The moral of this discussion, to be carried forward in the reader’s mind until we reach Chapters IV. and V., is that the price level is not mysterious, but is governed by a few, definite, analysable influences. Two of these, _n_ and _r_, are under the direct control (or ought to be) of the central banking authorities. The third, namely _k_ and _k´_, is not directly controllable, and depends on the mood of the public and the business world. The business of stabilising the price level, not merely over long periods but so as also to avoid cyclical fluctuations, consists partly in exercising a stabilising influence over _k_ and _k´_, in so far as this fails or is impracticable, in deliberately varying _n_ and _r_ so as to _counterbalance_ the movement of _k_ and _k´_. The usual method of exercising a stabilising influence over _k_ and _k´_ especially over _k´_, is that of bank-rate. A tendency of _k´_ to increase may be somewhat counteracted by lowering the bank-rate, because easy lending diminishes the advantage of keeping a margin for contingencies in cash. Cheap money also operates to _counterbalance_ an increase of _k´_, because, by encouraging borrowing from the banks, it prevents _r_ from increasing or causes _r_ to diminish. But it is doubtful whether bank-rate by itself is always a powerful enough instrument, and, if we are to achieve stability, we must be prepared to vary _n_ and _r_ on occasion. Our analysis suggests that the first duty of the central banking and currency authorities is to make sure that they have _n_ and _r_ thoroughly under control. For example, so long as inflationary taxation is in question _n_ will be influenced by other than currency objects and cannot, therefore, be fully under control; moreover, at the other extreme, under a gold standard _n_ is not always under control, because it depends on the unregulated forces which determine the demand and supply of gold throughout the world. Again, without a central banking system _r_ will not be under proper control because it will be determined by the unco-ordinated decisions of numerous different banks. At the present time in Great Britain _r_ is very completely controlled, and _n_ also, so long as we refrain from inflationary finance on the one hand and from a return to an unregulated gold standard on the other.[25] The second duty of the authorities is therefore worth discussing, namely, the _use_ of their control over _n_ and _r_ to counterbalance changes in _k_ and _k´_. Even if _k_ and _k´_ were entirely outside the influence of deliberate policy, which is not in fact the case, nevertheless _p_ could be kept reasonably steady by suitable modifications of the values of _n_ and _r_. [25] In the case of the United States the same thing is more or less true, so long as the Federal Reserve Board is prepared to incur the expense of bottling up redundant gold.