210 Journal of the Department of Agriculture. — Sept., 1922. 



At the end of May a uniform wetting of 4 inches was given to 

 all the plots, the water being measured by a right-angled V-notch in 

 the main furrow. Four beds were watered at a time — two on each side 

 of a lateral, the time taken to apply 4 inches of water varying with 

 the height of water over the notch, about 11 to 12 minutes being 

 required with an 8-incli flow over the notch. After irrigation the 

 beds were loosened by means of garden forks and seeded to " bird- 

 proof " wheat at the rate of 35 lb. per acre. 



Plate I. — Main Furrow, looking South, and. 

 showing V-notch in Centre. 



The reason for treating all the beds alike is to determine the 

 variability in natural productivity of the soil (other conditions being 

 the same), for it is obviously impossible to carry on experiments in 

 connection with the effect of different quantities of water on a crop 

 if the variability of the soil is an unknown quantity. This error can 

 be overcome to some extent by having a large number of replications. 

 It is better, however, to conduct uniformity tests first, and then by 

 plotting frequency curves to determine how reliable will be the results 

 of experiments on the particular plot of ground. 



