One of the major challenges of any historical science is the role of forecasting. Biology is a historical science, because organisms can only be understood in a temporal, phylogenetic and ontogenetic, perspective. In particular, the time of biology, and therefore of correlated sciences (ecology etc.), is a time of change in the "space of possibilities" (of "phases" as we say in physics, of ecosystems and species in biology), punctuated by rare events-evolutionary novelties, speciation. In physics, the space (of phases) is fixed as "a priori" of knowledge, as "condition of possibility" for "writing equations", explain Newton and Kant: it a priori contains all possible trajectories-unpredictability is within these trajectories (the randomness of a die concerns six possibilities, no more, no less). In biology, to physical randomness is added the unpredictability of changes of the space of possibilities and of rare events, to which one cannot even assign probability values, see (1). The historian of human affairs will recognize there elements of his theorization, although, of course, symbolic culture, in all its forms, imposes an important change in the tools of analysis. Faced with the intrinsic unpredictability of the history of the living, should we remain silent? No, science is not (or not only) the analytical game of "experience/observation, theory, prediction, verification/ falsification", but first of all a construction of objectivity, or even objects of knowledge, through difficult operations of "carving-out" (separating/distinguishing) and "qualification" of reality. This is how Darwinian theory offers us a remarkable historical theory of the living, of "species", a notion that is constantly being re-carved-out and re-qualified. This theory does not allow us to predict, but, by making us understand, it allows us to act, if we assume the risk of relying on the best available knowledge. We then decide to measure "biodiversity", an admittedly arbitrary partition of species and life forms, which is always open to discussion and revision; to assess the impact of man on an ecosystem, a difficult qualification of the consequences of activities that are sometimes centuries old. We can also give ourselves a measure of the notion of 'epidemic' and draw the historical diagram shown here (''The Evolution of the number of epidemics of infectious diseases in the world from 1950 to 2010'').