300 



canal water for infection, give but slowly a moderately developed bacteria film but 

 it grows during a very long time. The film on the citrate is rich in spirilla and 

 the Azotobacter form found in it differs in many respects from the ordinary varieties. 

 In all these cases the quantity of bacteria grown during the first 2 or 3 weeks, is 

 still too slight to necessitate a determination of the nitrogen, and could by a rough 

 comparison with former computations be valued at some tenths of milligrams N 2 per 

 gram of the dissolved lime salt. After a long time however, the fixation of nitrogen 

 with these salts is also considerable. 



With calciumglycolate in absence of nitrogen compounds no growth of microbes 

 could be observed at all. 



2. Quantity of the fixed nitrogen. 



Neglecting for the moment the volatile acid, to which we shall return below, the 

 analysis of the cultures is performed as follows. 



The whole quantity of the liquid, in which are present the calciumcarbonate 

 formed by oxidation from the malate or the other organic salt, besides the as yet not 

 decomposed malate, the salt of the volatile acid, and the bacteria, is treated with a 

 known quantity of normal hydrochloric acid by which the carbonic acid is expelled 

 on heating; a then following titration with normal alkali and phenolphtaleine as indi- 

 cator, shows how much calciumcarbonate is produced and consequently how of the 

 organic salt is oxidised. 



After addition of a little sulphuric acid the liquid is evaporated to dryness and 

 after K j e 1 d a h 1's method examined on nitrogen, while in each of the materials 

 used the rate of nitrogen is stated separately. The calciummalate of Merck, 

 Darmstadt, proved nearly free from nitrogen. 



Now follows a table of some analyses 1 ) which give an idea of the amount of 

 nitrogen fixed through Asotobacter, when organic salts are used as carbon food. 

 (See table.) 



These numbers show that the amount of nitrogen which can be fixed in the 

 crude culture is at most 4.9 and 2,8 m.g. per gram of oxidised calciumsalt, obtained 

 respectively with calciumpropionate and calciumacetate (experiment 10 and n), 

 \\hile, per gram of calciummalate was fixed about 2,6 m.g. (experiment 2), and per 

 gram of lactate 1,8 m.g. (experiment 9). It seems that the fixation goes on more 

 rapidly at the beginning than later in the course of the experiment, whence it fol- 

 lows that when little of the organic salt is used proportionately more nitrogen is fixed 

 than by larger amounts. This should be taken into consideration in judging the 

 favourable results obtained with propionate and acetate, for then solutions were used 

 with only i% of the salt. As to these salts, they have proved to be in general an 

 unfavourable source of carbon for Azotobacter if the rapidity of the growth is taken 

 as indicator of the process, and only then to be able to give good results, when for 

 the inoculation, cultures in malate solutions are used, in which a certain variety of 

 our species is present. But also then, as observed above, already at the first passage 



') I owe to Mr. D. C. J. Mink man, assistant to my laboratory the determina- 

 tions here referred to. 



