Sartre's foisting of the 'illusion of immanence' on to empiricist accounts of imaging is premature on two counts. Not only does he not show that empiricists are restricted to immanent ideas as determinants of intentional objects (Hume, significantly, talks of images as incidental to thought, in that they are 'excited' by 'discourse', and that they occur 'in thinking and reasoning',! rather than thinking and reasoning in them), he fails to convince us that, as far as the image-experience goes, he is not still restricted to such ideas himself. Indeed, if only an illusion of transcendence can cause these items to drop out, his failure to convince will be complete. For on his own account this illusion should be impossible on principle, as betokening a blindness of the mind to its own transactions with the world. The point of course is that the imaginary appearance is not illusorily immanent. It is there, and what makes imaging possible. And the criticism applies even if we accept that these items are not objects of the order of the objects that appear directly to consciousness in perception and imaginarily in imaging. For the only conclusion we can draw from the relevant considerations here, as we saw, is that images are not the same kind of thing as the objects we directly perceive. And although this seems to be a point on which Sartre and, for example, Hume genuinely differ, it does not explain why Sartre should see images and physical objects as being absolutely distinct types of thing; that is, it does not explain why images should not be conceived as a species of object at all, but as acts.