48 TOXIC LIMITS OF ALKALI 



it is seen that the variations are wide. In another experi- 

 ment with lupine, where growth was prevented by the 

 salts contained in the solutions, the magnesium salts were 

 not so toxic as the carbonates of sodium, and the mag- 

 nesium sulphate was the least toxic of all salts. This 

 shows that very wide differences might be expected ac- 

 cording to the method employed. 



True (30), using the above method for obtaining the 

 toxic limit of lupine in sodium chloride solutions found 

 it to be 3625 parts per million, which again shows the 

 possible error. Coupin (5) allowed the plants to remain 

 in the solutions until the whole plant showed the salts 

 to be causing injury. His limits for lupine using sodium 

 chloride, magnesium chloride, and magnesium sulphate 

 were 12,000, 8000, and 10,000 parts per million for the 

 respective solutions, which is about the same as the 

 above results where growth was prevented. 



The resistance here is several times that found by Harter 

 where the first injury was the point of indication rather 

 than the death of the plant. Allowing the roots to remain 

 in the salt solution twenty-one days and then weighing, 

 the author (10) found wheat seedlings to produce about 

 one-half as much as the check in the solutions containing 

 5000 parts per million of sodium carbonate, or in those 

 containing over 10,000 parts per million of sodium chloride 

 or sodium sulphate. Haselhoff (9) concluded that growth 

 might be inhibited with a 5ooo-parts-per-million solution 

 of sodium chloride and injury would result in the presence 

 of 500 parts per million. 



Hansteen (8) states that 5000 parts per million of salts 

 other than calcium are injurious when used singly, but 

 when combined with lime the injury is greatly diminished. 

 Others have found the same antagonistic effects of dif- 



