The natural result of this system of insufficiency is the creation of a whole series of makeshifts, or, as they are called by philologues, orthographical expedients. A few instances of this kind are enough to show how they work. A letter is wanting. This means that one sign has to do the work of two. Nothing illustrates this better than certain details in the history (and it forms a history by itself) of the letter c. In the French language it reigns predominant. It excludes k altogether. There was no k in Latin, and, as the French keeps up its classical traditions, there is none in the language of France. To a great extent, the rule that c before e, i, or y is pronounced as s is sufficiently regular and general to help the learner; though why there should be such a rule at all is not very clear. In like manner, the rule, that c before a, o, or u is sounded as k, is valid up to a certain point; for the sound of s before a broad vouel (a, o, u) in a word, which, etymologically, ought to be spelt with c, is rare. Still it occurs. To shew, then, that c is not c but s, we make a mark like a comma under it; which really means that we make a new letter though not one recognised in the alphabet and in the arrangement of dictionaries. It is not c; which, in such cases, is k. It is nots; because it is, as a letter, a mere modification of c. What then is it? It is c in a non-natural sense; a makeshift; an orthographical expedient.

;

On the other hand, the rule that c before a small vowel is s is valid up to a certain point: for the sound of k, which, before a small vowel (e, i, or y) ought to be spelt ke etc., is comparatively rare. Still it occurs. But the French cannot write it. In Italian we might write it chi but in French the combination ch is used with another power -used-up, so to say. It stands for the English sh; as in charade, chaise, etc. We have, then, no resource in the h. So the spelling is done by means of q+u; which gives us que, qui, quillet, etc. In Spanish, otherwise a well-spelt language, we have the same difficulty. How can a Spaniard express the sound of k before i? Not by c; for that has a different sound. Not by ch, for that, as in French, is used with another power. What, then, can the Spaniard do? He must even do as the Frenchman does-have recourse to q+u, and a very indifferent one it is.

Now the word chimera is, at once, English, French, Spanish, Latin, and Greek. In the last named language it is spelt xaipa, in Latin chimæra, in English chimera, in all of which languages the ch is pronounced as k. In French it is chimere; where the ch=sh. But in Spanish it is pronounced as in Greek, etc,. and retains either its true sound, or a near approximation to it. Yet in Spanish it is spelt quimera: of all impossible spellings for a Greek, the most impossible.

Such is one out of many of the long list of orthographical expedients. How little the system favors a true etymological representation is easily seen, Yet it is out of a supposed adherence to etymology than it grew. The preference of c to k is etymological or nothing.

It will be shown in the sequel that, mutatis mutandis, the series of insufficient makeshifts which has been illustrated by a reference to c repeats itself with g; where in words like rogue the u is needed in order to show that the o is long. Write it rog, and the vowel runs the risk of being sounded short (rog). Write it roag and you disguise its Latin origin, from rogo. Write it roge, and the g may be sounded as j, or as the ge in George.

This is sufficient to show what is meant by an orthographical expedient, and how it is connected with the insufficiency of signs; or, in other words, with the incompleteness of the alphabet.

SECTION X.

WRONG CLASSIFICATION-DISTURBANCE AND CONFUSION, ETC. To one of the three preceding heads more than three-fourths of the redundancies, deficiencies, inconsistencies, and other admitted faults of alphabets in general may be reduced. The minor faults may be noted in a more summary manner; or be indicated as they

Singularity in the use of any particular letter can scarcely be blamed when we consider an alphabet as what it is when taken by itself. There is no principle generally recognised which binds the speakers and spellers of one language to use the same letters that are used by others and still less to use them with the same power. Every alphabet must be considered on its own merits. If the Englishman chooses to use c where the German uses k, it is no fault of the alphabet of either of them. The practice of either one or the other may be peculiar, exceptional, or even eccentric. Still, this is. no reason against it. It is no part of one language to suit its spelling to that of another: though, at the same time, the greater the agreement between them the better. Absolute uniformity amounts, of course, to a universal alphabet; an admirable thing in itself, but one which we must wish for rather than expect. Nevertheless, extreme eccentricity in the use of a letter is a blemish. Nor is it a very common one. P, b, t, d, k, s, l, m, n, r are used with great uniformity throughout all the alpabets of Latin origin. The Hungarian, however, though in many respects a model alphabet, stands, I believe, alone in its use of s. In Hungarian it stands for sh as in shire; the s as in sire being represented by sz.

The misconception of the relation of sounds to one another, is a more serious evil. The English stands almost alone in treating the i in fine as the long sound of the i in fin: it is, really, that of short ee in feet; whereas our long i represents a shortened form of the combination of a (as in father) with y=ay: the sound of the ai in the German word Baiern Bavaria. This is, doubtless, broader than the ei in meine mine, but is equally compound; a compound made out of the same elements. So, too, with the ou, in noun, etc. Its real ele

ments are a +u (oo); or, as some say, o+u. In this mistaken view of the relation of certain vowels to each other, and the erroneous view as to the composition of our diphthongs, the English language is more than usually blameworthy. Then there are the faults which arise from the external relations rather than the intrinsic demerits of an orthography; such as extraordinary accumulations of words of foreign origin, mixture of dialects, intrusion of a strange orthography, and the like. These are the elements of what we have already called the wear-and-tear of a language, and of these the English has had more than its fair proportion.

Such, with the exception of a few additional minutiæ, are the chief reasons why alphabets and systems of spelling (we cannot always call them orthographies, but rather the contrary,) are insufficient for the purpose for which they are naturally intendedthe representation or reproduction of language. They will now be summarised, and that in the reverse order to the one in which they have been exhibited.

1. Upon wear-and-tear, in a general way there is nothing more to be said.

2. Upon the original incompleteness of the alphabet, and the want of any uniform principle in its application, thus much can be said; namely, that it is remediable. This can be said without reservation; and, for the purpose under notice, it is saying everything; for it means that a palpable and notorious evil ought to be remedied.

3. The etymological principle should, perhaps, have been put in a more general form. The inordinate preponderance, however, of the etymological objection over all others, especially in reference to the influence of the quarter from which it proceeds, makes it the representative principle of its class. A little consideration, however, shows us that the system upon which we distinguish the meanings of rite, right, write, and wright by the spelling, (our object being to indicate to the eye a difference of import when the sounds are identical,) comes under the same category-only its definition must be widened. Let us say, then, that the sacrifice of pure and proper Phonetic representation, when made for either the purpose of showing the origin of a word, or manifesting the difference of import between words similarly pronounced, (the differentiation principle,) are members of a higher class, the principle of which lies in the attempt to make spelling available for secondary and illegitimate objects, or to apply it to purposes for which it is neither intended nor appropriate.

It is possible, indeed, that by exploring the whole domain of language a few instances may be found where such a secondary object may, when not bought at the price of the sacrifice of anything else, be worthy of consideration. The cases, however, are yet to be found, and when found, have to be considered as exceptional, and (at best) excusable.

4. The extent to which the retention of old forms in the spelling,

when they have ceased to exist in the speaking, of a language, is desirable, is the question which is the least capable of being determined off-hand. As long as change goes on, there is always a period when it is difficult to say when the change on one side ought to be adopted to meet a change on the other. Sooner or later it may be made, but the right moment is difficult to determine. Here, then, more than elsewhere, are the innovators bound to place themselves, as much as possible, in the same light as the conservators.

If these divisions be natural, it is clear that there are degrees in the validity of the objections on one side, and in the claims for alteration on the other. The completion, however, of the alphabet, and the uniformity of its application are primary and absolute necessities. The limitation of the alphabet to its true function of representing a language differs from this so slightly that the difference has only been indicated for the sake of showing that it has not been overlooked. The fixation of a fluctuating pronunciation is a matter which must be left to settle itself. It is really a question of orthoepy rather than orthography.

SECTION XI.

AN ENGLISH GOSPEL TRANSLITERATED INTO THE LANGUAGE

OF THE KORAN.

With all these elements of imperfection developed in an inordinate degree, it is not strange that the art of learning to read English should be a difficult one. An Englishman, indeed, is apt to underrate its difficulties. Foreigners, however, are generally candid enough to own that, what with the multiplicity of its rules and what with the number of its irregularities, the approaches to it are, to say the least, discouraging.

As an instance of this I will lay before the reader an account of an attempt to teach it by the method of transliteration; and I will tell the story as slowly as I can, in order that, between the beginning and the end of the narrative, he may exercise his ingenuity in guessing at the explanation of it. I have never yet found anyone to whom I told it succeed in doing so.

About fifteen years ago a friend sent me a copy of one of the Gospels (John) in the ordinary English text but in Arabic characters. I took it, at first, for a mere curiosity, though without pretending to see my way to the object of it. Even a professed

Arabic scholar could not have read it off-hand. Letter for letter he could have spelt it, and would thus have arrived at something like something he had heard or seen before. However, the Preface told me what this was, and what was stranger still, told me that the work was one of a real practical value; that it was meant to be useful to certain men who, so far as the Arabic language was con

cerned, were no Arabs at all, but, on the contrary, as far as the English went, very good Englishmen ; men, indeed, of one language only, and that the English. This they spoke as their mother tongue, and they spoke nothing else. They could not, however, be taught to read it in English letters, and according to the English orthography. Put the words in Arabic characters, and they could read with pleasure and profit. I repeat it-I never found anyone who could guess who these strange English Arabs might be. Yet the explanation was simple enough. How much Mahometanism there is in the purely African and Negro parts of Africa is well known. The Mandingos and the Fulahs are more Mahometan than the men of Mecca themselves. In Bornu and Howssa the Arabic is to the native what Latin used to be to the Pole and Hungarian. It is, to them, not only the learned language, but, practically, the only written one. Those who use it are, of course, Mahometans, and where there is Mahometanism there is the Koran. And now we have only to think of the slave trade, and to turn to the Southern States of North America. Most of the negros of the plantations are pagans, and the descendants of pagans, men of many different languages and little knowledge of either reading or writing. Some, however, are either Mahometans, or of Mahometan blood; and for these the Gospel was thus transliterated from English into Arabic, in order to secure the means of representing its pronunciation.

SECTION XII.

THE ALPHABET IN ITS EARLIEST FORMS.

Let us now ask how far we can trace the imperfection of our spelling backwards; not, however, with an intention of giving anything like a natural history of the alphabet. This would lie beyond the scope of the present pages. Something, however, like a general view of the conditions under which the alphabets which have had the closest connection with our own were developed, may help us in the path of improvement, by showing how and when and in what directions the straight line of progress was abandoned.

The primary or primitive alphabet (the word "original" is often used in a different sense,) is generally attributed to the Phoenicians. We find it in a fuller form in the Hebrew, and hence it will often be called "Hebrew" in the present work. How are we to denominate such an alphabet? Was it an invention, or was it a discovery? Was it a mere application of something previously applied to something else? It was none of these things exclusively. The first sign, or letter, if made out of the first author's brain, and out of nothing else, was more than an invention. It was a creation. If the sign, however, existed with another import, it was an application. It is only, however, in respect of the letters themselves that