The Greeks, then, could not be consistent in the representation of the longness and the shortness of their vowels; and being unable to be consistent, had recourse to two principles

a. Sometimes, the longness of a syllable was shown by spelling it as a diphthong; and

b. Sometimes, the shortness of a syllable was shown by doubling the consonant that concluded it.

Both these methods are English, inasmuch as we not only write such combinations as oe, &c., to show that a vowel is long, but we also write such combinations as ock (okk), &c., to show that a vowel is short. This has been stated already. The antiquity of the practice is stated now.

§ 317. The Etymological Principle,-how far was it Greek? —That the greater part of our chief orthographic expedients are anything but novelties has been shown; for it has been shown that they can be traced up to the earliest period of our alphabetic history. How far is this the case with the Etymological principle? Did the Romans recognise it? Yes. Did the Greeks? No. Such is our answer; affirmative to the one question, negative to the other. Nor is it difficult to find the reason for the difference. The Roman literature began betimes to be affected by Greek influences, and Greek words at an early period found their way into the Latin language.

Now between the Greek and the Phoenician there was no such close literary contact, no great influence, no great borrowing of words; besides which the difference between the alphabets was greater than that between the Roman and the Greek. Neither Greeks nor Romans paid much attention to the study of the languages of the countries with which they came in contact; indeed, they despised them as barbarous. Of the two, however, the Romans paid the most. They studied the Greeks, and kept up a connection with their literature. The Greeks, on the other hand, when they had got what they wanted, in the shape of an alphabet, from the East, seem to have done with their instructors in the art of writing, to have dispensed with them entirely, to have distanced them in the race of Literature, and to have ignored all subsequent influences. Hence, if they did borrow a word from Asia, they took no more pains to keep up for it its Asiatic character, than we do

to indicate by the outward and visible sign of a peculiar orthography the Chinese, or Indian origin of such words as tea, or shawl.

The Greeks, then, ignored the etymological principle in the case of words of foreign origin. The Greeks ignored, but the Romans recognised, it; so that it is to the Roman period (and no further) that we can trace it.

Did the Greeks ignore the etymological principle in toto, i.e. in the spelling of words wherein one sound was changed for another according to the rules of their own grammar? Thus-if the present tense of a word was tup, and the preterite tuf, how would they spell the latter? Would they concoct any means for retaining the original letter p (or some part thereof) in the derivative form, in order to show the connection? We do it. So do many others. But did the Greeks? No. If one sound was changed into another, they took the change as they found it, and expressed it accordingly. Thus, one form of τύπτω (typto), was written τέτυφα (tetyfa), another ετύφθην (etyfpén). They might, however, have written rúttw, tétut'a, and εrúπ'r'nv—by which process the etymological connection between the and 7 on the one side, and the π and ▾ on the other, would have been all the clearer. Nevertheless, the Greeks did nothing of the kind, and they were wise in not doing it. As languages get modern they tend toward the contrary extreme, and so far over-do their etymologies as to conceal difference in order to exhibit likeness-a proceeding generally unnecessary, rarely successful, always and essentially untrue.

T

The general features, then, of our orthography, are as old as the Greek period. The most important details we get from

§ 318. The Roman.-Here we begin with the notices of such particular letters as claim attention; first of which

comes

C, the third letter of the Roman alphabet as it is of our own; but not the third of the Greek, nor yet the third of the Hebrew. The chief points of its history have already been given; the change, in power, from g to k, being Etruscan. Now, without saying that the Romans took their alphabet exclusively from the Etruscan, I hold that they did so as far

as the letter c is concerned; in other words, I doubt whether the Romans would ever have given it its modern power if it had not been for the Etruscans having done so before them, in their neighbourhood. But the Etruscans did so because the nature of their language (whose sound-system wanted b, g, and d) suggested it. Why do I indicate this? I do it for the sake of showing how little there is of what is arbitrary in the structure of the alphabet; which has grown after the fashion of a tree, rather than been built up after the manner of a house.

"This letter," writes Professor Key, "is derived from the Latin alphabet, in which it first appears. But even in that alphabet it originally possessed the power of g, as pronounced in goose. Thus the Roman proper names Caius and Cneius, which retained this sound, are correctly represented in the Greek character by Taos and гos; and the Duilian inscription, the orthography of which, however, seems to belong to a later date than the events celebrated in it, presents macestratus, leciones, pucnandod, ecfociont, in the place of the modern forms, magistratus, legiones, pugnando, ecfugiunt. Indeed the poet Ausonius expressly states that C once performed the duty of G; Gamma vice functa prius C. (See also Festus, vv. Prodigia, Orcum.) This medial pronunciation corresponds with the power of the letters which occupy the third place in the Greek and Hebrew alphabets, gamma and gimel; and the identity of the letters is confirmed by the similarity of the forms.

a, o, u.

с

"The letter c in English is pronounced as s before i, e, and k before This variety in the power of the letter seems difficult to account for; but it may be observed that i, e belong to one end of the vowel series, a, o, u to the other; and it is further to be noticed that the vowels i and e, when they precede vowels, have a power approaching to that of y in youth, and that if, in addition to this, c or g precede, there often results a sound like that at the beginning of the words church and John, and this sound of ch is not very different from a sibilant. The vowels i and e produce a similar sound when preceded by a d or t and followed as before by a vowel. Thus from ration the Italians have obtained ragione; and from radio, raggio ; from Diana the rustics of ancient Italy made Jana. These considerations are perhaps supported by the employment of the little mark called cedilla in the French language, which is used to denote that c is to be pronounced as an s even before the other vowels, as ça; for the mark appears to have been originally an i."

In this extract, the statement that c originally possessed the power of the g in goose requires qualification. As the sentence stands, it looks as if all the old c's of the Latin language were once sounded as g's, or, at any rate, that the normal and primary power of the letter was that of G, that of C being derivative, secondary, aberrant, and exceptional.

Of course, in the Greek and Phoenician, all of what was afterwards C was, originally, G. In the Latin, however, I find no reason for believing that the sound of G was the older. On the contrary, I believe that that of C was. How came it, then, ever to be sounded as G?

I imagine that the same Etruscan influence which gave the third letter with its peculiar power to the Roman alphabet, also determined either the non-adoption or the non-invention of any representative of the sound of G—in the first instance, at least. Yet as this sound of G was Roman, the sign that should indicate it was necessary. In due course of time it came into existence. Previous, however, to its development, I hold that C did double duty-partly as C, and partly as G-primarily, however, as the former, and not as the latter, letter.

Neither do I deny that at a period thus early its third power of s (or some modification of s) may have existed.

I cannot, however, think that at the time when legio and pugno were spelt lecio and pucno, that such a word as canis was either pronounced ganis, or spelt kanis-admitting, however, the possibility of the latter alternative.

F.-The history of this letter is given in the sketch of the Etruscan stage. It arose out of Vau. But what if the sound of Vau (v) existed in the language as well? A difficulty would arise. Either a new sign must be developed to represent it, or an orthographic expedient must be resorted to. This bears on the history of our letter v-now at the end, once at the beginning, of the alphabet.

I see in this modification of the power of Vau a confirmation of the views already expressed concerning the influence of the Etruscan alphabet on the Latin. Why should the Romans change the power of either Gamma or Vau, when they had the sounds of both G and V? A fortiori, why

change both? If we look at the sound-sytem of the language, the alteration is pre-eminently gratuitous. But it is not gratuitous if we look at the sound-system of the Etruscans, and, along with that, the extent to which the Romans borrowed from Etruria.

"G. This letter is derived to us from the Latin alphabet, in which it first appears. In the Greek alphabet its place is supplied by zeta. If, as seems probable, the sound of this Greek letter was the same as the consonantal sound at the beginning of the word judge, it may perhaps be inferred that the hissing sound now given to the letter g existed already in some dialect of ancient Italy. The sound at any rate is familiar to the modern Italian. The sound of the letter g in the English language is two-fold. Before a, o, and u, and occasionally before i and e, it is the medial letter of the guttural order. The other sound, which it possesses only before i and e, is one of the medials of the sibilant series, and is also represented by the letter j as pronounced by the English. The sibilant sound is written in Italian by two letters gi, as Giacomo, Jacob, or by gg, as oggi, to-day. The two-fold nature of the sound corresponds to the double sound of the letter c, which is sometimes a k, sometimes an 8."-KEY's Alphabet, p. 63.

A great deal of what applies to C applies to G also; and the history of the two letters is closely connected; indeed the histories of C, G, and K may be said to be complementary to each other.

Add to this that the histories of Zeta and Vau are, notwithstanding certain details, alike. Each was taken into the Latin alphabets from the Phoenician, either directly or indirectly. Each was changed in power. Each reappears-the one as v, the other as z, and each at the end of the alphabet.

Place for place, however, and sign for sign, we must remember that F is V; and also we must remember that, place for place and sign for sign, G is Z.

That the sibilant power of G (=dzh) is of the antiquity ascribed to it by Professor Key is probable. If so, the double power of g (as in gun and gibbet dzhibbet) is no novelty.

Neither (if the view concerning c be correct) is the double power of that letter (as in cat and city=sity) a novelty.

H.-The effect of the Latin orthography upon this letter