ON VAEIATIONS IN THE VALUE OF THE MONETARY STANDAED. 207 



appears that 72 is the figure eleventh in the order of magnittide : that is 

 the Median. 



This is the Simple or Unweighted Median. There is a variety consti- 

 tuted by assigning special importance to those returns which we have 

 reason to suppose are specially good representatives of the changes 

 aflPecting the value of money. If, as in the writer's Memorandum often 

 referred to,' we take mass of commodity as the principle of ponderation, we 

 shall have to proceed as follows with our twenty-one articles : 



As before, make three compartments for returns below 70, for those 

 Ijetween 70 and 80, and for those above 80 respectively. Write down in 

 the first and third compartments the returns in the order in which they 

 occur (in any order) ; but in the central compartment in the order of 

 magnitude.^ In the second column of each compartment write the figures 

 representing the relative precision assigned to each return. If these esti- 

 mates of precision are based upon the quantities of commodities, it is 

 recommended that they should be equal to, or rather less than, the square 

 roots of the proportionate masses. Accordingly 2 has been put for the 

 square root of 6, 1"5 for the square root of 2^, and so on. Add together 



' Sect. ix. of 'Memorandum' in Heport of Brit. Assoc, 1887. 

 ^ It will probably be convenient to write these returns first in the order of their 

 occurrence, and then rearrange them. 



