In the first place there would be an elimination of risk. If the average loss from exchange on funds sent out to India for the busy season works out at (say) 2 per cent per annum, the Banks, in order to recompense themselves for the risk of fluctuations beyond the average, would be able to make a difference of more than 2 per cent between the current Indian and English rates. In the case of funds borrowed in terms of rupees and repayable in terms of rupees, this element of risk is absent; and the elimination of it provides a source of net gain. If the effect of Government lending in India were to mitigate the seasonal stringency there, some lowering of the normal upper limit of fluctuation of exchange might result. In so far as this was the case, in normal years the consequences would be outwardly similar to those of the first alternative, discussed and rejected above, whilst the Government would not have bound themselves by any undertaking capable of turning out burdensome. Secondly, the rate of interest which the Secretary of State can earn on loans in London is appreciably lower, on account of the short period for which he lends and the nature of the security he requires, than the normal rate at which the Exchange Banks would raise their funds there, and a good deal lower than what would be obtained by direct lending in India. (It should be admitted, on the other hand, that the practice of lending funds in India would probably involve some sacrifice of perfect safety as compared with the present arrangements.) And, thirdly, it is not clear that it might not sometimes be feasible to lend out in India sums additional to those which would in fact be released under the present system, so that there would be some net addition to the resources available in India.