incompetent tutor; he will give you a fair outline of the general meaning, but when you come to question him closely, why is this particular word used here? what would this construction be in ordinary Greek? why does this collocation of words mean so and so when it usually means something else? he cannot go on for two lines without stumbling. Now, of course we do not mean to compare Horace with Sophocles for difficulty; yet there are many latent niceties (dodges and catches as they would be called in Cantab slang) all through the Odes, and the very fact that they have the reputation with most students of being easy is the strongest argument in support of our position. The Satires, on the contrary, are considered hard, and it is just for this reason because their difficulties are appreciated that our students on the whole know them better than any portion of any author read in our colleges.

We did not intend to make any particular remarks on individual notes in this work; but a single one which has struck our eye we cannot forbear commenting on briefly. At v. 6 of the Epistle to the Pisos (usually known as the Ars Poetica), Prof. Lincoln says "isti tabulæ. Such a picture as that: isti expresses contempt." We do not believe that iste in classical Latin ever expresses anything of the sort. There was a dictum of the old grammarians to that effect; and it is because it was one of the things particularly impressed upon us at school, and because we not only read but wrote a good deal of Latin before discovering the error, that we are anxious to correct it in others whenever the opportunity presents itself. (still represented by ese in Spanish and cotesti in Italian) is the demonstrative pronoun referring to the second person, as hic refers to the first person and ille to the third; hic, this by me, iste, this or that by you, this of yours; ille, that Cat a distance from both of us). The idea of implied contempt probably originated thus; in an advocate's speech, iste, your man, would be the term naturally applied to the client of the opposite counsel, and as "your man" was pretty sure to be well abused before the speech was through, grammarians fancied that the word had a bad sense and denoted a contemptible object in itself. So far all is tolerably plain sailing; but besides this there is a secondary and loose use of iste to denote a subject

of previous conversation or allusion; this that we have been talking about (we might construe this between us, to carry out the locative discrimination between the meanings of the pronouns) as in the passage before us, isti tabula, this imaginary picture that I have been telling you about, or in one word, such a picture. If it be asked why the word may not also express contempt here, since the imaginary picture is certainly ridiculous and contemptible enough, we answer simply because iste is found in other places referring to antecedents anything but contemptible. Thus in our very author, Epist. I. 6. 67: "Si quid novisti rectius istis, Candidus imperti, si non his utere mecum;" where Horace cannot mean to throw contempt on his own precepts, while he is challenging his reader to produce any better ones. So also Cicero in Catil. I. 9: "Utinam tibi islam mentem dii immortales duint." Would that the immortals could give you such a disposition (to go into exile).

LATIN PRONUNCIATION.*

Literary World, July 1851.

THE little pamphlet with this long title is really multum in parvo one of those books that suggests the perpetration of an article on it longer than the work itself. Professing to be merely a guide to scientific students who are not scholars, it opens out into a discussion of all the doubtful questions in Greek and Latin pronunciation. These questions cannot fail to be of interest to every scholar, particularly an English or American one, on account of the greater damage which the learned languages suffer in being subjected to the pronunciation of ours. Our difference from and inferiority

Elements of Latin Pronunciation, for the Use of Students in Law, Medicine, Zoology, Botany, and the Sciences generally, in which Latin words are used. By S. S. Haldeman, A. M., Professor of Natural History in the University of Pennsylvania. Philadelphia: Lippincott, Grambo & Co.

to the continental nations of Europe in this respect arises, not so much from the consonants tazé, taché, and tathé are probably as great variations from the original sound of tace as lasé is but from the continually recurring vowels a, i, u. It is not probable, however that the English will ever alter their habit of pronunciation, although it renders their attempts at conversing in Latin with German or Italian scholars difficult and ludicrous. In this country, where scholarship is more limited and more in its infancy, any attempt at such change might be more likely to succeed. Indeed it has partially succeeded in New York, where we have adopted a pronunciation of Latin and Greek nearly approaching the German; but the New Englanders still retain the English powers of the letters, with the additional ornament of as many false quantities as possible. Still it is not probable that even we shall generally adopt a new standard of classical pronunciation, because it is difficult to ascertain satisfactorily what the real standard was, for many reasons.

First, there are the natural caprices of language. Suppose the French should cease to be a living tongue, by what analogy or parity of reasoning from the other European languages should we be likely to arrive at its true pronunciation? If several files of the Charivari and other comic publications remained, the puns and rebuses might help us to find out some of its peculiarities, such as that it has more than a dozen combinations of letters to represent the simple sound of long o, but others, such as the pronunciation of the diphthong oi, we should never be likely to hit unless indeed we found in some English, German, or Italian author, French words written according to their sound in those languages and even not then with perfect accuracy, by reason of,

Secondly, the uncertainty with which sounds are rendered from one language to another. The Romans, we are expressly told by themselves, had no sound corresponding to the Greek upsilon and were obliged to invent a character for it; neither English, Spanish, nor Italian have a sound corresponding to the French u German teachers and German grammars will tell you that their ö is equivalent to the French au and their ü to the French u, which is contrary to the experience of every man's ears who has heard the two languages spoken

constantly, and also to the fact that such rhymes as schön and gehn, blick and zurück occur continually in the best German poets, whereas no Frenchman would think of rhyming cœur with amer or dure with pire. The Greek dipthong &, though generally expressed in Latin, by i, was in some well known words, as Medea, expressed by e, but this may have been owing to another cause, which brings us to,

Thirdly, the variableness of pronunciation in different parts of the same country, and by different people. In France and England there is but one standard, but equally well educated men in different parts of Germany will pronounce the past participle of the verb geben, ghegayben, yegayben, and yeyayben. The instability of some of the Greek diphthongs, particularly those of the long vowels with subscript, seems the only hypothesis capable of accounting for the contradictory modes in which they are expressed.

Fourthly, we have the difference of opinion among individuals themselves as to what sounds are different and what identical, what long and what short, what long and short sounds correlative. Thus Mr. Haldeman seems to consider the French u and German ü precisely equivalent to each, which we consider a want of discrimination. On the other hand, if asked the quanity of the vowel in art, we should say it was long like that of arm, for which he would reprehend us. And many people still maintain the (in our opinion) traditional infatuation of English lexicographers, that ai and long a have a different sound, e. g. that fair and fare are distinguishable in pronunciation. The great confusion of vowel and diphthong sounds, as well as of correlative short and long sounds, tends to destroy our accuracy of ear in this respect.

*

Fifthly, limited knowledge and the imperfect generalization consequently made from once or a few languages,

*In English two of the vowels (I U) have diphthongal power, one vowel (the continental U) can only be represented by a diphthong (00) and the ordinary long and short powers of every vowel belong to two different letters or a letter and a diphthong. Thus a in father and a in fate are the a and e of most languages, and so on throughout; in no one case is the short power of a vowel the sound of its long power shortened, or the short correlative of its long sound.

constitute a formidable impediment and source of error. When we are discussing what a sound ought to be or how one sound should fall into another on general principles of speech, it requires a most extensive knowledge of different languages to justify a dogmatic assertion of what those principles are. We have often seen and heard advanced in support of the uniformly hard pronunciation of C and G in Latin, the artifices used in the modern continental languages to preserve unchanged the sound of the radical consonant (e. g. manger makes mangeons not mangons in the first person plural, rico makes riquisimo, not ricisimo in the superlative) and the absence of any such artifice in the ancient tongues. But this rule does not hold in the language most immediately descended from the Latin; it requires no more than a fashionable operagoer's knowledge of Italian to observe that amico makes not amichi but amici in the. plural. Mind, we are not arguing against the uniform pronunciation of the Latin C and Gindeed the correspondence of the former to the Greek is a sufficient proof in its case at least - but only remarking that the analogy so often advanced to support it is imperfect and defective. And Mr. Haldeman quotes a ludicrous instance of a man's English associations misleading him, in Bonnycastle's argument that the Latin could not have been pronunced like w, because it is vulgar to say winegar in English!

The result of these difficulties (and we are not sure that we have enumerated all the difficulties of the case) is that it is not possible to determine satisfactorily the pronunciation of all the Latin letters so as to form a complete system which philologists will generally agree to. In the case of some letters we can attain.to absolute certainty; in the case of others, after all our speculations, we are left in absolute uncertainty; and there are various shades of probability between. Sometimes having decided one letter we can by means of it decide another with all the neatness and accuracy of a mathematical demonstration. The comparison of tu-tu to the cry of the screech-owl, the agreement of most modern languages in their pronunciation of u and its Greek equivalent ov, the absence of any contradictory evidence from any source, all unite to justify us in assuming that the Roman sound of this letter was our oo. Going a little further we find