HUME TO HEGEL would beneWt: instead of taking small forts on the frontiers of knowledge, we would now be able to ‘march up directly to the capital or centre of these sciences, to human nature itself ’ (T ) The Wrst book of the Treatise begins by setting out an empiricist classiWcation of the contents of the mind (‘perceptions’) This covers much of the same ground as Locke and Berkeley’s epistemology, but Hume divides perceptions into two classes, impressions and ideas Impressions are more forceful, more vivid, than ideas Impressions include sensations and emotions; ideas are perceptions involved in thinking and reasoning Hume treats in detail ideas of memory and imagination, and the association between them He endorses and reinforces Berkeley’s criticism of Locke’s abstract ideas After a second part devoted to the ideas of space and time1 Hume presents, in a section entitled ‘Of Knowledge and Probabilty’, his most original and inXuential thoughts All knowledge which extends beyond the immediate deliveries of the senses, Hume argues, depends upon the notions of cause and eVect: it is through those ideas that we discover what happened in the past and conjecture what will happen in the future We must therefore examine closely the origin of these ideas The idea of causation, he says, cannot arise from any inherent quality of object, because objects of the most diVerent kinds can be causes and eVects We must look, instead, for relationships between objects; and we Wnd that causes and eVects must be contiguous to each other, and that causes must be prior to their eVects Moreover, contiguity and succession are not enough for us to pronounce two objects to be cause and eVect, unless we see that objects of the two kinds are found in constant conjunction But that is not enough: if we are to infer an eVect from its cause, we feel, there must be a necessary connection between a cause and its eVect After many pages of artful argument, Hume leads us to an astonishing conclusion: it is not our inference that depends on the necessary connection between cause and eVect, but the necessary connection that depends on the inference we draw from one to the other Our belief in necessary connection is not a matter of reasoning, but of custom; and to wean us from the contrary doctrine Hume presents his own analysis of the relationship between reason and belief He rounds oV the book on See Ch below 82