14 one-syllable intervals, 76 two-syllable intervals, 85 of three syllables, 16 of four syllables. Stevenson, accordingly, has a great excess of iambic, or more accurately, "duple" feet. Cooper has more triple feet than duple; but the excess of the one over the other is not so marked as the excess of duple over triple in Stevenson. From these figures we conclude that Stevenson has a marked duple rhythm; that Cooper's rhythm is not as marked as Stevenson's, and that what there is of it is predominantly triple in character. It should be remembered, by the way, that we are here speaking only of phonetic or syllabic rhythm. We shall speak of other rhythms later. Comparing Johnson with Emerson in the same way, we find in Johnson 77 one-syllable intervals, 87 two-syllable, 78 three-syllable, 37 foursyllable; in Emerson 106 one-syllable, 111 two-syllable, 65 threesyllable, and 23 four-syllable. In neither of these two cases is the predominance of one foot over the other as clear as in the case of Stevenson; but see how the large feet come out in Johnson and the small in Emerson. Again, Burke's figures are 75, 98, 74, 33, but Ingersoll's are 135, 91, 55, 28-strong double rhythm in Ingersoll, weak triple rhythm in Burke. Ruskin's "Modern Painters" gives 141, 87, 66, 16; his "Sesame and Lilies" 95, 96, 64, 30-strong duple rhythm in the one, weaker rhythm in the other. Huxley's Preface to his Physiography gives 84, 105, 65, 34; his first chapter III, 120, 57, 22—the much greater rhythm of the first chapter is obvious. Spencer gives 51, 75, 63, 34; a clear expression of the commonly felt weakness of rhythm in Spencer.

The illustrations and figures that have been cited clearly do not prove that in one piece of prose we have pages of blank verse, and in another, pages of hexameters. They show only, that short flights of one kind of rhythm predominate in one style; and in another, flights of a different kind. The summation or resultant of these bits in a piece of prose gives it its characteristic syllabic rhythm.

But now let us turn to that other sort of rhythm mentioned a little while ago the rhythm that corresponds more closely to thought phrase, clause and sentence rhythm.

We have given a few illustrations of rhythm from the English

Bible. The same examples illustrate the larger rhythm, the rhythm of thought, that we now have in mind.

Mischief shall come upon mischief, and rumor shall be upon rumor.

The break in the middle, and the similarity of the thought in the two halves is plain. Biblical scholars call this "parallelism of the members," and Hebrew poetry throughout is constructed on this principle. The parallelism may be of various kinds.

How art thou fallen from Heaven, O Lucifer, son of the morning!

is evidently not a parallelism, like the first one cited, and

Be as the sands of the sea, which cannot be measured or numbered.

is different again, as is

Cease then and let me alone that I may take comfort a little.

Now, the principle of parallelism which to-day seems so obvious not only in the poetry of the Bible, but in a great deal of its prose, was not recognized and formulated by the translators of the King James version, and, considering the acknowledged influence of the King James Bible on English prose, it will not be too much to say that the principle of parallelism, which is the formal principle on which Hebrew poetry was constructed, has come to play an important part in the structure of English prose. But whether or not acquired in this way, the point here is, that the form which English rhetoricians call "balance" is in essence identical with that called by Biblical scholars "parallelism." Furthermore, this form of rhythm is far more prevalent in English than is commonly supposed.

When Macaulay says, "He has distanced all his competitors so decidedly, that it is not worth while to place them," there is no less parallelism than when Job says, "Cease then and let me alone, that I may take comfort a little." When Stevenson writes, "Nothing could overlie such a fellow; a kind of base success was written on his brow," the parallelism is as obvious as in the Psalm, "God is gone up with a shout, the Lord with the sound of a trumpet." And in the rest of Stevenson's paragraph from which this sentence is taken, the parallelism is no

less clear: "He was then in his ill days | but I can imagine him in Congress with his mouth full of bombast and sawder. As we moved in the same circle | I was brought necessarily into his society. I do not think I ever heard him say anything that was true, kind or interesting | but there was entertainment in the man's demeanor."

We can go farther, guided by a clue from the Bible. A Biblical parallelism may subsist between subject and predicate, indicated in the Hebrew text by an especial punctuation mark. "And the earth | was waste and void," "A laughing stock | has God made me," are examples. We may have the same thing in English. The middle point will be indicated by a pause and the two members will contain exactly, or very nearly, the same number of accents; as in these sentences from Lamb: "Awful ideas of the Tower | twined themselves about his presence;" "A captive, a stately being | let out of the Tower on Saturdays.” The balance between subject and predicate is especially marked when the subject is long. The voice in reading marks the point of division by a change of pitch and a suspension of tone, as may be felt clearly in this sentence by Ruskin: "A Claude, a Murillo, a Greuze and a Gainsborough | hung gracious in their chosen places." Rhythmical prose places corresponding or proportionate masses of sound, measured chiefly by accents, on each side of the dividing line, as it does also in the usually recognized cases of balance. So, for instance, in this double system from Macaulay: "The party whose principles afforded him no guarantee | would be attached to him by interest, | the party whose interests he attacked would be restrained from insurrection by principle."

It seems rash to suggest the existence of elements in prose analogous to the verse units of poetry, yet the thing has been done by an excellent authority, and close study shows that there is a good basis for the intimation. The prose units that have been likened to verses are phrasal sections.

We clearly indicate such sections in reading aloud, by pauses and intonations. There is no mistaking them in this sentence of Gibbon's: "He was severe to himself, indulgent to the imperfections of others, just and beneficent to all mankind." The

fact of an analogy between phrasal section in prose and verse in poetry suggests the question whether a uniform length of phrasal section is one of the devices of rhythmical prose. By length, we should not understand length in inches, in letters, or in syllables; but in number of accents. The use of this measure needs but little defense. It is the recognized measure in certain types of poetry. Anglo-Saxon poetry paid no attention to syllables, but only to accents. This is said to be the practice today in Palestine and Arabia; and Coleridge, as is well known, wrote "Christabel" on this principle.

The answer to the question put above may be sought by the statistical method. We take the texts that we have scanned and again go over them, marking the natural pauses which we assume indicate the termini of phrasal sections. We count the number of accents in each phrase and then the number of phrases having one, two, three accents, and so on, respectively. We finally calculate what percentage of all the phrases in each selection contain one accent, two accents, three accents, etc. It will be found that on the whole, phrases of two and three accents abound nearly equally, with rather more of three accented phrases than of two. So far, then, the evidence is inconclusive. There are, however, a few exceptions. Sir Thomas Browne's piece of prose will be found to have about 44 per cent of threeaccented phrases and only 24 per cent of two-accented. Contrast this with Ingersoll's 36 per cent of two-accented, and 26 per cent of three-accented phrases. Lamb, who modelled his style on Browne's, approximates him curiously with 28 per cent of twos and 41 per cent of threes; and Emerson, though so different in other rhythmical features, approaches these writers with 22 per cent of twos and 35 per cent of threes. From such results as these, and from what we have seen above of parallelism, it seems fair to conclude that phrases of equal accents do play a part in the rhythm of prose.

Among the varieties of phrase rhythm which we are cataloguing, should finally be mentioned those repetitions of similar forms with varied meanings; like Burke's "The unbought grace of life, the cheap defense of nations, the nurse of manly sentimen tsand heroic enterprise is gone" Reiteration of the same

part of speech, often buttressed with epithets, is another device. We find it in the flashiest and in the soberest writing. Thus, Ingersoll has, "I would rather have lived and died unnoticed and unknown except by those who loved me, and gone down into the voiceless silence of the dreamless dust, than to have been that imperial impersonation of force and murder who covered Europe with blood and tears. We get the same sort of thing more quietly in Howells, "and really the place was enchanting with its close-cropped, daisy-starred lawns, and the gay figures of polo players coming home from distant fields in the pale dusk of a radiant day of early June."

It is evident that rhythm in prose is a much more complicated affair than rhythm in poetry. The instrument, as has been said, is more difficult to play worthily. It requires a great master to extract all the music of which it is capable. Besides the syllabic rhythm, the phrase and word rhythm, the parallelism of clauses and phrases that have been mentioned, there are other effects rhythmic and melodic on which we cannot here dwell. For the sake of approximate completeness let it be added only, that the statistical method has been applied to sentences as wholes and that specific styles have been found to possess characteristic sentence lengths. There is, that is to say, also a sort of rough sentence-rhythm.

There will undoubtedly be persons who will think too much has been proved, or claimed, for prose rhythm. You might as well assert that all language is rhythmical, they will say, and be done with it. Something very nearly like this assertion is just what we should desire to make. All language is to a large extent rhythmical. Like all other bodily activity in which similar movements are repeated at brief intervals of time, speech tends towards rhythm and approaches regularity of rhythm as closely as the phonetic character of the words, all things considered, permits. The artistic prose writer moulds the material in which he works to heighten rhythmical effects and suppress roughnesses already there. He, like the poet, "so selects and arranges words, that the reader will find strongly stressed syllables coming naturally into the majority of the more prominent times of the desired rhythm." The difference be