At the beginning of Posterior Analytics 2.19 Aristotle reminds us that we cannot claim demonstrative knowledge (epistêmê apodeiktikê) unless we know immediate premisses, the archai of demonstrations. By the end of the chapter he explains why the cognitive state whereby we get to know archai must be Nous. In between, however, Aristotle describes the process of the acquisition of concepts, not immediate premisses. How should we understand this? There is a general agreement that it is Nous by means of which we acquire both premisses and concepts. I argue that this cannot be the case. Since concepts are simples while premisses are composites (predications of concepts), the two cannot be objects of the same cognitive state. I further argue that, whereas Nous is responsible for our grasp of concepts, the state Aristotle elsewhere calls non-demonstrable knowledge is the one whereby we get to know the premisses of demonstrations.