vivacity of expression, cannot be denied. But, as they give, when too frequent, a stiff and finical air to a performance; as they are not always explicit in the sense, nor agreeable in the sound; as they are apt to produce a confusion, or too great a multiplicity of images; as they tend to disfigure the language, and furnish a pretext for endless innovation; I would have them used sparingly; and those only used, which the practice of popular authors has rendered familiar to the ear, and which are in themselves peculiarly emphatical and harmonious. For I cannot think, with Dacier and Sanadon, that this well known verse in Horace's Art of Poetry,

Dixeris egregie, notum si callida verbum
Reddiderit junctura novum

gives any warrant, even to a Latin poet, for the formation of these compound words; which, if I mistake not, were more fashionable in the days of Ennius, than of Horace or Virgil.*

The criticks are divided about the meaning of this passage. Horace is speaking of new words; which he allows to be sometimes necessary: but which, he says, ought to be sparingly and cautiously introduced; In verbis etiam tenuis cautusque serendis; and then subjoins the words quoted in the text, Dixeris egregie, &c.

1. Some think, that this callida junctura refers to the

7. In the transformation of nouns into verbs and participles, our poetical dialect admits of greater latitude than prose. Hymn, pillow, curtain, story, pillar, picture, peal, surge, cavern,

formation of compound epithets, as velivolus, saxifragus, solivagus, &c. and that the import of the precept is this: "Rather than by bringing in a word altogether new, even "when a new word is necessary, you should express "yourself by two known words artfully joined together "into one, so as to assume a new appearance, and to ad"mit a new though analogical signification." This might no doubt be done with propriety in some cases. But I cannot think, that Horace is here speaking of compound words. For, first, this sort of words were much more suitable to the genius of the Greek than of the Latin tongue; as Quintilian somewhere insinuates, and every body knows who is at all acquainted with these languages. Secondly, we find in fact, that these words are less frequent in Horace and Virgil, than in the older poets; whence we may infer, that they became less fashionable as the Latin tongue advanced nearer to perfection. Thirdly, Virgil is known to have introduced three or four new words from the Greek, Lychni, Spelea, Thyas, &c. but it does not appear, that either Virgil or Horace ever fabricated one of these compound words; and it is not probable, that Horace would recommend a practice, which neither himself nor Virgil had ever warranted by example. Fourthly, our author, in his illustrations upon the precept in question, affirms, that new words will more easily obtain currency if taken from the Greek tongue; and Virgil, if we may judge of

honey, career, cincture, bosom, sphere, are common nouns; but, to hymn, to pillow, curtained, pillared, pictured, pealing, surging, cavern'd, honied, careering, cinctured, bosomed, sphered, would ap

his opinions by his practice, appears to have been of the same mind. And there was good reason for it. The Greek and Latin are kindred languages; and as the former was much studied at Rome, there was no risk of introducing any obscurity into the Roman language by the introduction of a Greek word. Lastly, it may be doubted, whether junctura, though it often denotes the composition of words in a sentence or clause (Quintil. ix. 4.), and sometimes arrangement or composition in general (Hor. Ar Poet. verse 242.) is ever used to express the union of syllables in a word, or of simple words in a compound epithet.

2. Other interpreters suppose, that this callida junctura refers to the arrangement of words in the sentence, and that the precept amounts to this: "When a new ex"pression is necessary, you will acquit yourself well, if

by means of an artful arrangement you can to a known "word give a new signification." But one would think, that the observance of this precept must tend to the utter confusion of language. To give new significations to words in present use, must increase the ambiguity of language, which in every tongue is greater than it ought to be, and which would seem to be more detrimental to eloquence and even to literature, than the introduction of many new words of definite meaning. Those who favour this interpretation give come sylvarum for folia, as a phrase to exemplify the precept. But

pear affected in prose, though in verse they are warranted by the very best authority.

Some late poets, particularly the imitators of Spenser, have introduced a great variety of un

be

the foilage of a tree is not a new idea, nor could there any need of a new word or new phrase to express it: though a poet, no doubt, on account of his verse, or on some other account, might choose to express it by a figure, rather than by its proper name. Come sylvarum for folia, is neither less nor more than a metaphor, or, if you please, a catachresis; but Horace is speaking, not of figurative language, but of new words. Both these interpretations suppose, that the words of our poet are to be construed according to this order: Dixeris egregie, si callida junctura reddiderit notum verbum novum. 3. The best of all our poet's interpreters, the learned Dr. Hurd, construes the passage in the same manner, and explains it thus: "Instead of framing new words, I "recommend to you any kind of artful management, by "which you may be able to give a new air and cast to "old ones." And this explication he illustrates most ingeniously by a variety of examples, that throw great light on the subject of poetical diction. See his notes on the Ars Poetica.

I should ill consult my credit, if I were to oppose my judgment to that of this able critick and excellent author. Yet I would beg leave to say, that to me the poet seems, through this whole passage, from vers. 46. to vers. 72. to be speaking of the formation of new words; a practice whereof he allows the danger, but proves the necessity. And I find I cannot divest myself of an old prejudice

common words, as certes, eftsoons, ne, whilom, transmew, moil, fone, losel, albe, hight, dight, pight, thews, couthful, assot, muchel, wend, arrear, &c. These were once poetical words, no doubt; but they are now obsolete, and to many readers unintelligible. No man of the present age, however conversant in this dialect, would naturally express himself in it on any interesting emergence; or, supposing this natural to the antiquarian, it would never appear so to the common hearer or reader. A mixture of these words, therefore, must ruin the pathos of mod

in favour of another interpretation, which is more obvious and simple, and which I considered as the best, long before I knew it was authorized by that judicious annotator Joannes Bond, and by Dryden in his notes upon the Eneid, as well as by the Abbe Batteux in his commentary on Horace's Art of Poetry. "New words (says the poet) are to be cautiously and sparingly in“troduced; but, when necessary, an author will do well "to give them such a position in the sentence, as that "the reader shall be at no loss to discover their mean"ing." For I would construe the passage thus, Dixeris egregie, si callida junctura reddiderit novum verbum notum. But why, it may be said, did not Horace, if this was really his meaning, put novum in the first line, and notum in the second? The answer is easy. His verse would not admit that order: for the first syllable of novum is short, and the first syllable of notum long.