With regard to the weak verbs which form preterites with ending -ede, the loss of the final e is somewhat more common, but it is usually retained, and sometimes it counts as a syllable in the verse. Where this is not the case, it is either elided in the usual way, or if it be dropped in writing, this is only under the conditions which apply to the verbs mentioned above, namely, before a vowel at the beginning of the succeeding word.

It is, however, noteworthy that the use of these forms, whether in -ede or -ed, is decidedly rare, and was avoided by our author even in cases where the -e would have been subject to elision. It is evident that he was always conscious of this ending, even if he did not always write it, and yet he felt that the two weak syllables ought not to have full value in the metre. The result was that he avoided the use of the form generally, so far as it was reasonably possible to do so. The whole number of these preterites in -ede, -ed to be found in the Confessio Amantis is surprisingly small, both actually and relatively, that is, taking account of the extent to which the verbs in question are employed in their other tenses. The method pursued is chiefly to substitute in narrative the present tense, or the perfect formed with ‘hath,’ for the 3rd person singular of the preterite, ‘Conforteth’ for ‘Confortede,’ ‘Hath axed’ for ‘axede,’ ‘feigneth’ for ‘feignede,’ and this apparently as a matter of habit and even in cases where a vowel follows. No doubt the use of the present tense in narrative is quite usual apart from this, but the extremely frequent combination of strong or syncopated preterites with the present tenses of verbs of this class seems to me to indicate clearly how the matter stood.

These examples will serve to illustrate a tendency which every reader will observe, when once his attention has been called to it. There are indeed many narrative passages in which nearly all the strong or syncopated verbs are used in the preterite, and all the others in the present, and it is evident that this cannot be accidentalZ.

There are, however, a certain number of instances of the use of weak preterites, indicative or subjunctive, and a few in which the final e (or -en) is sounded in the metre.

This list of examples, which is fairly complete up to v. 1970, will sufficiently show the manner in which -ede preterites are used. In more than three-fourths of the instances quoted the -e is subject to elision, and of those that remain nine are examples of the plural with -eden termination, and three only of the ending -ede, viz. ii. 2063, ‘Thogh that he lovede ten or tuelve,’ ii. 2453, ‘Supplantede the worthi knyht,’ and v. 1678, ‘He percede the harde roche,’ of which the first is really a case of syncope, ‘lov’de,’ as also ii. 502 (cp. vi. 882) and iv. 1593, whereas in ii. 2027 ‘lovede’ occurs unsyncopated but with -e elided. It will be noted that in the plural the form -eden is used regularly when the syllables are to be fully pronounced, though -ede can be used for the sake of elision.

The -ed form of preterite is less frequent than the other, and I am not aware of any clear example of its employment before a consonant or in rhyme. We have, for example, ‘And used it,’ i. 342, ‘Sche cleped him,’ i. 1535 (‘humbled him,’ i. 2065, is probably a participle, ‘to have humbled himself’), ‘pryded I me,’ i. 2372, ‘ne feigned I,’ ii. 2061, ‘the goddes ... Comanded him,’ iii. 2140 f., ‘Thei cleped him,’ v. 876, cp. 1057, &c. In iii. 1759, ‘The Gregois torned fro the siege,’ we have most probably a participle, ‘were torned.’ We may observe that the -ed form stands also in the plural.

Among weak preterites from originally strong verbs we may notice abreide, crepte (but past participle crope), foghte, fledde, schotte, slepte (also slep, with past participle slepe), smette (beside smot), wepte. The pret. satte in vii. 2282, ‘He satte him thanne doun,’ seems to arise from confusion of sat and sette.

Imperative. The Confessio Amantis is peculiarly rich in imperatives. Beside the regular imperative singular forms, e.g. ared, besech, behold, ches, com, do, forsak, griet, help, hier, hyd, kep, lef, ly, lei, lest, lep, prei, put, say, schrif, spek, tak, tell, thenk, understond, ȝif, &c., the MSS. give us also hyde, iii. 1502, seie, vii. 4084, speke, vii. 5422, take, iv. 2674, v. 6429, thenke, iii. 1083, but not in such positions as to affect the metre. The forms axe, herkne, loke, wite are regular, but lok also occurs (i. 1703, v. 1220).

In some instances the short form of imperative seems to be used as 3rd pers., e.g. ‘hold clos the ston,’ v. 3573, for ‘let him hold,’ ‘tak in his minde,’ viii. 1128, for ‘let him take,’ cp. viii. 1420. The singular and plural forms are often used without distinction, as v. 2333 ff., ‘Ches ... and witeth ... ches and tak ... goth ... taketh,’ v. 3986, ‘So help me nou, I you beseche,’ with ‘Helpeth,’ just above, several persons being addressed, and so ‘taketh hiede And kep conseil,’ viii. 1509 f., to one person. In the interchange of speech between the Confessor and the Lover, while sometimes the distinction is preserved, the Confessor saying tak, tell, understond, and the Lover telleth, axeth (e.g. i. 1395, 1875), at other times the Lover says lest, say, tell, lef, &c. (i. 1942, 1972, ii. 2074, iii. 841, &c.)AA.

Present Participle. The form of the present participle is the most characteristic part of Gower’s verb inflexion as compared (for example) with Chaucer’s. Chaucer seems regularly to have used the form in -inge (often with apocope -ing): Gower uses ordinarily the form -ende, and normally with the accent thrown on the termination, as i. 204, ‘To me spekende thus began,’ 236, ‘Whos Prest I am touchende of love,’ 428, ‘Stondende as Stones hiere and there,’ 633, ‘So that semende of liht thei werke,’ 1379 f., ‘That for I se no sped comende, ... compleignende,’ 1682, ‘Hangende doun unto the chin.’

Sometimes the same form is used with accent on the preceding syllable, and in this case the -e is systematically elided, e.g. Prol. 11, ‘In tyme comende after this,’ 259, ‘Belongende unto the presthode,’ i. 296, ‘As touchende of my wittes fyve’ (cp. 334, 742), 3025, ‘And wailende in his bestly stevene.’

In a relatively small number of instances the form -inge occurs either in rhyme, as i. 524, ‘So whan thei comen forth seilinge,’ in rhyme with ‘singe,’ i. 1710, ‘And liveth, as who seith, deyinge,’ in rhyme with ‘likynge’ (subst.), or with the accent thrown back, as i. 115, ‘Wisshinge and wepinge al myn one,’ v. 518, ‘Abidinge in hir compaignie,’ vi. 717, ‘I mai go fastinge everemo’; rarely out of rhyme and with accent, as i. 2721, ‘Mi fader, as touchinge of al.’

The final e is never lost in writing, but when the accent is thrown back it is always elided.

Past Participle. The -id termination of weak past participles is very rarely found in the Fairfax MS., except in the concluding passage, which is copied in a different hand from the rest. It occurs commonly in the Praise of Peace. Examples found elsewhere in F are weddid, iv. 650, medlid, iv. 1475.

From setten besides the regular past participle set there appears the form sete twice in rhyme, vii. 2864, forȝete: sete, and viii. 244, misgete (past partic.): upsete. This seems to be formed after the analogy of gete. On the other hand we have ferd, i. 445, &c., but also fare(n), iii. 2692, v. 3797, &c. The past participle of se is sen, sein, seie, but most commonly sene. In a few instances a final e is given by the MSS. in weak past participles, e.g. herde for herd, v. 4231, schope for schop, v. 4278, sette for set, vi. 10, wiste for wist, viii. 37.

The cases of weak past participles with plural inflexion (e.g. Prol. 300, i. 3246, iv. 2343, v. 6789) have already been mentioned in dealing with adjectives.

There is hardly any use of the prefix y- (i-), but we have ybore, ii. 499.

vi. Dialect. Gower’s language is undoubtedly in the main the English of the Court, and not a provincial dialect. Making allowance for the influences of literary culture and for a rather marked conservatism in orthography and grammatical inflexions, we can see that it agrees on the whole with the London speech of the time, as evidenced by the contemporary documents referred to by Prof. Morsbach. At the same time its tendencies are Southern rather than Midland, and he seems to have used Kentish forms rather more freely than Chaucer. This is shown especially (1) in the more extensive use of the forms in which e stands for O. E. y̆, as senne, kesse, pet, hell, &c.; (2) in the frequent employment of ie, both in French and English words, to represent ẹ̄, a practice which can hardly be without connexion with the Kentish cliene, diepe, diere, hier, hield, niede, &c.; (3) in the use of -ende as the normal termination of the present participle. (The Ayenbite regularly has -inde.) Probably also the preference shown by Gower for the close sound of ē, from O. E. ǣ, may be to some extent due to Kentish influence. Other points of resemblance between the language of Gower and that of the Ayenbite (for example) are the free use of syncopated forms in the 3rd pers. sing. of verbs and the regular employment of ous for us.

vii. Metre, &c. The smoothness and regularity of Gower’s metre has been to some extent recognized. Dr. Schipper in Englische Metrik, vol. i. p. 279, remarks upon the skill with which the writer, while preserving the syllabic rule, makes his verse flow always so smoothly without doing violence to the natural accentuation of the words, and giving throughout the effect of an accent verse, not one which is formed by counting syllables. Judging by the extracts printed in Morris and Skeat’s Specimens (which are taken from MS. Harl. 3869, and therefore give practically the text of Fairfax 3), he observes that the five principal licences which he has noted generally in the English verse of the period are almost entirely absent from Gower’s octosyllabics, and in particular that he neither omits the first unaccented syllable, as Chaucer so often does (e.g. ‘Be it rouned, red or songe,’ Hous of Fame, ii. 214, ‘Any lettres for to rede,’ iii. 51, ‘Of this hill that northward lay,’ iii. 62), nor displaces the natural accent (as ‘Of Decembre the tenthe day,’ Hous of Fame, i. 111, ‘Jupiter considereth wel this,’ ii. 134, ‘Rounede everych in otheres ere,’ iii. 954), nor slurs over syllables.

To say that Gower never indulges in any of these licences would be an exaggeration. Some displacement of the natural accent may be found occasionally, even apart from the case of those French words whose accent was unsettled, but it is present in a very slight degree, and the rhythm produced does not at all resemble that of the lines cited above from Chaucer: e.g. i. 2296, ‘Wher that he wolde make his chace,’ 2348, ‘Under the grene thei begrave,’ 2551, ‘“Drink with thi fader, Dame,” he seide.’ Such as it is, this licence is nearly confined to the first foot of the verse, and is not so much a displacement of the natural accent of the words as a trochaic commencement, after the fashion which has established itself as an admitted variety in the English iambic. We may, however, read long passages of the Confessio Amantis without finding any line in which the accent is displaced even to this extent.

Again, as to slurring of syllables, this no doubt takes place, but on regular principles and with certain words or combinations only. There are hardly more than three or four lines in the whole of the Confessio Amantis where a superfluous syllable stands unaccounted for in the body of the verse, as for example,

The writer seems to have no need of any licences. The narrative flows on in natural language, and in sentences and periods which are apparently not much affected by the exigencies of metre or rhyme, and yet the verse is always smooth and the rhyme never fails to be correct. If this is not evidence of the highest style of art, it shows at least very considerable skill.

In Gower’s five-accent line, as exhibited in the Supplication of viii. 2217-2300 and in the poem In Praise of Peace, Schipper finds less smoothness of metre, ‘owing perhaps to the greater unfamiliarity and difficulty of the stanza and verse’ (Englische Metrik, i. 483 ff.). His examples, however, are not conclusive on this point. Some of the lines cited owe their irregularity to corruptions of text, and others prove to be quite regularly in accordance with Gower’s usual metrical principles.

For instance, in viii. 2220 the true text is

which is a metrically perfect line. Again, in the Praise of Peace, l. 79,

it is impossible, according to Gower’s usage, that ‘heven’ should stand as a dissyllable. He wrote always ‘hevene,’ and the penultimate was syncopated. So also ‘levere’ in l. 340, ‘evere,’ l. 376. Hence there is no ‘epic caesura’ in any of these cases. Nor again in l. 164, ‘Crist is the heved,’ can ‘heved’ be taken as a dissyllable in the verse: it is always metrically equivalent to ‘hed.’ The only fair instance of a superfluous syllable at the caesura is in l. 66,

It seems that the trochee occurs more commonly here than in the short line. Such examples as Schipper quotes, occurring at the beginning of the line,

are of the same character as those which we find in the octosyllabics. Perhaps, however, a difference is afforded by the more frequent occurrence of the same licence in other parts of the verse, as,

The rhyming on words like ‘manhode,’ ‘axinge,’ &c., is in accordance with the poet’s general usage.

On the whole, the combination of the syllabic and the accentual system is effected in the five-accent line of these stanzas almost as completely as in the short couplet; and in his command of the measure, in the variety of his caesura, and the ease with which he passes without pause from line to line and rounds off the stanza with the matter, the author shows himself to be as fully master of his craft upon this ground as in the more familiar measure of the Confessio Amantis.

As regards the treatment of weak syllables in the metre, Gower’s practice, in accordance with the strict syllabic system which he adopted, is very different from Chaucer’s. The rules laid down by ten Brink, Chaucers Sprache, § 260, as to the cases in which weak final e is never counted as a syllable in the verse, except in rhyme, require some qualification even when applied to Chaucer (for example, ‘sone’ is certainly a dissyllable in Cant. Tales, A 1963, Hous of Fame, i. 218), and they are almost wholly inapplicable to Gower, as we shall see if we examine them. (α) Gower has the forms hire, oure, ȝoure, all occasionally as dissyllables apart from special emphasis or rhyme. (β) these, some, whiche are all sometimes dissyllables. (γ) The strong participles with short stems as come, drive, write as a rule have the final e sounded. (δ) The -e of the 2nd pers. sing. of the strong preterite may be sounded, e.g. iii. 2629 (but ‘Were thou,’ iv. 600). (ε) The form made, both singular and plural, regularly has -e sounded, were (pret.) usually, and wite sometimes. (ζ) sone, wone, schipe (dat.), and the French words in -ie (ye), &c., have -e regularly counted in the metre: so also beste, entente, tempeste. (η) before, tofore, there are used in both ways.

Syncope (so far as regards the metre) regularly takes place in the following: covere (discovere, &c.), delivere (but not deliverance, i. 1584, v. 1657), evene, evere, fievere, havene, hevene, levere, nevere, povere, sevene (also sefne), swevene (also swefne), and some other words of a similar kind, to which add heved, evel, devel. In these cases a final e is always pronounced unless elided, and in case of elision a word like hevene, nevere is reduced to a monosyllable, as

The following also are sometimes syncopated: lovede, loveden, ii. 502, vi. 882, but without syncope ii. 2027, beloved, i. 1928, belovëd, i. 1920 f., behovely, behovelich, iii. 1330, v. 4012, vii. 1949 (but unbehovëly, viii. 2884), leveful, v. 7053, Averil, vii. 1029, soverein, vii. 1776 (but usually three syllables, as Prol. 186, i. 1609, and sovereinete, five syllables, i. 1847), amorous, iii. 745 (but usually three syllables, as i. 1414), fader, ii. 2387, cp. fadre, ii. 2519 (but ordinarily a dissyllable), unkendeli, ii. 3124 (but unkindëly, iii. 2065), comelieste, comelihiede, v. 3048, 6734 (but comëly, ii. 441), namely, viii. 3041, also namly, ii. 47 (but usually three syllables, as Prol. 144, iii. 63), Termegis, iv. 2408. We may note, however, that this kind of syncope is less used by Gower than by Chaucer, and that chivalerie, chivalerous, foreward, foretokne, loveday, pilegrin, surquiderie, &c., are fully pronounced.

Elision of weak final e takes place regularly before a vowel or an unaspirated h. We must observe that several classical proper names ending originally in ē, as Alceone, Daphne, Progne, Phebe, have weak e and are subject to elision, and under this head it may be noted that Canace rhymes to place, whereas Chaucer (referring to Gower’s story) gives the name as Canacee, in rhyme with he. Also the combinations byme, tome, tothe, &c., have weak -e and are elided before a vowel.

An aspirated h prevents elision as effectively as any other consonant. We have ‘min holë herte,’ ‘gretë hornes,’ ‘Cadmë hyhte,’ ‘Mi Sonë, herkne,’ ‘proprë hous,’ ‘fastë holde’ (and even ‘othrë herbes,’ iv. 3008); but there are some words in which h is aspirated only when they are emphatic in sense or position, as have, hath, he, him, hire, how, &c. For example, elision takes place usually before have, he, how, but not so as a rule in cases where they are used in rhyme or with special emphasis, e.g. i. 2542, ‘Of such werk as it scholde have,’ ii. 2479, cp. v. 7766, ‘Wenende that it were he,’ iv. 3604, ‘And al the cause hou it wente.’ On the other hand, the preterite hadde seems to have an aspirated h even in unemphatic position, as ii. 589, ‘The Sceptre hadde forto rihte’: compare vii. 2364, ‘Victoire hadde upon his fo,’ with vii. 2392, ‘Thogh thou victoire have nou on honde.’ Elision also takes place before hierafter, though not before hiere.

There is one instance of hiatus, viii. 110, ‘That he his Sone Isaäc,’ and it may be noted that the same thing occurs with the same name in the Mirour, 12241, ‘De Isaak auci je lis.’

There is diaeresis regularly in such proper names as Theseüs, Peleüs, Tereüs, and also in Saül, Isaäc. We have Moïses usually, but Moises (dissyllable), iv. 648, Thaise usually, but Thaïsis in the epitaph, viii. 1536. One example occurs affecting the -ee termination, viz. Caldeë, v. 781 (usually a dissyllable), so Judeë, Galileë in Mirour, 20067, 29239. This is an essentially different case from that of degreës, which is found in Chaucer. The termination -ius is usually dissyllabic, but vii. 2967, ‘The god Mercurius and no man.’ The endings -ioun, -ious, -ien, &c., are always fully pronounced.

As regards accent, it has been already observed that the natural accent of words is preserved far better in Gower’s verse than in Chaucer’s. There are, however, a number of words of French origin, of which the accent was unsettled, and also some instances of English words in which a secondary syllable was capable of receiving the principal accent, either in case of composition, as in kingdom, knihthode, treweliche, or with a formative termination, as that of the superlative, fairéste, &c., or the present participle, as wepénde. In such cases the accent was often determined by the metre. Many Romance words are quite freely treated in the matter of accent, as for example folie, fortune, mercy, mirour, nature, parfit, preiere, resoun, science, sentence, tempeste. The terminations -hode, -hede, -inge, -liche, -ly, -nesse, -schipe are all capable of accent, and also the penultimate syllables of answere and felawe.

Nearly all that is important about rhyme has already been said under the head of Phonology. We may here remark on some of the instances in which the form of words is accommodated to the rhyme, these being sometimes cases where variants are supplied by neighbouring dialects. Thus we have aise for ese, ar for er, hair for heir, naght once for noght, fer once for fyr, hade, with the original long vowel, for hadde, geth (the originally correct form) for goth, fore for for; and alternatives such as moneie monoie, aweie awey away, seide saide, soverein soverain, are used in accordance with the rhyme, though it is difficult to say for certain in all cases whether there was difference of sound. Thus, while we have away as rhyme to day, awey is found rhyming to ey, i. 2545, said, saide rhyming with paid, Maide, while seide rhymes with alleide, obeide; we find soverein : aȝein, but brayn : soverain. The form yhe often varies to ÿe when in rhyme with -ie termination, as clergie : ÿe, Prol. 329 f., ÿe : agonie, i. 967 f. (but also yhe : pourpartie, i. 405 f., yhe : specefie, i. 571 f.). Sometimes however the other rhyme-word is modified to correspond to it, as pryhe : yhe, v. 469 f., and there was probably no perceptible difference of pronunciation in this case. So also the preterite lowh is written low when in rhyme with now, Prol. 1071, and similarly thou : ynou, vii. 2099 f. (but bowe: ynowhe, ii. 3225 f.). We have already seen that the use of such alternative forms as sinne senne, wile wole, lasse lesse, hedde hidde, -ende -inge is sometimes determined by the rhyme.

Alliteration is used by Gower in a manner which is especially characteristic of the new artistic style of poetry. It is sufficiently frequent, both in formal combinations, such as ‘cares colde,’ ‘lusty lif,’ ‘park and plowh,’ ‘swerd or spere,’ ‘lief and loth,’ ‘wel or wo,’ ‘dike and delve,’ ‘slepe softe,’ ‘spille ... spede,’ and as an element of the versification:

But it is not introduced in accordance with any fixed rules, and it often assists the flow of the verse without in the least attracting the attention of the reader. We do not find any examples of the rather exaggerated popular style which Chaucer sometimes adopts in passages of violent action, e.g. Cant. Tales, A 2604 ff. The whole subject of alliteration in Gower has been carefully dealt with by P. Höfer in his dissertation, Alliteration bei Gower, 1890, where a very large number of examples are cited and classified; and to this the reader may be referred.

viii. Text and Manuscripts. About forty manuscript copies of the Confessio Amantis are known to exist in public or private libraries or in the hands of booksellers, and probably there may be a few more in private possession, the existence of which has not yet been recorded. As the broad lines for their classification are necessarily laid down by the fact that the book was put forth by the author in several different forms, it is necessary, before proceeding further, to say something about this matter.

That the poem exists in at least two distinct forms, characterized by obvious differences near the beginning and at the end, has been matter of common knowledge. Even in Berthelette’s edition of 1532 the difference at the beginning was noted, and though the printer did not venture to deviate from the form of text which had been made current by Caxton, yet he gave in his preface the beginning of the poem as he found it in his manuscript. Dr. Pauli accordingly proceeded on the assumption that there were two normal forms, one having a dedication to Richard II at the beginning and a form of conclusion in which mention is made of Chaucer, and the other with a dedication to Henry of Lancaster and a conclusion in which Chaucer is not mentioned. Copies which do not conform to these standards are for him simply irregular. He is aware of the additional passages in Berthelette’s edition and in the Stafford MS., and in one place he speaks of three classes of MSS., but he does not know that there are any written copies except the Stafford MS. which contain the additional passages. If he had had personal knowledge of the manuscripts at Oxford and at Cambridge, instead of being satisfied to gather scraps of information about the former from Bodley’s Librarian and about the latter from Todd, he would have found the passages in question also in MS. Bodley 294 at Oxford and in the Trinity and Sidney MSS. at Cambridge.

There are then at least these three classes of manuscripts to be recognized even by a superficial observer, and we shall find that the more obvious differences which have been mentioned are accompanied by a number of others of less importance. The first recension according to our classification is that in which the conclusion of the poem contains praises of Richard II as a just and beneficent ruler and a presentation of the book for his acceptanceAB. The second has the additional passages of the fifth and seventh books, with a rearrangement of the sixth book which has not hitherto been noticed, while the conclusion of the poem has been rewritten so as to exclude the praises of the king, and in some copies there is also a new preface with dedication to Henry of Lancaster. The third exhibits a return to the form of the first as regards the additional passages, but has the rewritten preface and epilogue. Against this merely threefold division some objections might fairly be made. It might be pointed out that the so-called second recension includes at least two distinct forms, and moreover that upon further examination we see reason to divide the manuscripts of our first recension into two main groups, one exhibiting an earlier and the other a later text, this last being more in accordance generally with that of what we call the second and third recensions than with the earlier form of the first. For practical purposes, however, the division which has been laid down above may fairly be adopted. As regards the order of time, from the political tendency of the differences between them it is clear that what we call the first recension logically precedes the third. The intermediate position of the second is given chiefly by the fact that one of the seven existing manuscripts gives the earlier form of preface, and this may also have been the case with two others, which are defective at the beginningAC. However, as has been said, the name is used for convenience to cover a class of copies which, as regards the character of their text, do not all belong to the same period, and they must be looked upon as representing rather a concurrent variety of the first or the third recensionAD than as a type which is distinctly intermediate in order of time. At the same time the smaller variations of text exhibited by these seven MSS. in combination, as against all othersAE, mark them as really a family apart, more closely related to one another than to those that lie outside the group.

For the sake of clearness the manuscripts are in this edition regularly grouped according to this classification, and in the critical notes each class is cited by itself. At the same time it must not be assumed that the manuscripts of each recension stand necessarily by themselves, and that no connexion is traceable between one class and another. On the contrary, we shall find that many errors in the text of the first recension appear also in some copies of the second, and even of the third. The process by which this was brought about is made clearer to us by the fact that we have an example of a manuscript which has passed from one group into another partly by erasure and partly by substitution of leaves, apparently made under the direction of the author. This is MS. Fairfax 3, which forms the basis of our text, and the handwriting of some of the substituted pages is one which may be recognized as belonging to the ‘scriptorium’ of the poet.

The example is a suggestive one and serves to explain several things. It makes it easy to understand, for example, how the additional matter introduced into the second recension came to be omitted in the third. The author in this instance had before him a very fully revised and corrected copy of his first edition, and this by a certain amount of rewriting over erasure and by a substitution of leaves at the beginning and end of the poem was converted into a copy of what we call the third recension, which his scribes could use at once as an authoritative exemplar. The introduction of the additional passages in the fifth and seventh books could not have been effected without a process of recopying the whole book, which would have called for much additional labour of the nature of proof-reading on the part of the author, in order to secure its correctness. This argument would apply to a book which was intended to remain in the hands of the author, or rather of the scribes whom he employed, and to be used as an archetype from which copies were to be made. If a new book had to be specially prepared for presentation, the case would be different, and it might then be worth while to incorporate the additional passages with the fully revised and re-dedicated text, as we find was done in the case of the so-called Stafford MS.

Another matter which can evidently be explained in the same way is the reappearance in some copies of the second recension of errors which belong to the first. In producing the originals of such manuscripts as these, partially revised copies of the first recension must have been used as the basis, and such errors as had not yet received correction appear in the new edition.

The assumption that a certain number of errors are original, that is to say, go back either to the author’s own autograph or to the transcript first made from it, is in itself probable: we know in fact that some which appear in every copy, without exception, of the first and second recensions at length receive correction by erasure in Fairfax 3. So far as we can judge, the text of the Confessio Amantis during its first years exhibited a steady tendency to rid itself of error, and the process of corruption in the ordinary sense can hardly be said to have set in until after the death of the author. There are a large number of various readings in the case of which we find on the one side the great majority of first recension MSS., and on the other a small number of this same type together with practically the whole of the second and third recensions, as, for exampleAF: