603. We start from a region below the specialisation of visual faculty. The study of the successive dermal and nervous modifications which have led up to that faculty belongs to Biology, and all that our argument needs here is to point out that the very fact that this faculty has been developed in a germ, animated by metetherial life, indicates that some perceptivity from which sight could take its origin pre-existed in the originating, the unseen world. The germ was so constructed ab initio that it could develop in this and in other ways: whether we assume that each specific modifiability existed (and might have been discerned in the germ by an all-wise spectator) from the first, or that only a ground-plan existed, to which, in successive generations, fresh elements of determination and precision were added from the world of Life. We know vaguely how vision differentiated itself peripherally, with the growing sensibility of the pigment-spot to light and shadow. But there must have been a cerebral differentiation also, and also a psychological differentiation, namely, a gradual shaping of a distinct feeling from obscure feelings, whose history we cannot recover.

Yet I believe that we have still persistent in our brain-structure some dim vestige of the transition from that early undifferentiated continuous sensitivity to our existing specialisation of sense. Probably in all of us, though in some men much more distinctly than in others, there exist certain synoesthesia or concomitances of sense-impression, which are at any rate not dependent on any recognisable link of association. A second sense sometimes reacts automatically to a stimulus which seems fitted to excite a single sense alone. I do not merely mean that the dog's bark calls up the look of the dog, while his look suggests his bark; that is an association formed by the mere experience of life. But for a true synæs-thetic or "sound-seer," - to take the commonest form of these central repercussions of sensory shock, - there is a connection between sight and sound which is instinctive, complex, and yet for our intelligence altogether arbitrary. In some cases, indeed, these chromatisms can be watched in development, if not in origination, and may be referred to some odd chance of fanciful association.

But this first group of cases of sound-seeing melts into a second, where the chromatisms seem to be determined before birth, and to have preceded conscious mentation, in all their meaningless precision of correspondence, between, say, a particular note played on the piano and a particular tint of apple-green. The specimens given in 603 A and B will show something of this irrational complexity. My present point is that these synsesthesiæ stand on the dividing line between percepts externally and internally originated. These irradiations of sensitivity, sometimes, as I have said, apparently congenital, cannot, on the one hand, be called a purely mental phenomenon. Nor again can they be definitely classed under external vision; since they do sometimes follow upon a mental process of association. It seems safer to term them entencephalic, on the analogy of entoptic, since they seem to be due to something in brain-structure, much as entoptic percepts are due to something in the structure of the eye.

604. I will, then, start with the synaesthesiæ as the most generalised form of inward perception, and will pass on to other classes which approach more nearly to ordinary external vision.

From these entencephalic photisms we seem to proceed by an easy transition to the most inward form of unmistakable entoptic vision - which is therefore the most inward form of all external vision - the flash of light consequent on electrisation of the optic nerve. Next on our outward road we may place the phosphenes caused by pressure on the optic nerve or irritation of the retina. Next Purkinje's figures, or shadows cast by the blood-vessels of the middle layer upon the bacillary layer of the retina. Then muscoe volitantes, or shadows cast by motes in the vitreous humour upon the fibrous layer of the retina.

605. Midway, again, between entoptic and ordinary external vision we may place after-images,- which, although themselves perceptible with shut eyes, presuppose a previous retinal stimulation from without; - forming, in fact, the entoptic sequelae of ordinary external vision.

606. Next comes our ordinary vision of the external world - and this, again, is pushed to its highest degree of externality by the employment of artificial aids to sight. He who gazes through a telescope at the stars has mechanically improved his end-organs to the furthest point now possible to man.

607. And now, standing once more upon our watershed of entencephalic vision, let us trace the advancing capacities of internal vision. The forms of vision now to be considered are virtually independent of the eye; they can persist, that is to say, after the destruction of the eye, if only the eye has worked for a few years, so as to give visual education to the brain. We do not, in fact, fully know the limits of this independence, which can only be learnt by a fuller examination of intelligent blind persons than has yet been made. Nor can we say with certainty how far in a seeing person the eye is in its turn influenced by the brain. I shall avoid postulating any "retropulsive current" from brain to retina, just as I have avoided any expression more specific than "the brain" to indicate the primary seat of sight. The arrangement here presented, as already explained, is a psychological one, and can be set forth without trespassing on controverted physiological ground.

We may take memory-images as the simplest type of internal vision. These images, as commonly understood, introduce us to no fresh knowledge; they preserve the knowledge gained by conscious gaze upon the outer world. In their simplest spontaneous form they are the cerebral sequelae of external vision, just as after-images are its entoptic sequelae. And we find that in some cases these two classes of vision are confounded (see 607 A). But we see that into the cerebral storage of impressions one element habitually enters which is totally absent from the mere retinal storage, namely, a psychical element - a rearrangement or generalisation of the impressions retinally received.

608. Next we come to a common class of memory-images, in which the subliminal rearrangement is particularly marked. I speak of dreams - which lead us on in two directions from memory-images; - in the direction of imagination-images, and in the direction of hallucinations. Certain individual dreams, indeed, of rare types point also in other directions which later on we shall have to follow. But dreams as a class consist of confused memory-images, reaching a kind of low hallucinatory intensity, a glow, so to say, sufficient to be perceptible in darkness.

609. I will give the name of imagination-images to those conscious recombinations of our store of visual imagery which we compose either for our mere enjoyment, as " waking dreams," or as artifices to help us to the better understanding of facts of nature confusedly discerned. Such, for instance, are imagined geometrical diagrams; and Watt, lying in bed in a dark room and conceiving the steam-engine, illustrates the utmost limit to which voluntary internal visualisation can go.