Sale by a New York banker of a draft on Paris, say, for francs 25,250, at 5.17-1/2, bringing him in $4,879.23. Purchase by same banker of a draft on London for £1,000, at 4.84, costing him $4,840. Instructions by the American banker to his London correspondent to buy a check on Paris for francs 25,250 in London, and to send it over to Paris for the credit of his (the American banker's account). Such a draft, at 25.25 would cost just £1,000. The circle would then be complete. The American banker who originally drew the francs 25,250 on his Paris balance would have replaced that amount in his Paris balance through the aid of his London correspondent. The London correspondent would have paid out £1,000 from the American banker's balance with him, a draft for which amount would come in the next mail. All parties to the transaction would be satisfied--especially the banker who started it, for whereas he paid out $4,840 for the £1,000 draft on London, he originally took in $4,879.23 for the draft he sold on Paris. Between such cities as have been used in the foregoing illustrations rates are not apt to be wide enough apart to afford any such actual profit, but the chance for arbitraging does exist and is being continuously taken advantage of. So keenly, indeed, are the various rates in their possible relation to one another watched by the exchange men that it is next to impossible for them to "open up" to any appreciable extent. The chance to make even a slight profit by shifting balances is so quickly availed of that in the constant demand for exchange wherever any relative weakness is shown, there exists a force which keeps the whole structure at parity. The ability to buy drafts on Paris relatively much cheaper at London than at New York, for instance, would be so quickly taken advantage of by half a dozen watchful exchange men that the London rate on Paris would quickly enough be driven up to its right relative position.