my purpose, which is to fix the sound of (uu) for them all.

52 This (uu) con

firmed by Dutch and German.

And if my readers are not tired of Dutch and German, I will call their attention to the

Thus, as in the case of our i words, the Dutch and

German languages lend an emphatic confirmation to the

evidence of our own almost universal tradition as to the sound of these words with (uu).

duplicate pre

terit.

53 Did not a re- (The settlement of the ancient pronunciation of do as (duu), as at present, enables us to get rid of the erroneous notion that did is a reduplicate preterit. If, as I believe with Mr. Ellis, the A.S. y was sounded (y), dyde, the old preterit of dón, was simply a weak preterit, regularly formed except as to modification of the vowel by the "umlaut," precisely as in jung, jünger, &c. It is not really the same but a different question whether in weak verbs generally the termination -de is derived from the verb do. This notion is based on the fact that there are two ds in the dual and plural of the MæsoGothic preterit; the terminations of that tense being

But in the M.G. for to do the preterit in full runs thus—

=

=

And here we lose the apparent reduplication, or approach to reduplication, which we have in the English dyde or dide, the root tau and the termination dedu being by no means so similar. Moreover, if that is the derivation, why should rodida mean "I did speak" rather than "I do speak"? And the word di-de itself when so explained becomes do + di-de do + do + di-de do+do+do+ di-de, and so on: a manifest absurdity. It seems far more satisfactory to consider the dental d (or t) of the preterit akin to the dental d (or t or the cognate dental-nasal-liquid n) of the past participle. It is true that the French express "I will speak" by "I have to speak,” parler-ai, and "we have to speak" by parler-(av)ons, and have in course of time run together these and similar pairs

of words into single words; but we are not warranted in pushing this analogy so far as to contend that every inflexion in every language was originally a separate and distinct word. I should much rather believe that, letters having a force of their own (as the sibilant in eis and pós for example suitably expresses the idea of motion combined with that of the radical ev or πpó), the explosive t or d or the kindred nasal indicated interjection-wise a sense of relief when the action was finished and the work accomplished; and this equally in åɣaπîós and amatus, geliebt and loved (pt.), liebte, lufode, dyde, done and TUTTÓμe-v-os. It would be digressing too far to trace a like adaptation of sound to sense in amant-, amand-, TUTTOVт- (tiip'toond), TUTTEVT- (tii pend), liebend, lufiand, &c.)

cer, yet distinct

in sound.

54 A.S. 4 and 6 We have already seen one class of words in both o in Chau- which the A.S. á stands for (00) or (00), and now we find the A.S. 6 in another considerable class represents the sound of (uu); and, as I believe, these words were always so pronounced. Mr. Ellis, however, imagines that do, to, schoo, doom, soone, &c., were pronounced with (oo), like go, so, mo, stoon, noon (adj.), &c. These two classes of words are, in fact, totally distinct; but misled by the mere written language, and too implicitly believing that "the orthography shows the sound" (p. 255, heading), Mr. Ellis has confounded them, regardless of the distinction in their pronunciation now and certainly for some centuries, and of the distinction. in their orthography in A.S., and utterly deaf to the clamorous protests of their continental kinsmen. Chaucer it is true these classes of words are spelt alike, but pronounced alike they are not. Innumerable in Chaucer are the rhymes of go, i-go, so, also, woo (= woe), tho, mo, foo, fro, too (=toe), slo (= sloe), who, two (=two), ho, no, flo (= arrow); and very numerous those of do, i-do, fordo, to, thereto, and schoo. Once only in all the Canterbury Tales does do rhyme with so, once only i-do with ho, once therto with mo; but the numeral two seems somewhat shifting towards its present pronunciation, for twice it

In

rhymes with do and twice with i-do. Again we have brode, loode (s), glod, bistrood, rhyming together, and rood (vb.), abode, prentyshood, are of the same class: but not once in all the Cant. Ta. do these rhymes with food, stood, understode, mood, wode, hode, blode, flood. Again, cloke, poke (s.), broke, smoke rhyme, as do strook and oak; but none of these rhyme with schook, cooke, took, wook, awook, quook, forsook, look, pook. Again, oon, aloon, anoon, echoon, ilkoone, everychon, bone, gone, agoon, crone, schon (vb.), ton (= toes), lone (loan), moone (= moan), persone, stone, and some proper names, furnish an immense number of rhymes; doon, i-doon, soone, boone, moone (= luna), spoon, noon (= midday), also a large number: only four imperfect rhymes are there, and for these doon is responsible. Goos rhymes with schoos and with loos (adj.); but not once with loos (= laus), cloos (adj.), close (vb.), toos (=toes), glose, "chose," rose, hose, nose, pose, suppose, purpose, dispose. Swoote, Chaucer's epithet of April showers, and the pronunciation of which is tolerably fixed by the Dutch zoet and German süss, rhymes with roote, bote (= remedy), and foot; none of these rhyme even once with noote, rote, coote (= coat), bot (vb.), throte, hote, woot, noot (vb.), boot (= boat), wroot, goot, ote-s, smoot. Lastly, with sooth we have tooth rhyming, and doth (now dŭth); but not goth (= goeth), cloth, loth, wroth, bothe, oth: once only forsothe rhymes with bothe. So perfectly distinct were the (oo) and the (uu) words in Chaucer's language, however spelt.

An examination of the first five thousand lines of Roberde of Brunne's Handlyng Synne, for the o words, gives just such results as are derived from Chaucer. Mo, go, oo (= aye), wo, slo (= slay), þo (adv.), þo (pron.), fro, rhyme with one another exclusively: do rhymes regularly with to and its compounds: two rhymes once with slo, once with do; but so and also, curiously enough, and quite contrary to Chaucer's usage, rhyme only with do and to, except once only with the doubtful numeral two.*

* In like manner

* As so, also, two, all had similar forms in A.S., swá, alswá, twá, they might have been expected to undergo like changes. In fact the w tended to

noon, soon, shoon rhyme with don (inf.) and done (part.), never with bone, stone, gone, one with its compounds, &c. Once only the part. done rhymes with none, and twice with nouns of French derivation in -un (which R. of Br. writes more commonly than -on or -oun). So fote (= foot), boot (= remedy), rhyme with each other, but never with hote (adj.), hote (= promise), smote, grote, wrote, wote, note, prote, and so on, though a bad rhyme, such as come with gone, goste with hast, occurs here and there.

ceptionsmistaken

55 Occasional ex- As to one imperfect rhyme here and there, by Mr. Ellis for any reader of modern English verse might well be surprised if there were not in Chaucer any

the rule.

such macula

quas aut incuria fudit,

Aut humana parum cavit natura.*

With such imperfect rhymes Chaucer seems to have been content in dealing with proper names and foreign words. Thus while Amazone and Salamon alone occur, rhyming with stone, &c.; we have not only Palamon rhyming with anoon, &c., but also Palamoun rhyming with doun and toun (eleven of the former rhymes, eighteen of the latter). Plato rhymes once with tho, once with to; Funo with fordo; principio with schoo; Cupido, Placebo, change the (o) into (u) in all of them. They all hesitated, two finally gave way, but so and also stood firm in the original sound after ejecting the semivowel.

* A lady has kindly collected for me a few such faulty rhymes from some of our 19th century poets :—

KEATS: wood, flood; loll, poll; Arabian, man; trees, essences; these, offices; exhalations, cons; beautiful, cull; strawberries, butterflies.

SHELLEY: hail, majestical; death, path; shun, on; now, glow; feet, yet; abode, brotherhood; burning, morning.

COLERIDGE: guest, dismist; hear, Mariner; groan, one; fear, were; full, dull; fair, are; humming, women.

WORDSWORTH: flood, wood; gone, alone; dead, laid; ere, near; how, fro; long, hung; forth, earth; now, low; road, abroad; come, home; groves, loves; breath, underneath; year, fair.

TENNYSON early, barley; weary, airy; brow, snow; close (vb.), house; ran, swan; was, pass; wood, bud.

W. MORRIS: afar, war; were, near; heard, afeard; bear, rear; stood, blood; gone, alone; throne, upon; below, bow (vb.); here, artificer.