all changes of pronunciation are usually made by fmail degrees; and therefore when we find that a great number of thofe words which in Chaucer's time ended in e originally ended in a, we may reasonably prefumē that our ancestors firft paffled from the broader found of a to the thinner found of feminine, and not at once from a to e mute. Befides, if the final e in fuch words was not pronounced why was it added? From the time that it has confeffedly ceafed to be pronounced it has been gradually omitted in them, except where it may be fuppofed of use to lengthen or foften (70) the preceding fyllable, as inhope, name, &c. But according to the ancient orthography it terminates many words of Saxon original where it cannot have been added for any fuch purpose, as berte, childe, olde, wilde, &c. In thefe therefore we must fuppofe that it was Pronounced as an e feminine, and made part of a fecond fyllable, and fo, by a parity of reafon, in all others in which, as in thefe, it ap pears to have been fubftituted for the Saxon e,

Upon the fame grounds we may presume that in words terminated according to the Saxon form in eil, fuch as the infinitive modes and plural numbers of verbs, and a great variety of adverbs and prepofitions, then only was at firft thrown away, and the

(70) In most of the words in which the final e has been omitted its use in lengthening or foftening the preceding fyilable has been supplied by an alteration in the orthography of that fyl lable; thus in--grete, mete, fiele, rede, dere--in which the firft e was orginially long as clofing a fyllable, it has (fince they have been pronounced as monofyllables) been changed either into ea, as in-great, meat, fteai, read, dear; or into ce, as ingreet, meet, feel, reed, deer. In like manner the o in--bote, fole dore, gode, mone, has been changed either into oa, as in---boat, foal; or into 00, as in---door, good, moen..

which then became final, continued for a long time o be pronounced as well as written..

Thefe confiderations feem fufficient to make us believe that the pronunciation of the e feminine is Founded on the very nature of both the French and Saxon parts of our language; and therefore though we may not be able to trace the reasons of that pronunciation in all cafes fo plainly as in thofe which have been just mentioned, we may fafely, I think, conclude with the learned Wallis (71), that what is generally confidered, as an e-mute in our language,

(71) Gram. Ling. Ang. c. i. §. 2. “ Originem vero hujus e muti, " nequis mirętur unde devenerit, hanc effe judico; nempe, “quod antiquitus pronunciatum fuerit, fed obfcuro fono, ficut "Gallorum e fœmininum." He afterwards adds, Certifmum "autem hujus rei indicium eft ex antiquis poetis petendam ; "apud quos reperitur illude promifcuè vel conftituere vel non "conftituere novam fyllabam, prout ratio carminis poftulave"rit." Sothat, accordingto this judicious writer, (who has confelfedly fearched much deeperintothe formation of vocal founds in general, and the pronunciation of the English language in particular, than any of our other grammarians) I might have affumed as certain the point which I have been labouring in the text (by arguments drawn from reafon and analogy) to render probable. -There is much more to this purpose in Wallis, loc. cit, which I should transcribe, if I did not fuppofe that his book is in the hands of every one who is likely to be curious, upon this fubject. I will only take notice of one pafLage which may be wrefted to his difadvantage. From confider ing the gradual extinction of the e feminine in our language, and obferving that the French, with whom he converfed, very often fuppreffed it in their common fpeech, he has been led to predict that the pronunciation of it would perhaps shortly be difuted among then as among ourfelves. The prediction has certainly failed, but notwithstanding I will venture to say that at the time when it was made it was not unworthy of Wallis fagacity. Unluckily for its fuccefs a number of eminent writers happened at that very time to be growing up in France,

either at the end or in the middle of words (72), was anciently pronounced, but obscurely, like the e feminine of the French.

§ 17. The third kind of irregularity to which an English verfe is liable is from the accents being mifplaced. The reftoring of Chaucer's words to their juft number of fyllables, by the methods which have whofe works having fince been received as ftandards of ftyle muft probably fix for many centuries the ancient usage of the e feminine in poetry, and of course give a confiderable check to the natural progrefs of the language. If the age of Edward III. had been as favourable to letters as that of Louis XIV.; if Chaucer and his contemporary poets had acquired the fame authority here that Corneille, Moliere, Racine, and Boileau, have obtained in France; if their works had been published by themselves and perpetuated in a genuine ftate by printing, I think it probable that the e feminine would ftill have preferved its place in our poetical language at leaft, and certainly without any prejudice to the fmoothness of our verification.

(72) The reasoning in the text concerning the final e is equally applicable to the fame vowel in the middle of words. Indeed (as Wallis has obferved, loc. cit.) " vix ufpiam in me"dio dictionis reperitur e mutum, quod non ab origine fuerit "finale." If therefore it was pronounced while final, it would probably continue to be pronounced notwithstanding the addition of a fyllable: if it was pronounced in frete, trewe, large, riche, it would be pronounced in fwetely, trewely, largely, richely. [See v. 123 and 3219, v. 775 and 3692, v. 2740 and 3034, v. 1014 and 1913.] In another very numerous fet of words (French verbals ending in ment) the pronunciation of this middle e is countenanced not only by analogy, but alfo by the ftill fubfifting practice in the French language: fo Chaucer certainly pronounced the words jugement, ver. 780, 807, 820, commandement, ver. 2871, 2981, amendement, ver. 4183, pavement, avifement, ver. 4505, 4506. Even Spenfer in the fame canto (the 8th of B. v,) ufes atonement and avengement as words of four fyilables, [ft. 21. 8-30. 5.] and Wallis takes notice that the middle e in commandement was pronounced in his time.

been pointed out above, will often be of fignal fervice in reftoring his accents alfo to their proper places; but further, in many words we must be cautious of concluding too haftily that Chaucer accented the fame fyllables that we do; on the contrary, I am perfuaded that in his French words he moft commonly laid his accent, according to the French custom, (upon the lift fyllable, or the last but one, in words ending in e fer minine) which, as is well known, is the very reverse of our practice. Thus in ver.3,he udes licoûr for líquour s ver. II, coráges for courages; ver. 22, again corage for courage; ver. 37, refón for réafon ; ver. 77, viage for,váyage; ver, 109, 10, vifage, usage, for vifage úfage ver. 140, manére for manner; ver. 186, laboure for lá hour; ver. 204, prelát for prélates ver. 211, langáge for language; ver. 212, mariage for márriages ver. 216, contrée for country; and fo through the whole work.

In the fame manner he accents the laft fyllable of the participle prefent, as ver. 885, 6, wedding, coming, for wedding, coming; ver. 903, living for living ver. 907, 8, coming, crying, for coming, crying ver. 998, brenning for brenning, &c.; and as he does this in words of Saxon as well as of French growth, I fhould fuppofe that the old participle of the prefent tense ending in and was originally: accented upon that fyllable, as it certainly continued to be by the Scottish poets a long time after Chaucer. See Bp. Douglas, Virg. p. 18, ver. 18, spryngánd; ver. 51, bevand; p. 27, ver. 49, fleand; p. 29, ver. 10, feand.

Thefe inftances are all taken from the rhyining fyllables (where a strong accent is indifpenfably neceffary) in order to prove beyond contradiction that Chaucer frequently accented his words in the French manner. But if he followed this practice at the end. of his verfes, it is more than probable that he did

the fame in the middle, whenever it gave a more har monious flow to his metre; and therefore in ver. 4, instead of vértue, I fuppofe he pronounced vertue; in ver II, instead of náture, nature; in ver. 25, instead of aventure, aventure; in ver. 46, inftead of bónour, bonour, c.

It may be proper however to obferve, that we are not to expect from Chaucer that regularity in the dif pofition of his accents which the practice of our greateft poets in the laft and the prefent century has taught us to confider as effential to harmonious (73) verification: none of his mafters, either French or Italian, had fet him a pattern of exactness (74) in this refpect; and it is rather furprising that, without rule or example to guide him, he has fo feldom failed to place his accents in fuch a manner as to produce the cadence beft fuited to the nature of his verfe.

(73) It is agreed, I believe, that in our heroick metre those verfes (confidered fingly) are the moft harmonious in which the accents fail upon the even fyllables; but it has never (that I know) been defined how far a verfe may vary from this its most perfect form and yet remain a verse. On the tenth (or rhyming) fyllable a ftrong accent is in all cafes indifpenfably required, and in order to make the line tolerably harmonious, it seems neceffary that at leaft two more of the even fyllables fhould be accented, the fourth being (almost always) one of them. Milton, however, has not fubjected his verse even to thefe rules; particularly (either by negligence or defign) he has frequently put an unaccented fyllable in the fourth place. See P. L. b. iii. 36, 586, b. v. 413, 750, 874.

(74) It has been suggested above that Chaucer probably copied his heroick metre from Boccace; but neither Boccace nor any of the older Italian poets are exact in the difpofition of their accents. Though their hendecafyllable metre is allowed by the best criticks to be derived from the trimeter iambick catalectick, the perfection of it has never been determined (like that of our heroick metre) to confift in the conformity of its