IN LOAM SOIL 57 



nitrate, sodium carbonate, potassium carbonate, sodium 

 sulphate, potassium sulphate, and magnesium sulphate. 

 This order does not hold when the concentration is 

 determined by an analysis of the soil. The anions 

 were found to affect the toxicity more than the cation, 

 the chloride being the most toxic anion and sodium the 

 most toxic cation. 



Bancroft (i), in his work with beans growing in large 

 pots to which alkali was added from below after the plants 

 were growing until they wilted and died, found the fol- 

 lowing quantities of salts just killed the plants: magnesium 

 chloride, 2640 parts per million; sodium carbonate, 2710 

 parts per million; sodium nitrate, 3700 parts per million; 

 sodium chloride, 5660 parts per million; magnesium sul- 

 phate, 5820 parts per million; sodium sulphate, 6810 parts 

 per million; and sodium bicarbonate, 12,300 parts per 

 million. 



In germination tests on sugar-beet seed by Headden 

 (Co'lo. Sta. Bui. 46) it was found that while 1000 parts 

 per million of sodium carbonate permitted the seed to 

 germinate freely, 5000 parts per million was injurious. 

 The limit for sodium sulphate was about 8000 and for a 

 mixture of the two about the same as the sodium carbonate. 



From the foregoing discussion of the various experi- 

 ments with alkali under different conditions and from the 

 results given in Chapter XIV on crops for alkali land, it 

 is seen that the limits vary so widely because of the dif- 

 ferent methods of arriving at these limits, that unless the 

 conditions can be duplicated, considerable error might 

 result from estimates secured by different experimenters. 

 The estimates under field conditions would be expected 

 to range through a wider limit because of the complicated 

 changes within the soils and because of differences in de- 



