Among Cicero's numerous works on rhetoric the chief is the 'De Oratore.' Actually composed in 55 B.C., it is a dialogue, the scene set in 91 B.C., the characters being the chief Roman orators of that day. L. Crassus, who plays the host gracefully at his Tusculan country-seat, is also the chief speaker. These men were all known to Cicero in his boyhood, but most of them perished soon after in the Marian proscriptions. Of real character-drawing there is little, and all alike speak in graceful Ciceronian periods. The exposition of the technical parts of rhetoric goes on in leisurely wise, with copious illustrations and digressions. There is much pleasant repetition of commonplaces. Wilkins's edition of the 'De Oratore' is a good but not an ideal one. The introductions are most helpful. Countless discussions on etymology, etc., in the notes, should be relegated to the dictionaries. Instead, we crave adequate cross-references to passages in this and other works. The notes seem to be written too largely piecemeal, each with the single passage in mind. In Cicero's 'Brutus,' written in 46 B.C., Cicero, Brutus, and Atticus carry on the conversation, but it is mostly a monologue of Cicero and a historical sketch of Roman oratory. The affected modesty of the autobiographic parts is diverting. Brutus was the chief exponent of a terse, simple, direct, oratory,--far nearer, we judge, to English taste than the Ciceronian; and the opposition between them already appears. A convenient American edition is that by Kellogg (Ginn). The opposition just mentioned comes out more clearly in the 'Orator.' This portrays the ideal public speaker. His chief accomplishments are summed up in versatility,--the power to adapt himself to any case and audience. An interesting passage discusses the rhythms of prose. This book has been elaborately edited by J.E. Sandys. In these three dialogues Cicero says everything of importance, at least once; and the other rhetorical works in the Corpus may be neglected here, the more as the most practical working rhetoric among them all, the 'Auctor ad Herennium,' is certainly not Cicero's. It is probably by Cornificius, and is especially important as the _first_ complete prose work transmitted to us in authentic Latin form. (Cato's 'De Re Rustica' has been "modernized.") The later history of the Ciceronian correspondence is a dark and much contested field. (The most recent discussion, with bibliography, is by Schanz, in Iwan Mueller's Handbuch, Vol. viii., pp. 238-243.) Probably Cicero's devoted freedman Tiro laid the foundations of our collections. The part of Petrarch in recovering the letters during the "Revival of Learning" was much less than has been supposed.