Analyzing the main question two propositions are certain. Firstly, the long line can never charge _more_ than the short line; whence it follows that as the short line reduces its rate, the long one must accept that rate as made; and, secondly, the long line, costing more to operate, is, in the process of reduction of rates, bound to be the first to strike the bed-rock of cost incident to that particular service. To go below this point of particular cost would obviously be indefensible from every point of view. The general rule, then, is that "the short line rules the rate." This is accepted widely in practice, as for example throughout trunk line territory and between the so-called Missouri-Mississippi river points, where the short line from Hannibal to St. Joseph determines all rates by longer routes.[243] But the problem yet remains unsolved. The long line may never be able to charge _more_ than the short line--may it, however, charge _less_ under certain conditions? The moment it is enabled to do so, the long line and not the short line, for the moment, "controls the rate." If, now, we use the technically proper terms, the question becomes this: Under what circumstances is average cost of service in railway competition set aside in favor of other considerations; or, otherwise stated, when may a line, operated under a disability as to cost, properly give a lower rate than its competitors notwithstanding? Does disability justify a handicap or the reverse? This was the form which the question assumed in the notable Milk Rate case: as to whether the weaker lines in respect of distance or grades should be allowed compensation therefor by permission to charge higher rates. One of the common instances of rate control by a line operating under a disability as to distance or normal cost is the competition of a bankrupt with a solvent property. The "roundabout" line, like the Erie or the old New York and New England, having repudiated its fixed charges, undoubtedly "makes" the rate which the other roads must meet or lose the traffic. Usually they prefer to absorb or control it otherwise, financially, thus substituting monopoly for a ruinous condition of competition. Yet such instances resolve themselves, evidently, into questions of relative cost of operation after all. The bankrupt road holds the whip hand, because, having repudiated its fixed charges, its average costs of operation are correspondingly reduced. The validity of operating cost as a basis of charges is surely not shaken by this exceptional case.