But German economists are of interest, inasmuch as they have established a new school who urge the use of the historical method in political economy, and it is about the question of method that much of the interest of to-day centers. In 1814 Savigny introduced this method into jurisprudence, and about 1850 it was applied to political economy. The new school claim that the English “orthodox” writers begin by an _a priori_ process, and by deductions reach conclusions which are possibly true of imaginary cases, but are not true of man as he really acts. They therefore assert that economic laws can only be truly discovered by induction, or a study of phenomena first, as the means of reaching a generalization. To them Bagehot(80) answers that scientific bookkeeping, or collections of facts, in themselves give no results ending in scientific laws; for instance, since the facts of banking change and vary every day, no one can by induction alone reach any laws of banking; or, for example, the study of a panic from the concrete phenomena would be like trying to explain the bursting of a boiler without a theory of steam. More lately,(81) since it seems that the new school claim that induction does not preclude deduction, and as the old school never intended to disconnect themselves from “comparing conclusions with external facts,” there is not such a cause of difference as has previously appeared. Doubtless the insistence upon the merits of induction will be fruitful of good to “orthodox” writers, in the more general resort to the collection of statistics and means of verification. It is suggestive also that the leaders of the new school in Germany and England have reached no different results by their new method, and in the main agree with the laws evolved by the old English school. The economist does not pretend that his assumptions are descriptions of economic conditions existing at a given time; he simply considers them as forces (often acting many on one point or occasion) to be inquired into separately, inasmuch as concrete phenomena are the resultants of several forces, not to be known until we know the separate operation of each of the conjoined forces.