METHODS 273 



a coarsely granular structure extending to the margin. The colonies 

 rapidly increase in size, and after a week or more become brown at 

 the center with concentric rings alternating dark and white to the 

 circumference and darker streaks radiating from the center outward. 

 In old cultures, where the agar has partly dried up, the cells are often 

 united in sarcina-like packets; the cell walls are much swollen and 

 the contents are aggregated to a small ball at the center. At the 

 same time giant cells, both round and elongated and filled with oil 

 drops, can be seen. Often a number of involution forms are seen, 

 drawn out with long threads and false septa. By successive 

 dilutions and transfers, it may be obtained in pure culture, although 

 at times considerable difficulty is experienced in freeing it from a 

 small organism B. radiobacter. 



Several different methods have been used for studying its nitro- 

 gen-fixing powers: 



(a) Seeding into 100 c.c. of the medium given above and after a 

 certain time determining the nitrogen. 



(6) The use of the same medium, but the addition of sufficient 

 sand for the formation of sand slopes on which the organism can 



grow. 



(c) The addition of a definite quantity of a carbohydrate to a 

 soil and the incubation of this. 



Each of these methods has its value. The first is much more 

 readily handled in the final Kjeldahl determination, but the others 

 give much higher results. 



Freudenreich found that when Azotobacter are grown upon gyp- 

 sum, the gain in nitrogen is considerably in excess of that assimilated 

 in the liquid media. Krainsky found Azotobacter to utilize from 100 

 to 200 gm. of sugar in the assimilation of 1 gm. of nitrogen when 

 grown in solution, but when grown on sand it required only 11 to 

 30 gm. for the same fixation. Many other workers have noted 

 similar variation when grown in the soil. Where the organisms 

 have been grown on gypsum or soil, we may attribute the stimula- 

 tion to certain soluble constituents, yet this explanation scarcely 

 seems plausible when considered in relation to sand cultures. Three 

 strains of Azotobacter were grown in Ashby's mannite solution 

 and sand (nearly pure silicon dioxid) to which Ashby's solution was 

 added, with the following results: 



Moreover, arsenic is very toxic in the solution, whereas when 

 added to the soil or to pure quartz, in small quantities, it stimulates. 

 Although the total quantities of nitrogen fixed under the two 

 18 



