Several of these words are also written with the ending -e for -ie, as accide, famile, encorde.

Such words are similarly treated in Gower’s English lines, e.g.

cp. Chaucer’s usual treatment of words like victorie, glorie, which are not used in that form by Gower.

(c) In come (comme), sicome, and ove the final e never counts as a syllable in the metre. They are sometimes written com and ou. In another word, ore, the syllable is often slurred, as in Mir. 37, 1775, 3897, &c., but sometimes sounded, as 4737, 11377, Bal. xxviii. 1. So perhaps also dame in Mir. 6733, 13514, 16579, and Bal. ii. 3, xix. 3, xx. 2, &c.

x. The insertion of a parasitic e in connexion with r, and especially between v and r, is a recognized feature of the Anglo-Norman dialect. Examples of this in our texts are avera, devera, saveroit, coverir, deliverer, overir, vivere, livere, oevere, overage, povere, yvere, &c. As a rule this e is not sounded as a syllable in the metre, and in most of these words there is an alternative spelling, e.g. avra, savra, covrir, delivrer, ovrir, vivre, oevre, &c., but it is not necessary to reduce them to this wherever the e is mute. Less usually the syllable counts in the verse, e.g. overaigne in Mir. 3371, overage 8914, enyverer 16448, avera 18532, deveroit, beveroit in 20702 ff. viverai, vivera in Bal. iv.* 1, Mir. 3879, descoverir in Bal. ix. 1.

xi. About the consonants not much need be said.

(a) Initial c before a varies in some words with ch, as caccher, caitif, camele, camp, carboun, castell, catell, by the side of chacer, chaitif, chameal, champ, charboun, chastel, chateaux; cp. acater, achater. Before e, i, we find sometimes an interchange of c and s, as in ce for se in Mir. 1147, Bal. xviii. 3; c’il for s’il in Mir. 799 &c.; and, on the other hand, sent for cent in Bal. xli. 2, si for ci in the title of the Cinkante Balades, sil for cil in Bal. xlii. 3, sercher for cercher in Mir. 712 &c., also s for sc in septre, sintille, and sc for s in scilence.

(b) We find often qant, qe, qelle, qanqe, &c., for quant, que, &c., and, on the other hand, the spelling quar for the more usual car. In words like guaign, guaire, guaite, guarant, guarde, guarir, guaster, u is very frequently omitted before a, also occasionally before other vowels, as gile, 21394, for guile: w is used in warder, rewarder, way.

(c) The doubling of single consonants, especially l, m, n, p, s, is frequent and seems to have no phonetic significance. Especially it is to be observed that ss for s at the end of a word makes no difference to the quantity or quality of the syllable, thus, whether the word be deces or decess, reles or reless, engres or engress, bas or bass, las or lass, huiss or huis, the pronunciation and the rhyme are the same. The final s was sounded in both cases, and not more when double than when single. The doubling of r in futures and conditionals, as serray, dirray, &c., belongs to the Norman dialect.

(d) The final s of inflexion is regularly replaced by z after a dental, as courtz, desfaitz, ditz, excellentz, fitz, fortz, regentz, seintz, and frequently in past participles of verbs (where there is an original dental), as perturbez, enfanteez, rejoïz, perduz; but also elsewhere, especially with the termination -able, as refusablez, delitablez, in rhyme with acceptables. Sometimes however a dental drops out before s, as in apers, desfais, dis, dolens, presens. In all these cases however the difference is one of spelling only.

(e) Lastly, notice may be directed to the mute consonants either surviving in phonetic change or introduced into the spelling in imitation of the Latin form. The fourteenth century was a time when French writers and copyists were especially prone to the vice of etymological spelling, and many forms both in French and English which have been supposed to be of later date may be traced to this period. I shall point out some instances, etymological and other, most of which occur in rhyme.

Thus b is mute in doubte (also doute) rhyming with boute, and also in debte beside dette, soubdeinement beside soudeinement, &c.:

p in temps, accompte, corps, hanaps, descript, rhyming with sens, honte, tors, pas, dit, and in deceipte beside deceite;

d before s in ribalds rhyming with vassals;

t before z in such words as fortz, courtz, certz, overtz, fitz, ditz, aletz, decretz, rhyming with tors, destours, vers, envers, sis, dignités, ées;

s in such forms as dist, promist, quidasmes, &c., in rhyme with esjoït, espirit, dames; possibly however the 3 pers. sing. pret. of these verbs had an alternative pronunciation in which s was sounded, for they several times occur in rhyme with Crist, and then are always written -ist, whereas at other times they vary this freely with -it.

g in words like baraign, pleigne, soveraigne, rhyming with gain, peine;

c before s in clercs (also clers) rhyming with vers;

l in almes, ascoulte, moult, which rhyme with fames, route, trestout and in oultrage, estoultie, beside outrage, estoutie.

On the other hand v is sounded in the occasional form escrivre, the word being rhymed with vivre, in Mir. 6480.

As regards the Vocabulary, I propose to note a few points which are of interest with reference chiefly to English Etymology, and for the rest the reader is referred to the Glossary.

A certain number of words will be found, in addition to those already cited in the remarks on Phonology, § v, which appear in the French of our texts precisely as they stand in modern English, e.g. able, annoy, archer, carpenter, claret, courser, dean, draper, ease, fee, haste, host, mace, mess, noise, soldier, suet, treacle, truant, &c., not to mention ‘mots savants’ such as abject, absent, official, parable, and so on.

The doubling of consonants in accordance with Latin spelling in accepter, accord, accuser, commander, commun, &c., is already common in these texts and belongs to an earlier stage of Middle English than is usually supposed.

ambicioun: note the etymological meaning of this word in the Mirour.

appetiter: Chaucer’s verb should be referred directly to this French verb, and not to the English subst. appetit.

assalt: usually assaut in 14th cent. French and English.

audit: the English word is probably from this French form, and not directly from Latin: the same remark applies to several other words, as complet, concluder, curet, destitut, elat, &c.

avouer: in the sense of ‘promise.’

begant, beggerie, beguyner, beguinage: see New Eng. Dict. under ‘beg.’ The use of beguinage here as equivalent to beggerie is confirmatory of the Romance etymology suggested for the word: begant seems to presuppose a verb beg(u)er, a shorter form of beguiner; cp. beguard.

braier, M.E. brayen, ‘to bray in a mortar.’ The continental form was breier, Mod. broyer.

brusch: the occurrence of this word in a sense which seems to identify it with brusque should be noted. The modern brusque is commonly said to have been introduced into French from Italy in the 16th century. Caxton however in 1481 has brussly, apparently equivalent to ‘brusquely’; see New Eng. Dict.

buillon, in the sense of ‘mint,’ or ‘melting-house,’ is evidently the same as ‘bullion’ in the Anglo-Norman statutes of Edward III (see New Eng. Dict.). The form which we have here points very clearly to its derivation from the verb builer, ‘boil,’ as against the supposed connexion with ‘bulla.’

chitoun, ‘kitten.’ This is used also in Bozon’s Contes Moralizés. It seems more likely that the M.E. kitoun comes from this form of chatton with hardening of ch to k by the influence of cat, than that it is an English ‘kit’ with a French suffix.

Civile, i.e. ‘civil law’: cp. the use of the word as a name in Piers Plowman.

eneauer, ‘to wet,’ supplies perhaps an etymology for the word enewing or ennuyng used by Lydgate and others as a term of painting, to indicate the laying on or gradation of tints in water-colour, and illustrates the later Anglo-French words enewer, enewage, used apparently of shrinking cloth by wetting; see Godefroy (who however leaves them unexplained).

flaket, the same as the M.E. flakett, flacket (French flaschet). The form flaquet is assumed as a Northern French word by the New Eng. Dict., but not cited as occurring.

leisour, as a variation of loisir, leisir.

lusard: cp. Piers Plowman, B. xviii. 335.

menal, meynal, adj. in the sense of ‘subject.’

nice: note the development of sense from ‘foolish,’ Mir. 1331, 7673, to ‘foolishly scrupulous,’ 24858, and thence to ‘delicate,’ ‘pleasant,’ 264, 979.

papir, the same form that we find in the English of Chaucer and Gower.

parlesie, M.E. parlesie, palesie.

perjurie, a variation of perjure, which established itself in English.

phesant: early M.E. fesaun, Chaucer fesaunt.

philosophre, as in M.E., beside philosophe.

queinte, a(c)queintance: the forms which correspond to those used in English; less usually quointe, aquointance.

reverie, ‘revelry,’ which suggests the connexion of the English word with rêver, rather than with reveler from ‘rebellare.’ However, revel and reveller occur also in our texts.

reviler. Skeat, Etym. Dict., says ‘there is no word reviler or viler in French.’ Both are used in the Mirour.

rewarder, rewardie, rewardise, in the sense of the English ‘reward.’

sercher, Eng. ‘search,’ the more usual form for cercher.

somonce: this is the form required to account for the M.E. somouns, ‘summons.’

traicier, traiçour, names given (in England) to those who made it their business to pack juries.

trote, used for ‘old woman’ in an uncomplimentary sense.

université, ‘community.’

voiage (not viage): this form is therefore of the 14th century.

MIROUR DE L’OMME.

Authorship.—The evidence of authorship rests on two distinct grounds: first, its correspondence in title and contents with the description given by Gower of his principal French work; and secondly, its remarkable resemblance in style and substance to the poet’s acknowledged works.

We return therefore to the statement before referred to about the three principal books claimed by our author: and first an explanation should be made on the subject of the title. The statement in question underwent progressive revision at the hands of the author and appears in three forms, the succession of which is marked by the fact that they are connected with three successive editions of the Confessio Amantis. In the two first of these three forms the title of the French work is Speculum Hominis, in the third it is Speculum Meditantis, the alteration having been made apparently in order to produce similarity of termination with the titles of the two other books[H]. We are justified therefore in assuming that the original title was Speculum Hominis, or its French equivalent, Mirour de l’omme. The author’s account, then, of his French work is as follows:

‘Primus liber Gallico sermone editus in decem diuiditur partes, et tractans de viciis et virtutibus, necnon et de variis huius seculi gradibus, viam qua peccator transgressus ad sui creatoris agnicionem redire debet recto tramite docere conatur. Titulus (que) libelli istius Speculum hominis (al. meditantis) nuncupatus est.’

We are here told that the book is in French, that it is divided into ten parts, that it treats of vices and virtues, and also of the various degrees or classes of people in this world, and finally that it shows how the sinner may return to the knowledge of his Creator.

The division of our Mirour into ten parts might have been a little difficult to make out from the work itself, but it is expressly indicated in the Table of Contents prefixed:

‘Cy apres comence le livre François q’est apellé Mirour de l’omme, le quel se divide en x parties, c’est assavoir’ &c.

The ten parts are then enumerated, six of them being made out of the classification of the different orders of society.

The contents of the Mirour also agree with the author’s description of his Speculum Hominis. After some prefatory matter it treats of vices in ll. 841-9720 of the present text; of virtues ll. 10033-18372; of the various orders of society ll. 18421-26604; of how man’s sin is the cause of the corruption of the world ll. 26605-27360; and finally how the sinner may return to God, or, as the Table of Contents has it, ‘coment l’omme peccheour lessant ses mals se doit reformer a dieu et avoir pardoun par l’eyde de nostre seigneur Jhesu Crist et de sa doulce Miere la Vierge gloriouse,’ l. 27361 to the end. This latter part includes a Life of the Virgin, through whom the sinner is to obtain the grace of God.

The strong presumption (to say no more) which is raised by the agreement of all these circumstances is converted into a certainty when we come to examine the book more closely and to compare it with the other works of Gower. Naturally we are disposed to turn first to his acknowledged French writings, the Cinkante Balades and the Traitié, and to institute a comparison in regard to the language and the forms of words. The agreement here is practically complete, and the Glossary of this edition is arranged especially with a view to exhibit this agreement in the clearest manner. There are differences, no doubt, such as there will always be between different MSS., however correct, but they are very few. Moreover, in the structure of sentences and in many particular phrases there are close correspondences, some of which are pointed out in the Notes. But, while the language test gives quite satisfactory results, so far as it goes, we cannot expect to find a close resemblance in other respects between two literary works so different in form and in motive as the Mirour and the Balades. It is only when we institute a comparison between the Mirour and the two other principal works, in Latin and English respectively, which our author used as vehicles for his serious thoughts, that we realize how impossible it is that the three should not all belong to one author. Gower, in fact, was a man of stereotyped convictions, whose thoughts on human society and on the divine government of the world tended constantly to repeat themselves in but slightly varying forms. What he had said in one language he was apt to repeat in another, as may be seen, even if we leave the Mirour out of sight, by comparison of the Confessio Amantis with the Vox Clamantis. The Mirour runs parallel with the English work in its description of vices, and with the Latin in its treatment of the various orders of society, and apart from the many resemblances in detail, it is worth while here to call attention to the manner in which the general arrangement of the French work corresponds with that which we find in the other two books.

In that part of the Mirour which treats of vices, each deadly sin is dealt with regularly under five principal heads, or, as the author expresses it, has five daughters. Now this fivefold division is not, so far as I can discover, borrowed from any former writer. It is of course quite usual in moral treatises to deal with the deadly sins by way of subdivision, but usually the number of subdivisions is irregular, and I have not found any authority for the systematic division of each into five. The only work, so far as I know, which shares this characteristic with the Mirour is the Confessio Amantis. It is true that in this the rule is not fully carried out; the nature of the work did not lend itself so easily to a quite regular treatment, and considerable variations occur: but the principle which stands as the basis of the arrangement is clearly visible, and it is the same which we find in our Mirour.

This is a point which it is worth while to exhibit a little more at large, and here the divisions of the first three deadly sins are set forth in parallel columns:

In the latter part of the Confessio Amantis the fivefold division is not strictly observed, and in some books the author does not profess to deal with all the branches; but in what is given above there is quite enough to show that this method of division was recognized and that the main headings are the same in the two works.

Next we may compare the classes of society given in the Mirour with those that we find in the Vox Clamantis. It is not necessary to exhibit these in a tabular form; it is enough to say that with some trifling differences of arrangement the enumeration is the same. In the Vox Clamantis the estate of kings stands last, because the author wished to conclude with a lecture addressed personally to Richard II; and the merchants, artificers and labourers come before the judges, lawyers, sheriffs, &c., because it is intended to bring these last into connexion with the king; but otherwise there is little or no difference even in the smallest details. The contents of the ‘third part’ of the Mirour, dealing with prelates and dignitaries of the Church and with the parish clergy, correspond to those of the third book of the Vox Clamantis; the fourth part, which treats of those under religious rule, Possessioners and Mendicants, is parallel to the fourth book of the Latin work. In the Mirour as in the Vox Clamantis we have the division of the city population into Merchants, Artificers and Victuallers, and of the ministers of the law into Judges, Advocates, Viscounts (sheriffs), Bailiffs, and Jurymen. Moreover what is said of the various classes is in substance usually the same, most notably so in the case of the parish priests and the tradesmen of the town; but parallels of this kind will be most conveniently pointed out in the Notes.

To proceed, the Mirour will be found to contain a certain number of stories, and of those that we find there by much the greater number reappear in the Confessio Amantis with a similar application. We have the story of the envious man who desired to lose one eye in order that his comrade might be deprived of two (l. 3234), of Socrates and his scolding wife (4168), of the robbery from the statue of Apollo (7093), of Lazarus and Dives (7972), of Ulysses and the Sirens (10909), of the emperor Valentinian (17089), of Sara the daughter of Raguel (17417), of Phirinus, the young man who defaced his beauty in order that he might not be a temptation to women (18301), of Codrus king of Athens (19981), of Nebuchadnezzar’s pride and punishment (21979), of the king and his chamberlains (22765). All these are found in the Mirour, and afterwards, more fully related as a rule, in the Confessio Amantis. Only one or two, the stories of St. Macaire and the devil (12565, 20905), of the very undeserving person who was relieved by St. Nicholas (15757), of the dishonest man who built a church (15553), together with various Bible stories rather alluded to than related, and the long Life of the Virgin at the end of the book, remain the property of the Mirour alone.

If we take next the anecdotes and emblems of Natural History, we shall find them nearly all again in either the Latin or the English work. To illustrate the vice of Detraction we have the ‘escarbud,’ the ‘scharnebud,’ of the Confessio Amantis, which takes no delight in the flowery fields or in the May sunshine, but only seeks out vile ordure and filth (2894, Conf. Am. ii. 413). Envy is compared to the nettle which grows about the roses and destroys them by its burning (3721, Conf. Am. ii. 401). Homicide is made more odious by the story of the bird with a man’s features, which repents so bitterly of slaying the creature that resembles it (5029, Conf. Am. iii. 2599); and we may note also that in both books this authentic anecdote is ascribed to Solinus, who after all is not the real authority for it. Idleness is like the cat that would eat fish without wetting her paws (5395, Conf. Am. iv. 1108). The covetous man is like the pike that swallows down the little fishes (6253, Conf. Am. v. 2015). Prudence is the serpent which refuses to hear the voice of the charmer, and while he presses one ear to the ground, stops the other with his tail (15253, Conf. Am. i. 463). And so on.

Then again there are a good many quotations common to the Mirour and one or both of the other books, adduced in the same connexion and sometimes grouped together in the same order. The passage from Gregory’s Homilies about man as a microcosm, partaking of the nature of every creature in the universe, which we find in the Prologue of the Confessio and also in the Vox Clamantis, appears at l. 26869 of the Mirour; that about Peter presenting Judea in the Day of Judgement, Andrew Achaia, and so on, while our bishops come empty-handed, is also given in all three (Mir. 20065, Vox. Cl. iii. 903, Conf. Am. v. 1900). To illustrate the virtue of Pity the same quotations occur both in the Mirour and the Confessio Amantis, from the Epistle of St. James, from Constantine, and from Cassiodorus (Mir. 13929, 23055 ff., Conf Am. vii. 3149*, 3161*, 3137). Three quotations referred to ‘Orace’ occur in the Mirour, and of these three two reappear in the Confessio with the same author’s name (Mir. 3801, 10948, 23370, Conf. Am. vi. 1513, vii. 3581). Now of these two, one, as it happens, is from Ovid and the other from Juvenal; so that not only the quotations but also the false references are repeated. These are not by any means all the examples of common quotations, but they will perhaps suffice.

Again, if we are not to accept the theory of common authorship, we can hardly account for the resemblance, and something more than resemblance, in passages such as the description of Envy (Mir. 3805 ff., Conf. Am. ii. 3095, 3122 ff.), of Ingratitude (Mir. 6685 ff., Conf. Am. v. 4917 ff.), of the effects of intoxication (Mir. 8138, 8246, Conf. Am. vi. 19, 71), of the flock made to wander among the briars (Mir. 20161 ff., Conf. Am. Prol. 407 ff.), of the vainglorious knight (Mir. 23893 ff., Conf. Am. iv. 1627 ff.), and many others, not to mention those lines which occur here and there in the Confessio exactly reproduced from the Mirour, such as iv. 893,

compared with Mir. 5436,

Conf. Am. Prol. 213,

compared with Mir. 18675,

the context in this last case being also the same.

The parallels with the Vox Clamantis are not less numerous and striking, and as many of them as it seems necessary to mention are set down in the Notes to the Mirour, especially in the latter part from l. 18421 onwards.

Before dismissing the comparison with the Confessio Amantis, we may call attention to two further points of likeness. First, though the Mirour is written in stanzas and the Confessio in couplets, yet the versification of the one distinctly suggests that of the other. Both are in the same octosyllabic line, with the same rather monotonous regularity of metre, and the stanza of the Mirour, containing, as it does, no less than four pairs of lines which can be read as couplets so far as the rhyme is concerned, often produces much the same effect as the simple couplet. Secondly, in the structure of sentences there are certain definite characteristics which produce themselves equally in the French and the English work.

Resemblances of this latter kind will be pointed out in the Notes, but a few may be set down here. For example, every reader of Gower’s English is familiar with his trick of setting the conjunctions ‘and,’ ‘but,’ &c., in the middle instead of at the beginning of the clause, as in Conf. Am. Prol. 155,

and similarly in the Balades, e.g. xx. i,

Examples of this are common in the Mirour, as l. 100,

cp. 415, 4523, 7739, 7860, &c.

In other cases too there is a tendency to disarrangement of words or clauses for the sake of metre or rhyme, as Mir. 15941, 17996, compared with Conf. Am. ii. 2642, iv. 3520, v. 6807, &c.

Again, the author of the Confessio Amantis is fond of repeating the same form of expression in successive lines, e.g. Prol. 96 ff.,

Cp. Prol. 937, v. 2469, &c.

This also is found often in the Mirour, e.g. 4864-9:

and 8294-8304,

The habit of breaking off the sentence and resuming it in a different form appears markedly in both the French and the English, as Mir. 89, 17743, Conf. Am. iv. 2226, 3201; and in several passages obscure forms of expression in the Confessio Amantis are elucidated by parallel constructions in the Mirour.

Finally, the trick of filling up lines with such tags as en son degré, de sa partie, &c. (e.g. Mir. 373, 865), vividly recalls the similar use of ‘in his degree,’ ‘for his partie,’ by the author of the Confessio Amantis (e.g. Prol. 123, 930).

The evidence of which I have given an outline, which may be filled up by those who care to look out the references set down above and in the Notes, amounts, I believe, to complete demonstration that this French book called Mirour de l’omme is identical with the Speculum Hominis (or Speculum Meditantis) which has been long supposed to be lost; and, that being so, I consider myself at liberty to use it in every way as Gower’s admitted work, together with the other books of which he claims the authorship, for the illustration both of his life and his literary characteristics.

Date.—The Speculum Hominis stands first in order of the three books enumerated by Gower, and was written therefore before the Vox Clamantis. This last was evidently composed shortly after the rising of the peasants in 1381, and to that event, which evidently produced the strongest impression on the author’s mind, there is no reference in this book. There are indeed warnings of the danger of popular insurrection, as 24104 ff., 26485 ff., 27229 ff., but they are of a general character, suggested perhaps partly by the Jacquerie in France and partly by the local disturbances caused by discontented labourers in England, and convey the idea that the writer was uneasy about the future, but not that a catastrophe had already come. In one passage he utters a rather striking prophecy of the evil to be feared, speaking of the strange lethargy in which the lords of the land are sunk, so that they take no note of the growing madness of the commons. On the whole we may conclude without hesitation that the book was completed before the summer of the year 1381.

There are some other considerations which will probably lead us to throw the date back a little further than this. In 2142 ff. it seems to be implied that Edward III is still alive. ‘They of France,’ he says, ‘should know that God abhors their disobedience, in that they, contrary to their allegiance, refuse by way of war to render homage and obedience to him who by his birth receives the right from his mother.’ This can apply to none but Edward III, and we are led to suppose that when these lines were written he was still alive to claim his right. The supposition is confirmed by the manner in which the author speaks of the reigning king in that part of his work which deals with royalty. Nowhere does he address him as a child or youth in the manner of the Vox Clamantis, but he complains of the trust placed by the king in flatterers and of the all-prevailing influence of women, calling upon God to remedy those evils which arise from the monstrous fact that a woman reigns in the land and the king is subject to her (22807 ff.). This is precisely the complaint which might have been expected in the latter years of Edward III. On the other hand there is a clear allusion in one place (18817-18840) to the schism of the Church, and this passage therefore must have been written as late as 1378, but, occurring as it does at the conclusion of the author’s attack upon the Court of Rome, it may well have been added after the rest. The expression in l. 22191,

refers to the Pope and the Emperor, not to the division of the papacy. Finally, it should be observed that the introduction of the name Innocent, l. 18783, is not to be taken to mean that Innocent VI, who died in 1362, was the reigning pope. The name is no doubt only a representative one.

On the whole we shall not be far wrong if we assign the composition of the book to the years 1376-1379.

Form and Versification.—The poem (if it may be called so) is written in twelve-line stanzas of the common octosyllabic verse, rhyming aab aab bba bba, so that there are two sets of rhymes only in each stanza. In its present state it has 28,603 lines, there being lost four leaves at the beginning, which probably contained forty-seven stanzas, that is 564 lines, seven leaves, containing in all 1342 lines, in other places throughout the volume, and an uncertain number at the end, probably containing not more than a few hundred lines. The whole work therefore consisted of about 31,000 lines, a somewhat formidable total.

The twelve-line stanza employed by Gower is one which was in pretty common use among French writers of the ‘moral’ class. It is that in which the celebrated Vers de la Mort were composed by Hélinand de Froidmont in the twelfth century, a poem from which our author quotes. Possibly it was the use of it by this writer that brought it into vogue, for his poem had a great popularity, striking as it did a note which was thoroughly congenial to the spirit of the age[I]. In any case we find the stanza used also by the ‘Reclus de Moiliens,’ by Rutebeuf in several pieces, e. g. La Complainte de Constantinoble and Les Ordres de Paris, and often by other poets of the moral school. Especially it seems to have been affected in those ‘Congiés’ in which poets took leave of the world and of their friends, as the Congiés Adan d’Arras (Barb. et Méon, Fabl. i. 106), the Congié Jehan Bodel (i. 135), &c. As to the structure of the stanza, at least in the hands of our author, there is not much to be said. The pauses in sense very generally follow the rhyme divisions of the stanza, which has a natural tendency to fall into two equal parts, and the last three lines, or in some cases the last two, frequently contain a moral tag or a summing up of the general drift of the stanza.

The verse is strictly syllabic. We have nothing here of that accent-metre which the later Anglo-Norman writers sometimes adopted after English models, constructing their octosyllable in two halves with a distinct break between them, each half-verse having two accents but an uncertain number of syllables. This appears to have been the idea of the metre in the mind of such writers as Fantosme and William of Waddington. Here however all is as regular in that respect as can be desired. Indeed the fact that in all these thousands of lines there are not more than about a score which even suggest the idea of metrical incorrectness, after due allowance for the admitted licences of which we have taken note, is a striking testimony not only of the accuracy in this respect of the author, but also to the correctness of the copy which we possess of his work. The following are the lines in question:

This, it will be allowed, is a sufficiently moderate total to be placed to the joint account of author and scribe in a matter of more than 28,000 lines—on an average one in about 1,500 lines. Of these more than half can be corrected in very obvious ways: in 276, 397, we may read ‘grantment’ as in 8931; in 2955, 4832, we should read ‘deliverer,’ and in 9786 ‘metteroit,’ this e being frequently sounded in the metre, e.g. 3371, 16448, 18532; we may correct 3160, 9617, by altering to ‘mal,’ ‘autre’; in 4745 ‘plussoudeinement’ is certainly meant; 13503 is to be corrected by reading ‘en la fin,’ as in 15299, for ‘en fin,’ 19108 by substituting ‘avoltre’ for ‘avoltire,’ and 27598 by reading ‘angel,’ as in 27731 and elsewhere, for ‘angle.’ Of the irregularities that remain, one, exemplified in 3116 and 14568, consists in the introduction of an additional foot into the measure, and I have little doubt that it proceeds from the scribe, who wrote ‘predicacioun’ and ‘contemplacioun’ for some shorter word with the same meaning, such as ‘prechement’ and ‘contempler.’ In the latter of these cases I have corrected by introducing ‘contempler’ into the text; in the former, as I cannot be so sure of the word intended, the MS. reading is allowed to stand. There is a similar instance of a hypermetrical line in Bal. xxvii. 1, and this also might easily be corrected. The other irregularities I attribute to the author. These consist, first, in the use of ‘dame’ in several lines as a monosyllable, and I am disposed to think that this word was sometimes so pronounced, see Phonol. § ix (c); secondly, in the introduction of a superfluous unaccented syllable at a pause after the second foot, which occurs in 10623, 10628 (and perhaps 3160); thirdly, in the omission of the unaccented syllable at the beginning of the verse, as:

Considering how often lines of this kind occur in other Anglo-Norman verse, and how frequent the variation is generally in the English octosyllables of the period, we may believe that even Gower, notwithstanding his metrical strictness, occasionally introduced it into his verse. It may be noted that the three lines just quoted resemble one another in having each a pause after the first word.

With all this ‘correctness,’ however, the verses of the Mirour have an unmistakably English rhythm and may easily be distinguished from French verse of the Continent and from that of the earlier Anglo-Norman writers. One of the reasons for this is that the verse is in a certain sense accentual as well as syllabic, the writer imposing upon himself generally the rule of the alternate beat of accents and seldom allowing absolutely weak syllables[J] to stand in the even places of his verse. Lines such as these of Chrétien de Troyes,