In this chapter I turn to discussing the properties of icons. I examine iconic or analog representations highlighting their main properties. I concentrate on their significant difference from symbolic representations concerning both the way icons refer, and the way iconic representations compose to form more complex iconic representation. I also attempt to elucidate the notion of iconic representations because the discussions in the literature tend to be, to say the least, chaotic and perplexing. Researchers usually fail to distinguish between external and mental representations, between representational vehicles and contents, and between individual representations and representational systems, Lyons (Asian J Philos 1(50), 2022) being a notable exception. Before proceeding, I discuss the reasons why icons, qua representational signs, succeeded the non-representational indices, that is, what made possible the emergence of representations in the first place. Then, having examined the properties of icons, I turn to addressing the question of when hominins acquired the cognitive arsenal that allowed them to acquire the capacity to understand and employ iconic representations, and also the question why was the development of these capacities necessary, evolutionary speaking. This will allow us to pinpoint the cognitive capacities that agents must possess in order to be able to form iconic representations of their environment.