Institutio Oratoria

E has a flatter, i a narrower sound, and consequently such blemishes are less noticeable where they are concerned. It is a less serious fault to place short vowels after long, a statement which applies even more strongly to placing short vowels before long. But the least unsatisfactory combination is that of two short vowels. And in all conjunctions of vowels, the resulting sound will be proportionately soft or harsh according as they resemble or differ from each other in the method of utterance.

On the other hand, hiatus is not to be regarded as so very terrible a crime: in fact I do not know which is the worse fault in this connexion, carelessness or a pedantic

On the other hand Demosthenes and Cicero show a sense of proportion in the way in which they face the problem. For the coalescence of two letters, known as συναλοιφή, may make our language run more smoothly than if every word closed with its own vowel, while sometimes hiatus may even prove becoming and create an impression of grandeur, as in the following case, pulchra oratione ista iacta te. [*](Boast yourself of that fine speech of yours.) For syllables which are naturally long and rich in sound gain something from the time which intervenes between two vowels, as though there were a perceptible pause.

I cannot do better than quote the words of Cicero [*](Or. xxiii. 77. ) on this subject. Hiatus, he says, and the meeting of vowels produce a certain softness of effect, such as to suggest a not unpleasing carelessness on the part of the orator, as though he were more anxious about his matter than his words. But consonants also are liable to conflict at the juncture of words, more especially those letters which are comparatively harsh in sound; as for instance when the final s of one word clashes with x at the opening of the next. Still more unpleasing is the hissing sound produced by the collision between a pair of these consonants, as in the phrase ars studiorum.

This was the reason why Servius, as he

Hence we get forms such as belligerare and pomeridiem, to which the diee hanc [*](i.e. for belligerares, postmeridiem and diem hanc. ) of Cato the Censor, where the final m is softened into an e, presents an analogy. Unlearned readers are apt to alter such forms when they come across them in old books, and in their desire to decry the ignorance of the scribes convict themselves of the same fault.

On the other hand, whenever this same letter m comes at the end of a word and is brought into contact with the opening vowel of the next word in such a manner as to render coalescence possible, it is, although written, so faintly pronounced ( e.g. in phrases such as nultum ille and quantum erat ) that it may almost be regarded as producing the sound of a new letter. [*]( A very probable account is that -m was reduced through the lips not being closed to pronounce it. If, instead of closing the lips all that were done were to drop the uvula, a nasal sound would be given to the following initial vowel, so that fine onerat would be pronounced finewonerat with a nasalized o. Lindsay, Lat. Langu. p. 62. It is this sound which Quintilian describes as almost the sound of a new letter. ) For it is not elided, but merely obscured, and may be considered as a symbol occurring between two vowels simply to prevent their coalescence.

Care must also be taken that the last syllables of one word are not identical with the opening syllables of the next. In case any of my readers should wonder that I think it worth while to lay down such a rule, I may point out that Cicero makes such a slip in his Letters, in

1. Ofortunatam natam me consule Romam.

See XI. i. 24. [*](O happy Rome, born in my consulship.)

Again it is a blemish to have too many monosyllables in succession, since the inevitable result is that, owing to the frequency of the pauses, the rhythm degenerates into a series of jerks. For the same reason we must avoid placing a number of short verbs and nouns in succession; the converse also is true as regards long syllables, since their accumulation makes our rhythm drag. It is a fault of the same class to end a number of successive sentences with similar cadences, terminations and inflexions.

It is likewise inartistic to accumulate long series of verbs, nouns or other parts of speech, since even merits produce tedium unless they have the saving grace of variety.

The principles by which the connexion of words is guided are not sufficient in the case of commata and cola, though even here beginnings and ends should harmonise; but our structural effect will very largely depend on the relative order of these two types of clause. For in the following instance [*](Phil. Il. xxv. 63. By his vomiting he filled his lap and the whole judgement seat with fragments of undigested food. ) vomens frustis esculentis gremium suum el totum tribunal implevit [the order is satisfactory, since the fact of his having filled the whole judgement seat with his vomiting is the more important of the two]. On the other hand (for I shall repeat the same illustrations for different purposes to make them more familiar) in the following passage, [*](pro Arch. viii. 19. Rocks and solitude answer to the human voice and wild beasts are often pacified and brought to a halt by the influence of music. ) saxa alque solitudines voci respondent, bestiae

All combination, arrangement and connexion of words involves either rhythms (which we call numeri ), or metres, that is, a certain measure. Now though both rhythm and metre consist of feet, they differ in more than one respect.

For in the first place rhythm consists of certain lengths of time, while metre is determined by the order in which these lengths are arranged. Consequently the one seems to be concerned with quantity and the other with quality. Rhythm may depend on equal balance,

as in the case of dactylic rhythm, where one long syllable balances two short, (there are it is true other feet of which this statement is equally true, but the title of dactylic has been currently applied to all, [*]( For purely rhythmical purposes the term dactyl is arbitrarily used by the rhetoricians to include anapaests as well. See below. ) while even boys are well aware that a long syllable is equivalent to two beats and a short to one) or it may consist of feet in which one portion is half as long again as the other, as is the case with paeanic rhythm (a paean being composed of one long followed by three shorts, three shorts followed by one long or with any other arrangement preserving the proportion of three beats to two) or finally one part of the foot may be twice the length of the other, as in the case of the iambus, which is composed of a short followed by a long, or of the choreus consisting of a long followed by a short.

These feet are also employed by metre, but with this difference, that in rhythm it does not matter whether the two shorts of the dactyl precede or

Further, the laws of metre not merely refuse the substitution of one foot for another, but will not even admit the arbitrary substitution of any dactyl or spondee for any other dactyl or spondee. For example, in the line

1. Panditur snterea domus omnipotentis Olympi

Aen. x. 1. [*](Meanwhile Olympus' halls omnipotent / Are wide unbarred.)