72 FERTILISERS AS AN AID TO 



other factors which have to be provided for in 

 actual practice. This is confirmed 



^ Mr ' A ' D ' Hall > Wh is Careful 

 to point out that the " experi- 

 mental error " in ordinary farming in this 

 country must be considered as ten per cent., 

 which means that unless there is a greater 

 difference between results than this, they are 

 not to be accepted as conclusive evidence. 

 This is particularly applicable to experiments 

 which are carried on only for one year. When 

 they extend over a series of years at one or 

 more centres, the " experimental error " becomes 

 so much reduced as to be practically negligible. 

 Therefore in comparing results of 

 var i us nianurings this factor must 

 on no account be lost sight of. 

 In the experiments quoted in these pages we 

 have been careful only to exhibit those which 

 fulfil these conditions, and in every case they 

 will be found to be well over the required ten 

 per cent, margin. 



It is of course impossible to lay down absolute 

 rules for each individual case, as to the most 

 economical and suitable methods to pursue on 

 all varieties of soils, and under varying condi- 

 tions of rainfall, climate and exposure. A 

 most important factor, as all practical farmers 

 know, is the manner in which the previous 

 crop or crops have been manured, but the 



