Forecasts are answers to questions of the form "What will happen to x if p is the case ?", "When will x happen if p obtains ?", and the like. In science such answers are called predictions and are contrived with the help of theories and data: scientific prediction is, in effect, an application of scientific theory. Prediction enters our picture of science on three counts: (i) it ">anticipates fresh knowledge and therefore (ii) it is a test of theory and (iii) a guide to action. In this chapter we shall be concerned with the purely cognitive function of prediction, i.e. with foresight. The methodological aspect of prediction (its test function) will concern us in Ch. 15, and the practical side (planning) in the next chapter.