398 



Journal of Agricultural Research 



Vol. VI. No. II 



Table VI. — Quantity of nitrogen {in milligrams) fixed in loo c. c. of nutritive solution 

 ■with and without the addition of arsenic 



Treatment. 



Inoculated 

 •with Azotobac- 

 ter vinelandii. 



Soil +0.0728 



gm. of sterilized 



lead arsenate. 



Soil +0.0728 



gm. of unsteril- 



ized lead 



arsenate. 



Nutritive solution + i-5 gm- of mannite. . . . 

 Nutritive solution + i-5 gm. of mannite and 



0.0728 gm. of lead arsenate 



Nutritive solution + 0.0728 gm. of lead 



arsenate 



Nutritive solution + i-5 gm. of mannite 



and 0.0272 gm. of arsenic trisulphid 



Nutritive solution + 0.0272 gm. of arsenic 



trisulphid 



14. 12 

 c 

 o 



•5 



15.16 

 14.79 



1-45 



5-98 



.28 



15-77 

 13-72 



•52 

 2.05 



.08 



After the first series had been completed, it was thought possible that 

 the heat in the autoclave had changed the solubility of the arsenical 

 compounds and that this was the reason there was no fixation in the 

 solution with arsenic. For this reason analyses were made of the soluble 

 arsenic in 100 c. c. of the nutritive solution containing arsenic both 

 before and after autoclaving. The determinations were made as pre- 

 viously outlined. The lead arsenate yielded 0.91 mgm. of soluble arsenic 

 before autoclaving and 0.85 mgm. after autoclaving. The arsenic tri- 

 sulphid yielded 0.40 mgm. before autoclaving and 0.42 mgm. after 

 autoclaving. 



The results indicate conclusively that the toxicity of the compound is 

 not due to a difference in the solubility of the compound produced by the 

 heat. In order to make sure of this, a series was arranged in which the 

 arsenic was added just before inoculation and after the solution had been 

 autoclaved. These results are given in the last column of Table VI 

 and are slightly lower than those previously obtained with the arsenic. 

 The A. vinelandii fixed no nitrogen in the presence of the arsenic. Even 

 where the soil was used as the inoculating medium, the lead arsenate 

 retarded nitrogen fixations to a certain extent. The toxic influence of 

 the arsenic sulphid is very pronounced. These results show the care 

 which must be used in drawing conclusions from the Remy-solution 

 method as to what is to be expected in soils. They greatly strengthen 

 the contention of Jonsson (1896) that the fact that Nobbe (1884) found 

 arsenic solutions to be toxic to seedlings in water culture and concluded 

 that arsenic, even in small quantities, is extremely toxic to plants does 

 not indicate that these solutions will be toxic when in the soil. The 

 results herein reported show arsenic to be extremely toxic to nitrogen- 

 fixing organisms while in solution, but the same concentration in the 

 soil is not only devoid of toxicity but acts as a powerful stimulant. This 

 therefore establishes for the bacteria what Kanda (1904, p. 16) found 

 to be true for the higher plants — namely, that dilute solutions of sub- 



