word between vowels, and when final had a very weak and indistinct sound, particularly in the earlier and later language of the common people.

We might proceed to particularize further differences, not only in respect to the consonants, but also the vowels and diphthongs. But we have carried our comparison as far as we need to do for our present purpose,-far enough to show from valuable authority that the argument of those who would inspire us with courage to pronounce Latin "as the Romans did" is met by the fatal difficulty, that it is by no means agreed how the Romans pronounced Latin. Not even this new and valuable contribution of Professor Corssen's can be accepted as settling all the various and minute questions involved,-certainly not until it has been itself subjected to a general and thorough examination and criticism. It can itself rather be brought forward to show that the case is not yet closed.

"Grammatici certant et adhuc sub judice lis est."

We had designed to present at length some specimens of Professor Corssen's method of investigation, but this is rendered unnecessary by the publication of the Messrs. Tafel's abstract of his discussion of the alphabet, which we have had an opportunity to examine cursorily since writing all but a few lines of the preceding pages of this Article.* To this abstract we are glad to be able to refer those who have not Mr. Corssen's work at hand. Its object is "to give to the English students who have no access to the original work, a clear and succinct statement of all the results at which Mr. Corssen has arrived, so far as they have reference to the language itself, and excluding the part on quantity." It is not a consecutive translation, nor is it a mere compendium, but it is a combined exposition and criticism of Professor Corssen's views. It is an interesting publication, not only because of the valuable contents referred to, but because the editors, while they exercise their judgment independently, and so differ with Professor Corssen on several elementary points of pronunciation, yet close their little vol

* Latin Pronunciation and the Latin Alphabet. By Dr. LEONARD TAFEL, of Philadelphia, and Professor RUDOLPH L. TAFEL, A. M., St. John's College, Annapolis. Philadelphia: I. Kohler. New York: B. Westermann & Co. pp. 172.

ume with a new appeal to American scholars to reform their pronunciation of Latin, on the same ground as that on which Professor Richardson urges his Plea,-namely, that "the method by which the Romans pronounced the Latin, we [may] consider now as fairly established."* We, on the other hand, should take the very book they give us, as evidence that it is not established,-as evidence that within twelve months, new discussions of new editors and authors may defend new posi

The following table represents the sounds of the letters in "the practical Roman method," proposed by the Messrs. Tafel:

1. Sounds of the Vowels.

A long, as in the English father. A short, the same sound shorter.
E long, as in the English fate. E short, as in then and dishes.

I long, as in machine and caprice. I short, as in sit.

Rem Ai in ain, ais, ei in dein, oi in proin, are to be separated.

3. The Semivowels.

Jin jacio, like y in year.

V, like English v.

4. The Consonants.

a. Gutturals.

C, like k; before e, i (y), like ts, as in glacies, facis, etc.

Qu, before a and o, like k followed by a mute u (in put); before ae, e and i, like k followed by v.

G, always as in good and get. H, as in English.

b. Labials.

P, B, F, and Ph, as in English.

C. Dentals.

T, as in English. 7, when followed by another vowel, as in natio, like ts; except when preceded by 8, x, or another t, as in tristior, mixtio, Bruttii, in old infinitives in er, as nitier, quatier, in Greek words such as Miltiades, Bacotia, Aegyptii, and at the beginning of words, as in tiara, in all which cases it is pronounced like t. D, as in English.

d. Liquids.

L. R, M, N, and NG, as in English.

e. Sibilants.

S and X, as in English. Z like the Italian z, viz: ts or ds.

A little observation in the application of this system will show that it makes great havoc with that proposed by Professor Richardson.

tions and theories, which, according to their ingenuity and ability, may influence larger or narrower circles of adherents, and so introduce new disagreements in practice.

We say, therefore, that we are not yet ready to give up our system of pronunciation. It is almost universal among scholars in England, and quite as much so in New England. Out of New England, in America, it is more prevalent than any other one system. A large majority, therefore, of English-speaking scholars, have a system in which they agree and in which they understand each other. It is also founded on English analogy, and is therefore natural to the English learner and is easily retained. More than ten thousand valid objections to it can be brought by those who insist that, according to the light we have, we ought to pronounce as the Romans did. But objections scarcely less numerous can be brought against the pronunciation of our own good English tongue. Shall we set ourselves to a classical reformation of that? We think that the time has not yet come.

But shall we give up all thought of ever returning in our pronunciation of Latin to the true method of the ancient Romans? We are not prepared to say that, we only say that the way is not yet surely prepared for the great departure which is proposed. It may be that the day is not very distant when the venture might safely be made with a fair prospect of more good than evil results. But it is not clear to our minds that it has already come.

But why shall we not at least adopt the Roman pronunciation of the vowels, as the nations on the continent of Europe have done? To this we reply, first, that there is a want of agreement among scholars respecting the Roman method of distinguishing long and short vowels;-next, that continental scholars pronounce the Roman vowels as they pronounce the vowels of their own several languages, and therefore do not in practice agree; and, finally, we wish to ask what is the object of this whole proposed reform? This third point deserves serious consideration.

We suppose that when the sounds which represent, for instance, the Latin word jam, are uttered in the hearing of a Latin scholar, with the pronunciation with which he is fa

much the same as If this is so, then

miliar the impression on the mind is if he had read the word with his eye. the Englishman receives the same mental impression from the English pronunciation of the word that the German receives from the German utterance of the same word. The same would be the case with the Frenchman and the Italian and the Spaniard, however much or little their modes of pronunciation might differ. They all would think without hesitation of the same word, and would point out the same on the printed or written page. What difference does it make, then, whether the Englishman says jam or yam? We do not mean by this question to suggest that all pronunciation is a thing of indifference; for the Englishman has it to say,-and so has the German and the Italian,-that he pronounces as he does in obedience to the inscrutable laws of language, which, in the progress of the ages, have made his mother tongue with the pronunciation of it what it is. He does not, therefore, pronounce arbitrarily.

We repeat the question, then,-Why should the Englishman say yam? For the sake of agreeing with the continental scholar? But he does agree with him respecting the word in everything but the sound of it, he writes it in the same way, and when seen it suggests to both precisely the same idea. And as for oral communication, the occasion for it is comparatively rare,—so rare that no one would think it worth while to adopt a new system of pronunciation for that reason alone. We think, then, that the argument derived from the mere desirableness of uniformity among scholars of all nations, has, in itself, very little weight.

Why, then, should the Englishman say yam for jam? Is it for etymological reasons? We acknowledge that there is in many cases some force in this; and yet, to how many will the thought of etymological affinity between jam and id and ita, etc., be suggested by yam sooner than by jam? Surely the human mind is not all ear, notwithstanding the fact that language is primarily conveyed by sound, and that the mental operations are carried on by words. Who would be helped in his etymology by say ing dokility, because e in doctrine has the sound of k ?--or who is troubled in his etymology by the varying sound of g in

legal and legislate? Or who, we may ask further, that is making etymological investigations, let him be a beginner or a proficient in such studies, finds any difficulty in applying to a word under consideration all the sounds which may belong to it according to any theory of pronunciation? Etymology is rather learned by reflection than by intuition, and although many affinities in language would doubtless be more obvious at first sight to a novice with a different system of pronunciation, still a little practice causes the obscurity resulting from our present system to disappear and be forgotten.

But some one is ready to answer the question which we have repeatedly proposed, by saying that we should say yam because it is right. He insists that it is absurd in us, when we know how the Romans pronounced a word, not to pronounce it in the same way when we have occasion to use it. When we profess to utter Latin, let us utter the real language of Rome as the Romans uttered it, and not substitute for it a "barbarous jargon."

In reply, we would say, that if the ideal which our objector is in pursuit of, is the living pronunciation of the Latin language, such a pronunciation as charmed the hearers of Cicero, in the old Forum, or in the villa at Tusculum, or as conveyed the thoughts of the Romans in their ordinary life, then he cannot hope to realize his ideal, so long as there is doubt respecting any of the elementary sounds of the language. For suppose, for instance, that the Romans always gave e the sound of k, and t the sound it has in the English to, and that our modern scholars have decided that they are sibilated before i followed by a vowel; will not those two errors alone make all his attempts to reproduce the living language, an utter failure? Or, suppose that he errs in the sound of one of the vowels, for instance, e short, will not that error, reappearing, on an average, once in every ten syllables of the language, corrupt the whole, and make all his attempted imitations a mere caricature, which a Roman audience would hiss out of hearing? If, as Cicero says in a passage referred to in the Article in the North American Review, a failure to make a long syllable quite long enough, or a short syllable quite short enough, caused whole audiences in Rome to shout