bably not regret, that he did not do more towards correcting the text of Chaucer.

"In this state the Canterbury Tales remained* till the edition undertaken by Mr. Urry, which was published, some years after his death, in 1721. I shall say but little of that edition, as a very fair and full account of it is to be seen in the modest and sensible preface prefixed to it by Mr. Timothy Thomas, upon whom the charge of publishing Chaucer devolved, or rather was imposed, after Mr. Urry's death. The strange license, in which Mr. Urry appears to have indulged himself, of lengthening and shortening Chaucer's words according to his own fancy, and of even adding

"It may be proper just to take notice, that Mr. Speght's edition was reprinted in 1687, with an advertisement at the end, in which the editor pretended to publish from a MS. the conclusion of the Coke's Tale, and also of the Squires Tale, which in the printed books are said to be lost or never finished by the author. These conclusions may be seen in the Preface to Ed. Urr. Whoever the editor was, I must do him the justice to say, that they are both really to be found in MS. The first is to be found in MS. Ba. and the other in MS. B. d. from which Hearne has also printed it, as a choice discovery, in his letter to Bagford. App. to R. G. p. 601. If I thought the reader had any relish for such supplements to Chaucer, I could treat him from MS. B. a. with at least thirty more lines, which have been inserted in different parts of the Cook's Tale, by the same hand that wrote this Conclusion.

words of his own, without giving his readers the least notice, has made the text of Chaucer in his edition by far the worst that was ever published."

PLAN OF THE PRESENT EDITION.

During the latter half of the twelfth century and the earlier part of the thirteenth, the language spoken by our Saxon forefathers was rapidly breaking up, and losing its original grammatical inflections, and much of its characteristic phraseology. Books or songs written in English during this period were intended for the edification of the lower classes, or for the bourgeoisie, which still retained its Saxon habits. Great changes in language are generally coeval with political movements and convulsions, and the character of our language was completely changed by the baronial wars of the thirteenth century, which brought into prominence the Anglo-Saxon portion of the population, and made its language fashionable in high society. The consequence was, that it went through further changes in form, and became largely mixed with words having a French (or Anglo-Norman) origin. About the end of the reign of Edward I, the English language took a definite shape, which continued during the fourteenth century with very little alteration in its grammatical forms, and the only alterations in other

respects arising from words becoming obsolete, and from the facility with which French or AngloNorman words were adopted or received at the will of the author, and according to the class of society in which he moved and for which he wrote. This arose from the circumstance that English and the form of French spoken here were co-existent in our island as the languages of common life. This form of the English language was that of the author of Piers Ploughman, and of Geoffrey Chaucer, the former representing the popular feelings and containing fewest French words, while Chaucer, as the poet of the higher society, uses French words in much greater abundance. In our lan-guage of the present day, we have lost as much of the English of Piers Ploughman, as we have of the French of the Canterbury Tales.

The general character, and the grammatical constructions, of the English of the fourteenth century, were preserved during the opening years of the fifteenth, but they soon began to break up more rapidly even than in the thirteenth century, until, at the time of the Reformation, our language took nearly its modern form, the orthography excepted.

The language in which any man wrote could only be preserved correctly in manuscripts written in his own time, or very near it; for we find by

experience that copyists invariably altered what they copied to the form of the language at the time in which they wrote, and, which is still more embarrassing, to the local dialect of the county in which they lived. It is evident, therefore, that the plan of forming the text of any work of the periods of which we are speaking, from a number of different manuscripts, written at different times and different places, is the most absurd plan which it is possible to conceive. Yet this was the method professedly followed by Tyrwhitt, in forming a text of the Canterbury Tales of Chaucer. He even did worse: for he seems to have taken for his foundation merely one of the old editions, printed at a time when all the grammatical forms were lost, changing words or lines for others which pleased him better, from any manuscript which happened to contain them. It is true that he has given a list of manuscripts, in which he points out those which he considers the best, and which he followed in preference to others; but Tyrwhitt was so entirely unacquainted with the palæographical and philological knowledge necessary for the appreciation of them, that he places among his manuscripts of "highest authority," copies on paper of the latter part of the fifteenth century, while excellent manuscripts of an earlier date are looked upon with indifference. The more caution is

necessary in this respect with the text of Chaucer, because the greater number of the manuscripts are of the latter part or middle of the fifteenth century, when the language was very much changed from that of Chaucer's time.

Tyrwhitt's entire ignorance of the grammar of the language of Chaucer is exhibited in almost every line, few of which could possibly have been written by the poet as he has printed them. It need only be stated, as an instance of this, that in the preterites of what the modern Teutonic philologists term the strong verbs (which our common grammarians distinguish by the unfortunate title of irregular verbs), Tyrwhitt has invariably placed a verb in the plural with a noun in the singular. This is explained by the circumstance that, in our modern form of the language, the ancient plural of the preterite has been adopted for singular as well as plural. Examples of this (in the verbs to bear, of which the correct forms were, sing. bar, pl. bare; to come, s. cam, pl. come; to swear, s. swor, pl. swore; to give, s. gaf, pl. gave; to speak, s. spak, pl. spake; to rise, s. ros, roos, pl. rose; to take, s. took, pl. toke; &c.) occur almost in every sentence. In the verb to sit, of which the pret. s. and pl. was sette, Tyrwhitt has substituted set, a form which did not exist; and in the same manner, in the verb to creep, he has given pret. s. crept, when the forms were