250 REPORT— 1887. 



amount of money wHch at the posterior epoch is equal to a unit of money 

 at the prior epoch is of the following form : [^ (quantity of commodity A 

 consvimed per unit of time at prior epoch + quantity of commodity 

 A consumed per unit of time at posterior epoch) x average price of 

 commodity A at posterior epoch + ^ (quantity of B consumed per unit 

 of time at prior epoch + quantity of B consumed per unit of time 

 posterior epoch) X average price of B at posterior epoch + &c.] 

 divided by [-5 (quantity of commodity A consumed per unit of time 

 at posterior epoch + quantity of commodity A consumed per unit of 

 time at prior epoch) X average price of A at prior epoch + ^ (quantity 

 of B consumed per unit of time at prior epoch + quantity of B consumed 

 per unit of time at posterior epoch) x average price of B at prior epoch 

 + &c.] 



The difficulty that new kinds and qualities of articles are continually 

 entering into consumption is to be met in the manner suggested by 

 Professor Marshall.' Frequent revisions of the 'standard' are to be 

 made, say once a year, the purchasing power of money in each year being 

 continually compared with what it was in the preceding year, after the 

 manner above indicated. As soon as any new species of ware has made its 

 appearance in two successive years, as soon as it figures both in a ' prior ' 

 and a ' posterior epoch,' the consumed quantity thereof (presumably not 

 enormous within a year after the introduction of the new article) is to be 

 entered as one of the items on which our calculation is based. Perhajis, 

 however, there will not be much need of this refined adjustment for the 

 rough standard which, by way of a first essay, we propose to construct 

 for Great Britain. 



In choosing the commodities proper to this purpose, we may take as 

 our guide the ' Report on the Appropriation of Wages,' drawn up 

 by Professor Levi for this Association in the year 1881. From the 

 articles of national consumption specified in that Report, there should 

 be selected those which have a certain degree of importance in respect 

 both of bulk and also of the precision with which the returns of quantity 

 and of price are in each case ascertainable in the existing state of 

 statistics. The following list is provisionally offered : bread, potatoes, 

 butcher's meat, bacon, ham and pork, fish, cheese and butter, 7nilk, fruit, 

 sugar, tea and. coffee, beer, spirits, lolnes, tobacco, boots and shoes, cotton goods, 

 woollen goods, coal for domestic purposes. 



We reserve for a futui'e Report the task of discussing each of these 

 items on its merits ; and of determining what finished products may be 

 represented by means of articles which enter into their production, such 

 as coal and iron. It must suffice for the present to lay down the general 

 principle that, in estimating the precision of the ' average price,' regard is 

 to be had not only to the accuracy of the particular price-returns (of the 

 same commodity at different places or times) which are combined into 

 the average, but also to the applicability and worth of the mean so 

 formed. 



It may happen, as Professor Marshall has lately pointed out, that the 

 simple average (or Arithmetical Mean) of the particular price-returns may 

 be extremely unsnited to the purpose in hand. Thus, in the case which he 

 puts of strawberries, the price both in May and July might be Qd., but in 

 June od. per lb. To take hd. as the mean price might be very mislead- 

 ing. An undue weight is given to the particular price-i'eturns for May 

 ' Covtemporary Review, March 1887, 



