The use of the Theory of Information as a mathematical tool for the understanding of complex systems allows the application of the MaxEnt Principle as a criterion for the prediction of the final stationary states. The processes of growth and differentiation are subject to a similar treatment by means of a new entropic magnitude. The iterative construction of fractals suggests a mathematical model of the hierarchic processes of growth and differentiation whose final stationary states are defined by the MaxEnt Principle. This can be interpreted in a more general context concerning the concept of Adaptation.