deny the independent and separate existence of the one as well as of the other: they admit the two only as relative to each other, or as reciprocally implicated in the indivisible fact of cognition. The reasoning of Aristokles thus goes to prove the opinion which he is trying to refute. Most of the arguments, which Sextus adduces in favour of the Kyrenaic doctrine, show forcibly that the Objective Something, apart from its Subjective correlate, is unknowable and a non-entity; but he does not include in the Subjective as much as ought to be included; he takes note only of the present sensation, and does not include sensations remembered or anticipated. Another very forcible part of Sextus's reasoning may be found, vii. sect. 269-272, where he shows that a logical Subject _per se_ is undefinable and inconceivable--that those who attempt to define Man (_e.g._) do so by specifying more or fewer of the predicates of Man--and that if you suppose all the predicates to vanish, the Subject vanishes along with them.] CHAPTER IV. XENOPHON. [Side-note: Xenophon--his character--essentially a man of action and not a theorist--the Sokratic element in him an accessory.] There remains one other companion of Sokrates, for whom a dignified place must be reserved in this volume--Xenophon the son of Gryllus. It is to him that we owe, in great part, such knowledge as we possess of the real Sokrates. For the Sokratic conversations related by Xenophon, though doubtless dressed up and expanded by him, appear to me reports in the main of what Sokrates actually said. Xenophon was sparing in the introduction of his master as titular spokesman for opinions, theories, or controversial difficulties, generated in his own mind: a practice in which Plato indulged without any reserve, as we have seen by the numerous dialogues already passed in review.