FIELD EXPERIMENTS IN AGRICULTURE. 501 



years can eliminate this source of uncertainty in the interpretation of 

 the results. 



As regards the practicability of the method of exact field experi- 

 ments as developed by Drechsler, it is worthy of remark that, while 

 results which conform to its criteria are trustworthy, a considerable pro- 

 portion of his own experiments have simply succeeded in demonstrating 

 that the soil was too unequal to admit of successful field trials. 



Wagner has attacked the problem in a different way His first at- 

 tempt was to make field experiments upon very small plots, only two 

 or three square metres in area, separated from each other by walls of 

 masonry, and to compensate for the small size of the plots by the care 

 with which they were treated. He also adopted the plan of repeating 

 each manuring several times, as described above. The results were 

 not satisfactory, however, owing largely to the unequal distribution of 

 water among the plots, and after numerous experiments he has aban- 

 doned this method and adopted that of pot experiments. His pots are 

 cylindrical zinc vessels, fifty centimetres (nineteen inches and a half) 

 high and twenty-five centimetres (nine inches and three quarters) in 

 diameter. These are uniformly filled with the carefully mixed soil, 

 and are provided with an arrangement by which the water of the soil 

 is automatically replaced as fast as it evaporates. The small size of 

 the pots permits the use of pure materials as fertilizers, while for the 

 same reason duplicate trials can easily be multiplied. The method 

 in Wagner's hands has proved a practical one, and has already yielded 

 some very interesting results. 



It may seem that such a method as this is too far removed from 

 the conditions of actual practice to afford results of any practical value. 

 There is a degree of truth in this criticism. The conditions in such an 

 experiment are different from those in the field. Wagner's method 

 has one inestimable advantage, however, viz., that all the conditions 

 of the experiment are under control. The importance of this is strik- 

 ingly shown by considering the ejects of a deficient supply of water, 

 such as is liable to occur in any field experiment. 



It is a well-known law of vegetable growth that that factor which 

 is present in the least quantity in comparison to the amount needed — 

 i. e., which is present in relatively the minimum quantity — is the one 

 which chiefly regulates the amount of production. If, in a field trial, 

 the supply of water holds this position, as it easily may, it and not the 

 diverse manuring will determine the amount of crop. Moreover, as 

 the plants grow larger and expose more leaf-surface, they exhale more 

 water, and it might very well happen that a supply of water which 

 was sufiicient for a plot lightly manured might not be enough to sup- 

 ply the exhalation from the more luxuriant plants on a better-manured 

 plot. The result would be, that the growth on the latter plot would 

 be hindered, and the manure would not have a chance to sho^i^ its full 

 Value. In pot experiments conducted according to Wagner's plan, 



