of the abdomen, and then in its very gradually falling inward, in prolonged expiration:—the chest making no action downwards, but merely subsiding a little, as the bulk of the lungs diminishes.

In cases of pulmonary weakness, the very opposite of this mode of expiration is generally found to be habitual. Remove the error of respiration, and the lungs will recover their strength.

Stammerers almost always have their respiration, thus, the reverse of natural. The regulation of the breathing is to them the most important, and, generally, the most difficult part of the process of cure.

VOWELS.

THE glottis produces voice: the shape of the mouth gives vowel character to the voice. Variations of musical pitch, of acuteness and gravity in the sounds, are caused, in part at least, by variations in the glottis; but all vowel varieties are caused by changes in the shape of the vocal passage. If this theory is correct, the reed vibration* ought to be capable of being modified into the different vowel sounds. It is. The mere action of the hand enclosing the open end of the reed or quill modifies the sound. sufficiently to prove the effect of similar modification on the glottal sounds. Close the hand around the quill, so as to leave a very contracted aperture for the passage of the sound, and then expand the fingers, and the vowels oo and ah will be produced. Reiterate the actions rapidly, and the hand will give out no bad imitation of a cat's wawling-w-ah-oo-w-ah-oo -w-ah-oo. The apparatus of the mouth is wonderfully calculated to effect the most minute and delicate changes with definiteness and precision. Nature must, in this case, ever be infinitely superior to the most plastic power of art. Yet art has accomplished the mechanism of the vowels in various ways, and has even effected intelligible imitations of all the elements of speech. De Kempelen constructed a speaking machine; and, recently, Mr Faber's highly ingenious speaking automaton was exhibited in this country. Mr Willis, another philosophical inquirer into the mysteries of this subject, found that the vowel sounds might be imitated by drawing out a long straight tube

* Page 14.

from the vibrating reed.

"In this experiment he arrived at a curious result: with a tube of a certain length the series of vowels

was obtained by gradually drawing it out; and if the length was increased to a certain point, a farther gradual increase produced the same sequence in an inverted order, u-c—a—e—i; a still farther increase produced a return to the first scale, and so on."

Our own experiments on the mouth corroborate this as the natural order of these vowels; but we have been led to carry out the principle of vowel sequence much farther. We have been enabled to construct a scheme which includes, in regular progression, all the vowels in our language, besides several others,—characteristic of dialects, and of the French and other languages; and to which any other peculiar formations might be added, so as to form a complete scale of natural or possible vowel sounds.

If the second of Mr Willis's series, [e=a(le)] we reasoned, can be obtained by mere elongation of the sound conductor, beyond its dimensions for the production of the first [i-ee(1)], the change from i to e will probably be gradual; and, if so, the interval between the two sounds must yield some intermediate varieties of vowel quality. It should be possible, we thought, to pass from sound to sound by such slow progression, as to exhibit vowels in the same softly blending relation that is so beautifully seen in colours, where melting shades almost imperceptibly lead the eye from one to another of the prismatic series. And this is possible.

The following simple but conclusive experiment was one of our early landmarks in the discovery of vowel principles; and it may serve to give the student a clearer idea than lengthened theorizing could, of the mechanism of vowels, and of the vowel unity of the voice as emitted from the glottis.

Prolong with open mouth the vowel ah, and, while doing so, gradually cover the mouth with the hand. At every stage of this process, the ear will recognise a change of vowel quality; the sound will in progression become

U(rn), A(11), O(re), O(we),

Oo(ze),

by the mere contraction of the external aperture, while the

* The numbers refer to our English Vowel Scheme, page 31.

internal channel of the mouth remains uniformly and equally extended.

There are two great agents in vowel modification, the lips and the tongue. The lips, by their approximation, externally contract the oral aperture; and the tongue, by its elevation towards the palate, internally diminishes the oral channel. The effect of the labial approximation is, what we have seen to result from covering the mouth with the hand, viz. modification of the vowel quality from ah to oo. The effect of the lingual approximation is, similarly to modify the sound from ah to ee.

The arrangement of the lips, then, produces one set of vowels, and that of the tongue, another; though, perhaps, few of them owe their formation to either organ independently of the other. The labial vowels require an expanded internal channel; to maintain which the tongue is slightly depressed at the root, as the labial aperture contracts; and the lingual vowels require a clear and broad external aperture; to maintain which the lips are gradually elongated as the tongue rises within the arch of the palate.

From the mutual independence of these vowel modifiers—the lips and tongue,—it will be obvious that their vowel positions may be assumed simultaneously, or variously combined. This is an important and, hitherto,—so far as we are aware,- —an unnoticed fact, to the discovery of which we were led in our experimental endeavours to find the exact formation of the vowel in sir, her, &c. and of a peculiar, close sound, which some Irish pupils gave for the vowel oo. When the principle of separate and simultaneous labial and lingual vowel formation revealed itself, these and all other tested sounds found at once their proper place in the triple Vowel scale.

Equal combinations of labial and lingual forms produce a set of vowels to which we shall give the name of labio-lingual vowels. In this class will be recognised a few familiar sounds characteristically distinct from those of the two other classes: but, with the exception of the sound in sir, her, &c. the labio-lingual class contains no genuine English vowel.

FIRST VOWEL POSITION.

The first and last of Mr Willis's series, are the close labial and lingual vowels ee(1) and oo(ze.) The approximation of the organs in forming these vowels is so close, that any further contraction of the vocal aperture creates a vibratory effect upon the tongue or

lips, and so converts the vowel ee into the articulation Y, and the vowel oo into the articulation W.

The simultaneous formation of ee and oo produces the peculiar Irish sound above mentioned, which is heard in some of the Irish dialects, instead of oo.

EE, then, is the 1st lingual vowel; oo the 1st labial vowel; and the Irish sound, combining the qualities of ee and oo, the 1st labiolingual vowel.

SECOND VOWEL POSITION.

The tongue a little depressed from its elevated position at ee(1,) gives a vowel intermediate in form and effect to ee(1,) and a(le). This is the sound of i as in ill, is, it, &c. which is therefore the 2nd lingual vowel.

The lips slightly separated from their close position at oo(ze,) produce a sound intermediate to oo(ze) and o(ld), which is heard in some English dialects instead of o(ld); as when a Lancashireman says, "Put some coal" (almost, but not quite, cool)" on the fire." This, then, is the 2nd labial vowel.

These two formations combined, produce an appreciably different sound from the first labio-lingual vowel-intermediate to it, and the next vowel u(ne.) This is the 2nd labio-lingual vowel.

THIRD VOWEL POSITION.

A further slight enlargement of the oral apertures, by the depression of the tongue, and separation of the lips, produces, by the former action a(le,) the 3rd lingual, and by the latter o(ld,) the 3rd labial vowel.

The union of these formations gives the French sound of u, as in une, būt, lū, &c., which is therefore the 3rd labio-lingual vowel.

It is to be remarked of the two correspondent sounds a(le) and o(ld), as a curious peculiarity, that in English usage they are both diphthongally terminated with the close vowel of their respective classes, a with e, and o with oo. The omission of this final element of these beautiful vowels is a marked provincialism.

FOURTH VOWEL POSITION.

A farther slight opening of the vowel apertures from the 3rd lingual position, produces a sound heard in Scotland instead of the 2nd lingual, in such words as ill, in, sit, &c.; and, from the labial formation, produces the monophthongal sound of o as heard. in English before r, in such words as ore, four, soar, &c.

Ꭰ

The labio-lingual vowel resulting from the combination of these forms, occurs as a provincial and rustic peculiarity in England, instead of the more open vowel correctly heard in such words as sir, her, &c.

FIFTH VOWEL POSITION.

An increased depression of the tongue gives the formation of the sound heard in e(re,) ell, end, &c., the 5th lingual vowel: and a correspondent increase of the labial aperture from o(re) gives the vowel heard in all, saw, on, &c.,-the 5th labial formation.

From the combination of these positions results the vowel represented by eu in French, and by oe and ö, in German.

SIXTH VOWEL POSITION.

The next English degree of openness produces, in the lingual series, the sound heard in an, at, &c.; and in the labial series, a correspondent enlargement, produces the vowel uh as it is pronounced in Scotland, in such words as up, urge, &c.

The combination of these positions gives the peculiar English sound heard in sir, her, earn, &c.

We before observed, that few of the vowels owe their formation to labial or lingual position alone; there is for every vowel a necessary arrangement of the whole mouth: but the preceding sounds are formed by so evident a proportion of the one over the other, that their being called respectively labial or lingual vowels, will be perfectly intelligible. The sounds which follow, however, are dependent chiefly on the internal arrangement of the mouth, and do not so obviously fall under the same classification. The lips are well spread and open, and the tongue well depressed, so that the changes of organic arrangement are less manifest; but the vowels are all in regular progression, from close labial and close lingual forms, and do, therefore, truly belong to one or other of these classes. Positions intermediate to any two, likewise, may still be formed, though, from the necessarily slight differences between their effects, ears untrained to very accurate observation, may think them, in their separate utterance, "distinctions without difference." On such minute distinctions, however, often depends the very important difference between a cultivated speaker and an uneducated or a provincial one.