23 



As there is probably but a small difference in the amount of sodium 

 ions yielded by sodium chloride and by sodium hydrogen carbonate, 

 at the dilutions here involved, the difference in their toxicity observed 

 must in all probability be ascribed mainly to the anions. 



It is likely that the great toxicity of normal sodium carbonate is 

 largely due to the hydroxyl ions resulting from the hj^drolysis of 

 this salt. In the case of the bicarbonate of sodium in all the experi- 

 ments involving its use, and described in this paper, hydrolysis was 

 avoided by dissolving carbon dioxide in the solution in amounts suf- 

 ficient to prevent any inversion to the normal carbonate, a reaction 

 which would necessarih' result were hydrolj'sis of the bicarbonate 

 permitted.^ Since it seems reasonably certain that HCO3 ions are not 

 toxic, the toxic influence of the sodium bicarbonate solutions could 

 be safely attributed to the sodium ion alone were it not for the fact 

 that toxic solutions of this salt produce the peculiar " clearing " effect 

 upon plant tissues which is well known in the case of the normal car- 

 bonate of sodium and of the hydrates of potassium and of sodium. 

 This effect is vevy different from that caused by other salts of sodium, 

 e. g., the sulphate and the chloride. 



Calcium chloride was found to be ten times less injurious than is 

 sodium chloride. For this reason, and because it rarelj' predominates 

 in areas of any considerable size, this salt can not be regarded as, 

 under ordinary circumstances, a dangerous component of alkali soils. 

 As we shall presenth' see, there is reason to believe that it can in 

 many cases be a highly beneficial component of the soil. 



Attention should be directed to the fact that the figures given in 

 the above table represent only approximate results, the determination 

 of the absolute limit for each salt depending theoretically upon the 

 testing of an almost infinite number of concentrations. Thus, as a 

 rule, solutions of a concentration of 0.2, 0.15, 0.1, 0.075, 0.050, etc., 

 normal were emjjloyed, although more numerous intermediate concen- 

 trations, e. g., of 0.2000, 0.1825, 0.1750, etc., normal could have been 

 tested. However, it is doubtful whether the reaction upon plant 

 tissues of finer differences could be detected, and it is believed that 

 for all practical purposes a sufficient number of concentrations was 

 used. As has already been noted, the limits of endurance in the case 

 of different salts are not of lorecisel}^ equal value, the roots not sur- 

 viving in all in exactly the same condition. Thus roots which survived 

 after twenty-four hours in a 0.005 normal solution of sodium carbonate 

 presented a perfect appearance and grew vigorously in distilled water 

 during a subsequent period of twent}' -four hours. On the other hand, 

 roots which endured a 0.25 normal solution of calcium chloride pre- 

 sented a markedly abnormal aspect at the end of twenty-four hours, and 

 made little subsequent growth when transferred to water. Likewise 



^See paper on Equilibrium between Normal Carbonates and Bicarbonates in 

 Aqueous Solutions, Cameron and Briggs. Bui. 18, Div. Soils, U. S. Department 

 of Agriculture (1901); Jour.. Physical Chem., 5, 537 (1901). 



