Tales, viz.:-The Cook's Tale has in Rawl. poet. 141 four additional lines, no Tale of Gamelyn. It has in Bodley 686 twelve additional lines; see Urry and Tyrwhitt. A link CookGamelyn of four lines is in Lansdowne; one of two lines is in Royal 18, Sloane 1685, Barlow, Hatton, Laud 739, Mm, Petworth, Egerton 2863, Hodson-Ashburnham. (The conditions in Ii, Christ Church, Trinity 49, are not known to me.) A link Merchant-Wife of 16 lines is in Barlow, Laud 739, Royal 18. A link Franklin-Doctor of six lines is in Harley 7335. A link Canon's Yeoman-Doctor of 14 lines is in Selden, Royal 17, Royal 18, Rawl. poet. 149, Petworth, Mm, Hatton, Sloane 1685, Barlow, Egerton 2863, Laud 739; one of 16 lines is in Lansdowne; there is a gap in Harley 1758 at this point. There is a spurious prologue to Thopas, four lines, in R 3, 3. A link of 12 lines Pardoner-Shipman is in Harley 1758, Rawl. poet. 149, Petworth, Mm, Hatton, Ii, Sloane 1685, Barlow, Laud 739, Royal 18, Egerton 2863; a link of six lines is in Lansdowne. There are eight spurious lines concluding the Squire's Tale in Selden and in Lansdowne; the latter has four more introducing the Wife of Bath, who follows.

The spurious link Cook-Gamelyn (Lansdowne) is printed in Wright's Cant. Tales I: 175; the 16-line link MerchantWife, ibid. I 245-6 note; the spurious (Lansdowne) link Pardoner-Shipman, ibid. II: 283 note; the link Canon's Yeoman-Doctor is printed ibid. II: 245 note; the connection SquireWife (Lansdowne) ibid. II: 157 and I: 246 note.

The Host. Another point of superiority which the Tales possess over any similar set of stories is the figure of the Host. Already in the Biographia Britannica, 1747, art. Chaucer, note S, we find:-"Above all the character of the Host, who acts as a kind of Chorus in the ancient Drama, is most admirably kept up, and the same wit, spirit, and humour is preserved through the whole journey, that strikes and astonishes the reader so much at the very beginning, where the original character of this incomparable person is drawn at full length." Compare: "This type of a sturdy, well-to-do burgess, at a time when England still deserved the name of 'Merry England', was excellently fitted to play the part of the chorus in the varied drama of the procession to Canterbury. He represents most perfectly the magnanimous toleration, the serene benevolence, the easy and humane disposition which lend such a refreshing effect to Chaucer's magnificent poem." Ten Brink, Hist. Eng. Lit. II 147. And, still better: "The hearty positive genius of the English people keeps watch, throughout this final masterpiece of Chaucer, upon the irresponsible literary and poetic instincts of rhetoric, sentiment, romance, which in later days have

habitually scorned its control." Herford, introd. to English Tales in Verse, 1902, pp. xxix-xxx. On the Host see Todd p. 265 ff.; Skeat V : 58 ff.

It has been noted, though not yet sufficiently discussed, that in Chaucer's Links, and especially in the comments, are to be found the beginnings of English literary criticism. Compare Courthope's remark, Hist. of Eng. Poetry I: 259-60, and Saintsbury, Hist. of Criticism I: 450; both of these observations are based upon the Rime of Sir Thopas, while a mass of other material remains untouched.

The "Fragments" of the Canterbury Tales. Chaucer's plan for the Canterbury pilgrimage included, as it is sketched in the Prologue, two tales for each of the thirty pilgrims on the outward journey and two on the return; this scheme he later modified (Parson's headlink lines 15-16) by reducing it onehalf; but of even this diminished plan only some twenty-four tales remain to us, several of which are incomplete. We cannot doubt that he intended to arrange the tales in a certain definite order, so as to obtain contrast and variety; note how the Knight, the most dignified person in company, is followed by the drunken and irresponsible Miller instead of by the person next in rank, and how the Miller is angrily answered by the Reeve. The whole body of tales, if the plan had been fully worked out, would have been bound together by these Links, and the Prologue at the beginning would have been balanced by an account of the supper to the winner at the end of the return journey. But in the incomplete sketch which remains to us of this extensive plan, the twenty-four tales which we have are not connected into a coherent whole; at a number of places a gap occurs owing to the absence of any Link. Thus, the Prologue passes directly into the Knight's Tale; the Knight is followed at once by the Miller, and he by the Reeve; the Cook then comments on the Reeve's tale and begins one of his own, which however is left unfinished by Chaucer. We have therefore no certainty whose tale was to have come next, there being no link; and this group of tales is generally spoken of as Fragment A of the Canterbury Tales. The remaining stories are similarly bound together into "fragments"; e. g., there is a link binding the tale of the Pardoner to that of the Doctor which precedes it, but no link shows who preceded the Doctor or who followed the Pardoner. And the whole mass of the Canterbury Tales thus falls into ten fragments.

How Chaucer would finally have arranged these fragments which remain we cannot with absolute certainty say. We can, of course, recognize the first, or A-fragment, because of its inseparable connection with the Prologue; and it is definitely

stated in the Parson's prologue that his Tale is the last of the outward journey. But the sequence of the other eight fragments is matter for discussion. These fragments are, adopting for convenience the lettering and order favored by the Chaucer Society:

A. The General Prologue, the tales of Knight, Miller, Reeve, and Cook, with the links connecting the several tales. The spurious tale of Gamelyn is in many MSS found just after the Cook's fragmentary tale.

B'. The Man of Law's tale.

B2. The tales of Shipman, Prioress, the Rime of Sir Thopas, the tale of Melibeus, of the Monk, and of the Nun's Priest.

C. The tales of Doctor and Pardoner.

D. The tales of Wife of Bath, Friar, and Summoner.

E. The tales of Clerk and Merchant.

F. The tales of Squire and Franklin.

G. The tales of Second Nun and Canon's Yeoman.

H. The Manciple's tale.

I. The Parson's tale.

In the groups E, F, and G, the first of the two tales which each comprises is spoken of as E1, F1, or G1, the second as E2, F2, or G2.

Our uncertainty as to Chaucer's probable final arrangement of the Tales is caused by the confusion which we find existing in the manuscripts. One large body of texts has B1 followed by F1; another, represented by the noble Ellesmere codex, has D following B1, and so on. We observe, however, that these differences in arrangement are displayed by groups of MSS, and that the codices do not vary irregularly from one another, and in heterogeneous medley of Tales, but rather that they show a systematic confusion by Fragments, or fascicules, as we may also term them. In fact, if we try to imagine how Chaucer worked over the Canterbury Tales, we are compelled to believe that he must have had his papers in much such booklets or fascicules as we now call "fragments." No poet could have used as working copy a single large volume already bound; his changes and additions could not have been carried out. The differences among the existing MSS in arrangement of Tales are, it seems to me, exactly such as would arise if the Canterbury Tales were put into the hands of scribes in booklets, arranged and numbered tentatively by their author, but recombined by copyists in accordance with their own judgment. See discussion under Relative Dates of the Tales, Section III F below.

In somewhat similar manner, the sequence adopted by modern editors is based upon time and place allusions within the Tales

and the Links, without regard to the order seen in the MSS; this arrangement is not represented in any one manuscript, and depends for its validity upon the assumption that all these allusions were correct and final in Chaucer's intention. But in view of the occasional inconsistencies of the Tales, the feminine pronouns of the Shipman's Tale, the Second Nun's terming herself a "sone of Eve”, the Man of Law's announcing that he is about to speak in prose, we can hardly assert that the time and place allusions are above doubt.

Evidence as to the Arrangement of the Tales. In a discussion of the probable order of these fragments, the following kinds of evidence are available:

(1) Allusions to place and time within Tales or Links.

(2) Allusions in Tales or Links to Tales in other fragments. (3) The order as seen in the manuscripts.

I.

[Diagram of the Road to Canterbury.]

"Lo Depeford and it is halfway prime." (between 6 and 9 a. m.) Frag. A, Miller-Reeve link.

2. "Lo Rouchestre stant here fast by." Frag. B2, Monk's prologue.

3.

er I come to Sidingborne." Frag. D, Wife of Bath's prologue. The Summoner is speaking; cp. the last line of his tale, "we ben almost at toune."

4. “At Boughton-under-Blee us gan atake.” Frag. G, introd. to Canon's Yeoman.

Note that there is no place-evidence for the relative position of the C, E, or F fragment; but that such place-evidence as is given argues a sequence of fragments A, B, D, G, H.

2.

Allusions to Time or to Tales in Other Fragments.

Note on the Duration of the Journey. Tyrwhitt, in a note on the opening of the Canon's Yeoman's prologue, line 16023 in his edition, remarked that some difficulties could be avoided by considering that the journey occupied more than one day; but he did no more than suggest. Furnivall, Temp. Pref. as cited below, decided for 31⁄2 days, though at first favoring 2, see his note on p. 59 of Koch's Chronology. Koch there argues for 3; and see his ed. of the Pard. Tale p. xxi. Morley, Eng. Writers V: 310, still speaks of the journey as of one day; but see ibid. p. 312. And Skeat V : 132 seems to admit the one-day theory, though in III: 375-6 and V : 415 he follows Furnivall. Pollard, Chaucer Primer and Globe Chaucer, takes the opinion of Furnivall. Part of the evidence for the longer duration of the pilgrimage comes from the records of medieval journeys, part from the allusions in the Tales and Links themselves. For this latter sort of evidence, see below; for the former, see Temporary Preface 13-15, 41; Stanley's Historical Memorials of Canterbury, p. 237; Archaeologia 35: 461; Notes and Queries 1892 I: 522, reprinted in Koch's Chronology p. 89; Oxford Chaucer V: 415; Academy 1896 II: 14; Flügel in Anglia 23: 239; the Analogues of the Pilgrimage, Ch. Soc. 2d Series No. 36. See also under General Prologue, Section III G below.

It is generally considered that Chaucer intended all the tales for the outward journey; but Tyrwhitt, ibid., said "I have sometimes suspected that it was the intention of Chaucer to begin the journey from Canterbury with the Nonnes tale.” Ten Brink, Hist. Eng. Lit. II: 182, thought that the Manciple's tale was written for the beginning of the return journey.

First Day.

The Tales begin at once after a very early start, Prologue 822. "Lo Depeford and it is halfway prime", Frag. A, MillerReeve link.

Second Day according to Furnivall.

The hour 10 in the morning is given, introd. to the Man of Law's Tale, lines 1-14,-Frag. B1.

"I can right now no thrifty tale seyn",-Man of Law's prol. line 46, compare "This was a thrifty tale for the nones", in the link between Man of Law and Shipman, line 3.

This agreement of the word thrifty furnishes Bradshaw, Furnivall, Skeat, and Shipley (as below), with strong argument for placing B2 after B1, an arrangement still more strongly supported by the fact that B2 contains an allusion to Rochester which seems to place it before D with its reference to Sitting