64 Mineral Nutrition of Plants 



colloid and electrolyte. These data, like those obtained potentiometri- 

 cally, were generally interpreted on the assumption that all exchange 

 cations of a given kind on a given colloid were held by the same bond- 

 ing energy. Some evidence at variance with this simple concept grad- 

 ually accumulated but was not regarded as sufficiently general to in- 

 validate it. 



Three main items may be mentioned, (a) Certain potentiometric 

 titration curves of clays showed more than one inflection, {b) The last 

 traces of exchangeable hydrogen were found, in general, to be extremely 

 difficult to replace by metallic cations, (c) Hysteresis effects were found 

 in certain exchange reactions; that is, the equilibrium was different 

 when approached from different sides. 



The logical consequences of a variation in the bonding energy of a 

 given cation upon a given soil colloid were first adequately discussed 

 by Jarusov (4), who presented clear experimental evidence that the 

 replaceability of metallic cations varied with the degree of saturation. 

 He emphasized also the importance of the nature of the accompanying 

 ions in determining the replaceability of a given cation, thus fore- 

 shadowing the important work of Jenny and Ayers (5) on the comple- 

 mentary ion principle. 



The work summarized below was made possible by the development 

 in the Missouri Experiment Station, of membrane electrodes which, 

 acting in a somewhat similar fashion to the glass electrode, render the 

 determination of the activities of single cations feasible. The sequence 

 of events was briefly as follows. 



Thin plates cut from crystals of cation exchange minerals were first 

 employed. Below a certain concentration they functioned as electrodes 

 sensitive to a variety of cations. Their preparation was difficult. The 

 next materials tried were clay films produced by the evaporation of 

 colloidal suspensions of high base-exchange clays. By preliminary heat 

 treatments these membranes were found to acquire good mechanical 

 stability and at the same time their electrochemical properties were im- 

 proved (//). In the case of the hydrogen montmorillonite, it was found 

 that membranes could be prepared which were either sensitive both to 

 mono- and divalent cations (pretreatment below 450 ° C.) or which 

 were sensitive only to monovalent cations (pretreatment above 450 



