The positive achievements of English literature, during the period with which this volume deals, are not at first sight great; and all the more finished literary production of the time, till the extreme end of it, was in French and Latin. But the work done during this time in getting the English language ready for its future duties, in equipping it with grammar and prosody, in preparing, so to speak, for Chaucer, is not only of the first importance intrinsically, but has a value which is almost unique in general literary history as an example. Nowhere else have we the opportunity of seeing a language and a literature in the process of gestation, or at least of a reformation so great as to be almost equal to new birth. Of the stages which turned Latin through the Romanic vulgar tongues into Spanish, Italian, Portuguese, Provençal, French, we have the very scantiest remains; and though the Strasburg oaths and the Eulalia hymn are no doubt inestimable in their way, they supply exceedingly minute and precarious stepping-stones by which to cross from Ausonius to the Chanson de Roland. From the earliest literary stages of the Teutonic tongues we have, except in the case of Anglo-Saxon and Icelandic, very little wreckage of time; and Anglo-Saxon at least presents the puzzling characteristic that its earliest remains are, cœteris paribus, nearly as complete and developed as the earliest remains of Greek. In German itself, whether High or Low, the change from oldest to youngest is nothing like the change from the English of Beowulf to the English of Browning. And though the same process of primordial change as that which we have seen in English took place certainly in German, and possibly in the Romance tongues, it is nowhere traceable with anything like the same clearness or with such gradual development. By the eleventh century at latest in France, by the end of the twelfth in Germany, verse had taken, in the first case fully, in the second almost fully, a modern form. In England it was, during the two hundred years from 1150 to 1350, working itself steadily, and with ample examples, from pure accent to accentual quantity, and from alliteration to rhyme. Of this process, and those similar to it in other countries, we shall give an account which will serve for the whole in the latter part of this chapter; the actual production and gradual transformation of English language and literature generally may occupy us in the earlier part.

It is to be hoped that by this time a middle way, tolerably free from molestation, may be taken between those historians of English who would have a great gulf fixed before Chaucer, and those who insist upon absolute continuity from Cædmon to Tennyson. There must surely be something between dismissing (as did the best historian of the subject in the last generation) Anglo-Saxon as "that nocturnal portion of our literature," between calling it "impossible to pronounce with certainty whether anything in it is artistically good or bad,"[88] and thinking it proper, as it has sometimes been thought, in an examination in English literature, to give four papers to Cædmon, Ælfric, and Wulfstan, and one to the combined works of Addison, Pope, Johnson, and Burke. Extravagances of the latter kind have still, their heyday of reaction not being quite past, a better chance than extravagances of the former. But both may surely be avoided.

The evidence is rendered more easy in the present connection by the fact, recognised by the most competent authorities in First English or Anglo-Saxon itself, that for some time before the arbitrary line of the Conquest the productive powers of the literature had been failing, and the language itself was showing signs of change. No poetry of the first class seems to have been written in it much after the end of the ninth century, little prose of a very good class after the beginning of the eleventh; and its inflexions must in time have given way—were, it is said by some, actually giving way—before the results of the invasion and assimilation of French and Latin. The Conquest helped; but it did not wholly cause.

This, however, is no doubt open to argument, and the argument would have to be conducted mainly if not wholly on philological considerations, with which we do not here meddle. The indisputable literary facts are that the canon of pure Anglo-Saxon or Old-English literature closes with the end of the Saxon Chronicle in 1154, and that the "Semi-Saxon," the "First Middle English," which then makes its appearance, approximates, almost decade by decade, almost year by year, nearer and nearer to the modern type. And for our purpose, though not for the purpose of a history of English Literature proper, the contemporary French and Latin writing has to be taken side by side with it.

It is not surprising that, although the Latin literary production of the time, especially in history, was at least equal to that of any other European country, and though it is at least probable that some of the greatest achievements of literature, French in language, are English in nationality, the vernacular should for long have been a little scanty and a little undistinguished in its yield. Periods of moulting, of putting on new skins, and the like, are never periods of extreme physical vigour. And besides, this Anglo-Saxon itself had (as has been said) been distinctly on the wane as a literary language for more than a century, while (as has not yet been said) it had never been very fertile in varieties of profane literature. This infertility is not surprising. Except at rare periods literature without literary competition and comparison is impossible; and the Anglo-Saxons had absolutely no modern literature to compare and compete with. If any existed, their own was far ahead of it. On the other hand, though the supposed ignorance of Latin and even Greek in the "dark" ages has long been known to be a figment of ignorance itself, circumstances connected with, though not confined to, the concentration of learning and teaching in the clergy brought about a disproportionate attention to theology. The result was that the completest Anglo-Saxon library of which we can form any well-based conception would have contained about ten cases of religious to one of non-religious books, and would have held in that eleventh but little poetry, and hardly any prose with an object other than information or practical use.

It could not be expected that the slowly changing language should at once change its habits in this respect. And so, as the century immediately before the Conquest had seen little but chronicles and homilies, leechdoms and laws, that which came immediately afterwards gave at first no very different products, except that the laws were wanting, for obvious reasons. Nay, the first, the largest, and almost the sole work of belles lettres during the first three-fourths of our period, the Brut of Layamon, is a work of belles lettres without knowing it, and imagines itself to be a sober history, while its most considerable contemporaries, the Ormulum and the Ancren Riwle, the former in verse, the latter in prose, are both purely religious. At the extreme end of the period the most important and most certain work, Robert of Gloucester's, is, again, a history in verse. About the same time we have, indeed, the romances of Havelok and Horn; but they are, like most of the other work of the time, translations from the French. The interesting Poema Morale, or "Moral Ode," which we have in two forms—one of the meeting-point of the twelfth and thirteenth centuries, one fifty years later—is almost certainly older than its earliest extant version, and was very likely pure Saxon. Only in Nicholas of Guilford's Owl and Nightingale, about 1250, and perhaps some of the charming Specimens of Lyric Poetry, printed more than fifty years ago by Mr Wright, with a very few other things, do we find pure literature—not the literature of education or edification, but the literature of art and form.

Yet the whole is, for the true student of literature, full enough of interest, while the best things are not in need of praising by allowance. Of Layamon mention has already been made in the chapter on the Arthurian Legend. But his work covers very much more than the Arthurian matter, and has interests entirely separate from it. Layamon, as he tells us,[89] derived his information from Bede, Wace, and a certain Albinus who has not been clearly identified. But he must have added a great deal of his own, and if it could be decided exactly how he added it, the most difficult problem of mediæval literature would be solved. Thus in the Arthurian part, just as we find additions in Wace to Geoffrey, so we find additions to Wace in Layamon. Where did he get these additions? Was it from the uncertain "Albinus"? Was it, as Celtic enthusiasts hold, that, living as he did on Severn bank, he was a neighbour of Wales, and gathered Welsh tradition? Or was it from deliberate invention? We cannot tell.

Again, we have two distinct versions of his Brut, the later of which is fifty years or thereabouts younger than the earlier. It may be said that almost all mediæval work is in similar case. But then the great body of mediæval work is anonymous; and even the most scrupulous ages have not been squeamish in taking liberties with the text of Mr Anon. But the author is named in both these versions, and named differently. In the elder he is Layamon the son of Leovenath, in the younger Laweman the son of Leuca; and though Laweman is a mere variant or translation of Layamon, as much can hardly be said of Leovenath and Leuca. Further, the later version, besides the changes of language which were in the circumstances inevitable, omits many passages, besides those in which it is injured or mutilated, and alters proper names entirely at discretion.

The only explanation of this, though it is an explanation which leaves a good deal unexplained, is, of course, that the sense both of historical criticism and of the duty of one writer to another was hardly born. The curiosity of the Middle Ages was great; their literary faculty, though somewhat incult and infantine, was great likewise: and there were such enormous gaps in their positive knowledge that the sharp sense of division between the certain, the uncertain, and the demonstrably false, which has grown up later, could hardly exist. It seems to have been every man's desire to leave each tale a little richer, fuller, handsomer, than he found it: and in doing this he hesitated neither at the accumulation of separate and sometimes incongruous stories, nor at the insertion of bits and scraps from various sources, nor, it would appear, at the addition of what seemed to him possible or desirable, without troubling himself to examine whether there was any ground for considering it actual.

Secondly, Layamon has no small interest of form. The language in which the Brut is written has an exceedingly small admixture of French words; but it has made a step, and a long one, from Anglo-Saxon towards English. The verse is still alliterative, still destitute of any fixed number of syllables or syllabic equivalents. But the alliteration is weak and sometimes not present at all, the lines are of less extreme lawlessness in point of length than their older Saxon representatives, and, above all, there is a creeping in of rhyme. It is feeble, tentative, and obvious, confined to ostentatious pairs like "brother" and "other," "might" and "right," "fare" and "care." But it is a beginning: and we know that it will spread.

In the last comparison, that of matter, Layamon will not come out ill even if he be tried high. The most obvious trial is with the work of Chrestien de Troyes, his earlier, though not much earlier, contemporary. Here the Frenchman has enormous advantages—the advantage of an infinitely more accomplished scheme of language and metre, that of some two centuries of finished poetical work before him, that of an evidently wider knowledge of literature generally, and perhaps that of a more distinctly poetical genius. And yet Layamon can survive the test. He is less, not more, subject to the cliché, the stereotyped and stock poetical form, than Chrestien. If he is far less smooth, he has not the monotony which accompanies and, so to speak, dogs the "skipping octosyllable"; and if he cannot, as Chrestien can, frame a set passage or show-piece, he manages to keep up a diffused interest, and in certain instances—the story of Rouwènne (Rowena), the Tintagel passage, the speech of Walwain to the Emperor of Rome—has a directness and simple appeal which cannot be slighted. We feel that he is at the beginning, while the other in respect of his own division is nearly at the end: that he has future, capabilities, opportunities of development. When one reads Chrestien or another earlier contemporary, Benoît de Sainte-More, the question is, "What can come after this?" When one reads Layamon the happier question is, "What will come after this?"

The Ormulum and the Ancren Riwle appear to be—the former exactly and the latter nearly of the same date as Layamon, all being near to 1200. But though they were "good books," their interest is by no means merely one of edification. That of the Ormulum[90] is, indeed, almost entirely confined to its form and language; but it so happens that this interest is of the kind that touches literature most nearly. Orm or Ormin, who gives us his name, but of whom nothing else is known, has left in ten thousand long lines or twenty thousand short couplets a part only of a vast scheme of paraphrase and homiletic commentary on the Four Gospels (the "four-in-hand of Aminadab," as he calls them, taking up an earlier conceit), on the plan of taking a text for each day from its gospel in the calendar. As we have only thirty-two of these divisions, it is clear that the work, if completed, was much larger than this. Orm addresses it to Walter, his brother in the flesh as well as spiritually: the book seems to be written in an Anglian or East Anglian dialect, and it is at least an odd coincidence that the names Orm and Walter occur together in a Durham MS. But whoever Orm or Ormin was, he did two very remarkable things. In the first place, he broke entirely with alliteration and with any-length lines, composing his poem in a metre which is either a fifteen-syllabled iambic tetrameter catalectic, or else, as the reader pleases, a series of distichs in iambic dimeters, alternately acatalectic and catalectic. He does not rhyme, but his work, in the couplet form which shows it best, exhibits occasionally the alternation of masculine and feminine endings. This latter peculiarity was not to take hold in the language; but the quantified or mainly syllabic arrangement was. It was natural that Ormin, greatly daring, and being almost the first to dare, should neither allow himself the principle of equivalence shortly to distinguish English prosody from the French, which, with Latin, he imitated, nor should further hamper his already difficult task with rhyme. But his innovation was great enough, and his name deserves—little positive poetry as there is in his own book—high rank in the hierarchy of British poets. But for him and others like him that magnificent mixed harmony, which English almost alone of languages possesses, which distinguishes it as much from the rigid syllabic bondage of French as from the loose jangle of merely alliterative and accentual verse, would not have come in, or would have come in later. We might have had Langland, but we should not have had Chaucer: we should have had to console ourselves for the loss of Surrey and Wyatt with ingenious extravagances like Gawain Douglas's Eighth Prologue; and it is even possible that when the reaction did come, as it must have come sooner or later, we might have been bound like the French by the rigid syllable which Orm himself adopted, but which in those early days only served to guide and not to fetter.

His second important peculiarity shows that he must have been an odd and crotchety creature, but one with sense in his crotchets. He seems to have been annoyed by mispronunciation of his own and other work: and accordingly he adopts (with full warning and explanation) the plan of invariably doubling the consonant after every short vowel without exception. This gives a most grotesque air to his pages, which are studded with words like "nemmnedd" (named), "forrwerrpenn" (to despise), "tunderrstanndenn" (to understand), and so forth. But, in the first place, it fixes for all time, in a most invaluable manner, the pronunciation of English at that time; and in the second, it shows that Orm had a sound understanding of that principle of English which has been set at nought by those who would spell "traveller" "traveler." He knew that the tendency, and the, if not warned, excusable tendency, of an English tongue would be to pronounce this traveeler. It is a pity that knowledge which existed in the twelfth century should apparently have become partial ignorance close to the beginning of the twentieth.

The Ancren Riwle[91] has no oddities of this kind, and nothing particularly noticeable in its form, though its easy pleasant prose would have been wonderful at the time in any other European nation. Even French prose was only just beginning to take such form, and had not yet severed itself from poetic peculiarities to anything like the same extent. But then the unknown author of the Ancren Riwle had certainly four or five, and perhaps more, centuries of good sound Saxon prose before him: while St Bernard (if he wrote French prose), and even Villehardouin, had little or nothing but Latin. I have called him unknown, and he neither names himself nor is authoritatively named by any one; while of the guesses respecting him, that which identifies him with Simon of Ghent is refuted by the language of the book, while that which assigns it to Bishop Poore has no foundation. But if we do not know who wrote the book, we know for whom it was written—to wit, for the three "anchoresses" or irregular nuns of a private convent or sisterhood at Tarrant Keynes in Dorsetshire.

Later this nunnery, which lasted till the dissolution, was taken under the Cistercian rule; but at first, and at the time of the book, it was free, the author advising the inmates, if anybody asked, to say that they were under "the rule of St James"—i.e., the famous definition, by that apostle, of pure religion and undefiled. The treatise, which describes itself, or is described in one of its MSS., as "one book to-dealed into eight books," is of some length, but singularly pleasing to read, and gives evidence of a very amiable and sensible spirit in its author, as well as of a pretty talent for writing easy prose. If he never rises to the more mystical and poetical beauties of mediæval religion, so he never descends to its ferocities and its puerilities. The rule, the "lady-rule," he says, is the inward; the outward is only adopted in order to assist and help the inward: therefore it may and should vary according to the individual, while the inward cannot. The outward rule of the anchoresses of Tarrant Keynes was by no means rigorous. They were three in number; they had lay sisters (practically lady's-maids) as well as inferior servants. They are not to reduce themselves to bread-and-water fasting without special direction; they are not to be ostentatious in alms-giving; they may have a pet cat; haircloth and hedgehog-skins are not for them; and they are not to flog themselves with briars or leaded thongs. Ornaments are not to be worn; but a note says that this is not a positive command, all such things belonging merely to the external rule. Also they may wash just as often as it is necessary, or as they like!—an item which, absurd as is the popular notion of the dirt of the Middle Ages, speaks volumes for the sense and taste of this excellent anonym.

This part is the last or eighth "dole," as the sections are termed; the remaining seven deal with religious service, private devotion, the Wesen or nature of anchorites, temptation, confession, penance, penitence, and the love of God. Although some may think it out of fashion, it is astonishing how much sense, kindliness, true religion, and useful learning there is in this monitor of the anchoresses of Tarrant Keynes, which place a man might well visit in pilgrimage to do him honour. Every now and then, rough as is his vehicle of speech—a transition medium, endowed neither with the oak-and-rock strength of Anglo-Saxon nor with the varied gifts of modern English—he can rise to real and true eloquence, as where he speaks of the soul and "the heavy flesh that draweth her downwards, yet through the highship [nobleness] of her, it [the flesh] shall become full light—yea, lighter than the wind is, and brighter than the sun is, if only it follow her and draw her not too hard to its own low kind." But though such passages, good in phrase and rhythm, as well as noble in sense, are not rare, the pleasant humanity of the whole book is the best thing in it. M. Renan oddly enough pronounced Ecclesiastes, that voice of the doom of life, to be "le seul livre aimable" which Judaism had produced. The ages of St Francis and of the Imitation do not compel us to look about for a seul livre aimable, but it may safely be said that there is none more amiable in a cheerful human way than the Ancren Riwle.

It would serve no purpose here to discuss in detail most of the other vernacular productions of the first half of the thirteenth century in English.[92] They are almost without exception either religious—the constant rehandling of the time cannot be better exemplified than by the fact that at least two paraphrases, one in prose, one in verse, of one of the "doles" of the Ancren Riwle itself exist—or else moral-scientific, such as the Bestiary,[93] so often printed. One of the constantly recurring version-paraphrases of the Scriptures, however—the so-called Story of Genesis and Exodus,[94] supposed to date from about the middle—has great interest, because here we find (whether for the first time or not he would be a rash man who should say, but certainly for almost, if not quite, the first) the famous "Christabel" metre—iambic dimeter, rhymed with a wide licence of trisyllabic equivalence. This was to be twice revived by great poets, with immense consequences to English poetry—first by Spenser in the Kalendar, and then by Coleridge himself—and was to become one of the most powerful, varied, and charming of English rhythms. That this metre, the chief battle-ground of fighting between the accent-men and the quantity-men, never arose till after rhymed quantitative metre had met accentual alliteration, and had to a great extent overcome it, is a tell-tale fact, of which more hereafter. And it is to be observed also that in this same poem it is possible to discover not a few very complete and handsome decasyllables which would do no discredit to Chaucer himself.

But the Owl and the Nightingale[95] is another kind of thing. In the first place, it appears to be (though it would be rash to affirm this positively of anything in a form so popular with the French trouvères as the débat) original and not translated. It bears a name, that of Nicholas of Guildford, who seems to be the author, and assigns himself a local habitation at Portesham in Dorsetshire. Although of considerable length (nearly two thousand lines), and written in very pure English with few French words, it manages the rhymed octosyllabic couplet (which by this time had become the standing metre of France for everything but historical poems, and for some of these) with remarkable precision, lightness, and harmony. Moreover, the Owl and the Nightingale conduct their debate with plenty of mother-wit, expressed not unfrequently in proverbial form. Proverbs. Indeed proverbs, a favourite form of expression with Englishmen at all times, appear to have been specially in favour just then; and the "Proverbs of Alfred"[96] (supposed to date from this very time), the "Proverbs of Hendyng"[97] a little later, are not likely to have been the only collections of the kind. The Alfred Proverbs are in a rude popular metre like the old alliteration much broken down; those of Hendyng in a six-line stanza (soon to become the famous ballad stanza) syllabled, though sometimes catalectically, 8 8 6 8 8 6, and rhymed a a b c c b, the proverb and the coda "quod Hendyng" being added to each. The Owl and the Nightingale is, however, as we might expect, superior to both of these in poetical merit, as well as to the so-called Moral Ode which, printed by Hickes in 1705, was one of the first Middle English poems to gain modern recognition.

As the dividing-point of the thirteenth and fourteenth centuries approaches, the interest of literary work increases, and requires less and less allowance of historical and accidental value. This allowance, indeed, is still necessary with the verse chronicle of Robert of Gloucester,[98] the date of which is fixed with sufficient certainty at 1298. This book has been somewhat undervalued, in point of strict literary merit, from a cause rather ludicrous but still real. It will almost invariably be found that those mediæval books which happen to have been made known before the formal beginning of scholarship in the modern languages, are underrated by modern scholars, who not unnaturally put a perhaps excessive price upon their own discoveries or fosterlings. Robert of Gloucester's work, with the later but companion Englishing of Peter of Langtoft by Robert Manning of Brunne, was published by Hearne in the early part of the last century. The contemporaries of that publication thought him rude, unkempt, "Gothick": the moderns have usually passed him by for more direct protégés of their own. Yet there is not a little attraction in Robert. To begin with, he is the first in English, if not the first in any modern language, to attempt in the vernacular a general history, old as well as new, new as well as old. And the opening of him is not to be despised—

And in the same good swinging metre he goes on describing the land, praising its gifts, and telling its story in a downright fashion which is very agreeable to right tastes. Like almost everybody else, he drew upon Geoffrey of Monmouth for his early history: but from at least the time of the Conqueror (he is strongly prejudiced in the matter of Harold) he represents, if not what we should call solid historical knowledge, at any rate direct, and for the time tolerably fresh, historical tradition, while as he approaches his own time he becomes positively historical, and, as in the case of the Oxford town-and-gown row of 1263, the first Barons' Wars, the death of the Earl-Marshal, and such things, is a vigorous as well as a tolerably authoritative chronicler. In the history of English prosody he, too, is of great importance, being another landmark in the process of consolidating accent and quantity, alliteration and rhyme. His swinging verses still have the older tendency to a trochaic rather than the later to an anapæstic rhythm; but they are, so to speak, on the move, and approaching the later form. He is still rather prone to group his rhymes instead of keeping the couplets separate: but as he is not translating from chanson de geste form, he does not, as Robert of Brunne sometimes does, fall into complete laisses. I have counted as many as twenty continuous rhymes in Manning, and there may be more: but there is nothing of that extent in the earlier Robert.

Verse history, however, must always be an awkward and unnatural form at the best. The end of the thirteenth century had something better to show in the appearance of romance proper and of epic. When the study of any department of old literature begins, there is a natural and almost invariable tendency to regard it as older than it really is; and when, at the end of the last century, the English verse romances began to be read, this tendency prevailed at least as much as usual. Later investigation, besides showing that, almost without exception, they are adaptations of French originals, has, partly as a consequence of this, shown that scarcely any that we have are earlier than the extreme end of the thirteenth century. Among these few that are, however, three of exceptional interest (perhaps the best three except Gawaine and the Green Knight and Sir Launfal) may probably be classed—to wit, Horn, Havelok, and the famous Sir Tristram. As to the last and best known of these, which from its inclusion among Sir Walter Scott's works has received attention denied to the rest, it may or may not be the work of Thomas the Rhymer. But whether it is or not, it can by no possibility be later than the first quarter of the fourteenth century, while the most cautious critics pronounce both Havelok the Dane and King Horn to be older than 1300.[99]

It is, moreover, not a mere accident that these three, though the authors pretty certainly had French originals before them, seem most likely to have had yet older English or Anglo-Saxon originals of the French in the case of Horn and Havelok, while the Tristram story, as is pointed out in the chapter on the Arthurian Legend, is the most British in tone of all the divisions of that Legend. Havelok and Horn have yet further interest because of the curious contrast between their oldest forms in more ways than one. Havelok is an English equivalent, with extremely strong local connections and identifications, of the homelier passages of the French chansons de geste. The hero, born in Denmark, and orphan heir to a kingdom, is to be put away by his treacherous guardian, who commits him to Grim the fisherman to be drowned. Havelok's treatment is hard enough even on his way to the drowning; but as supernatural signs show his kingship to Grim's wife, and as the fisherman, feigning to have performed his task, meets with very scant gratitude from his employer, he resolves to escape from the latter's power, puts to sea, and lands in England at the place afterwards to be called from him Grimsby. Havelok is brought up simply as a rough fisher-boy; but he obtains employment in Lincoln Castle as porter to the kitchen, and much rough horse-play of the chanson kind occurs. Now it so happens that the heiress of England, Goldborough, has been treated by her guardian with as much injustice though with less ferocity; and the traitor seeks to crown his exclusion of her from her rights by marrying her to the sturdy scullion. When the two rights are thus joined, they of course prevail, and the two traitors, after a due amount of hard fighting, receive their doom, Godard the Dane being hanged, and Godric the Englishman burnt at the stake. This rough and vigorous story is told in rough and vigorous verse—octosyllabic couplets, with full licence in shortening, but with no additional syllables except an occasional double rhyme—in very sterling English, and with some, though slight, traces of alliteration.

Horn (King Horn, Horn-Child and Maiden Rimnilde, &c.) is somewhat more courtly in its general outlines, and has less of the folk-tale about it; but it also has connections with Denmark, and it turns upon treachery, as indeed do nearly all the romances. Horn, son of a certain King Murray, is, in consequence of a raid of heathen in ships, orphaned and exiled in his childhood across the sea, where he finds an asylum in the house of King Aylmer of Westerness. His love for Aylmer's daughter Rimenhild and hers for him (he is the most beautiful of men), the faithfulness of his friend Athulf (who has to undergo the very trying experience of being made violent love to by Rimenhild under the impression that he is Horn), and the treachery of his friend Fikenild (who nearly succeeds in making the princess his own), defray the chief interest of the story, which is not very long. The good steward Athelbrus also plays a great part, which is noticeable, because the stewards of Romances are generally bad. The rhymed couplets of this poem are composed of shorter lines than those of Havelok. They allow themselves the syllabic licence of alliterative verse proper, though there is even less alliteration than in Havelok, and they vary from five to eight syllables, though five and six are the commonest. The poem, indeed, in this respect occupies a rather peculiar position. Yet it is all the more valuable as showing yet another phase of the change.

The first really charming literature in English has, however, still to be mentioned: and this is to be found in the volume—little more than a pamphlet—edited fifty years ago for the Percy Society (March 1, 1842) by Thomas Wright, under the title of Specimens of Lyric Poetry composed in England in the Reign of Edward the First, from MS. 2253 Harl. in the British Museum. The first three poems are in French, of the well-known and by this time far from novel trouvère character, of which those of Thibaut of Champagne are the best specimens. The fourth—

is English, and is interesting as copying not the least intricate of the trouvère measures—an eleven-line stanza of eight sevens or sixes, rhymed ab, ab, ab, ab, c, b, c; but moral-religious in tone and much alliterated. The fifth, also English, is anapæstic tetrameter heavily alliterated, and mono-rhymed for eight verses, with the stanza made up to ten by a couplet on another rhyme. It is not very interesting. But with VI. the chorus of sweet sounds begins, and therefore, small as is the room for extract here, it must be given in full:—

The next, "With longyng y am lad," is pretty, though less so: and is in ten-line stanzas of sixes, rhymed a a b, a a b, b a a b. Those of VIII. are twelve-lined in eights, rhymed ab, ab, ab, ab, c, d, c, d; but it is observable that there is some assonance here instead of pure rhyme. IX. is in the famous romance stanza of six or rather twelve lines, à la Sir Thopas; X. in octaves of eights alternately rhymed with an envoy quatrain; XI. (a very pretty one) in a new metre, rhymed a a a b a, b. And this variety continues after a fashion which it would be tedious to particularise further. But it must be said that the charm of "Alison" is fully caught up by—

by a sturdy Praise of Women which charges gallantly against the usual mediæval slanders; and by a piece which, with "Alison," is the flower of the whole, and has the exquisite refrain—

Here is Tennysonian verse five hundred years before Tennyson. The "cry" of English lyric is on this northern wind at last; and it shall never fail afterwards.

This seems to be the best place to deal, not merely with the form of English lyric in itself, but with the general subject of the prosody as well of English as of the other modern literary languages. A very great[100] deal has been written, with more and with less learning, with ingenuity greater or smaller, on the origins of rhyme, on the source of the decasyllabic and other staple lines and stanzas; and, lastly, on the general system of modern as opposed to ancient scansion. Much of this has been the result of really careful study, and not a little of it the result of distinct acuteness; but it has suffered on the whole from the supposed need of some new theory, and from an unwillingness to accept plain and obvious facts. These facts, or the most important of them, may be summarised as follows: The prosody of a language will necessarily vary according to the pronunciation and composition of that language; but there are certain general principles of prosody which govern all languages possessing a certain kinship. Historical retrospect. These general principles were, for the Western branches of the Aryan tongues, very early discovered and formulated by the Greeks, being later adjusted to somewhat stiffer rules—to compensate for less force of poetic genius, or perhaps merely because licence was not required—by the Latins. Towards the end of the classical literary period, however, partly the increasing importance of the Germanic and other non-Greek and non-Latin elements in the Empire, partly those inexplicable organic changes which come from time to time, broke up this system. Rhyme appeared, no one knows quite how, or why, or whence, and at the same time, though the general structure of metres was not very much altered, the quantity of individual syllables appears to have undergone a complete change. Although metres quantitative in scheme continued to be written, they were written, as a rule, with more or less laxity; and though rhyme was sometimes adapted to them in Latin, it was more frequently used with a looser syllabic arrangement, retaining the divisional characteristics of the older prosody, but neglecting quantity, the strict rules of elision, and so forth.

On the other hand, some of the new Teutonic tongues which were thus brought into contact with Latin, and with which Latin was brought into contact, had systems of prosody of their own, based on entirely different principles. The most elaborate of these probably, and the only one from which we have distinct remains of undoubtedly old matter in considerable quantities, is Anglo-Saxon, though Icelandic runs it close. A detailed account of the peculiarities of this belongs to the previous volume: it is sufficient to say here that its great characteristic was alliteration, and that accent played a large part, to the exclusion both of definite quantity and of syllabic identity or equivalence.

While these were the states of things with regard to Latin on the one hand, and to the tongues most separated from Latin on the other, the Romance languages, or daughters of Latin, had elaborated or were elaborating, by stages which are almost entirely hidden from us, middle systems, of which the earliest, and in a way the most perfect, is that of Provençal, followed by Northern French and Italian, the dialects of the Spanish Peninsula being a little behindhand in elaborate verse. The three first-named tongues seem to have hit upon the verse of ten or eleven syllables, which later crystallised itself into ten for French and eleven for Italian, as their staple measure.[101] Efforts have been made to father this directly on some classical original, and some authorities have even been uncritical enough to speak of the connection—this or that—having been "proved" for these verses or others. No such proof has been given, and none is possible. What is certain, and alone certain, is that whereas the chief literary metre of the last five centuries of Latin had been dactylic and trisyllabic, this, the chief metre of the daughter tongues, and by-and-by almost their only one, was disyllabic—iambic, or trochaic, as the case may be, but generally iambic. Rhyme became by degrees an invariable or almost invariable accompaniment, and while quantity, strictly speaking, almost disappeared (some will have it that it quite disappeared from French), a syllabic uniformity more rigid than any which had prevailed, except in the case of lyric measures like the Alcaic, became the rule. Even elision was very greatly restricted, though cæsura was pretty strictly retained, and an additional servitude was imposed by the early adoption in French of the fixed alternation of "masculine" and "feminine" rhymes—that is to say, of rhymes with, and rhymes without, the mute e.