Each of the first six lines is a group: it is uttered as one thing, and it has one center of interest (star, eagle, spring, dew, wind, bubbles). Against them all stands the next group, "Even such is man", setting forth man as the real subject of reflection, for whose sake all the previous matters have received attention; and the first member is brought to an end with the word to-night, the last group giving the ground for the comparison. Then, "The wind" makes a group, to which "blows out" answers, and the same relation is maintained between bubble and dies, spring and autumn, and the rest; and finally, "Man-forgot" completes the second large member, and brings the sentence to its close.

Now it is not difficult to consider the structure of this sentence under the three conceptions, unity, coherence, and emphasis. Its theme is distinct, its orderly progress markedly clear, and its points of culmination are conspicuous. At the same time, here again that which is emphasis in the part is the means of coherence for the whole; the stressed words of the smaller groups serve as means by which the relation of the groups is made plain. The dominant emphasis, furthermore, on man and forgot is indeed emphasis, but it is also the direct expression of the fundamental idea, or directing unity, of the sentence. Any other well constructed sentence would illustrate with similar distinctness the propositions which have already been put forward under so many aspects.

The physical means by which the formal logical relations of sentences are expressed to the ear, and through the ear to the mind, would make the same contentions almost mechanically obvious. It would be possible to separate these physical means of expression from articulate words, or in other words to utter the melody of the sentence; and if that were done with exactness, the whole logic of the sentence as regards the relations of the parts would be completely and precisely expressed. To do it, it would be necessary to utilize every means always necessary for the existence of a melody;-relative duration, pause, volume, discrete change of pitch, and in addition the sliding change of pitch characteristic of speech as distinct from song. Each element has a special function

as a means of expression; but the emphasis, the means of distinguishing the most significant parts, is obviously nothing but the culmination of a process which runs through the whole melody. If, for example, the emphasis is gained, as it mainly is, by a change in the direction of the slides of the voice, the change is as a physical fact inseparable from the slides on the other syllables. Whatever means of emphasis is in any given case employed, it is always tangibly a part of the whole melody; in other words it is a part of that which expresses the logical relations of the parts, or in rhetorical phraseology the coherence of the sentence. And as a whole, a sentence as a thing heard and addressing itself through the senses to the rational mind is a tune which sounds complete; that is to say, it is one as a process realized through the physical manifestations of its coherence and its emphasis.

The accepted formula will not quite work. It is an imperfectly fitting master-key; sometimes it turns a lock easily; in the next lock it must be humored a little; in another lock it scrapes and grates; and in another it refuses to budge. "Very well," it is natural to say, "after all, why not? These terms were not proposed as philosophic verities, but as convenient rough practical classifications. Professor Wendell found a bewildering confusion of empirical directions for the ordering of material in written compositions, and reduced the chaos to manageable simplicity. The formulas may not provide examples of clean logical division, they may occasionally be contradictory or inapplicable, but on the whole they work. Young people do not take their rhetoric so seriously as to be much hurt by the categories of the books. Why not let well alone?"

The whole course of the present discussion leaves the practical convenience of the three terms where it was. Nothing that is here said gives any grounds for abandoning them; coherence and emphasis remain definite and useful conceptions whether they are thought of as manifestations of unity or as principles co-ordinate with it. Emphasis, moreover, is a real and significant thing, whatever it is; and difficult cases are made not less but more manageable by an accurate defini

tion of it. The fundamental immediate aims of this paper are to do something toward breaking up superficial views upon all aspects of the structure of written composition, and especially to aid in establishing the organic conception of unity for the mechanical conception of it; to cause unity to be thought of as a force and not as a fact, as an energy and not as a mere absence of transgression. Thus only can the mind be led to a recognition of the higher aspects of unity. Important among these, important even in elementary practice as well as in the development of the power of appreciation, is harmony of tone. A crude lack of sensibility to inappropriateness of diction is one of the most frequent faults of young writers, and one of the most difficult to combat. The most lucid and the most cogent method of dealing with this matter is by a thorough explication of the idea of constructive unity, which will also lay a solid basis for the recognition of the high achievement of those writers who have succeeded in developing a higher unity from incongruity itself. Now, there is no place for this aspect of writing as a phase of structure under any of the three heads. There is no place for that imaginative power which is most manifest and most triumphant in reducing opposites to harmony, for the "electric power of combination" which finds in distant places and strange associations and apparent contradictions the support for the highest unity, the unity of conquest. Surely, though such matters may have little place in direct teaching, they should not be out of the range of our general views of style. The conception that a great plot, or a great character, or a great metaphor is made one by a conquering force and not simply that it is one as a fact is worth having for its own sake, and it does not fit into the scheme of the three terms. The idea of unity and variety as exhibited in formal beauty, the beauty of rhythm, the beauty of varied recurrences of any type is out of the range of the ordinary definitions. Yet the theory of prose cannot leave out this aspect of unity in small structures or in large;-in a well balanced play, or a finely turned argument, or a well built period, or a graceful phrase. It is, moreover, a barren theory which excludes

them from its fundamental principles, and only considers them as superficial ornament. Such a theory must either reject beauty as illegitimate, or overvalue mere external adornment.

In all this there is nothing mystical. Rather the contrary. What has been said urges that rhetorical terminology should conform to the usage of ordinary language. In common usage, unity includes severalty, singleness of aim, system, continuity of substance, consistency of temper, definiteness of outline, harmony of impression.

Technical terms derive their justification from their precision and adequacy; and a technical term which falls below common consciousness in precision and adequacy can only increase difficulties, not solve them. The current application of the accepted principles of structure is too crude, too remote from reality, too superficially simple. A theory is needed which, without losing breadth and simplicity in its main ideas, shall be more inclusive in its scope, and more delicately adaptable to the complex phenomena of style in detail. In this paper the attempt has been made to indicate some of the main lines which such a theory must follow. The author hopes that later he may carry on the application of these conceptions to a consistent practical system.

BEOWULF AND THE NIEBELUNGEN COUPLET 1

WILLIAM ELLERY LEONARD

I

Tinker's dissertation, The Translations of Beowulf (Yale Studies in English, XVI, 1903) records six English translations in prose, inclusive of Thorpe's line for line parallel printing, and five in verse. Of the verse translations two are in rhyme. Wackerbarth's (1849), though readable in itself, is in the pseudo-ballad verse of Scott's Marmion, essentially a metrical sophistication in the eighteenth century tradition, and in its musical superficiality as unsuited to the rugged manner of Beowulf in one direction as its surprising adoption by Conington, however skilful, proved it unsuited to the charm, pathos, and intellectual depth of the Aeneid in another direction. Lumsden's in seven-accent couplets, iambic in movement, with caesura frequently varying to either end of the movement, is the measure of Chapman's Iliad, and forceful though it be in retelling the Germanic folk-epic, it has all the metrical associations of the Elizabethan Renaissance -as well as all the unBeowulfian qualities of "ballad verse.'' Three are in "imitative verse," which, if my ear as a versewriter does not deceive me, are not verse, and which, if my judgment as a scholar has not gone all astray, are not imitative either (see Section VI)-a criticism that must apply, if correct in principle, to the otherwise fine work and workmanship of the subsequent version of Gummere. To complete the

The following metrical study forms the Appendix to a new version of Beowulf, to be published shortly.

21881, 2nd Ed. 1883.

So Tinker; and Gummere, The Translation of Beowulf, AJP, VII, 1886. 'Garnett's, J. L. Hall's, and Morris'.

The Oldest English Epic, 1909.