1914] 



OSTERHOUT— CRITERIA OF ANTAGONISM 



183 



It would be possible to determine this additive curve experi- 



mentally, and then to 

 example, at the point 

 P it would be ex- 



express antagonism quantitatively; 



for 



pressed as 



UT 

 TP 



But 



the labor would be 

 much greater than by 

 the method of mixing 

 equally toxic solu- 

 tions. The additive 

 curve would be deter- 

 mined by growing 

 plants, not in mix- 

 tures of A with B, 

 but in mixtures of A 

 with another solution 

 of A having the same 

 toxicity as B. Or we 

 might use mixtures 

 of B with another 

 solution of B having 

 the same toxicity as 

 A . The two methods 

 might not give 

 exactly the same re- 



sult. This is 



N 







P 



a 



R 



A 100 



75 



50 



25 







o 



B 



25 



50 



75 



100 



Fig. 3. — Curves showing growth in mixtures of 



an unequally toxic solutions: the ordinates express 



additional argument growth; the abscissas express the composition of the 

 in favor of using mixtures as in fig. 1; the dotted line VTW expresses 



equally toxic solu- 

 tions. 



the growth which would occur if there were no antago- 

 nism (additive effect); VUW, antagonism curve; 

 VSW, curve expressing increased toxicity (opposite of 



An illustration of antagonism); the quantitative expression of antago- 



UT 



TP 



this method is found nism at the point P is 



in the results given 

 in table II. The growth in the various mixtures (additive and 

 antagonistic) was in part determined directly and in part was 

 calculated from results obtained by growing plants in mixtures 



