This paper is a second half of a longer argument.¹ In its predecessor,² I argued that Kant saw the gist of the mathematical method in what essentially amounts to instantiation rules, i.e., what he himself characterized as arguing in terms of particular representatives of general concepts.³ But when can such an anticipatory introduction of representatives of general concepts yield synthetic knowledge a priori? Kant's transcendental viewpoint commits him to answering: Only in so far as we have ourselves put the relations and properties we are arguing about into objects. Then our mathematical knowledge does not pertain to things, but only to the structure of our processes of coming to know them.