1914] 



OSTERHOUT— CRITERIA OF ANTAGONISM 



185 



solutions of copper salts, while the roots (which are immersed in 

 the solution) do not grow at all. The reason is that as the solution 

 passes through the roots to the leaves a great part of the copper 

 is removed either by adsorption or by chemical combination. 3 



Finally, it 



m 



be 



noted that quantitative 

 results are more reliable 

 and much more easily ob- 

 tained when uniform 



material is used. 





Pure 



line" wheat and other 

 seeds may now be ob- 

 tained and should be used 

 whenever possible. When 

 they are not obtainable 

 the following method 

 (suggested by Professor 

 Johannsen) may be em- T 

 ployed. Several heads of 



B 



Fig. 4. — Types of antagonism curves: the 

 ordinates express growth; the abscissa? express 



wheat are taken. One 



grain of wheat from each the composition of the mixtures as in fig. 1. 



is placed in each solution 



(and these grains should resemble each other as much as possible). 



In this way each solution receives the same kinds of wheat, and an 



average of the growth of all the plants in any solution may be 



safely used for comparison with the average in any other solution. 



It is desirable to employ this method even when pure line seeds 

 are used. 



Summary 



method 



criterion of antagonism, since we know at the outset just what 



mixture 



Mixtures of two equally toxic solutions must have precisely the 

 same effect on growth as the pure solutions themselves, provided 

 that the effects of the salts are additive. If antagonism exists there 



3 Cf. Bot. Gaz. 44:268. 1907. 



