De Quincey in the language of elaborate stateliness, Macaulay in the language of brilliant energy.

Here it may be well to point out-and the caution is of such importance that it may have to be repeated—that the divisions in the following analysis are not, in the language of the logicians, mutually exclusive. Following Professor Lain's Rhetoric we consider style under three different aspects-approach it from three different sides; but we do not treat of different things. In each of the divisions the same things are examined, only from different points of view. Each of these divisions, were our examination to be ideally thorough, should exhibit every possible excellence and defect of style. We might take up all the notable points in an author's style under what we have called the "Elements of Style" -the choice of words, plain and figurative, and the arrangement of these in sentences and paragraphs. We might, again, take up everything remarkable under the "Qualities of Style"-simplicity, clearness, and so forth a style is good or bad according as it produces, or fails to produce, certain effects. Finally, we might comprehend the whole art of style under the "Kinds of Composition": every excellence of style is either good description, good narration, good exposition, good persuasion, or good poetry. The divisions are far from being mutually exclusive. Were we to say in one department all that might be said, we should leave nothing for the others. The sole justification of having three, and not one, is practical convenience. There must of necessity be occasional repetitions, but each department has certain arts of style that are best regarded from its own particular point of view.

THE SENTENCE.

The construction of sentences is an important part of style. Sometimes, indeed, it is expressed by the word style, as if it constituted the whole art. With a nearer approach to accuracy, it is sometimes called the mechanical part of style. This designation may be allowed, if sentence-building is loosely taken to include the construction of paragraphs and the general method of a discourse. It is probably true that the construction of sentences and of paragraphs, in so far as they are intended for the communication of knowledge, may be subjected to more precise rules than any other processes of the art of composition. The principles on which these rules are founded are capable of extension to the method of whole chapters or essays. But it must be borne in mind that a writer can benefit from direct precept chiefly as regards the easy, clear, and complete communication of what is in his thoughts; for any effect of style beyond this, precepts are of comparatively little service.

SPECIAL ARTIFICES OF CONSTRUCTION.-One may doubt whether it would be practicable to make anything like a comprehensive

collection of all the forms of sentence possible in English. At any rate, it has not yet been done. Writers on composition have hitherto attempted nothing more than to distinguish a few wellmarked modes of construction.

I. The Periodic Structure.-"A period," says Campbell, "is a complex sentence, wherein the meaning remains suspended till the whole is finished. .. The criterion of a period is this: If you stop anywhere before the end, the preceding words will not form a sentence, and therefore cannot convey any determined sense." This is the common definition of a period, and it is probably difficult to go farther without committing one's self to general statements that will not apply to every period. At the risk of being slightly inaccurate, it might be well to go a little deeper into the substance of the periodic structure. What exactly do we imply by saying that the meaning is suspended till the close? We imply that the reader's interest is kept in suspense till the close. And how is this done? Generally, it may be said, by bringing on predicates before what they are predicated of, and, which is virtually a similar process, qualifications before what they qualify; letting us know descriptive adjuncts, results, conditions, alternatives, oratorical contrasts, of subjects, states, or actions, before we formally know the particular subjects, states, or actions contemplated by the writer. Thus, in the following sentence

"On whatever side we contemplate Homer, what principally strikes us is his invention;'

the subject-in this case the key-word-is reserved to the last, and the adverbial adjuncts of the predicate are stated before the predicate itself. A statement is made in a form showing that it has a bearing upon something to follow, and our curiosity is awakened to know what that something is. "On whatever side we contemplate Homer." The next statement, "what principally strikes us," contracts our curiosity into a more definite field, and thereby sharpens our interest. Still it points us forward. There is a progress from the indefinite to the definite, and, in the case of this particular period, a growing interest, which is not relieved till we reach the very last word. In a loose structure of sentence, which may be called the natural or usual structure in English, the predicate follows the subject, and qualifying adjuncts follow what they qualify we know the subject before we know the attribute predicated of it or annexed to it. In a period, on the other hand, the writer, stating the predicate or qualifying adjuncts of a word before the word itself, may be said to circumvent that word—to make (as the Greek periodos signifies) a "circuit" about it, to bring its predicate or its adjuncts, as the case may be, from behind it and place them before it.

Campbell speaks of the period as a "complex" sentence. If the above view of the period is accepted as substantially correct, "complexity," in the grammatical sense, must be regarded as an accident of the period, and not part of its essence. The statements of other writers on composition warrant us in applying the term period to sentences that are not complex. Professor Bain simply says that, "in a periodic sentence, the meaning is suspended until the close," and makes no mention of a periodic sentence being necessarily complex. And Whately gives, as an example of periodic structure, the following "simple" sentence: "One of the most celebrated of men for wisdom and for prosperity was Solomon."

It would be well if the application of the term periodic were a little extended. When qualifying adjuncts are brought in before their exact bearing is known, and in such a way as to stimulate curiosity, a peculiar effect is produced; and we should be justified by the derivation of the word "periodic" in applying it to all marked cases of such anticipation. Practically, indeed, the word is applied in the wider sense. If Campbell's definition were rigorously adhered to, the term periodic could be applied only to sentences that keep the reader in suspense up to the very last word. But, as a matter of fact, the term is applied much more widely. We speak of writers as having a periodic style, although their works contain few complete periods, according to Campbell's "criterion of a period." Since, therefore, the narrow definition of the term is practically disregarded, it would be well to come to a formal understanding of its latitude. The term "period" might still be retained for a periodic sentence, rigorously complete or nearly so. But it would probably better suit the prevailing application of the term "periodic" to accept it as a name for such anticipations as I have roughly indicated to call every style "periodic" where such anticipations habitually occur. this periodic style, the most eminent of modern masters is De Quincey.

Of

In the loose sentence - - in a sentence so constructed as to be noticeably "loose"-qualifying and explanatory adjuncts are tacked on after the words they refer to. This might be copiously exemplified from the writings of Carlyle, and, in a less degree, from Addison.

The effect of the periodic structure is to keep the mind in a state of uniform or increasing tension until the dénouement. This is the effect stated in its ultimate and most general form. The effect that a reader is conscious of receiving varies greatly with the nature of the subject-matter. When the subject is easy and familiar, the reader, finding the sentence or clause come to an end as soon as his expectations are satisfied, receives an agreeable impression of neatness and finish. When the subject-matter is un

familiar, or when the suspense is unduly prolonged, the periodic structure is intolerably tedious, or intolerably exasperating, accord ing to the temper of the reader. In impassioned writing the period has a moderating effect, the tension of the mind till the key-word is reached preventing a dissipation of excitement.

Dr Blair says that the periodic style is "the most pompous, musical, and oratorical manner of composing," and that it "gives an air of gravity and dignity to composition." The Doctor probably had in his eye such periodic writers as Hooker, Sir Thomas Browne, and Johnson. Undoubtedly long periodic sentences give great scope for pomp, music, gravity, dignity, and such effects, but these are not necessary attributes of the period. A period, as we have defined it, need not be long; and a lively interest may be sustained as well as a grave interest.

Advantages and disadvantages of the periodic structure.-To some extent we have anticipated these in considering the effect of the period.

In light subjects, neatness or finish is generally regarded as an advantage. Yet even in this a caution is needed; rounded neatness, if it recurs too frequently, may become tiresome. The caution can probably be given in no more definite form than Hamlet's: "Be not too periodic neither, but let your own discretion be your tutor."

In unfamiliar subjects, care must be taken that the considerations kept in suspense be not too numerous or too abstruse. De Quincey has vividly described "the effect of weariness and repulsion which may arise from this single vice of unwieldy comprehensiveness in the structure of sentences." "Those who are not accustomed to watch the effects of composition upon the feelings, or have had little experience in voluminous reading pursued for weeks, would scarcely imagine how much of downright physical exhaustion is produced by what is technically called the periodic style of writing. It is not the length, the deparтoλoyía, the paralytic flux of words. It is not even the cumbrous involution of parts within parts, separately considered, that bears so heavily upon the attention. It is the suspense, the holding on of the mind until what is called the ȧmódonis, or coming round of the sentence, commences. This it is which wears out the faculty of attention. A sentence, for example, begins with a series of ifs; perhaps a dozen lines are occupied with expanding the conditions under which something is affirmed or denied. Here you cannot dismiss and have done with the ideas as you go along; for as yet all is hypothetic-all is suspended in air. ditions are not fully to be understood until you are acquainted with the dependency: you must give a separate attention to each clause of this complex hypothesis, and yet, having done that by a

The con

painful effort, you have done nothing at all; for you must exercise a reacting attention through the corresponding latter section, in order to follow out its relations to all parts of the hypothesis which sustains it." These remarks point to the abuse rather than to the use of the periodic style, and were directed against a prevailing style in newspaper "leaders." It is obviously necessary, for the avoiding of perplexity, not to bring in too many or too abstruse considerations before their bearing is made known. A writer with the least regard for his readers, should see that by so doing he exacts too severe an effort of attention. It may safely be laid down that the longer a period is, the simpler should be both the language and the matter of the suspended clauses. Still more must this be kept in view when the principle of the periodic structure is extended to paragraphs or chapters.

Mr Herbert Spencer in his Essay on the Philosophy of Style,' and Professor Bain in his Rhetoric, advocate what we have defined as periodic structure, on the ground that it enables us to apprehend the meaning of a complex statement with, less risk of confusion. The advantage of placing qualifying words before the object that they qualify is briefly stated in Bain's Rhetoric, under the "order of words."

The legitimate use of the periodic structure in impassioned prose is best seen in the so-called "prose fantasies" of De Quincey.

II. Sentences studiously Long and studiously Short.-No small element in the mechanical art of sentence-building is the adjustment of the length of the sentence. One of the greatest faults in our early writers is that their sentences are too long. They did not know when to stop. They seem to have been afraid to let a sentence out of their hands till they had tacked on all the more important qualifications of the main statement. They thus frequently ran on to a most cumbrous length; and when they did proceed to a new sentence, frequently took no pains to connect it with the preceding main statement, but started off in pursuit of some subordinate idea suggested by one of the qualifying statements. So defective, indeed, were they in sentence-structure, that it is dangerous for a beginner in composition to spend much time in their company. And one great part of this deficiency was, that they did not know when to end a sentence, or, in other words, that they had not the art of beginning a new sentence at the proper point.

It would be absurd to prescribe a definite limit for the length of sentences, or even to say in what proportion long and short should be intermixed. Here, too, discretion is the tutor. Only it must be borne in mind that a long series either of very short sentences or of very long sentences is tiresome.

The distinction between the "periodic style" (style périodique)