1882.] AGRICULTURAL EXPERIMENTS. 45 



20. Co-ordination, or the bringing- of influencing causes into har- 



monious action, and the effect produced by varying factors 

 of disturbance. 



2 1 . Interference, or the changes which are produced in the factors 



regulating plant growth and cropping through the acts of 

 man, as in influencing water-supply, checking evaporation, 

 root-pruning, leaf -pruning, harvesting methods, etc. 



THE EXPERIMENTER. 



Through a study of these complex conditions, agricultural 

 investigation should be able to supply to the experimenter seed of 

 a known potency of production, and uniformity of character. 

 Without a seed which under uniform conditions shall furnish a 

 uniform cropping, we are in the condition of a chemist whose 

 reagents are impure and never twice of the same exact composi- 

 tion. The seed obtained, a portion should be planted by itself on 

 a plot under maximum conditions of fertility, and under uniform 

 culture as determined by experience in order to afford a verifica- 

 tion of the quality of the seed, and the produce carefully studied 

 should serve as a check upon the results obtained by the second 

 portion used in the experiment proper. Upon carefully selected 

 plots, upon land considered uniform, and whose history is to some 

 extent known, the seed corn should be planted in a duplicate 

 series, together with the quality and quantity of manures whose 

 effects are to be studied in order to answer the farmer's question of 

 "how much and what kind of manure can I profitably apply ? " 

 One of each series should be unfilled, the other tilled uniformly 

 according to the best dictates of judgment. There should be sev- 

 eral plots left without manuring, and these should receive different 

 kinds of cultivation, one certainly no cultivation whatsoever, and 

 another as much intercultiiral tillage as seems desirable. In the 

 duplicate plots which receive manuring, one portion should be un- 

 tilled. Under this method, we may hope to receive tolerably correct 

 replies, if the comparisons of results be skillfully made so as to ehmi- 

 nate the sources of misinterpretation; for we have as factors to use: 



1. In the verification plot we obtain a measure of the uniform- 

 ity and potency of the seed under the known favorable conditions. 



2. In the no-manure plots we obtain a measure of the condition 

 of the land in its relations to the seed, under two conditions, the 

 one favorable and the other unfavorable to the crop. 



