110 



100,000]." The tabulated results of comparisons of this new method 

 with the Crum mercury method as modified by Franklaud and Arm- 

 strong on numerous samples of water, show the new method to give 

 results in most cases slightly higher than the mercury method ; this 

 error amounted in only one instance to 0.4 part of nitrogen in 100,000, 

 and was usually below 0.2 i)art. 



The advantage claimed for the test are rapidity, simi»licity, and ease 

 of execution, and the small quantity of water required for the deter- 

 mination, 20 c. c. being suflQcient. The only precaution mentioned is to 

 avoid the addition of large quantities of zinc dust, as thoy decolorize 

 the solution. 



Absorption of atmospheric ammonia by arable soils, T. Schlo- 

 sing [Compt. rend, 110 (iS\)0), pp. 421) and 41>i)). — Previous experiments 

 by the author and by other investigators have indicated that arable 

 soils absorb considerable quantities of ammonia from the air, but the 

 conclusions have not been universally accepted. During the years 

 1880-90, twenty-five experiments were instituted by the author for the 

 further study of the problem. 



Specimens of ditlerent kinds of soil were placed in shallow, circular, 

 flat-bottomed glass vessels with a superficial area of 2 square dm., and 

 a depth varying with the weight of the soil. As a certain amount of 

 nitrogen would be brought to the soils in atmospheric dust, one vessel 

 was left empty and the nitrogen determined in the dust collected in it. 

 All these vessels were i)laced inside an ai)paratus arranged to provide 

 for circulation of air. Two other vessels of the same kind eoiitaijiiug 

 dilute sulphuric acid were placed one inside the apparatus and the other 

 in open air. The amounts of ammonia absorbed by the acid in the 

 two vessels were compared to obtain an indication of the amounts of 

 atmospheric ammonia which would come in contact with tlie soil of the 

 experiments as compared with the amounts which would come in con- 

 tact with ordinary cultivated soils. In estimating the amount of nitro- 

 gen gained by the different soils under experiment, a correction was 

 made for the amount of nitrogen in the atmospheric dust. It was found 

 that the acid in the vessel placed near those containing the soils under 

 experiment inside the apparatus, through which a more or less constant 

 current of air was passing, absorbed more ammonia than liad been 

 found to be absorbed by acid exposed to the open air. The inference 

 was that more atmospheric ammonia had been brought in contact with 

 these soils than would be the case with ordinary cultivated soils. A 

 correction for ajumonia contact was therefore made in the figures for 

 total gain of nitrogen by the soils. This correction, which varied with 

 the different soils, was much larger than that for atmospheric dust. 

 The figures quoted below for the annual gain of nitrogen i)er hectare 

 are those obtained after the subtractions for both atmospheric dust and 

 ammonia coutact. 



