122 Journal of the Asiatic Society of Bengal. [March, 1910. 



same chart shows at a glance the effect of this weighting. The 

 manner in which the general index number line can by weight- 

 ing be deflected from the low import index number line to the 

 high food-grains index number line is only a graphic illustration 

 of the difficulty that occurs in all discussions on prices to de- 

 cide the degree of importance that should be attached to the 

 food-grains. It also serves to indicate that extreme caution is 

 needed in basing a theory on any particular index numbers, 

 unless the system on which they have been prepared is clearly 

 understood. The Statistical Department's index numbers are 

 compiled on the exceedingly simple principle of calling the price 

 of each commodity for the year 1873-100, and reducing all 

 prices to a percentage of the price of that year and then divid- 

 ing the sum of the converted prices for any particular year by 

 the number of the commodities. 



No doubt the ideal index number would be on a consump- 

 tion basis, but the matter is not as simple as it at first sounds. 

 We should first have to divide society up into typical classes, a 

 task as difficult as forming electoral colleges : we should then 

 have to frame for the average individual of each class a budget 

 estimate of expenditure, a task only one degree simpler than 

 the framing of the Imperial Budget Estimate. We should then 

 arrive at a number of index numbers which might not help us 

 very much on account of their mutual incompatibility. The 

 fact is that the attempt to frame index numbers on a consump- 

 tion basis seems fated to end in dissipated energy and dis- 

 appointment. It is like piling Pelion on Ossa and Ossa on 



Pelion. 



One fact, however, is obvious, — that so far as three quar- 

 ters of the population of India is concerned, the importance of 

 the food-grains is paramount. There is, in other words, a very 

 strong case for weighting the food-grains, as Mr. Atkinson has, 

 in fact, done. If, however, we weight the food- grains to any- 

 thing like an adequate extent, we shall obtain an index line 

 which differs little from the index line for food- grains. 



But it must not be inferred that the latter line, as drawn on 

 the principles adopted by the Statistical Department, is in- 

 tended to represent in any sense the cost of living, for this is 

 very far from the fact. If anyone were found so impatient of 

 theoretical exactitude as to wish to frame his daily bill of diet 

 on the proportions of the staples that enter into the food-grains 

 index number, it would be necessary to add to that number a 

 substantial item for medical fees. All that we can really claim 

 for the figure is that it affords us a rough-and-ready means of 

 comparison of the prices of food-grains in different years. 



The food-grains included in this index number are wheat 

 and rice, the two most important, and then jawar, bajra, gram, 

 barley, and ragi. In this special index number the prices for 

 the seven food-grains are retail prices, whereas, in the special 



