Septembee 14, 1917] 



SCIENCE 



253 



ative actions ; and an organism capable of vig- 

 orous activity in pure culture may be almost 

 inactive in the presence of its natural rivals. 

 Laboratory tests, therefore, give but indirect 

 evidence at their best. Indirect evidence has 

 its value; but it is futile to draw conclusions 

 from it unless results obtained by one method 

 are confirmed by those obtained in some other 

 way. 



Similar difficulties in regard to pathogenic 

 bacteria caused the literature of the early nine- 

 teenth century to abound in misstatements as 

 to the relation of certain bacteria to certain 

 diseases. Gradually, however, it came to be 

 recognized that neither the constant presence 

 of a given microorganism in a certain disease, 

 nor its ability to produce a similar disease in 

 lower animals proves it to be the causal agent 

 of a human disease. These ideas were put in 

 concise form by Koch when he restated and 

 emphasized the requirements originally sug- 

 gested by Henle as necessary steps in proving 

 a given organism to be the cause of a given 

 disease. These postulates, as stated by Koch, 

 are as follows: (1) The organism must be 

 shown to be present in abundance in the tis- 

 sues, blood, or discharges of animals suffering 

 from the disease; (2) it must be isolated and 

 studied in pure culture; (3) it must be shown 

 capable of producing the same disease in 

 healthy animals; (4) it must subsequently be 

 foimd again in abundance in the experiment- 

 ally inoculated animals. 



Really the case of bacterial activities in 

 soil is analogous. The constant presence of 

 a certain organism in manured soil, for in- 

 stance, does not prove that it decomposes the 

 manure any more than the constant presence 

 of an organism in a given disease proves its 

 causal relation. Neither does the fact that 

 an organism ammonifies laboratory media prove 

 that it ammonifies organic matter in soil, any 

 more than the fact that an organism produces 

 a certain disease in a lower animal proves that 

 it produces a similar disease in man. Al- 

 though this fact may be recognized in a gen- 

 eral way by soil bacteriologists, a little 

 thought will show that no rules as strict as 

 Koch's postulates have ever been followed in 



establishing the agency of bacteria in any soil 

 activity — with the exception of the bacteria of 

 legume nodules. Even in regard to the nitri- 

 fiers — certain as we may be of their agency in 

 converting ammonium salts into nitrates — we 

 do not have the complete proof. This thought 

 is somewhat disconcerting and shows the need 

 of drawing up strict rules to apply to the 

 activities of soil microorganisms. Koch's pos- 

 tulates can not be applied directly to soil 

 microorganisms, because the latter operate 

 under quite different conditions from patho- 

 genic bacteria; but it is possible to modify 

 his rules to fit soil conditions. 



The first postulate is that the organism must 

 be shown to be present in abimdance in an- 

 imals suffering from the disease in question. 

 It is equally necessary to show that an organ- 

 ism is present in abundance in soil in which a 

 certain biological activity is going on — in 

 fact that it is more abundant in such soil 

 than in similar soil in which the activity is 

 not taking place — ^before asserting that the or- 

 ganism in question is the causal agent. It is 

 also necessary to show that the organism is 

 present in such soil in active form. This 

 is necessary because at least three groups of 

 soil microorganisms have inactive as well as 

 active forms : namely, protozoa, molds and 

 spore-bearing bacteria. If the organism in 

 question belongs to one of these three groups, 

 the mere demonstration of its presence is not 

 enough, but it must be shown to be present in 

 active form. In other words, Koch's first pos- 

 tulate must be expanded as follows when ap- 

 plied to soil conditions : The organism in ques- 

 tion must be shown to be present in active 

 form when the chemical transformation under 

 investigation is taking place ; and must also be 

 shown to be present in larger numbers in such 

 soil than in similar soil in which the chemical 

 change is not taking place. These two steps 

 have seldom been carried out in investigating 

 the cause of any biological activity in soil, 

 but they are nevertheless as important as 

 Koch's first postvdate in regard to pathogenic 

 bacteria. They are perhaps a little more 

 stringent than the first postulate of Koch's; 

 but special stringency is necessary here in view 



