PHYSICAL COMPOSITION OF SOILS. 9I 



in it the upward current ascends in a conical glass tube, (see 

 figure 6) entering through a narrow, curved inlet tube, in which 

 the soil sample is kept agitated by the current itself. The 

 objection to this plan is twofold: first, the narrow^, curved 

 inlet-tube is readily clogged by the soil mass at the lower 

 velocities, which are thereby changed, so that, unless a very 

 small amount of soil only is employed, the whole mass is not 

 kept properly stirred ; second, the circulating currents brought 

 about by the conical shape of the tube cause the sediment-par- 



FIG. 7. The Churn Elutriator (Hilgard's) for the physical analysis of soils. 



tides to coalesce into complex, larger ones (floccules), \vhich 

 will then settle down and fail to pass over at the current- 

 velocity corresponding to their individual component parts. 



Churn Elutriator with Cylindrical Tube. The errors just 

 alluded to are obviated by an arrangement devised by the 

 writer, in which a rapidly-revolving stirrer, placed at the base 

 of a cylindrical tube in which the washing process is conducted 



