156 REPORT — 1889. 



expected, or indeed in any considerable degree. The tables referred to 

 should be compared with the Appendix at p. 159, as strikingly illustrating 

 how different principles of averaging bring out the same mean result ; in 

 short, that in our sort of work it is not very easy to go wrong. 



Among imports, grain and timber are suspicious. But with regard to 

 timber Sir Rawson Rawson shows that, though the ' weight ' (determined 

 by its bulk) is large, yet it is not materially different from what it ought to 

 be as determined by value. However, he is no doubt judicious in excluding 

 such-like items from his final Index-number. 



With regard to the inequality of values, as this has not proved fatal 

 to Mr. Giffen's and the cognate methods, there is ci fortiori less reason to 

 apprehend it in the case of an argument which is based on a greater 

 number of independent items. However, it might be well to examine 

 specially the influence of cotton. 



There remains to be considered the remainder, which is made up of 

 differences between the density of packing in different years. It is natural 

 to suppose that these should compensate each other except so far as in the 

 course of years a general tendency to increased economy of room makes 

 itself felt. Sir Rawson Rawson sets off against this tendency the increase 

 of passenger traffic ; a quantity which he has abundantly shown to be of 

 an order which may be neglected. For short periods, at any rate, the new 

 method appears to constitute an important adjunct to, if not a complete 

 substitute for, the received methods. 



Sir Rawson Rawson's method may be regarded in another aspect as 

 affording a measure of the level of prices in different years. If the 

 hypotheses made in the earlier part of this pajDer are conceded, no ad- 

 ditional remark is called for here. We have simply to write Index-number 

 for level of prices in year y as com pared with x -- average price in y 



. „„„ ^ • • Value in v Value in x [ Value in -?/ 



-f- average price m x ^ — ^ ^-J— _:_ , ( or =; 



Volume in y ' Volume in .c \ Value in x 



-7- -T^^n -■ — ~ 1 ; where the values are given figures and the volumes are 



Volume va. x) o o 



proportioned to the respective tonnages. We thus obtain a new and 



remarkably easy solution of cur problem. 



Section VI. 



Tlie present Writer's Method. 



We have so far been supposing that the importance attached to each 

 variation in price is, or ought to be, proportioned to the value of the 

 corresponding article. But we have now to entertain a different sup- 

 position and distinct method. We are now to imagine a general change 

 coming over the monetary world — or some zone of it like wholesale 

 prices — like a general variation in temperature or atmospheric pressure 

 over a physical region which is not perfectly level and uniform in its 

 conditions. In reading a barometer or thermometer in any particular 

 place with a view of ascertaining the fact and amount of a general 

 change it would not be appropriate to attach importance to the mere 

 size of the tube and quantity of the rising or falling liquid. In fact the 

 smaller thermometer has so far the preference, as it takes on more 

 quickly changes of temperature in the surrounding medium. Sensitive- 

 ness, not size, is the criterion of these indicators. So also, in virtue of 



