370 



Journal of Agricultural Research voi. xviii. No. ? 



The average absolute dry-weight yields of wheat roots for each of the 

 three different degrees of moisture employed in the sand cultures are 

 shown graphically in figure 4, in the same manner as were the correspond- 

 ing yields of tops in the graphs of figure 2. 



The dry-weight yields of the series employing a medium moisture con- 

 tent (series B) were arranged in the order of their values, beginning 

 with the highest. These are represented in the upper graph of figure 4. 

 The yield values of series A and series C were plotted in the same order, 



45562 34 52346214 38 1712321117613243521 

 542366435532742316 133425722182411111. 



Fig. 4. — Average absolute yields of wheat roots for low, medium, and high moisture content of sand 



cultures. 



and these are represented by the graphs indicated by the continuous and 

 the broken line, respectively. As in the graphs representing the yields of 

 tops, there is a decided tendency for each of these graphs to slope down- 

 ward to the right, thus indicating the difiference in the yield values brought 

 about mainly by the variations in the salt proportions. But this tendency 

 is much more marked in series B than it is in either of the other two series. 

 The medium moisture content (series B) shows the highest absolute dry 

 weights of roots throughout the entire series. It is to be noted also that 

 the highest yields in the two series with the extremes in the moisture con- 

 tent employed are nearly equal in value and very much lower than the 

 highest yield produced with medium moisture content. 



The differences between the graphs representing the root yields from 

 the two series of cultures with the lowest and the highest moisture con- 

 tent are not pronounced; these two series resemble each other with re- 



