868 REPORT—1885. 
from increased sales of 2,733,0007. was lessened by 9,604,0001. fron» 
lower prices, the actual difference adverse to last year being 6,871,8981. 
Of the present year, only two-thirds having elapsed, it is impossible: 
to speak definitely, but assuming that the next four months will follow 
the same course as those which are gone, it is probable that there will be: 
a further decay of 21,000,000/. on the total of last year. Of this a 
rough calculation points to one-third as the probable loss from quantity;. 
and two-thirds the failure in prices. 
Although the purport of this paper has been to explain the method of 
investigation rather than the results to be obtained by its use, it would 
scarcely be right to leave wholly untouched some of the evidence it elicits; 
or the problems whose existence may thus be manifested, though not 
solved. Had time permitted, illustrations might have been drawn from 
the import records which admit of exactly the same treatment that has 
been bestowed upon the exports. Before this, however, it may be well 
to allude again to the differences of the two methods as accounting for: 
the figures evolved by the one process differing from those produced by the- 
other. 
It has been previously noticed that the selection by Mr. Giffen of 
1861 as a basis narrows the list of articles which can be specifically 
calculated by so many as were not enumerated in that year, and so throws. 
the calculation on the whole year to be made upon a smaller proportion. 
In 1883, 61° instead of 70-4, leaving 39: to be averaged instead of 29°6. 
But it does more than this, for many of the classes left out in the later 
year are really included in the earlier. For instance, the total value of’ 
iron exported in 1861 was 10,000,0001., and in 1883, 29,000,0001., but 
only 6:4 represented the latter, whilst 5°6 stood for the former in building 
up the index numbers, owing to the exclusion of descriptions separately 
shown in 1883, but included under more general heads in 1861. This. 
applies still more to earlier years. 
Then the choosing for the index number a percentage of the actual: 
value of its own year, instead of some fixed datum, rendered the figures 
of unequal values in the different years. Thus 61, the index for 1883,. 
meant a of 240,000,000 = 146,000,0001., whilst 71:1, for 1861, meant. 
ee of 125,000,000 = 89,000,000/., each unit of the index being of 
double value in one year to what it bore in the other. 
Again, taking the proportions of 1875 as those to which the increased 
prices should be reckoned as affecting the index of all years alike, could 
only be correct for such years as happened to have the same relative 
quantities as in that particular year, and must be particularly injurious. 
where a very high or low price happened to be coincident with a greatly 
varying quantity. This too would be aggravated or minimised by the 
proportions being those of value rather than of quantity. Thus the 
proportions of cotton yarn for 1865, 1875, 1883 stood as 104 : 216 : 265,. 
but by value as 10 : 13 : 14, and tke percentages of increase or decrease 
from the standard of 1861 were as + 91:23 : + 41°63 : — 2:31. Tt is. 
difficult to see how any combination of these factors, so widely differing in 
their ratios, can bring about the result that the index numbers for cotton 
yarn should be altered as +5°38 : +1:00 : — 0:'14.as shown in the Board 
of Trade tables. 
The effect of these three several arbitrary departures, as they seem to. 
