the natural order of the vowel sounds, as is now admitted,

is

(i) (e) (a) (A) (0) (u) (y);

and the extremes meet in some way that has not yet, so far as I know, been explained,* so that our Miller and the German Müller are pronounced almost alike: the one sound passes with great facility into the other. And in the words referred to, while the English word has the first sound (i) of the above natural series, its congeners are furnished from the other end of the series with (y) or even (u), the latter especially in Dutch. Thus hüten (Du. hoeden, Kil.) heed, kühn (Du. koen) = keen, grün (Du. groen) = green, süss (Du. zoet) = sweet, grüssen (Du. groeten) = greet, fühlen (Du. voelen) = feel, Füsse (Du. voeten) = feet, &c.

=

In like manner it may be argued that the congeners in other languages of many of the words which Mr. Ellis would pronounce with (e), are all found vowelled from the other end of the system. Thus knee has for kinsmen the Greek yów (with yvus, πpóxvu, iyvúa), Lat. genu, Skt. jánu, Zend zenu, M.G. kniu. In O.N. alone is there any (e), but then accompanied by (i), hnè being pronounced, and sometimes written, hnie (Hniee). Indeed scarcely any congeners can be found with (e) for any of the words above given (§ 78), except only sehen and -zehn given in the last list.

The conclusion to which I am forced by this evidence from various sources—confirmed as it is to a certain extent by the testimony of Ben Jonson for the 17th century, and Palsgrave for the 16th (see § 86)-is that all these words have been sounded with (ii) in every age of our language, the 14th century of course included. And with these go many other words whose final syllable has a long e for its vowel, as the rhymes of the poets prove beyond all doubt. 84 EA words had But Mr. Ellis finds reasons for believing— (ee) or (ee): they and I have arrived at the same conclusionnever rhymed with EE words. that most, or perhaps all, of the words which in Chaucer's time were spelt with a simple e, but which two centuries later were spelt with ea, were at this later period * I am told that Mr. Melville Bell has thrown light on this.

pronounced with (ee) or (ee), while those that continued to be spelt with e or ee were at that later time pronounced with (ii). Such words are sea, flea, each, teach, preach, reach, beast, feast, read, lead (vb.), mead, sheaf, leaf, weak, speak, meal, deal, beam, dream, stream, bean, mean, lean, clean, heap, rear, tear (s.), tear (vb.), eat, heat, meat, wheat, heath, wreath, leave, weave, please, ease, tease release, cease. (I have not had time to make a complete list; though I should like to have done so, for such inquiries are, to a much greater extent than Mr. Ellis seems to suspect, inquiries about individual words.) All of these are often-perhaps most commonly-pronounced in the western counties with (ee)— say, vlay, aych, taych, and so on; but almost all of them (though flea has the same form in A.S.) are derived either from A.S. words with è or a (?), or from O.N. words with ei, or from French words with some modification of (e). These words therefore having been formerly pronounced with (ee) or (ee)—rædan, tæcan, hæp, veikr, prescher, aise, &c.-but being now pronounced with (ii), have at some time or other undergone a change; and I agree with Mr. Ellis that the change (at least in our southern dialects) has taken place later than the middle of the 17th century. In Ben Jonson the words deem, seem, esteem, redeem, rhyme with one another, but do not once in all his poems rhyme with dream, stream, moonbeam; feel, steel, eel, heel, wheel, do not once rhyme with veal, seal, steal, weal, deal (portion), deal (board), meal, heal, conceal, reveal, zeal; geese, piece, Greece, fleece, do not once rhyme with peace, increase, cease, release; deep, sleep, weep, keep, peep, steep, creep, sheep, not once with cheap, reap, heap, leap; and so on. In Spenser, so far as I have examined, the same distinction is observed, though I have found speed once rhyming with dread, and peer with ear, as occasional imperfect rhymes must be expected.* I have also examined the whole of Sir Philip Sidney's rhymes, and all of Heywood's

But as to peer, ear, if the latter was (eer), we may remember that the former is from the French pair, and Spenser may possibly have used the word with the ancient sound, spelling notwithstanding.

rhymes in his Proverbs and Epigrams (1562), and with

like result.

85 Nor do these But now I must recall attention to the re-
words rhyme in mark I made at starting as to the importance
Chaucer, though
spelt alike. of not confounding written language with the
spoken, which alone is language proper. It is the spoken
language with which we are primarily concerned; and Mr.
Ellis has been seriously misled through his attending too
exclusively to the written symbols of language. I shall
doubtless astonish him when I assert, and demonstrate,
that the very same distinction that exists between these
classes of words in Ben Jonson, Spenser, Sir Philip Sidney,
Heywood, and other poets of that age, exists also in
Chaucer, clearly and strongly marked, though disguised by
the spelling. What Mr. Ellis, justly for the most part, calls
an innovation, namely the spelling of words of the latter
class with ea, in Chaucer's time was yet unknown. But for
all that the words, though spelt alike, were not spoken
alike. The "so sharp distinction" which Mr. Ellis imagines
(p. 242) between the English of Chaucer and that of Spenser
does not exist. This must be looked at more in detail.
There is not indeed in the case of the
86 Final
Chaucer was (ii). accented final e any distinction between (ii)
and (ee) words,—I assume for the moment that the two
classes may be correctly thus designated-; and I shall
endeavour to prove that all belong to the former class.
The only word which for reasons already indicated we
might expect to find pronounced with (ee) is the noun sea,
in the Devonshire dialect say. But it had in A.S. not only
the form sæ, but also se (sé?) and seo (Bosw.); and Chaucer
seems to have retained only these. He uses the word
rhyming with he, see, tree, &c.

e in

Now Ben Jonson lays down the rule that "When e is the last letter, and soundeth, the sound is sharp, as in the French i." In Palsgrave, a century earlier, we do not find this stated as a rule, nor have we a right to expect it; but all the examples he gives are in accordance with it-bee (s.), fee, and also "dyvers other pronownes ending in e,

87

as we, me, the, he, she, and suche lyke." All of these he sounds like the French or Italian i. But was the usage the same in Chaucer's time? I claim the right to affirm, on the ground of the vis inertia of language (see §§ 6 to 9), that it was the same, unless the contrary can be proved; and the only arguments to prove the contrary are, first, the pure assumption-and a highly improbable one too-that in common English speech foreign words (such as the Latin benedicite, and the French magesté, degré, &c.) were not anglicized; and secondly, the use of one actual French word.

And this one French word, pardé-to deal with it firstPardé. in fact only confirms my conclusion, if at least the final i in French was sometimes sounded (i) (see § 13). For pardi is the common form in French, as used by Voltaire (quoted by Littré) and at the present day. Chaucer uses the word both as parde and perdy. Spenser and Shakspeare also use the latter form, Shakspeare making it rhyme with fly. I suspect it had both sounds in French (ii) and (əi). (If pardé existed in French in Chaucer's time-but I cannot find it, though I do find dé = God-we must simply consider the pronunciation as anglicized.)

And as to anglicized pronunciation, even if we did not 88 Tendency to find mention in Chaucer of French spoken anglicize foreign words.

"After the scole of Stratford atte Bowe,"

we might expect such anglicizing from the tendency continually exemplified around us to pronounce foreign words in the easiest manner. Not only do we hear Mounseer for Monsieur, and (sendziindiieek) for St. Jean d'Acre; but witness the recognised pronunciation of chagrin, bombazine, chenille, patty, bergamot, and of military terms as enfilade, calibre, &c., and of geographical names as Mexico, Saragossa, Sherry, Canton, Sedan, Paris, &c. Especially might we look for such modifications of foreign sounds in an age when there was hardly any travelling, and when there was therefore no motive for preserving them with exactness.

A curious instance of this anglicizing is found in the

§ 89] ECCLESIASTICAL PRONUNCIATION OF LATIN.

65

Rom. of the Rose (p. 164 in Bell's edition), where parcuere, i.e. par cœur, is made to rhyme with lere. The latter is probably in this case an (ee) word (as we shall presently find that it is, very exceptionally), but even then the vowel is sufficiently remote from the French cuer, coer, or queur, all of which I believe to have been merely different modes of representing the same sound as caur represents.*

I have therefore not the slightest difficulty in believing that mageste, equite, and such like words, when adopted into English, assumed the common English pronunciation of the final e, i.e. (ii) or (ż).

89 Latin no exception to the

rule.

As to Latin however Mr. Ellis seems to think himself warranted in assuming that the vowels were sounded in England in the continental mode; nor is he alone in supposing that the priesthood in this island had a traditional pronunciation of ecclesiastical Latin in which an approximation to the Italian pronunciation was maintained. So far as I can learn, this notion is simply a delusion. Two learned Catholic Doctors of my acquaintance, one the president of St.--'s College, and the other the Prior of inform me that such traditional pronunciation has no existence. The late Cardinal Wiseman endeavoured with considerable success to introduce the Italian pronunciation of Latin among English priests, but before his time there prevailed-and still largely prevails—a mongrel pronunciation, half French half English. The French element was due to the dispersion of the priests at the time of the Persecution (what we call the Reformation), when many of them took refuge in St. Omer and other places in France; but prior to the Persecution there was only the English pronunciation of Latin in this country. I have not had the opportunity of referring to

* In Stanza xxii. of the Chanson de Roland we find several words with oe in assonance with fieus (plur. of fieu = fiefs), as well as oilz = yeux, while one of these oe words, estoet il faut, is found in Mätzner's Altfr. Lieder, xxxiv.

26, in the form estuet.

=

+ Thus Mr. Payne says: "The assumption with which I commence is that the literary pronunciation of Church Latin in the thirteenth century was a tradition of ages," p. 369.

F