KOCH'S POSTULATES 47 



fleeted and transmitted light alternately. Having determined the 

 presence of two or more different kinds of colonies, make a ring with 

 wax pencil around one or more of each kind of colony, numbering them. 

 The slides or culture tubes used in determining the species of organism 

 present in the plate should bear the same number as that of the colony 

 from which the material was taken. A convenient procedure is to put 

 a loopful of water on a clean cover-glass and to emulsify material from 

 a colony in it. Then invert over a concave slide without vaselining 

 the circumference of the concavity. After examining for motility, 

 smear out and dry the bacterial preparation. Then fix in the flame 

 and stain with aniline gentian violet for two to five minutes. Wash 

 and mount the preparation in water. Afterward pass through the usual 

 Gram technic. 



After this inoculate the various culture media from similar colonies. 

 One may inoculate a tube of bouillon from a single colony and later 

 on inoculate the other culture tubes. 



In testing for gas production it is better to use the Durham fermenta- 

 tion tube as small amounts of gas may not be easily detected with deep 

 stab cultures into glucose or lactose agar. 



If a Durham or Smith tube be not at hand the production of gas may be deter- 

 mined by observing bubble formation on the surface of the sugar bouillon culture. 

 As none of the pathogenic cocci produce gas, fermentation tubes are unnecessary 

 where cocci are to be studied. The litmus milk tube gives data as to acid production. 



An important point is to wait at least forty-eight hours (in the case 

 of M. melitensis, four to seven days) before reporting on the cultural 

 findings on the agar or blood-serum slant or plate upon which the 

 material is smeared (pus, exudate, blood, etc.). 



Should an organism be encountered in original investigations these 

 requirements as to etiological relationship should be carried out (Koch's 

 postulates), i. The organism should be constantly present in that 

 particular pathological condition. 2. Such bacteria should be isolated 

 in pure culture from the pathological material. 3. Such pure cultures 

 when inoculated into suitable animals should reproduce the pathologi- 

 cal conditions and should be capable of a second isolation in pure culture 

 from such an experimental animal. For various reasons, such as unsuit- 

 able animals or artificial media, these requirements are impossible of 

 execution with several organisms which are generally recognized as the 

 causes of certain diseases. 



The experimental animals most frequently employed in the diagno- 



