T HE Keynesian concept of liquidity-preference is not a single idea, but like a prism has several facets. The views produced by a prism depend on the way it is turned and whether it is in motion or perfectly still. Users of the concept of liquidity-preference, like a person with a prism in his hand, may unconsciously turn it around and produce confused and changing images when they are seeking precision and clarity. This kind of effect appears to have been produced by Mr. James Tobin in his article, "Liquidity Preference and Monetary Policy." ' However, the problem which Mr. Tobin attempts to clarify changes in the desire to hold cash balances, and their bearing on the effectiveness of monetary policy in maintaining or expanding money national income is one which needs the best possible illumination. In this article I propose to examine four aspects of this problem. First, I shall distinguish between the three chief facets of the concept of liquidity-preference, and show that failure to recognize the differences between these concepts has led Mr. Tobin, as well as his predecessors, into an ambiguity in his discussion of the "L" function, expressing the demand for cash balances as a function of the rate of interest. Along with this discussion I shall mention some inadequacies in the Keynesian analysis which have an important bearing on Mr. Tobin's presentation or on his criticism of my views. Second, I shall elaborate the nature of the liquidity-preference assumption underlying my "Monetary Theory of Deficit Spending," 2 inasmuch as Mr. Tobin's presentation of my view is inaccurate. Third, I shall examine the statistical data used by Mr. Tobin to support his position, and present some additional data bearing on the problem. Fourth, I shall add a few comments on the aims and techniques of monetary policy.