Page:Essays and Addresses.djvu/384

 to the words, the, he is tacitly contrasting his own practice with that of Herodotus, the only conspicuous example in this department. If his statement were developed in this light, it might be paraphrased thus:—Thucydides says: (1) I have not introduced a speech except when I had reason to know that a speech was actually made: unlike Herodotus, when he reports the conversation between Croesus and Solon, the debate of the Persian conspirators, the discussion in the cabinet of Xerxes. (2) I do not pretend to give the exact form of the speeches made: as a writer implies that he does when, without warning the reader, he introduces a speech with the formula, "He said these things", instead of "He spoke to this effect". (3) On the other hand, I have faithfully reproduced the speaker's general line of argument, the purport and substance of his speech, so far as it could be ascertained. Herodotus disregards this principle when he makes Otanes,