Page:Essays and Addresses.djvu/443

 the antiquarian turn of Claudius, Tacitus might easily have concluded that the Emperor's speech would dwell largely on historical precedents; but it seems more likely that he knew, from oral or written report, the substance of what Claudius had said, and worked up this in his own way. Here, then, is a rough gauge of the approximation which might be made to the truth by a historian who composed a speech based on "the general sense of what was really said." Thucydides and Polybius, Sallust and Tacitus, are widely removed from writers who introduce harangues merely as opportunities of display. The latter tendency prevailed in what Gibbon calls "the elaborate and often empty speeches of the Byzantine historians ." The Latin chroniclers of the middle ages rarely ventured on such ambitious efforts. But at the revival of letters the classical practice of inserting speeches was revived by historical writers, whether they wrote in Latin