The Physiological Scale of Vowels.

If we recall the process by which the semi-vowels were formed in the three principal classes, and if, instead of stopping the vocal sound by means of that slight remnant of consonantal contact or convergence, which characterized the formation of the semi-vowels, we allow the full volume of breath to pass over the point of contact and there to vibrate and sound, we get three pure vowel sounds, guttural, palatal, and dental, which can best be expressed by the Italian A, I, U, as heard in psalm, ravine, flute.

Formation of the Labial Vowel.

Let us pronounce the labial semi-vowel, the English w in woe, and, instead of stopping the vocal sound as it approaches the labial point of contact, emit it freely through the rounded aperture of the lips, and we have the vowel u. Here also the experiment of the candle will elucidate the process that takes place, but of which we are hardly conscious. The mere semi-vowel w, not followed by any vowel, should not produce any disturbance in the flame; at least not more than might be occasioned by the motion of the lips, which is the same for all consonants. The labial flatus, f, on the contrary, will disturb the flame considerably, and the vowel u may extinguish it.

Formation of the Palatal Vowel.

The same process which changes w into u, changes the guttural semi-vowel 'h into a, and the palatal semi-vowel y into i. Let us pronounce the y in yea without any vowel after it, and it only requires the removal of that stoppage of sound which takes place between tongue and palate, in order to allow the vowel i, as in ravine, to be heard distinctly.

Formation of the Guttural Vowel.

Let us pronounce the guttural semi-vowel as heard in the Dutch dag or the Hebrew 'hain, and, if we try to change this semi-vowel gradually into the vowel a, we feel that what we effect is merely the removal of that stoppage which in the formation of the semi-vowel takes place at the very point of guttural contact.

The vowels, as was said before, are formed by the voice modified, but not interrupted, by the various positions of the tongue and the lips. "Their differences depend," as Professor Wheatstone adds, "on the proportions between the aperture of the lips and the internal cavity of the mouth, which is altered by the different elevations of the tongue."

Succession of Vowels, natural and artificial.

The organic succession of vowel sounds is the same as for consonants,—guttural, palatal, labial, a, i, u. The succession of vowel sounds produced by the gradual lengthening of a cylindrical tube joined to a reed organ-pipe, as described by Professor Willis^*, is an interesting experiment as to the scale of vowels in the abstract. It gives, or, at least, is reported to give,

But as these pipes are round and regular, while the construction of the pipe formed by larynx, throat, palate, jaws, and lips is not, the succession of vowels given by these pipes cannot be expected to correspond with the local succession of vowels as formed by the organs of speech.

Kempelen states that if we pay attention to the successive contraction of the throat only, we shall find, indeed, that the aperture of the throat is smallest if we pronounce the Italian i, and that it gets gradually larger as we pronounce e, a, o, u; while if we pay attention to the successive contraction of the lips, which is quite as essential

^* Transactions of the Cambridge Philosophical Society, vol. iii. paper 10. 1828-29.

page 19 to the formation of the vowels as the contraction of the throat, the scale of vowels is a different one. Here the aperture of the lips is largest if we pronounce the a; and it gradually decreases as we go on to the e, i, o, and u.

It has been remarked by Professor Purkinje, that the conditions for the formation of some of the vowels, particularly of a and e, as heard in far and name, have not been quite correctly stated by Kempelen. The production of both these sounds depends principally on the form of the cavity of the throat between the root of the tongue and the larynx; in both cases this space is large, but largest in the pronunciation of e. The size of the opening of the mouth is the same in the two cases; not different, as Kempelen states. The position which he ascribes to the lips in pronouncing o is also unnecessary. ^*

The experiments of Professor Willis show that, if we look on the instrument by which the vowels are formed as a vibrating membranous tongue, with one tube prefixed, and another added below the tongue, the shortest length of the tube gives i; the longest, u; and an intermediate one, a. But as the human organ of speech is not a regular tube, we must insist on this, that in the mouth the shortest length is indicated by the point of palatal contact, the longest by the point of labial, and the intermediate by the point of guttural contact; and that here, by the simultaneous operation of the guttural and labial aperture, the vowels i, u, and a are formed.

The Lingual and Dental Vowels.

Besides the three vowels struck at the guttural, palatal, and labial points of contact, the Sanskrit, in strict analogy, forms two peculiar vowels as modifications of the lingual and dental semi-vowels. R and L, subjected to the same process which changes 'h into a, y into

^* Sec J. Müller, Elements of Physiology, p. 1047.

page 20 i, and w into u, become ri, li, or rĕ and lĕ. At least these sounds ri and li, approach as near to the original value of the Indian vowels as with our alphabet we can express it. According to their origin, they may be described as r and I opened and vocalised.

Unmodified Vowels.

If we attempt in singing to pronounce no particular vowel, we still hear the vowel-sound of the Italian a. This vowel expresses the quality of the musical vibrations emitted from the human larynx and naturally modified by a reverberation of the palate. Hut if we arrest the vibrations before they pass the guttural point of contact—if, either in a whispered or a vocalised shape, we emit the voice without allowing it to strike against any part of the mouth—we hear the unmodified and primitive sound as in but, bird, lull. It is the sound which, in Professor Willis's experiments, "seems to be the natural vowel of the reed," or, according to Mr. Ellis, "the voice in its least modified form." We hear it also if we take the larynx of a dead body, and blow through it while compressing the chordæ vocales.

In these experiments it is impossible to distinguish more than one sound; and most people admit but one unmodified vowel in English. According to Sir John Herschell, there is no difference in the vowels of the words spurt, assert, dirt, virtue, dove, double, blood. Mr. Ellis considers the u in cur as the corresponding long vowel. Other writers, however, as Sheridan and Smart, distinguish between the sounds of bird and work, of whirl'd and world; and in some languages this difference requires to be expressed. It is a very delicate difference, but may be accounted for by a slight palatal and labial pressure, by which this obscure sound is affected after having escaped the guttural reverberation.

In English almost every vowel is liable to be absorbed by this obscure sound; as beggar, offer, bird, work, but. It is sometimes pronounced between two consonants, though not expressed in writing; as el-m, mar-sh, schis-m, rhyth-m. Here it is the breath inherent in continuous consonants. In French it is the e muet, as in entendre, Londres. In German it is doubtful whether the same sound exists at all, though I think it may be heard occasionally in such words as leber, leben.

Quantity of Vowels.

All vowels may be short or long, with the exception of the unmodified breathing (Rapp's "Urlaut"), which is always short. The sound of the long a we have in psalm, messa (It.); short, in Sam.

The sound of the long i we have in neat, Italia; short, in knit.

The sound of the long u we have in fool, usarono (It.); short, in full^*

The sound of ĕ we have in bird.

The sound of ŏ we have in work.