-a most unhappy adjective). Logical necessity is the necessity of consistency, of coherence in thought. Consistency is, as I have said, our negative test of reality. Complete consistency (to us a mere ideal) would be a positive test. To say that through natural selection our minds have become such that they require to believe in the uniformity of nature, is to give what seems the most probable historical account of the development of human thought; but it is no logical solution of the question about the ultimate relation of thought to reality, it only restates the fact that Nature, in the only sense which it can have for the scientific man, is a rational system, whose "laws" are the logical necessities of our thought. If we ask why this should be so, there is first of all, an answer with which some persons profess to be satisfied: The necessity, it is held, is merely subjective; what "nature" really is, we do not know. Observe, it is generally assumed, that we know somehow (but how?) that there is a reality "behind" the phenomena-a survival from old metaphysics in the minds of those who boast that they are not metaphysicians. It seems to me an odd use of the word "real" to make it mean what stands in no relation to our lives. Secondly, dissatisfied with this "positivism" or subjective idealism, with or without a hypothetical "realism" in the background, we may be content to fall back upon the dualism of popular language, and assume a "parallelism" between our thought and the nature of things. But the demands of philosophical criticism will hardly allow us to rest satisfied with this metaphor of parallelism, which easily leads us back into the other metaphors of mirrors, waxen tablets, etc. Knowledge is not possible, if thought and things confront each other as alien substances. Is it not the better hypothesis (not to call it anything more) that our thought is in some way identical with the reason in things? Logical necessity is the necessity, not merely of this or that individual mind, nor of a sum of individual minds, but of thought as such, and therefore of the universe.

(3) In regard to human actions, Mill shows quite well (Logic, Book VI. ch. ii.) that the necessity presupposed by any science such as psychology or sociology, is, as with physical phenomena, simply necessity of inference. On the other hand, moral necessity (if the phrase be used for a moral precept, "You must, i.e., ought to, do so and so ") is a totally different thing. A political or moral law is the statement of an ideal; it puts forward one motive for conduct among other motives. It is through association with this sense of "must" (implying a command with threats of penalties) that our understanding of physical necessity is apt to be vitiated (see below, pp. 174-177).

IV.

ON PLATO'S PHAEDO.1

I.

BEFORE We can answer the questions: "What are Plato's arguments about the soul's nature and destiny?" "What is their relation to one another?" "What is their value?" we are obliged to consider how far the expressions used by him are to be understood literally.

Plato's visions of another world have fixed themselves indelibly in the common consciousness of Western civilisation. We hardly know, without the most careful examination, how many of those beliefs that are often spoken of as if they were peculiar to Christianity, are due directly or indirectly to Platonic influence. Thus, even if it should be the case that the mythical element in Plato is (as Hegel' holds) quite unessential in his philosophy, or (as Teichmüller holds) not believed in at all by Plato him

3

1 Read before the Aristotelian Society (London) on Nov. 30, 1885, and published in Mind, Vol. XI., No. 43.

2 Geschichte der Phil., ii. 207 ff.

3 Studien zur Gesch. der Begriffe and Ueber die Unsterblichkeit der Seele. Some criticisms of the late Prof. Teichmüller's on this paper of mine will be found in his Religionsphilosophie (Breslau,

self, this mythical element would still deserve the attention of all students of human thought, both as taking up previous Pythagorean, Orphic, probably Egyptian and perhaps Indian ideas, and as influencing all the Hellenic and Roman world, i.e., what we commonly call the whole world. And, in any case, the mythical form of expression must throw some light on Plato's habitual manner of thinking; for we cannot abstractly separate form and content, expression and thought.

Let us take the three characteristic Platonic "doctrines" of Recollection, Pre-existence and Transmigration, and endeavour to discover in what sense they are to be understood.

1. The doctrine of Recollection (aváμvnois) occurs both in the Meno and the Phado. "Knowing is remembering." This theory seemed to obviate the Sophistic puzzle about the impossibility of learning : -We either learn what we already know or what we don't know in the first case we don't learn; in the second case, we can't (cf. Meno, 80 E). This is just one of those instances where the Aristotelian distinction of potentiality and actuality comes at

1886), p. 502. Prof. Vera in his monograph entitled Platone e l'immortalità dell' anima (Naples, 1881), to which Dr. J. Hutchison Stirling kindly directed my attention, maintains a theory very much like Teichmüller's, except that he would treat Plato's "myths" as approximations, "serious jests "the world itself being such. "Il giuoco è l'involucro, la parvenza dell' idea, cioè la poesia." But, as he says philosophy is essentially "esoteric," his general attitude to Plato comes to be the same as that of Teichmüller.

once to our help, We learn what we are capable of knowing we cannot learn what is quite alien to us. But the knowledge, which in some form is there already, is there only virtually, and requires the effort of what we call learning to become actual, to be realised, to become what we can properly call knowledge. Plato in the Theatetus (which in many respects may be called the most "modern" of all his dialogues, for in it he discusses, not the usual ancient question of Being, but the modern question of Knowing) does arrive at this Aristotelian distinction in his recognition of the difference between "possessing" and "having or holding," illustrating it by the birds in a cage (Theat., 197); but it remained for Aristotle to grasp the full significance of this distinction, which has become so much a commonplace of our language and our thought that it requires an effort to see its importance and to understand how the problems of knowledge presented themselves before the time of Aristotle. Now, this is just the philosophic truth of Plato's theory of Recollection: in learning the mind is not filled with something alien to it,-as popular language, now as then, is inclined to assume, and as even some philosophers have been apt to suppose, e.g., when they ask how Mind can know Matter, after defining Matter in such a way that it is of its very essence, as the exact antithesis of Mind, that it cannot be known. According to Plato, in learning the soul recovers its own. This is more than a theory of knowledge merely. In the Phædrus it becomes a theory of art and morality as well. The