That doctrine, of course, holds that the extended can be actual, entirely apart from every other quality. But extension is never so given. If it is visual, it must be coloured; and if it is tactual, or acquired in the various other ways which may fall under the head of the “muscular sense,”—then it is never free from sensations, coming from the skin, or the joints, or the muscles, or, as some would like to add, from a central source. And a man may say what he likes, but he cannot think of extension without thinking at the same time of a “what” that is extended. And not only is this so, but particular differences, such as “up and down,” “right and left,” are necessary to the terms of the spatial relation. But these differences clearly are not merely spatial. Like the general “what,” they will consist in all cases of secondary quality from a sensation of the kinds I have mentioned above. Some psychologists, indeed, could go further, and could urge that the secondary qualities are original, and the primary derivative; since extension (in their view) is a construction or growth from the wholly non-extended. I could not endorse that, but I can appeal to what is indisputable. Extension cannot be presented, or thought of, except as one with quality that is secondary. It is by itself a mere abstraction, for some purposes necessary, but ridiculous when taken as an existing thing. Yet the materialist, from defect of nature or of education, or probably both, worships without justification this thin product of his untutored fancy.
“Not without justification,” he may reply, “since in the procedure of science the secondary qualities are explained as results from the primary. Obviously, therefore, these latter are independent and prior.” But this is a very simple error. For suppose that you have shown that, given one element, , another, , does in fact follow on it; suppose that you can prove that comes just the same, whether is
