ON VARIATIONS IN THE VALUE OF THE MONETARY STANDARD. 



205 



pared results might have been predicted by theory, and may be predicted 

 safely of adjacent cases. 



III. We come next to the index- number of Jevons : the Geometric 

 Mean of the price-variations appertaining to a number of groups. The 

 definition of these groups is not wholly iiTespective of their importance to 

 the consumer and producer. There is evinced more or less concern that 

 each article of equal importance should ' count for one ' in the composi- 

 tion of the index-number. But Jevons does not affect precision of weight. 

 Pepper, for instance, forms one of the constituent thirty-nine articles.' 



The analogue of this operation for our materials appears to be the 

 Simple Geometric Mean of the price-variations for each of the articles 

 specified in our scheme ; except, indeed, those to which a very small 

 weight, namely 1, has been assigned. Accordingly Petroleum, Indigo,. 

 Palm Oil, and Caoutchouc may, with propriety, be lumped into one group, for 

 which the mean price-variation is to be ascertained geometrically. For 

 the sake of comparison with Mr. Sauerbeck's result, Caoutchouc (not 



' In the ' Serious Fall,' republished in Currency and, Finance, p. 44. In the 

 •Variation of Prices ' (i&ifZ., p. 142) Jevons seems to' have employed the practice of 

 weighting rather more extensively. He says, ' Several qualities of one commodity 

 have been joined and averaged before being thrown as one unit into larger groups' 

 — in the case of certain articles which are not very clearly indicated. For the period 

 after 1844 the [unweighted] ' average prices, as calculated from the price-lists of the 

 JEconomist . . . were mostly used.' 



