In an earlier article,1 an attempt was made to establish the process by means of which Keynes' equation between the rate of interest and the marginal efficiency of capital (or expected rate of profit) would be brought about. It was established that the conditions under which such an equality could be exactly achieved were somewhat restrictive and various simplifying assumptions had to be employed-rational behaviour by businessmen, perfect foresight, a constant demand for money to hold, interest-elastic investment, perfect shiftability of funds, a single rate of interest, and perfect mobility of resources other than money. Even under these conditions, the level at which this equation would be achieved would certainly vary from time to time. When, however, this framework of simplifying assumptions was removed, there was good reason to believe that on many occasions the influence of the interest factor is definitely swamped by other more obvious considerations . If we allow that the businessman does pay heed, even though only to a limited extent, to the costs of raising finance, it is necessary to emphasise, first, that no two entrepreneurs will have access to financial resources on exactly the same terms and, second, even the same entrepreneur will generally find that these costs will be conditioned by the purpose for which he borrows. (In simple terms, rates will usually be different for short-, mediumand long-term borrowings.) Furthermore, not only is there a whole range of interest rates, but also a variety of conditions which may be attached to a loan. Each transaction, therefore, has its unique characteristics. As a result, for each borrowing operation there is one particular rate of interest, which is the Relevant Rate of Interest, and, to the extent that the entrepreneur acts rationally, he will attempt to secure an equation at the margin between this rate and his expected rate of profit. When developing this analysis, no specific mention was made of credit rationing. My attention was drawn to this lacuna in the analysis by my colleague Professor F. W. Paish. I was disposed to believe at the time that the introduction of this complication would require no material modification of the argument. Upon reflection, however, there would seem to be certain peculiarities which do call for further investigation. Sufficient was said in the earlier article about the Relevant Rate of Interest for us to be tolerably clear about its definition. Can we