prairie, to vamose.

Even these are fewer than is sometimes taken for granted. But I think some fair defence may be made against the charge of vulgarity. Properly speaking, vulgarity is in the thought, and not in the word or the way of pronouncing it. Modern French, the most polite of languages, is barbarously vulgar if compared with the Latin out of which it has been corrupted, or even with Italian. There is a wider gap, and one implying greater boorishness, between ministerium and métier, or sapiens and sachant, than between druv and drove, or agin and against, which last is plainly an arrant superlative. Our rustic coverlid is nearer its French original than the diminutive coverlet, into which it has been ignorantly corrupted in politer speech. I obtained from three cultivated Englishmen at different times three diverse pronunciations of a single word, -cowcumber, coocumber, and cucumber. Of these the first, which is Yankee also, comes nearest to the nasality of concombre. Lord Ossory assures us that Voltaire saw the best society in England, and Voltaire tells his countrymen that handkerchief was pronounced hankercher. I find it so spelt in Hakluyt and elsewhere. This enormity the Yankee still persists in, and as there is always a reason for such deviations from the sound as represented by the spelling, may we not suspect two sources of deriva

tion, and find an ancestor for kercher in couverture rather than in couvrechef? And what greater phonetic vagary (which Dryden, by the way, called fegary) in our lingua rustica than this ker for couvre? I copy from the fly-leaves of my books where I have noted them from time to time a few examples of pronunciation and phrase which will show that the Yankee often has antiquity and very respectable literary authority on his side. My list might be largely increased by referring to glossaries, but to them every one can go for himself, and I have gathered enough for my purpose.

I will take first those cases in which something like the French sound has been preserved in certain single letters and diphthongs. And this opens a curious question as to how long this Gallicism maintained itself in England. Sometimes a divergence in pronunciation has given us two words with different meanings, as in genteel and jaunty, which I find coming in toward the close of the seventeenth century, and wavering between genteel and jantee. It is usual in America to drop the u in words ending in our, a very proper change recommended by Howell two centuries ago, and carried out by him so far as his printers would allow. This and the corresponding changes in musique, musick, and the like, which he also advocated, show that in his time the French accent indicated by the superfluous letters

strong an accent There is plenty

(for French had once nearly as as Italian) had gone out of use. of French accent down to the end of Elizabeth's reign. In Daniel we have riches' and counsel, in Bishop Hall comet', chapelain, in Donne pictures', virtue', presence, mortal', merit, hainous', giant, with many more, and Marston's satires are full of them. The two latter, however, are not to be relied on, as they may be suspected of Chaucerizing. Herrick writes baptime. The tendency to throw the accent backward began early. But the incongruities are perplexing, and perhaps mark the period of transition. In Warner's "Albion's England" we have creator' and creature' side by side with the modern creator and creature. Envy and envying occur in Campion (1602), and yet envy' survived Milton. In some cases we have gone back again nearer to the French, as in revenue for revenue. I had been so used to hearing imbecile pronounced with the accent on the first syllable, which is in accordance with the general tendency in such matters, that I was surprised to find imbecile in a verse of Wordsworth. The dictionaries all give it so. I asked a highly cultivated Englishman, and he declared for imbeceel'. In general it may be assumed that accent will finally settle on the syllable dictated by greater ease and therefore quickness of utterance. Blasphemous, for example, is more rapidly pro

nounced than blasphem'ous, to which our Yankee clings, following in this the usage of many of the older poets. American is easier than American, and therefore the false quantity has carried the day, though the true one may be found in George Herbert, and even so late as Cowley.

To come back to the matter in hand. Our " uplandish men" retain the soft or thin sound of the u in some words, such as rule, truth (sometimes also pronounced truth, not trooth), while he says noo for new, and gives to view and few so indescribable a mixture of the two sounds with a slight nasal tincture that it may be called the Yankee shibboleth. In rule the least sound of a precedes the u. I find reule in Pecock's "Repressor." He probably pronounced it rayoole, as the old French word from which it is derived was very likely to be sounded at first, with a reminiscence of its original regula. Tindal has rueler, and the Coventry Plays have preudent. As for noo, may it not claim some sanction in its derivation, whether from nouveau or neuf, the ancient sound of which may very well have been noof, as nearer novus? Beef would seem more like to have come from buffe than from bœuf, unless the two were mere varieties of spelling. The Saxon few may have caught enough from its French cousin peu to claim the benefit of the same doubt as to sound; and our slang phrase a few (as "I licked

For

him a few") may well appeal to un peu for sense and authority. Nay, might not lick itself turn out to be the good old word lam in an English disguise, if the latter should claim descent as, perhaps, he fairly might, from the Latin lambere? The New England ferce for fierce, and perce for pierce (sometimes heard as fairce and pairce), are also Norman. its antiquity I cite the rhyme of verse and pierce in Chapman and Donne, and in some commendatory verses by a Mr. Berkenhead before the poems of Francis Beaumont. Our pairlous for perilous is of the same kind, and is nearer Shakespeare's parlous than the modern pronunciation. One other Gallicism survives in our pronunciation. Perhaps I should rather call it a semi-Gallicism, for it is the result of a futile effort to reproduce a French sound with English lips. Thus for joint, employ, royal we have jynt, emply, ryle, the last differing only from rile (roil) in a prolongation of the y sound. In Walter de Biblesworth I find solives Englished by gistes. This, it is true, may have been pronounced jeests, but the pronunciation jystes must have preceded the present spelling, which was no doubt adopted after the radical meaning was forgotten, as analogical with other words in oi. In the same way after Norman-French influence had softened the out of would (we already find woud for veut in N. F. poems), should followed the example, and then