one feels in reading him — the new idea, the new logical distinction, pressing in before the old one is comfortably disposed of. He is by nature 'Semper instans sibi' (Quintilian). A certain freedom in grammar is common to all Greek, probably to all really thoughtful and vivid, writers: abstract singular nouns with plural verbs, slight anacolutha, intelligible corripressions of speech. But what is not explicable in Thucydides is that he should have fallen into the absolute hodge-podge of ungrammatical and unnatural language, the disconcerting trails of comment and explanation, which occur on every third page.
Not explicable if true; but is it true? The answers arise in a storm. "No; our text is utterly corrupt." "It is convicted of gross mistakes by contemporary inscriptions. It is full of glosses. It has been filled with cross-references and explanatory interpolations during its long use as a school-book." "Intentional forgers in late times have been at it" (Cobet, Rutherford). " One of them was * blood-thirsty,' and one talked 'like a cretin'!" (Miiller-Striibing). " Nay, the work itself being notoriously unfinished, it was edited after the author's death by another " (Wilamowitz); or by various others, who interpolated so freely, and found the MSS. in such a state of confusion, that the "unity of authorship is as hopelessly lost in the Thucydidean question as in the Homeric" (Schwartz).
Against this onslaught, it is not surprising that the average scholar has taken refuge in deafness, or looked on with sympathetic hope while Herbst does his magnificent gladiator-work in defence of everything that he believed in the happy sixties — the time, as he says plaintively, when he felt, in opening his Thucydides, that he was "resting in Abraham's bosom." It is not sur-
