and those, who have heard them pronounced by a native, will readily understand the immense resources, which may thus be placed within the reach of language. I am not, however, aware that these differences of tone have ever been applied to the purposes of construction. There does not seem to have been any relative and subordinate intonation in a sentence; a word had its tone fixed, and this it retained, whatever its position.

Whether the metrical arsis heightened the tone of the syllable on which it fell, has been doubted. Bentley thought it did; but later critics have seen reason to question his opinion; and as it must often interfere with the verbul tone, their objections are entitled to much weight. There are, however, passages in the old grammarians, which favour the notion of there having been some change in the voice. May not the arsis have been marked by a stress, resembling our modern accent? If this were so, the change from the temporal to the accentual rhythm, in the fourth century, would be natural and easy; the same syllable taking the accent in the new rhythm, which (according to Bentley and Dawes) received the arsis in the old. With this exception (if it be one), I know no instance in the Greek and Latin, where an alteration either in the tone or loudness of the voice, has been used for purposes of construction or of rhythm. The tone seems to have been a mere accident of the word; and had no influence on the sentence, further than as it contributed to its harmony. The stress of the voice seems to have been employed solely for the purposes of emphasis; and was certainly considered by Quintilian as reducible to no system, for he leaves the learner to gather from experience, "quando attollenda vel submittenda sit vox." Had the stress of voice been in any way dependent on the construction, its laws might have been readily explained; and would have certainly fixed the attention of a people who scrutinized the peculiarities of their language with so much care.

But though I can find no system of accents like our own, in these kindred languages, yet there are reasons for believing, that our present accentuation has been handed down to us from a very remote antiquity. We find it reduced to a system in our Anglo-Saxon rhythms; and its wide prevalence in the other Gothic dialects, points clearly to an origin of even earlier date. The precision of the laws, which regulated the accents in AngloSaxon verse, is one of the most striking features of their poetry. We find none of those licentious departures from rule, which are so common in the old English, and are occasionally met with, even in our later dialect. It may be questioned, if any primary accent were doubtfult in the Anglo-Saxon; at any rate, the limits of uncertainty must have been extremely narrow.

In modern usage, we sometimes hear a word accented, though it immediately adjoin upon an accented syllable; especially when it contains a long vowel-sound. The rhythm of Sackville's line,

Their great cruelty and the deepe bloodshed

Of friends

is not without example, in the every-day conversation of many persons, who have accustomed themselves to a slow and emphatic mode of delivery. Were this practice generally sanctioned by that of our earlier and more perfect dialect, we might infer, with some plausibility, that our English accents were at one time, like those of the Greek and Latin, strictly verbal; and that the sectional pause was a consequence, which followed naturally from the system of accentuation, originally prevalent in our language.

But

*The widest departure from the common rhythm of the language which the Anglo-Saxon poet allowed himself, was owing to the frequent use of the sectional pause. We shall have more to say on this head shortly.

†There are perhaps instances, in which the same sentence has been differently accentuated. But this may be owing to a difference of dialect. The Anglo-Saxon author is, I believe, always consistent with himself.

there are grounds for believing, that in the Anglo-Saxon the stress on the adjective was always subordinate to that on the substantive. In nine cases out of ten, it was clearly subordinate; in no case is it found predominant ;* and when with the aid of the sectional pause, it takes the accent, there is, in the great majority of cases, an evident intention on the part of the poet, to use the pause for the purposes of emphasis-the substantive, in all probability, still keeping the stronger accent. There are, indeed, instances of the sectional pause, where it is certainly not used as an emphatic stop; but these, I believe, are, for the most part, found in poems of inferior merit, or in those artificial rhythms † which were probably invented in the course of the ninth and tenth centuries. They may perhaps be laid to the account of carelessness or of incapacity, and ranked with those cases, where the ordinary rhythm of the language has been made to yield to the rhythm of its poetry. These exceptions may shake, but I do not think they are sufficiently numerous to overturn, the hypothesis that has been started.

Having thus given the reasons, which incline me to the opinion already stated as to the origin of the pause, I shall now proceed to range in order, those sections into which it enters. If we consider the pause as filling the place of an unaccented syllable, we may use nearly the same notation to indicate the rhythm, as hitherto. We have merely to show the presence of the pause, by the addition of a p. Thus the section we have already quoted

from Sackville,

Their great crueltie.

would be represented by the formula, 5 ll. p.

* When the adjective has a stronger accent than its substantive, it always forms part of a compound, and is no longer subject to inflexion.

Conybeare's rhiming poem, for example.

291

THE SECTION 1p. OF TWO ACCENTS.

Sections, which admit the pause, may be divided into two classes, accordingly as they contain two or three accents. When the section contains only two, the pause cannot change its position, for it must fall between the accented syllables; but as the section may vary both its beginning and its end no less than three different ways, it admits of nine varieties. Of these six have established themselves in English literature, to wit, 1. p. 1 l. p. 1 ll. p. 5. p. 5 l. p. 5 ll. p.

Whether the section 1. p. were known in Anglo-Saxon, is a matter of some doubt. In Beowulf, there is the couplet,

Spræc tha ides Scyldinga.

Spake then the Scylding's Lady

and in Cædmon, 148, we have,

Thy læs him westengryre,

Har hæth: holmegum wederum
Oferclamme.

Lest them the desert-horror

The hoar heath-with deluging storms
O'erwhelm.

The lengthened section, 1 l. p. is somewhat more com

The section 1 p. was never common. It was chiefly used by our dramatists; and more particularly in their faëry dialect.

This is the only instance of the section in Milton, who doubtless borrowed it from Fletcher. The propriety of Shakespeare's rhythm will be better understood, if we suppose (what was certainly intended) that the fairy is pouring the love-juice on the sleeper's eye, while he pronounces the words, "Thou tak'st." The words form, indeed, the fairy's "charm," and the rhythm is grave and emphatic as their import. I cannot think, with Tyrwhitt, that the line would be improved, "both in its measure and construction, if it were written thus:

See thou tak'st]."

I know not how the construction is bettered, and the correspondence, no less than the fitness of the numbers, is entirely lost. Seward, in like manner, took compassion upon the halting verses of Fletcher. His corrections afford us an amusing specimen of conjectural criticism.

Let your odour drive | from henee

All mistes that dazzle sense!