The second defect of convertible treasury notes is the fact that they impose upon the government the necessity of keeping on hand a reserve of standard coin sufficient to meet at any and all times the demand for redemption. There are two objections to the assumption by government of such an obligation.

The first is that it may be a source of great expense. The coin reserve must be accumulated by taxation or loans, and in either case the people must pay the bill. If, as has been proposed, in order partially to overcome this difficulty, the notes are never paid out except in exchange for standard coin, their utility as a means of expanding the currency is destroyed, the quantity of notes put into circulation being exactly equal to the amount of coin withdrawn.

It has been claimed that the expense of a coin reserve is more than offset by the profit accruing to the government from the payment of a portion of its obligations by means of non-interest-bearing notes only partially covered by the reserve. This seems to be the case when we consider simply the beginning of the history of the notes. If the reserve amounts to but one-third of the issues, there is obviously at the beginning a clear profit amounting to two-thirds of their face value. Suppose, however, that the redemption of the entire issue should be demanded three times in the course of a year. The original reserve would be exhausted when one-third of the notes had been redeemed once, and must be replenished from one of three sources: either from money coming into the treasury through the channels of its ordinary receipts, from the product of a special tax, or from the sale of bonds. The first source obviously cannot be relied upon. It is rarely practicable for the government to exact standard coin for all payments due to it, especially when it is trying to circulate its own notes. It would be the merest chance, therefore, if the requisite amount of standard coin were paid in by the government's creditors. But certainty on this point would not remove all difficulties, since the coin paid in might be required for ordinary expenses, in which case it would not be available as a redemption fund. Even the adoption of the policy of surplus financiering would not insure the existence of a coin surplus just at the times when redemption of notes might be demanded. The levy of a special tax for the replenishment of an exhausted reserve is quite out of the question. The machinery of taxation is necessarily slow in its operation, and cannot be relied upon to furnish funds for sudden exigencies, such as the one we are considering.

The sale of bonds in such an emergency seems, therefore, to be necessary, and this involves the payment of an annual interest charge which may much more than counterbalance the profit accruing from the original issue of the notes. If, according to our assumption, redemption three times in the course of a year were demanded, and bonds payable one year or more after date were issued to procure the necessary coin, in the course of a year the face value of the bonds outstanding would be three times as great as that of the notes, and the expense incurred by the government would be treble the ordinary rate of interest on the face value of the notes.

If we assume with the advocates of this form of currency, that redemption in large quantities is an unusual and rare occurrence, and consequently that the expense involved in the issue of bonds for the replenishment of the reserve is not often incurred, we must still remember that danger from this source is always imminent, and hence that such notes at the best always introduce an element of uncertainty into the finances of the government. Whenever for any reason the reserve falls below its ordinary limit, the ability of the government to maintain specie payments is placed in question. No one knows what the Secretary of the Treasury may do or what action Congress may take. A "panicky" feeling is, therefore, liable to be produced, which is certain to affect the value of all sorts of negotiable securities, to render business unstable, and, if other disturbing circumstances concur, to produce a commercial crisis.

Our own experience with these notes illustrates the objections which have been mentioned, and furnishes the best argument against their use. Though the amount of the gold reserve was not fixed by law, it early became customary to keep on hand at least a hundred millions of dollars. Between 1879, the date of the resumption of specie payments, and 1890 the Secretaries of the Treasury were aided in the maintenance of this reserve by an annual surplus of revenues over receipts and by a constantly increasing demand for currency, occasioned in part by our rapidly expanding commerce and in part by a decrease in the circulation of bank-notes. In spite of these favouring circumstances, however, in 1885 and 1886* the reserve fell to near the danger point and occasioned anxiety. In 1890 occurred two events which soon disclosed the danger to which Congress had subjected the treasury and the country when it assumed the obligation to maintain a currency of convertible treasury notes. One was the passage of the Sherman act, to which reference has already been made, and the other was the discontinuance of the policy of surplus financiering occasioned by the passage of the so-called McKinley bill. The former act added in three years $150,000,000 of new notes to the $346,000,000 of greenbacks, thus dangerously increasing the burden placed upon the gold reserve, and the latter soon substituted an annual deficit for the former surplus. The gold reserve showed the effect of the changed conditions. It fell in 1892 to $114,000,000, in October, 1893, to $81,551,385, and in 1894 to $68,000,000.+ The "panicky" conditions thus produced, combined with various other circumstances, brought on a crisis in 1893, which was accompanied by a large foreign demand for gold. The result was the establishment of the so-called "endless chain," which revealed in the most complete fashion the true nature of our currency of treasury notes. The gold reserve was first reduced far below the $100,000,000 limit by the redemption of notes. The treasurer was, then, compelled to sell gold bonds in order to replenish it, and the redeemed notes were put again into circulation in the usual manner. This process was four times repeated, and it was discovered that, in some instances at least, the very firms who had purchased bonds and thus transferred gold to the Government withdrew it very soon thereafter by presenting notes for redemption. In all, bonds to the amount of nearly $300,000,000 were issued, and the endless chain* was finally broken by a contract with a syndicate of foreign bankers by which they agreed to supply gold to the treasury in the payment of bonds sold them at profitable rates, and to prevent, so far as possible, the presentation of notes for redemption. The danger of a repetition of this humiliating and expensive experience was partially removed in 1900 by an act which fixed the minimum gold reserve at $150,000,000 and took from the Secretary of the Treasury the power to reissue redeemed notes except in return for gold coin. As has already been remarked, the efficiency of a regulation of this sort depends upon the extent to which it transforms the notes into ordinary gold certificates, and its necessity accordingly is a proof of the practical impossibility of maintaining a currency of this sort without endangering the safety of our financial institutions and the credit of the government.

* Taussig's "Silver Situation."

+ Noyes's "Thirty Years of American Finance."