moderns especially, whose style, by inverfions too violent, is rendered harsh and obfcure. Sufpenfion of the thought till the close of the period, ought never to be preferred before perfpicuity. Neither ought fuch fufpenfion to be attempted in a long period; because in that cafe the mind is bewildered amidst a profufion of words: a traveller, while he is puzzled about the road, relishes not the finest profpect:

All the rich presents which Aftyages had given him at parting, keeping only fome Median horses, in order to propagate the breed of them in Perfia, he distributed among his friends whom he left at the court of Ecba

tana.

Travels of Cyrus, Book 1.

The foregoing rules concern the arrangement of a fingle period: I add one rule more concerning the diftribution of a difcourfe into different periods. A fhort period is lively and familiar a long period, requiring more attention, makes an impreffion grave and folemn*. In general, a writer ought to study a mixture of long and short periods, which prevent an irksome uniformity, and entertain the mind with variety of impreffions. In particular, long periods ought to be avoided till

F 4

* Demetrius Phalereus (of Elocution, fect. 44.) obferves, that long members in a period make an impreffion of gravity and importance. The fame obfervation is applicable to periods.

till the reader's attention be thoroughly engaged; and therefore a difcourfe, especially of the familiar kind, ought never to be introduced with a long period. For that reafon, the commencement of a letter to a very young lady on her marriage is faulty:

Madam, The hurry and impertinence of receiving and paying visits on account of your marriage, being now over, you are beginning to enter into a course of life, where you will want much advice to divert you from falling into many errors, fopperies, and follies, to which your fex is fubject.

Swift.

See another example ftill more faulty, in the commencement of Cicero's oration, Pro Archia poeta.

Before proceeding farther, it may be proper to review the rules laid down in this and the preceding fection, in order to make fome general obfervations. That order of the words and members of a period is juftly termed natural, which correfponds to the natural order of the ideas that compose the thought. The tendency of many of the foregoing rules is to fubftitute an artificial arrangement, in order to catch fome beauty either of found or meaning for which there is no place in the natural order. But feldom it happens, that in the fame period there is place for a plurality of thefe

these rules: if one beauty can be retained, another must be relinquished; and the only queftion is, Which ought to be preferred? This question cannot be refolved by any general rule: if the natural order be not relished, a few trials will discover that artificial order which has the best effect; and this exercife, fupported by a good tafte, will in time make the choice easy. All that can be faid in general is, that in making a choice, found ought to yield to fignification.

The tranfpofing words and members out of their natural order, fo remarkable in the learned languages, has been the subject of much speculation. It is agreed on all hands, that fuch tranfpofition or inverfion beftows upon a period a very fenfible degree of force and elevation; and yet writers feem to be at a lofs how to account for this effect. Cerceau afcribes fo much power to inverfion, as to make it the characteristic of French verfe, and the fingle circumftance which in that language diftinguishes verse from prose: and yet he pretends not to fay, that it hath any other effect but to raife furprife; he must mean curiofity, which is done by fufpending the thought during the period, and bringing it out entire at the clofe. This indeed is one effect of inverfion; but neither its fole effect, nor even that which is the most remarkable, as is made evident above. But waving cenfure, which is not an a

greeable

* Reflections fur la poëfie Françoise.

greeable task, I enter into the matter; and begin with obferving, that if conformity between words and their meaning be agreeable, it must of course be agreeable to find the fame order or arrangement in both. Hence the beauty of a plain or natural ftyle, where the order of the words correfponds precifely to the order of the ideas. Nor is this the fingle beauty of a natural ftyle it is alfo agreeable by its fimplicity and perfpicuity. This obfervation throws light upon the subject: for if a natural ftyle be in itfelf agreeable, a tranfpofed ftyle cannot be fo; and therefore its agreeableness muft arise from admitting fome pofitive beauty that is excluded in a natural ftyle. To be confirmed in this opinion, we need but reflect upon fome of the foregoing rules, which make it evident, that language by means of inverfion, is fufceptible of many beauties that are totally excluded in a natural arrangement. From thefe premifes it clearly follows, that inverfion ought not to be indulged, unless in order to reach fome beauty fuperior to thofe of a natural ftyle. It may with great certainty be pronounced, that every inverfion which is not governed by this rule, will appear harsh and ftrained, and be difrelished by every one of tafte. Hence the beauty of inverfion when happily conducted; the beauty, not of an end, but of means, as furnishing opportunity for numberlefs ornaments that find no place in a natural ftyle: hence the force, the elevation,

the

the harmony, the cadence, of fome compofitions : hence the manifold beauties of the Greek and Roman tongues, of which living languages afford but faint imitations.

SECT. III.-Beauty of Language from a Refemblance between Sound and Signification.

A

Refemblance between the found of certain words and their fignification, is a beauty that has escaped no critical writer, and yet is not handled with accuracy by any of them. They have probably been of opinion, that a beauty so obvious to the feeling, requires no explanation. This is an error; and to avoid it, I fhall give examples of the various resemblances between found and fignification, accompanied with an endeavour to explain why fuch refemblances are beautiful. I begin with examples where the refemblance between the found and fignification is the most entire; and next examples where the refemblance is lefs and lefs fo.

There being frequently a ftrong resemblance of one found to another, it will not be furprifing to find an articulate found resembling one that is not articulate thus the found of a bow-ftring is imitated by the words that express it:

The