THK MECHANICAIy ANALYSIS OF AKll) SOILS 



41 



of Hall and of Atterberg, but it is uvident that to obtain a separation up to about '002 nun. 



it is not necessary to carry the time of subsidence as far as prescribed by Hall. Many of influence of 



our results were obtained by the use of a column of 20 centimetres heieht, but we have '"^'S^t of 



•' ^ ' column 



abandoned this method for the reason that the shortening of the number of decantations 

 was not what was expected. It would appear that the longer the column the more the 

 particles which subside carry others with them. A better result is had by employing two 

 vessels with columns one half the height (10 cm.) the liquid being divided equally beween 

 them ; but such a multiplication of vessels does not' appear necessary if the soil be properly 

 puddled as described. 



An examination of the particles deposited from the 20 centimetre colunms proved them 

 to be but very slightly larger than those of the 10 or 8^ centimetre columns, provided 

 always that the time of subsidence be carried to 24 hours. The following figures show how 

 small is the effect of the height of the liquid on the results obtained from the mechanical 

 analysis of Sudan soils. The duration of subsidence was, as stated, 2'4 hours. 



Determination of the "Silt" 



When a number of soils have to be examined at the same time, we have found the use 

 of the apparatus shown in Fig. 5 to effect a great saving of time and trouble. The 



residues from the clay 

 determinations (made by 

 difference) are washed into 

 the cylindrical beakers (or 

 museum jars) with distilled 

 water containing the usual 

 small amount of sodium 

 carbonate.^ After standing 

 for a length of time de- 

 pendent upon the size of 

 the particles adopted as 

 representing the "silt" 

 division, the liquid is 

 poured away and rejected. 

 This preliminary treat- 

 ment should be made with distilled water, since the soil still contains particles in size 

 bordering on those of the clay division. The subsequent operations may, as a rule, be 

 carried out with clear tap water. 



' Since the heating of the soil residue in the oven usually causes It to cake together, it is advisable to add to 

 it a small amount of water containing sodium carbonate and bring almost or quite to boiling. It is then treated 

 as described above. 



Fig. 5. — Apparatus tor the determiaation of ttUt 



•Silf 



