with the Greek in place. It agrees, to some extent, in form. In power it disagrees altogether. Nevertheless, the two signs are connected in origin, though signs of a different import. X, x, iz the Greek & (xi) the Hebrew Samech, which, in the first instance, the Latin alphabet either ignored or neglected to keep. It took the form, however, of x, x (Khi,) the letter with which it corresponded in place, though with a different import. For the two remaining letters,,, and V, v and Y, v, U, u we may probably claim an independent origin in each alphabet, or, at any rate, an early one of obscure origin; for notwithstanding the extent to which U, u and V, v seem to be mere varieties of the same letter, (one for the purposes of ordinary writing, the other for inscriptions,) I cannot but think that they stand in the same relation to one another as v and in Greek.

More important, however, than the consideration of the order, is that of

SECTION XXIV.

THE MERITS OF THE LATIN ALPHABET.

The merits of the Latin Alphabet are as follows:

1. It emancipated itself from the connexion with the numeral system; but the freedom thus created for the classification of the letters according to their affinities was never carried onward toward its legitimate results. A thorough classification of this kind is found in the Sanskrit only.

2. It improved the diacritical mark (*) as the sign of a breathing, or as an aspirate, into the truly alphabetic letter H, h. This is the only genuine new letter it has given us. It was, however, bought at a price. By gaining a sign for the aspirate they lost one for the short e (Epsilon.)

3. It kept the letter wanted thus preserving, for subsequent use in Western Europe, the letter F,f; which was the Hebrew Vau, which the Greeks allowed to become obsolete as a letter, though they kept it as a numeral.

4. It rejected, in the first instance at least, the compendium X,x: though, under Greek influence, it took it back afterwards. Upon Q, q more will be said hereafter.

Subject to these reservations, all these were movements in the right direction.

SECTION XXV.

THE DEMERITS OF THE LATIN ALPHABET.

The very fact of new letters being introduced from the Greek for the purpose of spelling words of Greek origin, tells us, in unmistakeable language, that the Etymological Principle has now been recognized.

The loss of the sign for the Greek Epsilon was, as has just been stated, the price of the letter H, h. The difference between the long and short O, o; the Omikron and Omega of the Greek, was in like manner left unexpressed.

The letter H h, so long as it kept its proper place, was a good servant; when it got beyond it, a bad master. The misapplication of it has been the source of three serious evils; for it has spread from the Latin to most of the languages derived from it; where it has affected not only the native words, but even such Greek ones as may have been introduced into it ;-words in the spelling of which it is singularly inappropriate.

1. Simple misrepresentation.-For this it is when we imagine that the sound of the ph in philosophy, etc., is really the result of a bona fide combination of p+h, etc. That it is allied to the sound of p is true; but it is equally true that, neither wholly nor in part, is it the same. Nor is the difference the result of any addition of h; though the prolongation of the breathing has something (much indeed) to do with it. The sound, in short, is a simple one; one incapable either of being decomposed into its parts, or built-up out of the combination of any two independent articulations. If it were otherwise, and if the letter h accurately represented the difference, the sound which stands in the same relation to b should be spelt on the same principle; and v be expressed by bh; in which case vase would be written bhase even as fase is phase. We know that, practically, this is not the case with b; but we should, also, know that it is not the case, theoretically, with p.

2. The diversion of the combination from its real power.-When the real sound of a consonant followed by h has to be represented, confusion arises. Such is the case in words like haphazard, inkhorn, nuthook and hogshead; where the second element begins with an aspirate. That this ambiguity can be abated by the insertion of a hyphen between the two contiguous letters is certain; for we can write hap-hazard, nut-hook, and the like. The expedient, however, simple though it be, is one which is unnecessarily forced upon us.

In compounds where not only the second element begins with an h, but the first ends in one, the objection is stronger. In words like Bathampton (so far as the elements are Bath and Hampton), Southampton, the h does double duty; and, in more cases than one, uncertainty as to the true elements of the compound has arisen. 3. The establishment of a precedent for digraphs.-This is

the head and front of the offending.

The preceding evils have been mere matters of detail. The one now under notice is the establishment of a vicious and pernicious principle. The combinations ph, th, and ch, as the Latin equivalents to •,, and X, are the fathers of all subsequent digraphs; the protoplasts of the family of the Makeshifts.

SECTION XXVI.

THE SUBSTITUTION OF THE LETTER C FOR K-K AND S

AS SOUNDS.

How c came to be used in the Latin Alphabet to the practical exclusion of k is a matter connected with the history of the alphabet which need not at present be gone into. We shall best appreciate the full import of the substitution by seeing what it has led to.

C by no means stands alone. This we cannot too closely attend to. What applies to c applies to other letters as well: in short, c as a letter, is pre-eminently a representative one.

In both the compendiums of the English alphabet, q and x, the sound of c enters.

In most digraphs we have either c or h, or both.

In the system of orthographical expedients, c is more conspicuous than all the other letters put together.

Let us begin with what is a good groundwork in all questions of the kind, provided that we can get it; a fact of Language; of Language itself as opposed to spelling, or the mere representation of language; a fact in the history of speaking, not merely of writing, a fact appertaining to the real object rather than to the picture of it. Let the language be what it may, it is a fact that wherever we have the sound of the k as in king, it is always likely, sooner or later, to be converted into the sound of the s in sing; or if not this exactly, into something akin to it,-into that of the ch in chest, or the j in jest, or something wherein the sound of s or its fellow-sibilant sh, enters. For what is the ch in chest but tsh, and what is the jin jest but dzh; and what is sh but s with a modification, or zh but a modification of z which is a sonant s? To s, then, in some shape or other every sound of k in existence has a tendency to be reduced. The process may be slow, or it may be quick. There are words in which it has not yet been completed; there are words in which it has not yet begun; and there are words in which it never may begin, or words which will be sounded with k until the language to which they belong is extinct. Still there is the tendency; while, on the other hand, there are words in which the k may have been changed three thousand years ago, or before the oldest alphabetical record in existence.

The change, then, or the tendency towards it, is a fact in language; the representation of it is a fact in orthography. The two may or may not coincide. If they do not, there is the risk of confusion sooner or later. At present, however, the fact in language is the only one under notice.

The first step in the investigation of this lies in the difference between the broad k and the small vowels-a, o, u on the one side, and e i, (and y) on the other. We know what happens to these empirically. Before the broad ones, cis sounded as k; before the small

ones as S. But they were not always so pronounced. If they were, why was the s used in spelling? The s sign existed. Why was the e necessary? Because words which once had the sound of k no longer retained it; and because words which now have that of s had it not when they were first spelt. There is something, then, non-natural in this use of cs: and the reason of it lies in the fact that the change of sound and the expression of it in spelling have not coincided. At present, however, the difference between a broad and a small vowel upon the sound of the letter by which they are preceded is the question in hand.

K before a small vowel has a tendency to become s. Hask the same tendency before a broad one? I think not. Kop will not directly become sop, shop, or tshop (chop); and the same applies to ka and ku-not directly. But here comes in the influence of the semivowel y. Now there is a tendency to say kyard for kard (card); and kyind for kind, even with a small vowel. The result may be a vulgarism, a Cockneyism, or the like. But, be it what it may, the change may be either introduced or kept up by so many speakers as to constitute a difference of dialect; and if that dialect happen to become the dialect out of which the literary language is developed, it becomes an error which corrects itself, a wrong which, by precedent and prescription, ends in constituting a right. It is a prophecy which fulfils its own accomplishment. Such, with the letter h, is the case with no smaller a language than the Italian. The literary Italian is the Florentine or Tuscan. But the Florentines (so to say) dropped their h's (Aitches). Before, however, the practice was noted and condemned as a vulgarism the dialect had become predominant, and the practice established. Hence, while it is a shocking thing to" exasperate" an aitch in English, it is equally objectionable to sound one in Italian.

But y after k comports itself as a small vowel; so that, when once kard (card) is sounded kyard, it is in the same predicament as kird. That the subsequent details of the change are different for the two combinations will be shown in the sequel: nor will the whole of them be exhibited. Kyard, does not, directly, become sard. It rather becomes ksard and tshard. It is submitted, however, that, as a fact in language, this is an adequate notice of the principle by which it is determined.

What we have now to investigate is the result of these tendencies, and the extent to which they may affect a language. This depends, mainly, upon the proportion which the sounds of k and s bear to those of the rest of the alphabet. The greater the share they take in the formation of any particular tongue the greater is the amount of their possible changes; so that here again we are about to be engaged with a fact of language as opposed to one of orthography.

Chas already been called a representative letter. But it is this mainly on the strength of its two-sided relations towards k and s.

It is in these that we must seek the realities of the question before us. K and s represent actual articulations, true elementary sounds, genuine consonants. C merely stands for s or k as the case may be. C, taken by itself, has no reality; and, except so far as its sounds are those of k or s it has no relations to any other letters. On the other hand, however, the relations of k and s are those of c also: and we shall now see that these are numerous. Khas its congeners, and so has s, so that each forms part of a system.

Herein, k stands to g (as in gate), as p stands to b, and as t to d.(4) But Ρ and b, t and d have, respectively, and as pairs, certain relations to f, v and th (both in thin and thine). Such relations, also, have k and g to a specific pair of sounds, to which they stand, each to each, as p to f, and b to v, etc. These sounds are not found in the English language: neither are they the sounds of the so-called gutturals ch and gh. What they are will soon be seen.

The P-series, as we may conveniently call it, runs through v into w, and thence, into u and the broad vowels.

The K-series runs, through the two un-English sounds, into y, and thence, into i and the small vowels.

More than this, the aspirate h is allied to both k and 9. In English, then, we have the sequence ha, ka, ga, (—), (—), and ya; allied sounds. They are five in number; and, if we had the analogues of f and v, they would amount to seven. Now every member of this group shares with k its tendency to change according to the character of the sounds with which it comes into contact.

S stands alone as little as k; being part of the series sa, sha, za, zha.

yoo.

Followed by y, s and z have a tendency to become sh and zh; as syoor shure, zyoor-zhure. This is common in English, though the spelling conceals it. U, however, in sure, and z in azure, are sounded K+ Y, and g+y have a tendency to become ksh and gzh. This, however, is not well exemplified in English; though the change is so important that it will command much of our attention in the sequel. T+y has a tendency to become tsh. The u in nature=yoo, and the sound is na-tshur.

D+y has a tendency (though not so strong as it was in the preceding instance) to become dzh (or j). The diphthong ew, when pronounced yoo, gives us not unfrequently the sound jew for dew. This is a vulgarism; but the allied change in nature is very good English, or if not, the change by which it is brought about is a genuine process of language in general, and not a peculiarity of any one dialect or language in particular.

Such, then, is the basis in philology of the changes which the sound of k may undergo, and of the extent to which an adequate or inadequate, accurate or inaccurate, representation of them by letters may affect the orthography of a language. It is manifest that in

4. Here, as elsewhere, there is sacrifice to conciseness. K, g, or p, etc., means the sound of k, g, or p, etc.