the tendencies of those languages which, having once reached a full inflectional development, have been broken up, to be reconstructed in new forms. In such cases, the tendency of languages to retrace the steps by which they originally advanced, and to return to the simpler grammar of primitive times, is certainly a remarkable, as it is in general an undoubted, fact.

The third part of this Grammar, on the Phonology of the English language, is a most valuable, and in some respects, a highly original, contribution to this branch of English grammatical science, a branch which has usually been almost as much neglected as that of the historical relations of the language. We cannot follow the author through the details of this part; but as we differ from him and his school in several important particulars, we shall take leave to present our own views in an independent form.

We venture to propose, without further apology, a system of consonant and vowel sounds for the English, and in general for the Teutonic and Pelasgic languages, in the accompanying tabular form. (See next page.)

The consonant system and the vowel system are to be conceived as having each its independent origin; as starting from opposite poles, as it were, and meeting each other in certain points of contact; and not as being continuously developed one from the other. They must, therefore, be treated separately.

We subjoin, first, the following notes and explanations of the system of consonants.

1. The progress in a vertical direction through any part of the table corresponds, step by step, with the transition through the organs, from one extreme to the other, between the throat and the lips.

2. The progress from left to right is, throughout, from surd consonants, through an increasing degree of vocalization, until we reach the transition to pure vowel sounds.

3. We have proposed to give, in our table, a separate representation to all those consonant sounds, and to those only, which are simple, that is, which are produced with a single effort of the organs, and, when protracted, retain unchanged the same character throughout their whole extent.

4. Although p and b are proper labials, their aspirates ph

Lene. Aspirate. Lene. Aspirate. Lene. Aspirate. Lene. Aspirate. Fundamental.

Alphabetical.

Usual Variations.

ORGANS.

and bh (or ƒ and v) have, in actual utterance, become labiodentals; but this is merely a result of accommodation for greater facility in articulating them. In Spanish and German, these sounds are still produced by many persons between the lips. It may be added, that kh and gh are easily made more perfectly guttural than their corresponding lenes.

5. Qk, is followed only by u, and either requires the u, as in French and Spanish, to be silent, or, as generally in English and German, reduces the u or oo to a semivowel equivalent to w. In Anglo-Saxon, it was represented by c=k, followed by w.

6. The consonants, it will be perceived, are distributed into four classes, called surds, sonants, liquids, and semivowels. Each of these four classes is divided into two branches, which we have denominated lenes and aspirates. By lene, we mean a determinate consonant sound defined by a simple contact or particular position of the organs; and by aspirate we mean, in each case, the result of bringing the organs nearly into the same contact or position, and then continuing to expel the breath for an indefinite length of time. In every case, the aspirate combines a breathing with its corresponding lene, and the relation of the lene to the aspirate is the same throughout the table; so that, k to kh, as g to gh, as d to dh, as b to bh, as w to wh, &c.

7. The aspirates are all, in their nature, continuous; and of the surds and sonants, the lenes are all explosive, or incapable of prolongation. From this characteristic, these two classes may be put under the common term mutes,

8. The term surd is applied to those mutes which can only be whispered, and are sometimes called sharp; sonant to those which are accompanied with tone or voice, and are sometimes called flat.

9. Of the liquids, the lenes as well as the aspirates are continuous; and this arises from the peculiar characteristic which distinguishes them as a class from all other consonants, and to which we would direct special attention; namely, that while, like the lene mutes, they are defined by a perfect contact of the organs, and are, therefore, not aspirates, they allow the breath to flow easily through some other passage, as through the nose in the case of the nasal liquids, or around the points of contact in the case of the palatals.

of

10. The semi-vowels (lene) may be described as a sort of fulcrum, or pivot of articulation, in passing from the English e (or i short) to any closely subjoined vowel sound, in the case y; and from u or oo to any such vowel sound, in the case of w. Thus, in yarn, wit, we may give first the full sounds ee-'arn, oo-'it, where, between the initial vowel sound ee, oo, and the following vowel sounds, the organs pass through a certain momentary but definite position, which gives the character of a consonant sound, and which we have denominated a fulcrum or pivot. If now the vowel part, the ee or oo-sound, be reduced to a minimum, and we begin immediately upon this pivot or fulcrum, and pronounce yard, wit, we shall have the y and w representing sounds of a proper consonant character. These sounds may be, etymologically and historically, the result of thus minimizing a vowel sound, and may then be called semi-consonants; or they may be the result of reducing a full consonant sound, as kh, g, gh, or l, for example, to this mere pivot or fulcrum, and then be called semi-vowels, which only comes to the same thing. They are, in short, the points of contact between the consosant and vocal systems.

11. Each of the aspirates might have been represented by a single character; but, as h represents a simple breathing or aspiration, and as all the aspirates are similarly combined with such a breathing, and those of them which are used in English are generally so represented, we have chosen to represent them all as combined with h. We do not mean by this to intimate that the sound of h is added to the re

spective lenes for in that case the aspirates would not be simple sounds but that it is combined with them throughout their whole extent. They are simple, therefore, under our definition; and, if in any sense compound, they are so by a sort of chemical composition, in distinction from a mechanical aggregate or mixture. Kh, for example, is not equal to k+h, but to kxh. This we consider a true aspiration; while the sound of h, added after a consonant, no more renders that consonant a true aspirate, than it does the following consonant or vowel. We do not doubt there are such aspirates ("so called ") in other languages, as in the Sanscrit, for example; but we here speak of the strict propriety of the term.

12. The ch (in the German ich) and jh (g in the German siege) are more palatal than their corresponding lenes k and g, and thus constitute a transition to the surd and sonant palatals, respectively, of the next set. In their most attenuated and softened form, they may also become mere aspirates of the semi-vowel y.

13. C (ch in chin) is manifestly a simple elementary consonant, and a lene. It is produced by placing a certain portion of the tongue near the tip, but not the tip itself, against a certain part of the palate, and, after pressure, suddenly withdrawing it with a violent emission of breath. It has no t-sound in its composition, for neither the tip of the tongue nor the teeth are used in its production. Neither does it end in an sh-sound, for, in that case, it could be prolonged ad libitum, which the true c (ch English) cannot be. Moreover, it does not begin with any one sound, and end with another, but is the same simple sound throughout its whole extent.

It may be shown by a similar experiment, and proof, that j is a simple elementary sound. It bears the same relation to c (ch) that g does to k, or any other lene sonant to its corresponding lene surd.

14. Sh is not the aspirate of s, is not related to s as th to t, ph to p, &c., as any one may ascertain by a simple experiment of pronunciation. S is more dental than palatal; sh is not dental at all. But sh is related to c (ch) precisely as any other aspirate to its lene; that is, if you place the organs as if to produce c, but instead of bringing them into perfect