also can be united with various other consonants, without any vowel at all; as in vz, dhr; ls, lz; rs, rz, &c., &c.; and even w consonant, and y consonant, and h, which are the three aspirates, can be sounded alone, and without a vowel, though in words they always precede a vowel.(6)

“The truth is, that the possession of a vocal sound is not the characteristic distinction of a vowel, for we can articulate vowels in whispering as well as consonants, though we then use only non-vocal breath, and not vocalized breath, or voice, at all.(7) Each vowel, as it is called, has, therefore, a specific sound of its own, arising from the mere expiration of breath, though not vocalized, through the peculiar mechanism in the mouth, used as its articulation; and hence, in this point of view, there is no real distinction whatever between vowels and consonants, both being, in whispering, affections of mere breath; but in common speaking, where voice is always used, there is this distinction between vowels and consonants, that, in the former, we use a distinct musical sound in the larynx along with each of them, and emit it through their peculiar mechanisms, while, even in the vocal consonants, there is only a non-musical murmur of the voice."

We have italicized the last few lines, for the purpose of taking a slight exception. The sound in the larynx is not necessarily musical; in speaking it is generally not musical (in the sense given to this word by Sir J. Herschell, see suprà, p. 12, last par.); and, secondly, that the spoken consonants are not accompanied merely by a "non-musical murmur," but by a clear vocal sound, or by none at all.

We have thus glanced generally over our subject, and pointed out the peculiar difficulties under which an analysis of spoken sounds necessarily labours. As regards the vowels, they may be safely said to be entirely subjective; and that any analysis that can be offered, can only be considered as that of the author's own sensations, which may differ from those

(6.) The aspirate is found without a vowel in Sanscrit; and the weakened consonants (see infrà, chap. 8, sect. 4) seem to imply to imply the existence of y (or yh, chap. 6, sect. 1) not followed by a vowel.

66

(7.) The absoluteness of this assertion may be doubted. That vowels can be distinguished in what is commonly termed "whispering," cannot be denied; but ? do we mean, in common language, by this word the total absence of vocalized breath. We think that it will be usually found that a 'whisper" contains a very perceptible, although small, amount of vocalized breath. We must be more strict in our language; in this work we have employed a whisper" for the absolutely unvocalized breath. But even this whisper will be modified in a similar manner to that in which the voice is modified, when passing through the peculiar modifying cavities. Hence differences may be perceived corresponding to, but far from being identical with, the vowel qualities.

of the readers. It will not, therefore, follow that the analysis is false, but simply that the subject-matter is different in the two cases. The difference will not be found to be very great, however, in any case. But, ? how is the idea of the sounds and their modifications to be conveyed to the reader. This is another source of difficulty. The only means which we possess is to present the reader with a certain number of words (key words, as they may be called), in different languages, which, as we pronounce them, produce in us the sensations which we shall analyse. ? How can we guarantee that our readers shall pronounce them in such a manner (we do not say as to experience perfectly similar sensations, for of that we could never be sure, but) that we, if we heard them, should experience the same sensations as we did when we pronounced the words. We can only hope that they will pronounce them nearly in the same manner, and we shall try to select such words as are (according to our experience) most generally pronounced alike by the majority of speakers. These fears principally regard the vowels, for the distinctions between the consonants are less fine; these do not, like the vowels, glide into one another by almost imperceptible degrees. We shall recur to this subject, and consider it at due length in Part II., Chap. 1; and shall now proceed to the detailed examination of the elements of speech.

CHAPTER 5.-ON THE VOWELS IN DETAIL.

SECTION 1.-On the Long Vowels.

IN Prof. Willis's experiments, already given (Chap. 3), we are furnished with something approaching to a standard for measuring vowel sounds. The shortest length of his pipes gives e in me, and the longest, previous to the recurrence of the vowels, oo in moo; while an intermediate length gave a in mar. We may express this by saying, that the latter sound, a, is intermediate to the two former; and generally when we say that any sound is intermediate to two other sounds, we shall mean that they require pipes of intermediate lengths to generate them. We shall represent these sounds by ī, ū, ā, respectively. These sounds, considered in relation to the human organs of speech, correspond respectively to the greatest diminution of the aperture of the throat by means of the tongue, to nearly the greatest diminution of the external orifice of the mouth by means of the lips, and to nearly the greatest opening of the mouth and throat. These sounds are, in different languages, designated in different manners;

but the signs we have chosen to represent them may be considered as the normal forms assigned to them in all languages for which the Roman alphabet characters are used.

Intermediate to i (which corresponds to 38 inches on Prof. Willis's pipe, see p 35, suprà) and ā (2·2 inches), lies ē (1 inch), as represented by ay in may. Intermediate to a and u, is o (4.7 inches), as represented by ow in mow. These five sounds, in the order ī, ē, ā, o, ū, constitute what are commonly called the five vowels, to which number the vowel sounds are usually supposed to be restricted.

The vowels ē, o, are apt to be pronounced impurely in England. When the sound of è is lengthened, the tongue is liable to approach gradually, and almost imperceptibly, to the position requisite for the pronunciation of i; similarly, o, upon being lengthened, degenerates into u. By care and trial, this defective mode of speaking can be remedied; and we particularly mention it, because, if our readers suppose that we are speaking of such impure sounds, they will naturally disagree with much that we shall have to say.

Between a and o, we have ō (3.05, or 3.8 inches), as represented by aw in maw; and, according to Prof. Willis, at some point between o and ū, is another vowel, which we will call e. Of this vowel, however, which corresponds to y in myrrh, he remarks, that it seems to be the natural vowel of the reed," (suprà, p. 34). It is, we believe, the voice in its least modified form, being produced while the speaking apparatus is in its most natural condition. If we pronounce the vowels ī, ē, ā, ō, o, ū, in this order, we see clearly that they follow one another in a natural manner. ? Where are we, then, to insert e. According to Prof. Willis, between o and u; but, on repeating ī, ē, ā, ō, o, e, ū, there is a clear break in the continuity of change in the position of the vocal organs, the mouth being more contracted for o than for e, being most contracted for i, gradually widening up to ō, and then as gradually closing towards u. Various places have been proposed for this vowel. Dr. Orpen (1) places it between ā and ō; others insist that its proper place is between ē and a. To ourselves, the most natural place appears to be between ō and o; so that the series of seven vowels will be i, ē, ā, ō, e̟, o̟, ū.

For the circumstance of e occurring in the organ-pipe between o and ū, or rather just before u is heard, we have no means of accounting; but the following considerations serve to reconcile the other diversities of opinion. The various vowel-modifications of the voice are divisible into three classes, the contracted guttural, open guttural, and contracted labial.

(1.) Pestalozzian Primer, pp. 77-79.

We will designate these by the old English letters, í, a, u, respectively. Then to the class í, belong i, ē; to a, belong ā, ō, v; and to u, o̟, ū. The primary or typical vowel in í, is ī; in a, e; and in u, ū. The nonprimary vowels in each class have a tendency, upon prolongation, to pass into the typical vowel: this we have already remarked for ē and o; and the reader will discover that the same is true for a and ō, each of which, when prolonged, have a tendency to terminate in . Consequently, it is difficult to determine the relative position of e with regard to ā and ō, because it is the chief element of each of them. As, however, e naturally follows when a and ō are prolonged, we think it best placed after them, and we have, therefore, adopted this arrangement.

This tendency of vowel in each class to terminate in the typical vowels of that class, has been considered so important by Mr. Cull, that, in his article STAMMER (Penny Cyclopædia), he grounds a division of the vowels upon it, classing some as monophthongal, (2) and others as diphthongal. The monophthongal he considers to be those expressed by the italic letters in end, eel, her, in, and ooze, of which the first and fourth will be considered hereafter; while the second, third, and fifth correspond to our i, ?, and ū, respectively. The other vowels he calls diphthongs. We shall recur to this classification in a future chapter.

It may be mentioned that, in the Sanscrit language, ē, o are considered as diphthongs compounded of a (?) and ī, and of ā (??) and ù respectively.(3) Bopp compares this with the French apparent diphthongs, ai, auê, o. We may compare our digraphs ai (in pain)=ē, and au (in maul)=ō.

Thus far we have considered the vowels as acoustical phenomena, quite independently of the means by which they are produced by the human organs of voice. We subjoin the following particular account of the mechanisms by which they are produced in this case, extracted from Dr. Orpen's instructions for teaching the deaf to speak:—(4)

"In pronouncing the sound (ā), we should make him (the deaf pupil) observe that our mouth is pretty wide open; that our tongue lies flat in

(2.) From μóvos (monos, or monos, in M. G.), single, and płóyyos (fthongos) voice, comes monophthongal (monophthongǝl), single-voiced, an epithet corresponding to our "simple vowels." Diphthongal is from dis (dis, or dhis, in M. G.), twice, and φθόγγος.

a

(3.) Bopp, Vergleichende Grammatik, sec. 2. He calls the first element " short a," which we know, from Sir W. Jones (Works, vol. iii., p. 270), means 66 a short " (or ǝ) in Sanscrit.

(4.) Pestalozzian Primer, pp. 145-150.

the bottom of our mouth, neither retracted, nor pushed forward, neither dilated, nor contracted, but with its tip just behind the back of the lower front teeth, and its sides just touching the insides of the lower side teeth, at both sides. Thus we see that, by the under jaw being lowered, to open the mouth, the two rows of front teeth are separated from each other about three-fourths of an inch, and the upper surface of the tongue, which itself descends, of course, with the under jaw, is distant from the arch of the hard palate about an inch. While these parts of the mouth are in this position, we cause vocalized breath, or voice, to be formed in the larynx, by the vibration of its sides from the outward current of air, and emit it through this mechanism, which thus articulates the sound of ā.(5) And as this sound is not either as bass, as the vowel sound ō, nor as treble as that of i, the larynx is not either much depressed towards the chest, or much elevated towards the throat; the former, viz., depression of the larynx, being essential to a bass sound, and the latter, viz., its elevation, being essential to a treble sound.

"The mechanism of ē is produced by raising the jaw, so as to make the two rows of front teeth come within half an inch of each other, and the upper surface of the tongue, of course, come within about three quarters of an inch of the arch of the hard palate, and by emitting vocalized breath, or voice, through this mechanism; the larynx being a little more raised towards the throat, as the sound is a little more treble, than in ā.

"In i, the lower jaw is so much raised, that the two rows of front teeth are only about a quarter of an inch separate, and the tongue is also, of course, so much raised, that its upper surface only leaves a shallow channel, from back to front, about a quarter of an inch deep, between it and the arch of the hard palate. The larynx is also raised, as far as it can be, towards the throat, as this is the most treble of all the vowels, and the vocalized breath, now formed in the larynx, is emitted through the mechanism of the mouth just described.

"The other trio of vowels, ō, o, u, are all, in some measure, essentially bass sounds, and, therefore, the larynx is depressed in them all towards the chest, which both shortens and widens the windpipe; and also, which is its chief object, enlarges, backwards, the cavity of the mouth, into which the sound formed in the larynx enters.

"In all these sounds, too, the tongue is drawn back in the mouth, which must be the case whenever the larynx, to which the root of the tongue is more or less attached, is drawn down as just described; but the chief object of it is, to increase still more the cavity of the mouth, in which the sound is revcberated and made bass.

(5.) Not to confuse the reader, we use our own symbols throughout.