On the subject of the first reason, he says, that the figure which we perceive as associated with colour, is the real, and not the apparent figure. 'Is there,' he asks, 'the slightest consciousness of a perception of visible figure, corresponding to the affected portion of the retina?' To which, though he seems to think an affirmative answer impossible, we cannot hesitate to reply, that there is undoubtedly such a consciousness; that though obscured by being made the ground of habitual inference as to the real figure, this consciousness is constantly referred to by the draughtsman, and easily recalled by any one. We may separate colour, he says {318} again[16\4], from the figures on the retina, as we may separate it from length, breadth, and thickness, which we do not see. But this is altogether false: we cannot separate colour from length, breadth, and thickness, _in any other way_, than by transferring it to the visible figure which we do see. He cannot, he allows, separate the colour from the visible form of the trunk of a large oak; but just as little, he thinks, can he separate it from the convex mass of the trunk, which (it is allowed on all hands) he does not immediately see. But in this he is mistaken: for if he were to make a _picture_ of the oak, he would separate the colour from the convex shape, which he does not imitate, but he could not separate it from the visible figure, which he does imitate; and he would then perceive that the fact that he _has not_ an immediate perception of the convex form, is necessarily connected with the fact that he _has_ an immediate perception of the apparent figure; so far is the rejection of immediate perception in the former case from being a reason for rejecting it in the latter. [Note 16\4: _Lectures_, vol. ii. p. 84.]