296 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 1935 



Table 4. — The toxicity of selenium for reheat seedlings in relation to the sul- 

 phur (as maffncsium and ammonium sulphate) in the nutrient solution 

 {injury indicated hi/ plus signs) 



The protective action of sulphur in the several such experiments 

 may perhaps be most strikingly summarized by saying that the 

 quantity of selenium necessary to kill the plants varied from about 

 one part in a million, where no sulphate was added, to nearly 100 

 times as much where there was the most sulphate — i. e., 192 parts 

 of sulphur per million. Moreover, there was a definite relation be- 

 tween the quantity of sulphate sulphur in the nutrient solution and 

 the quantity of selenium required to produce a given effect on the 

 plant (table 4). Thus the presence of 12 times as much sulphate 

 sulphur as selenium always prevented injury regardless of the abso- 

 lute amount of selenium supplied. Where there was only about 10 

 times as much sulphate sulphur as selenium, slight stunting and 

 traces of the white chlorosis appeared on the leaves. "VVliere there 

 was but twice as much sulphur as selenium, the plants eventually 

 died/ 



In these experiments (table 4 and pi. 4) the sulphur (as sulphate) 

 of a given series was the same in each flask, the selenium content 

 being varied. They were repeated with the selenium constant in 



■^ These ratios cannot be used to predict toxicity in soils because the soil analyses on 

 which they would of necessity be based do not show what concentrations are actually 

 available to the plant. 



