236 ALFEED W. BENN : Theophrastus quoted in Diels' Doxographi, pp. 499 sqq. If so I must remind him that the opinions given there relate not to know- ledge but to sensation (aio-^o-ts). It is of no importance what views Heracleitus held about the senses and their mode of action. We are only concerned about his theory of knowledge, and as to that we have the evidence of Aristotle, who tells us in the De Animd that according to him the soul is, like the universe, in a state of flux, "for the moving is known by the moving" (TO Se Kivov/jifvov /avou/xeVu> ymoo-Kccrtfai). 1 And this is confirmed by the sub- stantially equivalent statement of other authorities that Heracleitus represented the soul as fiery. . Moreover we have it on his own direct authority that " the dry soul is the wisest and best " wisest because most like the elemental fire. And drink makes men foolish by moistening their souls, bringing them, that is to say, into a state opposite to the reality of things. Now on Mr. Taylor's theory of Heracleiteanism true wisdom would consist in following the prescription of the hermit-sage, and having not only " some beer " but a good deal of it. With Anaxagoras the case is rather different, and I must con- fess that I overlooked him. He affirms that the cosmic Nous is unlike everything and knows everything. But Aristotle tells us that in taking this view he stood alone, and that he neither gives nor suggests any explanation as to how this knowledge is obtained. 2 And even this exception is only partial. For Anaxagoras would not have denied that we know the cosmic Nous as well as the scattered portions of it in other men, in animals, and in plants by the like nous in ourselves. So far as knowledge in the Greek sense goes there is no question of a school, nor of an elaborately worked out doctrine of generation by opposites. Diogenes of Apollonia, who seems to have set up a fashionable Anaxagorean school at Athens, abandoned this part of the master's theory, and by identifying the Nous with air restored the principle of cognition by likeness. It appears then that by writing " almost unanimous " for " unanimous " my statement would be made strictly accurate. Mr. Taylor, I suspect, would have to give the printer much more trouble if he tried to bring his criticism into accordance with fact and logic. 3 I remarked that Plato would not have agreed with Descartes in holding that the idea of perfection involves that of existence ; and Mr. Taylor " entirely fails to see how this is to be reconciled with" a passage he quotes from the Sophistes (p. 1). I have studied the passage long and earnestly but " entirely fail to see " what it has 1 405 a, 27. 2 Loc ciL, &., 20. 3 While I am about it, I wish to take this opportunity of correcting another regrettable inaccuracy. In the article referred to I quoted Plato as saying that he " had never met a mathematician who could reason " (p. 89). I should have written with Jowett " hardly ever" (/zdXa ye rivts oXt'yot. Rep., 531 E).
