possible. A long strip of land is likely to be more rep- 

 resentative in character than a square piece, as it will 

 contain more of the inequalities of the soil, and for this 

 reason is to be preferred for the purpose of these experi- 

 ments. The land should be as level as possible, and if 

 not level, the plots should be so located that the fertilizers 

 cannot be carried by rain from one plot to another. No 

 part of the strip should be shaded by trees. A convenient 

 size would be 33 feet wide by 416 feet long divided into 

 6 equal plots each 66 feet long, with a path 4 feet wide 

 between the plots. Each plot, therefore, would be 33x66 

 feet and measuring exactly one-twentieth of an acre. 

 Each plot should be separated from its neighbor plots, 

 as well as from adjacent cultivated ground, by a 4 ft. 

 path, so that the roots of the plants grown on it can get 

 no fertilizer that is not intended for them. Of course 

 these paths or borders should be kept reasonably free 

 from grass and weeds, which would otherwise feed upon 

 the fertilizers intended for the plants in the plots. 

 Having divided up the plots as indicated and marked 

 them by numbers from 1 to 6 inclusive, prepare the soil 

 thoroughly in the usual way, after applying the fer- 

 tilizers broadcast as follows : 



Plot 1. No fertilizer. 



Plot 2. Nitrate of soda 20 Iba. 



Acid phosphate 60 " 



Plot 3. Nitrate of soda 20 ** 



Muriate of potash 16 



Plot 4. Acid phosphate 60 



Muriate of potash 16 



Plot 5. Nitrate of soda 20 '' 



Acid phosphate 60 '* 



(87) 



(4 

 ( ( 

 ( ( 



