216

A TREATISE ON MONEY

BK. Ill

For example, let us imagine an automatic inter-national gold system in which the quantity of moneyis solely determined by the amount of gold in theCentral Bank , and Bank-rate is left to find its own levelas a result of free competition between borrowers forthe amount of bank-money available on the basis ofthe amount of gold in the Central Bank . It thenfollows from the above that in such conditions (assum-ing an absence of economic frictions and of time-lag,and in particular that the rates of money-earnings arefree to move in response to the competition of entre-preneurs for the services of the factors of production)an amount of gold will flow in which is just sufficientto establish a price-level and a bank-rate at homewhich shall be such, relatively to price-levels andbank-rates abroad, as to keep both S = I and L = Bat the same time.

(v.) The Relation of Bank-rate to the Quantityof Money

In a free loan-market ( i.e. in the absence of therationing of loans) a given level of bank-rate, takenin conjunction with all other relevant factors, mustbe uniquely correlated, if it is to be effective, with agiven quantity of bank-money. That is to say, everyeffective alteration of bank-rate must be associated,except in so far as it is balanced by simultaneousalterations in other factors, by some alteration in thequantity of bank-money.

But there is no simple or invariable relation be-tween the effect of an alteration of bank-rate on theprice-level, whether of liquid consumption-goods or ofoutput as a whole, and the associated alteration inthe quantity of bank-money. The volume of bank-money must be in due relation to the volume of out-put, the rate of earnings, the rate of profit, thevelocities of different classes of deposits, and the