The spread of a classification,--that is to say, the graduation of rates as between all kinds of goods, from fine silks to lime and sand, or from aeroplanes, "set up," to pig iron, "knocked down,"--is not constant. How shall this be theoretically justified? At first sight it would appear as if the relativity of charges between different things, as determined by cost or value of service, ought to remain fixed; that is to say, for example, that rates on raw hides fairly standing at one-half of the charge for shoes, ought to remain always and everywhere at this ratio. Advocates of a rigid classification prescribed by public authority seem often to assume that this could be brought about. But a moment's consideration of the nature of a tariff as it has already been described will show that this is impossible. The spread or gradation, far from being fixed, must in the nature of things ever vary from place to place with change of trade conditions. The rate on raw hides relatively to that on shoes in New England--the centre of manufacture for footwear--should be very different at Kansas City or Chicago, whence the raw hides are derived: different alone, if for no other reason than because hides, moving east, progressively add the cost of carriage the farther they go; while with shoes the augmentation of value goes on in the opposite direction, geographically. True as between commodities, the same inconstancy of ratio also holds good as between different points along a given line. The rate from New York to Durham, North Carolina, for example, on first-class freight may be fifteen per cent. above that for freight of the second class; the second class maybe twenty per cent. above that of third class for this distance, etc.; yet the divergence between these same classes for another distance, as between New York and Jacksonville, Florida, may be quite different,--twenty per cent. between first and second class, twenty-seven per cent. between second and third, and so on. This is indeed rather a difficult matter to understand.