212 Journal of Agricultural Research voi. iv. No. 3 



between the —CI, — SO4, and — NO3 ions; that such antagonism, how- 

 ever, is not as marked as that between cations; and that the antagonism 

 of one ion to another — for example, — SO4 to — CI — may be greater than 

 under opposite conditions — e. g., —CI to — SO4. Likewise, as between 

 — NO3 and — SO4, the former ion has the superior power to neutralize 

 the toxic effects of the other. In so far as antagonism between the anions 

 is observed by Miyake, he confirms by less striking examples in culture 

 solutions what we have shown takes place in soil cultures. We can not 

 agree, however, on the basis of our results that antagonism between 

 anions is more feeble than that between cations. The differences found 

 by Miyake in the power of two anions to counteract mutually each other's 

 toxicity has been pointed out in relation to the nitrifying bacteria by 

 the senior writer and Burgess in another place (5). 



Our great caution in pointing out differences between soil and solution 

 cultures is explained in several different publications, some of which are 

 cited in this paper. It must be remembered, moreover, that on the basis 

 of direct comparison of the soil and solution cultures Kearney and 

 Cameron (2) pointed out several years ago the very material differences 

 obtaining between all phases of salt effects in solutions and in field 

 experiments. The importance of this point in investigations of salt 

 effects on living organisms which are intended ultimately for practical 

 application can not be overemphasized. 



MAINTENANCE OF THE ALKALI CONTENT IN THE EXPERIMENTAL SOIL 



In anticipation of queries with reference to the maintenance of the 

 original "alkali" concentration in the soils described throughout the 

 experiment the following statement is made. The irrigation was so 

 carried out that drainage from the soils was never noted. In other words, 

 enough water was supplied to provide the plants mth all the moisture 

 necessary, but no excess was employed. By keeping glazed plates 

 beneath the pots it was possible to note constantly the lack of percolation 

 from the pots. Moreover, samples of soil were removed from the pots at 

 the end of the second season of growth and analyzed for "alkali." 

 It was always possible to recover all or very nearly all the sodium chlorid 

 and sodium sulphate that had been originally added. Sometimes the 

 quantity recovered showed slightly less and at other times slightly more 

 than was added. These irregularities are doubtless due to the slightly 

 imperfect mixing of the salts with the soil or are errors inherent in the 

 method of determination employed. 



Quite the contrary was true, however, of the pots receiving sodium 

 carbonate. Not only was it impossible to recover all the sodium carbonate 

 that had been added at the commencement of the experiment, but it was 

 actually possible to recover very little of that salt, the highest percentage 

 recovered being about 25 per cent of the amount first added. This would 



