256 



SCIENCE 



[N. S. Vol. XLI. No. 1050 



ened the writer's confidence in the hjrpotliesis 

 and led to further investigations. One of 

 these which was of special interest related to 

 acids. For a number of reasons it was sup- 

 posed that acid would not cause a decrease 

 of permeability. But investigation showed 

 that such a decrease actually occurred in the 

 presence of HCl and it was then a simple 

 matter to predict that antagonism would be 

 found between NaOl and HCl. This turned 

 out to be the case, the amount of antagonism 

 corresponding to the amount of decrease of 

 permeability.® 



The hypothesis was further tested by inves- 

 tigations on other salts, the most interesting 

 of which are those which (in contrast to those 

 just mentioned) are more effective than CaClj 

 in decreasing permeability, such as La^CNOs)^ 

 CejCNOj),,, etc. Here also it was found that 

 the degree of antagonistic action could be 

 foretold by observing the amount of decrease 

 of permeability produced by the pure salts. 

 The results of these investigations aSord 

 strong support to the hypothesis. 



It seems to the writer that the hypothesis 

 offers a rational esplanation of antagonism by 

 showing that salts antagonize each other be- 

 cause they produce opposite effects on the pro- 

 toplasm and by stating definitely what these 

 effects are (it should be noted that they have 

 been measured with considerable accuracy). 



The soundness of this point of view is indi- 

 cated not only by the fact that we are able to 

 predict both qualitatively and (to a consider- 

 able extent) quantitatively the effect of com- 

 binations of salts^ but also by the very signif- 



6 The Journal of Biochemistry, 19, 1914. 



' It should be noted that mixing solutions of 

 two salts which belong to different classes does not 

 produce an effect which is merely intermediate be- 

 tween the two. For example, tissue may be killed 

 by an exposure of 24 hours in NaCl or in CaClj but 

 remain normal in a mixture of these in the proper 

 proportions. Cf. Pringsheim's Jahrh. f. wiss. Bot., 

 54, 645, 1914. 



The writer has found cases in which two sub- 

 stances which can decrease permeability are able 

 to antagonize each other. So far as the writer's 

 experiments with Laminaria have gone there is no 

 great amount of antagonism in such cases and 



icant fact that we are able to extend this con- 

 ception to organic compounds and to show 

 that non-electrolytes which decrease perme- 

 ability can also antagonize such substances as 

 NaCl. These facts indicate that the hypothesis 

 may be applied in a general manner so as to 

 include both electrolytes and non-electrols^tes. 

 w. j. v. osterhout 

 Harvard University, 

 Laboratory of Plant Physiology 



ISOLATION OP bacillus EADICICOLA PROM SOIL 



Ever since the epoch-making achievement 

 of Hellriegel and Wilf arth, reported in 1887, 

 which established the symbiotic relationship 

 between bacteria and legumes in the fixation 

 of atmospheric nitrogen, the legume bacteria, 

 named in 1901, Bacillus radicicola by Beije- 

 rinck, have been the object of numerous in- 

 vestigations in all parts of the world. These 

 investigations have assumed a variety of forms 

 and were planned from both the economic and 

 pure science points of view. There has ever 

 remained, nevertheless, the unsolved problem 

 of the direct isolation of Bacillus radicicola 

 from the soil. Sporadic attempts, rather few in 

 number, have been made to attain that end, but, 



what there is may perhaps be correlated with the 

 fact that all substances which decrease permeabil- 

 ity do not act alike, some producing a much 

 greater decrease than others. Moreover these sub- 

 stances will, if the exposure be sufficiently pro- 

 longed, alter their action and increased permeabil- 

 ity. The rapidity of this change varies with dif- 

 ferent substances and this may be related to the 

 fact that some of these substances antagonize each 

 other to some degree. This will be more fully dis- 

 cussed in a subsequent paper. 



Experiments on some plants (in which the cri- 

 terion of antagonism is not electrical resistance 

 but growth) show a fairly strong antagonism be- 

 tween magnesium and calcium. It is possible that 

 for these plants magnesium belongs in the first 

 class. 



It will be noted that the hypothesis, as here set 

 forth, says nothing about the mutual relations of 

 substances belonging to the same class but merely 

 states that substances of one class will antagonize 

 those of the other. In this form the hypothesis 

 is completely justified by all the experiments, in- 

 cluding those on organic substances. 



