21 



Experiments to ascertain the limit of endurance in pure solutions 

 were also made with seedlings of alfalfa {Medicago sativa). Although 

 absolute limits for this plant have not, as yet, been determined, they 

 appear to be somewhat lower for every salt than inthecaseof ivi<.pmws 

 albus^hwt more than one-half as high. Thus for magnesium sulphate 

 the limit appears to lie between 0.000625 and 0.00125 normal, while for 

 magnesium chloride the limit Avill be found between 0.00125 and 0.0025 

 normal. 



A glance at the preceding table shows very clearly that it is the 

 basic rather than the acid radicle of the salts used which chiefly deter- 

 mines their relative toxicitj^ In other words, the cathions derived 

 from these salts are A^ery much more active in their effect upon plant 

 tissues than are the anions. This is strikingly brought out \>y a 

 comparison among themselves of the three chlorides of magnesium, 

 sodium, and calcium, on the one hand, and of the chlorides and 

 sulphates of magnesium and sodium, respectively, on the other. In 

 the former case, although the anions (CI) are identical in kind Ave find 

 magnesium chloride eight times as toxic as sodium chloride, and one 

 hundred times as toxic as calcium chloride. In the latter case, mag- 

 nesium sulphate is only twice as toxic as magnesium chloride, A\iiile 

 sodium sulphate is little more than two and one-half times as injurious 

 as the corresponding chloride. 



The results Avith salts of magnesium, as compared Avith those of 

 sodium, confirm the results obtained by W. Wolf, Loew, and others 

 as to the strongly poisonous qualities of the former base. 



All four of the salts of sodium with which experiments were 

 made are widely distributed and often A^ery abundant in the alkali 

 regions of the Avestern United States. As Avas to be expected, sodium 

 carbonate or black alkali was found to be the most harmful of these, 

 but it is not much more injurious than sodium sulphate. That the 

 latter is much more poisonous than sodium chloride is a result not 

 altogether anticipated at the beginning of the investigation.^ As Avas 

 predicted, sodium bicarbonate proved to be somewhat less toxic than 

 sodium chloride.^ 



As a matter of fact, the limit of endurance in a solution of sodium 

 bicarbonate is not much higher in i)arts of salt per 100,000 of Avater 



'Stewart [Ninth Ann. Rep. Utah Agr. Exp. Sta. p. 26 (1898)] found sodium 

 chloride more injurious than sodium sulphate to germinating seeds of legumes and 

 cereals. 



-Very different results from these here recorded as to the relative toxicity of the 

 carbonate and bicarbonate of sodium were obtained by Coupin [Rev. Gen. de 

 Botanique, 12, 180 (1900) J. Experimenting with seedlings of wheat, this author 

 found that the least concentrated fatal solution (" equivalent toxique'') is 1.1 grams 

 per 100 of water for sodium carbonate, while for the bicarbonate it is 0.6 gram. 

 Hence the latter would be twice instead of one-fourth as poisonous as the former. 

 Sigmund [Landw. Versuchsst., 47, 2 (1896)] found that while NajCOgat a concen- 

 tration of O.ij per cent killed germinating seedlings of vetch and rape and retarded 

 the development of wheat seedlings, NaHCOg at the same concentration was 

 harmless. 



