an was put into could, where it does not belong, to satisfy the logic of the eye, which has affected the pronunciation and even the spelling of English more than is commonly supposed. I meet with eyster for oyster as early as the fourteenth century. I find dystrye for destroy in the Coventry Plays, viage in Bishop Hall and Middleton the dramatist, bile in Donne and Chrononhotonthologos, line in Hall, ryall and chyse (for choice) in the Coventry Plays. In Chapman's "All Fools" is the misprint of employ for imply, fairly inferring an identity of sound in the last syllable. Indeed, this pronunciation was habitual till after Pope, and Rogers tells us that the elegant Gray said naise for noise just as our rustics still do. Our cornish (which I find also in Herrick) remembers the French better than cornice does. While, clinging more closely to the AngloSaxon in dropping the g from the end of the present participle, the Yankee now and then pleases himself with an experiment in French nasality in words ending in n. It is not, so far as my experience goes, very common, though it may formerly have been more so. Capting, for instance, I never heard save in jest, the habitual form being kepp'n. But at any rate it is no invention of ours. In that delightful old volume, "Ane Compendious Buke of Godly and Spirituall Songs," in which I know not

b

whether the piety itself or the simplicity of its expression be more charming, I find burding, garding, and cousing, and in the State Trials uncerting used by a gentleman. The n for ng I confess preferring.

Of Yankee preterites I find risse and rize for rose in Middleton and Dryden, clim in Spenser, chees (chose) in Sir John Mandevil, give (gave) in the Coventry Plays, shet (shut) in Golding's Ovid,* het in Chapman and in Weever's Epitaphs, thriv and smit in Drayton, quit in Ben Jonson and Henry More, and pled in the fastidious Landor. Rid for rode was anciently common. So likewise was see

for saw, but I find it in no writer of authority, unless Chaucer's seie was so sounded. Shew is used by Hector Boece, Giles Fletcher, and Drummond of Hawthornden. Similar strong preterites, like snew, thew, and even mew, are not without example. I find sew for sowed in Piers Ploughman. Indeed, the anomalies in English preterites are perplexing. We have probably transferred flew from flow (as the preterite of which I have heard it) to fly because we had another preterite in fled. Of weak preterites the Yankee retains growed, blowed, for which he has good authority, and less often knowed. His sot is merely a broad sounding of sat, no more inelegant than the com*Cited in Warton's Obs. Faery Q.

mon got for gat, which he further degrades into gut. When he says darst, he uses a form as old as Chaucer.

The Yankee has retained something of the long sound of the a in such words as axe, wax, pronouncing them exe, wex (shortened from aix, waix). He also says hev and hed (have, had) for have and had. In most cases he follows an Anglo-Saxon usage. In aix for axle he certainly does. I find wex and aisches (ashes) in Pecock, and exe in the Paston Letters. Chaucer wrote hendy. Dryden rhymes can with men, as Mr. Biglow would. Alexander Gill, Milton's teacher, in his "Lagonomia" cites hez for hath as peculiar to Lincolnshire. I find hayth in Collier's "Bibliographical Account of Early English Literature" under the date 1584, and Lord Cromwell so wrote it. Sir Christopher Wren wrote belcony. Thaim for them was common in the sixteenth century. We have an example of the same thing in the double form of the verb thrash, thresh. While the New-Englander cannot be brought to say instead for instid (commonly 'stid where not the last word in a sentence), he changes the i into e in red for rid, tell for till, hender for hinder, rense for rinse. I find red in the old interlude of "Thersytes," tell in a letter of Daborne to Henslowe, and also, I shudder to mention it, in a letter of the great Duchess of Marlborough, Atossa herself! It oc

curs twice in a single verse of the Chester Plays, which I copy as containing another Yankeeism:

"Tell the day of dome, tell the heames blow."

From this word blow is formed blowth, which I heard again this summer after a long interval. Mr. Wright explains it as meaning "a blossom." With us a single blossom is a blow, while blowth means the blossoming in general. A farmer would say that there was a good blowth on his fruit-trees. The word retreats farther inland and away from the railways, year by year. Wither rhymes hinder with slender, and Lovelace has renched for rinsed. In "Gammer Gurton" is sence for since; Marlborough's Duchess so writes it, and Donne rhymes since with Amiens and patience, Bishop Hall and Otway with pretence, Chapman with citizens, Dryden with providence. Indeed, why should not sithence take that form?

E sometimes takes the place of u, as jedge, tredge, bresh. I find tredge in the interlude of "Jack Jugler," bresh in a citation by Collier from "London Cries" of the middle of the seventeenth century, and resche for rush (fifteenth century) in the very valuable "Volume of Vocabularies" edited by Mr. Wright. Resce is one of the Anglo-Saxon forms of the word in Bosworth's A. S. Dictionary. The

* Dictionary of Obsolete and Provincial English.

Yankee always shortens the u in the ending ture, making ventur, natur, pictur, and so on. This was common, also, among the educated of the last generation. I am inclined to think it may have been once universal, and I certainly think it more elegant than the vile vencher, naycher, pickcher, that have taken its place, sounding like the invention of a lexicographer with his mouth full of hot pudding. Nash in his " Pierce Penniless" has ventur, and so spells it, and I meet it also in Spenser, Drayton, Ben Jonson, Herrick, and Prior. Spenser has tort'rest, which can only be contracted from tortur and not from torcher. Quarles rhymes nature with creator, and Dryden with satire, which he doubtless pronounced according to its older form of satyr.

I shall now give some examples which cannot so easily be ranked under any special head. Gill charges the Eastern counties with kiver for cover, and ta for to. The Yankee pronounces both too and to like ta (like the tou in touch) where they are not emphatic. In that case, both become tu. In old spelling, to is the common (and indeed correct) form of too, which is only to with the sense of in addition. I suspect that the sound of our too has caught something from the French tout, and it is possible that the old too-too is not a reduplication, but a reminiscence of the feminine form of the same word (toutes) as anciently pronounced,