Table Showing Rainfall and Depth of Water in inches used by Crops 



during a Dry Season. 



That is, these crops, during their period of growth, used approxi- 

 mately two and one-quarter times as much water as fell in rain. Since 

 the plants did not grow as large or strong as those in actual field condi- 

 tions, we are safe in assuming that field crops used as much as or even 

 more than those in the crqcks. Last season was about an average one. 

 Thus we see that under ordinary conditions, if the crops are to be supplied 

 with all the water they need, there must be a great store of it in the soil 

 from which they may draw. Hence, in anticipation of an average or dry 

 season, our treatment of the soil must put it in such a condition that it 

 will retain a great deal of the spring, winter and autumn precipitation. The 

 crocks in 1905 were set on the roof of the annex to our building during 

 the early part of the season ; but it was thought that possibly the loss 

 there was very much in excess of what it would be at the ground. So 

 about the middle of the season half the crocks were removed to the gar- 

 den, part being set on the ground and part in the ground about level 

 with it. Between the losses from the former and the latter, the scales 

 showed no difference, though the loss from those on the roof was slightly 

 greater than from those in the garden. But the first tests in any experi- 

 ment are seldom made in just the same way as subsequent ones. This 

 season (1906) all crocks were set in the ground in a field of barley, a 

 path leading into the grain, and the crocks being set back in it on either 

 side of the path. 



The crocks had a capacity of four gallons, were 10 inches in dia- 

 meter, and about 12 or 13 inches deep, and caught all the rain that fell. 

 The day they were set outside a very heavy rain fell, and having weighed 

 them just before the rain, we weighed them again just after, and com- 

 pared the result with our rain-gauge. It was found that the crocks had 

 absorbed the whole shower. None of the rain was lost by drainage. We 

 had a drainage tube in the bottom of each crock, but not once during the 

 whole season was the rain sufficient to saturate the soil and cause perco- 

 lation. The quantity of rain required to saturate the soil depends on the 

 amount of moisture in the soil when the rain comes. We tested that 

 point once during the season. We let the soil dry out until the grain 

 began to wilt. The amount of water in the soil at wilting point varies 

 in different soils. This was a loam, and by actual test was found to 

 contain 7.3 per cent, water when the plants wilted. Water was added to 



