At times it is inevitable that cost of service and value of service considerations come flatly into opposition. Usually, as in the California raisin case or in the grant of low rates on Oregon lumber east bound about 1893, they reinforce one another; that is to say, the lower rate given to build up business obtains on a service given at lower cost. But it sometimes happens that shipments of the same commodities over a line in opposite directions may occur and that the lower rate applies to the (presumably) more costly service. In 1906 a manufacturer in Menasha, Wis., complained to the Interstate Commerce Commission[155] that his rates on woodenware to the Pacific slope were ten cents per 100 lbs. higher than were rates on the same goods between the same points east bound, _notwithstanding_ the fact that the empty car mileage west bound was then three times as great as in the contrary direction. The movement of empties west bound would certainly seem to justify as low if not lower rates on the basis of comparative cost of operation, supposing that there was coincidence in time. Only one satisfactory explanation for this apparent anomaly suggests itself; viz., that this low eastbound rate was given to build up a new industry in the West. In other words, the cost of service, a dependable guide for a road in a static condition, failed of effect upon a line possessed of great dynamic possibilities. Occasionally opposition of principles like this may occur in questions of classification. It may temporarily be worth while, in order to build up a new industry, to accord a lower rating to a commodity actually more valuable or more expensive to handle than others. Here again the dynamic force in the value of service principle out-weighs all other considerations of relative cost of service. The value of service principle in general fails, not only in the determination of absolutely reasonable rates, but it is inadequate also to the solution of perhaps the more difficult problem of relative rates. This question of relativity is twofold; first as between different places, and secondly as between different commodities. These are, in other words, the problems respectively of distance tariffs and of classification. The manner in which distance tariffs evolve, has already been discussed, and it is evident that the cost of service principle is of fundamental importance, even though it be tempered by considerations of commercial expediency, that is to say, by the necessity of at all times under stress of competition, charging only what the traffic will bear. But while the value of service principle--charging according to demand in other words--applies at the competitive points, the other principle of relative cost should be the fundamental one in fixing upon the scale of local non-competitive rates.