readers puzzled, by the uncertainty of types; and until some general principles are followed, no success in this course can be hoped for.

Such principles, however, seem to lie even on the surface of the languages most important to us; nor does it seem at all difficult to embrace, by slight modifications of our common type, nearly every letter of the Sanscrit, the Armenian, the Arabic, the Amharic; which seem to include all those consonants which are likely to be met, or for which at least we need at present make provision.

We distinguish the following as cardinal consonants; D T, G K, B P, Z S, M N, L R, H,-and Y, W, V, F, more or less supernumerary. Another aspirate is wanting, to express the Arabic Ain; and it seems to me highly desirable, if not essential, to cut a new type, substantially similar to that of the Arabic Ain final, only so far modified as to be similar in style to European letters. Such a new type, Professor Lassen has, at my suggestion, preferred to any other method for printing this sound in the Berber language, in a recent number of his Oriental Journal; but the type-cutter has been by no means successful. Perhaps it may hereafter appear, that, modified according to the same laws as the other letters, it may serve to express the curious clicks of which we are told in the Hottentot language.

The cardinal consonants above recited need to receive three principal changes. If this be but admitted (and I think it is a broad fact), it follows that the same mark should be added to every letter, to denote a change of the same kind; after which it is only a secondary question what that particular mark should be.

The principle has already been in part acted on. To express what is called the Aspiration of consonants, Bopp and others have superadded the Greek spiritus asper to t, d, &c. I understand, also, that Lassen has set an accent over letters, to denote their hard or palatal nature, in the Sanscrit. A third modification has been made less prominent, viz. the softening of letters, which, in imitation of the Portuguese ñ, I have thought might be denoted by a circumflex over any consonant. Such appear to be the three necessary modifications-Aspiration, Softening, and Hardening (Thickening or Strengthening?). Aspiration is to be distinguished from the mere annexa

tion of the h-sound to a consonant, such as we hear in the compound words, inhabit, dishearten, &c. Nor does there seem any reason why such compound sounds as the Irish th, fh, bh, &c. should not be represented by these true equivalents. But the very fact that such double sounds exist as the true th, ph, &c. makes it improper to represent such as the Greek 0 and English f by th and ph; for a legitimate use of the latter forms is preoccupied. This is no idle refinement. Serious inconvenience must result from the uncertainty whether such a word as sepher represented sefer with three consonants, or sep-her with four; and would quite lead astray a person seeking information in a strange language. Besides, when a consonant is doubled, as in afford, it would be very cumbrous to have to write it aphphord, or, if this were endurable, yet how are we then to aspirate by h the aspirates? An Irishman says Fhair for Fair; who would wish to have to write it Phhair?

σκι,

σι,

It has been usual to regard the sounds which the French write ch and j as Aspirates of s and z. But I apprehend they really are s and z softened, being nothing but a slightly degenerate pronunciation of si and zi before a vowel. This is seen in the words enunciate and persuasion (si for zi). The Scandinavian tongues write, I believe, sj for our sh, which recognizes the fact; nor is the Italian sci very far from it. The modern Greeks sometimes write σk, and sometimes o, before a vowel, for the same sound. I see no serious objection to devoting j to the sole office of softening consonants (as sj, zj, dj, tj, kj, &c.), except the cumbrousness of it, provided that Y, not J, were used for the isolated consonant, as in English. One remark only need be added; that D and G, T and K, when softened, merge in only two sounds-Dj and Tch, as we write them. This is a familiar fact to grammarians, in such words as Giorno from Diurnus; it has been less insisted on in the sounds Tio and Cio, perhaps because the Tio so often has also passed into the sound of Sho. Etymology must decide (if possible) whether a given sound should be written by softened G or D, by softened K or T.

Hard or Thick consonants are known to us as provincial peculiarities, but not as legitimate or radical sounds of the language. The thick T of the Irish must have been heard by every Englishman. It is not difficult to conceive of a similar variation of every other consonant. As regards the mark to be used

to denote a thickening of the sound, it ought, if possible, to be such as not to have any other use. The accent is forestalled for far too important a function to be superseded, which seems to make some other mark desirable. In an English Hebrew work of the last century, I have seen a cross placed over a hor k, to denote the strengthening or deepening of the sound; and a very similar artifice is employed by M. Delaporte, in recent specimens of the Berber language. Finally, it should be kept in mind that it may be necessary simultaneously to soften and thicken the same consonant, as in the thick or emphatic Tch of the Amharic and Armenian. The marks, however, which denote the two changes separately, might in any case be combined, if both were to be placed above the consonant.

After these preliminaries, the following scheme will be easily understood,

No cognizance is here taken of differences which are solely euphonic, as that between the French and English T, the soft and hard German ch, the Greek y and Arabic Ghain, the

X

Amharic and the Arabic Qaf. To attend to such minutiæ would produce needless embarrassment. It is possible that a fourth kind of mark may hereafter be found needful, to express the nasal modification of letters. The nasal n in French is wholly euphonic, as was the final m nasal (?) of the Romans. The Turkish grammars assign a nasal n as an alphabetic letter in that language; but the sound is in practice often neglected, and it does not appear, from the imperfect dictionary to which alone I have at present access, that this sound is ever significant, in contrast to the common n. If otherwise, a new notation will be needed. It is easy by trial to satisfy oneself that every consonant of our language admits of being enunciated nasally, in a manner analogous to the French n. The sound of the English ng is also unprovided for ; nor do I know how this is to be done on any general principle, nor whether the defect is of much importance in the Sanscrit and Celtic tongues.

F. W. NEWMAN.

P.S. Since the above was written, I have become acquainted with the Galla Grammar of Karl Tutschek, from which it appears that the Gallas have a modification of d and of t different from those known to Sanscrit, Arabic, and Amharic. Nevertheless, their peculiar D may be referred to the last column above, and their T, perhaps, to the second; on the ground that the D is very backward on the palate, as and its collateral consonants, while T is dental, and approaches the sound of ts, and is like the aspirates in being forward in the mouth. But on this matter it is perhaps impossible to judge without hearing the sound. The Gallas have also a peculiar nasal n, differing essentially from their common n.

F. W. N.

XXV.

THE PRACTICE OF WRITING ENGLISH IN CLASSIC METRES.

ENGLISH Hexameters and Sapphics have ever been a fertile theme for ridicule and contempt, and when they have been employed in earnest, they have given occasion to burlesques, which have remained in the memory long after their immediate cause has been forgotten. Canning's Knifegrinder is the first poem that presents itself to the mind, if classical lyrics in the English language are mentioned.

But yet it must be admitted, by all who look on the subject with an unprejudiced eye, and who bring with them a feeling for the tone of classic originals, that we must, in some way or other, approach the ancient method of treatment, if we do not exactly follow it. The recent adoption of sundry species of blank lyrics, which, while they are not antique, have never become familiar to the English ear, is a recognition of the fact, that the mode of translating in vogue during the last century will no longer meet the exigencies of the present.

Indeed, when we take up such a book as-for instanceFrancis's Horace, and look over a single page, we have a difficulty in understanding for what purpose it was intended, and what notion the whiter attached to the word "translation." Writing in a form the rules of which were exceedingly rigid and exacting, and which had not the slightest affinity with those of the original, he was constantly tempted to allow himself the greatest license with respect to his author, and was encouraged in his deviations by the prevalent desire to bring all poetry to one level mode of expression. The result of confinement on the one hand, and latitude on the other, was an English poem written indeed on the subject of the original, but different in every other respect.

That there is now a much more correct notion of translation than a century ago, there can be no doubt, but still there is a want of fixed principle, and we have no work of magnitude that answers the object of one of the German translations of