JEFFERY ON MECHANICAL ANALYSIS OF SOILS. 



91 



in llu' soil, and with the pore space determined, K (in the formula) is 

 found by reference to a tahle. 



On the other end of tlie tube (a) is now screwed a gauze cap, the 

 tube is inverted, the solid cap removed, and the tube is screwed into 

 phice. The wei.uht d is next susp(Mi(h'd and this actinj; u[)on the bell 

 causes the aspiialion of Hie air throujih the soil. The minoraeter 

 indicates the jtressure under which the air is di-awn through the soil. 

 Usually the initial and linal pressui-es are i-ecorded, and the average 

 of the two taken as p in the formula. The pressure should not much 

 exceed tln*ee centimeters of water. 



The initial and final time are noted and the time rcipiired to aspirate 

 one or more liters of air through the soil is det(n-mined by subtracting 

 the initial fr(un the linal time, and is ('X[>i-essed in seconds. If only one 

 liter is aspirated then the time must be multiplied by five. Substitut- 

 ing these values in the formula, (J is readily determined. 



The (juestion naturally arises how nearly do the determined effective 

 size of soil grain compare with the actual size. Here is a classifica- 

 tion of a series of d.eterminations of sizes of sands made both by aspira- 

 tion and by counting and weighing. This classification is taken from 

 the l!Mh annual report of the United States Geological Survey, p. 224, 

 with a column added to show the difference. 



The sands were building sands, graded by means of soil sieves having 

 20, 40, 00, 80, 100 meshes to the inch, respectively. 



It is seen that with graded sands fairly close results are gotten bv 

 the two methods. Witli the ground glass the marvel is that so nearly 

 close results were obtained, since the particles were anything but 

 spherical in form. 



With the finer soils where counting and weighing are impossible we 

 ha])pen to have a check which seems to indicate that the effective size 

 of grain obtained by aspiration is far more nearly the true size of 

 grain than anv size determined bv anv of the old methods. 



Having given the size of spheres it is an easy matter to determine 

 their surface. 



In an experiment conducted by myself in thesis work, the total sur- 

 faces of weighed (juantities of samples of clay soil were determined, 

 first using the diameters of grains as determined by the old method, 



