In this paper certain basic assumptions about causation will be made explicit and then an attempt will be made to show, on the basis of those assumptions, that causation must be supplemented by an extra-causal operation of a selective nature. It is hoped that our assumptions will have in their own right a certain general acceptability, but since assumptions or “postulate sets”, whether antecedently acceptable or not are usually chosen for the purpose of proving something already deemed desirable, a word should be said as to why it was thought desirable to argue for selectivity.