It would still form only a small part of the total cost of producing and marketing them. But to add anything like that sum to the cost of one hundred pounds of salt or cement would put an end to the business at once. Only about so much can in practice be added to the price of any given commodity for freight without widely limiting the area of its available market. Thus raw cotton seems to be able to bear an addition of about fifty cents per hundred pounds for freight to its total cost. Experience demonstrates that anything more than this one-half cent per pound charged on cotton, entails more loss than gain. In the case of fancy groceries or fine furniture, there may be no considerable demand in any event above a certain ascertainable level of prices. For boots and shoes or cut building stone it may be that competition from some other centre of production nearby, precludes any great addition to the price for freight. The business simply will not bear more than a certain proportion of charge. Not only would the rigid application of the cost of service principle hinder all transportation of low-grade traffic; it would also prevent any development of long distance business. It is indubitable that sole reliance upon cost of service as a basis for rate making is theoretically unsound, and impossible of practical application.[142] Cost of service, while unsound as a sole reliance, nevertheless affords an important check upon the value of service principle. Without it there is always grave danger that traffic managers, seeking to enlarge their revenues, may push rates unreasonably high. At first sight it would appear as if this could not occur, inasmuch as an inordinately high rate would immediately reduce the volume of business offered. It is constantly alleged by railway men that this must of necessity occur. And it would indeed follow, were it not that the incidence of the rate is rarely upon the actual shipper. He merely pays it, and at once shifts it to the consumer. For low-grade or staple goods like cement or kerosene, where transportation charges form a large part of the total cost of production, it is conceivable that higher freight rates might so far increase the price as to check consumption. Five cents a hundredweight higher freight means $1.25 per 1,000 ft. added to the price of soft lumber, $2 to hard lumber; three cents per bushel added to the price of wheat, and $1 to the ton of pig iron or coal. Such substantial additions might readily reduce the demand. Yet even this would not be true of necessities of life like anthracite coal or sugar, on which latter the freight rate amounts to about one-half cent per pound. Is five cents a barrel added to the price of flour likely to decrease the consumption of that staple commodity? Yet the enhancement of railway revenues would indeed be enormous from such an increase of freight rates. For these necessities of life, an increased