NlTKATES. 103 



as sulphates (most frequently calcium sulphate), in organic com- 

 pounds, and occasionally also as sulphides (e.g., iron pyrites). 

 The sulphuric acid, existing as such, may be determined by 

 digestion of the soil with dilute hydrochloric acid, nitration, 

 and precipitation of the solution with barium chloride in the 

 usual way employed for sulphuric acid determinations. The 

 total sulphur may be sufficiently well determined by digestion 

 with concentrated nitric acid for six hours, diluting with water, 

 filtering, and precipitating the filtrate with barium chloride. 



Determination of nitrates. The amount of nitrate present in 

 a soil at any given time is usually very small and it is not, as a 

 rule, necessary to determine it. 



Sometimes, however, a determination is made for special 

 purposes. 



If such be required.it is important that the soil be air-dried 

 as soon after its collection as possible, otherwise the process 

 of nitrification may continue and increase the amount of 

 nitrates. It is advisable to accelerate the process of drying 

 by heating the soil to 30 or 40 C. and drawing a current of 

 warm air over it. An arrangement which the author finds 

 very convenient for drying soils before analysis is described 

 on p. 91. With this apparatus the operation can be rapidly 

 completed without the possibility of contamination by dust, 

 products of combustion, &c., and with little opportunity for 

 nitrification to appreciably increase the quantity of nitrates in 

 the soil. 



From the air-dried soil the nitrates are extracted by washing 

 with water. Many methods may be used : 



(1) 1000 grammes are treated with 2 litres of pure water and 



allowed to stand for 48 hours, W 7 ith occasional shaking. 

 1 litre of the liquid is then filtered, mixed with a little 

 pure sodium carbonate, and evaporated to small volume 

 on the water-bath, any precipitate which may form 

 being removed by filtration. 



(2) A cylindrical funnel is made by removing the bottom from 



a " Winchester quart" bottle, placing a disc of copper 



gauze on the shoulder, and covering this with two filter 



, papers. The bottle is then connected by means of a 



cork and tube to a filtering flask and from 200 to 500 



