364 



REPORT OF THE COMMISSTONT^R OF AGRICULTURE. 



making a bad sliowing and tending to a constantly declining pro- 

 duct. The State returns are from this or similar cause of declension 

 usually 10 per cent, too low. This is so in other countries and by vari- 

 ous methods. For instance, the Agricultural Gazette of London for 

 August 16, 1886, makes the following report of harvest estimates, the 

 figures representing the number of returns: 



Reducing these returns to their percentage values (on the basis 

 of numbers alone, without reference to area represented by each re- 

 porter), we have for wheat 5.2 per cent, of them over average, 24.5 

 per cent, average, and 70.3 under average; a very bad showing, seven- 

 tenths being under average, and the remainder scarcely more than 

 average. 



This for a single year. The returns of a series of years, if aver- 

 aged, should give a statement in which the percentage of returns 

 "above" and "below" should be equal. The same paper gives the 

 data for such a test, eight consecutive annual returns, of wheat as 

 follows: 



Averaging these eight years, only about one-sixth of these returns 

 are above an "average," while nearly one-half of all are below aver- 

 age. Conceding eight years to be a period long enough to construct 

 an average, this result upsets the accuracy of the returns and estab- 

 lishes the tendency to underestimate. 



The percentages of all other crops show the same tendency except 

 those relating to barley. The averages for the period are as follows: 



Here the percentage "under average " for the period is nearly doublo 

 that of the ' ' over average " returns, when they should be equal. If 



