Manchester Memoirs, Vol. xli. (1897), No. 12. 5 



which is but small, are massed together and entered by 

 value only. In all such cases of entry by value alone, we 

 cannot pretend to measure the price-variation from the 

 information supplied. The method adopted is to deter- 

 mine the average price-variation for the goods which 

 permit of it, and to assume an equal price-variation for 

 the remainder. It may be said that the assumption is 

 one which might be expected, a priori, to be false. It is 

 at any rate capable of being contended that it is equally 

 likely to be correct, and there is, moreover, no means 

 available for determining what proportion of price- 

 variation should be applied if the average actually proved 

 to hold for the tested part of the trade be not used. 



When the total cost of one year's imports or exports 

 has been determined at the preceding year's prices, a 

 comparison with the actual cost of the preceding year's 

 trade enables us at once to obtain a figure showing the 

 variation in volume (as the phrase is used) of the trade 

 between the one year and the preceding. 



The process just described is nothing more nor less 

 than the construction of a very elaborately weighted index 

 number for the single year's variation. When we proceed 

 to the next year, the weights associated with each price- 

 variation will be changed, and thus the process differs 

 from that of the construction of an ordinary index 

 number with fixed weightings. It is obvious that we 

 may compound the successive yearly variations and 

 obtain an estimate of the total variation both in price 

 and in quantity between any two dates for which these 

 calculations exist. There is the advantage that the 

 accidental price-proportions of a reference-period are not 

 perpetuated in their influence on the measurement of all 

 subsequent movements, as in ordinary index numbers. 

 The impossibility of evaluating the whole of the trade 

 imparts some uncertainty to the result. It is, however, 



