fand v top and b, etc., we should have a full and perfect quarternion; or four series of four similarly modified forms. As it is, we have only an approach to it. The class, however, as far as it goes, is natural. The sounds which constitute it are more closely allied with each other than they are with the sounds of any other group; yet, at the same time they have relations with the rest of the alphabet V, related to f, on the one side, is related to w on the other; through w to u; and, finally, through u to o and the vowels in general. Then there is an affinity between m and b; so that the full sequence runs m, b, p, f, v, w, u, etc.,

with a liquid at one end and a vowel at the other.

We have nearly the same sequence in the series beginning with k; and, if we had, on our language, the continuants of k and 9, we should have it entirely. Now what are these sounds? They are not those of the so-called gutterals, i.e., kh (or ch) and gh; sounds which, though foreign to our own language are common enough elsewhere as in the German, the Welsh, and the Gaelic. These are not the continuants here under notice; though often confounded with them. The place of the continuants in the sequence is that of f and v in the p series; and that is between the explosives and the semi-vowels; a fact which connects them with y, just as f and v are connected with w. Now these are conditions which are not fulfilled by the so-called gutturals.

The sound is a rare one. It is certainly found as one of the pronunciations of the Norwegian kj. It may be heard occasionally, and sporadically, in England. There are certain speakers who, when they use the word h-u-m-o-r never say, exactly, either "hew-mor," or "yoo-mor," but something like oo preceded by a sound intermediate to that of k and y. With stammerers, as far as my own observation goes, this practice is common. Now this is the sound in question, or the continuant of k. It is not, of itself, an easy sound; still less so when we have to compare it with its nearest congener, the continuant of g. In fact, the latter generally runs into y. It may be heard, however, in more than one of the Low German dialects. Like ky and gy, to which it is closely allied, it is a very unstable sound; and, as such, rare.

Thus much has been written upon the question of Unstable Combinations; and it has been written with a special view to the following question. Admitted, that an alphabet with its corresponding orthography is, in and of itself, an undeniable benefit to a language, can there be such a thing as certain times, dates, or stages in the growth, or development of that language in which the introduction of it is less opportune at one season than at another? Instances have been given to show that something of this kind exists. The next question is whether they be the only ones; and then follows the question, as to the extent to which they occur in English; and, if they do, what they account for in the way of objectionable spelling.

SECTION XXXV.

THE ENGLISH LANGUAGE AS ADAPTED FOR A PHONETIC

ORTHOGRAPHY.

Over and above the particular stage of a language into which a phonetic alphabet is introduced, is there anything in the structure of the language itself, upon which either an objection or a recommendation can be grounded?

We have said already much about the facts of language; treating them as potent, influential, and sometimes irresistible forces which determine certain results independent of any artificial or spontaneous conventionalities in the way of representation. It would, then, be a suicidal argument to hold that any language in the world was so constituted as to make a natural and simple orthography, an orthography limited to the mere representation of sounds and combinations as they actually exist, impossible. On the other hand, however, it would be a wilful neglect of patent and well-known facts to deny that, in the adaptibility of languages to such an expression, there are degrees. Some languages lend themselves to it spontaneously, others are not amenable to its treatment; except at the price of a contrast between one modification of a word and another. Thus, a language that, for instance, forms its genitive in the syllable es or is is less disturbed in the spelling of its inflections than one which merely adds the single letter s. This is because, in the former case, it does not matter in what kind of a consonant the radical part of the word ends. Whether the root end in b orp, tor d, k or g, we can add is without superinducing any further change. If, however, we merely add the sound of s, a change must take place, and one of the two consonants must be accommodated or assimilated to the other; in other words, we must, if the word be slab, write either slabz or slaps; if knot either knots or knodz. Combinations like knotz or knods can be, without doubt, written and seen, but they can neither be heard nor uttered. Such is our first instance of the Law of Assimilation, and we shall see that it goes a long way. Now the operation of this gives us one of the commonest and most prominent facts in the English language. It is the rule of the formation of both the genitive case and the plural number of our substantives. Yet we dispense with the phonetic exhibition of it, and write stags for stagz. Here, then, is a fact which would undoubtedly complicate the rules for the formation of two of the most important inflections in our language if they extended to our spelling; but it is, also, one which, as far as several centuries of experience teach us, we may ignore in our speaking. Doubtless, it is neither more nor less than an instance of the insufficiency of our orthography. As far, however, as it goes, it is evident that we can get a large amount of work out of an insufficient one; and this is an argument that a conservative may fairly make use of. But the antagonism

may go farther. There may be cases where the letters (that is, the sound represented by it) may be tacked-on, immediately, to a word in b or d, while the sound of z is wholly wanting to the language itself. In such a case it is evident that the latter of the two sounds cannot exist; so that no change, accommodation or assimilation can be effected in that direction. Hence the b must be changed into p, and the d into t. But this affects the radical, or fundamental, part of the word, and, by so doing, does much towards disguising its nature; since the combination must now come out slaps, etc. The English language does not, indeed, go so far as this. If it did, however, the result would be as has been indicated. Now this is the way by which the last part of a word, if such be the habit of the language, may be metamorphosed.

Let us now ask whether the middle part of a word can be thus affected? The best way of illustrating this is to take some language which gives us such affections as a general rule. By this we measure the extent of its influence. The following is from Euren's "Finlandic Grammar." It is not exactly an extract; it is rather an abstract. It gives us, however, the real details of a very important and relevant process.

The hard (this is the Swedish term) vowels are, a, o, u; the mild, ä, ö, y; the light, e, i.

The Vocal Harmony (vokal harmoni) means that the hard vowels (a, o, u) can never follow or precede any mild one (ä, ö, y), in a non-compound word. The light ones (e, i) can follow or precede both; but a light vowel in the radical part of the word requires to be followed by a mild one: as talolla, ottavat; tylsällä, kyntävät; and, from tie, mies, tie-llä, miehe-nä. Hence, words have two kinds of endings; one with a hard, and another with a mild vowel; i.e., with hard vowels in the root a hard ending, and with mild ones either a mild or a light

one.

The light vowels are our old friends the small or slender ones. Here the character of the vowel which follows is determined by that of the vowel which precedes.

But in the Gaelic the influence is reversed, and small endings give rise to small vowels in the root: a process which affects the middle syllable. Thus

...

A, o, are called broad vowels, e and i small. The poets, in latter ages, devised a rule, which prescribes that the vowel which goes before a consonant, must be of the same class with that which follows that consonant, i.e., both broad or both small. Neilson's Introduction to the Irish Language.

That this assimilation is not found in the older manuscripts is specially stated; but it is not stated that this is not a fact in the spoken language as well as in the existing orthography. And in both it is a fact. The same change, in a more artificial form, occurs in the German under the names of " umlaut," or ""aboutsound;" and this, like the Irish assimilation, is not found in the Moso-gothic, or the German in its oldest known form.

This, then, tells us how the middle part of a word may be affected.

Now let us ask what the Keltic languages teach us as to the alterations of the next beginnings of words. Any elementary work on the Welsh (where there are no true case-endings) tells how the first consonant of a word is modified according to its place in a sentence. Thus

and so on, with variations according to the consonant with which the word begins. Now the Welsh, phonetically, is one of the best-spelt languages in the world. But, if we can bring ourselves to imagine what it would have been had it been written in the manner of the English, and (what is a lighter effort) imagine that the reform had been deferred till the present moment, it is evident that the Welsh conservative might appeal to the etymological doctrine with more cogency than the English one. There are degrees, then, in the validity, or, as the Phonetician would say, in the plausibility of the objections on the score of etymology; the etymology being that which is limited to the single language to which it applies. There is nothing in any one of these instances which touches the question of the propriety of spelling city with a c on the ground of its representing the c in the foreign word civitas. Of the ground, however, to which it is limited, it covers a great deal.

Now the orthography of an allied language, the Irish, has not been reformed; and here we find an attempt to combine the etymological principle with the phonetic. The Welsh sacrifices the former to the latter, and changes the radical consonant. The Irish preserves it; but prefixes the letter, or its equivalent, (which in Welsh displaces it,) and gives us a rule by which it is said to be eclipsed; i.e., written, but treated in speaking, as if it were non-existent. Hence, the following table,

Now, however meanly we may think of the value of the etymology which connects two words from different languages, like city and civitas; we must allow that, when we come to changes within the same languages, the etymological objection improves; and when we see such transformations in the radical parts of words as the pre

ceding, we must admit that its validity, or, to say the least, its plausibility, may vary with the language. I am not writing this to show that in any language it is actually valid. I am only showing that different languages are in different predicaments in respect to its applicablility. And I only do this with the view of asking how the English stands in this respect. I submit that in respect to any of the above-mentioned languages, or as tested by any of the above-mentioned processes, it stands high. So far as the mixed character of its vocabulary goes, it is in a worse position. The phonetician condemns the etymological system in toto. It is well, however, that he should know how his own language, in the eyes of an opponent, and from his point of view, comports itself. The English is fitter for the application of the phonetic system than most languages; and, even when we weigh its demerits against its merits, as fit as any.

to

SECTION XXXVI.

ANALYSIS OF THE ETYMOLOGICAL PRINCIPLE.

The bearing of the instability of certain combinations, and certain points connected with the structure of numerous and important languages, has now received its full share of consideration; and, it is evident that upon the most important of the current objections to Phonetic Spelling-the Etymological-they have a very decided bearing. The number of scholars who hold that, though the phonetic system may suit some languages it is but ill-fitted for the English, is considerable. Now just so far as the English is a language in which there is a foreign element, this objection is plausible; and, as there is an inordinate proportion of foreign words in our tongue it has, say the least, sufficient importance to command attention. Beyond this, there is but little in it. I have, perhaps, gone farther in this same etymology than usual; but I now wish to show that I have not gone farther than is necessary. Much that has been written, as the contents of the last four sections, has, perhaps, appeared irrevelant. But it is not this. The facts exhibited and the trains of reasoning suggested may, perhaps, have some import in themselves. But the main object of their exhibition is to show that they are only parts of a whole; only means to an end. However much they may, at the first view, appear to be something different, they are, in fact, preliminaries to the question as to the value of the etymological objection.

It will now be submitted to the reader that the ordinary objec tions which are involved in the word etymological, as it is generally applied, are mere fractions of the import of the term in its more general and more legitimate sense. The current objections are etymological so far as they hang loosely on the great subject of etymology; but as the representatives of etymology in its wider sense