzas, so that the present metre of the Knight's Tale presents no difficulty. Of course it will be understood that there is, in these stanzatales, some of Chaucer's latest work, but I shall presently shew that this late work is easily picked out.

The above canon is due to no fancy, but to the simple fact, that Chaucer cannot be proved to have used his couplets till he was well advanced in composition. Indeed, it has always been remarked that no English poet before him ever dreamt of such a metre, and it has been a source of wonder, for hundreds of years, whence he derived it. To say that it was derived from the French ten-syllable verse is not a complete solution of the mystery; for nearly all such verse is commonly either in stanzas, or else a great number of successive lines are rimed together. What we desire is to find a specimen of French ten-syllable verse in which only two successive lines are rimed together; and these, I believe, are very scarce. After some search I have, however, fortunately lighted upon a very interesting specimen, among the poems of Guillaume de Machault, a French writer whom Chaucer is known to have imitated 1, and who died in 1377. In the edition of Machault's poems edited by Tarbé, Reims and Paris, 1849, p. 89, there is a poem of exactly this character, of no great length, but fortunately dated; for its

1 See Specimens from Chaucer's Book of the Duchess as compared with some from Machault's Remède de Fortune in Furnivall's Trial Forewords p. 47, where he quotes from Étude sur G. Chaucer, by M. Sandras, p. 290. The obligations to the Remède de Fortune are somewhat doubtful (Trial Forewords, p. 115): but there are other instances which go to shew that Chaucer had read Machault; see Professor Ten Brink's note (at the same reference) and the last note in Tyrwhitt's notes to the Canterbury Tales.

title is 'Complainte écrite après la bataille de Poitiers et avant le siège de Reims par les Anglais' (1356-1358). The first four lines run thus:

'A toy, Henry, dous amis, me complain,

Pour ce que ne cueur ne mont ne plein1;
Car a piet suy, sans cheval et sans selle,
Et si n'ay mais esmeraude, ne belle.'

The last couplet (the second line of which has two examples of the fully-sounded final e) is as follows:—

Et que jamais ne feray chant ne lay,
Adieu te di: car toute joie lay.'

Now as Chaucer was taken prisoner in France in 1359, he had an excellent opportunity for making himself acquainted with this poem, and with others, possibly, in a similar metre which have not come down to us. It is also almost certain that the earliest attempt to use this metre in English was made by Chaucer, in his Legend of Good Women, commenced, according to Professor Ten Brink, in the year 1385 (Furnivall's Trial Forewords, p. 111). Surely this date is one of considerable importance; for we at once derive from it the probability that all of the Canterbury Tales written in this metre were written after 1385, whilst those not in this metre were probably earlier. With this to guide us, I can now proceed to discuss separately such of the Canterbury Tales as are printed in the present volume.

Man-of-Law Head-link. This is an important passage, as it gives the date (April 18) of one of the days of the pilgrimage, and a list of the Tales which Chaucer meant to include in his Legend of Good Women. These points are discussed in the notes to ll. 3 and 61, which see. The metre, by the canon, shews late or new work, as the subject-matter proves.

Prologue to the Man of Law's Tale. The metre would, by the canon, indicate early work, yet it is obviously not so.

1 Observe particularly this rime of complain with plein. This shews whence Chaucer derived such rimes as seke, seke; Prol. 17, 18. There is a poem of 92 lines called Le Dit de la Harpe, printed in Bartsch's Crestomathie Française, p. 408, in which more than half the rimes are of this character.