142 BEPORT— 1889. 



It is submitted that this course commits us to some such hypothesis 

 as the following. If, in order to compare the volume of trade for a series 

 of years, we assign a weight to the price of each article proportioned to 

 the importance of that article, we must regard the relative importance of 

 each article as constant for that series of years. If, then, the relative 

 importance of each article is to be measured by the pecuniary value of 

 the quantity bought or sold, the proportions which the value of each 

 article bears to the value of any other article, or to the total value of all 

 the articles, ought to be pretty constant during the whole series of years. 

 This assumption is strikingly verified by Mr. Giifen's Table II. Again, if 

 the proportionate amount expended on each article is pretty constant from 

 year to year, we may conceive a purchasing public (whether the com- 

 munity in whose interest the computation is being made, or the foreigners 

 with whom they deal) constant as to the nature of their wants [though 

 it may be increasing in numbers in the course of years]. Accordingly 

 the rates of exchange between the different articles ought to be constant. 

 In other words, the ratio between the prices of the different articles 

 ought to be constant during the series of years. This assumption is 

 verified as well as could be expected by Mr. Giffen's Table I. and Table 

 III., A.i 



The values and prices being constant, it is implied that the propor- 

 tionate quantities also, the number of tons, or it may be gallons, of each 

 article exported or imported, have a degree of constancy. This propo- 

 sition also may be verified by glancing at the quantity columns in Mr. 

 Giffen's Table IV., or the same figures in the statistical abstract. 



These assumptions as to the steadiness of the course of foreign trade 

 being admitted, a definite interpretation may be assigned to the otherwise 

 vague idea of increase in the volume of exports and imports. Or rather 

 two or three definitions become possible. The primary significance of an 

 increase in the volume of foreign trade is as a measure of the benefit 

 which the community desires from foreign trade.^ This conception is 

 particularly germane to the case where the articles on which the compu- 

 tation is based are commodities imported for the consumption of the 

 community. 'In some countries,' writes Mr. Giffen, 'the whole imports 

 less the re-exports may be treated as imports for final consumption.' If 

 the imports are materials as distinguished from finished products, still the 



• There are reasons why Mr. Giffen's table of price variations (Table III., A) should 

 present the appearance of stability in a less degree than his table of proportionate 

 values (Table II.) First, each entry in the former table is obtained by comparing 

 one item with another item, viz., the price of an article in any year with the price 

 of the same article in 18G1 ; whereas each entry in the latter table is obtained by 

 comparing an item (the value of an article) with an aggregate (the total value), 

 which of course is apt to be more stable than an item. If the suggestion made 

 below of referring each price to the mean price of the article for adjacent years were 

 adopted this contrast would doubtless be diminished. 



- Tlie variation in the volume of trade as thus conceived is very similar to Coumot's 

 definition of ' real gain,' or loss, of social revenue {Recherches sur Ics Principes 

 Mathematupies de la Thiome des Rlchesses, ch. x. ; and later redactions). But Cournot, 

 who seems not to have seized the idea of ' final utility,' strains the monetary measur- 

 ing rod beyond its legitimate application when he propounds his paradox that freeing 

 a commodity from a prohibition results in a loss of real gain to the country which 

 becomes an importer thereof {Ibid. Ait. 89). For this case implies a change in the 

 qvalitg of trade, a diversion of the streams of commodity into new channels with 

 which our methods are unable to deal, through failure of the hypothesis enunciated 

 in the text. 



