192 LITERATURE OF ANCIENT GREECE prising that conservative editors have even adopted the extraordinary theory — merely in defence against the development theories of Ullrich, Kirchoff, and Cwiklinski — that Thucydides did not write a word betweeen 432 and 404, and then apparently did the whole book at a sitting. This is not the place to discuss the text, except in the broadest manner, and for the sake of its signifi- cance in the history of literature and in our conception of Thucydides. In the first place, the general line of Cobet followed by Rutherford, that the text is largely defaced by adscripts and glosses, and that Thucydides, a trained stylist at a time when style was much studied, did not, in a work which took twenty-nine years' writing, mix long passages of masterly expression with short ones of w^hat looks like gibberish — thus much seems morally certain. The mere comparison of the existing MSS. and the study of Thucydides's manner show it. But that takes us very little way. Dr. Rutherford's valuable edition of Book IV., attempting to carry these results to a logical conclusion, has produced a text which hardly a dozen scholars in Europe would accept. We can see that the original wording has been tampered with ; we can see to a certain extent the lines of the tampering. We cannot from that restore the original. But we have some concrete facts by which to estimate our tradition. We have part of the original text of one of Thucydides's documents extant on an Attic stone. ^ We have some significant quotations in the late geographer Stephen of Byzantium. The inscription, according to Kirchofif, taking the twenty-five lines alone, but allowing for restorations, 1 The treaty, Thuc. v. 47 = C. I. A. iv. 46 b.
