A 
MONEY 
fall need not be more than about 10 inches, and 
Jones finds it expedient to abandon the argument 
that nothing has been removed, and to find some 
other defence for his parent. 
Commodities and services are SO numerous in 
kind and the kinds shade into each other so gradually, 
that to take into account the price of all of them is 
much like taking into account the level of every part 
of a rough field, when smoothing it is not to be 
thought of. We cannot do it literally, and must be 
content with taking a sufficient number of measure- 
ments at points selected without bias. The ordinary 
person’s impression about a general change of prices 
fs much like Smith’s measurement of the level of his 
field “ by the eye”; itis likely that he will be able 
to recognize a large change of prices—probably 
anything over 25 per cent., just as Smith is 
likely to be able to detect a fall of 10 feet in the 
general level of his field. When the change is not 
great, he is just as likely as Jones to be misled by 
bias into denying its existence, and in all cases bias 
is likely to mislead him, as it led Smith, into very 
faulty estimates. To arrive at agreement it is neces- 
ary, as in the case of the disturbed field, to introduce 
statistical methods, and this is done by the construc- 
tion of what are called “ index numbers *’ of prices. 
The prices of a large number of commodities at some 
particular date, called for this purpose the “ base 
year ”’ or the * standard year,” are collected, and the 
prices of the same commodities at subsequent (or 
earlier) dates are represented as percentages of the 
prices of the base year. If beef cost 10d. per lb. in 
the base year and 13d. at some later date, it is put 
down at roo for the first and 130 for the second 
period, since if it takes 134. to buy what formerly 
could be got for 1od., it takes 130d. to buy what 
could formerly be got for 100. The prices of a