(b) The second colon tends to be more compact than the first. Four points may be adduced to show this: (1) 2579 words occur all told (each word is counted as many times as it occurs); of these 1354 are in the first colon, but 1225 (viz. 129 less) in the second colon. (2) The total number of hexasyllables used is 6 and all of them stand in the second colon. (3) The total number of pentasyllables used is 19, of which all but 7 stand in the second colon. (4) The total number of monosyllables used 423. If our thesis is true, the proportions as compared with the foregoing should here be reversed, and this turns out to be the fact, for 293 stand in the first colon and only 130 in the second. A Lesser Asclepiad then, despite its fixed sequence of feet, accords with other kinds of verse, which as a rule exhibit greater compactness, less opportunity for substitutions, and more regular sound-effects in the final part. The poet's feeling as revealed in Lesser Asclepiads accords with the feeling of the ancients about the arrangement and relative length of cola in a well constructed oration. See Cicero De Or. III, 48 fin. 50.

(c) The compactness of the verse as a whole varies somewhat according to the structure of the strophe into which it enters. Two points of evidence will make this clear. (1) Sense-pauses, so far as they are revealed by punctuation (see p. 290), occur after the 1st syl. of the verse only in those kinds of strophes where three or more Lesser Asclepiads come in succession, viz. in strophes I and III. (2) Strophe I shows Lesser Asclepiads that are more compact than those figuring in strophes II, III, or IV. The Lesser Asclepiads in strophe IV, for example, when compared with those forming strophe I, show relatively

Io per cent more cases where a word ends with the 1st syl.

2 per cent more cases where a word ends with the

9th syl.

Only the 3d syl. shows a decrease and that a slight one. The above thesis is explained, in part, by the fact that into the said three strophes there enters a variety of verses, as regards the number and arrangement of feet, whereas strophe I is composed exclusively of Lesser Asclepiads, complex and intricate metrical demands being most readily satisfied with short words. Contrast for example the elegiac couplet and the heroic verse: the preponderance of long words is generally to be found in the latter.

(d) The compactness of the verse varies according to the period in Horace's life when it was composed. It becomes gradually less compact, the change affecting first the forward colon (viz. in Book III) and finally the second colon (viz. in Book IV).

The Lesser Asclepiads in Book III, when compared with those in Book I, show an appreciable tendency to have words end more frequently with the

This more than offsets the fact that words end less frequently in the

The change, then, from Book I to Book III is toward a less compact verse, especially in the first colon, diaereses and caesuras increasing particularly after the 2d and 4th syllables.

The Lesser Asclepiads in Book IV, when compared with those in Book III, show an appreciable tendency to have words end more frequently with the

This far more than offsets the fact that words end less frequently with the 8th syl. (decrease: 8 per cent). The change, then, from Book III to Book IV is toward a less compact verse, especially in the second colon, diaereses and caesuras increasing particularly in

the 7th, 9th, 10th, and 11th syllables. This thesis may properly be considered in locating an ode of doubtful date.

(e) Taking into account all possible arrangements of diaereses and caesuras in a Lesser Asclepiad, we find that Horace chose to employ but few of them. The first colon, for example, in the form monosyllable, quadrisyllable, monosyllable was avoided. (For the reason, see p. 292). Only three of the 509 verses begin with a monosyllable followed by a quadrisyllable, and in two of these Horace does not write another monosyllable as the third word but a thing that nowhere else occurs in his Lesser Asclepiadshe omits the diaeresis between the cola. These two cases are in II, 12, 25 and IV, 8, 17. The latter verse is by many regarded as spurious, and the former is partly justified by the fact that the place for the diaeresis corresponds to the division between the parts of a prepositional compound word (de-torquet).

The fact is, 2048 different arrangements of diaereses and caesuras are possible in any verse of twelve syllables. (The formula for finding this number in the case of any verse is 2"-1, n being the number of syllables in the given verse). Yet Horace among 509 verses chose to employ simply 142 out of the 2048, confining himself indeed, as a rule, to the use of only 12. For convenience, we shall represent this aspect of the subject as follows. "3-6-912" designates a verse whose diaereses and caesuras, as the case may be, fall after the third, sixth, ninth, and twelfth syllables. Horace's favorite arrangements are:

Examples of these twelve types are:

Maecenas atavis edite regibus.
quicquid de Libycis verritur areis.
late conspicuum tollere verticem.
obstrictis aliis praeter Iapyga.
ignem fraude mala gentibus intulit.
iracunda diem proferet Ilio.

sic fratres Helenae lucida sidera.

mittamus, scelerum si bene paenitet.
numquam demoveas, ut trabe Cypria.
qui semper vacuam semper amabilem.
nec tristis Hyadas nec rabiem Noti.
turparunt umeros immodicae mero.

(f) Why were these arrangements preferred? They seem to embody word-groups easily enunciated by the reader, symmetry in the divisions of the verse, the requisite variety of sound, and a certain subtle word-rhythm that plays, now within, now without the rhythm of the feet. To express it negatively, they avoid unwieldy and unbeautiful word-lengths and word-combinations; they avoid weak and unmusical verse-endings, unpleasant monotony, such as a verse composed exclusively of monosyllables or of dissyllables or one wherein the diaereses are relatively too numerous or one wherein ictus too often coincides with word-accent.

(g) The ends of words fall so that they do not generally coincide with ends of feet; otherwise expressed, caesura is more common than diaeresis. The degree in which this is true may be gathered from the following facts. The word sequence wherein a single word builds each foot is indicated by the scheme 2-5-69-11-12. This form, however, nowhere occurs. Somewhat similar to it though are the following six forms, each occurring but once: 2-5-6-12, 2-5-6-10-12, 2-5-6-8-10-12, 2-5-6-8-12, 2-5-6-8-9-12, 2-5-6-9-10-12. Even in these rarely occurring forms it is to be observed that when the poet allows coincidence in the first colon, he generally avoids it in the second-a phase of the subject that is more fully treated below.

(h) Monosyllabic words are not evenly distributed throughout the verse. A monosyllabic word stands in

Five of the verses have a monosyllable standing in the twelfth place, but three of these are strengthened by an elided est or et; so only two out-and-out cases remain and these are in IV, 13, where Horace's work seems less careful than in the general level of Book III. In only one of these five cases is the final monosyllable preceded by another monosyllable.

The table shows that monosyllables are more numerous in the first than in the second colon. They are massed in each case in the forward part. In a general way, each colon becomes more and more compact, as the reader moves from the beginning to the end of it. A colon accordingly takes on, to some extent, the characteristics of a verse. This is confirmed by other features that are common to the Lesser Asclepiadean verse and colon: the end of either regularly coincides with the end of a word; the rhythmic pause after either is often made the place for a sense pause. On the whole, the position and relative frequency of monosyllables harmonize with the foregoing theses.

There are certain points in the verse where monosyllables do not often occur, for the reason that the presence there of one monosyllable necessarily entails another in adjacent position. Such are the 2d, 5th, 8th, and 11th places. The poet generally avoided a heaping up of monosyllables, except for some special effect of sense or sound, the reason being that a series of monosyllables does not carry the rhythm well and besides produces monotony.

What is the prevailing quantity of the monosyllables? We find that the ratio between the total number of long and the total number of short monosyllables used by Horace does not correspond to the ratio between the total number of long and the total number of short monosyllables commonly occurring in Latin prose. In choice and arrangement of monosyllables he shows a strong preference for those fulfilling the conditions of long quantity. The underlying reason seems to be that monosyllables, for purposes of versification, are wanting in fluent and rhythmic properties, particularly so when they are short in quantity. In the latter case, each word lasts only one mora of time and is not reinforced by an ictus.

The heaping up of monosyllables at the outset of the cola is partially accounted for by the idiom of the Latin language, whereby many a common collocation of words begins with a monosyllable. The explanation, however, probably lies in part