If the Government’s Reserve is held purely to support the currency, then the maximum volume of rupees and notes, which could, so far as one can anticipate, be spared from the circulation and tendered to the Government for exchange, sets an upper limit to the necessary amount of this Reserve. If, on the other hand, it is intended to act as a banking reserve and to ensure India’s ability to meet her international obligations at all times, then its upper limit is set by the probable maximum amount of the adverse balance which could arise against India for immediate payment. 18. I will begin by discussing this question on the first hypothesis—that what the Government has been accumulating is intended to serve as a currency reserve only—and will return later to the problem of a reserve held for wider purposes, and of the possible magnitude of the balance of international indebtedness against India. 19. To estimate the demand that the reserves might have to meet merely in order to support the currency, the existing volume of currency is what we chiefly require to know. For this sets, or suggests, a limit to the maximum amount which can possibly be spared from the active circulation. Attempts to estimate the rupee circulation of India have been the occasion of some very interesting calculations. For many years past (since 1875) an annual census of rupees has been taken by examining in each Government Treasury a bag containing 2000. This enabled Mr. F. C. Harrison, when he was Comptroller of Currency, to apply the Jevonian method very fully; and he was also able to corroborate his estimates by reference to the numbers of the older issues, 1835 and 1840 (_e.g._), actually withdrawn from circulation on the occasions when the Mint recalled them. Mr. Harrison’s results were checked by the labours of a later Comptroller of Currency, Mr. Adie, who applied to the same material two alternative methods of much greater technical complexity than Mr. Harrison’s.[66]