152 MODERN MONETARY SYSTEMS 
statistical data and in view of the uncertain intervals in 
which the supposed reactions show themselves, it does not 
seem possible to draw scientific conclusions from a 
detailed comparison of the curves. But certain results, 
such as the recovery in the foreign trade of France after 
1920 following a serious fall in the exchange, are in 
favour of the reasonable assumption that the exchange 
has a regulating effect on the balance of payments, and 
Vice versa. 
And yet this tendency towards equilibrium cannot be 
considered demonstrable in all circumstances. In the first 
place the idea that the foreign trade of a country varies 
solely with prices is crude. The most prosperous country 
in the world can only give what it has got, and a premium 
on the exchange will not allow of an indefinite development of 
exports if the means of production are inadequate. On the 
other hand, iz may happen—and it certainly has happened— 
that the exporter who derives profit from the premium on 
exchange, owing to the difference between the external 
depreciation and the minimum internal depreciation of the 
currency, will abandon a part of that profit to his foreign 
purchaser. Tt is true that nevertheless exports will be 
thereby stimulated; but goods will be sold at a lower 
price in foreign currency—say in gold—than that at which 
it would have been sold with a normal exchange, and the 
value of exports will not increase in proportion to their 
volume. But since the balance of payments depends on 
an increase not in the weight or tonnage but in the value 
of exports, it is easy to see that on this assumption, which 
is often in fact realised, the fall in the exchange will not be 
enough to restore the balance of payments. The auto- 
matic machinery contemplated above may very well not 
come into play. 
Lastly, the whole theory has been built up as though 
the rate of exchange were bound up solely with the balance 
of payments. But as has been pointed out, this statement, 
while it is approximately true when the exchange is 
restricted by the gold points, becomes inaccurate when 
we are dealing with abnormal exchanges. Of course, the