1914] 



OSTERHOUT— ANTAGONISM 



2 75 



>m 



measunn 



ordinate at the required point, subtracting 55 which is the amount 

 of growth in the pure solutions, and in this case (since all the pure 

 solutions are equally toxic) the amount of growth which would 

 occur if the toxic actions of 

 the salts were additive (that 

 is, if each salt exerted its 

 own toxic effect independ- 

 ently of the other salts); 

 the result should then be 

 divided by 55. 5 



In this case the additive 



effect 



is represented by a 



plane surface parallel to the 



rms 



figu 



of the model. The height 

 of this plane is indicated by 

 the shading in the 



Other methods (as mix- 

 ing unequally toxic solutions 

 or keeping the concentration 

 of one salt constant while 

 varying that of the others) 



will 



give for the additive 



effect a curved surface very 

 difficult to determine and 

 not easily represented or 

 measured on the model. 



K 



Na 



Ca 



With 



more 



than 



expressed 



Fig. 2. — Solid model showing the forms of 

 the antagonism curves in all possible mixtures 

 of XaCl 0.12 M, KC1 0.13 M, and CaCL 

 o. 164 M. 



lid model 

 following 



Let us suppose that equally toxic solutions 



mixed 



mixture 



may 



(diff 



called solution 2 7 etc.). To solution 1 various amounts of D 



s For a fuller discussion see Bot. Gaz. 58: 178. 1914- 



