equilibrium, _caeteris paribus_, will be at a higher price, to be sure, but with the same amount supplied and demanded. "Demand" is a term which carries the connotation of motivating power in economic theory. Through demand run the forces which regulate production and supply. The function of increased demand is to induce increased supply. But the value concept, and the assumption of a fixed value of money, are needed to preserve this part of the doctrine. Without them we have no way of distinguishing a _real_ increase in demand in the active sense, which does modify the adjustments in production, and alter the proportions of different supplies, from a _nominal_ increase in demand in the active sense, which merely raises a money-price, without affecting supply.[45] Another approach will lead to the same conclusion. Demand and supply-curves are not to be understood merely in terms of brute, physical quantities. They are rather curves expressing economic _significances_, manifesting _psychological_ forces which lie behind them. No considerations of mere physical quantity will explain why one demand-curve should be "elastic" and another inelastic,--each curve has its own peculiarities, which are not mechanical in their nature. Demand-curves express the diminishing economic significance of goods as their quantity is increased. How economic significance is to be interpreted need not be argued here. I have elsewhere undertaken to show that the utility theory of value does not explain the economic significance which demand-curves express--that demand-curves are not utility curves. My own theory is that demand-curves are to be explained only in terms of a social psychology, that demand-curves are social-value curves. But my argument at this point does not rest on the particular type of causal theory of value one chooses. It is enough that the demand-curve be recognized as expressing economic significance, and diminishing economic significance.[46] But for the demand-curve to express variation in economic significance of a good, there is need for a unit in which to express that variation. That unit is the economic significance of the dollar, itself assumed to be invariable--as all measures must be assumed to be invariable if measurement is to mean anything. If the unit chosen vary in the course of a given investigation, the curve tells you nothing at all.