In the illustration above we have assumed that a monopoly was starting _de novo_. Where a Trust is formed, as is commonly the case, by an amalgamation of existing capitals largely embodied in plant and machinery of production, it will probably not pay to limit production to a very small output, even though the largest proportionate margin of profit might seem to stand there. For the interest upon the closed mills and other idle capital should be reckoned among the expenses of production for the purposes of determining the profitable price. Thus where large means of production are owned by a monopoly it will seldom pay to sell a very small supply at a very high price. So far we have treated of absolute monopolies, eliminating all consideration of competition. We have found that the supply and the price of an article of absolute monopoly is determined by the relation between expenses of production and flexibility of demand. Although a new invention or a wide expansion of market may alter so considerably the expenses of production of the several quantities of supply as to materially affect monopoly-supply and prices, it is the latter influence, that of flexibility of demand, that directly in each specific case determines whether a Trust's prices shall be high or low. When we find the Standard Oil Trust maintaining a low level of prices, or the Western Union Telegraph Company charging low rates, we shall find the explanation in the character of the public demand for oil and telegraphic messages. § 7. A number of considerations relating to "demand" limit the economic power of monopolies to charge high prices. A monopoly price, as we have seen, exactly measures the marginal utility of the supply, as indicated by the quantity of money which the purchaser of the last increment of supply is just willing to pay for it. When this marginal utility sinks fast with an increase of supply the monopoly price will be high for it, and it will pay the monopolist better to restrict the output and sell the limited supply at a high price, because a large reduction of price will not stimulate a proportionably large increase of consumption. So where the marginal utility sinks slowly, it will pay to increase the supply and lower the price, for each fall of price will stimulate a large increase of consumption. Since the marginal utility of a number of increments of supply will not be the same in the case of any two commodities, it is evident that the determination of monopoly prices is a very delicate operation. It is not possible to present even an approximately accurate classification of commodities in relation to the powers of a Trust or Monopoly. But the following considerations will assist us to understand why in some cases a Trust appears to raise prices, in others to keep them as they were, and in others even to lower them:--