106. The O offers less to remark on than the other vowels. It has been the most stable member of our vowel-system, and that in which we are most in harmony with other nations.

107. Of the U, it is very obscure what has led to its name. The pronunciation of the u as yew can hardly be of EastAnglian growth, though natives of that province sometimes bring in the sound unexpectedly. When they utter the words rule, truth, Jerusalem, with energy, they have been observed to convert them into ryule, tryewth, Jeryewsalem. I have seen it somewhere suggested that possibly this peculiar vowel-sound has risen out of a distorted effort to imitate the inimitable French U. There is perhaps something in this idea. A very peculiar u exists in Devonshire, one which is near the French, and one which would seem to have been inherited from British pronunciation, if we may judge from its proximity to the Welsh U. Now this Devonshire u is not at all yew, but it has been often so reported of, and tourists tell how in that strange land they heard the natives say byewts, myewn, for boots, moon. I do not believe they ever heard any such thing, and I take their evidence to be good only to shew that there is some point of contact between the French u and the yew sound, at least on the ear. Thus the idea that our yew grew out of the French u is plausible. But I do not think it to be correctly stated in this form, and for the following reason:-the sound recurs in many independent and external places. The Dutch nieuw indicates by its orthography the same sound as our new. The Danish lys (light) is pronounced lyews, and in Swedish it is phonetically so written, namely ljus. The tree which in English is called yew was in Saxon written iw, from which we gather that the pronunciation is unaltered. These instances seem to shew that the sound we are treating of was an anciently inherited one, and if French influence had

I

anything to do with putting it on our u, it only caused the extension of a sound already domestic and familiar.

To so great a length have I pursued this subject of the naming of our vowels, because it is in fact a most exceptional and insular phenomenon. As a criterion of the whole case we might refer to the designations of the five vowels in French or German, and the reasonableness of those designations. If this were done, the result would be something as follows. The French and Germans have named the vowels, but the English have nick-named them. When a man is called a king or a servant, he is characterised by what may properly be called a name. But if we call him Longshanks or Peach-blossom, we nick-name him. And this is analogous to what we have done with the vowels. They have been named, not after their proper functions or chief characteristics, but after some anomaly or adventitious oddity which has attracted a too pointed attention.

Of the Vowel Functions.

108. The tendency of observations like the above, arising out of the arbitrary naming of our vowels, is to create in the mind an impulse such as that which is attributed to the etymologists of a past age, to put the vowels aside as if they were hopelessly beyond the reach of scientific method. Each vowel sign has such a variety of sounds in English, and each sound has such a variety of vowel signs, and these so cross each other's track, that anything like disentanglement and orderly arrangement might well be despaired of, if there were no help to be found beyond the limits of the single language. But much of that which is arbitrary or accidental may be eliminated by the process of comparing two dialects together, and then a third with the results of

the first comparison, and so on; sifting each time the net product to a clearer expression; till we at length reach the conclusion that a phonology or science of vocal sounds is possible. It is found that there are three principal sounds, which are those of ‘a,' ‘i,' ‘u'—that is to say, not according to the value of these signs in the English naming, Ae, Igh, Yew, but according to the value which they most commonly represent in European languages, and which we may spell thus, ah, ee, o0. It is the sound of 'a' in arm, father, of ‘i' in dig, and of 'u' in full. It will be convenient to distinguish these signs by quotation marks, when we use them for the true and principal sounds. That these are the cardinal vowels can be shewn in two ways.

109. Either we may observe the organs of speech, or we may examine those languages in which the vowel system is most robust and symmetrical. There is one dialect of our family which is distinguished for such a vocalism, and that is the Mosogothic. In this dialect, all the vocalic and diphthongal sounds are so regularly derivable from these three, that we are compelled to regard the 'a,' 'i' and ‘u' as fundamental, at least for that particular language. Other languages are found to contribute, some more some less, to the general adoption of this trio of vowel-sounds as the basis of phonology.

A like result is obtained by physical analysis of the sounds, and the acoustical study of the organs of speech. Experiments of exquisite ingenuity and delicacy have been made by Helmholz and Koenig on the musical contents of the several vowels, and by these it has been established, that U is, musically speaking, the lowest, I the highest, and A the central of all the vowels. Thus these vowels appear by a novel kind of evidence as the three Cardinal Vowels. (122.)

A.

110. Of this central vowel, Mr. Hullah says:'On one vowel only is the timbre of the human voice to be heard in its highest perfection—the vowel a pronounced as in the English word father. And again: Recent physiological researches have justified the choice of aa not merely as the vowel on which the voice is heard to the greatest advantage, but also as that on which, with a view to its improvement, it should be most exercised1.' There is no doubt that the a in Saxon writing represented this 'a' sound, sometimes short as in van, sometimes long as in father. But this 'a' had already in Saxon times lost much of the ground it once occupied, especially the short ‘a.' And many examples which then existed are now lost. (We will consider the losses first, and the compensations afterwards. 112.)

The single instance of -as, the plural form of an increasing group of substantives, presents a great amount of loss in regard to this principal vowel-sound. The 'a' is lost in every one of those instances; and words which were written dagas, endas, fixas, pathas, smithas, stanas, are now written with a toneless e as in fishes, or a merely orthographic e as in stones; or else, and this is the commonest result, it has left no trace behind, as in smiths, But then it is in flexional terminations

days, ends, paths.

that the vowels degenerate most rapidly, and we must not hastily conclude that the 'a' is becoming a stranger to our language, as some phonologists seem almost to do, when they speak of this cardinal sound as 'the Italian A.'

:

111. Words in which the Saxon 'a' is fully retained addle, adesa adze, ancra anchor, and, anfilt anvil, ask, assa ass, awul awl, alr alder, apul apple, blac black, brand

The Cultivation of the Speaking Voice, Clarendon Press Series, ch. vii.

(fire-), candel candle, cat, crabbe crab, fann fan (vannus), gader gather, gang wæg gang-way, ganra gander, garleac garlic, galga gallow, halgian hallow, hand, lamb, land, malwe mallow, man, panne pan, plant, ramm ram, sadol saddle, sand, span (subst.), stand, swalewe swallow, tan, wann wan (colour).

Words in which the character is preserved but the sound altered to ae:-apa ape, cara care, cran crane, cafer chafer, capun capon, cradel cradle, faran fare, hara hare, nihtscada nightshade, raca rake, sala sale, scamu shame, spada spade, sam same, tam tame, wacian wake.

Words in which 'a' has become o:-camb comb. clað cloth, fald fold, gast ghost, halig holy, lang long, mapu moth, rap rope, sang song, strang strong, tacen token, tange tongs.

Words in which it has become o with subscript e:ban bone, dran drone, ham home, lar lore, mara more, rah roe, rap rope, sar sore, sla sloe, stan stone, spaca spoke (of a wheel). The Saxon ma (more) became mo and moe.

Words in which the Saxon 'a' has become oa:-âc oak, âð oath, âr oar, bât boat, brâd broad, gâd goad, gât goat, hâr hoar, hlaf loaf, lâd load, lâm loam, râd road, wad woad. The original 'a' in all these cases was long; but the Saxon long 'a' did not always produce English oa, thus bân bone, stân stone.

In one instance this oa has drawn in a cockney r, namely hâs hoarse. In Devonshire the true analogy is preserved, and this word is pronounced hoase or hoaze.

112. As we have thus seen that the Saxon 'a' has broken and dissipated itself into a variety of modifications, so now on the other hand we must see what compensation there has been that this chief vowel should not perish out of the language. We shall find that many words which in Saxon. had not 'a,' but some weaker and softer vowel, have now by