320 cox. 



From Table XII it will be seen that if the pressure in the apparatus is ad- 

 justed by means of the stopcock so that the height of the water in the piezo- 

 meter is 1.83 centimeters and the flow is maintained sufficiently long all silt and 

 clay particles of less diameter than 0.01 millimeter will be floated out of the 

 apparatus at the orifice y. About 4 liters of water are required to complete 

 this operation. This water is allowed to stand 24 hours in 20-centimeter columns 

 and that portion which does not settle is siphoned oflF and called clay, whereas 

 the residue is termed fine silt. The largest particles of the former are about 

 0.002 millimeter in diameter. The clay may be directly determined by evaporat- 

 ing to dryness and weighing, or by evaporating only an aliquot part, but in 

 most cases it is sufficiently aeciu-ately determined by diff'erenee. 



If the pressure is now changed so that the height of the water in the piezo- 

 meter is 21.24 centimeters all silt and sand particles of less diameter than 0.1 

 millimeter will be removed from the tube A. \N\iQn this is complete the stopcock 

 is closed, the piezometer replaced by a solid rubber stopper and tube A removed 

 from the system and the stopcock x directly connected with tube B. When 

 the piezometer is again in place the stopcock is next opened so that the pressure 

 in the piezometer is 39.6 centimeters and the flow continued until all particles 

 less than 0.0.5 millimeter in diameter pass out at the orifice y when they are 

 combined with those which passed out at 21.24 centimeters pressure. We have 

 then separated our soil into the following sizes : 



Clay (diameter less than 0.002 millimeter) which was carried over 



at a pressure of 1.83 centimeters but remained in suspension. 



Fine silt (diameter 0.01 to 0.002 millimeter) which was carried over 



at a pressure of 1.83 centimeters but settled out in 24 hours. 



Silt (diameter 0.0.5 to 0.01 millimeter) which was carried over at a 



pressure of 1.83 to 39.6 centimeters. 

 Fine sand (diameter 0.1 to O.Oo millimeter) which was carried out 

 of tube A and remained in tube B at a pressure of 39.6 milli- 

 meters. 

 Sand (diameter greater than 0.1 millimeter) which remained in 

 tube A at a pressure of 21.24 millimeters." 

 Each of these fractions is united in or transferred to a single beaker and after 

 the clear water from each has been siphoned off they are transferred to porce- 

 lain crucibles, evaporated, dried in an air bath at 110°,*° cooled in a desiccator, 

 and weighed promjDtly. The sand is usually further divided by passing through 

 sieves which retain the grains of 0.5 and 0.25 millimeter in diameter respectively. 

 These fractions are weighed and the sand of 0.25 to 0.1 millimeter diameter 

 determined by diff'erenee. This is sufficiently accurate because grains of these 

 sizes fioeculate only slightly if any on drying. 



The soil must foe thoroughly disintegrated. — Complete disintegration of the 

 aggregates can be the only basis of comparison, for the aggregates in soil are not 

 constant. In general there are many substances, foremost among which are 

 acids, which promote flocculation, while alcohol, ether, alkalies, etc., inhibit it." 



" This may be further divided by passing it under water through sieves of 

 desired sizes. 



'" Three to four hours is enough to insure complete drying. 



" There are several theories to account for the formation and breaking down 

 of these flocules. Ostwald, Wo. Grundriss der Kolloidchemie, Dresden. (1910), 

 380 et seq. Hardy, W. B. Proc. Roy. Soc. (1899), 66, 95, 112. Puchner, Hein- 

 rich. Landiv. Vers. 8ta. 70, 249. Fickendey, E. Konigsberg i. Pr. J. Landw. 

 54, 343. V. 8. Geol. Surv. Bull. (1909), 388, 15. 



