And first we must examine mere _extension_ as such, which distinguishes our active dealings with visual and audible sensations from our passive reception of the sensations of taste and smell. For while in the case of the latter a succession of similar stimulations affects us as "more taste of strawberry" or "more smell of rose" when intermittent, or as a vague "there _is_ a strong or faint taste of strawberry" and a "there is a smell of lemon flower"--when continuous; our organ of sight being mobile, reports not "more black on white" but "so many inches of black line on a white ground," that is to say reports a certain _extension_ answering to its own movement. This quality of extension exists also in our sound-perceptions, although the explanation is less evident. Notes do not indeed exist (but only sounding bodies and air-vibrations) in the space which we call "real" because our eye and our locomotion coincide in their accounts of it; but notes are experienced, that is thought and felt, as existing in a sort of imitation space of their own. This "musical space," as M. Dauriac has rightly called it, has limits corresponding with those of our power of hearing or reproducing notes, and a central region corresponding with our habitual experience of the human voice; and in this "musical space" notes are experienced as moving up and down and with a centrifugal and centripetal direction, and also as existing at definite spans or _intervals_ from one another; all of which probably on account of presumable muscular adjustments of the inner and auditive apparatus, as well as obvious sensations in the vocal parts when we ourselves produce, and often when we merely think of, them. In visual perception the sweep of the glance, that is the adjustment of the muscles of the inner eye, the outer eye and of the head, is susceptible of being either interrupted or continuous like any other muscular process; and its continuity is what unites the mere successive sensations of colour and light into a unity of extension, so that the same successive colour-and-light-sensations can be experienced either as _one_ extension, or as two or more, according as the glance is continuous or interrupted; the eye's sweep, when not excessive, tending to continuity _unless a new direction requires a new muscular adjustment._ And, except in the case of an _extension_ exceeding any single movement of eye and head, a new adjustment answers to what we call _a change of direction. Extension_ therefore, as we have forestalled with regard to sound, has various modes, corresponding to something belonging to ourselves: a _middle,_ answering to the middle not of our field of vision, since that itself can be raised or lowered by a movement of the head, but to the middle of our body; and an _above_ and _below,_ a _right_ and a _left_ referable to our body also, or rather to the adjustments made by eye and head in the attempt to see our