order to justify any one in saying that a bimetallic circulation could exist in spite of it. Therefore it seems to me that the more accurately the bimetallic system was established the more delicate and more easily overthrown it would be, while if it was not accurately established it would not come about at all. I submit that such a result is one of the notes of an absurdity in any science.
An analogy has been suggested in illustration and support of the bimetallic theory that two vessels of water connected by a tube tend to preserve a level. I have already indicated my suspicion of all analogies, but I will alter this one to make it fit my idea of bimetallism. Suppose two vessels capable of expansion and contraction to a considerable degree, under the operation of forces which act entirely independently of each other, so that the variations in shape and capacity of each may have all conceivable relations to the corresponding variations of the other. Suppose further that each is fed by a stream of water, each stream being variable in its flow and the variations of each having all possible relations to the variations of the other. The fluctuations in capacity may represent fluctuations of demand, and the fluctuations of inflow, fluctuations of supply. Would the water in the two vessels stand at the same level except temporarily and accidentally, even though the two vessels were connected by a tube? The analogy of the connecting tube could not be admitted even then, because it brings into play the natural law of the equilibrium of fluids, to which the legal tie between the metals is not analogous. If we desire to make the analogy approximately just, in this respect, we may suppose that each vessel has an outlet and that a man is stationed to open the outlet of the vessel in which the water is at the higher point so as to try to keep them both at a level. It is evident that his utmost vigilance would be unavailing to secure the object proposed. I do not borrow the analogy
