Investment is a term used very loosely in the economic literature and all too frequently without any attempt at definition. Yet it is possible to defend an attitude which refuses to regard precise definition as an end in itself. The businessman, for example, rightly regards ‘investment’ not as a logical category with strictly definable limits, but as a rough description of a type of activity. To the businessman and to the economist, it is the purpose of any particular expenditure which matters. And, indeed, if we take a representative sample of definitions of investment, they all have this in common. Fraser, for instance, defines investment in a money economy—which is the sort of economy we are concerned with—as ‘the use of monetary resources for the acquisition of wealth of a relatively illiquid type’ 2 and Keynes obviously had a similar definition in mind when he divided current output into ‘(a) the flow of liquid goods and services which are in a form available for immediate consumption and (b) the net flow of increments (after allowing for wastage) to capital goods and loan capital . . . which are not in a form available for consumption’. 3 But what are goods of ‘a relatively illiquid type’? What is ‘capital’? Must we not define ‘capital’ before we can define ‘investment’ and as a condition of defining investment?