394 Journal of Agricultural Research voi. vi, no. n 



(3) Do the arsenic and arsenic compounds act as a source of energy to 

 the nitrogen-fixing organisms or do they so influence the soil flora that 

 it can utilize more economically the carbon compounds available? (4) 

 What nitrogen-fixing organisms are there in the soil which are influenced 

 by arsenic ? 



In order to find whether arsenic influences the nitrogen-fixing powers 

 of other soils in a similar manner, three other soils were tested with and 

 without arsenic. The soils vary greatly in chemical and physical com- 

 position. Soil A is a black loam of very light texture and, for an arid 

 soil, high in nitrogen and humus. It is well supplied with phosphorus, 

 potassium, and calcium carbonate and grew potatoes for 23 years. 

 After this it was planted to oats for 2 years, and during the past 4 years 

 has been planted in alfalfa. It has received some manure. Soil B is a 

 sandy loam of much lighter color than soil A and contained much less 

 humus and nitrogen, but an abundance of other elements. It has been 

 cultivated for 28 years and during this time has been fallowed two sum- 

 mers. The remainder of the time it has been planted in wheat. Soil C 

 is a heavy clay almost devoid of humus. The nitrogen is low, but the 

 soil is well supplied with phosphorus, potassium, and calcium carbonate. 

 While wet it is exceedingly sticky, and on drying it bakes like adobe. It 

 has been tilled for 23 years, and during this time it has been fallowed for 

 3 years. The remainder of the time it has been in wheat. While it has 

 received no manure during this time, it is still very productive. All of 

 the soils are very fertile and well supplied with Azotobacter, and previous 

 work has shown them to have high nitrogen-fixing powers. 



The soils were all air-dried in the dark for 24 hours, ground in a mortar, 

 sieved, weighed, and placed in sterile tumblers. Some were mixed with 

 mannite and arsenic, others with mannite, while still others received 

 only arsenic. They were all incubated in the regular manner, and the 

 nitrogen determined as in the previous series. The results are given in 

 Table III. Each reported result is the average of six closely agreeing 

 determinations. 



A marked stimulation is found in every case where the arsenic and 

 mannite were applied to the soil, as compared with the results obtained 

 where the mannite only was applied. The action of the various arsenical 

 compounds follows the same order in each of these soils that it did in the 

 first soil tested, being greatest with the lead arsenate and least with the 

 sodium arsenate. The nitrogen fixed in the presence of arsenic but in 

 the absence of mannite is usually considerably higher than that fixed in 

 the presence of mannite and absence of arsenic. It would not be right 

 to conclude from these results that the arsenic compounds furnish a 

 source of energy to the nitrogen-fixing organisms, for these soils (Greaves, 

 1 914, p. 456) have been found to fix appreciable quantities of nitrogen 

 when incubated with an optimum moisture content without the addition 

 of any carbon compound. It is likely that the arsenic makes the nitrogen- 



