PROSE STYLE IN THE 'ORATORS' 331 rhythm and hiatus, just as it explains many details in the system of punctuation — for instance, the dash below the Hne which warns you beforehand of the approach of the end of the sentence. We are but little sensible to rhythm and less to hiatus or the clashing of two vowel-sounds without a dividing consonant ; we are keenly alive to rhyme. The Greeks generally did not notice rhyme, but felt rhythm strongly, and abhorred hiatus. In poetry hiatus was absolutely forbidden. In careful prose it was avoided in varying degrees by most writers after about 380 B.C. Isocrates is credited with introducing the fashion. He was followed by all the historians and philosophers and writers of belles lettres, and even, in their old age, by Plato and Xenophon.^ The orators who 'published' generally felt bound to preserve the prevailing habit. In the real debates of the Assembly, of course, such refinement would scarcely be either attainable or notice- able, but a published speech had to have its literary polish. A written speech, however, was an exceptional thing. The ordinary orators — Callistratus, Thrasybulus, Leodamas — were content simply to speak. Even Demosthenes must have spoken ten times as much as he wrote. The speeches we possess are roughly of three kinds. First, there are the bought speeches preserved by the client for whom they were written : such are ox? seven
- There is indeed some dou^* ;!juut this avoidance of hiatus. Our earliest
papyri give texts wbAc*" :-w.nit hiatus freely. The funeral speech of Hyperides, for instance, abounds in harsh instances, and the pre-Alexandrian papyri of Plato have more hiatus than our ordinary MSS. Does- this mean that the Alexandrian scholars deliberately doctored their classical texts and removed hiatus ? Or does it mean that our pre-Alexandrian remains are generally in- accurate ? The former view must be dismissed as flatly impossible, though there are some difficulties in the latter.
