16. Let us consider first how heavy a loss and expense the popularity of a gold currency might involve. During the last twelve years the Government have been able to accumulate a sum of about £21,000,000 sterling from the profits of rupee coinage; and the interest on the invested portion of the Paper Currency Reserve is now about £300,000 annually. Thus the annual income, derivable from the interest on the sums set free by the use of cheap forms of currency, amounts already to about £1,000,000. With the rapidly increasing use of notes, this income should show a steady growth in the future. Both these sources of profit would be gravely jeopardised if the introduction of an Indian gold coin were to meet with any considerable measure of success. It would be specially unfortunate if a competitor to the paper currency were to be introduced, before the virtual abolition of the system of circles has had time to have its full effect in the direction of popularising the use of notes. 17. Advocates of a gold currency, however, would not, I think, deny that it might involve the country in some extra expense. They support their policy on the ground that it would do a great deal to ensure the stability of the currency system, and that it is worth while to incur some expense for this object. I think it is possible to show that such a policy is likely on the whole to have an exactly opposite effect. It is suggested that the currency should be composed of rupees, gold, and paper, with rupees still predominating, but consisting of gold in a considerably higher proportion than at present. This greater infusion of gold would necessarily be at the expense either of the Currency Reserve or of the Gold Standard Reserve. If the gold replaced notes, the former would be diminished, and, if it replaced rupees, the latter. It is tacitly assumed that the greater part of what has to be withdrawn from the circulation at a time of crisis would come from the gold portion of the circulation. This assumption seems to me to be unwarranted and contrary to general experience. At a time of crisis it is the fiduciary coins with which the public are most eager to part. Bankers and others would keep as much of their surplus currency as they possibly could in the form of gold, and it would be rupees (in great part) and not gold that would be paid into the Government Treasuries. Thus the infusion of more gold into the circulation would necessarily weaken the existing reserves and would not correspondingly reduce the amount of such reserves which Government ought in prudence to keep. When it became necessary to contract the volume of currency, Government would be in a _worse_ position than at present, unless the greater part of what was withdrawn came from the gold portion of the circulation and not from the rupee or paper portion. This is not an expectation upon which it would be prudent to act.