Haigh — Estimation of Hygroscopic Moisture in Soils. 531 



"When an estimation is to be made, excess of powdered carbide (about 

 three times the weight of the soil used) is put into tlie bent tube, and the tube 

 is tilted until the carbide falls into the bulb. A quantity of soil is weiglied into 

 the small tube which is slipped inside the larger, the upper portion of the latter 

 being icept horizontal. The cork is tlien replaced tiglitly, the tube is connected 

 with tlie nitrometer, and the pressure is equalized. The tube is tilted until 

 the inner tube slips down to the bend, and the soil empties itself on to the 

 carbide ; it is then sliaken to mix the carbide and soil, the former rapidly 



C. Powdered Calcium Caibide. 



S. Small sample-tube containing a weighed quantity of soil. 



taking up moisture from tlie soil and evolving acetylene, which is collected in 

 the nitrometer. The gas is given off rapidly, and after a few minutes the 

 reaction ceases. If much moisture is present, some heat is generated, and it 

 is necessary to allow the gas to regain the normal temperature of the room 

 before measuring. 



It has been found by several observers that the volume of acetylene given 

 off from commercial carbide when allowed to react with water is slightly 

 less than that required theoretically. For every 18 milligrammes of water 



