All together, these findings must be placed in the context of the recent realization, based on complete genome sequences, that evolution toward a higher level of complexity does not rely on the dramatic acquisition of new regulatory functions, or even on a large increase in gene number. It has become clear that Nature uses over and over again a limited number of biochemical modules, such as hh/Shh or dpp/BMP, to achieve various simple patterning tasks. It appears now that patterning of complex tissues also reuses networks of these modules simply by changing the rules of their interaction. For instance, changes in the promoters of hh and dpp as well as changes in the targets of these signaling molecules through the influence of the genes that determine the nature of the structure (master regulators) might allow the development of new structures during evolution.It is very likely that the primitive ancestral unit from which eyes evolved was simply a photoreceptive cell that used Pax6 to activate expression of an opsin. As different eyes became more complex, Pax6 became restricted to progenitor cells within the retina. Then, a convergent strategy that used the interactions between several preexisting biochemical modules such as hh, dpp, and N was superimposed onto the Pax6 module to pattern the crystalline arrays of photoreceptors (Figure 3(Figure 3). Differences between phyla in the details of the regulation of such conserved genes and networks support the idea that they have been recruited independently through evolution to elaborate new eye structures. It is striking that very complex and distinct developmental processes are achieved with a relatively restricted and simple number of interlinked genetic modules such as the hh, dpp, N, and EGFR pathways. This leads to the notion that similar strategies do not necessarily imply common ancestry of the organ, but instead reflect the reuse of an efficient mechanism invented for a similar task.