DAVID HUME has described his theory of causation as the 'chief argument' of his Treatise of Human Nature. The broad lines of that argument are well known, and need not be detailed here. Hume's conclusion is that causation is not a ' power ' in the cause but a ' felt compulsion ' in the mind- an expectation that a certain event will be followed by a certain other event of the type habitually associated with the first in our experience. Both events are perceptions; the first is an impression, the second a believed idea: and causation is the association of the two. Furthermore, Hume not only expresses his notion of causation in terms of the theory of impressions and ideas; he also argues to it by means of this theory. I take the liberty of making these assertions without argument in the belief that they will be generally accepted: but I should like to argue the following propositions: I. that the theory of impressions and ideas, which Hume utilises in his later discussion of causation, is based on that precise theory of causation which the later discussion is designed to refute. II. that Hume's discussion of association makes appeal to a variety of causal theories. We shall consider these in turn.