686 Journal of Agricultural Research vol. vi, No. is 



Mannite soil extract, which has been used with very satisfactory 

 results by the senior author (13) for more than 10 years for the study 

 of nitrogen-fixing and other soil organisms, unquestionably has the dis- 

 advantage of a varying and partially unknown chemical composition. 

 However, beef broth and similar substrates are liable to the same objec- 

 tion and yet are generally used in bacteriological laboratories. Never- 

 theless it would be preferable to have media at our disposition of the 

 same favorable qualities but of well-defined chemical composition. 

 Practically all of the many artificial substrates recommended in the 

 bacteriological literature are much too concentrated, especially for soil 

 organisms. For many of our experiments we used the following min- 

 eral solution to good advantage: 



Monopotassium phosphate neutralized to phenolphthalein 



by sodium hydroxid 0.02 per cent 



Magnesium sulphate 02 per cent 



Sodium chlorid 02 per cent 



Calcium sulphate 01 per cent 



Ferric chlorid, 1 per cent solution 2 drops per 100 c. c. 



As carbonaceous material for B. azotobacier, 1 per cent mannite was 

 used. The further addition of 0.02 per cent of potassium nitrate or 

 peptone proved to be beneficial though not necessary. All these solu- 

 tions are entirely clear and therefore especially suitable for microscop- 

 ical studies. If a strip of filter paper sufficiently long to reach about 

 1 inch out of the solution is placed into the test tube before sterilizing, 

 a luxuriant growth of the large forms belonging to types A and B quickly 

 spreads on the part above the liquid. 1 In old solutions the symplasm 

 frequently develops to such a degree that it becomes clearly visible to the 

 naked eye as white flakes or slimy threads. Figure 14 (PI. C) is a repro- 

 duction of the end of such an enormous accumulation of living material. 



The last three pictures of our Azotobacter series (fig. 22 to 24 of PI. D) 

 illustrate one of the comparatively rare direct connections between the 

 small and the large forms. We have here before us the exact counter- 

 part of the alteration shown in figure 20 (PI. D). Certainly this direct 

 growing up of the small organisms to forms belonging to type B, the 

 forthcoming of threads of type G in the germinating process, and the 

 unusual appearance of their budding, by which the small forms are 

 regenerated, deserve our full attention. However, this kind of develop- 

 ment seems to be a rather rare exception to the rule. These forms also 

 are much inclined to turn into the symplastic stage. A photographic 

 picture of this occurrence will be published later. 



The close study of this side connection, however, led us to another dis- 

 covery which we had failed to make before, although our other preparates, 



1 The arrangement mentioned above is very helpful for obtaining pure cultures of Azotobacter from the 

 soil. At the same time it allows the motility of an organism to be determined macroscopically. One of 

 our strains crept up 20 cm. in 10 days on long paper strips in large test tubes. 



