4. If we compare "pit, pet, pat, pot, put (substantive), put (verb)," we shall find that they all contain the same consonants, but have each a different vowel, and each vowel is stopped—a fact which we may determine by endeavouring to prolong any one of them, and failing of doing so. It is this impossibility of separating the vowel from its consonant which has rendered the preceding analysis necessary. The stopped vowel is like a chemical substance, the existence of which we can feel and demonstrate, but which cannot be exhibited except in conjunction with some other substance. Thus the foregoing experiments have proved the existence of vowel sounds in conjunction with consonants, essentially differing from the vowel sounds previously considered, and have shewn that their appearance does not depend upon any particular consonant, because the same vowel may be produced before different consonants, and different vowels before the same consonant. Hence although we cannot exhibit the stopped vowel except in conjunction with a consonant, we have, as we think, satisfactorily shewn its necessarily independent existence.

What is the exact relation between the full and stopped vowels, it is not very easy to state; the feeling of different persons on this point differs considerably, and there does not appear to be any objective standard to which we can refer.(7) In the following table we give the long, short, and

(7.) Some writers assert that three of our stopped vowels (those marked Ĭ, ĕ, g) have no corresponding long vowel whatever, but are vowels incapable of fulness— i. e., of being pronounced independently of a following consonant. Mr. Knowles only allows of nine distinct and separate vowels in our language, corresponding to our ō, ā, ē, i, o, u, ĕ, I, g, admitting ŏ, ă, ů, not as distinct vowels, but as the short (stopped) sounds of ō, ā, ū, (foreigners would probably disagree with him as to ă, at any rate, without further explanation), and his scheme is, therefore, peculiar in leaving e, i, o without corresponding stopped, and ě, I, g without corresponding long vowels; whereas, in our opinion, o is the only vowel which does not occur in a double (or rather triple) form in our language. The following is the description Mr. Knowles gives of the pronunciation of ě, I, g (Exposition of the Principles of Speech, p. 6, col. 2, prefixed to his Pronouncing Dictionary; we have, as usual, translated his notation) ::-"In uttering the sounds ě, I, g, he (a teacher, or critic, who looks into a child's mouth while it is repeating these sounds deliberately, or places himself before a magnifying glass and looks into his own mouth) will perceive that the sound, represented by ě, is produced by a movement of the larynx and glottis, which throws out a short and quick percussion of the voice upon the tongue and palate, which (? the tongue) is directed towards the front of the mouth; that the extremity of the tongue is pointed strongly down to the lower palate, or gums; that it is considerably raised in the mouth, but not drawn back; that it rises towards the back part, so as to receive the impulse of the voice, and throw it forward as I have said. In uttering the sound represented by I, he will perceive that the position of the tongue is the same as in the former sound; but that the tongue raises itself more and throws the sound to the roof of the mouth. In uttering the sound represented by g, he will perceive that the

stopped vowels in three columns, endeavouring as far as possible to give such examples as contain the three kinds of vowel in conjunction with the same consonant, in order that the reader may be better able to compare the several sounds together and thereby form his own conclusions. We have, however, of course adopted our own arrangement as the basis of our notation, which is shewn in the three last columns of the table, corresponding respectively with the three first.

Combined Table of Long, Short, and Stopped Vowels.

poddymac' (pl.) (8) pomocy (pl.) Towarszystwo (pl.)

We will now examine these vowels in order. First, ĭ is considered by almost every writer to be the corresponding stopped vowel to i. Mr. Knowles, as we have seen (Note 7), holds it to be an independent sound,

end or point of the tongue is looser in the mouth than it was in the former sounds, and drawn a little further back from the lower gums, though it rests upon the lower palate; that it is drawn back so much as to leave a considerable space between it and the palate; and that, by a strong movement, it raises the back part of it up, so as to throw the voice against the extreme back part and sides of the palate."

(8.) All words are English, except otherwise specified. The names of languages are, in this table, contracted thus: pl. Polish, fr. French, it. Italian, ger. German. (9.) We think that these sounds occur in the English words "fairy, ferry.” (10.) When a space is left in this manner, it shews that we are not acquainted with any word exemplifying the sound, which sound is, nevertheless, perfectly conceivable.

(11.) This belongs to a class of words (balm, calm, palm, &c.) which some persons pronounce with à, but which we pronounce (as most persons that we have heard speak English do) with the same vowel as a in father.

(12.) If be recognized as distinct from o, its three signs will be 9, 9, ọ.

having no connection with i; and Dr. Young, in the passage lately quoted, seems to agree with him. Our own observations, added to the testimony of several eminent phoneticians,(13) lead us to pronounce unhesitatingly that ĭ is the stopped sound of i.

Concerning ì we cannot feel very sure, as we have had but very slender opportunities of receiving instruction in Polish pronunciation; to our ears it partakes much of the nature of ě.

According to Rapp,(14) our word "debt" has the vowel è and not ě; in this case we are unable to conceive what sound he attributes to ě (in his notation é), but the difference appears to be so essential, that he invariably distinguishes the two cases, and even grounds some distinctions in German dialectic pronunciations upon them. We seem to hear a broader sound in the Italian " ben," which, we believe, contains the true è. Latham says that he has convinced himself that e in bed, "although both broad and slender, is incapable of becoming independent" (i. e., full).(15) As, in our opinion, ê occurs before r in such words as fairy=fêri, it would appear that è must be the vowel in ferry=fèri; but of this we are yet uncertain. Great doubts may be entertained concerning the propriety of classing ǎ with a. Rapp makes it, in his scheme of English vowels, (16) correspond with our ê. The sound is peculiar to the English language, in which it replaces the foreign à. As a is less broad than â, so we have thought it best to place the slender ǎ to ā, and leave the comparatively broad à to â. Foreigners are apt to confuse ǎ either with à or è.

The vowel à is not generally recognised as existing in our language, but some theorists require it in such words as after, ask, laugh, aunt, path, balm, &c., and generally before the continuous sibilants, writing such words-aftǝr, àsk, làf, ànt, pàth, bàm, &c. They certainly approach this sound much more nearly than that of ǎ, as ǎfter, åsk, ănt, păth, which some other theorists propose, and which is common among the upper classes in Yorkshire. If we had only to choose between the vowel characters, ā and ǎ, we should certainly select the former as the representative of this sound, in preference to the latter, although a is by some considered broad and vulgar; and Judge Haliburton, by writing "larf" for

(13.) Latham, On the English Language, p. 112; Rapp, Physiologie der Sprache, vol. 3, pp. 164, 174; Bp. Wilkins, Essay on a Real Character, p. 363; to whom might be added many others.

(14.) Phys. d. Spr., v. 3, p. 174, compared with vol. 1, p. 23. Klopstock makes ĕ the stopped sound of ē, and does not admit of any stopped value of ê; see suprà, p. 62, note 6.

(15.) Engl. Lang., p. 112.

(16.) Phys. d. Spr., v. 3, p. 174.

"laugh," in his Sam Slick's conversation, clearly wishes to point out the pronunciation "laf" in contradistinction either to làf or laf. Theoretically, ǎ occurs before an r which is not followed by a vowel; but, practically, its place is always usurped by a, at any rate in our language.

There cannot be much doubt as to the correspondence of ŏ with ō; but it is curious that while ŏ is of general occurrence, ō is only found, as a recognized sound, in English and Swedish. This ŏ is, in most languages, used as the stopped vowel of o, with which it has no real connection. The Italian ò does not differ much from it, being only somewhat broader. We have before stated (§ 1) that, to us, many persons appear to pronounce croft, loft, cross, &c., as crôft, lôft, crôs; at times, however, this sound (which is very uncertain in the mouths of English persons) seems rather to approach the stopped sound cròft, lòft, cròs. For practical purposes, the full vowel, crōft, lōft, crōs, appears to be the best notation. Among all European languages, the English is peculiar in possessing a stopped sound of ǝ, namely, g. Foreigners are continually confounding it with ŏ, or, more properly, o, to which it bears a strong resemblance. The difference between and o may be well felt from hearing a foreigner's mispronunciation of some such word as bun, for which he will almost Invariably (at any rate, during his first trials) say bọn, instead of ben.(17) It is important that readers should learn to distinguish with accuracy between g and, for, strange to say, most orthoëpists have confounded them. Knowles, for example, marks cur to be pronounced with the vowel in but, and curry to have the same vowel in its first syllable. Dr. Orpen says that is always short (stopped) in English; and Bishop Wilkins states the same. Such words as cur, curry, occur, occurrence=ker, keri, ŏker, ŏkerǝns, should, however, have led them to a different conclusion; they should have shown them that the long effect in such words as cur, was not entirely due to the disturbing action of the following r, as the same long effect is not produced in curry, where r also follows. The only doubt that can be raised is, as we mentioned in the last section, as to whether is the pure long vowel corresponding phonetically (as it does

(17.) We find M. Volney making this error (L' Alfabet Européen, p. 33; Phonotypic Journal, vol. 3, p. 110). He gives, as examples of "clair ou bref, petit o" (clear or short, little o), in French, odorat, hotte (d'osier), molle (cire), sol; in English, rod, gut, nut, cut, lull; in German, ochs, oft; which words we believe to be pronounced respectively, ŏdorā, Họt, mọl, sŏl, röd, gęt, nyt, kęt, lęl, öks, ŏft. But this is not the only instance of M. Volney's defective ear, for, just above, he confuses ǎ with à, and â with ō. It should be observed that, in the first five editions of Mr. Pitman's Phonography, e was considered as the stopped sound of o, or as identical with o; this error has been corrected in the 6th edition.

practically) with the stopped vowel and the short vowel ǝ, which two undoubtedly belong to one another.

The stopped sound does not differ much from g, and is, perhaps, the nearest approach to this sound which the generality of Germans are able to effect.

The stopped sound o does not occur in English-at any rate, according to the recognized pronunciation; but we have thought that we sometimes heard the word wholly pronounced Holi, instead of holli (with the double, or dinounced 1), to distinguish it from Holi (holy).

No doubt can exist as to the relation of ŭ to ū.

The Germans have a great habit of confounding û with ī, and ù with ĭ;(18) this will serve to give an Englishman (who confounds them with ū and ŭ respectively) some clue to the pronunciation of this sound, in order to produce which it is necessary to pout the lips.

We do not attempt to describe the mechanism of these sounds; to be at all understood, they must be heard continually during some months, till the ear becomes gradually able to appreciate and imitate them. We have, therefore, contented ourselves in general (19) with giving a list of words in which, as we utter them, the sounds occur.

The orthography and pronunciation of our language are so opposed to one another, that we must not be surprised at meeting with many anomalies in the common mode of referring the stopped to the full vowels in English. We are accustomed to consider ai and yū (20) as pure vowels, because they are the alphabetical names of the vowels i, u. Grammatically, then, the full and stopped vowels in the English language would have to be arranged thus, ai, i; i, č; ē, ă; o, ŏ; yū, g; no proper place

(18.) It is on the banks of the Rhine where this pronunciation is in full force. In the same part of Germany the final n is frequently dropped. Thus we recollect a servant girl at Weinheim (=Vainнaim), in Baden (=Bādən), asking for some teacups, “tsům shpīlə,” for play, instead of “tsům shpûlən,” for rincing. In Tyrol and Austria, ū ŭ take the place of û ù, and ê è of y e. Thus we were once directed to an inn at Innsbruck (=ĭnzbrük, which should have been Innsbrücke=inzbrùkə, the bridge over the river Inn), by the name "tsům göldnən lêbən,” the sign of the golden LIFE, instead of “tsům göldənən lêvən,” the sign of the golden LION!

(19.) In giving Dr. Orpen's account of the mechanism by which the seven principal long vowels, and Mr. Knowles's, of that in which the three stopped vowels, ĭ, ě, g, are produced, we have departed from this rule; and we leave the reader to judge from these specimens of the degree of satisfaction which he would have experienced from reading similar descriptions of the remaining 29 vowels; for we reckon 13 long, short, and stopped, in all 39 vowel sounds.

(20.) The nature of our work has frequently obliged us to anticipate the meaning of our symbols. The diphthongs will be explained in the next chapter. Here we may remark that ai represents i in tide, and yū, u, in use.