10 



classifying the organisms as separate species. Beyerinck 

 proposed the name Bacillus radicicola, for these nodule pro- 

 ducing bacteria; as to whether there are more than one species, 

 authorities are still undetermined. Hereby is evolved a 

 rational system for the continuous addition of nitrogen to the 

 soil, an increase which can not only be enjoyed and ap- 

 propriated by the leguminous plants, but likewise by succeed- 

 ing vegetable growth. 



The non-symbiotic fixation of nitrogen possesses the im- 

 portant feature of having more universal application. The 

 following aerobic species are the most vigorous nitrogen fixing 

 organisms hitherto discovered: A. agilis, A. chroococcum, 

 A. vinelandi, A. beyerincki, A. vitreum, and A. woodstowni. 

 Of these A. chroococcum is in all probability the most common 

 in our soil. I have isolated this organism from many parts of 

 the state. 



STUDIES ON IMPURE CULTURES 



In order to determine the relative nitrogen-fixing power of 

 our soils, the aforementioned samples were inoculated into Ash- 

 by's medium, and the folio wing process and technique followed: 

 One hundred cubic centimeters of Ashby's medium was meas- 

 ured into 250 ccErlenmeyer flasks, and sterilized: these flasks, 

 therefore, each contained two grams of mannite; the medium 

 occupying about three-fourths of an inch in depth in the 

 bottom of the flask, there remained above the surface an 

 abundant air space. These flasks were each inoculated with 

 one gram of soil, and incubated at room temperature, which 

 prevailed at about 33C, for twenty-one days. At the end of 

 this period the entire contents of these flasks were transferred 

 to the Kjeldahl apparatus and the total nitrogen content 

 determined by the Kjeldahl method, the ammonia being 

 distilled over into tenth-normal sulphuric acid and titrated 

 back with tenth-normal sodium hydroxide, using congo red as 

 indicator. The original nitrogen content of the soil was 

 determined by the Kjeldahl method, using ten grams of 

 sample. 



Each operation during the investigation was carefully 

 checked in order to reduce the possible error to the limit of 

 experimental manipulation. Samples number 48 and 49 were 



