drought and from waterlogging. The weight of grain and 

 straw of the cereal and also of the leguminous crop should be 

 noted. If the NaNo ;3 plot shows special increase in the case 

 of the cereal but not so in the case of the pulse crop the soil 

 should be considered rather poor in nitrogen, especially if the 

 yield of the cereal in the unmanured plots is found to be in- 

 variably less than that of the plot manured wirh NaNo ;j . 



If the Sodium phosphate does not show any benefit, the 

 soil should be considered rich in phosphates. 



If the plot manured with CaO shows better yield espe- 

 cially in the case of the pulse crop as compared to yield from 

 unmanured plots, the land should be considered deficient in 

 lime. 



If the potassium -chloride plot shows no benefit in either 

 case the land should be considered rich in potash. 



If the cereal is benefited chiefly in the weight of the straw 

 by the NaNo 3 the soil should be considered poor in N. If 

 the Sodium phosphate plot shows better yield of grain in the 

 case of the cereal, the soil should be considered wanting in 

 phosphoric acid. 



65. Na and Cl should be chosen as these elements are not 

 so essential to plant life and do not affect the yield. The object 

 of having buffer plots without manure, is to keep the effects 

 of the different manures quite distinct. The plots should be 

 protected from heavy rain or the experiment conducted in 

 the Rabi season, lest there should be overflowing from one 

 plot to the other. Watering should be done gently and care 

 should be taken that there is no mixing up of soils, weeds, 

 and crops of different plots. This ten-plot experiment is 

 recommended for practical purposes in judging of the che- 

 mical value of soils. It will not give absolutely correct ideas 

 as to the potentiality of a soil, but it will give a very fair 

 idea of the available plant foods. If the plots are quite 

 detached one unmanured plot will be found sufficient and in 

 that case it will be a five-plot experiment. As experiments 



