words as cherish, merit, very, merry, etc., and lyric, spirit, virulent, mirror, etc., the vowels have their regular "second" sounds, and the r is a consonant. Established custom is, unfortunately, inconsistent in reference to words containing er and ir. Thus the sound in the radical word is retained in the derivatives, in stir, stirring; prefer, preferring; but not in err, erring, error, errant, etc. No wonder, therefore, that speakers are often perplexed as to the pronunciation of such words as sirrah, stirrup and sirup; some giving the "second" sound of i, and others the sound of ir.

Another class of words, respecting which principle and practice are considerably at variance, consists of adopted foreign words. These, when first introduced into our language, naturally come in the phonetic dress of their native tongue; but when the words have become naturalised by use, their foreign garments are laid aside, and replaced by the ordinary attire of English sounds. The change is merely a question of time. To an English complexion all such words must come at last. Whether, therefore, we say toma(h)to or tomāto, bana(h)na or banāna, promena(h)de or promenāde, chara(h)de or charāde, depends on whether we consider the words as foreigners, or as naturalised citizens. Fashion and affectation preserve the foreign sounds of words long after the terms have been incorporated in our familiar vocabularies. It seems to be thought more scholarly to speak of a "va(h)se" or "va(w)se," than a vāse; and of an "envelope" (French) than of an envelope; but the words must ultimately come to match with other words of the same orthography; and individual example can only delay and not avert the transformation.

The speaker's guide is expressed by Pope:

"In words, as fashions, the same rule will hold,—

Alike fantastic, whether new or old;

Be not the first by whom the new are tried,

Nor yet the last to lay the old aside.”

In application of this principle, and in view of the ultimate destiny of imported words, the rule may be adopted to Anglicise the pronunciation of words wherever the doing so is free from the objection of singularity.

An anomaly has recently been introduced of pronouncing such words as Celt, celtic, Cicero, with the sound of k instead of the ordinary "soft" sound of c. This is, no doubt, a reversion to the original pronunciation of the words, and therefore unobjectionable; but those who adopt it are bound to change the orthography to Kelt, keltic, Kikero, and not do needless violence to one of the few phonetic rules of our language. The spelling "Cicero" with the pronunciation "Kikero" is a monstrosity.

A most remarkable theory of elementary pronunciation was propounded in a recent report of one of the American Institutions for the Deaf and Dumb. The statement by the principal of the institution is that he had made a discovery, the description of which follows in his own words:

"This principle can be briefly stated to be, that aspirate sounds are made by exhalation, while vocal sounds are made by inhalation. Thus the sound of p is made by closing the lips, parting them, and expelling the breath. The sound of b, on the contrary, is produced by the attempt to draw in the breath, while the lips are closed. The sound of m is made by closing the lips and expanding the chest by drawing in air through the nose.”

The application of this theory would undoubtedly revolutionise the teaching of articulation and speech-reading, by rendering speech altogether impossible either to the deaf or the hearing. It is difficult to believe in the reality of such an absurd promulgation, more especially by a teacher; but there it stands where any one may read it, in the report for 1882 of the "New York Institution for the

Instruction of the Deaf and Dumb." If any one fact is established beyond controversy, it is that all speech, whether vocal or whispered, is formed of breath in outward flow; and that the replenishment of the lungs by influx of air takes place during the pauses of sound. The important practical point to be attended to, is to take care that this necessary influx shall be silent, regular, and sufficient.*

The principle must never be lost sight of that consonants require the articulating organs to be separated in order to finish the element. Attention to this point will secure to speech the beauty of distinctness, a quality wanting which the finest composition loses its effect, and the highest oratorical talent is obscured. The Rev. Mr. Austin, in his admirable book, the "Chironomia," compares the words of a good speaker to "beautiful coins newly issued from the mint, deeply and accurately impressed, perfectly finished, neatly struck by the proper organs, distinct, sharp, in due succession, and of due weight." To realise this description is to perfect the mechanical part of speech.

One broad principle of organic action remains to be noticed. The lips and the tongue are apt to be pushed from position to position with an ungainliness of effect unless they receive room for a lightness of play by a preliminary separation of the organs. The precept, then, is : Always open the mouth before you begin to speak. Whatever initial element you propose to pronounce, the first action of the organs, essential to distinctness and lightness of articulation, is an opening of the mouth.

Italian teachers of singing have long found fault with their English pupils that they would persist in singing through their teeth. Speaking through the teeth is still a

*See "Respiration in Speech."

prevalent fault in England, much more so than in America. But it is not to this that reference is made. Many of the actions of the mouth require the teeth to be brought very close, although never quite in contact. Let the jaws approximate without restraint when necessary, but give them always a free opening at the commencement of any utterance. Suppose the word you are going to pronounce to be a cake of gingerbread; open the mouth as if to prepare for a good bite, and then proceed until you come to a pause. This simple action will cure speaking through the teeth, and contribute sharpness, grace, and other good qualities to pronunciation.

IV. ENGLISH PHONETIC ELEMENTS. The following arrangement exhibits all the English Phonetic Elements, in a scheme of Roman letters, by means of which every detail of English pronunciation may be exactly represented in ordinary type.

The mark (-) over vowels denotes the "long" or namesounds of the letters; the mark (v) denotes their second or "short" sounds; the mark (^) denotes the sounds of the vowel-letters before r; and a dot under vowels denotes "obscure," unaccented sounds. The digraphs ah, ay, aw, oo, ow, oy are associated with their most usual sounds, so as to make phonetic transcription as little as possible different from ordinary orthography.