32 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



[Vol. 6 



Owing to an accident in the introduction of the alkali into 

 the Kjeldahl flasks to the set of 4 cultures, 45-48, these nitro- 

 gen determinations were lost. In spite of these irregularities 

 the experiment shows undoubted benefit resulting from the 

 addition of tricalcium phosphate, the same material filtered 

 off in experiment 1 above. It is interesting to note that only 

 when both the carbonate and the phosphate of calcium were 

 added was good growth obtained. 



0.2 gm. 



MEDIA WHICH FORMED NO PRECIPITATE 



Medium of Lohnis and Smith. — Lohnis and Smith ( '16, p. 

 686) state that a medium of the following composition is ex- 

 cellent for supporting the growth of Azotobacter and remains 

 perfectly clear: 



Dextrose 20 gms. 



Dipotassium phosphate 1 0.2 gm. 



Sodium chloride 



Magnesium sulphate 0.2 gm. 



Calcium sulphate 0.1 gm. 



10% ferric chloride 2 drops 



Distilled water 1000 cc. 



This medium is essentially Ashby's solution, hence we felt 

 sure it would yield a precipitate. When the above materials 

 were dissolved in the cold the solution was almost but not 

 quite clear. On heating to boiling a slight flocculent precipi- 

 tate formed. After the solution had cooled this precipitate 

 was filtered off and 25-cc. portions of the perfectly clear fil- 

 trate pipetted into 1000-cc. Erlenmeyer flasks. Two series 

 of flasks of 6 each were prepared, the one series receiving a 



pinch of calcium carbonate per flask, the other not. The 

 method of inoculation was the same as that used in the two 

 previous experiments, the same suspension being used. Two 

 flasks of each series were placed on the rotator, and the re- 

 maining ones on the shelf near by. The incubation conditions 



were the same as in previous experiments. The results are 

 given in table x. 



1 Lohnis and Smith used monopotassium phosphate neutralized to phenol- 

 phthalein with sodium hydroxide. 



