PRECAUTIONS NECESSARY IN MAKING EXPERIMENTS 37 



have a stake at each corner of the different plots, which 

 should be marked with the number of the plot. 



5. It is best to have the experimental plots long and 

 narrow, because thus they will average up for unevenness 

 of soil. 



6. It is best to separate plots by paths, in order to 

 prevent roots of plants of one plot from feeding on the 

 fertilizer supplied to the adjoining plots. 



7. Avoid windy days in spreading the fertilizers, so 

 that they may not be blown and scattered unevenly over 

 the plots. 



8. All the plots must be treated alike in every respect, 

 except as to the amount and kind of fertilizer applied. The 

 same kind and quality of seed must be used over the whole 

 area. The planting or sowing on all the plots must be done 

 the same day. (If a part be planted before and a part after a 

 rain, the experiment may become valueless.) Use every pre- 

 caution necessary to secure a full stand of plants, and if a 

 uniform stand has not been secured at the first planting, 

 plow up the whole field and plant over again. Arrange the 

 same number of rows on each plot, and the same number 

 of hills and plants (as nearly as possible) in each row. The 

 plots should be plowed and cultivated alike, and whatever 

 operation is needed in the experimental field, should be 

 carried out uniformly all over the plots. 



9. The harvesting of the crop and weighing of yields 

 must be accurate. Experiments are usually made on plots 

 of one-tenth or one-twentieth of an acre, and a mistake will 



