[Vol. 6 



22 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



method is that proteins are removed by the Michaelis-Rona 

 colloidal iron precipitation, and the centrifuge used for clari- 

 fying the solution and recovery of the cuprous oxide, which 

 is then determined by Bertrand's method. 



In our first experiments dextrose was determined on one 

 culture and nitrogen on its duplicate; later the procedure 

 was modified so that sugar and nitrogen were determined on 

 the same culture with reasonable accuracy. The procedure 

 finally adopted follows : The culture medium is acidified with 

 N/2 HC1 and warmed till mineral salts are in solution and 

 the proteins dispersed to an opalescent solution. After cool- 

 ing the material is transferred to a 100-cc. or a 250-cc. volu- 

 metric flask, depending on the size of the culture, and made 

 to the mark. An aliquot of this suspension is transferred to 

 a 100-cc. volumetric flask, the volume made to approximately 

 75 cc, and a pinch of sodium acetate added to reduce the hy- 

 drogen ion concentration. Five cc. of Merck's colloidal iron 

 are then added, the suspension well mixed, and approximately 

 0.2 gm. Na2S0 4 added, and water added to the mark. The 

 suspension is again well mixed, poured into a 100-cc. centri- 

 fuge tube, and centrifuged for 15 minutes. Duplicate 20-cc. 

 portions of the clear supernatant liquid are then transferred 

 to 50-cc. centrifuge tubes. The procedure from this point on 

 is the same as that outlined by Shaffer. 



The method of Shaffer is really a ' 'micro" method pro- 

 posed for the determination of sugar in blood, where only 

 small samples of a tissue low in sugar can be analyzed. The 

 conditions worked out by Shaffer cover naturally a compara- 

 tively narrow range of dextrose amounts; hence in working 

 with cultures of Azotobacter which contain, in the controls 

 at least, very large amounts of sugar, it is easily possible to 

 draw off in aliquoting a too large amount of dextrose in solu- 

 tion. On the other hand, it is just as easy to remove an aliquot 

 so small that the error of the analysis is multiplied by a too 

 large factor in computing the amount of sugar in the portion 

 of culture medium under examination. The latter error was 

 made in some of our determinations and probably accounts for 

 the results indicated below as being of questionable value. 



