There can be no question that for an indispensable public service like transportation, conducted under monopolistic conditions, the ideal system of charges would be to ascertain the cost of each service rendered and to allow a reasonable margin of profit over and above this amount. To the application of this principle alone, however, there are several insuperable objections both theoretical and practical. Such cost is practically indeterminate, being joint for all services in large part, as we have seen: it is highly variable, being perhaps never twice the same, as circumstances change from time to time; cost is unknown until volume is ascertained, and volume is ever fluctuating; the cost of service, obviously, could never be ascertained until after the service had been rendered, while, of course, the schedule of rates must be known in advance, in order that the shipper may calculate his probable profits; and finally the principle of increasing returns, flowing from the dependence of cost upon volume of traffic, imposes such an incentive for development of new business, which in turn depends for its volume upon the rate charged, that cost of service is subordinated at once to other considerations in practice. Of these objections to rate making upon the principle of cost of service alone, it would indeed appear as if the first should be conclusive. If the cost is simply indeterminable, why bother about any further refutation of the principle at all? But the persistency of the idea that somehow railway operations are analogous to the business of an ordinary merchant; and that cost and profits are ascertainable; renders it necessary to pile proof upon proof of the limitations upon its applicability to real conditions in service. Not only is the mere cost of service indeterminable; if it could be ascertained, it would not establish the chargeable rate in many instances. The freight service of a railway comprises the carriage of all kinds of goods simultaneously, from the most valuable high-priced commodities, such as silks and satins, down to lumber, coal, cement, and even sand.[141] To compel each of these classes of goods to bear its proportionate share of the cost of carriage, would at once preclude the possibility of transporting low-priced goods at all. One dollar a hundred pounds may not be too much to add to the price of boots and shoes for transportation from Boston to Chicago.