THE THEORY OF EXCHANGE 143 
At the present time when exchange crises are so numer- 
ous, we are far from being able to observe that the curves 
of exchanges and of circulation are consistently parallel. 
An exchange will move considerably upwards and down- 
wards while the circulation is constant, and conversely. 
Lastly, it should be observed that the fact of currency and 
exchange movements being parallel is not by itself con- 
clusive. In Germany, particularly since 1920, the increase 
in the volume of currency has lagged behind the fall in the 
exchange instead of preceding it and really amounts to a 
modified contraction in view of the level of prices due to 
the exchange position. 
Moreover, it would not be surprising, even if variations 
in the circulation affected the exchange, that this influence 
should not show itself by simultaneous variations in the 
curves of the exchange and circulation, as the latter in any 
case depends upon other factors such as the balance of 
payments, which may in its turn be affected by the cir- 
culation (but may also fluctuate for many other reasons) 
and by all the psychological imponderabilia which in- 
fluence a market exposed to speculation. Therefore in the 
absence of any means of directly proving and measuring 
the relation between exchange and circulation, we must 
confine ourselves to a careful appreciation of the argu- 
ments advanced in favour of its possibility. 
We come back, therefore, to the arguments enumerated 
above when, in seeking the initial cause of an exchange 
crisis, we discussed whether we were right in seeking 
“Le papier-monnaie,” by Subercaseaus, contains (Ch. ITI) various 
instances drawn from the 19th and early 20th centuries which completely 
invalidate the belief in a constant relation between the exchange rates and 
the volume of the circulation and incidentally show that in the case of 
certain countries such as Chili from 1905 to 1908 the coincidence of the 
two phenomena can be explained by the influence of factors which reacted 
both on the note issue and on the balance of payments. Finally, it should 
be observed that even in the time of Ricardo, who did so much to popularise 
this simple theory, the strict relation of cause and effect between inflation 
and the rate of exchange was denied by some technical writers (see the 
“collection des grands économistes,” Vol. XIII, “Le haut prix des lingots” 
and the annexed documents). See also, in spite of his drawing the opposite 
conclusion, Mr. Cannan, “The Paper Pound of 1797 to 1821.”