Some writers have satisfied themselves with general opinions — guesses, I am obliged to call them — that if the fluctuations were kept within certain limits the concurrent circulation would stand. They probably rely on an element analogous to friction which unquestionably acts in economy and finance. This element consists of habit, prejudice, passion, dislike of trouble. It acts with great force in retail trade, and in individual cases, and in small transactions. Its force diminishes as we go upwards towards the largest transactions, where the smallest percentages give very appreciable sums. It seems to me that the bimetallic system reduces this friction to a minimum. If a man has to spend a dollar he does not go to a broker to buy a trade dollar with a greenback dollar, and save a cent or two, but if he has both a gold dollar and a silver dollar in his pocket (and, under the bimetallic system, the chances are that when he has two dollars he will have one of each), it needs only the lightest shade of difference in value to determine him which to give and which to hold. A bank of issue, holding equal amounts of the two metals with which to redeem its notes, would find an appreciable profit in giving one and holding the other, and it would require nothing but a word of command to the proper officer, involving no risk at all. Hence I say this friction would be reduced to its minimum under the bimetallic system. It is astonishing what light margins of profit suffice to produce financial movements nowadays; and the tendency is to make the movements turn on smaller and smaller margins. Five in the thousand above par carries gold out of this country. Four in the thousand carries it from England to France. When the French suspended specie payments a depreciation of two in the thousand on the paper sufficed to throw gold out of circulation. A variation in the ratio of metals from 15.5:1 to 15.6: 1 is a variation of six and one-half in the thousand. I do not see how small a variation must be in
